A B C D E F G H I J K L M N O P Q R S T U V W X _

A

abs() - Method in class neureka.math.Functions

abs() - Method in interface neureka.Tensor

This method is a functionally identical to the following alternatives:

ABSOLUTE - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

Absolute - Class in neureka.backend.main.operations.functions

Absolute() - Constructor for class neureka.backend.main.operations.functions.Absolute

AbstractBaseDevice<V> - Class in neureka.devices

AbstractBaseDevice() - Constructor for class neureka.devices.AbstractBaseDevice

AbstractComponentOwner<C> - Class in neureka.common.composition

Together with the Component interface, this class defines a simple component system in which implementations of the Component interface are managed by extensions of this AbstractComponentOwner.

AbstractComponentOwner() - Constructor for class neureka.common.composition.AbstractComponentOwner

AbstractCPUConvolution - Class in neureka.backend.main.implementations.convolution

AbstractCPUConvolution() - Constructor for class neureka.backend.main.implementations.convolution.AbstractCPUConvolution

AbstractDevice<V> - Class in neureka.devices

This is the abstract precursor class providing some useful implementations for core concepts which are most likely applicable to most concrete implementations of the Device interface.

AbstractDevice() - Constructor for class neureka.devices.AbstractDevice

AbstractDeviceAlgorithm<C extends DeviceAlgorithm<C>> - Class in neureka.backend.api.template.algorithms

This is a partial implementation of the Algorithm interface which implements the component system for implementation instances of the ImplementationFor interface.

AbstractDeviceAlgorithm(String) - Constructor for class neureka.backend.api.template.algorithms.AbstractDeviceAlgorithm

AbstractDeviceData<T> - Class in neureka.devices

AbstractDeviceData(AbstractBaseDevice<?>, Object, DataType<T>, Runnable) - Constructor for class neureka.devices.AbstractDeviceData

AbstractFunAlgorithm - Class in neureka.backend.api.template.algorithms

AbstractFunAlgorithm(String) - Constructor for class neureka.backend.api.template.algorithms.AbstractFunAlgorithm

AbstractFunDeviceAlgorithm<C extends DeviceAlgorithm<C>> - Class in neureka.backend.api.template.algorithms

This is the base class for implementations of the Algorithm interface.

AbstractFunDeviceAlgorithm(String) - Constructor for class neureka.backend.api.template.algorithms.AbstractFunDeviceAlgorithm

AbstractImplementationFor<D extends Device<?>> - Class in neureka.backend.api.template.implementations

AbstractImplementationFor(ImplementationFor<D>, int) - Constructor for class neureka.backend.api.template.implementations.AbstractImplementationFor

AbstractNDC - Class in neureka.ndim.config

The following is an abstract implementation of the NDConfiguration which offers a basis for instantiation and caching of concrete implementations extending this abstract class.

AbstractNDC() - Constructor for class neureka.ndim.config.AbstractNDC

AbstractOperation - Class in neureka.backend.api.template.operations

This abstract Operation implementation is a useful template for creating new operations.

AbstractOperation(OperationBuilder) - Constructor for class neureka.backend.api.template.operations.AbstractOperation

accept(Class<? extends Operation>, Class<? extends DeviceAlgorithm>, Class<? extends D>, Function<LoadingContext, ImplementationFor<D>>) - Method in interface neureka.backend.api.ini.ImplementationReceiver

access(Tensor<T>) - Method in class neureka.devices.AbstractDevice

This method exposes the tensor access API for reading from or writing to a tensor stored on this device.

access(Tensor<T>) - Method in interface neureka.devices.Device

This method exposes the tensor access API for reading from or writing to a tensor stored on this device.

access(Tensor<T>) - Method in class neureka.devices.file.FileDevice

act(ADTarget<?>) - Method in interface neureka.autograd.ADAction

The auto-differentiation forward or backward pass of an ADAction propagate partial differentiations forward along the computation graph.

activationCode() - Method in interface neureka.backend.main.implementations.fun.api.ScalarFun

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarAbsolute

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarCbrt

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarCosinus

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarExp

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarGaSU

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarGaTU

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarGaussian

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarGaussianFast

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarGeLU

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarIdentity

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarLog10

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarLogarithm

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarQuadratic

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarReLU

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarSeLU

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarSigmoid

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarSiLU

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarSinus

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarSoftplus

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarSoftsign

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarSqrt

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarTanh

activationCode() - Method in class neureka.backend.main.implementations.fun.ScalarTanhFast

actualize() - Method in interface neureka.devices.Device.Access

ADAction - Interface in neureka.autograd

This interface is the declaration for lambda actions for both the ADAction.act(ADTarget) method of the ADAction interface.

ADAction(Function, ExecutionCall<? extends Device<?>>) - Static method in class neureka.backend.api.template.algorithms.FallbackAlgorithm

ADActionSupplier - Interface in neureka.backend.api.fun

Implementations of this functional interface ought to return a new instance of the ADAction class responsible for performing automatic differentiation both for forward and backward mode differentiation.

AdaGrad<V extends java.lang.Number> - Class in neureka.optimization.implementations

Adaptive Gradients, or AdaGrad for short, is an extension of the gradient descent optimization algorithm that adjusts the step size for each parameter based on the squared gradients seen over the course of previous optimization steps.

AdaGrad - Static variable in interface neureka.optimization.Optimizer

AdaGradFactory - Class in neureka.optimization.implementations

AdaGradFactory() - Constructor for class neureka.optimization.implementations.AdaGradFactory

ADAM<V extends java.lang.Number> - Class in neureka.optimization.implementations

ADAM (short for Adaptive Moment Estimation) is an adaptive learning rate optimization algorithm that utilises both momentum and scaling, combining the benefits of RMSProp and SGD with respect to Momentum.

ADAM - Static variable in interface neureka.optimization.Optimizer

ADAMFactory - Class in neureka.optimization.implementations

ADAMFactory() - Constructor for class neureka.optimization.implementations.ADAMFactory

add() - Method in class neureka.math.Functions

addAssign() - Method in class neureka.math.Functions

addChild(Tensor<V>) - Method in class neureka.framing.Relation

Addition - Class in neureka.backend.main.operations.operator

Addition() - Constructor for class neureka.backend.main.operations.operator.Addition

addOperation(Operation) - Method in class neureka.backend.api.BackendContext

This method registers Operation implementation instances in this BackendContext which is the thread local execution context receiving and processing Tensor instances...

addPending(Set<GraphNode<V>>) - Method in class neureka.autograd.JITProp

addPermuteRelationFor(Tensor<V>, int[]) - Method in class neureka.framing.Relation

When creating permuted versions of slices then there must be a translation between the shape configuration between this new slice and the original parent tensor from which both slices have been derived.

addToGradient(Tensor<T>) - Method in interface neureka.MutateTensor

This method takes the provided Tensor instance and adds its contents to the contents of the Tensor which is set as gradient of this very Tensor.

ADSupportPredicate - Interface in neureka.backend.api.fun

A ADSupportPredicate lambda checks which auto differentiation mode can be performed for a given ExecutionCall.

ADTarget<V> - Class in neureka.autograd

This is simply a wrapper for useful information needed by implementations of the ADAction and ADAction interfaces to perform error propagation.

Algorithm - Interface in neureka.backend.api

This class is the middle layer of the 3 tier compositional architecture of this backend, which consists of Operations, Algorithms and in case of a DeviceAlgorithm also ImplementationFor.

all(Call.TensorCondition) - Method in class neureka.backend.api.Call.Validator

all(Call.TensorCompare) - Method in class neureka.backend.api.Call.Validator

all(V) - Method in class neureka.fluent.building.NdaBuilder

all(V) - Method in interface neureka.fluent.building.states.IterByOrIterFromOrAll

This method creates and return a Tensor instance which will be homogeneously filled by the the provided value irrespective of the previously defined shape.

all(V) - Method in interface neureka.fluent.building.states.IterByOrIterFromOrAllTensor

This method creates and return a Tensor instance which will be homogeneously filled by the the provided value irrespective of the previously defined shape.

all() - Method in class neureka.fluent.slicing.AxisSliceBuilder

all() - Method in interface neureka.fluent.slicing.states.FromOrAt

This is a convenience method replacing "from(0).to(axisSize-1)", meaning that it simply slices the whole current axis from the original tensor.

all() - Method in interface neureka.fluent.slicing.states.FromOrAtTensor

This is a convenience method replacing "from(0).to(axisSize-1)", meaning that it simply slices the whole current axis from the original tensor.

allAliasGetter(Supplier<List<Object>>) - Method in class neureka.framing.fluent.AxisFrame.Builder

allAliasGetterFor(Function<Integer, List<Object>>) - Method in class neureka.framing.fluent.AxisFrame.Builder

allMetaArgs() - Method in class neureka.backend.api.Call

allNotNull(Call.TensorCondition) - Method in class neureka.backend.api.Call.Validator

allNotNullHaveSame(Call.TensorProperty) - Method in class neureka.backend.api.Call.Validator

allocate(DataType<T>, NDConfiguration) - Method in interface neureka.devices.Device

allocate(DataType<T>, int) - Method in interface neureka.devices.Device

allocate(DataType<V>, NDConfiguration) - Method in class neureka.devices.file.FileDevice

allocate(Class<T>, Object) - Method in class neureka.devices.host.CPU

allocate(Class<T>, int, Object) - Method in class neureka.devices.host.CPU

allocate(DataType<T>, NDConfiguration) - Method in class neureka.devices.host.CPU

allocate(DataType<T>, NDConfiguration) - Method in class neureka.devices.opencl.OpenCLDevice

allocateFromAll(DataType<T>, NDConfiguration, Object) - Method in interface neureka.devices.Device

allocateFromAll(DataType<T>, NDConfiguration, Object) - Method in class neureka.devices.file.FileDevice

allocateFromAll(DataType<T>, NDConfiguration, Object) - Method in class neureka.devices.host.CPU

allocateFromAll(DataType<T>, NDConfiguration, Object) - Method in class neureka.devices.opencl.OpenCLDevice

allocateFromOne(DataType<T>, NDConfiguration, T) - Method in interface neureka.devices.Device

allocateFromOne(DataType<V>, NDConfiguration, V) - Method in class neureka.devices.file.FileDevice

allocateFromOne(DataType<T>, NDConfiguration, T) - Method in class neureka.devices.host.CPU

allocateFromOne(DataType<T>, NDConfiguration, T) - Method in class neureka.devices.opencl.OpenCLDevice

allowsBackward() - Method in enum neureka.backend.api.AutoDiffMode

allowsForward() - Method in enum neureka.backend.api.AutoDiffMode

allShare(Function<Tensor<?>, T>) - Method in class neureka.backend.api.Call.Validator

and(F) - Method in interface neureka.backend.main.algorithms.internal.AndBackward

andArgs(List<Arg>) - Method in class neureka.backend.api.Call.Builder

andArgs(Arg<?>...) - Method in class neureka.backend.api.Call.Builder

andArgs(List<Arg>) - Method in class neureka.backend.api.ExecutionCall.Builder

andArgs(Arg<?>...) - Method in class neureka.backend.api.ExecutionCall.Builder

AndBackward<F> - Interface in neureka.backend.main.algorithms.internal

andFill(V...) - Method in class neureka.fluent.building.NdaBuilder

andFill(V...) - Method in interface neureka.fluent.building.states.IterByOrIterFromOrAll

Provide an array of values which will be used to fill the Tensor instance returned by this last fluent builder method.

andFill(List<V>) - Method in interface neureka.fluent.building.states.IterByOrIterFromOrAll

Provide a list of values which will be used to fill the Tensor instance returned by this last fluent builder method.

andFill(V...) - Method in interface neureka.fluent.building.states.IterByOrIterFromOrAllTensor

Provide an array of values which will be used to fill the Tensor instance returned by this last fluent builder method.

andFill(List<V>) - Method in interface neureka.fluent.building.states.IterByOrIterFromOrAllTensor

Provide a list of values which will be used to fill the Tensor instance returned by this last fluent builder method.

andFillFrom(V) - Method in class neureka.fluent.building.NdaBuilder

andFillFrom(V) - Method in interface neureka.fluent.building.states.IterByOrIterFromOrAll

This part of the builder API allows for specifying a range which starts from the provided value and will end at the value specified in the next builder step returned by this method.

andFillFrom(V) - Method in interface neureka.fluent.building.states.IterByOrIterFromOrAllTensor

This part of the builder API allows for specifying a range which starts from the provided value and will end at the value specified in the next builder step returned by this method.

andImplementation(ImplementationFor<CPU>) - Method in interface neureka.backend.main.implementations.CPUImplementation.AndImplementation

andSeed(Object) - Method in class neureka.fluent.building.NdaBuilder

andSeed(Object) - Method in interface neureka.fluent.building.states.IterByOrIterFromOrAll

This method creates and return a Tensor instance which will be filled based on the provided seed object.

andSeed(Object) - Method in interface neureka.fluent.building.states.IterByOrIterFromOrAllTensor

This method creates and return a Tensor instance which will be filled based on the provided seed object.

andWhere(Filler<V>) - Method in class neureka.fluent.building.NdaBuilder

This method receives an Filler lambda which will be used to populate the Tensor instance produced by this API with values.

andWhere(Filler<V>) - Method in interface neureka.fluent.building.states.IterByOrIterFromOrAll

Pass a lambda to this method which will be used to populate the Tensor built by this fluent builder API based on the indices of the tensor.

andWhere(Filler<V>) - Method in interface neureka.fluent.building.states.IterByOrIterFromOrAllTensor

Pass a lambda to this method which will be used to populate the Tensor built by this fluent builder API based on the indices of the tensor.

any(Call.TensorCondition) - Method in class neureka.backend.api.Call.Validator

any(String...) - Static method in interface neureka.devices.Device

This method returns Device instances matching the given search parameter.

any(Predicate<V>) - Method in interface neureka.Nda

Iterates over every element of this nd-array, and checks whether any element matches the provided lambda.

any(Predicate<Integer>) - Method in interface neureka.Shape

anyNotNull(Call.TensorCondition) - Method in class neureka.backend.api.Call.Validator

applyGradient() - Method in interface neureka.Tensor

If this tensor owns a gradient tensor as component, then it can be applied by this method.

approve(ExecutionCall<? extends Device<?>>) - Method in class neureka.devices.AbstractDevice

This method plays an important role in approving a provided ExecutionCall. When implementing custom operations or such for the backend of this library, then one may use this in order to check if the provided call is suitable for this Device.

approve(ExecutionCall<? extends Device<?>>) - Method in interface neureka.devices.Device

This method is used internally to give Device implementations the opportunity to perform some exception handling before the ExecutionCall will be dispatched.

approve(ExecutionCall<? extends Device<?>>) - Method in class neureka.devices.file.FileDevice

architecture - Variable in class neureka.devices.host.machine.CommonMachine

Arg<T> - Class in neureka.math.args

Extend this class to define additional meta arguments for Functions.

Arg(T) - Constructor for class neureka.math.args.Arg

Arg.Axis - Class in neureka.math.args

Arg.Derivative<V> - Class in neureka.math.args

Arg.DerivIdx - Class in neureka.math.args

This is an import argument whose role might not be clear at first : An operation can have multiple inputs, however when calculating the derivative for a forward or backward pass then one must know which derivative ought to be calculated.

Arg.Ends - Class in neureka.math.args

Arg.Indices - Class in neureka.math.args

Arg.Layout - Class in neureka.math.args

Arg.MinRank - Class in neureka.math.args

Arg.Offset - Class in neureka.math.args

Arg.Seed - Class in neureka.math.args

Arg.Shape - Class in neureka.math.args

Arg.Stride - Class in neureka.math.args

Arg.TargetDevice - Class in neureka.math.args

Arg.VarIdx - Class in neureka.math.args

The following argument is relevant for a particular type of operation, namely: an "indexer".

Args - Class in neureka.math.args

Args() - Constructor for class neureka.math.args.Args

arity() - Method in class neureka.backend.api.Call

arity(int) - Method in class neureka.backend.api.template.operations.OperationBuilder

as(Class<D>) - Method in interface neureka.Data

This returns the underlying raw data object of a nd-array or tensor.

asDerivative(Function[], int) - Method in interface neureka.backend.api.Operation

Operation implementations and Function implementations are in a tight relationship where the Function describes an abstract syntax tree based on the syntactic information provided by the Operation (through methods like Operation.getOperator() or Operation.getIdentifier()).

asDerivative(Function[], int) - Method in class neureka.backend.api.template.operations.AbstractOperation

Operation implementations and Function implementations are in a tight relationship where the Function describes an abstract syntax tree based on the syntactic information provided by the Operation (through methods like Operation.getOperator() or Operation.getIdentifier()).

asDerivative(Function[], int) - Method in interface neureka.backend.api.template.operations.OperationBuilder.Derivation

asDerivative(Function[], int) - Method in class neureka.backend.main.operations.functions.Logarithm

asDerivative(Function[], int) - Method in class neureka.backend.main.operations.operator.Addition

asDerivative(Function[], int) - Method in class neureka.backend.main.operations.operator.Division

asDerivative(Function[], int) - Method in class neureka.backend.main.operations.operator.Modulo

asDerivative(Function[], int) - Method in class neureka.backend.main.operations.operator.Multiplication

asDerivative(Function[], int) - Method in class neureka.backend.main.operations.operator.Power

asDerivative(Function[], int) - Method in class neureka.backend.main.operations.operator.Subtraction

asImage(Tensor.ImageType) - Method in interface neureka.Tensor

Turns this tensor into a BufferedImage based on the provided Tensor.ImageType formatting choice.

asInlineArray() - Method in interface neureka.ndim.config.NDConfiguration

This method returns an array of flattened arrays which define this nd-configuration in a compact manner.

assign(T) - Method in interface neureka.MutateNda

Use this to assign the provided item to all elements of this nd-array! This method is an inline operation which changes the underlying data of the nd-array.

assign(Nda<T>) - Method in interface neureka.MutateNda

Use this to assign the provided nd-array to this nd-array! This method is an inline operation which changes the underlying data of the nd-array.

assign(T) - Method in interface neureka.MutateTensor

assign(Nda<T>) - Method in interface neureka.MutateTensor

AssignLeft - Class in neureka.backend.main.operations.other

AssignLeft() - Constructor for class neureka.backend.main.operations.other.AssignLeft

assumptionBasedOn(String) - Static method in class neureka.math.parsing.ParseUtil

This method tries to find the next best operation String the user might have meant.

asType(Class<T>) - Method in interface neureka.Tensor

at(int) - Method in interface neureka.devices.Device.Writer

Writes whatever kind of data was previously specified, to the tensors' data at the position targeted by the provided index.

at(String) - Static method in class neureka.devices.file.FileDevice

at(int) - Method in class neureka.fluent.slicing.AxisSliceBuilder

This method returns an instance of this very AxisSliceBuilder instance disguised by the AxisOrGet interface.

at(int) - Method in interface neureka.fluent.slicing.states.FromOrAt

This is a convenience method replacing "from(i).to(i)", meaning that it simply slices a single axis from the original tensor at the specified index.

at(int) - Method in interface neureka.fluent.slicing.states.FromOrAtTensor

This is a convenience method replacing "from(i).to(i)", meaning that it simply slices a single axis from the original tensor at the specified index.

At<K,R> - Interface in neureka.framing.fluent

at(K) - Method in interface neureka.framing.fluent.At

at(int...) - Method in interface neureka.MutateNda

Exposes the MutateNda.Item interface which allows you to get or set individual nd-array items.

at(int...) - Method in interface neureka.Nda

This method exposes the Nda.Item API which allows you to get or set individual items within this nd-array targeted by an array of provided indices.

atAxis(Object) - Method in class neureka.framing.NDFrame

A NDFrame exposes aliases for axes as well as aliases for individual positions within an axis.

atIndexAlias(Object) - Method in class neureka.framing.fluent.AxisFrame

autoDelete(Tensor<?>...) - Static method in class neureka.backend.main.memory.MemUtil

This method will try to delete the provided array of tensors if the tensors are not important computation graph components (like derivatives for example).

AutoDiffMode - Enum in neureka.backend.api

autoDiffModeFrom(ExecutionCall<? extends Device<?>>) - Method in interface neureka.backend.api.fun.ADSupportPredicate

Implementations of this ought to check which auto differentiation mode can be performed for a given ExecutionCall.

autoDiffModeFrom(ExecutionCall<? extends Device<?>>) - Method in class neureka.backend.api.template.algorithms.AbstractFunAlgorithm

autoDiffModeFrom(ExecutionCall<? extends Device<?>>) - Method in class neureka.backend.api.template.algorithms.AbstractFunDeviceAlgorithm

autoDiffModeFrom(ExecutionCall<? extends Device<?>>) - Method in class neureka.backend.api.template.algorithms.FallbackAlgorithm

autograd() - Method in class neureka.Neureka.Settings

autograd(Object) - Method in class neureka.Neureka.Settings

This allows you to configure Neureka using a Groovy DSL.

AutoGrad() - Constructor for class neureka.Neureka.Settings.AutoGrad

autogradMode() - Method in class neureka.backend.api.ExecutionCall

This method queries the underlying Operation for a suitable Algorithm for this ExecutionCall to see what kind of auto differentiation can be performed.

axis(int) - Method in class neureka.fluent.slicing.AxisSliceBuilder

This method returns an instance of the AxisSliceBuilder targeted by the provided index.

axis(int) - Method in class neureka.fluent.slicing.SliceBuilder

This method returns an instance of the AxisSliceBuilder disguised by the FromOrAt interface.

axis(int) - Method in interface neureka.fluent.slicing.states.AxisOrGet

Slicing a tensor ultimately means slicing one or more of its axes! This method allows one to specify which axis should be sliced next.

axis(int) - Method in interface neureka.fluent.slicing.states.AxisOrGetTensor

Slicing a tensor ultimately means slicing one or more of its axes! This method allows one to specify which axis should be sliced next.

AxisFrame<G,V> - Class in neureka.framing.fluent

This class represents the labeled axis of an NDFrame.

AxisFrame.Builder<SetType,GetType,ValueType> - Class in neureka.framing.fluent

AxisFrame.Set<V> - Interface in neureka.framing.fluent

AxisOrGet<V> - Interface in neureka.fluent.slicing.states

This is the starting point of the call transition graph exposed by the slice builder API.

AxisOrGetTensor<V> - Interface in neureka.fluent.slicing.states

AxisSliceBuilder<V> - Class in neureka.fluent.slicing

AXPY - Class in neureka.backend.main.operations.linear.internal.blas

The ?axpy routines perform a vector-vector operation defined as y := a*x + y where: a is a scalar x and y are vectors each with a number of elements that equals n.

AXPY() - Constructor for class neureka.backend.main.operations.linear.internal.blas.AXPY

B

backend() - Method in class neureka.Neureka

BackendContext - Class in neureka.backend.api

Instances of this class are execution contexts hosting Operation instances which receive Tensor instances for execution.

BackendContext() - Constructor for class neureka.backend.api.BackendContext

This creates a new context which is completely void of any Operation implementation instances.

BackendContext.Runner - Class in neureka.backend.api

This is a very simple class with a single purpose, namely it exposes methods which receive lambda instances in order to then execute them in a given BackendContext, just to then switch back to the original context again.

BackendExtension - Interface in neureka.backend.api

Implementations of this might introduce CUDA or ROCM to Neureka.

BackendExtension.DeviceOption - Class in neureka.backend.api

This class describes an available Device implementation found for a given BackendExtension.

BackendLoader - Interface in neureka.backend.api.ini

BackendRegistry - Class in neureka.backend.api.ini

backward(Tensor<V>) - Method in class neureka.autograd.GraphNode

This method is the entry-point for the back-propagation process.

backward(Tensor<V>) - Method in interface neureka.Tensor

Tensors which are used or produced by the autograd system will have a GraphNode component attached to them.

backward(double) - Method in interface neureka.Tensor

Tensors which are used or produced by the autograd system will have a GraphNode component attached to them.

backward() - Method in interface neureka.Tensor

Use this to back-propagate an error signal of 1.0 through the recorded computation graph.

backwardJIT(Tensor<V>) - Method in class neureka.autograd.GraphNode

This method is called only when JITProp is active.

BAD - Static variable in interface neureka.backend.api.fun.SuitabilityPredicate

badIfAll(Call.TensorCondition) - Method in class neureka.backend.api.Call.Validator.Estimator

badIfAll(Call.TensorCompare) - Method in class neureka.backend.api.Call.Validator.Estimator

badIfAny(Call.TensorCondition) - Method in class neureka.backend.api.Call.Validator.Estimator

badIfAnyNonNull(Call.TensorCondition) - Method in class neureka.backend.api.Call.Validator.Estimator

BasicMachine - Class in neureka.devices.host.machine

How much memory, and how many threads share that memory.

BasicMachine(long, int) - Constructor for class neureka.devices.host.machine.BasicMachine

basicSuitability() - Method in class neureka.backend.api.Call.Validator

The validity as float being >0/true and 0/false.

belongsToGraph() - Method in interface neureka.Tensor

Tensors which are used or produced by the autograd system will have a GraphNode component attached to them.

BiElementwise - Class in neureka.backend.main.algorithms

BiElementwise() - Constructor for class neureka.backend.main.algorithms.BiElementwise

BiScalarBroadcast - Class in neureka.backend.main.algorithms

BiScalarBroadcast() - Constructor for class neureka.backend.main.algorithms.BiScalarBroadcast

boolToDouble(boolean[]) - Static method in class neureka.common.utility.DataConverter.Utility

boolToFloat(boolean[]) - Static method in class neureka.common.utility.DataConverter.Utility

bootstrapTip() - Method in enum neureka.devices.opencl.utility.Messages.Tips

borrow(Tensor<V>, Tensor<V>...) - Method in interface neureka.devices.Device

This is a very simple fluent API for temporarily storing a number of tensors on this Device, executing a provided lambda action, and then migrating all the tensors back to their original devices.

Broadcast - Class in neureka.backend.main.algorithms

Broadcast() - Constructor for class neureka.backend.main.algorithms.Broadcast

bufferType - Variable in enum neureka.Tensor.ImageType

build() - Method in class neureka.backend.api.template.operations.OperationBuilder

build() - Method in class neureka.framing.fluent.AxisFrame.Builder

builder() - Static method in interface neureka.backend.api.Operation

builder() - Static method in class neureka.framing.fluent.AxisFrame

buildFunAlgorithm() - Method in class neureka.backend.api.template.algorithms.AbstractFunAlgorithm

buildFunAlgorithm() - Method in class neureka.backend.api.template.algorithms.AbstractFunDeviceAlgorithm

byDefaults() - Method in interface neureka.view.NdaAsString.Builder

byteToBigInteger(byte[]) - Static method in class neureka.common.utility.DataConverter.Utility

byteToDouble(byte[]) - Static method in class neureka.common.utility.DataConverter.Utility

byteToFloat(byte[]) - Static method in class neureka.common.utility.DataConverter.Utility

byteToInt(byte[]) - Static method in class neureka.common.utility.DataConverter.Utility

byteToLong(byte[]) - Static method in class neureka.common.utility.DataConverter.Utility

byteToShort(byte[]) - Static method in class neureka.common.utility.DataConverter.Utility

C

Cache<O> - Class in neureka.common.utility

This is a simple, fixed size cache for immutable objects which are shared throughout the library runtime...

Cache(int) - Constructor for class neureka.common.utility.Cache

cache - Variable in class neureka.devices.host.machine.CommonMachine

The size of one top level (L3 or L2) cache unit in bytes.

Cache.LazyEntry<K,V> - Class in neureka.common.utility

Lazy cache entries are entries whose values will be calculated only when the entry is being stored in the cache.

calculate(double[], int, int, Function[]) - Method in interface neureka.backend.api.Operation

This method mainly ought to serve as a reference- and fallback- implementation for tensor backends and also as the backend for handling the calculation of scalar inputs passed to a given abstract syntax tree of Function instances...

calculate(double, boolean) - Method in interface neureka.backend.main.implementations.fun.api.ScalarFun

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.indexer.Product

calculate(double[], int, Function[]) - Static method in class neureka.backend.main.operations.indexer.Product

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.indexer.Summation

calculate(double[], int, Function[]) - Static method in class neureka.backend.main.operations.indexer.Summation

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.linear.Convolution

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.linear.DotProduct

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.linear.MatMul

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.linear.XConvLeft

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.linear.XConvRight

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.operator.Addition

calculate(double[], int, Function[]) - Static method in class neureka.backend.main.operations.operator.Addition

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.operator.Division

calculate(double[], int, Function[]) - Static method in class neureka.backend.main.operations.operator.Division

calculate(double[], int, Function[]) - Static method in class neureka.backend.main.operations.operator.Modulo

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.operator.Modulo

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.operator.Multiplication

calculate(double[], int, Function[]) - Static method in class neureka.backend.main.operations.operator.Multiplication

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.operator.Power

calculate(double[], int, Function[]) - Static method in class neureka.backend.main.operations.operator.Power

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.operator.Subtraction

calculate(double[], int, Function[]) - Static method in class neureka.backend.main.operations.operator.Subtraction

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.other.AssignLeft

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.other.Cat

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.other.DimFit

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.other.DimTrim

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.other.Max

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.other.Min

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.other.Permute

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.other.Randomization

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.other.ReLayout

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.other.Reshape

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.other.Slice

calculate(double[], int, int, Function[]) - Method in class neureka.backend.main.operations.other.Sum

call(Supplier<T>) - Method in class neureka.backend.api.BackendContext.Runner

Use this method to supply a lambda which will be executed in the BackendContext which produced this very BackendContext.Runner instance.

Call<D> - Class in neureka.backend.api

Instances of this class model simple execution calls to the backend.

Call(Tensor<?>[], D, List<Arg>) - Constructor for class neureka.backend.api.Call

call(int) - Method in class neureka.devices.opencl.KernelCaller

call(long[], long[]) - Method in class neureka.devices.opencl.KernelCaller

Use this to call the kernel with 2 long arrays defining how the kernel should be indexed and parallelized.

call(double) - Method in interface neureka.math.Function

Invokes this Function with the provided scalar as a single input and returns the scalar result.

call(double[], int) - Method in interface neureka.math.Function

Invokes this Function with the provided array of inputs ad an index for input dependent indexing.

call(double...) - Method in interface neureka.math.Function

Invokes this Function with the provided array of inputs.

call(Call.Builder<T, D>) - Method in interface neureka.math.Function

Use this for more control over the execution, which is especially useful when interfacing with more complex types of operations, requiring more context information.

call(Call<D>) - Method in interface neureka.math.Function

Use this for more control over the execution, which is very helpful when interfacing with more complex types of operations, requiring more context information.

call(Args, Tensor<T>...) - Method in interface neureka.math.Function

Use this to call this Function alongside with some additional meta-arguments which will be passed to the underlying Operation(s).

call(Tensor<T>) - Method in interface neureka.math.Function

call(List<Tensor<T>>) - Method in interface neureka.math.Function

call(Tensor<T>[], int) - Method in interface neureka.math.Function

call(Tensor<T>...) - Method in interface neureka.math.Function

call(Tensor<T>...) - Method in interface neureka.math.Function.Callable

call(double[], int) - Method in class neureka.math.implementations.FunctionConstant

call(double[], int) - Method in class neureka.math.implementations.FunctionInput

call(double[], int) - Method in class neureka.math.implementations.FunctionNode

call(double[], int) - Method in class neureka.math.implementations.FunctionVariable

Call.Builder<V,T extends Device<V>> - Class in neureka.backend.api

Call.DeviceCondition - Interface in neureka.backend.api

Call.Else<T> - Interface in neureka.backend.api

Call.OperationCondition - Interface in neureka.backend.api

Call.TensorCompare - Interface in neureka.backend.api

Call.TensorCondition - Interface in neureka.backend.api

Call.TensorProperty - Interface in neureka.backend.api

Call.TensorsCondition - Interface in neureka.backend.api

Call.Validator - Class in neureka.backend.api

This is a simple nested class offering various lambda based methods for validating the tensor arguments stored inside this ExecutionCall.

Call.Validator.Estimator - Class in neureka.backend.api

canAccessOpenCL() - Method in class neureka.Neureka

canAccessOpenCLDevice() - Method in class neureka.Neureka

canBeDeleted() - Method in class neureka.autograd.GraphNode

Cat - Class in neureka.backend.main.operations.other

Cat() - Constructor for class neureka.backend.main.operations.other.Cat

CBRT - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

Cbrt - Class in neureka.backend.main.operations.functions

Cbrt() - Constructor for class neureka.backend.main.operations.functions.Cbrt

cbrt() - Method in class neureka.math.Functions

cbrt() - Method in interface neureka.Tensor

This method is a functionally identical to the following alternatives:

change() - Method in class neureka.devices.ReferenceCounter.ChangeEvent

ChangeEvent(ReferenceCounter.ChangeType, int, int) - Constructor for class neureka.devices.ReferenceCounter.ChangeEvent

check(Device<?>) - Method in interface neureka.backend.api.Call.DeviceCondition

check(Operation) - Method in interface neureka.backend.api.Call.OperationCondition

check(Tensor<?>, Tensor<?>) - Method in interface neureka.backend.api.Call.TensorCompare

check(Tensor<?>) - Method in interface neureka.backend.api.Call.TensorCondition

check(Tensor<?>[]) - Method in interface neureka.backend.api.Call.TensorsCondition

checkArity() - Method in class neureka.backend.api.ExecutionCall

childCount() - Method in class neureka.framing.Relation

CLBackend - Class in neureka.backend.ocl

This is an OpenCL context component for any given BackendContext which extends a given backend context instance for additional functionality, which in this case is the OpenCL backend storing platform and device information.

CLBackend() - Constructor for class neureka.backend.ocl.CLBackend

Use this constructor if you want to create a new OpenCL world in which there are unique OpenCLPlatform and OpenCLDevice instances.

CLBiElementwise - Class in neureka.backend.main.implementations.elementwise

CLBiElementwise(String, String, String) - Constructor for class neureka.backend.main.implementations.elementwise.CLBiElementwise

CLBiElementwiseAddition - Class in neureka.backend.main.implementations.elementwise

CLBiElementwiseAddition(String) - Constructor for class neureka.backend.main.implementations.elementwise.CLBiElementwiseAddition

CLBiElementwiseDivision - Class in neureka.backend.main.implementations.elementwise

CLBiElementwiseDivision(String) - Constructor for class neureka.backend.main.implementations.elementwise.CLBiElementwiseDivision

CLBiElementwiseModulo - Class in neureka.backend.main.implementations.elementwise

CLBiElementwiseModulo(String) - Constructor for class neureka.backend.main.implementations.elementwise.CLBiElementwiseModulo

CLBiElementwiseMultiplication - Class in neureka.backend.main.implementations.elementwise

CLBiElementwiseMultiplication(String) - Constructor for class neureka.backend.main.implementations.elementwise.CLBiElementwiseMultiplication

CLBiElementwisePower - Class in neureka.backend.main.implementations.elementwise

CLBiElementwisePower(String) - Constructor for class neureka.backend.main.implementations.elementwise.CLBiElementwisePower

CLBiElementwiseSubtraction - Class in neureka.backend.main.implementations.elementwise

CLBiElementwiseSubtraction(String) - Constructor for class neureka.backend.main.implementations.elementwise.CLBiElementwiseSubtraction

CLBroadcast - Class in neureka.backend.main.implementations.broadcast

CLBroadcast(String, String, String) - Constructor for class neureka.backend.main.implementations.broadcast.CLBroadcast

CLBroadcastAddition - Class in neureka.backend.main.implementations.broadcast

CLBroadcastAddition(String) - Constructor for class neureka.backend.main.implementations.broadcast.CLBroadcastAddition

CLBroadcastDivision - Class in neureka.backend.main.implementations.broadcast

CLBroadcastDivision(String) - Constructor for class neureka.backend.main.implementations.broadcast.CLBroadcastDivision

CLBroadcastModulo - Class in neureka.backend.main.implementations.broadcast

CLBroadcastModulo(String) - Constructor for class neureka.backend.main.implementations.broadcast.CLBroadcastModulo

CLBroadcastMultiplication - Class in neureka.backend.main.implementations.broadcast

CLBroadcastMultiplication(String) - Constructor for class neureka.backend.main.implementations.broadcast.CLBroadcastMultiplication

CLBroadcastPower - Class in neureka.backend.main.implementations.broadcast

CLBroadcastPower(String) - Constructor for class neureka.backend.main.implementations.broadcast.CLBroadcastPower

CLBroadcastSubtraction - Class in neureka.backend.main.implementations.broadcast

CLBroadcastSubtraction(String) - Constructor for class neureka.backend.main.implementations.broadcast.CLBroadcastSubtraction

clConfigOf(Tensor<?>) - Method in class neureka.devices.opencl.OpenCLDevice

clConfigOf(NDConfiguration) - Method in class neureka.devices.opencl.OpenCLDevice

clContextCouldNotFindAnyDevices() - Static method in class neureka.devices.opencl.utility.Messages

clContextCreationFailed() - Static method in class neureka.devices.opencl.utility.Messages

CLConvolution - Class in neureka.backend.main.implementations.convolution

CLConvolution(String) - Constructor for class neureka.backend.main.implementations.convolution.CLConvolution

CLDot - Class in neureka.backend.main.implementations.linear

Performs a dot product on two vectors using OpenCL.

CLDot() - Constructor for class neureka.backend.main.implementations.linear.CLDot

cleanedHeadAndTail(String) - Static method in class neureka.math.parsing.ParseUtil

cleanup(Runnable) - Method in interface neureka.devices.Device.Access

Use this to perform some custom memory cleanup for when the accessed Tensor gets garbage collected.

CLElementwiseFunction - Class in neureka.backend.main.implementations.elementwise

CLElementwiseFunction(ScalarFun) - Constructor for class neureka.backend.main.implementations.elementwise.CLElementwiseFunction

CLFunctionCompiler - Class in neureka.devices.opencl.utility

Turns a Function into OpenCL kernel code to make optimized just in time compilation possible.

CLFunctionCompiler(OpenCLDevice, Function, String) - Constructor for class neureka.devices.opencl.utility.CLFunctionCompiler

CLGEMM - Class in neureka.backend.main.operations.linear.internal.opencl

CLGEMM() - Constructor for class neureka.backend.main.operations.linear.internal.opencl.CLGEMM

CLImplementation - Class in neureka.backend.main.implementations

This class is the ExecutorFor < OpenCLDevice > implementation used to properly call an OpenCLDevice instance via the ExecutionOn < OpenCLDevice > lambda implementation receiving an instance of the ExecutionCall class.

CLImplementation(ImplementationFor<OpenCLDevice>, int) - Constructor for class neureka.backend.main.implementations.CLImplementation

CLMatMul - Class in neureka.backend.main.implementations.matmul

CLMatMul() - Constructor for class neureka.backend.main.implementations.matmul.CLMatMul

clone() - Method in class neureka.backend.api.BackendContext

This method produces a shallow copy of this BackendContext.

clone() - Method in class neureka.view.NDPrintSettings

CLRandomization - Class in neureka.backend.main.implementations.elementwise

CLRandomization() - Constructor for class neureka.backend.main.implementations.elementwise.CLRandomization

CLReduce - Class in neureka.backend.main.operations.linear.internal.opencl

CLReduce(CLReduce.Type) - Constructor for class neureka.backend.main.operations.linear.internal.opencl.CLReduce

CLReduce.Type - Enum in neureka.backend.main.operations.linear.internal.opencl

CLScalarBroadcast - Class in neureka.backend.main.implementations.broadcast

CLScalarBroadcast(String, String, String) - Constructor for class neureka.backend.main.implementations.broadcast.CLScalarBroadcast

CLScalarBroadcastAddition - Class in neureka.backend.main.implementations.broadcast

CLScalarBroadcastAddition(String) - Constructor for class neureka.backend.main.implementations.broadcast.CLScalarBroadcastAddition

CLScalarBroadcastDivision - Class in neureka.backend.main.implementations.broadcast

CLScalarBroadcastDivision(String) - Constructor for class neureka.backend.main.implementations.broadcast.CLScalarBroadcastDivision

CLScalarBroadcastIdentity - Class in neureka.backend.main.implementations.broadcast

CLScalarBroadcastIdentity(String) - Constructor for class neureka.backend.main.implementations.broadcast.CLScalarBroadcastIdentity

CLScalarBroadcastModulo - Class in neureka.backend.main.implementations.broadcast

CLScalarBroadcastModulo(String) - Constructor for class neureka.backend.main.implementations.broadcast.CLScalarBroadcastModulo

CLScalarBroadcastMultiplication - Class in neureka.backend.main.implementations.broadcast

CLScalarBroadcastMultiplication(String) - Constructor for class neureka.backend.main.implementations.broadcast.CLScalarBroadcastMultiplication

CLScalarBroadcastPower - Class in neureka.backend.main.implementations.broadcast

CLScalarBroadcastPower(String) - Constructor for class neureka.backend.main.implementations.broadcast.CLScalarBroadcastPower

CLScalarBroadcastSubtraction - Class in neureka.backend.main.implementations.broadcast

CLScalarBroadcastSubtraction(String) - Constructor for class neureka.backend.main.implementations.broadcast.CLScalarBroadcastSubtraction

CLScalarFunction - Class in neureka.backend.main.implementations.scalar

CLScalarFunction(ScalarFun) - Constructor for class neureka.backend.main.implementations.scalar.CLScalarFunction

CLSettings - Class in neureka.backend.ocl

OpenCL related settings for the CLBackend extension.

CLSettings() - Constructor for class neureka.backend.ocl.CLSettings

CLSum - Class in neureka.backend.main.operations.linear.internal.opencl

CLSum() - Constructor for class neureka.backend.main.operations.linear.internal.opencl.CLSum

CommonMachine - Class in neureka.devices.host.machine

Stuff common to Hardware and ConcreteMachine.

CommonMachine(Hardware, Runtime) - Constructor for class neureka.devices.host.machine.CommonMachine

CommonMachine(String, BasicMachine[]) - Constructor for class neureka.devices.host.machine.CommonMachine

new MemoryThreads[] { SYSTEM, L3, L2, L1 } or new MemoryThreads[] { SYSTEM, L2, L1 } or in worst case new MemoryThreads[] { SYSTEM, L1 }

compareTo(Hardware) - Method in class neureka.devices.host.machine.Hardware

compileAdHocKernel(String, String) - Method in class neureka.devices.opencl.OpenCLDevice

This method compiles so called "ad hoc" kernel.

compileAndGetAdHocKernel(String, String) - Method in class neureka.devices.opencl.OpenCLDevice

This method compiles and returns the KernelCaller for a so called "ad hoc" kernel.

Component<O> - Interface in neureka.common.composition

This interface alongside the AbstractComponentOwner class define a simple component system.

Component.IsBeing - Enum in neureka.common.composition

Entries of this enum represent events describing updates to the state of the owner of a given Component instance.

Component.OwnerChangeRequest<O> - Interface in neureka.common.composition

Component.OwnerChangeRequest implementation instances will be passed to the Component.update(OwnerChangeRequest) method which inform a given component about a state change related to said component.

ComponentOwner<C> - Interface in neureka.common.composition

A component owner is a thing holding components which can be accessed by their type class.

concat() - Method in class neureka.math.Functions

concatAt(int, Nda<V>, Nda<V>...) - Method in interface neureka.Nda

This method concatenates the provided nd-arrays together with this nd-array along a specified axis.

concatAt(int, Nda<V>) - Method in interface neureka.Nda

This method concatenates the provided nd-array together with this nd-array along a specified axis.

concatAt(int, Nda<V>, Nda<V>...) - Method in interface neureka.Tensor

This method concatenates the provided nd-arrays together with this nd-array along a specified axis.

concatAt(int, Nda<V>) - Method in interface neureka.Tensor

This method concatenates the provided nd-array together with this nd-array along a specified axis.

ConcreteMachine - Class in neureka.devices.host.machine

confidence() - Method in class neureka.backend.api.BackendExtension.DeviceOption

configure(Object) - Static method in class neureka.Neureka

This allows you to configure Neureka using a Groovy DSL.

construct(int[], int[], int[]) - Static method in class neureka.ndim.config.types.permuted.Permuted1DConfiguration

construct(int[], int[], int[]) - Static method in class neureka.ndim.config.types.permuted.Permuted2DConfiguration

construct(int[], int[], int[]) - Static method in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

construct(int[], int[], int[]) - Static method in class neureka.ndim.config.types.permuted.PermutedNDConfiguration

construct() - Static method in class neureka.ndim.config.types.simple.Simple0DConfiguration

construct(int[], int[]) - Static method in class neureka.ndim.config.types.simple.Simple1DConfiguration

construct(int[], int[]) - Static method in class neureka.ndim.config.types.simple.Simple2DConfiguration

construct(int[], int[]) - Static method in class neureka.ndim.config.types.simple.Simple3DConfiguration

construct(int[], int[]) - Static method in class neureka.ndim.config.types.simple.SimpleNDConfiguration

construct(int[], int[]) - Static method in class neureka.ndim.config.types.sliced.Sliced0DConfiguration

construct(int[], int[], int[], int[], int[]) - Static method in class neureka.ndim.config.types.sliced.Sliced1DConfiguration

construct(int[], int[], int[], int[], int[]) - Static method in class neureka.ndim.config.types.sliced.Sliced2DConfiguration

construct(int[], int[], int[], int[], int[]) - Static method in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

construct(int[], int[], int[], int[], int[]) - Static method in class neureka.ndim.config.types.sliced.SlicedNDConfiguration

construct(int[]) - Static method in class neureka.ndim.config.types.views.virtual.VirtualNDConfiguration

contains(Tensor<V>) - Method in class neureka.devices.AbstractBaseDevice

contains(Tensor<V>) - Method in interface neureka.devices.Storage

contains(Tensor<V>) - Method in interface neureka.Tensor

This method name translates to the "in" keyword in Kotlin! The same is true for the "isCase" method in Groovy.

conv() - Method in class neureka.math.Functions

conv(Tensor<V>) - Method in interface neureka.Tensor

This method performs convolution between this tensor and the one passed as argument.

convDot(Tensor<V>) - Method in interface neureka.Tensor

This method performs a convolutional based dot product between the last dimension of this tensor and the first dimension of the passed tensor.

convert(Object, Class<T>) - Method in class neureka.common.utility.DataConverter

This method embodies the purpose of this class.

convertToHolder(Object) - Method in class neureka.dtype.custom.F32

convertToHolder(Object) - Method in class neureka.dtype.custom.F64

convertToHolder(Object) - Method in class neureka.dtype.custom.I16

convertToHolder(Object) - Method in class neureka.dtype.custom.I32

convertToHolder(Object) - Method in class neureka.dtype.custom.I64

convertToHolder(Object) - Method in class neureka.dtype.custom.I8

convertToHolder(Object) - Method in class neureka.dtype.custom.UI16

convertToHolder(Object) - Method in class neureka.dtype.custom.UI32

convertToHolder(Object) - Method in class neureka.dtype.custom.UI64

convertToHolder(Object) - Method in class neureka.dtype.custom.UI8

convertToHolder(Object) - Method in interface neureka.dtype.NumericType

This method is a generic converter from any object to an instance of the HolderType parameter specified by an implementation of this interface.

convertToHolderArray(Object) - Method in class neureka.dtype.custom.F32

convertToHolderArray(Object) - Method in class neureka.dtype.custom.F64

convertToHolderArray(Object) - Method in class neureka.dtype.custom.I16

convertToHolderArray(Object) - Method in class neureka.dtype.custom.I32

convertToHolderArray(Object) - Method in class neureka.dtype.custom.I64

convertToHolderArray(Object) - Method in class neureka.dtype.custom.I8

convertToHolderArray(Object) - Method in class neureka.dtype.custom.UI16

convertToHolderArray(Object) - Method in class neureka.dtype.custom.UI32

convertToHolderArray(Object) - Method in class neureka.dtype.custom.UI64

convertToHolderArray(Object) - Method in class neureka.dtype.custom.UI8

convertToHolderArray(Object) - Method in interface neureka.dtype.NumericType

This method is a generic converter from any object to an instance of the HolderType parameter specified by an implementation of this interface.

convertToTarget(Object) - Method in class neureka.dtype.custom.F32

convertToTarget(Object) - Method in class neureka.dtype.custom.F64

convertToTarget(Object) - Method in class neureka.dtype.custom.I16

convertToTarget(Object) - Method in class neureka.dtype.custom.I32

convertToTarget(Object) - Method in class neureka.dtype.custom.I64

convertToTarget(Object) - Method in class neureka.dtype.custom.I8

convertToTarget(Object) - Method in class neureka.dtype.custom.UI16

convertToTarget(Object) - Method in class neureka.dtype.custom.UI32

convertToTarget(Object) - Method in class neureka.dtype.custom.UI64

convertToTarget(Object) - Method in class neureka.dtype.custom.UI8

convertToTarget(Object) - Method in interface neureka.dtype.NumericType

This method is a generic converter from any object to an instance of the TargetType parameter specified by an implementation of this interface.

convertToTargetArray(Object) - Method in class neureka.dtype.custom.F32

convertToTargetArray(Object) - Method in class neureka.dtype.custom.F64

convertToTargetArray(Object) - Method in class neureka.dtype.custom.I16

convertToTargetArray(Object) - Method in class neureka.dtype.custom.I32

convertToTargetArray(Object) - Method in class neureka.dtype.custom.I64

convertToTargetArray(Object) - Method in class neureka.dtype.custom.I8

convertToTargetArray(Object) - Method in class neureka.dtype.custom.UI16

convertToTargetArray(Object) - Method in class neureka.dtype.custom.UI32

convertToTargetArray(Object) - Method in class neureka.dtype.custom.UI64

convertToTargetArray(Object) - Method in class neureka.dtype.custom.UI8

convertToTargetArray(Object) - Method in interface neureka.dtype.NumericType

This method is a generic converter from any object to an instance of the TargetArrayType parameter specified by an implementation of this interface.

Convolution - Class in neureka.backend.main.operations.linear

Convolution() - Constructor for class neureka.backend.main.operations.linear.Convolution

ConvUtil - Class in neureka.backend.main.operations

ConvUtil() - Constructor for class neureka.backend.main.operations.ConvUtil

COPY - Class in neureka.backend.main.operations.linear.internal.blas

The ?copy routines perform a vector-vector operation defined as y = x, where x and y are vectors.

COPY() - Constructor for class neureka.backend.main.operations.linear.internal.blas.COPY

copyOf(T[]) - Static method in class neureka.backend.main.operations.linear.internal.blas.COPY

cores - Variable in class neureka.devices.host.machine.CommonMachine

The total number of processor cores.

cos() - Method in class neureka.math.Functions

cos() - Method in interface neureka.Tensor

This method is a functionally identical to the following alternatives:

COSINUS - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

Cosinus - Class in neureka.backend.main.operations.functions

Cosinus() - Constructor for class neureka.backend.main.operations.functions.Cosinus

count() - Method in class neureka.devices.ReferenceCounter

count(Predicate<V>) - Method in interface neureka.Nda

Iterates over every element of this nd-array, and counts the number of times the provided lambda matches the items of this array.

count(Predicate<Integer>) - Method in interface neureka.Shape

CPU - Class in neureka.devices.host

The CPU class, one of many implementations of the Device interface, is simply supposed to be an API for dispatching threaded workloads onto the CPU as well as reading from or writing to tensors it stores.

CPU.IndexedWorkload - Interface in neureka.devices.host

CPU.JVMExecutor - Class in neureka.devices.host

The CPU.JVMExecutor offers a similar functionality as the parallel stream API, however it differs in that the CPU.JVMExecutor is processing CPU.RangeWorkload lambdas instead of simply exposing a single index or concrete elements for a given workload size.

CPU.RangeWorkload - Interface in neureka.devices.host

A simple functional interface for executing a range whose implementations will either be executed sequentially or they are being dispatched to a thread-pool, given that the provided workload is large enough.

CPUBackend - Class in neureka.backend.cpu

This class loads the CPU operations into the Neureka library context.

CPUBackend() - Constructor for class neureka.backend.cpu.CPUBackend

CPUBiElementWise - Class in neureka.backend.main.implementations.elementwise

CPUBiElementWise() - Constructor for class neureka.backend.main.implementations.elementwise.CPUBiElementWise

CPUBiElementWiseAddition - Class in neureka.backend.main.implementations.elementwise

CPUBiElementWiseAddition() - Constructor for class neureka.backend.main.implementations.elementwise.CPUBiElementWiseAddition

CPUBiElementWiseDivision - Class in neureka.backend.main.implementations.elementwise

CPUBiElementWiseDivision() - Constructor for class neureka.backend.main.implementations.elementwise.CPUBiElementWiseDivision

CPUBiElementWiseModulo - Class in neureka.backend.main.implementations.elementwise

CPUBiElementWiseModulo() - Constructor for class neureka.backend.main.implementations.elementwise.CPUBiElementWiseModulo

CPUBiElementWiseMultiplication - Class in neureka.backend.main.implementations.elementwise

CPUBiElementWiseMultiplication() - Constructor for class neureka.backend.main.implementations.elementwise.CPUBiElementWiseMultiplication

CPUBiElementWisePower - Class in neureka.backend.main.implementations.elementwise

CPUBiElementWisePower() - Constructor for class neureka.backend.main.implementations.elementwise.CPUBiElementWisePower

CPUBiElementWiseSubtraction - Class in neureka.backend.main.implementations.elementwise

CPUBiElementWiseSubtraction() - Constructor for class neureka.backend.main.implementations.elementwise.CPUBiElementWiseSubtraction

CPUBiFun - Interface in neureka.backend.main.implementations.fun.api

CPUBroadcast - Class in neureka.backend.main.implementations.broadcast

CPUBroadcast() - Constructor for class neureka.backend.main.implementations.broadcast.CPUBroadcast

CPUBroadcastAddition - Class in neureka.backend.main.implementations.broadcast

CPUBroadcastAddition() - Constructor for class neureka.backend.main.implementations.broadcast.CPUBroadcastAddition

CPUBroadcastDivision - Class in neureka.backend.main.implementations.broadcast

CPUBroadcastDivision() - Constructor for class neureka.backend.main.implementations.broadcast.CPUBroadcastDivision

CPUBroadcastModulo - Class in neureka.backend.main.implementations.broadcast

CPUBroadcastModulo() - Constructor for class neureka.backend.main.implementations.broadcast.CPUBroadcastModulo

CPUBroadcastMultiplication - Class in neureka.backend.main.implementations.broadcast

CPUBroadcastMultiplication() - Constructor for class neureka.backend.main.implementations.broadcast.CPUBroadcastMultiplication

CPUBroadcastPower - Class in neureka.backend.main.implementations.broadcast

CPUBroadcastPower() - Constructor for class neureka.backend.main.implementations.broadcast.CPUBroadcastPower

CPUBroadcastSubtraction - Class in neureka.backend.main.implementations.broadcast

CPUBroadcastSubtraction() - Constructor for class neureka.backend.main.implementations.broadcast.CPUBroadcastSubtraction

CPUBroadcastSummation - Class in neureka.backend.main.implementations.broadcast

CPUBroadcastSummation() - Constructor for class neureka.backend.main.implementations.broadcast.CPUBroadcastSummation

CPUConvolution - Class in neureka.backend.main.implementations.convolution

CPUConvolution() - Constructor for class neureka.backend.main.implementations.convolution.CPUConvolution

CPUDot - Class in neureka.backend.main.implementations.linear

CPUDot() - Constructor for class neureka.backend.main.implementations.linear.CPUDot

CPUElementwiseAssignFun - Class in neureka.backend.main.implementations.elementwise

CPUElementwiseAssignFun() - Constructor for class neureka.backend.main.implementations.elementwise.CPUElementwiseAssignFun

CPUElementwiseFunction - Class in neureka.backend.main.implementations.elementwise

CPUElementwiseFunction(ScalarFun) - Constructor for class neureka.backend.main.implementations.elementwise.CPUElementwiseFunction

CPUFun - Interface in neureka.backend.main.implementations.fun.api

CPUImplementation - Class in neureka.backend.main.implementations

This class is a wrapper class for the ImplementationFor<CPU> interface which enables a functional style of implementing the backend API!
It is used merely as a simple formality and implementation type specification.

CPUImplementation.AndImplementation - Interface in neureka.backend.main.implementations

This is represents the second step in the simple builder API for CPUImplementation instances.

CPUMatMul - Class in neureka.backend.main.implementations.matmul

This is a library internal class, do not depend on this.

CPUMatMul() - Constructor for class neureka.backend.main.implementations.matmul.CPUMatMul

CPURandomization - Class in neureka.backend.main.implementations.elementwise

CPURandomization() - Constructor for class neureka.backend.main.implementations.elementwise.CPURandomization

CPUReduce - Class in neureka.backend.main.operations.other.internal

An implementation of the min and max algorithm running on the CPU.

CPUReduce(CPUReduce.Type) - Constructor for class neureka.backend.main.operations.other.internal.CPUReduce

CPUReduce.Type - Enum in neureka.backend.main.operations.other.internal

CPUScalaBroadcastPower - Class in neureka.backend.main.implementations.broadcast

CPUScalaBroadcastPower() - Constructor for class neureka.backend.main.implementations.broadcast.CPUScalaBroadcastPower

CPUScalarBroadcast - Class in neureka.backend.main.implementations.broadcast

CPUScalarBroadcast() - Constructor for class neureka.backend.main.implementations.broadcast.CPUScalarBroadcast

CPUScalarBroadcastAddition - Class in neureka.backend.main.implementations.broadcast

CPUScalarBroadcastAddition() - Constructor for class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastAddition

CPUScalarBroadcastDivision - Class in neureka.backend.main.implementations.broadcast

CPUScalarBroadcastDivision() - Constructor for class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastDivision

CPUScalarBroadcastFunction - Class in neureka.backend.main.implementations.scalar

CPUScalarBroadcastFunction(ScalarFun) - Constructor for class neureka.backend.main.implementations.scalar.CPUScalarBroadcastFunction

CPUScalarBroadcastIdentity - Class in neureka.backend.main.implementations.broadcast

CPUScalarBroadcastIdentity() - Constructor for class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastIdentity

CPUScalarBroadcastModulo - Class in neureka.backend.main.implementations.broadcast

CPUScalarBroadcastModulo() - Constructor for class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastModulo

CPUScalarBroadcastMultiplication - Class in neureka.backend.main.implementations.broadcast

CPUScalarBroadcastMultiplication() - Constructor for class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastMultiplication

CPUScalarBroadcastSubtraction - Class in neureka.backend.main.implementations.broadcast

CPUScalarBroadcastSubtraction() - Constructor for class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastSubtraction

CPUScalarFunction - Class in neureka.backend.main.implementations.scalar

CPUScalarFunction(ScalarFun) - Constructor for class neureka.backend.main.implementations.scalar.CPUScalarFunction

CPUSum - Class in neureka.backend.main.operations.other.internal

An implementation of the sum and may algorithm running on the CPU.

CPUSum() - Constructor for class neureka.backend.main.operations.other.internal.CPUSum

create(Tensor<V>) - Method in class neureka.optimization.implementations.AdaGradFactory

create(Tensor<V>) - Method in class neureka.optimization.implementations.ADAMFactory

create(Tensor<V>, Tensor<V>) - Method in class neureka.optimization.implementations.ADAMFactory

create(Tensor<V>) - Method in class neureka.optimization.implementations.MomentumFactory

create(Tensor<V>) - Method in class neureka.optimization.implementations.RMSPropFactory

create(Tensor<V>) - Method in class neureka.optimization.implementations.SGDFactory

create(Tensor<V>) - Method in interface neureka.optimization.OptimizerFactory

createDeconvolutionFor(String) - Static method in class neureka.backend.main.operations.ConvUtil

CSVHandle - Class in neureka.devices.file

This class is one of many extensions of the AbstractFileHandle which is therefore ultimately an implementation of the FileHandle interface.

CSVHandle(String, Map<String, Object>) - Constructor for class neureka.devices.file.CSVHandle

currentCount() - Method in class neureka.devices.ReferenceCounter.ChangeEvent

D

D1C - Class in neureka.ndim.config.types

An abstract class for NDConfigurations which are representing tensors of rank 1, meaning the name of this class translates to "Dimension-1-Configuration".

D1C() - Constructor for class neureka.ndim.config.types.D1C

D2C - Class in neureka.ndim.config.types

An abstract class for NDConfigurations which are representing tensors of rank 2, meaning the name of this class translates to "Dimension-^2-Configuration".

D2C() - Constructor for class neureka.ndim.config.types.D2C

D3C - Class in neureka.ndim.config.types

An abstract class for NDConfigurations which are representing tensors of rank 3, meaning the name of this class translates to "Dimension-3-Configuration".

D3C() - Constructor for class neureka.ndim.config.types.D3C

Data<V> - Interface in neureka

A wrapper type for the raw data array of a tensor/nd-array, which is typically provided by implementations of the Device interface.

dataArrayType() - Method in class neureka.dtype.DataType

DataConverter - Class in neureka.common.utility

This class is a singleton.

DataConverter.ForTensor - Class in neureka.common.utility

This is a stateful and parallelized converter for converting the internal data array of a tensor to another data array based on a provided lambda.

DataConverter.Utility - Class in neureka.common.utility

This is a static utility class containing the actual conversion logic which is usually referenced by the Converter lambdas via method signatures...

dataType() - Method in interface neureka.Data

dataType() - Method in class neureka.devices.AbstractDeviceData

DataType<T> - Class in neureka.dtype

This class is a Multiton implementation for wrapping and representing type classes.

dataType - Variable in enum neureka.Tensor.ImageType

debug() - Method in class neureka.Neureka.Settings

debug(Object) - Method in class neureka.Neureka.Settings

This allows you to configure Neureka using a Groovy DSL.

Debug() - Constructor for class neureka.Neureka.Settings.Debug

decrement() - Method in class neureka.devices.ReferenceCounter

decrement(int[], int[]) - Static method in class neureka.ndim.config.NDConfiguration.Utility

decrement() - Method in interface neureka.ndim.iterator.NDIterator

decrement() - Method in class neureka.ndim.iterator.types.permuted.Permuted2DCIterator

decrement() - Method in class neureka.ndim.iterator.types.permuted.Permuted3DCIterator

decrement() - Method in class neureka.ndim.iterator.types.simple.Simple1DCIterator

decrement() - Method in class neureka.ndim.iterator.types.simple.Simple2DCIterator

decrement() - Method in class neureka.ndim.iterator.types.simple.Simple3DCIterator

decrement() - Method in class neureka.ndim.iterator.types.sliced.Sliced1DCIterator

decrement() - Method in class neureka.ndim.iterator.types.sliced.Sliced2DCIterator

decrement() - Method in class neureka.ndim.iterator.types.sliced.Sliced3DCIterator

decrement() - Method in class neureka.ndim.iterator.types.sliced.SlicedNDIterator

decrement() - Method in class neureka.ndim.iterator.types.virtual.VirtualNDIterator

decrementUsageCount() - Method in class neureka.devices.AbstractDeviceData

decrementUsageCount() - Method in interface neureka.devices.DeviceData

deepClone() - Method in interface neureka.Tensor

This is almost identical to the Tensor.deepCopy() method except that the returned tensor will have autograd support, meaning that the cloning will be part of the autograd computation graph, and backpropagation will traverse the cloned tensor as well.

deepCopy() - Method in interface neureka.Nda

This method creates and returns a new nd-array instance which is not only a copy of the configuration of this nd-array but also a copy of the underlying data array.

deepCopy() - Method in interface neureka.Tensor

This method creates and returns a new nd-array instance which is not only a copy of the configuration of this nd-array but also a copy of the underlying data array.

delete() - Method in interface neureka.MutateTensor

Although tensors will be garbage collected when they are not strongly referenced, there is also the option to manually free up the tensor and its associated data in a native environment.

dependsOn(int) - Method in interface neureka.math.Function

Use this to determine if this function directly or indirectly references an input with the provided index.

dependsOn(int) - Method in class neureka.math.implementations.FunctionConstant

dependsOn(int) - Method in class neureka.math.implementations.FunctionInput

dependsOn(int) - Method in class neureka.math.implementations.FunctionNode

dependsOn(int) - Method in class neureka.math.implementations.FunctionVariable

derivationCode() - Method in interface neureka.backend.main.implementations.fun.api.ScalarFun

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarAbsolute

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarCbrt

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarCosinus

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarExp

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarGaSU

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarGaTU

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarGaussian

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarGaussianFast

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarGeLU

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarIdentity

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarLog10

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarLogarithm

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarQuadratic

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarReLU

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarSeLU

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarSigmoid

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarSiLU

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarSinus

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarSoftplus

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarSoftsign

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarSqrt

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarTanh

derivationCode() - Method in class neureka.backend.main.implementations.fun.ScalarTanhFast

derive(int[], Tensor[], Function<Integer, Tensor<?>>) - Static method in class neureka.backend.main.operations.operator.Multiplication

derive(double[], int, int) - Method in interface neureka.math.Function

Calculates the derivative of a particular input with respect to the output of this Function based on the provided array of inputs, an index targeting the input to be derived and an index for input dependent indexing.

derive(double[], int) - Method in interface neureka.math.Function

Calculates the derivative of a particular input with respect to the output of this Function based on the provided array of inputs and an index targeting the input to be derived.

derive(Tensor<T>[], int, int) - Method in interface neureka.math.Function

derive(Tensor<T>[], int) - Method in interface neureka.math.Function

derive(List<Tensor<T>>, int, int) - Method in interface neureka.math.Function

derive(List<Tensor<T>>, int) - Method in interface neureka.math.Function

derive(double[], int) - Method in class neureka.math.implementations.FunctionConstant

derive(double[], int, int) - Method in class neureka.math.implementations.FunctionConstant

derive(double[], int) - Method in class neureka.math.implementations.FunctionInput

derive(double[], int, int) - Method in class neureka.math.implementations.FunctionInput

derive(double[], int, int) - Method in class neureka.math.implementations.FunctionNode

derive(double[], int) - Method in class neureka.math.implementations.FunctionNode

derive(double[], int) - Method in class neureka.math.implementations.FunctionVariable

derive(double[], int, int) - Method in class neureka.math.implementations.FunctionVariable

detach() - Method in interface neureka.MutateTensor

This method detaches this tensor from its underlying computation-graph or simply does nothing if no graph is present.
Nodes within a computation graph are instances of the "GraphNode" class which are also simple components of the tensors they represent in the graph.

detached() - Method in class neureka.fluent.slicing.AxisSliceBuilder

detached() - Method in class neureka.fluent.slicing.SliceBuilder

detached() - Method in interface neureka.fluent.slicing.states.AxisOrGet

This method concludes the slicing API by performing the actual slicing and returning the resulting Tensor instance based on the previously specified slice configuration...

detached() - Method in interface neureka.fluent.slicing.states.AxisOrGetTensor

This method concludes the slicing API by performing the actual slicing and returning the resulting Tensor instance based on the previously specified slice configuration...

detached() - Method in interface neureka.fluent.slicing.states.StepsOrAxisOrGetTensor

This method concludes the slicing API by performing the actual slicing and returning the resulting Tensor instance based on the previously specified slice configuration...

detached() - Method in interface neureka.Tensor

This method returns a new tensor detached from any underlying computation-graph or simply does nothing if no graph is present.
Nodes within a computation graph are instances of the "GraphNode" class which are also simple components of the tensors they represent in the graph.

device() - Method in class neureka.backend.api.BackendExtension.DeviceOption

Device<V> - Interface in neureka.devices

Implementations of this represent computational devices for storing tensors (instances of the Tensor<V> class), which may also expose a useful API for executing operations on tensors (used in backend operations).

Device.Access<V> - Interface in neureka.devices

Implementations of this represent the access to tensors stored on a device in order to read from or write to said tensor.

Device.In - Interface in neureka.devices

The second part of the method chain of the fluent API for executing tensors on this Device temporarily.

Device.Writer - Interface in neureka.devices

Instances of this complete a request for writing to an accessed tensor stored on a device.

DeviceAlgorithm<C extends DeviceAlgorithm<C>> - Interface in neureka.backend.api

A DeviceAlgorithm is an advanced form of Algorithm which delegates the execution to implementations of ImplementationFor specific Device types.

DeviceCleaner - Interface in neureka.devices

DeviceData<V> - Interface in neureka.devices

A sub-interface of the Data interface providing more device specific methods.

DeviceOption(Device<?>, double) - Constructor for class neureka.backend.api.BackendExtension.DeviceOption

DeviceQuery - Class in neureka.devices.opencl.utility

A program that queries and prints information about all available devices.

DimFit - Class in neureka.backend.main.operations.other

DimFit() - Constructor for class neureka.backend.main.operations.other.DimFit

DimTrim - Class in neureka.backend.main.operations.other

DimTrim() - Constructor for class neureka.backend.main.operations.other.DimTrim

dimTrim() - Method in class neureka.math.Functions

dimtrim() - Method in interface neureka.Tensor

This creates a new tensor with the same underlying Data and whose shape is trimmed.

dispatch(Function, ExecutionCall<? extends Device<?>>) - Method in class neureka.backend.api.template.algorithms.FallbackAlgorithm

dispose() - Method in interface neureka.backend.api.BackendExtension

Tells this extension to dispose itself.

dispose() - Method in class neureka.backend.api.template.operations.OperationBuilder

dispose() - Method in class neureka.backend.cpu.CPUBackend

dispose() - Method in class neureka.backend.ocl.CLBackend

This method will free all the resources occupied by this context, meaning that all platforms and their devices will be disposed.

dispose() - Method in interface neureka.devices.Device

This method signals the device to get ready for garbage collection.

dispose() - Method in class neureka.devices.file.FileDevice

dispose() - Method in class neureka.devices.host.CPU

This method will shut down the internal thread-pool used by this class to execute JVM/CPU based operations in parallel.

dispose() - Method in class neureka.devices.opencl.OpenCLDevice

This method tells the to restore all tensors stored on it and release all resources.

dispose() - Method in class neureka.devices.opencl.OpenCLPlatform

div() - Method in class neureka.math.Functions

div(Tensor<V>) - Method in interface neureka.Tensor

This method will produce the quotient of two tensors with the same rank (or two ranks which can be made compatible with padding ones), where the left operand is this Tensor instance and the right operand is the tensor passed to the method.

div(V) - Method in interface neureka.Tensor

divAssign() - Method in class neureka.math.Functions

divAssign(Tensor<T>) - Method in interface neureka.MutateTensor

divide(int, CPU.RangeWorkload) - Method in class neureka.devices.host.concurrent.WorkScheduler.Divider

divide(int, int, CPU.RangeWorkload) - Method in class neureka.devices.host.concurrent.WorkScheduler.Divider

Divider(ExecutorService) - Constructor for class neureka.devices.host.concurrent.WorkScheduler.Divider

Division - Class in neureka.backend.main.operations.operator

Division() - Constructor for class neureka.backend.main.operations.operator.Division

doesNotExist() - Method in interface neureka.Nda.Item

DOT - Class in neureka.backend.main.operations.linear.internal.blas

The ?dot routines perform a vector-vector reduction operation defined as Equation where xi and yi are elements of vectors x and y.

DOT() - Constructor for class neureka.backend.main.operations.linear.internal.blas.DOT

dot() - Method in class neureka.math.Functions

dot(Tensor<V>) - Method in interface neureka.Tensor

Performs a dot product between the last dimension of this tensor and the first dimension of the provided tensor.

DotProduct - Class in neureka.backend.main.operations.linear

DotProduct() - Constructor for class neureka.backend.main.operations.linear.DotProduct

DotProductAlgorithm - Class in neureka.backend.main.algorithms

DotProductAlgorithm() - Constructor for class neureka.backend.main.algorithms.DotProductAlgorithm

doubleToBigInteger(double[]) - Static method in class neureka.common.utility.DataConverter.Utility

doubleToBool(double[]) - Static method in class neureka.common.utility.DataConverter.Utility

doubleToByte(double[]) - Static method in class neureka.common.utility.DataConverter.Utility

doubleToFloat(double[]) - Static method in class neureka.common.utility.DataConverter.Utility

doubleToInt(double[]) - Static method in class neureka.common.utility.DataConverter.Utility

doubleToLong(double[]) - Static method in class neureka.common.utility.DataConverter.Utility

doubleToShort(double[]) - Static method in class neureka.common.utility.DataConverter.Utility

dtype() - Method in class neureka.Neureka.Settings

dtype(Object) - Method in class neureka.Neureka.Settings

This allows you to configure Neureka using a Groovy DSL.

DType() - Constructor for class neureka.Neureka.Settings.DType

E

elements() - Method in interface neureka.Shape

ElementwiseAlgorithm - Class in neureka.backend.main.algorithms

This is lambda based Algorithm implementation providing some basic functionality for implementing custom activation functions.

ElementwiseAlgorithm() - Constructor for class neureka.backend.main.algorithms.ElementwiseAlgorithm

ElemWiseUtil - Class in neureka.backend.main.operations

Methods inside this utility class execute only some ExecutionCall arguments in groups if their total number exceeds the arity of an operation.

ElemWiseUtil() - Constructor for class neureka.backend.main.operations.ElemWiseUtil

endsFrom(int[]) - Static method in class neureka.backend.main.operations.other.DimTrim

ENVIRONMENT - Static variable in class neureka.devices.host.machine.ConcreteMachine

equals(Object) - Method in class neureka.common.utility.Cache.LazyEntry

equals(Object) - Method in class neureka.devices.host.machine.BasicMachine

equals(Object) - Method in class neureka.devices.host.machine.CommonMachine

equals(Object) - Method in class neureka.devices.host.machine.ConcreteMachine

equals(Object) - Method in class neureka.devices.host.machine.Hardware

equals(Object) - Method in class neureka.devices.opencl.KernelCode

equals(Object) - Method in class neureka.dtype.DataType

equals(Object) - Method in class neureka.ndim.config.AbstractNDC

equals(NDConfiguration) - Method in class neureka.ndim.config.AbstractNDC

equals(NDConfiguration) - Method in interface neureka.ndim.config.NDConfiguration

error() - Method in class neureka.autograd.ADTarget

errorCorrectionSupport() - Method in class neureka.devices.opencl.OpenCLDevice

Estimator(boolean) - Constructor for class neureka.backend.api.Call.Validator.Estimator

every(Predicate<V>) - Method in interface neureka.Nda

Iterates over every element of this nd-array, and checks whether all elements are true according to the provided lambda.

every(Predicate<Integer>) - Method in interface neureka.Shape

EXCELLENT - Static variable in interface neureka.backend.api.fun.SuitabilityPredicate

execute() - Method in class neureka.autograd.JITProp

This method triggers the continuation of the back-propagation which has been put on hold by saving the pending graph nodes inside this class.

execute(Function, ExecutionCall<? extends Device<?>>) - Method in interface neureka.backend.api.fun.Execution

execute(Function, ExecutionCall<?>) - Method in interface neureka.backend.api.Operation

execute(Function, ExecutionCall<? extends Device<?>>) - Method in class neureka.backend.api.template.algorithms.AbstractFunAlgorithm

execute(Function, ExecutionCall<? extends Device<?>>) - Method in class neureka.backend.api.template.algorithms.AbstractFunDeviceAlgorithm

execute(Function, ExecutionCall<? extends Device<?>>) - Method in class neureka.backend.api.template.algorithms.FallbackAlgorithm

execute(boolean, double[], double[], double[], int, int, int) - Static method in class neureka.backend.main.implementations.matmul.CPUMatMul

execute(boolean, float[], float[], float[], int, int, int) - Static method in class neureka.backend.main.implementations.matmul.CPUMatMul

execute(boolean, long[], long[], long[], int, int, int) - Static method in class neureka.backend.main.implementations.matmul.CPUMatMul

execute(boolean, int[], int[], int[], int, int, int) - Static method in class neureka.backend.main.implementations.matmul.CPUMatMul

execute(ExecutionCall<? extends Device<?>>) - Method in interface neureka.backend.main.internal.FinalExecutor

execute(Function, ExecutionCall<?>) - Method in class neureka.backend.main.operations.indexer.Product

execute(Function, ExecutionCall<?>) - Method in class neureka.backend.main.operations.indexer.Summation

execute(Function, ExecutionCall<?>) - Method in class neureka.backend.main.operations.linear.Convolution

execute(Function, ExecutionCall<?>) - Method in class neureka.backend.main.operations.linear.MatMul

execute(Function, ExecutionCall<?>) - Method in class neureka.backend.main.operations.operator.Addition

execute(Function, ExecutionCall<?>) - Method in class neureka.backend.main.operations.operator.Division

execute(Function, ExecutionCall<?>) - Method in class neureka.backend.main.operations.operator.Modulo

execute(Function, ExecutionCall<?>) - Method in class neureka.backend.main.operations.operator.Multiplication

execute(Function, ExecutionCall<?>) - Method in class neureka.backend.main.operations.operator.Power

execute(Function, ExecutionCall<?>) - Method in class neureka.backend.main.operations.operator.Subtraction

execute(Function, ExecutionCall<?>) - Method in class neureka.backend.main.operations.other.AssignLeft

execute(Function, ExecutionCall<?>) - Method in class neureka.backend.main.operations.other.Cat

execute(int) - Method in interface neureka.devices.host.CPU.IndexedWorkload

execute(int, int) - Method in interface neureka.devices.host.CPU.RangeWorkload

execute(Tensor<?>...) - Method in interface neureka.math.Function.Callable

Warning: Tensors returned by this method are eligible for deletion when consumed by other functions.

execute(Call<?>) - Method in interface neureka.math.Function

Warning: Tensors returned by this method are eligible for deletion when consumed by other functions. Use this to pass more context information for execution of input tensors.

execute(Args, Tensor<?>...) - Method in interface neureka.math.Function

Warning: Tensors returned by this method are eligible for deletion when consumed by other functions.
Use this to call this Function alongside with some additional meta-arguments which will be passed to the underlying Operation(s).

execute(Tensor<?>...) - Method in interface neureka.math.Function

Warning: Tensors returned by this method are eligible for deletion when consumed by other functions.

execute(Tensor<?>[], int) - Method in interface neureka.math.Function

Warning: Tensors returned by this method are eligible for deletion when consumed by other functions.

execute(Args, Tensor<?>...) - Method in class neureka.math.implementations.FunctionConstant

execute(Args, Tensor<?>...) - Method in class neureka.math.implementations.FunctionInput

execute(Args, Tensor<?>...) - Method in class neureka.math.implementations.FunctionNode

execute(Args, Tensor<?>...) - Method in class neureka.math.implementations.FunctionVariable

executeChange() - Method in interface neureka.common.composition.Component.OwnerChangeRequest

This method will trigger the actual state change identified by the Component.IsBeing constant returned by the Component.OwnerChangeRequest.type() method.

executeDerive(Tensor<?>[], int, int) - Method in interface neureka.math.Function

Warning: Tensors returned by this method are eligible for deletion when consumed by other functions.

executeDerive(Tensor<?>[], int) - Method in interface neureka.math.Function

Warning: Tensors returned by this method are eligible for deletion when consumed by other functions.

executeDeviceAlgorithm(ExecutionCall<? extends Device<?>>) - Static method in class neureka.backend.api.template.algorithms.AbstractDeviceAlgorithm

executeFor(Function, ExecutionCall<? extends Device<?>>, FinalExecutor) - Static method in class neureka.backend.api.template.algorithms.AbstractDeviceAlgorithm

executeOnCommonDevice(ExecutionCall<?>, Supplier<R>) - Static method in class neureka.backend.api.template.algorithms.AbstractDeviceAlgorithm

executeRecursively(String, ExecutionCall<? extends Device<?>>) - Static method in class neureka.backend.main.operations.ConvUtil

Execution - Interface in neureka.backend.api.fun

Implementations of this functional interface is supposed to be the final execution procedure responsible for dispatching the execution further into the backend.

ExecutionCall<D extends Device<?>> - Class in neureka.backend.api

This class is a simple container holding references to a targeted Device, Operation and maybe some case specific meta Args needed to execute an array of input tensors which are also wrapped by this.

ExecutionCall.Builder<D extends Device<?>> - Class in neureka.backend.api

ExecutionPreparation - Interface in neureka.backend.api.fun

An Algorithm will typically produce a result when executing an ExecutionCall.

exists() - Method in interface neureka.Nda.Item

EXP - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

Exp - Class in neureka.backend.main.operations.functions

Exp() - Constructor for class neureka.backend.main.operations.functions.Exp

exp() - Method in class neureka.math.Functions

exp() - Method in interface neureka.Tensor

This method is a functionally identical to the following alternatives:

extension() - Method in interface neureka.devices.file.FileHandle

The file ending which comes after the '.' character...

Extensions - Class in neureka.backend.api

This is an internal class for managing the extension of any given BackendContext class.

Extensions() - Constructor for class neureka.backend.api.Extensions

F

F32 - Class in neureka.dtype.custom

F32() - Constructor for class neureka.dtype.custom.F32

F64 - Class in neureka.dtype.custom

F64() - Constructor for class neureka.dtype.custom.F64

FACTORY - Static variable in interface neureka.devices.file.FileHandle

FallbackAlgorithm - Class in neureka.backend.api.template.algorithms

FallbackAlgorithm(String, int, Operation) - Constructor for class neureka.backend.api.template.algorithms.FallbackAlgorithm

fastGaus() - Method in class neureka.math.Functions

fastTanh() - Method in class neureka.math.Functions

FileDevice - Class in neureka.devices.file

The FileDevice is a Device implementation responsible for reading tensors from and or writing them to a given directory.

FileHandle<FinalType,ValType> - Interface in neureka.devices.file

fileHandleOf(Tensor<?>) - Method in class neureka.devices.file.FileDevice

Filler<T> - Interface in neureka.ndim

Implementations of this ought to map the index of a tensor entry to a value which should be placed at that entry position.

fillRandomly(T, Arg.Seed) - Static method in class neureka.backend.main.implementations.elementwise.CPURandomization

fillRandomly(T, String) - Static method in class neureka.backend.main.implementations.elementwise.CPURandomization

fillRandomly(T, long) - Static method in class neureka.backend.main.implementations.elementwise.CPURandomization

filter(Predicate<V>) - Method in interface neureka.Nda

A convenience method for stream().filter( predicate ).

filter(Predicate<Integer>) - Method in interface neureka.Shape

FinalExecutor - Interface in neureka.backend.main.internal

find(Class<E>) - Method in class neureka.backend.api.BackendContext

Returns an Optional instance of the provided BackendExtension type or an empty Optional if no extension of that type was found.

find(String) - Method in interface neureka.backend.api.BackendExtension

The BackendContext does not handle Device instances directly.

find(String) - Method in class neureka.backend.cpu.CPUBackend

find(String) - Method in class neureka.backend.ocl.CLBackend

find(Class<T>) - Method in class neureka.common.composition.AbstractComponentOwner

This method finds a component of the given type/class and returns it as an Optional which may or may not be empty.

find(Class<T>) - Method in interface neureka.common.composition.ComponentOwner

This method finds a component of the given type/class and returns it as an Optional which may or may not be empty.

find(String...) - Static method in interface neureka.devices.Device

This method returns Device instances matching the given search parameter.

find(Class<D>, String...) - Static method in interface neureka.devices.Device

This method returns Device instances matching the given search parameters.

findAdHocKernel(String) - Method in class neureka.devices.opencl.OpenCLDevice

findCaptured() - Method in interface neureka.autograd.ADAction

Finds captured Tensor instances in this current action using reflection (This is usually a partial derivative).

findComponentIn(String, int) - Static method in class neureka.math.parsing.ParseUtil

findOrCompileAdHocKernel(String, Supplier<String>) - Method in class neureka.devices.opencl.OpenCLDevice

findParametersIn(String, int) - Static method in class neureka.math.parsing.ParseUtil

findRootTensor() - Method in class neureka.framing.Relation

This method tries to find the root data ancestor of this tensor.

findTip() - Static method in class neureka.devices.opencl.utility.Messages

finishedCount() - Method in class neureka.autograd.JITProp

first(Call.TensorCondition) - Method in class neureka.backend.api.Call.Validator

flatMap(Function<V, Stream<R>>) - Method in interface neureka.Nda

A convenience method for nda.stream().flatMap( mapper ), which turns this Nda into a Stream of its items.

flatten(Function, ExecutionCall<D>) - Static method in class neureka.backend.api.template.algorithms.AbstractDeviceAlgorithm

flattenForIndexer(Function, ExecutionCall<D>) - Static method in class neureka.backend.api.template.algorithms.AbstractDeviceAlgorithm

floatToBigInteger(float[]) - Static method in class neureka.common.utility.DataConverter.Utility

floatToByte(float[]) - Static method in class neureka.common.utility.DataConverter.Utility

floatToDouble(float[]) - Static method in class neureka.common.utility.DataConverter.Utility

floatToInt(float[]) - Static method in class neureka.common.utility.DataConverter.Utility

floatToLong(float[]) - Static method in class neureka.common.utility.DataConverter.Utility

floatToShort(float[]) - Static method in class neureka.common.utility.DataConverter.Utility

forDevice(Class<? extends D>) - Method in class neureka.backend.api.ini.BackendRegistry

forEachDerivative(BiConsumer<GraphNode<V>, ADAction>) - Method in class neureka.autograd.GraphNode

forEachTarget(Consumer<GraphNode<V>>) - Method in class neureka.autograd.GraphNode

foreignHolderBytesToTarget(byte[]) - Method in class neureka.dtype.custom.F32

foreignHolderBytesToTarget(byte[]) - Method in class neureka.dtype.custom.F64

foreignHolderBytesToTarget(byte[]) - Method in class neureka.dtype.custom.I16

foreignHolderBytesToTarget(byte[]) - Method in class neureka.dtype.custom.I32

foreignHolderBytesToTarget(byte[]) - Method in class neureka.dtype.custom.I64

foreignHolderBytesToTarget(byte[]) - Method in class neureka.dtype.custom.I8

foreignHolderBytesToTarget(byte[]) - Method in class neureka.dtype.custom.UI16

foreignHolderBytesToTarget(byte[]) - Method in class neureka.dtype.custom.UI32

foreignHolderBytesToTarget(byte[]) - Method in class neureka.dtype.custom.UI64

foreignHolderBytesToTarget(byte[]) - Method in class neureka.dtype.custom.UI8

foreignHolderBytesToTarget(byte[]) - Method in interface neureka.dtype.NumericType

forInputs(Tensor<?>[], Supplier<Result>) - Static method in class neureka.backend.main.memory.MemValidator

format(String, Object...) - Static method in class neureka.common.utility.LogUtil

forOperation(Class<? extends Operation>) - Method in interface neureka.backend.api.ini.ReceiveForDevice

ForTensor(Tensor<?>) - Constructor for class neureka.common.utility.DataConverter.ForTensor

frame() - Method in interface neureka.Tensor

This is a functionally identical alternative to Tensor.getFrame().

free(Tensor<T>) - Method in interface neureka.devices.Device

Use this to remove the provided tensor from this Device!

free(Tensor<T>) - Method in class neureka.devices.file.FileDevice

free() - Method in interface neureka.devices.file.FileHandle

An implementation of this method ought to "free" up the memory used to store a tensor.

free(Tensor<T>) - Method in class neureka.devices.host.CPU

free(Tensor<T>) - Method in class neureka.devices.opencl.OpenCLDevice

from(int) - Method in class neureka.fluent.slicing.AxisSliceBuilder

This method returns an instance of this very AxisSliceBuilder instance disguised by the To interface.

from(int) - Method in interface neureka.fluent.slicing.states.FromOrAt

This is the starting point for defining the slice range of a specified axis within the method chain/graph exposed by the slice builder API.

from(int) - Method in interface neureka.fluent.slicing.states.FromOrAtTensor

This is the starting point for defining the slice range of a specified axis within the method chain/graph exposed by the slice builder API.

FromOrAt<V> - Interface in neureka.fluent.slicing.states

This is the starting point for defining the slice range of a specified axis within the call transition graph exposed by the slice builder API.

FromOrAtTensor<V> - Interface in neureka.fluent.slicing.states

fullDelete() - Method in class neureka.devices.ReferenceCounter

fully() - Method in interface neureka.devices.Device.Writer

A convenience method for specifying that the entire data array of the accessed tensor should be written to.

FunAlgorithm - Class in neureka.backend.api.template.algorithms

FunAlgorithm(String) - Constructor for class neureka.backend.api.template.algorithms.FunAlgorithm

Function - Interface in neureka.math

Besides the Tensor class, which is the core class of Neureka, this interface and its implementations represents the second most important feature of this library.

Function.Callable - Interface in neureka.math

An API for calling a Function after having specified a set of Arg instances through the Function.with(Args) method.

FunctionCache - Class in neureka.math

This class is part of a given BackendContext instance responsible for caching Function references based on their String representation generated by Object.toString() as well as caching of results for active functions.

FunctionCache() - Constructor for class neureka.math.FunctionCache

FunctionConstant - Class in neureka.math.implementations

Instances of this implementation of the Function interface are leave nodes within the abstract syntax tree of a function, representing constant numeric values to a function.

FunctionConstant(String) - Constructor for class neureka.math.implementations.FunctionConstant

FunctionInput - Class in neureka.math.implementations

Instances of this implementation of the Function interface are leave nodes within the abstract syntax tree of a function, representing inputs to a function.

FunctionNode - Class in neureka.math.implementations

The most common type of Function which references other Functions to form an abstract syntax tree.

FunctionNode(Operation, List<Function>, boolean) - Constructor for class neureka.math.implementations.FunctionNode

FunctionParser - Class in neureka.math.parsing

The FunctionParser takes a BackendContext instance based on which it builds Function implementation instances, usually by parsing Strings.

FunctionParser(BackendContext) - Constructor for class neureka.math.parsing.FunctionParser

Functions - Class in neureka.math

Functions(boolean) - Constructor for class neureka.math.Functions

FunctionVariable - Class in neureka.math.implementations

Instances of this implementation of the Function interface are leave nodes within the abstract syntax tree of a function, representing indexed inputs to a function.

FunctionVariable(String) - Constructor for class neureka.math.implementations.FunctionVariable

FunDeviceAlgorithm - Class in neureka.backend.api.template.algorithms

FunDeviceAlgorithm(String) - Constructor for class neureka.backend.api.template.algorithms.FunDeviceAlgorithm

G

GASU - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

GaSU - Class in neureka.backend.main.operations.functions

The Self Gated Softsign Unit is based on the Softsign function (a computationally cheap non-exponential quasi Tanh) making it a polynomially based version of the GaTU function which is itself based on the Tanh function.

GaSU() - Constructor for class neureka.backend.main.operations.functions.GaSU

GATU - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

GaTU - Class in neureka.backend.main.operations.functions

The Self Gated Tanh Unit is based on the Tanh making it an exponentiation based version of the GaSU function which is itself based on the Softsign function (a computationally cheap non-exponential quasi Tanh).

GaTU() - Constructor for class neureka.backend.main.operations.functions.GaTU

gaus() - Method in class neureka.math.Functions

GAUSSIAN - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

Gaussian - Class in neureka.backend.main.operations.functions

Gaussian() - Constructor for class neureka.backend.main.operations.functions.Gaussian

GAUSSIAN_FAST - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

GaussianFast - Class in neureka.backend.main.operations.functions

GaussianFast() - Constructor for class neureka.backend.main.operations.functions.GaussianFast

gaussianFrom(long, double[]) - Static method in class neureka.backend.main.implementations.elementwise.CPURandomization

GELU - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

gelu(double) - Static method in class neureka.backend.main.implementations.fun.ScalarGeLU

GeLU - Class in neureka.backend.main.operations.functions

The GELU activation function is based on the standard Gaussian cumulative distribution function and is defined as x Φ( x ) and implemented as x * sigmoid(x * 1.702).

GeLU() - Constructor for class neureka.backend.main.operations.functions.GeLU

gelu() - Method in class neureka.math.Functions

GEMM - Class in neureka.backend.main.operations.linear.internal.blas

A collection of primitive sub-routines for matrix multiplication performed on continuous arrays which are designed so that they can be vectorized by the JVMs JIT compiler (AVX instructions).

GEMM() - Constructor for class neureka.backend.main.operations.linear.internal.blas.GEMM

GEMM.VectorOperationF32 - Interface in neureka.backend.main.operations.linear.internal.blas

GEMM.VectorOperationF64 - Interface in neureka.backend.main.operations.linear.internal.blas

get() - Method in class neureka.backend.api.Call.Builder

get(Class<T>) - Method in class neureka.backend.api.Call

get() - Method in class neureka.backend.api.LazyRef

get() - Method in class neureka.backend.api.Result

get(Class<T>) - Method in class neureka.common.composition.AbstractComponentOwner

This method tries to find a component inside the internal component array whose class matches the one provided.

get(Class<T>) - Method in interface neureka.common.composition.ComponentOwner

Use this to get the component of the specified component type class.

get() - Static method in class neureka.common.utility.DataConverter

This method returns the singleton.

get() - Method in interface neureka.Data

This returns the underlying raw data object of a nd-array or tensor of a backend specific type (e.g.

get(String...) - Static method in interface neureka.devices.Device

This method returns Device instances matching the given search parameter.

get(Class<D>, String...) - Static method in interface neureka.devices.Device

This method returns Device instances matching the given search parameters.

get() - Static method in class neureka.devices.host.CPU

Use this method to access the singleton instance of this CPU class, which is a Device type and default location for freshly instantiated Tensor instances.

get(String) - Method in class neureka.devices.opencl.KernelCache

get(cl_device_id) - Method in class neureka.devices.opencl.OpenCLPlatform

get() - Method in class neureka.fluent.slicing.AxisSliceBuilder

get() - Method in class neureka.fluent.slicing.SliceBuilder

This method will create and return a new slice tensor based on the provided configuration through methods like AxisSliceBuilder.from(int), AxisSliceBuilder.to(int) and AxisSliceBuilder.at(int)...

get() - Method in interface neureka.fluent.slicing.states.AxisOrGet

This method concludes the slicing API by performing the actual slicing and returning the resulting Tensor instance based on the previously specified slice configuration...

get() - Method in interface neureka.fluent.slicing.states.AxisOrGetTensor

This method concludes the slicing API by performing the actual slicing and returning the resulting Tensor instance based on the previously specified slice configuration...

get() - Method in interface neureka.fluent.slicing.states.StepsOrAxisOrGetTensor

This method concludes the slicing API by performing the actual slicing and returning the resulting Tensor instance based on the previously specified slice configuration...

Get<ValueType> - Interface in neureka.framing.fluent

get() - Method in interface neureka.framing.fluent.Get

get(List<Object>) - Method in class neureka.framing.NDFrame

get(Object...) - Method in class neureka.framing.NDFrame

get() - Method in class neureka.math.args.Arg

get(String, boolean) - Method in class neureka.math.FunctionCache

get(int...) - Method in interface neureka.Nda

The following method enables access to specific scalar elements within the nd-array.

get(Object...) - Method in interface neureka.Nda

The following method enables the creation of nd-array slices which access the same underlying data (possibly from a different view).

get(int) - Method in interface neureka.Nda

This getter method creates and returns a slice of the original nd-array.

get(Number) - Method in interface neureka.Nda

This getter method creates and returns a slice of the original nd-array.

get(Object) - Method in interface neureka.Nda

This method enables nd-array slicing! It takes a key of various types and configures a slice nd-array which shares the same underlying data as the original nd-array.

get() - Method in interface neureka.Nda.Item

Get the value at the targeted position or throw an exception if the item does not exist.

get(int) - Method in interface neureka.ndim.iterator.NDIterator

get() - Method in interface neureka.ndim.iterator.NDIterator

get(int) - Method in class neureka.ndim.iterator.types.permuted.Permuted2DCIterator

get() - Method in class neureka.ndim.iterator.types.permuted.Permuted2DCIterator

get(int) - Method in class neureka.ndim.iterator.types.permuted.Permuted3DCIterator

get() - Method in class neureka.ndim.iterator.types.permuted.Permuted3DCIterator

get(int) - Method in class neureka.ndim.iterator.types.simple.Simple1DCIterator

get() - Method in class neureka.ndim.iterator.types.simple.Simple1DCIterator

get(int) - Method in class neureka.ndim.iterator.types.simple.Simple2DCIterator

get() - Method in class neureka.ndim.iterator.types.simple.Simple2DCIterator

get(int) - Method in class neureka.ndim.iterator.types.simple.Simple3DCIterator

get() - Method in class neureka.ndim.iterator.types.simple.Simple3DCIterator

get(int) - Method in class neureka.ndim.iterator.types.sliced.Sliced1DCIterator

get() - Method in class neureka.ndim.iterator.types.sliced.Sliced1DCIterator

get(int) - Method in class neureka.ndim.iterator.types.sliced.Sliced2DCIterator

get() - Method in class neureka.ndim.iterator.types.sliced.Sliced2DCIterator

get(int) - Method in class neureka.ndim.iterator.types.sliced.Sliced3DCIterator

get() - Method in class neureka.ndim.iterator.types.sliced.Sliced3DCIterator

get(int) - Method in class neureka.ndim.iterator.types.sliced.SlicedNDIterator

get() - Method in class neureka.ndim.iterator.types.sliced.SlicedNDIterator

get(int) - Method in class neureka.ndim.iterator.types.virtual.VirtualNDIterator

get() - Method in class neureka.ndim.iterator.types.virtual.VirtualNDIterator

get() - Static method in class neureka.Neureka

The Neureka class represents the configuration of this library.

get(int) - Method in interface neureka.Shape

get(int...) - Method in interface neureka.Tensor

The following method enables access to specific scalar elements within the nd-array.

get(Object...) - Method in interface neureka.Tensor

The following method enables the creation of nd-array slices which access the same underlying data (possibly from a different view).

get(int) - Method in interface neureka.Tensor

This getter method creates and returns a slice of the original nd-array.

get(Number) - Method in interface neureka.Tensor

This getter method creates and returns a slice of the original nd-array.

get(Object) - Method in interface neureka.Tensor

This method enables nd-array slicing! It takes a key of various types and configures a slice nd-array which shares the same underlying data as the original nd-array.

getAbs() - Method in class neureka.math.Functions

getActivation() - Method in interface neureka.backend.main.implementations.fun.api.ScalarFun

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarAbsolute

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarCbrt

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarCosinus

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarExp

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarGaSU

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarGaTU

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarGaussian

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarGaussianFast

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarGeLU

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarIdentity

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarLog10

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarLogarithm

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarQuadratic

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarReLU

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarSeLU

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarSigmoid

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarSiLU

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarSinus

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarSoftplus

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarSoftsign

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarSqrt

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarTanh

getActivation() - Method in class neureka.backend.main.implementations.fun.ScalarTanhFast

getActiveThreadCount() - Method in class neureka.devices.host.CPU.JVMExecutor

Returns the approximate number of threads that are actively executing tasks.

getAdd() - Method in class neureka.math.Functions

getAddAssign() - Method in class neureka.math.Functions

getAdHocKernel(String) - Method in class neureka.devices.opencl.OpenCLDevice

getAgentSupplier() - Method in class neureka.backend.api.Result

getAlgorithm() - Method in class neureka.backend.api.ExecutionCall

An ExecutionCall will either already have a targeted Algorithm defined at instantiation or otherwise it will query the associated Operation for an Algorithm best suitable for the state of this ExecutionCall.

getAlgorithm(Class<T>) - Method in interface neureka.backend.api.Operation

Operation implementations embody a component system hosting unique Algorithm instances.

getAlgorithm(Class<T>) - Method in class neureka.backend.api.template.operations.AbstractOperation

Operation implementations embody a component system hosting unique Algorithm instances.

getAlgorithmFor(ExecutionCall<?>) - Method in interface neureka.backend.api.Operation

Alongside a component system made up of Algorithm instances, implementations of this interface also ought to express a routing mechanism which finds the best Algorithm for a given ExecutionCall instance.

getAlgorithmFor(ExecutionCall<?>) - Method in class neureka.backend.api.template.operations.AbstractOperation

getAlgorithmName() - Method in interface neureka.backend.api.ini.LoadingContext

getAll(Class<T>) - Method in class neureka.common.composition.AbstractComponentOwner

This method tries to find all components inside the internal component array whose classes are sub types of the one provided.

getAll(Class<T>) - Method in interface neureka.common.composition.ComponentOwner

Use this to get all components of the specified component type class.

getAllAlgorithms() - Method in interface neureka.backend.api.Operation

getAllAlgorithms() - Method in class neureka.backend.api.template.operations.AbstractOperation

getAllAliases() - Method in class neureka.framing.fluent.AxisFrame

getAllAliasesForIndex(int) - Method in class neureka.framing.fluent.AxisFrame

getAllFunctions() - Method in interface neureka.math.Function

getAndRemovePendingError() - Method in class neureka.autograd.GraphNode

This method is called by the JITProp component.

getArchitecture() - Static method in class neureka.devices.host.machine.ConcreteMachine

getArity() - Method in interface neureka.backend.api.Operation

Arity is the number of arguments or operands that this function or operation takes.

getArity() - Method in class neureka.backend.api.template.operations.AbstractOperation

getArity() - Method in class neureka.backend.api.template.operations.OperationBuilder

getAsInt() - Method in enum neureka.devices.host.concurrent.Parallelism

getAt(int...) - Method in interface neureka.Nda

The following method enables access to specific scalar elements within the nd-array.

getAt(Number) - Method in interface neureka.Nda

This getter method creates and returns a slice of the original nd-array.

getAt(Object...) - Method in interface neureka.Nda

The following method enables the creation of nd-array slices which access the same underlying data (possibly from a different view).

getAt(int) - Method in interface neureka.Nda

This getter method creates and returns a slice of the original nd-array.

getAt(Map<?, Integer>) - Method in interface neureka.Nda

This method is most useful when used in Groovy where defining maps is done through square brackets, making it possible to slice nd-arrays like so:

getAt(List<?>) - Method in interface neureka.Nda

This method enables nd-array slicing! It takes a key of various types and configures a slice nd-array which shares the same underlying data as the original nd-array.

getAt(int...) - Method in interface neureka.Tensor

The following method enables access to specific scalar elements within the nd-array.

getAt(Number) - Method in interface neureka.Tensor

This getter method creates and returns a slice of the original nd-array.

getAt(Object...) - Method in interface neureka.Tensor

The following method enables the creation of nd-array slices which access the same underlying data (possibly from a different view).

getAt(int) - Method in interface neureka.Tensor

This getter method creates and returns a slice of the original nd-array.

getAt(Map<?, Integer>) - Method in interface neureka.Tensor

This method is most useful when used in Groovy where defining maps is done through square brackets, making it possible to slice nd-arrays like so:

getAt(List<?>) - Method in interface neureka.Tensor

This method enables nd-array slicing! It takes a key of various types and configures a slice nd-array which shares the same underlying data as the original nd-array.

getAutogradFunction() - Method in class neureka.backend.api.BackendContext

This method returns a Functions instance which wraps pre-instantiated Function instances which are configured to track their computational history.

getBackend() - Method in class neureka.Neureka

getCbrt() - Method in class neureka.math.Functions

getCellSize() - Method in class neureka.view.NDPrintSettings

A cell size refers to the number of characters reserved to the String representation of a single element.

getChildren() - Method in class neureka.autograd.GraphNode

The children are GraphNode instances which represent computations involving the payload of this very GraphNode instance.

getChildren() - Method in class neureka.framing.Relation

getCode() - Method in class neureka.devices.opencl.KernelCode

getColLabels() - Method in class neureka.devices.file.CSVHandle

getCompletedTaskCount() - Method in class neureka.devices.host.CPU.JVMExecutor

Returns the approximate total number of tasks that have completed execution.

getConcat() - Method in class neureka.math.Functions

getContext() - Method in class neureka.devices.opencl.OpenCLPlatform

getConv() - Method in class neureka.math.Functions

getCoreCount() - Method in class neureka.devices.host.CPU

Returns the number of CPU cores available to the Java virtual machine.

getCorePoolSize() - Method in class neureka.devices.host.CPU.JVMExecutor

Returns the core number of threads.

getCos() - Method in class neureka.math.Functions

getData() - Method in class neureka.common.utility.ListReader.Result

getData() - Method in interface neureka.MutateNda

At the heart of every tensor is the Data object, which holds the actual data array, a sequence of values of the same type.

getDataAs(Class<A>) - Method in interface neureka.MutateNda

This method returns the data of this nd-array as a Java array of the specified type.

getDataAs(Class<A>) - Method in interface neureka.Nda

Use this to get the items of the underlying Data buffer of this nd-array as a primitive array of the specified type.

getDataAt(int) - Method in interface neureka.Nda

Use this to access elements of the underlying data array without any index transformation applied to it.

getDataForWriting(Class<A>) - Method in interface neureka.MutateTensor

Use this to access the underlying writable data of this tensor if you want to modify it.

getDataSize() - Method in interface neureka.devices.Device.Access

getDataSize() - Method in class neureka.devices.file.CSVHandle

getDataSize() - Method in interface neureka.devices.file.FileHandle

This method returns the byte size of the data which is stored in the tensor of the file which is managed by this FileHandle.

getDataSize() - Method in class neureka.devices.file.IDXHandle

getDataType() - Method in class neureka.devices.file.CSVHandle

getDataType() - Method in interface neureka.devices.file.FileHandle

getDataType() - Method in class neureka.devices.file.IDXHandle

getDataType() - Method in class neureka.devices.opencl.KernelCode

getDataType() - Method in interface neureka.Tensor

This method returns the DataType instance of this Tensor, which is a wrapper object for the actual type class representing the value items stored inside the underlying data array of this tensor.

getDefaultAlgorithm() - Method in class neureka.backend.api.template.operations.AbstractOperation

getDefaultDataType() - Method in class neureka.Neureka.Settings.DType

getDefaultDataTypeClass() - Method in class neureka.Neureka.Settings.DType

The default data type is not relevant most of the time.

getDelimiter() - Method in class neureka.devices.file.CSVHandle

getDerivative() - Method in interface neureka.backend.main.implementations.fun.api.ScalarFun

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarAbsolute

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarCbrt

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarCosinus

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarExp

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarGaSU

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarGaTU

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarGaussian

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarGaussianFast

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarGeLU

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarIdentity

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarLog10

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarLogarithm

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarQuadratic

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarReLU

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarSeLU

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarSigmoid

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarSiLU

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarSinus

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarSoftplus

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarSoftsign

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarSqrt

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarTanh

getDerivative() - Method in class neureka.backend.main.implementations.fun.ScalarTanhFast

getDerivative(int) - Method in interface neureka.math.Function

This method builds a new Function which is the derivative of this Function with respect to the provided input index.

getDerivative(int) - Method in class neureka.math.implementations.FunctionConstant

getDerivative(int) - Method in class neureka.math.implementations.FunctionInput

getDerivative(int) - Method in class neureka.math.implementations.FunctionNode

getDerivative(int) - Method in class neureka.math.implementations.FunctionVariable

getDerivativeIndex() - Method in class neureka.backend.api.Call

getDerivator() - Method in class neureka.backend.api.template.operations.OperationBuilder

getDevice() - Method in class neureka.backend.api.Call

getDevice() - Method in interface neureka.Tensor

getDeviceFor(Class<T>) - Method in class neureka.backend.api.Call

getDevices() - Method in class neureka.devices.opencl.OpenCLPlatform

getDimTrim() - Method in class neureka.math.Functions

getDirectory() - Method in class neureka.devices.file.FileDevice

getDiv() - Method in class neureka.math.Functions

getDivAssign() - Method in class neureka.math.Functions

getDot() - Method in class neureka.math.Functions

getEstimation() - Method in class neureka.backend.api.Call.Validator.Estimator

getEstimator() - Method in class neureka.backend.api.Call.Validator

getExecutor() - Method in class neureka.devices.host.CPU

The CPU.JVMExecutor offers a similar functionality as the parallel stream API, however it differs in that the CPU.JVMExecutor is processing CPU.RangeWorkload lambdas instead of simply exposing a single index or concrete elements for a given workload size.

getExp() - Method in class neureka.math.Functions

getExtensions() - Method in class neureka.backend.api.BackendContext

getFastGaus() - Method in class neureka.math.Functions

getFastTanh() - Method in class neureka.math.Functions

getFileName() - Method in interface neureka.devices.file.FileHandle

getFrame() - Method in interface neureka.Tensor

getFunction() - Method in class neureka.autograd.GraphNode

Recorded Function which produced this GraphNode.

getFunction() - Method in class neureka.backend.api.BackendContext

This method returns a Functions instance which wraps pre-instantiated Function instances which are configured to not track their computational history.

getFunctionCache() - Method in class neureka.backend.api.BackendContext

getGaus() - Method in class neureka.math.Functions

getGelu() - Method in class neureka.math.Functions

getGradient() - Method in interface neureka.Tensor

getGraphNode() - Method in interface neureka.Tensor

getHasDerivatives() - Method in class neureka.view.NDPrintSettings

getHasGradient() - Method in class neureka.view.NDPrintSettings

getHasRecursiveGraph() - Method in class neureka.view.NDPrintSettings

getHasShape() - Method in class neureka.view.NDPrintSettings

getHasSlimNumbers() - Method in class neureka.view.NDPrintSettings

getHasValue() - Method in class neureka.view.NDPrintSettings

getId() - Method in class neureka.devices.opencl.OpenCLDevice

getId() - Method in class neureka.devices.opencl.OpenCLPlatform

getIdentifier() - Method in interface neureka.backend.api.Operation

Concrete Operation types ought to be representable by a function name.

getIdentifier() - Method in class neureka.backend.api.template.operations.AbstractOperation

getIdentifier() - Method in class neureka.backend.api.template.operations.OperationBuilder

getIdy() - Method in class neureka.math.Functions

getImplementationFor(Class<D>) - Method in interface neureka.backend.api.DeviceAlgorithm

An ImplementationFor a specific Device can be accessed by passing the class of the Device for which an implementation should be returned.

getImplementationFor(D) - Method in interface neureka.backend.api.DeviceAlgorithm

An ImplementationFor a specific Device can be accessed by passing the Device for which an implementation should be returned.

getImplementationFor(Class<D>) - Method in class neureka.backend.api.template.algorithms.AbstractDeviceAlgorithm

getIndent() - Method in class neureka.view.NDPrintSettings

getIndexAndIncrement() - Method in interface neureka.ndim.iterator.NDIterator

getIndexAtAlias(Object) - Method in class neureka.framing.fluent.AxisFrame

getIndexToIndexAccessPattern() - Method in interface neureka.ndim.config.NDConfiguration

getInt(cl_device_id, int) - Static method in class neureka.devices.opencl.OpenCLDevice.Query

Returns the value of the device info parameter with the given name

getInts(cl_device_id, int, int) - Static method in class neureka.devices.opencl.OpenCLDevice.Query

Returns the values of the device info parameter with the given name

getIsAutoConvertingExternalDataToJVMTypes() - Method in class neureka.Neureka.Settings.DType

This flag will determine if foreign data types will be converted into the next best fit (in terms of bits) or if it should be converted into something that does not mess with the representation of the data.

getIsCellBound() - Method in class neureka.view.NDPrintSettings

getIsDifferentiable() - Method in class neureka.backend.api.template.operations.OperationBuilder

getIsIndexer() - Method in class neureka.backend.api.template.operations.OperationBuilder

getIsInline() - Method in class neureka.backend.api.template.operations.OperationBuilder

getIsLegacy() - Method in class neureka.view.NDPrintSettings

This flag determines the usage of bracket types, where "[1x3]:(1, 2, 3)" would be the legacy version of "(1x3):[1, 2, 3]".

getIsMultiline() - Method in class neureka.view.NDPrintSettings

getIsOperator() - Method in class neureka.backend.api.template.operations.OperationBuilder

getIsScientific() - Method in class neureka.view.NDPrintSettings

getItem() - Method in interface neureka.Nda

Equivalent to the #item(0) and Nda.item().

getItems() - Method in interface neureka.Nda

A more verbose version of the Nda.items() method (best used by JVM languages with property support).

getItemsAs(Class<A>) - Method in interface neureka.Nda

Use this to get the items of this nd-array as a primitive array of the specified type.

getItemType() - Method in interface neureka.Nda

getItemTypeClass() - Method in class neureka.dtype.DataType

getKernel(ExecutionCall<OpenCLDevice>) - Method in class neureka.devices.opencl.OpenCLDevice

getKernel(String) - Method in class neureka.devices.opencl.OpenCLDevice

getKernel(String) - Method in class neureka.devices.opencl.OpenCLPlatform

getKernelCode() - Method in class neureka.backend.main.implementations.ParsedCLImplementation

getKernelCode() - Method in class neureka.backend.main.implementations.SimpleCLImplementation

getKernelCode() - Method in interface neureka.devices.opencl.StaticKernelSource

getKernelFor(ExecutionCall<OpenCLDevice>) - Method in class neureka.backend.main.implementations.ParsedCLImplementation

getKernelFor(ExecutionCall<OpenCLDevice>) - Method in class neureka.backend.main.implementations.SimpleCLImplementation

getKernelFor(ExecutionCall<OpenCLDevice>) - Method in interface neureka.devices.opencl.KernelSource

getLabel() - Method in class neureka.framing.NDFrame

getLabel() - Method in interface neureka.Nda

A nd-array can have a label.

getLayout() - Method in interface neureka.ndim.config.NDConfiguration

The layout of most tensors is either row major or column major.

getLearningRate() - Method in class neureka.optimization.implementations.ADAM

getLn() - Method in class neureka.math.Functions

getLoadable() - Method in class neureka.devices.file.FileDevice

getLoaded() - Method in class neureka.devices.file.FileDevice

getLoader() - Method in interface neureka.backend.api.BackendExtension

getLoader() - Method in class neureka.backend.cpu.CPUBackend

getLoader() - Method in class neureka.backend.ocl.CLBackend

getLocation() - Method in interface neureka.devices.file.FileHandle

getLog10() - Method in class neureka.math.Functions

getLong(cl_device_id, int) - Static method in class neureka.devices.opencl.OpenCLDevice.Query

Returns the value of the device info parameter with the given name

getLongs(cl_device_id, int, int) - Static method in class neureka.devices.opencl.OpenCLDevice.Query

Returns the values of the device info parameter with the given name

getLongs(int, ByteBuffer, long[]) - Static method in class neureka.devices.opencl.OpenCLDevice.Query

getMatMul() - Method in class neureka.math.Functions

getMax() - Method in class neureka.math.Functions

getMemory() - Static method in class neureka.devices.host.machine.ConcreteMachine

getMin() - Method in class neureka.math.Functions

getMinus() - Method in class neureka.math.Functions

getMinusAssign() - Method in class neureka.math.Functions

getMod() - Method in class neureka.math.Functions

getModAssign() - Method in class neureka.math.Functions

getMode() - Method in class neureka.autograd.GraphNode

This is the getter for an important GraphNode property which holds the auto-differentiation mode used by this instance to decide if a given error should be forward propagated backward propagated or not propagated at all.

getMomentum() - Method in class neureka.optimization.implementations.ADAM

getMul() - Method in class neureka.math.Functions

getMulAssign() - Method in class neureka.math.Functions

getMut() - Method in interface neureka.Nda

This method exposes an API for mutating the state of this tensor.

getMut() - Method in interface neureka.Tensor

This method exposes an API for mutating the state of this tensor.

getName() - Method in interface neureka.backend.api.Algorithm

The name of an Algorithm may be used for OpenCL kernel compilation or simply for debugging purposes to identify which type of algorithm is being executed at any given time...

getName() - Method in class neureka.devices.opencl.KernelCode

getNDConf() - Method in interface neureka.ndim.NDimensional

getNDPrintSettings() - Method in class neureka.Neureka.Settings.View

Settings for configuring how tensors should be converted to String representations.

getNeg() - Method in class neureka.math.Functions

getNewInstance() - Static method in interface neureka.devices.DeviceCleaner

getNewOwner() - Method in interface neureka.common.composition.Component.OwnerChangeRequest

getNumberOfColumns() - Method in class neureka.devices.file.CSVHandle

getNumberOfRows() - Method in class neureka.devices.file.CSVHandle

getNumericTypeTarget() - Method in interface neureka.dtype.NumericType

This method returns the NumericType representation of the target type of this class.

getOldOwner() - Method in interface neureka.common.composition.Component.OwnerChangeRequest

getOperation(int) - Method in class neureka.backend.api.BackendContext

This method queries the operations in this BackendContext by a provided index integer targeting an entry in the list of Operation implementation instances sitting in this execution context.

getOperation(String) - Method in class neureka.backend.api.BackendContext

This method queries the operations in this BackendContext by a provided identifier which has to match the name of an existing operation.

getOperation() - Method in class neureka.backend.api.ExecutionCall

This returns the operation which will ultimately process this execution call.

getOperation() - Method in interface neureka.math.Function

getOperation() - Method in class neureka.math.implementations.FunctionConstant

getOperation() - Method in class neureka.math.implementations.FunctionInput

getOperation() - Method in class neureka.math.implementations.FunctionNode

getOperation() - Method in class neureka.math.implementations.FunctionVariable

getOperationIdentidier() - Method in interface neureka.backend.api.ini.LoadingContext

getOperationLookupMap() - Method in class neureka.backend.api.BackendContext

This method returns an unmodifiable view of the mapping between the Operation.getIdentifier() / Operation.getOperator() properties and the Operation implementation instances to which they belong.

getOperations() - Method in class neureka.backend.api.BackendContext

This method returns an unmodifiable view of the list of Operation implementation instances managed by this context.

getOperator() - Method in interface neureka.backend.api.Operation

getOperator() - Method in class neureka.backend.api.template.operations.AbstractOperation

getOperator() - Method in class neureka.backend.api.template.operations.OperationBuilder

getOrNull() - Method in interface neureka.Data

This returns the underlying raw data object of a nd-array or tensor of a backend specific type (e.g.

getOrNull() - Method in class neureka.devices.AbstractDeviceData

getParent() - Method in class neureka.framing.Relation

getParents() - Method in class neureka.autograd.GraphNode

getPayload() - Method in class neureka.autograd.GraphNode

The value of a graph node is the tensor to which it belongs (is a component of).

getPayloadDataType() - Method in class neureka.autograd.GraphNode

getPayloadReferenceVersion() - Method in class neureka.autograd.GraphNode

This variable holds a copy of the version of the payload tensor recorded when this GraphNode instance is instantiated.

getPayloadShape() - Method in class neureka.autograd.GraphNode

Note: This method will never return null even if the actual payload tensor was garbage collected.

getPendingError() - Method in class neureka.autograd.GraphNode

Used by the Just-In-Time back-prop component.

getPermute() - Method in class neureka.math.Functions

getPermuteRelationFor(Tensor<V>) - Method in class neureka.framing.Relation

When creating permuted versions of slices then there must be a translation between the shape configuration between this new slice and the original parent tensor from which both slices have been derived.

getPlatform() - Method in class neureka.devices.opencl.OpenCLDevice

getPlatforms() - Method in class neureka.backend.ocl.CLBackend

getPlus() - Method in class neureka.math.Functions

getPlusAssign() - Method in class neureka.math.Functions

getPostfix() - Method in class neureka.view.NDPrintSettings

getPow() - Method in class neureka.math.Functions

getPowAssign() - Method in class neureka.math.Functions

getPrefix() - Method in class neureka.view.NDPrintSettings

getQuad() - Method in class neureka.math.Functions

getRandom() - Method in class neureka.math.Functions

getRank() - Method in interface neureka.ndim.NDimensional

getRawData() - Method in interface neureka.Nda

This returns an unprocessed version of the underlying data of this nd-array.

getRawItems() - Method in interface neureka.Nda

getRelayout() - Method in class neureka.math.Functions

getRelu() - Method in class neureka.math.Functions

getRepresentativeItemClass() - Method in interface neureka.Tensor

The Class returned by this method is the representative Class of the value items of a concrete AbstractNda but not necessarily the actual Class of a given value item, this is especially true for numeric types, which are represented by implementations of the NumericType interface.

getRepresentativeType() - Method in class neureka.dtype.DataType

getReshape() - Method in class neureka.math.Functions

getResult() - Method in class neureka.backend.main.memory.MemValidator

getRowLabels() - Method in class neureka.devices.file.CSVHandle

getRowLimit() - Method in class neureka.view.NDPrintSettings

Very large tensors with a rank larger than 1 might take a lot of vertical space when converted to a String.

getSelu() - Method in class neureka.math.Functions

The Scaled Exponential Linear Unit, or SELU, is an activation functions that induce self-normalizing properties.

getSettings() - Method in class neureka.backend.ocl.CLBackend

getShape() - Method in class neureka.common.utility.ListReader.Result

getShape() - Method in class neureka.devices.file.CSVHandle

getShape() - Method in interface neureka.devices.file.FileHandle

getShape() - Method in class neureka.devices.file.IDXHandle

getShape() - Method in interface neureka.ndim.NDConstructor

getShape() - Method in interface neureka.ndim.NDimensional

getSigmoid() - Method in class neureka.math.Functions

getSilu() - Method in class neureka.math.Functions

The SiLu activation function, also known as the swish function, is defined as x * sigmoid(x).

getSin() - Method in class neureka.math.Functions

getSize(cl_device_id, int) - Static method in class neureka.devices.opencl.OpenCLDevice.Query

Returns the value of the device info parameter with the given name

getSize() - Method in interface neureka.ndim.NDConstructor

getSize() - Method in interface neureka.ndim.NDimensional

getSizes(cl_device_id, int, int) - Static method in class neureka.devices.opencl.OpenCLDevice.Query

Returns the values of the device info parameter with the given name

getSoftplus() - Method in class neureka.math.Functions

SoftPlus is a smooth approximation to the ReLU function and can be used to constrain the output of a machine to always be positive.

getSoftsign() - Method in class neureka.math.Functions

The softsign function, defined as x / ( 1 + Math.abs( x ) ), is a computationally cheap 0 centered activation function which rescales the inputs between -1 and 1, very much like the Tanh function.

getSqrt() - Method in class neureka.math.Functions

getState() - Method in class neureka.framing.NDFrame

getString(cl_device_id, int) - Static method in class neureka.devices.opencl.OpenCLDevice.Query

Returns the value of the device info parameter with the given name

getString(cl_platform_id, int) - Static method in class neureka.devices.opencl.OpenCLDevice.Query

Returns the value of the platform info parameter with the given name

getStringifier() - Method in class neureka.backend.api.template.operations.OperationBuilder

getSubFunctions() - Method in interface neureka.math.Function

getSubFunctions() - Method in class neureka.math.implementations.FunctionConstant

getSubFunctions() - Method in class neureka.math.implementations.FunctionInput

getSubFunctions() - Method in class neureka.math.implementations.FunctionNode

getSubFunctions() - Method in class neureka.math.implementations.FunctionVariable

getSum() - Method in class neureka.math.Functions

getT() - Method in interface neureka.Tensor

A method which returns a new Tensor instance which is a transposed twin of this instance.
This is an alternative to the functionally identical Tensor.T() method.

getTanh() - Method in class neureka.math.Functions

getter(At<Object, Get<GetType>>) - Method in class neureka.framing.fluent.AxisFrame.Builder

getThreads() - Static method in class neureka.devices.host.machine.ConcreteMachine

getTime() - Method in class neureka.optimization.implementations.ADAM

getTotalNumberOfDevices() - Method in class neureka.backend.ocl.CLBackend

getTotalSize() - Method in class neureka.devices.file.CSVHandle

getTotalSize() - Method in interface neureka.devices.file.FileHandle

This method returns the number of bytes which are used to store the tensor in the file whose access is being managed by an implementation of th FileHandle interface.

getTotalSize() - Method in class neureka.devices.file.IDXHandle

getTraits() - Method in interface neureka.ndim.config.NDConfiguration

getTranspose2D() - Method in class neureka.math.Functions

getType() - Method in class neureka.common.utility.ListReader.Result

getTypeClassInstance(Class<T>) - Method in class neureka.dtype.DataType

getValOf(Class<T>) - Method in class neureka.backend.api.Call

getValue() - Method in class neureka.common.utility.Cache.LazyEntry

getValueSize() - Method in class neureka.devices.file.CSVHandle

getValueSize() - Method in interface neureka.devices.file.FileHandle

This method return the size of the value which is stored in the tensor of the file which is managed by this FileHandle.

getValueSize() - Method in class neureka.devices.file.IDXHandle

getVelocity() - Method in class neureka.optimization.implementations.ADAM

getVersion() - Method in interface neureka.Tensor

The version number is tracking how often this tensor has been mutated.

globalMemSize() - Method in class neureka.devices.opencl.OpenCLDevice

GOOD - Static variable in interface neureka.backend.api.fun.SuitabilityPredicate

goodIfAll(Call.TensorCondition) - Method in class neureka.backend.api.Call.Validator.Estimator

goodIfAll(Call.TensorCompare) - Method in class neureka.backend.api.Call.Validator.Estimator

goodIfAny(Call.TensorCondition) - Method in class neureka.backend.api.Call.Validator.Estimator

goodIfAnyNonNull(Call.TensorCondition) - Method in class neureka.backend.api.Call.Validator.Estimator

gradient() - Method in interface neureka.Tensor

This is a functionally identical alternative to the Tensor.getGradient() method.

gradientApplyRequested() - Method in interface neureka.Tensor

This flag works alongside two autograd features which can be enabled inside the library settings.

GraphNode<V> - Class in neureka.autograd

Instances of the GraphNode class are components of tensors (Tensor instances) which model and record computations / operations between them.

GraphNode(Function, ExecutionCall<Device<?>>, Supplier<Result>) - Constructor for class neureka.autograd.GraphNode

graphNode() - Method in interface neureka.Tensor

This is a functionally identical alternative to Tensor.getGraphNode().

GraphNode.Print - Enum in neureka.autograd

groupBy(String, String, String, String) - Static method in class neureka.math.parsing.ParseUtil

H

Hardware - Class in neureka.devices.host.machine

This models the cache levels and threads of a CPU using an array of where each entry represents a memory level.

Hardware(String, BasicMachine[]) - Constructor for class neureka.devices.host.machine.Hardware

new BasicMachine[] { SYSTEM, L3, L2, L1 } or new BasicMachine[] { SYSTEM, L2, L1 } or in worst case new BasicMachine[] { SYSTEM, L1 }

has(GraphNode<V>) - Method in class neureka.autograd.GraphNode

This method checks if a given graph node is an AD target of this node.

has(Class<E>) - Method in class neureka.backend.api.BackendContext

Checks if this context has an instance of the provided BackendExtension type.

has(Class<T>) - Method in class neureka.common.composition.AbstractComponentOwner

This method checks if a component identified by the passed Class instance is present inside the stored component collection.

has(Class<T>) - Method in interface neureka.common.composition.ComponentOwner

Use this to check if a component of the specified component type class is present.

has(O) - Method in class neureka.common.utility.Cache

has(Tensor<T>) - Method in class neureka.devices.AbstractBaseDevice

This method checks if the passed tensor is stored on this Device instance.

has(Tensor<T>) - Method in interface neureka.devices.Device

Use this to check if a tensor is stored on this Device!

has(Tensor<T>) - Method in class neureka.devices.file.FileDevice

has(String) - Method in class neureka.devices.opencl.KernelCache

has(cl_device_id) - Method in class neureka.devices.opencl.OpenCLPlatform

has(String, boolean) - Method in class neureka.math.FunctionCache

has(NDTrait) - Method in interface neureka.ndim.config.NDConfiguration

hasAdHocKernel(String) - Method in class neureka.devices.opencl.OpenCLDevice

hasChildren() - Method in class neureka.framing.Relation

hasDerivatives() - Method in class neureka.autograd.GraphNode

hasGradient() - Method in interface neureka.Tensor

Tensors can be components of other tensors which makes the implicitly their gradients.

hashCode() - Method in class neureka.common.utility.Cache.LazyEntry

hashCode() - Method in class neureka.devices.host.machine.BasicMachine

hashCode() - Method in class neureka.devices.host.machine.CommonMachine

hashCode() - Method in class neureka.devices.host.machine.ConcreteMachine

hashCode() - Method in class neureka.devices.host.machine.Hardware

hashCode() - Method in class neureka.devices.opencl.KernelCode

hashCode() - Method in class neureka.dtype.DataType

hashCode() - Method in class neureka.ndim.config.AbstractNDC

hashCode() - Method in interface neureka.ndim.config.NDConfiguration

hasImplementationFor(D) - Method in interface neureka.backend.api.DeviceAlgorithm

hasKernel(String) - Method in class neureka.devices.opencl.OpenCLPlatform

hasLabelsForAxis(Object) - Method in class neureka.framing.NDFrame

hasOperation(Operation) - Method in class neureka.backend.api.BackendContext

hasOperation(String) - Method in class neureka.backend.api.BackendContext

hasParent() - Method in class neureka.framing.Relation

holderArrayType() - Method in class neureka.dtype.custom.F32

holderArrayType() - Method in class neureka.dtype.custom.F64

holderArrayType() - Method in class neureka.dtype.custom.I16

holderArrayType() - Method in class neureka.dtype.custom.I32

holderArrayType() - Method in class neureka.dtype.custom.I64

holderArrayType() - Method in class neureka.dtype.custom.I8

holderArrayType() - Method in class neureka.dtype.custom.UI16

holderArrayType() - Method in class neureka.dtype.custom.UI32

holderArrayType() - Method in class neureka.dtype.custom.UI64

holderArrayType() - Method in class neureka.dtype.custom.UI8

holderArrayType() - Method in interface neureka.dtype.NumericType

The holder array type is the JVM type which can hold the data but not necessarily represent it (int[] cant represent uint[]).

holderType() - Method in class neureka.dtype.custom.F32

holderType() - Method in class neureka.dtype.custom.F64

holderType() - Method in class neureka.dtype.custom.I16

holderType() - Method in class neureka.dtype.custom.I32

holderType() - Method in class neureka.dtype.custom.I64

holderType() - Method in class neureka.dtype.custom.I8

holderType() - Method in class neureka.dtype.custom.UI16

holderType() - Method in class neureka.dtype.custom.UI32

holderType() - Method in class neureka.dtype.custom.UI64

holderType() - Method in class neureka.dtype.custom.UI8

holderType() - Method in interface neureka.dtype.NumericType

The holder type is the JVM type which can hold the data but not necessarily represent it (int cant represent uint).

HOW_TO_INSTALL_OPENCL - Variable in enum neureka.devices.opencl.utility.Messages.Tips

HOW_TO_INSTALL_OPENCL_DRIVERS - Variable in enum neureka.devices.opencl.utility.Messages.Tips

I

i() - Method in interface neureka.ndim.iterator.NDIterator

i() - Method in class neureka.ndim.iterator.types.permuted.Permuted2DCIterator

i() - Method in class neureka.ndim.iterator.types.permuted.Permuted3DCIterator

i() - Method in class neureka.ndim.iterator.types.simple.Simple1DCIterator

i() - Method in class neureka.ndim.iterator.types.simple.Simple2DCIterator

i() - Method in class neureka.ndim.iterator.types.simple.Simple3DCIterator

i() - Method in class neureka.ndim.iterator.types.sliced.Sliced1DCIterator

i() - Method in class neureka.ndim.iterator.types.sliced.Sliced2DCIterator

i() - Method in class neureka.ndim.iterator.types.sliced.Sliced3DCIterator

i() - Method in class neureka.ndim.iterator.types.sliced.SlicedNDIterator

i() - Method in class neureka.ndim.iterator.types.virtual.VirtualNDIterator

I16 - Class in neureka.dtype.custom

I16() - Constructor for class neureka.dtype.custom.I16

I32 - Class in neureka.dtype.custom

I32() - Constructor for class neureka.dtype.custom.I32

I64 - Class in neureka.dtype.custom

I64() - Constructor for class neureka.dtype.custom.I64

I8 - Class in neureka.dtype.custom

The following abstract class implements some basic logic which is applicable across all final concrete classes extending this abstract one.

I8() - Constructor for class neureka.dtype.custom.I8

IAXPY - Class in neureka.backend.main.operations.linear.internal.blas

The ?axpy routines perform a vector-vector operation defined as y := a*x + y where: a is a scalar x and y are vectors each with a number of elements that equals n.

IAXPY() - Constructor for class neureka.backend.main.operations.linear.internal.blas.IAXPY

id() - Method in interface neureka.backend.main.implementations.fun.api.ScalarFun

id() - Method in class neureka.backend.main.implementations.fun.ScalarAbsolute

id() - Method in class neureka.backend.main.implementations.fun.ScalarCbrt

id() - Method in class neureka.backend.main.implementations.fun.ScalarCosinus

id() - Method in class neureka.backend.main.implementations.fun.ScalarExp

id() - Method in class neureka.backend.main.implementations.fun.ScalarGaSU

id() - Method in class neureka.backend.main.implementations.fun.ScalarGaTU

id() - Method in class neureka.backend.main.implementations.fun.ScalarGaussian

id() - Method in class neureka.backend.main.implementations.fun.ScalarGaussianFast

id() - Method in class neureka.backend.main.implementations.fun.ScalarGeLU

id() - Method in class neureka.backend.main.implementations.fun.ScalarIdentity

id() - Method in class neureka.backend.main.implementations.fun.ScalarLog10

id() - Method in class neureka.backend.main.implementations.fun.ScalarLogarithm

id() - Method in class neureka.backend.main.implementations.fun.ScalarQuadratic

id() - Method in class neureka.backend.main.implementations.fun.ScalarReLU

id() - Method in class neureka.backend.main.implementations.fun.ScalarSeLU

id() - Method in class neureka.backend.main.implementations.fun.ScalarSigmoid

id() - Method in class neureka.backend.main.implementations.fun.ScalarSiLU

id() - Method in class neureka.backend.main.implementations.fun.ScalarSinus

id() - Method in class neureka.backend.main.implementations.fun.ScalarSoftplus

id() - Method in class neureka.backend.main.implementations.fun.ScalarSoftsign

id() - Method in class neureka.backend.main.implementations.fun.ScalarSqrt

id() - Method in class neureka.backend.main.implementations.fun.ScalarTanh

id() - Method in class neureka.backend.main.implementations.fun.ScalarTanhFast

identifier(String) - Method in class neureka.backend.api.template.operations.OperationBuilder

IDENTITY - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

Identity - Class in neureka.backend.main.operations.functions

Identity() - Constructor for class neureka.backend.main.operations.functions.Identity

IDOT - Class in neureka.backend.main.operations.linear.internal.blas

The ?dot routines perform a vector-vector reduction operation defined as Equation where xi and yi are elements of vectors x and y.

IDOT() - Constructor for class neureka.backend.main.operations.linear.internal.blas.IDOT

IDXHandle - Class in neureka.devices.file

This class is one of many extensions of the AbstractFileHandle which is therefore ultimately an implementation of the FileHandle interface.

IDXHandle(String) - Constructor for class neureka.devices.file.IDXHandle

IDXHandle(Tensor<Number>, String) - Constructor for class neureka.devices.file.IDXHandle

idy() - Method in class neureka.math.Functions

ifValid(T) - Method in class neureka.backend.api.Call.Validator

IGEMM - Class in neureka.backend.main.operations.linear.internal.blas

A collection of primitive sub-routines for matrix multiplication performed on continuous arrays which are designed so that they can be vectorized by the JVMs JIT compiler (AVX instructions).

IGEMM() - Constructor for class neureka.backend.main.operations.linear.internal.blas.IGEMM

IGEMM.VectorOperationI32 - Interface in neureka.backend.main.operations.linear.internal.blas

IGEMM.VectorOperationI64 - Interface in neureka.backend.main.operations.linear.internal.blas

image2DMaxHeight() - Method in class neureka.devices.opencl.OpenCLDevice

image2DMaxWidth() - Method in class neureka.devices.opencl.OpenCLDevice

image3DMaxDepth() - Method in class neureka.devices.opencl.OpenCLDevice

image3DMaxHeight() - Method in class neureka.devices.opencl.OpenCLDevice

image3DMaxWidth() - Method in class neureka.devices.opencl.OpenCLDevice

imageSupport() - Method in class neureka.devices.opencl.OpenCLDevice

ImplementationFor<D extends Device<?>> - Interface in neureka.backend.api

Generally speaking, this interface describes the functionality of an implementation of an execution procedure tailored to a specific Device (interface) instance and Algorithm (interface) instance! Instances of implementations of the ImplementationFor interface are components of instances of implementations of the Algorithm interface, which themselves are components of Operation implementation instances.

ImplementationReceiver - Interface in neureka.backend.api.ini

in(Supplier<R>) - Method in interface neureka.devices.Device.In

increment() - Method in class neureka.devices.ReferenceCounter

increment(int[], int[]) - Static method in class neureka.ndim.config.NDConfiguration.Utility

increment() - Method in interface neureka.ndim.iterator.NDIterator

increment() - Method in class neureka.ndim.iterator.types.permuted.Permuted2DCIterator

increment() - Method in class neureka.ndim.iterator.types.permuted.Permuted3DCIterator

increment() - Method in class neureka.ndim.iterator.types.simple.Simple1DCIterator

increment() - Method in class neureka.ndim.iterator.types.simple.Simple2DCIterator

increment() - Method in class neureka.ndim.iterator.types.simple.Simple3DCIterator

increment() - Method in class neureka.ndim.iterator.types.sliced.Sliced1DCIterator

increment() - Method in class neureka.ndim.iterator.types.sliced.Sliced2DCIterator

increment() - Method in class neureka.ndim.iterator.types.sliced.Sliced3DCIterator

increment() - Method in class neureka.ndim.iterator.types.sliced.SlicedNDIterator

increment() - Method in class neureka.ndim.iterator.types.virtual.VirtualNDIterator

incrementUsageCount() - Method in class neureka.devices.AbstractDeviceData

incrementUsageCount() - Method in interface neureka.devices.DeviceData

incrementVersion(ExecutionCall<?>) - Method in interface neureka.MutateTensor

This method is responsible for incrementing the "_version" field variable which represents the version of the data of this tensor.

indent(int) - Static method in class neureka.view.NdaAsString.Util

index() - Method in class neureka.math.implementations.FunctionInput

indexOfIndex(int) - Method in interface neureka.ndim.config.NDConfiguration

Use this to calculate the true index for an element in the data array (data array index) based on a provided "virtual index", or "value array index".

indexOfIndex(int) - Method in class neureka.ndim.config.types.permuted.Permuted1DConfiguration

Use this to calculate the true index for an element in the data array (data array index) based on a provided "virtual index", or "value array index".

indexOfIndex(int) - Method in class neureka.ndim.config.types.permuted.Permuted2DConfiguration

Use this to calculate the true index for an element in the data array (data array index) based on a provided "virtual index", or "value array index".

indexOfIndex(int) - Method in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

Use this to calculate the true index for an element in the data array (data array index) based on a provided "virtual index", or "value array index".

indexOfIndex(int) - Method in class neureka.ndim.config.types.permuted.PermutedNDConfiguration

Use this to calculate the true index for an element in the data array (data array index) based on a provided "virtual index", or "value array index".

indexOfIndex(int) - Method in class neureka.ndim.config.types.simple.Simple0DConfiguration

Use this to calculate the true index for an element in the data array (data array index) based on a provided "virtual index", or "value array index".

indexOfIndex(int) - Method in class neureka.ndim.config.types.simple.Simple1DConfiguration

Use this to calculate the true index for an element in the data array (data array index) based on a provided "virtual index", or "value array index".

indexOfIndex(int) - Method in class neureka.ndim.config.types.simple.Simple2DConfiguration

Use this to calculate the true index for an element in the data array (data array index) based on a provided "virtual index", or "value array index".

indexOfIndex(int) - Method in class neureka.ndim.config.types.simple.Simple3DConfiguration

Use this to calculate the true index for an element in the data array (data array index) based on a provided "virtual index", or "value array index".

indexOfIndex(int) - Method in class neureka.ndim.config.types.simple.SimpleNDConfiguration

Use this to calculate the true index for an element in the data array (data array index) based on a provided "virtual index", or "value array index".

indexOfIndex(int) - Method in class neureka.ndim.config.types.sliced.Sliced0DConfiguration

Use this to calculate the true index for an element in the data array (data array index) based on a provided "virtual index", or "value array index".

indexOfIndex(int) - Method in class neureka.ndim.config.types.sliced.Sliced1DConfiguration

Use this to calculate the true index for an element in the data array (data array index) based on a provided "virtual index", or "value array index".

indexOfIndex(int) - Method in class neureka.ndim.config.types.sliced.Sliced2DConfiguration

Use this to calculate the true index for an element in the data array (data array index) based on a provided "virtual index", or "value array index".

indexOfIndex(int) - Method in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

Use this to calculate the true index for an element in the data array (data array index) based on a provided "virtual index", or "value array index".

indexOfIndex(int) - Method in class neureka.ndim.config.types.sliced.SlicedNDConfiguration

Use this to calculate the true index for an element in the data array (data array index) based on a provided "virtual index", or "value array index".

indexOfIndex(int) - Method in class neureka.ndim.config.types.views.SimpleReshapeView

Use this to calculate the true index for an element in the data array (data array index) based on a provided "virtual index", or "value array index".

indexOfIndex(int) - Method in class neureka.ndim.config.types.views.virtual.VirtualNDConfiguration

Use this to calculate the true index for an element in the data array (data array index) based on a provided "virtual index", or "value array index".

indexOfIndex(int) - Method in interface neureka.ndim.NDimensional

This is a convenience method identical to ndArray.getNDConf().indexOfIndex(i).

indexOfIndices(int[]) - Method in interface neureka.ndim.config.NDConfiguration

The following method calculates the true index for an element in the data array based on a provided index array.

indexOfIndices(int) - Method in class neureka.ndim.config.types.D1C

indexOfIndices(int, int) - Method in class neureka.ndim.config.types.D2C

indexOfIndices(int, int, int) - Method in class neureka.ndim.config.types.D3C

indexOfIndices(int[]) - Method in class neureka.ndim.config.types.permuted.Permuted1DConfiguration

The following method calculates the true index for an element in the data array based on a provided index array.

indexOfIndices(int) - Method in class neureka.ndim.config.types.permuted.Permuted1DConfiguration

indexOfIndices(int[]) - Method in class neureka.ndim.config.types.permuted.Permuted2DConfiguration

The following method calculates the true index for an element in the data array based on a provided index array.

indexOfIndices(int, int) - Method in class neureka.ndim.config.types.permuted.Permuted2DConfiguration

indexOfIndices(int[]) - Method in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

The following method calculates the true index for an element in the data array based on a provided index array.

indexOfIndices(int, int, int) - Method in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

indexOfIndices(int[]) - Method in class neureka.ndim.config.types.permuted.PermutedNDConfiguration

The following method calculates the true index for an element in the data array based on a provided index array.

indexOfIndices(int[]) - Method in class neureka.ndim.config.types.simple.Simple0DConfiguration

The following method calculates the true index for an element in the data array based on a provided index array.

indexOfIndices(int[]) - Method in class neureka.ndim.config.types.simple.Simple1DConfiguration

The following method calculates the true index for an element in the data array based on a provided index array.

indexOfIndices(int) - Method in class neureka.ndim.config.types.simple.Simple1DConfiguration

indexOfIndices(int[]) - Method in class neureka.ndim.config.types.simple.Simple2DConfiguration

The following method calculates the true index for an element in the data array based on a provided index array.

indexOfIndices(int, int) - Method in class neureka.ndim.config.types.simple.Simple2DConfiguration

indexOfIndices(int[]) - Method in class neureka.ndim.config.types.simple.Simple3DConfiguration

The following method calculates the true index for an element in the data array based on a provided index array.

indexOfIndices(int, int, int) - Method in class neureka.ndim.config.types.simple.Simple3DConfiguration

indexOfIndices(int[]) - Method in class neureka.ndim.config.types.simple.SimpleNDConfiguration

The following method calculates the true index for an element in the data array based on a provided index array.

indexOfIndices(int[]) - Method in class neureka.ndim.config.types.sliced.Sliced0DConfiguration

The following method calculates the true index for an element in the data array based on a provided index array.

indexOfIndices(int[]) - Method in class neureka.ndim.config.types.sliced.Sliced1DConfiguration

The following method calculates the true index for an element in the data array based on a provided index array.

indexOfIndices(int) - Method in class neureka.ndim.config.types.sliced.Sliced1DConfiguration

indexOfIndices(int[]) - Method in class neureka.ndim.config.types.sliced.Sliced2DConfiguration

The following method calculates the true index for an element in the data array based on a provided index array.

indexOfIndices(int, int) - Method in class neureka.ndim.config.types.sliced.Sliced2DConfiguration

indexOfIndices(int[]) - Method in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

The following method calculates the true index for an element in the data array based on a provided index array.

indexOfIndices(int, int, int) - Method in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

indexOfIndices(int[]) - Method in class neureka.ndim.config.types.sliced.SlicedNDConfiguration

The following method calculates the true index for an element in the data array based on a provided index array.

indexOfIndices(int[]) - Method in class neureka.ndim.config.types.views.SimpleReshapeView

The following method calculates the true index for an element in the data array based on a provided index array.

indexOfIndices(int[]) - Method in class neureka.ndim.config.types.views.virtual.VirtualNDConfiguration

The following method calculates the true index for an element in the data array based on a provided index array.

indexOfIndices(int[]) - Method in interface neureka.ndim.NDimensional

This is a convenience method identical to ndArray.getNDConf().indexOfIndices(indices).

INDICES_MAPPER_ID - Static variable in class neureka.backend.main.operations.linear.internal.opencl.CLReduce

indicesMap() - Method in interface neureka.ndim.config.NDConfiguration

If one wants to for example access the fourth last item of all items within a tensor based on a scalar index x then the NDConfiguration.indicesMap() is needed as a basis for translating said scalar index x to an array of indices for every axis of the tensor represented by this NDConfiguration.

indicesMap(int) - Method in interface neureka.ndim.config.NDConfiguration

This method receives an axis index and return the indices mapping value of said axis to enable readable access to the indices map of this configuration.

indicesMap() - Method in class neureka.ndim.config.types.permuted.Permuted1DConfiguration

If one wants to for example access the fourth last item of all items within a tensor based on a scalar index x then the NDConfiguration.indicesMap() is needed as a basis for translating said scalar index x to an array of indices for every axis of the tensor represented by this NDConfiguration.

indicesMap(int) - Method in class neureka.ndim.config.types.permuted.Permuted1DConfiguration

This method receives an axis index and return the indices mapping value of said axis to enable readable access to the indices map of this configuration.

indicesMap() - Method in class neureka.ndim.config.types.permuted.Permuted2DConfiguration

If one wants to for example access the fourth last item of all items within a tensor based on a scalar index x then the NDConfiguration.indicesMap() is needed as a basis for translating said scalar index x to an array of indices for every axis of the tensor represented by this NDConfiguration.

indicesMap(int) - Method in class neureka.ndim.config.types.permuted.Permuted2DConfiguration

This method receives an axis index and return the indices mapping value of said axis to enable readable access to the indices map of this configuration.

indicesMap() - Method in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

If one wants to for example access the fourth last item of all items within a tensor based on a scalar index x then the NDConfiguration.indicesMap() is needed as a basis for translating said scalar index x to an array of indices for every axis of the tensor represented by this NDConfiguration.

indicesMap(int) - Method in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

This method receives an axis index and return the indices mapping value of said axis to enable readable access to the indices map of this configuration.

indicesMap() - Method in class neureka.ndim.config.types.permuted.PermutedNDConfiguration

If one wants to for example access the fourth last item of all items within a tensor based on a scalar index x then the NDConfiguration.indicesMap() is needed as a basis for translating said scalar index x to an array of indices for every axis of the tensor represented by this NDConfiguration.

indicesMap(int) - Method in class neureka.ndim.config.types.permuted.PermutedNDConfiguration

This method receives an axis index and return the indices mapping value of said axis to enable readable access to the indices map of this configuration.

indicesMap() - Method in class neureka.ndim.config.types.simple.Simple0DConfiguration

If one wants to for example access the fourth last item of all items within a tensor based on a scalar index x then the NDConfiguration.indicesMap() is needed as a basis for translating said scalar index x to an array of indices for every axis of the tensor represented by this NDConfiguration.

indicesMap(int) - Method in class neureka.ndim.config.types.simple.Simple0DConfiguration

This method receives an axis index and return the indices mapping value of said axis to enable readable access to the indices map of this configuration.

indicesMap() - Method in class neureka.ndim.config.types.simple.Simple1DConfiguration

If one wants to for example access the fourth last item of all items within a tensor based on a scalar index x then the NDConfiguration.indicesMap() is needed as a basis for translating said scalar index x to an array of indices for every axis of the tensor represented by this NDConfiguration.

indicesMap(int) - Method in class neureka.ndim.config.types.simple.Simple1DConfiguration

This method receives an axis index and return the indices mapping value of said axis to enable readable access to the indices map of this configuration.

indicesMap() - Method in class neureka.ndim.config.types.simple.Simple2DConfiguration

If one wants to for example access the fourth last item of all items within a tensor based on a scalar index x then the NDConfiguration.indicesMap() is needed as a basis for translating said scalar index x to an array of indices for every axis of the tensor represented by this NDConfiguration.

indicesMap(int) - Method in class neureka.ndim.config.types.simple.Simple2DConfiguration

This method receives an axis index and return the indices mapping value of said axis to enable readable access to the indices map of this configuration.

indicesMap() - Method in class neureka.ndim.config.types.simple.Simple3DConfiguration

If one wants to for example access the fourth last item of all items within a tensor based on a scalar index x then the NDConfiguration.indicesMap() is needed as a basis for translating said scalar index x to an array of indices for every axis of the tensor represented by this NDConfiguration.

indicesMap(int) - Method in class neureka.ndim.config.types.simple.Simple3DConfiguration

This method receives an axis index and return the indices mapping value of said axis to enable readable access to the indices map of this configuration.

indicesMap() - Method in class neureka.ndim.config.types.simple.SimpleNDConfiguration

If one wants to for example access the fourth last item of all items within a tensor based on a scalar index x then the NDConfiguration.indicesMap() is needed as a basis for translating said scalar index x to an array of indices for every axis of the tensor represented by this NDConfiguration.

indicesMap(int) - Method in class neureka.ndim.config.types.simple.SimpleNDConfiguration

This method receives an axis index and return the indices mapping value of said axis to enable readable access to the indices map of this configuration.

indicesMap() - Method in class neureka.ndim.config.types.sliced.Sliced0DConfiguration

If one wants to for example access the fourth last item of all items within a tensor based on a scalar index x then the NDConfiguration.indicesMap() is needed as a basis for translating said scalar index x to an array of indices for every axis of the tensor represented by this NDConfiguration.

indicesMap(int) - Method in class neureka.ndim.config.types.sliced.Sliced0DConfiguration

indicesMap() - Method in class neureka.ndim.config.types.sliced.Sliced1DConfiguration

If one wants to for example access the fourth last item of all items within a tensor based on a scalar index x then the NDConfiguration.indicesMap() is needed as a basis for translating said scalar index x to an array of indices for every axis of the tensor represented by this NDConfiguration.

indicesMap(int) - Method in class neureka.ndim.config.types.sliced.Sliced1DConfiguration

This method receives an axis index and return the indices mapping value of said axis to enable readable access to the indices map of this configuration.

indicesMap() - Method in class neureka.ndim.config.types.sliced.Sliced2DConfiguration

If one wants to for example access the fourth last item of all items within a tensor based on a scalar index x then the NDConfiguration.indicesMap() is needed as a basis for translating said scalar index x to an array of indices for every axis of the tensor represented by this NDConfiguration.

indicesMap(int) - Method in class neureka.ndim.config.types.sliced.Sliced2DConfiguration

This method receives an axis index and return the indices mapping value of said axis to enable readable access to the indices map of this configuration.

indicesMap() - Method in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

If one wants to for example access the fourth last item of all items within a tensor based on a scalar index x then the NDConfiguration.indicesMap() is needed as a basis for translating said scalar index x to an array of indices for every axis of the tensor represented by this NDConfiguration.

indicesMap(int) - Method in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

This method receives an axis index and return the indices mapping value of said axis to enable readable access to the indices map of this configuration.

indicesMap() - Method in class neureka.ndim.config.types.sliced.SlicedNDConfiguration

If one wants to for example access the fourth last item of all items within a tensor based on a scalar index x then the NDConfiguration.indicesMap() is needed as a basis for translating said scalar index x to an array of indices for every axis of the tensor represented by this NDConfiguration.

indicesMap(int) - Method in class neureka.ndim.config.types.sliced.SlicedNDConfiguration

This method receives an axis index and return the indices mapping value of said axis to enable readable access to the indices map of this configuration.

indicesMap() - Method in class neureka.ndim.config.types.views.SimpleReshapeView

If one wants to for example access the fourth last item of all items within a tensor based on a scalar index x then the NDConfiguration.indicesMap() is needed as a basis for translating said scalar index x to an array of indices for every axis of the tensor represented by this NDConfiguration.

indicesMap(int) - Method in class neureka.ndim.config.types.views.SimpleReshapeView

This method receives an axis index and return the indices mapping value of said axis to enable readable access to the indices map of this configuration.

indicesMap() - Method in class neureka.ndim.config.types.views.virtual.VirtualNDConfiguration

If one wants to for example access the fourth last item of all items within a tensor based on a scalar index x then the NDConfiguration.indicesMap() is needed as a basis for translating said scalar index x to an array of indices for every axis of the tensor represented by this NDConfiguration.

indicesMap(int) - Method in class neureka.ndim.config.types.views.virtual.VirtualNDConfiguration

This method receives an axis index and return the indices mapping value of said axis to enable readable access to the indices map of this configuration.

indicesMap() - Method in interface neureka.ndim.NDimensional

indicesOfIndex(int) - Method in interface neureka.ndim.config.NDConfiguration

The following method calculates the axis indices for an element in the nd-array array based on a provided "virtual index".

indicesOfIndex(int) - Method in class neureka.ndim.config.types.permuted.Permuted1DConfiguration

The following method calculates the axis indices for an element in the nd-array array based on a provided "virtual index".

indicesOfIndex(int) - Method in class neureka.ndim.config.types.permuted.Permuted2DConfiguration

The following method calculates the axis indices for an element in the nd-array array based on a provided "virtual index".

indicesOfIndex(int) - Method in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

The following method calculates the axis indices for an element in the nd-array array based on a provided "virtual index".

indicesOfIndex(int) - Method in class neureka.ndim.config.types.permuted.PermutedNDConfiguration

The following method calculates the axis indices for an element in the nd-array array based on a provided "virtual index".

indicesOfIndex(int) - Method in class neureka.ndim.config.types.simple.Simple0DConfiguration

The following method calculates the axis indices for an element in the nd-array array based on a provided "virtual index".

indicesOfIndex(int) - Method in class neureka.ndim.config.types.simple.Simple1DConfiguration

The following method calculates the axis indices for an element in the nd-array array based on a provided "virtual index".

indicesOfIndex(int) - Method in class neureka.ndim.config.types.simple.Simple2DConfiguration

The following method calculates the axis indices for an element in the nd-array array based on a provided "virtual index".

indicesOfIndex(int) - Method in class neureka.ndim.config.types.simple.Simple3DConfiguration

The following method calculates the axis indices for an element in the nd-array array based on a provided "virtual index".

indicesOfIndex(int) - Method in class neureka.ndim.config.types.simple.SimpleNDConfiguration

The following method calculates the axis indices for an element in the nd-array array based on a provided "virtual index".

indicesOfIndex(int) - Method in class neureka.ndim.config.types.sliced.Sliced0DConfiguration

The following method calculates the axis indices for an element in the nd-array array based on a provided "virtual index".

indicesOfIndex(int) - Method in class neureka.ndim.config.types.sliced.Sliced1DConfiguration

The following method calculates the axis indices for an element in the nd-array array based on a provided "virtual index".

indicesOfIndex(int) - Method in class neureka.ndim.config.types.sliced.Sliced2DConfiguration

The following method calculates the axis indices for an element in the nd-array array based on a provided "virtual index".

indicesOfIndex(int) - Method in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

The following method calculates the axis indices for an element in the nd-array array based on a provided "virtual index".

indicesOfIndex(int) - Method in class neureka.ndim.config.types.sliced.SlicedNDConfiguration

The following method calculates the axis indices for an element in the nd-array array based on a provided "virtual index".

indicesOfIndex(int) - Method in class neureka.ndim.config.types.views.SimpleReshapeView

The following method calculates the axis indices for an element in the nd-array array based on a provided "virtual index".

indicesOfIndex(int) - Method in class neureka.ndim.config.types.views.virtual.VirtualNDConfiguration

The following method calculates the axis indices for an element in the nd-array array based on a provided "virtual index".

indicesOfIndex(int) - Method in interface neureka.ndim.NDimensional

This is a convenience method identical to ndArray.getNDConf().IndicesOfIndex(i).

init(int, int[]) - Method in interface neureka.ndim.Filler

initialScramble(long) - Static method in class neureka.backend.main.implementations.elementwise.CPURandomization

input(int) - Method in class neureka.backend.api.Call

input(Class<V>, int) - Method in class neureka.backend.api.Call

inputIndex() - Method in class neureka.autograd.ADTarget

inputs() - Method in class neureka.backend.api.Call

INSTANCE - Static variable in interface neureka.devices.DeviceCleaner

intoRange(int, int) - Method in interface neureka.devices.Device.Writer

Writes whatever kind of data was previously specified, to the tensors' data into the range targeted by the provided start and limit.

intStream(int, int) - Static method in class neureka.common.utility.DataConverter.Utility

Use this to create a range based IntStream which is only parallel if the provided threshold smaller than the provided workload size.

intToBigInteger(int[]) - Static method in class neureka.common.utility.DataConverter.Utility

intToByte(int[]) - Static method in class neureka.common.utility.DataConverter.Utility

intToDouble(int[]) - Static method in class neureka.common.utility.DataConverter.Utility

intToFloat(int[]) - Static method in class neureka.common.utility.DataConverter.Utility

intToLong(int[]) - Static method in class neureka.common.utility.DataConverter.Utility

intToShort(int[]) - Static method in class neureka.common.utility.DataConverter.Utility

invert(int[]) - Static method in class neureka.backend.main.operations.other.Permute

invoke(Supplier<T>) - Method in class neureka.backend.api.BackendContext.Runner

Use this method to supply a lambda which will be executed in the BackendContext which produced this very BackendContext.Runner instance.

invoke(double, double) - Method in interface neureka.backend.main.implementations.fun.api.CPUBiFun

invoke(float, float) - Method in interface neureka.backend.main.implementations.fun.api.CPUBiFun

invoke(int, int) - Method in interface neureka.backend.main.implementations.fun.api.CPUBiFun

invoke(long, long) - Method in interface neureka.backend.main.implementations.fun.api.CPUBiFun

invoke(byte, byte) - Method in interface neureka.backend.main.implementations.fun.api.CPUBiFun

invoke(short, short) - Method in interface neureka.backend.main.implementations.fun.api.CPUBiFun

invoke(boolean, boolean) - Method in interface neureka.backend.main.implementations.fun.api.CPUBiFun

invoke(char, char) - Method in interface neureka.backend.main.implementations.fun.api.CPUBiFun

invoke(Object, Object) - Method in interface neureka.backend.main.implementations.fun.api.CPUBiFun

invoke(double) - Method in interface neureka.backend.main.implementations.fun.api.CPUFun

invoke(float) - Method in interface neureka.backend.main.implementations.fun.api.CPUFun

invoke(int) - Method in interface neureka.backend.main.implementations.fun.api.CPUFun

invoke(long) - Method in interface neureka.backend.main.implementations.fun.api.CPUFun

invoke(byte) - Method in interface neureka.backend.main.implementations.fun.api.CPUFun

invoke(short) - Method in interface neureka.backend.main.implementations.fun.api.CPUFun

invoke(boolean) - Method in interface neureka.backend.main.implementations.fun.api.CPUFun

invoke(char) - Method in interface neureka.backend.main.implementations.fun.api.CPUFun

invoke(Object) - Method in interface neureka.backend.main.implementations.fun.api.CPUFun

invoke(double[], int, double, double[], int, int, int) - Static method in class neureka.backend.main.operations.linear.internal.blas.AXPY

invoke(float[], int, float, float[], int, int, int) - Static method in class neureka.backend.main.operations.linear.internal.blas.AXPY

invoke(double[], int, double[], int, int, int) - Static method in class neureka.backend.main.operations.linear.internal.blas.DOT

invoke(float[], int, float[], int, int, int) - Static method in class neureka.backend.main.operations.linear.internal.blas.DOT

invoke(long[], int, long[], int, int, int) - Static method in class neureka.backend.main.operations.linear.internal.blas.DOT

invoke(int[], int, int[], int, int, int) - Static method in class neureka.backend.main.operations.linear.internal.blas.DOT

invoke(float[], float[], int, float[]) - Method in interface neureka.backend.main.operations.linear.internal.blas.GEMM.VectorOperationF32

invoke(double[], double[], int, double[]) - Method in interface neureka.backend.main.operations.linear.internal.blas.GEMM.VectorOperationF64

invoke(long[], int, long, long[], int, int, int) - Static method in class neureka.backend.main.operations.linear.internal.blas.IAXPY

invoke(int[], int, int, int[], int, int, int) - Static method in class neureka.backend.main.operations.linear.internal.blas.IAXPY

invoke(long[], int, long[], int, int, int) - Static method in class neureka.backend.main.operations.linear.internal.blas.IDOT

invoke(int[], int, int[], int, int, int) - Static method in class neureka.backend.main.operations.linear.internal.blas.IDOT

invoke(int[], int[], int, int[]) - Method in interface neureka.backend.main.operations.linear.internal.blas.IGEMM.VectorOperationI32

invoke(long[], long[], int, long[]) - Method in interface neureka.backend.main.operations.linear.internal.blas.IGEMM.VectorOperationI64

invoke(ExecutorService, int, int, int) - Method in class neureka.devices.host.concurrent.WorkScheduler

Synchronous execution - wait until it's finished.

invoke(Tensor<T>...) - Method in interface neureka.math.Function.Callable

This method is functionally identically to Function.Callable.call(Tensor[]), however it is best used in Kotlin, where one can omit the function name entirely and call this Function directly!

invoke(double) - Method in interface neureka.math.Function

Invokes this Function with the provided scalar as a single input and returns the scalar result.

invoke(double[], int) - Method in interface neureka.math.Function

Invokes this Function with the provided array of inputs ad an index for input dependent indexing.

invoke(double...) - Method in interface neureka.math.Function

Invokes this Function with the provided array of inputs.

invoke(Call.Builder<T, D>) - Method in interface neureka.math.Function

Use this to pass more context information for execution of input tensors.

invoke(Args, Tensor<T>...) - Method in interface neureka.math.Function

Use this to call this Function alongside with some additional meta-arguments which will be passed to the underlying Operation(s).

invoke(Tensor<T>) - Method in interface neureka.math.Function

This method is functionally identically to Function.call(Tensor), however it is best used in Kotlin, where one can omit the function name entirely and call this Function directly!

invoke(List<Tensor<T>>) - Method in interface neureka.math.Function

This method is functionally identically to Function.call(List), however it is best used in Kotlin, where one can omit the function name entirely and call this Function directly!

invoke(Tensor<T>[], int) - Method in interface neureka.math.Function

This method is functionally identically to Function.call(Tensor[], int), however it is best used in Kotlin, where one can omit the function name entirely and call this Function directly!

invoke(Tensor<T>...) - Method in interface neureka.math.Function

This method is functionally identically to Function.call(Tensor[]), however it is best used in Kotlin, where one can omit the function name entirely and call this Function directly!

is(Class<?>) - Method in interface neureka.Tensor

This method compares the passed class with the underlying data-type of this NDArray.

isAnOperation(String) - Static method in class neureka.math.parsing.ParseUtil

isApplyingGradientWhenRequested() - Method in class neureka.Neureka.Settings.AutoGrad

Gradients will only be applied if requested.

isApplyingGradientWhenTensorIsUsed() - Method in class neureka.Neureka.Settings.AutoGrad

Gradients will automatically be applied (or JITed) to tensors as soon as they are being used for calculation (GraphNode instantiation).

isAutoConvertToFloat() - Method in class neureka.backend.ocl.CLSettings

isBranch() - Method in interface neureka.Tensor

Tensors which are used or produced by the autograd system will have a GraphNode component attached to them.

isCase(Tensor<V>) - Method in interface neureka.Tensor

This method name translates to the "in" keyword in Groovy! The same is true for the "contains" method in Kotlin.

isCompact() - Method in interface neureka.ndim.config.NDConfiguration

NDConfiguration instance where this flag is true will most likely not be slices because they have no offset (all 0) and a compact spread / step array (all 1).

isCompatible(NDConfiguration.Layout) - Method in enum neureka.ndim.config.NDConfiguration.Layout

isDeleted() - Method in interface neureka.Tensor

This will check if the MutateTensor.delete() method was previously called on this tensor.

isDeletingIntermediateTensors() - Method in class neureka.Neureka.Settings.Debug

Function instances will produce hidden intermediate results when executing an array of inputs.

isDifferentiable() - Method in interface neureka.backend.api.Operation

Deprecated.

isDifferentiable() - Method in class neureka.backend.api.template.operations.AbstractOperation

isDifferentiable(boolean) - Method in class neureka.backend.api.template.operations.OperationBuilder

isDoingAD() - Method in interface neureka.math.Function

Only branch Functions can do autograd / 'Auto-Differentiation', meaning functions whose Function.isFlat() flag is set to false!

isDoingAD() - Method in class neureka.math.implementations.FunctionConstant

isDoingAD() - Method in class neureka.math.implementations.FunctionInput

isDoingAD() - Method in class neureka.math.implementations.FunctionNode

isDoingAD() - Method in class neureka.math.implementations.FunctionVariable

isDone() - Method in class neureka.autograd.JITProp

isEmpty() - Method in class neureka.devices.AbstractBaseDevice

A device is empty if there are no tensors stored on it.

isEmpty() - Method in interface neureka.devices.Storage

isEmpty() - Method in interface neureka.Tensor

A tensor is empty if it's Data storage is null.

isFirstColIsIndex() - Method in class neureka.devices.file.CSVHandle

isFirstRowIsLabels() - Method in class neureka.devices.file.CSVHandle

isFlat() - Method in interface neureka.math.Function

isFlat() - Method in class neureka.math.implementations.FunctionConstant

isFlat() - Method in class neureka.math.implementations.FunctionInput

isFlat() - Method in class neureka.math.implementations.FunctionNode

isFlat() - Method in class neureka.math.implementations.FunctionVariable

isFullSlice() - Method in interface neureka.Nda

If this nd-array is a full slice of a parent nd-array then this method will yield true.

isGraphLeave() - Method in class neureka.autograd.GraphNode

isIndexer() - Method in interface neureka.backend.api.Operation

This boolean property tell the Function implementations that this Operation ought to be viewed as something to be indexed.

isIndexer() - Method in class neureka.backend.api.template.operations.AbstractOperation

isIndexer(boolean) - Method in class neureka.backend.api.template.operations.OperationBuilder

isInline() - Method in interface neureka.backend.api.Operation

This flag indicates that the implementation of this Operation performs an operation which modifies the inputs to that operation.

isInline() - Method in class neureka.backend.api.template.operations.AbstractOperation

isInline(boolean) - Method in class neureka.backend.api.template.operations.OperationBuilder

isIntermediate() - Method in interface neureka.Tensor

Intermediate tensors are internal non-user tensors which may be eligible for deletion when further consumed by a Function.

isKeepingDerivativeTargetPayloads() - Method in class neureka.Neureka.Settings.Debug

Every derivative is calculated with respect to some graph node.

isL2Specified() - Method in class neureka.devices.host.machine.Hardware

isL3Specified() - Method in class neureka.devices.host.machine.Hardware

isLeave() - Method in class neureka.autograd.GraphNode

This node (and the corresponding tensor) was not created by a function! (it's a leave tensor)

isLeave() - Method in interface neureka.Tensor

Tensors which are used or produced by the autograd system will have a GraphNode component attached to them.

isLocked() - Method in class neureka.Neureka.Settings

Locked settings can only be read but not written to.

isOnlyUsingDefaultNDConfiguration() - Method in class neureka.Neureka.Settings.NDim

This flag determines which NDConfiguration implementations should be used for nd-arrays/tensors.

isOperator() - Method in interface neureka.backend.api.Operation

An operator is an alternative to a function like "sum()" or "prod()".

isOperator() - Method in class neureka.backend.api.template.operations.AbstractOperation

isOperator(boolean) - Method in class neureka.backend.api.template.operations.OperationBuilder

isOutsourced() - Method in interface neureka.Tensor

Outsourced means that the tensor is stored on a Device implementation instance which is not the CPU.

isPartialSlice() - Method in interface neureka.Nda

If this nd-array is a partial slice of a parent nd-array then this method will yield true.

isPartialSlice() - Method in interface neureka.Tensor

If this nd-array is a partial slice of a parent nd-array then this method will yield true.

isPreventingInlineOperations() - Method in class neureka.Neureka.Settings.AutoGrad

Inline operations are operations where the data of a tensor passed into an operation is being modified.

isReliesOnJustInTimeProp() - Method in class neureka.autograd.GraphNode

This flag is used for a performance optimization feature namely 'Just In Time Propagation'.

isRetainingPendingErrorForJITProp() - Method in class neureka.Neureka.Settings.AutoGrad

This flag enables an optimization technique which only propagates error values to gradients if needed by a tensor (the tensor is used again) and otherwise accumulate them at divergent differentiation paths within the computation graph.
If the flag is set to true
then error values will accumulate at such junction nodes.

isShallowCopy() - Method in interface neureka.Nda

If this nd-array is a shallow copy of a parent nd-array then this method will yield true.

isShallowCopy() - Method in interface neureka.Tensor

If this nd-array is a shallow copy of a parent nd-array then this method will yield true.

isSimple() - Method in interface neureka.ndim.config.NDConfiguration

The boolean returned by this method simply reports if this configuration is the most basic form of configuration possible for the given shape represented by this instance.

isSlice() - Method in interface neureka.Nda

If this nd-array is a slice of a parent nd-array then this method will yield true.

isSlice() - Method in interface neureka.Tensor

If this nd-array is a slice of a parent nd-array then this method will yield true.

isSliceParent() - Method in interface neureka.Nda

If slices have been derived from this nd-array then it is a "slice parent".

isSliceParent() - Method in interface neureka.Tensor

If slices have been derived from this nd-array then it is a "slice parent".

isSuitableFor(ExecutionCall<? extends Device<?>>) - Method in interface neureka.backend.api.fun.SuitabilityPredicate

When an ExecutionCall instance has been formed then it will be routed by
the given Operation instance to their components, namely :
Algorithm instances !
The ability to decide which algorithm is suitable for a given ExecutionCall instance
is being granted by implementations of the following method.

isSuitableFor(ExecutionCall<? extends Device<?>>) - Method in class neureka.backend.api.template.algorithms.AbstractFunAlgorithm

The SuitabilityPredicate checks if a given instance of an ExecutionCall is suitable to be executed in ImplementationFor residing in this Algorithm as components.

isSuitableFor(ExecutionCall<? extends Device<?>>) - Method in class neureka.backend.api.template.algorithms.AbstractFunDeviceAlgorithm

The SuitabilityPredicate checks if a given instance of an ExecutionCall is suitable to be executed in ImplementationFor residing in this Algorithm as components.

isSuitableFor(ExecutionCall<? extends Device<?>>) - Method in class neureka.backend.api.template.algorithms.FallbackAlgorithm

isUndefined() - Method in interface neureka.Tensor

A tensor is "undefined" if it has either no NDConfiguration implementation instance or this instance does not have a shape set for this Tensor which is needed for a tensor to also have a rank and dimensionality...

isUsedAsDerivative() - Method in class neureka.autograd.GraphNode

The chain-rule states that the derivative of f(x) = h(g(x)) with respect to x is: g'(x) * h'(g(x)) An example would be: f(x) = ((x*y)*z) f'(x) = (1*y) * (1*z) = z*y The values z,y or z*y must not be deleted as they are needed for back-propagation!

isValid() - Method in class neureka.backend.api.Call.Validator

isVirtual() - Method in interface neureka.ndim.config.NDConfiguration

isVirtual() - Method in class neureka.ndim.config.types.views.virtual.VirtualNDConfiguration

isVirtual() - Method in interface neureka.Tensor

A Virtual tensor is a tensor whose underlying data array is of size 1, holding only a single value.

isWronglyIntermediate() - Method in class neureka.backend.main.memory.MemValidator

isWronglyNonIntermediate() - Method in class neureka.backend.main.memory.MemValidator

item(int) - Method in interface neureka.Nda

The following method returns a single item within this nd-array targeted by the provided integer index.

item(int...) - Method in interface neureka.Nda

This method returns a raw value item within this nd-array targeted by an index array which is expected to hold an index for every dimension of the shape of this nd-array.

item() - Method in interface neureka.Nda

Equivalent to the #item(0) and Nda.getItem().

items() - Method in interface neureka.Nda

A more concise version of the Nda.getItems() method.

itemType() - Method in interface neureka.Nda

iterator() - Method in interface neureka.Shape

IterByOrIterFromOrAll<V> - Interface in neureka.fluent.building.states

IterByOrIterFromOrAllTensor<V> - Interface in neureka.fluent.building.states

J

JITProp<V> - Class in neureka.autograd

This class keeps track of graph nodes which require back-propagation in order to be able to continue the process at a later point in time (based on some configurable conditions).

JITProp(Set<GraphNode<V>>) - Constructor for class neureka.autograd.JITProp

JVMExecutor() - Constructor for class neureka.devices.host.CPU.JVMExecutor

K

K - Static variable in class neureka.devices.host.machine.CommonMachine

keep(Tensor<?>[], Supplier<T>) - Static method in class neureka.backend.main.memory.MemUtil

This method makes sure that the provided tensors do not get deleted by setting the Tensor.isIntermediate() flag to off during the execution of the provided Supplier lambda! In said lambda the supplied thing will ultimately be returned by this method...

keep(Tensor<?>, Tensor<?>, Supplier<T>) - Static method in class neureka.backend.main.memory.MemUtil

This method makes sure that the provided tensors do not get deleted by setting the Tensor.isIntermediate() flag to off during the execution of the provided Supplier lambda! In said lambda the supplied thing will ultimately be returned by this method...

KernelCache - Class in neureka.devices.opencl

A fixed sized cache for ad-hoc (just in time compiled) OpenCLDevice kernels.

KernelCache() - Constructor for class neureka.devices.opencl.KernelCache

KernelCaller - Class in neureka.devices.opencl

Instances of this class are utility factories provided by OpenCLDevice instances.

KernelCaller(cl_kernel, cl_command_queue) - Constructor for class neureka.devices.opencl.KernelCaller

KernelCode - Class in neureka.devices.opencl

KernelCode(String, String) - Constructor for class neureka.devices.opencl.KernelCode

KernelCode(String, String, DataType<?>) - Constructor for class neureka.devices.opencl.KernelCode

KernelSource - Interface in neureka.devices.opencl

Provides kernel source code for a provided ExecutionCall.

L

label(String) - Method in interface neureka.MutateNda

Sets the label of this nd-array.

label(String) - Method in interface neureka.MutateTensor

Sets the label of this nd-array.

label() - Method in interface neureka.Nda

A nd-array can have a label.

labelAxes(String[]...) - Method in interface neureka.MutateNda

This method receives a nested String array which ought to contain a label for the index of this tensor.

labelAxes(List<List<Object>>) - Method in interface neureka.MutateNda

This method receives a nested String list which ought to contain a label for the index of this tensor.

labelAxes(Map<Object, List<Object>>) - Method in interface neureka.MutateNda

This method provides the ability to label not only the indices of the shape of this tensor, but also the dimension of the shape.

labelAxes(String[]...) - Method in interface neureka.MutateTensor

This method receives a label for this tensor and a nested String array which ought to contain a label for the index of this tensor.

labelAxes(List<List<Object>>) - Method in interface neureka.MutateTensor

This method receives a nested String list which ought to contain a label for the index of this tensor.

labelAxes(Map<Object, List<Object>>) - Method in interface neureka.MutateTensor

This method provides the ability to label not only the indices of the shape of this tensor, but also the dimension of the shape.

last(Call.TensorCondition) - Method in class neureka.backend.api.Call.Validator

LazyEntry(K, Function<K, V>) - Constructor for class neureka.common.utility.Cache.LazyEntry

LazyRef<V> - Class in neureka.backend.api

This will simply fetch a variable from a lambda once and then continuously return this one value.

learningRate() - Method in class neureka.optimization.implementations.SGD

like(Tensor<V>) - Static method in interface neureka.Tensor

Use this factory method to instantiate a new tensor with the same data type, shape and memory location (Device instance) as the provided template tensor.

ListReader - Class in neureka.common.utility

This is a simple utility class which traverses nested data structures and converts them into information which can be used to instantiate a tensor, namely: A flat data array, a shape array and a type class.

ListReader.Result - Class in neureka.common.utility

ln() - Method in class neureka.math.Functions

ln() - Method in interface neureka.Tensor

This method is a functionally identical to the following alternatives:

load(BackendRegistry) - Method in interface neureka.backend.api.ini.BackendLoader

load() - Method in class neureka.devices.file.CSVHandle

load(String) - Method in class neureka.devices.file.FileDevice

load(String, Map<String, Object>) - Method in class neureka.devices.file.FileDevice

load() - Method in interface neureka.devices.file.FileHandle

An implementation of this method ought to create a new tensor instance containing the data which is stored in the file whose access this FileHandle manages.

load() - Method in class neureka.devices.file.IDXHandle

LoadingContext - Interface in neureka.backend.api.ini

loadProperties(Neureka) - Static method in class neureka.common.utility.SettingsLoader

localMemSize() - Method in class neureka.devices.opencl.OpenCLDevice

localMemType() - Method in class neureka.devices.opencl.OpenCLDevice

LOG10 - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

Log10 - Class in neureka.backend.main.operations.functions

Log10() - Constructor for class neureka.backend.main.operations.functions.Log10

log10() - Method in class neureka.math.Functions

log10() - Method in interface neureka.Tensor

This method is a functionally identical to the following alternatives:

LOGARITHM - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

Logarithm - Class in neureka.backend.main.operations.functions

Logarithm() - Constructor for class neureka.backend.main.operations.functions.Logarithm

LogUtil - Class in neureka.common.utility

A utility class for message formatting.

LogUtil() - Constructor for class neureka.common.utility.LogUtil

longToBigInteger(long[]) - Static method in class neureka.common.utility.DataConverter.Utility

longToByte(long[]) - Static method in class neureka.common.utility.DataConverter.Utility

longToDouble(long[]) - Static method in class neureka.common.utility.DataConverter.Utility

longToFloat(long[]) - Static method in class neureka.common.utility.DataConverter.Utility

longToInt(long[]) - Static method in class neureka.common.utility.DataConverter.Utility

longToShort(long[]) - Static method in class neureka.common.utility.DataConverter.Utility

M

makeFit(Tensor<?>[], boolean) - Static method in class neureka.backend.main.operations.other.Permute

makeSimple() - Static method in class neureka.devices.host.machine.Hardware

makeSimple(String, long, int) - Static method in class neureka.devices.host.machine.Hardware

map(Function<V, V>) - Method in interface neureka.Nda.Item

Maps this item to an optional value based on the provided lambda.

map(Function<V, V>) - Method in interface neureka.Nda

This method is a convenience method for mapping the items of this nd-array to another nd-array of the same type based on the provided lambda function, which will be applied to all items of this nd-array individually (element-wise).

map(int) - Method in interface neureka.ndim.config.NDConfiguration.IndexToIndexFunction

map(Function<Integer, Integer>) - Method in interface neureka.Shape

This method is used to transform a Shape into another Shape by applying a function to it.

map(Function<V, V>) - Method in interface neureka.Tensor

This method is a convenience method for mapping the items of this nd-array to another nd-array of the same type based on the provided lambda function, which will be applied to all items of this nd-array individually (element-wise).

mapTo(Class<T>, Function<V, T>) - Method in interface neureka.Nda

This is a convenience method for mapping a nd-array to a nd-array of new type based on a provided target item type and mapping lambda.

mapTo(Class<T>, Function<V, T>) - Method in interface neureka.Tensor

This is a convenience method for mapping a nd-array to a nd-array of new type based on a provided target item type and mapping lambda.

MatMul - Class in neureka.backend.main.operations.linear

MatMul() - Constructor for class neureka.backend.main.operations.linear.MatMul

matMul() - Method in class neureka.math.Functions

matMul(Tensor<V>) - Method in interface neureka.Tensor

This will produce the matrix product of two tensors with rank 2 (matrices), where the left operand is this Tensor instance and the right operand is the argument passed to the method.

MatMulAlgorithm - Class in neureka.backend.main.algorithms

MatMulAlgorithm() - Constructor for class neureka.backend.main.algorithms.MatMulAlgorithm

Max - Class in neureka.backend.main.operations.other

Max() - Constructor for class neureka.backend.main.operations.other.Max

max() - Method in class neureka.math.Functions

max() - Method in interface neureka.Tensor

Calculate the max value of all values within this tensor and returns it in the form of a scalar tensor.

maxAddressBits() - Method in class neureka.devices.opencl.OpenCLDevice

maxClockFrequenzy() - Method in class neureka.devices.opencl.OpenCLDevice

maxComputeUnits() - Method in class neureka.devices.opencl.OpenCLDevice

maxConstantBufferSize() - Method in class neureka.devices.opencl.OpenCLDevice

maxConstantBufferSizeKB() - Method in class neureka.devices.opencl.OpenCLDevice

maxItem(Comparator<V>) - Method in interface neureka.Nda

Returns the maximum item of this nd-array according to the provided Comparator.

maxMemAllocSize() - Method in class neureka.devices.opencl.OpenCLDevice

maxReadImageArgs() - Method in class neureka.devices.opencl.OpenCLDevice

maxWorkGroupSize() - Method in class neureka.devices.opencl.OpenCLDevice

maxWorkItemSimensions() - Method in class neureka.devices.opencl.OpenCLDevice

maxWorkItemSizes() - Method in class neureka.devices.opencl.OpenCLDevice

maxWriteImageArgs() - Method in class neureka.devices.opencl.OpenCLDevice

mean() - Method in interface neureka.Tensor

Calculate the mean value of all values within this tensor and returns it in the form of a scalar tensor.

memory - Variable in class neureka.devices.host.machine.BasicMachine

MemUtil - Class in neureka.backend.main.memory

Utility methods for deleting tensors or preventing thereof.

MemValidator - Class in neureka.backend.main.memory

This class validates the states of tensors with respect to memory management before and after a lambda executes a function or some kind of algorithm on said tensors.

Messages - Class in neureka.devices.opencl.utility

Messages.Tips - Enum in neureka.devices.opencl.utility

Min - Class in neureka.backend.main.operations.other

Min() - Constructor for class neureka.backend.main.operations.other.Min

min() - Method in class neureka.math.Functions

min() - Method in interface neureka.Tensor

Calculate the min value of all values within this tensor and returns it in the form of a scalar tensor.

minItem(Comparator<V>) - Method in interface neureka.Nda

Returns the minimum item of this nd-array according to the provided Comparator.

minus() - Method in class neureka.math.Functions

minus(Tensor<V>) - Method in interface neureka.Tensor

Performs subtraction on two tensors with the same rank (or two ranks which can be made compatible with padding ones), where the left operand is this Tensor instance and the right operand is the tensor passed to the method.

minus(V) - Method in interface neureka.Tensor

This method will create a new Tensor with the provided item subtracted from all elements of this Tensor.

minusAssign() - Method in class neureka.math.Functions

minusAssign(Tensor<T>) - Method in interface neureka.MutateTensor

minusAssign(T) - Method in interface neureka.MutateTensor

mod() - Method in class neureka.math.Functions

mod(Tensor<V>) - Method in interface neureka.Tensor

Produces the modulus of two tensors with the same rank (or two ranks which can be made compatible with padding ones), where the left operand is this Tensor instance and the right operand is the tensor passed to the method.

mod(int) - Method in interface neureka.Tensor

modAssign() - Method in class neureka.math.Functions

modAssign(Tensor<T>) - Method in interface neureka.MutateTensor

Modulo - Class in neureka.backend.main.operations.operator

Modulo() - Constructor for class neureka.backend.main.operations.operator.Modulo

Momentum<V extends java.lang.Number> - Class in neureka.optimization.implementations

Momentum - Static variable in interface neureka.optimization.Optimizer

MomentumFactory - Class in neureka.optimization.implementations

MomentumFactory() - Constructor for class neureka.optimization.implementations.MomentumFactory

mul() - Method in class neureka.math.Functions

mulAssign() - Method in class neureka.math.Functions

Multiplication - Class in neureka.backend.main.operations.operator

Multiplication() - Constructor for class neureka.backend.main.operations.operator.Multiplication

multiply(Tensor<V>) - Method in interface neureka.Tensor

This method is synonymous to the Tensor.times(Tensor) method.

multiply(V) - Method in interface neureka.Tensor

multiply(double) - Method in interface neureka.Tensor

mut() - Method in interface neureka.Nda

This method exposes an API for mutating the state of this tensor.

mut() - Method in interface neureka.Tensor

This method exposes an API for mutating the state of this tensor.

MutateNda<T> - Interface in neureka

Nd-arrays should be used as immutable data structures mostly, however sometimes it is important to mutate their state for performance reasons.

MutateNda.Item<V> - Interface in neureka

Instances of this are being returned by the Nda.at(int...) method, and they allow you to get or set individual nd-array items

MutateTensor<T> - Interface in neureka

Tensors should be considered immutable, however sometimes it is important to mutate their state for performance reasons.

N

name() - Method in class neureka.devices.opencl.OpenCLDevice

Nda<V> - Interface in neureka

Nda, which is an abbreviation of 'N-Dimensional-Array', represents a multidimensional, homogeneously filled fixed-size array of items.

Nda.Item<V> - Interface in neureka

Instances of this are being returned by the Nda.at(int...) method, and they allow you to get individual nd-array items

NdaAsString - Class in neureka.view

This class is in essence a simple wrapper class for a tensor and a StringBuilder Methods in this class use the builder in order to construct a String representation for said tensor.

NdaAsString.Builder - Interface in neureka.view

A builder interface providing multiple different options for building a NdaAsString instance in a fluent way.

NdaAsString.Util - Class in neureka.view

This class is a simple utility class which contains a collection of static and stateless methods containing useful functionalities for tensor stringification.

NdaBuilder<V> - Class in neureka.fluent.building

This is the implementation of the fluent builder API for creating Nda/Tensor instances.

NdaBuilder(Class<V>) - Constructor for class neureka.fluent.building.NdaBuilder

ndArrays(Consumer<NDPrintSettings>) - Method in class neureka.Neureka.Settings.View

This allows you to provide a lambda to configure how tensors should be converted to String instances.

NDConfiguration - Interface in neureka.ndim.config

This interface represents the access pattern configuration for the data array of a tensor.

NDConfiguration.IndexToIndexFunction - Interface in neureka.ndim.config

Implementations of this are produced and returned by the NDConfiguration.getIndexToIndexAccessPattern() and their purpose is to translate the item index of a tensor to the index of the item within the underlying data array of said tensor.

NDConfiguration.Layout - Enum in neureka.ndim.config

Types of common data layouts:
ROW_MAJOR

NDConfiguration.Utility - Class in neureka.ndim.config

This utility class provides static methods which are helpful for nd-configuration related operations like reshaping, incrementing or decrementing index arrays...

NDConstructor - Interface in neureka.ndim

NDConvolution - Class in neureka.backend.main.algorithms

NDConvolution() - Constructor for class neureka.backend.main.algorithms.NDConvolution

NDFrame<V> - Class in neureka.framing

Instances of this class are components of tensors, which store aliases for the indices of the tensor.

NDFrame(List<List<Object>>, Tensor<V>, String) - Constructor for class neureka.framing.NDFrame

NDFrame(Tensor<V>, String) - Constructor for class neureka.framing.NDFrame

NDFrame(Map<Object, List<Object>>, Tensor<V>, String) - Constructor for class neureka.framing.NDFrame

ndim() - Method in class neureka.Neureka.Settings

ndim(Object) - Method in class neureka.Neureka.Settings

This allows you to configure Neureka using a Groovy DSL.

NDim() - Constructor for class neureka.Neureka.Settings.NDim

NDimensional - Interface in neureka.ndim

This interface defines the most essential methods of the nd-array/tensor API, which describe them with respect to their dimensionality.

NDIterator - Interface in neureka.ndim.iterator

An NDIterator is used to iterate over n-dimensional arrays.

NDIterator.NonVirtual - Enum in neureka.ndim.iterator

Defines if a new NDIterator is allowed to be a VirtualNDIterator.

NDPrintSettings - Class in neureka.view

This is simply a mutable container for configuring how Tensor instances ought to be converted to Strings.

NDPrintSettings(Supplier<Boolean>) - Constructor for class neureka.view.NDPrintSettings

NDTrait - Enum in neureka.ndim.config

NDUtil - Class in neureka.ndim

Static utility methods for the NDArray.

NDUtil() - Constructor for class neureka.ndim.NDUtil

neg() - Method in class neureka.math.Functions

neg() - Method in interface neureka.Tensor

This method is a functionally identical to the following alternatives:

negative() - Method in interface neureka.Tensor

neureka - package neureka

Neureka - Class in neureka

Neureka is the key access point for thread local / global library settings ( seeNeureka.Settings) as well as execution contexts (see BackendContext) and pre-instantiated Functions.

neureka.autograd - package neureka.autograd

neureka.backend.api - package neureka.backend.api

neureka.backend.api.fun - package neureka.backend.api.fun

neureka.backend.api.ini - package neureka.backend.api.ini

neureka.backend.api.template.algorithms - package neureka.backend.api.template.algorithms

neureka.backend.api.template.implementations - package neureka.backend.api.template.implementations

neureka.backend.api.template.operations - package neureka.backend.api.template.operations

neureka.backend.cpu - package neureka.backend.cpu

neureka.backend.main.algorithms - package neureka.backend.main.algorithms

Everything in this package should be considered library-private! DO NOT DEPEND ON CLASSES INSIDE THIS PACKAGE! Code inside this package or any sub-packages might change frequently...

neureka.backend.main.algorithms.internal - package neureka.backend.main.algorithms.internal

neureka.backend.main.implementations - package neureka.backend.main.implementations

Everything in this package should be considered library-private! DO NOT DEPEND ON CLASSES INSIDE THIS PACKAGE! Code inside this package or any sub-packages might change frequently...

neureka.backend.main.implementations.broadcast - package neureka.backend.main.implementations.broadcast

neureka.backend.main.implementations.convolution - package neureka.backend.main.implementations.convolution

neureka.backend.main.implementations.elementwise - package neureka.backend.main.implementations.elementwise

neureka.backend.main.implementations.fun - package neureka.backend.main.implementations.fun

neureka.backend.main.implementations.fun.api - package neureka.backend.main.implementations.fun.api

neureka.backend.main.implementations.linear - package neureka.backend.main.implementations.linear

neureka.backend.main.implementations.matmul - package neureka.backend.main.implementations.matmul

neureka.backend.main.implementations.scalar - package neureka.backend.main.implementations.scalar

neureka.backend.main.internal - package neureka.backend.main.internal

Everything in this package should be considered library-private! DO NOT DEPEND ON CLASSES INSIDE THIS PACKAGE! Code inside this package or any sub-packages might change frequently...

neureka.backend.main.memory - package neureka.backend.main.memory

Everything in this package should be considered library-private! DO NOT DEPEND ON CLASSES INSIDE THIS PACKAGE! Code inside this package or any sub-packages might change frequently...

neureka.backend.main.operations - package neureka.backend.main.operations

Everything in this package should be considered library-private! DO NOT DEPEND ON CLASSES INSIDE THIS PACKAGE! Code inside this package or any sub-packages might change frequently...

neureka.backend.main.operations.functions - package neureka.backend.main.operations.functions

Everything in this package should be considered library-private! DO NOT DEPEND ON CLASSES INSIDE THIS PACKAGE! Code inside this package or any sub-packages might change frequently...

neureka.backend.main.operations.indexer - package neureka.backend.main.operations.indexer

Everything in this package should be considered library-private! DO NOT DEPEND ON CLASSES INSIDE THIS PACKAGE! Code inside this package or any sub-packages might change frequently...

neureka.backend.main.operations.linear - package neureka.backend.main.operations.linear

Everything in this package should be considered library-private! DO NOT DEPEND ON CLASSES INSIDE THIS PACKAGE! Code inside this package or any sub-packages might change frequently...

neureka.backend.main.operations.linear.internal.blas - package neureka.backend.main.operations.linear.internal.blas

Everything in this package should be considered library-private! DO NOT USE CLASSES INSIDE THIS PACKAGE!

neureka.backend.main.operations.linear.internal.opencl - package neureka.backend.main.operations.linear.internal.opencl

neureka.backend.main.operations.operator - package neureka.backend.main.operations.operator

Everything in this package should be considered library-private! DO NOT DEPEND ON CLASSES INSIDE THIS PACKAGE! Code inside this package or any sub-packages might change frequently...

neureka.backend.main.operations.other - package neureka.backend.main.operations.other

Everything in this package should be considered library-private! DO NOT DEPEND ON CLASSES INSIDE THIS PACKAGE! Code inside this package or any sub-packages might change frequently...

neureka.backend.main.operations.other.internal - package neureka.backend.main.operations.other.internal

neureka.backend.ocl - package neureka.backend.ocl

neureka.common.composition - package neureka.common.composition

neureka.common.utility - package neureka.common.utility

neureka.devices - package neureka.devices

neureka.devices.file - package neureka.devices.file

neureka.devices.host - package neureka.devices.host

neureka.devices.host.concurrent - package neureka.devices.host.concurrent

Everything in this package should be considered library-private! DO NOT USE CLASSES INSIDE THIS PACKAGE!

neureka.devices.host.machine - package neureka.devices.host.machine

Everything in this package should be considered library-private! DO NOT USE CLASSES INSIDE THIS PACKAGE!

neureka.devices.opencl - package neureka.devices.opencl

neureka.devices.opencl.utility - package neureka.devices.opencl.utility

neureka.dtype - package neureka.dtype

neureka.dtype.custom - package neureka.dtype.custom

Everything in this package should be considered library-private! DO NOT DEPEND ON CLASSES INSIDE THIS PACKAGE! Code inside this package or any sub-packages might change frequently...

neureka.fluent.building - package neureka.fluent.building

neureka.fluent.building.states - package neureka.fluent.building.states

neureka.fluent.slicing - package neureka.fluent.slicing

neureka.fluent.slicing.states - package neureka.fluent.slicing.states

neureka.framing - package neureka.framing

neureka.framing.fluent - package neureka.framing.fluent

neureka.math - package neureka.math

neureka.math.args - package neureka.math.args

neureka.math.implementations - package neureka.math.implementations

Everything in this package should be considered library-private! DO NOT DEPEND ON CLASSES INSIDE THIS PACKAGE! Code inside this package or any sub-packages might change frequently...

neureka.math.parsing - package neureka.math.parsing

Everything in this package should be considered library-private! DO NOT DEPEND ON CLASSES INSIDE THIS PACKAGE! Code inside this package or any sub-packages might change frequently...

neureka.ndim - package neureka.ndim

neureka.ndim.config - package neureka.ndim.config

neureka.ndim.config.types - package neureka.ndim.config.types

neureka.ndim.config.types.permuted - package neureka.ndim.config.types.permuted

neureka.ndim.config.types.simple - package neureka.ndim.config.types.simple

neureka.ndim.config.types.sliced - package neureka.ndim.config.types.sliced

neureka.ndim.config.types.views - package neureka.ndim.config.types.views

neureka.ndim.config.types.views.virtual - package neureka.ndim.config.types.views.virtual

neureka.ndim.iterator - package neureka.ndim.iterator

neureka.ndim.iterator.types.permuted - package neureka.ndim.iterator.types.permuted

neureka.ndim.iterator.types.simple - package neureka.ndim.iterator.types.simple

neureka.ndim.iterator.types.sliced - package neureka.ndim.iterator.types.sliced

neureka.ndim.iterator.types.virtual - package neureka.ndim.iterator.types.virtual

neureka.optimization - package neureka.optimization

neureka.optimization.implementations - package neureka.optimization.implementations

Neureka.Settings - Class in neureka

This class hosts the settings of the Neureka instance which will be used throughout the library.

Neureka.Settings.AutoGrad - Class in neureka

This class contains settings which are related to the automatic differentiation of tensors.

Neureka.Settings.Debug - Class in neureka

Neureka.Settings.DType - Class in neureka

Neureka.Settings.NDim - Class in neureka

Settings for configuring the access pattern of nd-arrays/tensors.

Neureka.Settings.View - Class in neureka

Settings for configuring how objects should be converted to String representations.

Neureka.Utility - Class in neureka

neureka.view - package neureka.view

newChildToParent(Tensor<T>) - Static method in class neureka.framing.Relation

newInstance() - Static method in interface neureka.Tensor

This static factory method creates and return a completely empty and undefined tensor which is void of any contents and meaning.

newParentToChildren() - Static method in class neureka.framing.Relation

newReshaped(int[]) - Method in class neureka.ndim.config.AbstractNDC

newReshaped(int[]) - Method in interface neureka.ndim.config.NDConfiguration

This method enables reshaping for NDConfiguration implementation instances.

newStridesFor(int[]) - Method in enum neureka.ndim.config.NDConfiguration.Layout

newTensorLike(Tensor<V>, double) - Static method in class neureka.backend.main.operations.ElemWiseUtil

newTensorLike(Class<V>, Shape, boolean, Device<Object>, double) - Static method in class neureka.backend.main.operations.ElemWiseUtil

node() - Method in class neureka.autograd.ADTarget

none() - Static method in interface neureka.Data

This is a static factory method which returns a Data object which does not contain any data.

none(Predicate<V>) - Method in interface neureka.Nda

Iterates over every element of this nd-array, and checks whether none of the elements match the provided lambda.

none() - Static method in interface neureka.ndim.config.NDConfiguration

NOT_GOOD - Static variable in interface neureka.backend.api.fun.SuitabilityPredicate

noteFinished(GraphNode<V>) - Method in class neureka.autograd.JITProp

nullArgCheck(T, String, Class<?>, String...) - Static method in class neureka.common.utility.LogUtil

numberOfArgs() - Method in interface neureka.math.Function

numberOfBytes() - Method in class neureka.dtype.custom.F32

numberOfBytes() - Method in class neureka.dtype.custom.F64

numberOfBytes() - Method in class neureka.dtype.custom.I16

numberOfBytes() - Method in class neureka.dtype.custom.I32

numberOfBytes() - Method in class neureka.dtype.custom.I64

numberOfBytes() - Method in class neureka.dtype.custom.I8

numberOfBytes() - Method in class neureka.dtype.custom.UI16

numberOfBytes() - Method in class neureka.dtype.custom.UI32

numberOfBytes() - Method in class neureka.dtype.custom.UI64

numberOfBytes() - Method in class neureka.dtype.custom.UI8

numberOfBytes() - Method in interface neureka.dtype.NumericType

numberOfChannels - Variable in enum neureka.Tensor.ImageType

numberOfDataObjects() - Method in class neureka.devices.AbstractBaseDevice

numberOfDataObjects() - Method in interface neureka.devices.Device

Note that this is not necessarily equal to Storage.numberOfStored(), because multiple tensors may share a single Data object.

numberOfOperationsWithin(List<String>) - Static method in class neureka.math.parsing.ParseUtil

numberOfStored() - Method in class neureka.devices.AbstractBaseDevice

numberOfStored() - Method in interface neureka.devices.Storage

NumericType<TargetType,TargetArrayType,HolderType,HolderArrayType> - Interface in neureka.dtype

This interface enables "Polymorphic" utility by defining common functionalities used for handling various numeric types.

O

objBooleansToPrimBooleans(Boolean[]) - Static method in class neureka.common.utility.DataConverter.Utility

objBytesToPrimBytes(Byte[]) - Static method in class neureka.common.utility.DataConverter.Utility

objCharsToPrimChars(Character[]) - Static method in class neureka.common.utility.DataConverter.Utility

objDoublesToPrimDoubles(Double[]) - Static method in class neureka.common.utility.DataConverter.Utility

objectsToBytes(Object[], int) - Static method in class neureka.common.utility.DataConverter.Utility

objectsToDoubles(Object[], int) - Static method in class neureka.common.utility.DataConverter.Utility

objectsToFloats(Object[], int) - Static method in class neureka.common.utility.DataConverter.Utility

objectsToInts(Object[], int) - Static method in class neureka.common.utility.DataConverter.Utility

objectsToLongs(Object[], int) - Static method in class neureka.common.utility.DataConverter.Utility

objectsToShorts(Object[], int) - Static method in class neureka.common.utility.DataConverter.Utility

objFloatsToPrimFloats(Float[]) - Static method in class neureka.common.utility.DataConverter.Utility

objIntsToPrimInts(Integer[]) - Static method in class neureka.common.utility.DataConverter.Utility

objLongsToPrimLongs(Long[]) - Static method in class neureka.common.utility.DataConverter.Utility

objShortsToPrimShorts(Short[]) - Static method in class neureka.common.utility.DataConverter.Utility

of(ADAction) - Static method in interface neureka.autograd.ADAction

of(Tensor<?>, ADAction) - Static method in interface neureka.autograd.ADAction

of(Tensor<?>...) - Static method in class neureka.backend.api.ExecutionCall

Use this factory method to build ExecutionCall instances in a readable fashion.

of(ImplementationReceiver) - Static method in class neureka.backend.api.ini.BackendRegistry

of(Supplier<V>) - Static method in class neureka.backend.api.LazyRef

of(Tensor<?>) - Static method in class neureka.backend.api.Result

of(Class<V>, V...) - Static method in interface neureka.Data

of(float...) - Static method in interface neureka.Data

of(double...) - Static method in interface neureka.Data

of(int...) - Static method in interface neureka.Data

of(long...) - Static method in interface neureka.Data

of(byte...) - Static method in interface neureka.Data

of(short...) - Static method in interface neureka.Data

of(boolean...) - Static method in interface neureka.Data

of(char...) - Static method in interface neureka.Data

of(String...) - Static method in interface neureka.Data

of(Class<T>) - Static method in class neureka.dtype.DataType

of(int) - Static method in class neureka.math.args.Arg.Axis

of(Tensor<V>) - Static method in class neureka.math.args.Arg.Derivative

of(int) - Static method in class neureka.math.args.Arg.DerivIdx

of(int[]) - Static method in class neureka.math.args.Arg.Ends

of(int...) - Static method in class neureka.math.args.Arg.Indices

of(NDConfiguration.Layout) - Static method in class neureka.math.args.Arg.Layout

of(int) - Static method in class neureka.math.args.Arg.MinRank

of(int...) - Static method in class neureka.math.args.Arg.Offset

of(String) - Static method in class neureka.math.args.Arg.Seed

of(long) - Static method in class neureka.math.args.Arg.Seed

of(int...) - Static method in class neureka.math.args.Arg.Shape

of(int...) - Static method in class neureka.math.args.Arg.Stride

of(Device<?>) - Static method in class neureka.math.args.Arg.TargetDevice

of(int) - Static method in class neureka.math.args.Arg.VarIdx

of(Arg<?>...) - Static method in class neureka.math.args.Args

of(String) - Static method in interface neureka.math.Function

This static factory method will return Function instances based on a provided mathematical String expression describing the function using 'I[0]', 'I[1]', 'I[2]'...

of(String, boolean) - Static method in interface neureka.math.Function

This static factory method will return Function instances based on a provided mathematical String expression describing the function using 'I[0]', 'I[1]', 'I[2]'...

of(String, boolean) - Static method in class neureka.math.implementations.FunctionInput

of(Class<V>) - Static method in interface neureka.Nda

This is the entry point to the fluent nd-array builder API for building Nda instances in a readable and type safe fashion.

of(double) - Static method in interface neureka.Nda

of(float...) - Static method in interface neureka.Nda

Constructs a vector of floats based on the provided array.

of(double...) - Static method in interface neureka.Nda

Constructs a vector of doubles based on the provided array.

of(byte...) - Static method in interface neureka.Nda

Constructs a vector of bytes based on the provided array.

of(int...) - Static method in interface neureka.Nda

Constructs a vector of ints based on the provided array.

of(long...) - Static method in interface neureka.Nda

Constructs a vector of longs based on the provided array.

of(short...) - Static method in interface neureka.Nda

Constructs a vector of shorts based on the provided array.

of(boolean...) - Static method in interface neureka.Nda

Constructs a vector of booleans based on the provided array.

of(T...) - Static method in interface neureka.Nda

Constructs a vector of objects based on the provided array.

of(Shape, double...) - Static method in interface neureka.Nda

Use this to construct and return a double based nd-array of the specified shape and initial values.

of(Shape, float...) - Static method in interface neureka.Nda

Use this to construct and return a float based nd-array of the specified shape and initial values.

of(Shape, byte...) - Static method in interface neureka.Nda

Use this to construct and return a byte based nd-array of the specified shape and initial values.

of(Shape, int...) - Static method in interface neureka.Nda

Use this to construct and return a int based nd-array of the specified shape and initial values.

of(Shape, long...) - Static method in interface neureka.Nda

Use this to construct and return a long based nd-array of the specified shape and initial values.

of(Shape, short...) - Static method in interface neureka.Nda

Use this to construct and return a short based nd-array of the specified shape and initial values.

of(Shape, boolean...) - Static method in interface neureka.Nda

Use this to construct and return a boolean based nd-array of the specified shape and initial values.

of(Shape, T...) - Static method in interface neureka.Nda

Use this to construct and return an object based nd-array of the specified shape and initial values.

of(Iterable<T>) - Static method in interface neureka.Nda

Constructs a vector of objects based on the provided iterable.

of(List<T>) - Static method in interface neureka.Nda

Constructs a vector of objects based on the provided list.

of(int[], int[], int[], int[], int[]) - Static method in interface neureka.ndim.config.NDConfiguration

of(Tensor<?>) - Static method in interface neureka.ndim.iterator.NDIterator

Use this to instantiate NDIterators optimized for the provided tensor.

of(Tensor<?>, NDIterator.NonVirtual) - Static method in interface neureka.ndim.iterator.NDIterator

Use this to instantiate NDIterators optimized for the provided tensor which may not be allowed to be a VirtualNDIterator instance.

of(NDConfiguration, NDIterator.NonVirtual) - Static method in interface neureka.ndim.iterator.NDIterator

Use this to instantiate NDIterators optimized for the provided NDConfiguration which may not be allowed to be a VirtualNDIterator instance.

of(int[], int[], int[], int[], int[]) - Static method in interface neureka.ndim.NDConstructor

of(NDConfiguration) - Static method in interface neureka.ndim.NDConstructor

of(Shape) - Static method in interface neureka.ndim.NDConstructor

of(int...) - Static method in interface neureka.ndim.NDConstructor

of(Optimization<T>) - Static method in interface neureka.optimization.Optimizer

of(List<? extends Number>) - Static method in interface neureka.Shape

This method is used to create a Shape instance from a list of numbers whose integer values are used to describe the shape of a nd-array.

of(Stream<? extends Number>) - Static method in interface neureka.Shape

This method is used to create a Shape instance from a stream of numbers whose integer values are used to describe the shape of a nd-array.

of(Iterable<? extends Number>) - Static method in interface neureka.Shape

This method is used to create a Shape instance from an iterable of numbers whose integer values are used to describe the shape of a nd-array.

of(int...) - Static method in interface neureka.Shape

This method is used to create a Shape instance from an array of integers.

of(Tensor<T>, char, Tensor<T>) - Static method in interface neureka.Tensor

Use this to conveniently operate on 2 tensors.

of(Tensor<T>, char, Tensor<T>, char, Tensor<T>) - Static method in interface neureka.Tensor

Use this to conveniently operate on 3 tensors.

of(String, Tensor<T>, String) - Static method in interface neureka.Tensor

Use this to conveniently operate on a tensor.

of(String, Tensor<T>, char, Tensor<T>, String) - Static method in interface neureka.Tensor

Use this to conveniently operate on 2 tensors.

of(String, Tensor<T>, String, Tensor<T>, String, Tensor<T>, String) - Static method in interface neureka.Tensor

Use this to conveniently operate on 3 tensors.

of(Object...) - Static method in interface neureka.Tensor

This static Tensor factory method tries to interpret the provided arguments to create the instance the use might wants.

of(Iterable<T>) - Static method in interface neureka.Tensor

Constructs a vector of objects based on the provided iterable.

of(List<Integer>, T) - Static method in interface neureka.Tensor

This is a convenient factory method for creating Tensor instances for values of type T based on a list of integers defining a shape made up of axes sizes as well as a scalar value of type T which will fill out the data array spanned by the provided shape information.

of(Shape, T) - Static method in interface neureka.Tensor

This is a convenient factory method for creating Tensor instances for representing items of type T.

of(List<? extends Number>, String) - Static method in interface neureka.Tensor

This factory method will create and return a Tensor instance based on a list of Number instances whose rounded values will be interpreted as the shape of this new Tensor instance and a seed which will serve as a source of pseudo randomness to generate the values for the new instance.

of(List<? extends Number>, List<V>) - Static method in interface neureka.Tensor

Creates a new Tensor instance based on a list of numbers representing the shape, and a list of values representing the value of the resulting tensor.

of(Shape, List<V>) - Static method in interface neureka.Tensor

Creates a new Tensor instance based on a shape tuple of numbers representing the nd-array shape, and a list of items representing the value of the resulting tensor.

of(List<Object>) - Static method in interface neureka.Tensor

This factory method will turn a list of values or nested lists of values into a Tensor instance with the corresponding rank and shape.

of(Class<T>, List<Object>) - Static method in interface neureka.Tensor

This factory method will turn a list of values or nested lists of values into a Tensor instance with the corresponding rank and shape and whose values are of the provided type.

of(Class<V>) - Static method in interface neureka.Tensor

This is the entry point to the fluent tensor builder API for building Tensor instances in a readable and type safe fashion.

of(double...) - Static method in interface neureka.Tensor

Constructs a vector of doubles based on the provided array.

of(double) - Static method in interface neureka.Tensor

of(float...) - Static method in interface neureka.Tensor

Constructs a vector of floats based on the provided array.

of(float) - Static method in interface neureka.Tensor

of(byte...) - Static method in interface neureka.Tensor

Constructs a vector of bytes based on the provided array.

of(byte) - Static method in interface neureka.Tensor

of(int...) - Static method in interface neureka.Tensor

Constructs a vector of ints based on the provided array.

of(int) - Static method in interface neureka.Tensor

of(long...) - Static method in interface neureka.Tensor

Constructs a vector of longs based on the provided array.

of(long) - Static method in interface neureka.Tensor

of(short...) - Static method in interface neureka.Tensor

Constructs a vector of shorts based on the provided array.

of(short) - Static method in interface neureka.Tensor

of(boolean...) - Static method in interface neureka.Tensor

Constructs a vector of booleans based on the provided array.

of(Class<V>, Shape, Arg.Seed) - Static method in interface neureka.Tensor

Use this to construct and return a seeded tensor of the specified type.

of(Shape, double) - Static method in interface neureka.Tensor

Use this to construct and return a homogeneously populated double tensor of the specified shape.

of(Shape, double[]) - Static method in interface neureka.Tensor

Use this to construct and return a double tensor of the specified shape and initial values.

of(Shape, int[]) - Static method in interface neureka.Tensor

Use this to construct and return an int tensor of the specified shape and initial values.

of(Shape, byte[]) - Static method in interface neureka.Tensor

Use this to construct and return a byte tensor of the specified shape and initial values.

of(Shape, long[]) - Static method in interface neureka.Tensor

Use this to construct and return a long tensor of the specified shape and initial values.

of(Shape, short[]) - Static method in interface neureka.Tensor

Use this to construct and return a short tensor of the specified shape and initial values.

of(Shape, float[]) - Static method in interface neureka.Tensor

Use this to construct and return a float tensor of the specified shape and initial values.

of(Shape, float) - Static method in interface neureka.Tensor

Use this to construct and return a homogeneously populated float tensor of the specified shape.

of(Shape, boolean[]) - Static method in interface neureka.Tensor

Use this to construct and return a boolean tensor of the specified shape and initial values.

of(Shape, Data<V>) - Static method in interface neureka.Tensor

Use this to construct and return a tensor of the specified shape and data object.
This method is typically used like this:

of(DataType<V>, Shape) - Static method in interface neureka.Tensor

Use this to construct and return a tensor of the specified type and shape.

of(Class<V>, Shape, Object) - Static method in interface neureka.Tensor

Use this to construct and return a tensor of the specified type, shape and data object.

of(Class<V>, List<Integer>, Object) - Static method in interface neureka.Tensor

Use this to construct and return a tensor of the specified type, shape and data object.

of(Class<V>, Shape, Number) - Static method in interface neureka.Tensor

Use this to construct and return a tensor of the specified type, shape and number.

of(Class<V>, List<Integer>, List<V>) - Static method in interface neureka.Tensor

Use this to construct and return a tensor of the specified type, shape and data object.

of(Class<V>, Shape, List<V>) - Static method in interface neureka.Tensor

Use this to construct and return a tensor of the specified type, shape and list of items.

of(DataType<V>, List<Integer>, List<V>) - Static method in interface neureka.Tensor

Use this to construct and return a tensor of the specified type, shape and data object.

of(DataType<V>, Shape, List<V>) - Static method in interface neureka.Tensor

Use this to construct and return a tensor of the specified type, shape and a list of items.

of(DataType<V>, Shape, Object) - Static method in interface neureka.Tensor

This factory method is among the most flexible and forgiving ways to create a Tensor instance.

of(DataType<V>, Device<N>, Shape, Object) - Static method in interface neureka.Tensor

This factory method is among the most flexible and forgiving ways to create a Tensor instance.

of(DataType<V>, NDConstructor, Data<V>) - Static method in interface neureka.Tensor

This factory method a raw tensor constructor which will not perform any type checking or data conversion on the data provided to it.

of(DataType<T>, List<Integer>, Filler<T>) - Static method in interface neureka.Tensor

This factory method allows the creation of tensors with an additional initialization lambda for filling the underlying data array with desired values.

of(DataType<T>, Shape, Filler<T>) - Static method in interface neureka.Tensor

This factory method allows the creation of tensors with an additional initialization lambda for filling the underlying data array with desired values.

of(Class<T>, Shape, Filler<T>) - Static method in interface neureka.Tensor

This factory method allows the creation of tensors with an additional initialization lambda for filling the underlying data array with desired values.

of(String, V...) - Static method in interface neureka.Tensor

This factory method allows for the creation and execution of Function instances without actually instantiating them manually, where the result will then be returned by this factory method.

of(String, List<Tensor<V>>) - Static method in interface neureka.Tensor

This factory method allows for the creation and execution of Function instances without actually instantiating them manually, where the result will then be returned by this factory method.

of(String, boolean, List<Tensor<V>>) - Static method in interface neureka.Tensor

This method takes a list of tensors and a String expression describing operations which ought to be applied to the tensors in said list.

of(String, Tensor<V>) - Static method in interface neureka.Tensor

This method takes a tensor and a String expression describing operations which ought to be applied to said tensor.

of(String, Tensor<V>...) - Static method in interface neureka.Tensor

This method takes an array of tensors and a String expression describing operations which ought to be applied to the tensors in said array.

of(String, boolean, Tensor<V>...) - Static method in interface neureka.Tensor

This method takes an array of tensors and a String expression describing operations which ought to be applied to the tensors in said array.

ofAny(Class<V>, Shape, Object) - Static method in interface neureka.Tensor

Use this to construct and return a tensor of the specified type, shape and data object.

ofBigDecimals() - Static method in interface neureka.Nda

This is a shortcut method for Nda.of(BigDecimal.class) used to build Ndas storing BigDecimals.

ofBooleans() - Static method in interface neureka.Nda

This is a shortcut method for Nda.of(Boolean.class) used to build Ndas storing Booleans.

ofBytes() - Static method in interface neureka.Nda

This is a shortcut method for Nda.of(Byte.class) used to build Ndas storing Bytes.

ofBytes() - Static method in interface neureka.Tensor

This is a simple convenience method which is simply calling the Tensor.of(Class) method like so: of(Byte.class).

ofChars() - Static method in interface neureka.Nda

This is a shortcut method for Nda.of(Character.class) used to build Ndas storing Characters.

ofDoubles() - Static method in interface neureka.Nda

This is a shortcut method for Nda.of(Double.class) used to build Ndas storing Doubles.

ofDoubles() - Static method in interface neureka.Tensor

This is a simple convenience method which is simply calling the Tensor.of(Class) method like so: of(Double.class).

ofFloats() - Static method in interface neureka.Nda

This is a shortcut method for Nda.of(Float.class) used to build Ndas storing Floats.

ofFloats() - Static method in interface neureka.Tensor

This is a simple convenience method which is simply calling the Tensor.of(Class) method like so: of(Float.class).

offset() - Method in interface neureka.ndim.config.NDConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset(int) - Method in interface neureka.ndim.config.NDConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset() - Method in class neureka.ndim.config.types.permuted.Permuted1DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset(int) - Method in class neureka.ndim.config.types.permuted.Permuted1DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset() - Method in class neureka.ndim.config.types.permuted.Permuted2DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset(int) - Method in class neureka.ndim.config.types.permuted.Permuted2DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset() - Method in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset(int) - Method in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset() - Method in class neureka.ndim.config.types.permuted.PermutedNDConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset(int) - Method in class neureka.ndim.config.types.permuted.PermutedNDConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset() - Method in class neureka.ndim.config.types.simple.Simple0DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset(int) - Method in class neureka.ndim.config.types.simple.Simple0DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset() - Method in class neureka.ndim.config.types.simple.Simple1DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset(int) - Method in class neureka.ndim.config.types.simple.Simple1DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset() - Method in class neureka.ndim.config.types.simple.Simple2DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset(int) - Method in class neureka.ndim.config.types.simple.Simple2DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset() - Method in class neureka.ndim.config.types.simple.Simple3DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset(int) - Method in class neureka.ndim.config.types.simple.Simple3DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset() - Method in class neureka.ndim.config.types.simple.SimpleNDConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset(int) - Method in class neureka.ndim.config.types.simple.SimpleNDConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset() - Method in class neureka.ndim.config.types.sliced.Sliced0DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset(int) - Method in class neureka.ndim.config.types.sliced.Sliced0DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset() - Method in class neureka.ndim.config.types.sliced.Sliced1DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset(int) - Method in class neureka.ndim.config.types.sliced.Sliced1DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset() - Method in class neureka.ndim.config.types.sliced.Sliced2DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset(int) - Method in class neureka.ndim.config.types.sliced.Sliced2DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset() - Method in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset(int) - Method in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset() - Method in class neureka.ndim.config.types.sliced.SlicedNDConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset(int) - Method in class neureka.ndim.config.types.sliced.SlicedNDConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset() - Method in class neureka.ndim.config.types.views.SimpleReshapeView

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset(int) - Method in class neureka.ndim.config.types.views.SimpleReshapeView

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset() - Method in class neureka.ndim.config.types.views.virtual.VirtualNDConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset(int) - Method in class neureka.ndim.config.types.views.virtual.VirtualNDConfiguration

The offset is the position of a slice within the n-dimensional data array of its parent tensor.

offset() - Method in interface neureka.ndim.NDimensional

The offset is the position of a slice within the n-dimensional data array of its parent array.

ofGradient(Optimization<T>) - Static method in interface neureka.optimization.Optimizer

ofInts() - Static method in interface neureka.Nda

This is a shortcut method for Nda.of(Integer.class) used to build Ndas storing Integers.

ofInts() - Static method in interface neureka.Tensor

This is a simple convenience method which is simply calling the Tensor.of(Class) method like so: of(Integer.class).

ofLongs() - Static method in interface neureka.Nda

This is a shortcut method for Nda.of(Long.class) used to build Ndas storing Longs.

ofNumbers() - Static method in interface neureka.Nda

This is a shortcut method for Nda.of(Number.class) used to build Ndas storing Numbers.

ofObjects() - Static method in interface neureka.Nda

This is a shortcut method for Nda.of(Object.class) used to build Ndas storing Objects.

ofRandom(Class<V>, int...) - Static method in interface neureka.Tensor

This factory method produces a randomly populated tensor of the provided type and shape using a hard coded default seed.

ofShorts() - Static method in interface neureka.Nda

This is a shortcut method for Nda.of(Short.class) used to build Ndas storing Shorts.

ofShorts() - Static method in interface neureka.Tensor

This is a simple convenience method which is simply calling the Tensor.of(Class) method like so: of(Short.class).

ofStrings() - Static method in interface neureka.Nda

This is a shortcut method for Nda.of(String.class) used to build Ndas storing Strings.

OKAY - Static variable in interface neureka.backend.api.fun.SuitabilityPredicate

on(D) - Method in class neureka.backend.api.ExecutionCall.Builder

on(Device<? super V>) - Method in class neureka.fluent.building.NdaBuilder

on(Device<? super V>) - Method in interface neureka.fluent.building.states.WithShapeOrScalarOrVectorOnDevice

Use this to specify the type onto which the tensor should be stored.

OpenCLDevice - Class in neureka.devices.opencl

This class models OpenCL supporting accelerator hardware like GPUs or FPGAs for storing tensors and executing operations on them.

OpenCLDevice.Query - Class in neureka.devices.opencl

OpenCLDevice.Type - Enum in neureka.devices.opencl

OpenCLPlatform - Class in neureka.devices.opencl

This class models the OpenCL concept of platforms, which refer to device vendors / or vendor OpenCL runtime drivers.

OpenCLPlatform(cl_platform_id) - Constructor for class neureka.devices.opencl.OpenCLPlatform

Operation - Interface in neureka.backend.api

This interface is part of the backend API, and it embodies the top layer of the 3 tier backend architecture.

OperationBuilder - Class in neureka.backend.api.template.operations

This builder class builds instances of the Operation interface.

OperationBuilder() - Constructor for class neureka.backend.api.template.operations.OperationBuilder

OperationBuilder.Derivation - Interface in neureka.backend.api.template.operations

OperationBuilder.Stringifier - Interface in neureka.backend.api.template.operations

operationForF32(boolean, long, long) - Static method in class neureka.backend.main.operations.linear.internal.blas.GEMM

operationForF64(boolean, long, long) - Static method in class neureka.backend.main.operations.linear.internal.blas.GEMM

operationForI32(boolean, long, long) - Static method in class neureka.backend.main.operations.linear.internal.blas.IGEMM

operationForI64(boolean, long, long) - Static method in class neureka.backend.main.operations.linear.internal.blas.IGEMM

operationName() - Method in class neureka.backend.api.template.operations.AbstractOperation

Override this if you want your operation to have a string representation with a custom prefix which is something other than the simple class name!

operator(String) - Method in class neureka.backend.api.template.operations.OperationBuilder

Optimization<V> - Interface in neureka.optimization

optimize() - Method in class neureka.devices.opencl.utility.CLFunctionCompiler

optimize(Tensor<V>) - Method in class neureka.optimization.implementations.AdaGrad

optimize(Tensor<V>) - Method in class neureka.optimization.implementations.ADAM

optimize(Tensor<V>) - Method in class neureka.optimization.implementations.Momentum

optimize(Tensor<V>) - Method in class neureka.optimization.implementations.RMSProp

optimize(Tensor<V>) - Method in class neureka.optimization.implementations.SGD

optimize(Tensor<V>) - Method in interface neureka.optimization.Optimization

optimizedFunctionOf(Function, String) - Method in interface neureka.devices.Device

This method tries to allow this device to produce an optimized Function based on the provided function.

optimizedOperationOf(Function, String) - Method in interface neureka.devices.Device

This method tries to allow this device to produce an optimized Operation based on the provided function.

optimizedOperationOf(Function, String) - Method in class neureka.devices.file.FileDevice

optimizedOperationOf(Function, String) - Method in class neureka.devices.host.CPU

optimizedOperationOf(Function, String) - Method in class neureka.devices.opencl.OpenCLDevice

Optimizer<V> - Interface in neureka.optimization

Optimizers are tensor components which implement the Optimization (functional) interface applying various optimization algorithms to the gradients of tensors.

OptimizerFactory - Interface in neureka.optimization

orElse(T) - Method in interface neureka.backend.api.Call.Else

orElse(V) - Method in interface neureka.Nda.Item

Get the value at the targeted position or return the provided default value if the item does not exist.

orElseNull() - Method in interface neureka.Nda.Item

Get the value at the targeted position or return null if the item does not exist.

OS_MEMORY_PAGE_SIZE - Static variable in class neureka.devices.host.machine.Hardware

Practically all architectures/OS:s have a page size of 4k (one notable exception is Solaris/SPARC that have 8k)

owner() - Method in interface neureka.Data

owner() - Method in class neureka.devices.AbstractDeviceData

P

pad(int, String) - Static method in class neureka.view.NdaAsString.Util

pad(String, int) - Static method in class neureka.view.NdaAsString.Util

Parallelism - Enum in neureka.devices.host.concurrent

A set of standard levels of parallelism derived from the number of available cores and optionally capped by reserving a specified amount of memory per thread.

parallelism(IntSupplier) - Method in class neureka.devices.host.concurrent.WorkScheduler.Divider

PARALLELIZATION_THRESHOLD - Static variable in class neureka.devices.host.CPU

parse(Operation, int, boolean) - Method in class neureka.math.parsing.FunctionParser

parse(String, boolean) - Method in class neureka.math.parsing.FunctionParser

ParsedCLImplementation - Class in neureka.backend.main.implementations

ParsedCLImplementation(ImplementationFor<OpenCLDevice>, int, String, String, String, String, Function<KernelCode, KernelCode[]>) - Constructor for class neureka.backend.main.implementations.ParsedCLImplementation

parsedOperation(String, int) - Static method in class neureka.math.parsing.ParseUtil

ParseUtil - Class in neureka.math.parsing

Utility for parsing function expressions.

partialDerivative() - Method in interface neureka.autograd.ADAction

pass(Tensor<T>) - Method in class neureka.devices.opencl.KernelCaller

This method passes 1 argument to the kernel.

pass(int) - Method in class neureka.devices.opencl.KernelCaller

pass(int...) - Method in class neureka.devices.opencl.KernelCaller

Use this to pass an array of int values to the kernel.

pass(float...) - Method in class neureka.devices.opencl.KernelCaller

Use this to pass an array of float values to the kernel.

pass(double...) - Method in class neureka.devices.opencl.KernelCaller

pass(short...) - Method in class neureka.devices.opencl.KernelCaller

pass(long...) - Method in class neureka.devices.opencl.KernelCaller

pass(byte...) - Method in class neureka.devices.opencl.KernelCaller

pass(float) - Method in class neureka.devices.opencl.KernelCaller

pass(double) - Method in class neureka.devices.opencl.KernelCaller

pass(short) - Method in class neureka.devices.opencl.KernelCaller

pass(long) - Method in class neureka.devices.opencl.KernelCaller

pass(byte) - Method in class neureka.devices.opencl.KernelCaller

pass(Number) - Method in class neureka.devices.opencl.KernelCaller

passAllOf(Tensor<Number>) - Method in class neureka.devices.opencl.KernelCaller

This method passes 2 arguments to the kernel.

passConfOf(Tensor<Number>) - Method in class neureka.devices.opencl.KernelCaller

This method passes the ND-Configuration in the form of a flattened int array to the kernel.

passLocalFloats(long) - Method in class neureka.devices.opencl.KernelCaller

pendingCount() - Method in class neureka.autograd.JITProp

PERFECT - Static variable in interface neureka.backend.api.fun.SuitabilityPredicate

Permute - Class in neureka.backend.main.operations.other

Permute() - Constructor for class neureka.backend.main.operations.other.Permute

permute() - Method in class neureka.math.Functions

permute(int...) - Method in interface neureka.Nda

Returns a view of the original tensor input with its dimensions permuted.
Consider a 3-dimensional tensor x with shape (2×3×5), then calling x.permute(1, 0, 2) will return a 3-dimensional tensor of shape (3×2×5).

permute(int...) - Method in interface neureka.Tensor

Returns a view of the original tensor input with its dimensions permuted.
Consider a 3-dimensional tensor x with shape (2×3×5), then calling x.permute(1, 0, 2) will return a 3-dimensional tensor of shape (3×2×5).

Permuted1DConfiguration - Class in neureka.ndim.config.types.permuted

Permuted1DConfiguration(int, int, int) - Constructor for class neureka.ndim.config.types.permuted.Permuted1DConfiguration

Permuted2DCIterator - Class in neureka.ndim.iterator.types.permuted

Permuted2DCIterator(Permuted2DConfiguration) - Constructor for class neureka.ndim.iterator.types.permuted.Permuted2DCIterator

Permuted2DConfiguration - Class in neureka.ndim.config.types.permuted

Permuted2DConfiguration(int[], int[], int[]) - Constructor for class neureka.ndim.config.types.permuted.Permuted2DConfiguration

Permuted3DCIterator - Class in neureka.ndim.iterator.types.permuted

Permuted3DCIterator(Permuted3DConfiguration) - Constructor for class neureka.ndim.iterator.types.permuted.Permuted3DCIterator

Permuted3DConfiguration - Class in neureka.ndim.config.types.permuted

Permuted3DConfiguration(int[], int[], int[]) - Constructor for class neureka.ndim.config.types.permuted.Permuted3DConfiguration

PermutedNDConfiguration - Class in neureka.ndim.config.types.permuted

PermutedNDConfiguration(int[], int[], int[]) - Constructor for class neureka.ndim.config.types.permuted.PermutedNDConfiguration

plus() - Method in class neureka.math.Functions

plus(Tensor<V>) - Method in interface neureka.Tensor

This method will produce the addition of two tensors with the same rank (or two ranks which can be made compatible with padding ones), where the left operand is this Tensor instance and the right operand is the tensor passed to the method.

plus(V) - Method in interface neureka.Tensor

This method will create a new Tensor with the provided double scalar added to all elements of this Tensor.

plusAssign() - Method in class neureka.math.Functions

plusAssign(Tensor<T>) - Method in interface neureka.MutateTensor

Performs an addition of the passed tensor to this tensor.

pow() - Method in class neureka.math.Functions

powAssign() - Method in class neureka.math.Functions

Power - Class in neureka.backend.main.operations.operator

Power() - Constructor for class neureka.backend.main.operations.operator.Power

power(Tensor<V>) - Method in interface neureka.Tensor

This will produce the power of two tensors with the same rank (or two ranks which can be made compatible with padding ones), where the left operand is this Tensor instance and the right operand is the tensor passed to the method.

power(V) - Method in interface neureka.Tensor

Raises all items of this tensor to the power of the provided value.

PREDEFINED - Static variable in class neureka.devices.host.machine.Hardware

Should contain all available hardware in ascending "power" order.

prefVecWidthChar() - Method in class neureka.devices.opencl.OpenCLDevice

prefVecWidthDouble() - Method in class neureka.devices.opencl.OpenCLDevice

prefVecWidthFloat() - Method in class neureka.devices.opencl.OpenCLDevice

prefVecWidthInt() - Method in class neureka.devices.opencl.OpenCLDevice

prefVecWidthLong() - Method in class neureka.devices.opencl.OpenCLDevice

prefVecWidthShort() - Method in class neureka.devices.opencl.OpenCLDevice

prepare(ExecutionCall<? extends Device<?>>) - Method in interface neureka.backend.api.fun.ExecutionPreparation

prepare(ExecutionCall<? extends Device<?>>) - Method in class neureka.backend.api.template.algorithms.AbstractFunDeviceAlgorithm

Preparing refers to instantiating output tensors for the provided ExecutionCall.

prepare(ExecutionCall<? extends Device<?>>) - Method in class neureka.backend.api.template.algorithms.FallbackAlgorithm

prepareAndExecute(ExecutionCall<? extends Device<?>>, FinalExecutor) - Static method in class neureka.backend.api.template.algorithms.AbstractDeviceAlgorithm

process(T) - Method in class neureka.common.utility.Cache

produceNDC(boolean) - Method in interface neureka.ndim.NDConstructor

Product - Class in neureka.backend.main.operations.indexer

This type of operation belongs to the same species as the Summation operation.

Product() - Constructor for class neureka.backend.main.operations.indexer.Product

propertyOf(Tensor<?>) - Method in interface neureka.backend.api.Call.TensorProperty

providesGradient() - Method in class neureka.math.implementations.FunctionInput

providesGradient() - Method in class neureka.math.implementations.FunctionVariable

put(String, OpenCLDevice.cl_ad_hoc) - Method in class neureka.devices.opencl.KernelCache

put(Function) - Method in class neureka.math.FunctionCache

putAt(Map<?, Integer>, Nda<T>) - Method in interface neureka.MutateNda

This method enables assigning a provided nd-array to be a subset/slice of this nd-array! It takes a key which is used to configure a slice sharing the same underlying data as the original nd-array.

putAt(int[], T) - Method in interface neureka.MutateNda

Use this to put a single item at a particular position within this nd-array.

putAt(int, T) - Method in interface neureka.MutateNda

Individual entries for value items in this nd-array can be set via this method.

putAt(List<?>, Nda<T>) - Method in interface neureka.MutateNda

This method enables injecting slices of nd-array to be assigned into this nd-array! It takes a key of various types which is used to configure a slice nd-array sharing the same underlying data as the original nd-array.

putAt(List<?>, T) - Method in interface neureka.MutateNda

Use this to place a single item at a particular position within this nd-array!

putAt(Map<?, Integer>, Nda<T>) - Method in interface neureka.MutateTensor

This method enables assigning a provided nd-array to be a subset/slice of this nd-array! It takes a key which is used to configure a slice sharing the same underlying data as the original nd-array.

putAt(int[], T) - Method in interface neureka.MutateTensor

Use this to put a single item at a particular position within this nd-array.

putAt(int, T) - Method in interface neureka.MutateTensor

Individual entries for value items in this nd-array can be set via this method.

putAt(List<?>, Nda<T>) - Method in interface neureka.MutateTensor

This method enables injecting slices of nd-array to be assigned into this nd-array! It takes a key of various types which is used to configure a slice nd-array sharing the same underlying data as the original nd-array.

putAt(List<?>, T) - Method in interface neureka.MutateTensor

Use this to place a single item at a particular position within this nd-array!

Q

quad() - Method in class neureka.math.Functions

QUADRATIC - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

Quadratic - Class in neureka.backend.main.operations.functions

Quadratic() - Constructor for class neureka.backend.main.operations.functions.Quadratic

Query() - Constructor for class neureka.devices.opencl.OpenCLDevice.Query

query() - Static method in class neureka.devices.opencl.utility.DeviceQuery

The entry point of this program

R

random() - Method in class neureka.math.Functions

Randomization - Class in neureka.backend.main.operations.other

This Operation takes an optional user seed, the shape of its input tensor, and the indices of individual elements within said tensor to generate floats or doubles with a gaussian distribution where the mean is 0 and the standard deviation is 1.

Randomization() - Constructor for class neureka.backend.main.operations.other.Randomization

rank() - Method in interface neureka.ndim.config.NDConfiguration

This method returns the number of axis of a nd-array / Tensor which is equal to the length of the shape of an nd-array / Tensor.

rank() - Method in class neureka.ndim.config.types.permuted.Permuted1DConfiguration

This method returns the number of axis of a nd-array / Tensor which is equal to the length of the shape of an nd-array / Tensor.

rank() - Method in class neureka.ndim.config.types.permuted.Permuted2DConfiguration

This method returns the number of axis of a nd-array / Tensor which is equal to the length of the shape of an nd-array / Tensor.

rank() - Method in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

This method returns the number of axis of a nd-array / Tensor which is equal to the length of the shape of an nd-array / Tensor.

rank() - Method in class neureka.ndim.config.types.permuted.PermutedNDConfiguration

This method returns the number of axis of a nd-array / Tensor which is equal to the length of the shape of an nd-array / Tensor.

rank() - Method in class neureka.ndim.config.types.simple.Simple0DConfiguration

This method returns the number of axis of a nd-array / Tensor which is equal to the length of the shape of an nd-array / Tensor.

rank() - Method in class neureka.ndim.config.types.simple.Simple1DConfiguration

This method returns the number of axis of a nd-array / Tensor which is equal to the length of the shape of an nd-array / Tensor.

rank() - Method in class neureka.ndim.config.types.simple.Simple2DConfiguration

This method returns the number of axis of a nd-array / Tensor which is equal to the length of the shape of an nd-array / Tensor.

rank() - Method in class neureka.ndim.config.types.simple.Simple3DConfiguration

This method returns the number of axis of a nd-array / Tensor which is equal to the length of the shape of an nd-array / Tensor.

rank() - Method in class neureka.ndim.config.types.simple.SimpleNDConfiguration

This method returns the number of axis of a nd-array / Tensor which is equal to the length of the shape of an nd-array / Tensor.

rank() - Method in class neureka.ndim.config.types.sliced.Sliced0DConfiguration

This method returns the number of axis of a nd-array / Tensor which is equal to the length of the shape of an nd-array / Tensor.

rank() - Method in class neureka.ndim.config.types.sliced.Sliced1DConfiguration

This method returns the number of axis of a nd-array / Tensor which is equal to the length of the shape of an nd-array / Tensor.

rank() - Method in class neureka.ndim.config.types.sliced.Sliced2DConfiguration

This method returns the number of axis of a nd-array / Tensor which is equal to the length of the shape of an nd-array / Tensor.

rank() - Method in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

This method returns the number of axis of a nd-array / Tensor which is equal to the length of the shape of an nd-array / Tensor.

rank() - Method in class neureka.ndim.config.types.sliced.SlicedNDConfiguration

This method returns the number of axis of a nd-array / Tensor which is equal to the length of the shape of an nd-array / Tensor.

rank() - Method in class neureka.ndim.config.types.views.SimpleReshapeView

This method returns the number of axis of a nd-array / Tensor which is equal to the length of the shape of an nd-array / Tensor.

rank() - Method in class neureka.ndim.config.types.views.virtual.VirtualNDConfiguration

This method returns the number of axis of a nd-array / Tensor which is equal to the length of the shape of an nd-array / Tensor.

rank() - Method in interface neureka.ndim.iterator.NDIterator

rank() - Method in class neureka.ndim.iterator.types.sliced.SlicedNDIterator

rank() - Method in class neureka.ndim.iterator.types.virtual.VirtualNDIterator

rank() - Method in interface neureka.ndim.NDimensional

read(List<Object>, Function<Object, Object>) - Static method in class neureka.common.utility.ListReader

Reads the provided data and turns it into a ListReader.Result object, containing a flattened list of the data alongside its shape and data type.

readAll(boolean) - Method in interface neureka.devices.Device.Access

Use this to read the full data array of the accessed tensor.

readAndConvertForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.F32

readAndConvertForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.F32

readAndConvertForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.F64

readAndConvertForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.F64

readAndConvertForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.I16

readAndConvertForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.I16

readAndConvertForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.I32

readAndConvertForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.I32

readAndConvertForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.I64

readAndConvertForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.I64

readAndConvertForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.I8

readAndConvertForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.I8

readAndConvertForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.UI16

readAndConvertForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.UI16

readAndConvertForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.UI32

readAndConvertForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.UI32

readAndConvertForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.UI64

readAndConvertForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.UI64

readAndConvertForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.UI8

readAndConvertForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.UI8

readAndConvertForeignDataFrom(DataInput, int) - Method in interface neureka.dtype.NumericType

This method expects the provided stream to spit out bytes which can be read as holder type elements.

readAndConvertForeignDataFrom(Iterator<T>, int) - Method in interface neureka.dtype.NumericType

This method expects the provided iterator to return elements which can be read as holder type elements.

readArray(Class<A>, int, int) - Method in interface neureka.devices.Device.Access

Use this to read an array of items from the accessed tensor by specifying a start position of the chunk of data that should be read.

readAt(int) - Method in interface neureka.devices.Device.Access

Find a particular tensor item by providing its location.

readForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.F32

readForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.F32

readForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.F64

readForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.F64

readForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.I16

readForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.I16

readForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.I32

readForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.I32

readForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.I64

readForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.I64

readForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.I8

readForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.I8

readForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.UI16

readForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.UI16

readForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.UI32

readForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.UI32

readForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.UI64

readForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.UI64

readForeignDataFrom(DataInput, int) - Method in class neureka.dtype.custom.UI8

readForeignDataFrom(Iterator<T>, int) - Method in class neureka.dtype.custom.UI8

readForeignDataFrom(DataInput, int) - Method in interface neureka.dtype.NumericType

This method expects the provided stream to spit out bytes which can be read as target type elements.

readForeignDataFrom(Iterator<T>, int) - Method in interface neureka.dtype.NumericType

This method expects the provided iterator to return elements which can be read as holder type elements.

readResource(String) - Method in class neureka.Neureka.Utility

Helper method which reads the file with the given name and returns the contents of this file as a String.

rearrange(int[], int[], int[]) - Method in enum neureka.ndim.config.NDConfiguration.Layout

rearrange(int[], int[]) - Static method in class neureka.ndim.config.NDConfiguration.Utility

rearrangeInputs(int...) - Method in class neureka.backend.api.Call

ReceiveForDevice<D extends Device<?>> - Interface in neureka.backend.api.ini

ReceiveForOperation<D extends Device<?>> - Interface in neureka.backend.api.ini

recompile() - Method in class neureka.devices.opencl.OpenCLPlatform

ReferenceCounter - Class in neureka.devices

ReferenceCounter(Consumer<ReferenceCounter.ChangeEvent>) - Constructor for class neureka.devices.ReferenceCounter

ReferenceCounter.ChangeEvent - Class in neureka.devices

ReferenceCounter.ChangeType - Enum in neureka.devices

register(Object, Runnable) - Method in interface neureka.devices.DeviceCleaner

Relation<V> - Class in neureka.framing

This class is an important tensor component responsible for managing the relationships between slices and the tensors from which they have been derived.

ReLayout - Class in neureka.backend.main.operations.other

ReLayout() - Constructor for class neureka.backend.main.operations.other.ReLayout

relayout() - Method in class neureka.math.Functions

RELU - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

ReLU - Class in neureka.backend.main.operations.functions

ReLU() - Constructor for class neureka.backend.main.operations.functions.ReLU

relu() - Method in class neureka.math.Functions

relu() - Method in interface neureka.Tensor

This method is a functionally identical to the following alternatives:

rem(int) - Method in interface neureka.Tensor

This method is synonymous to the Tensor.mod(int) method.

remove(Class<T>) - Method in class neureka.common.composition.AbstractComponentOwner

This method removes a component identified by the passed Class instance if found in the stored component collection.

remove(Class<T>) - Method in interface neureka.common.composition.ComponentOwner

Use this to remove a component of the specified component type class.

removeChild(Tensor<V>) - Method in class neureka.framing.Relation

replace(Object) - Method in class neureka.framing.fluent.AxisFrame

Replace<ValueType,ReplacementType,ReturnType> - Interface in neureka.framing.fluent

replace(ValueType) - Method in interface neureka.framing.fluent.Replace

replacer(Replace<Object, Object, NDFrame<ValueType>>) - Method in class neureka.framing.fluent.AxisFrame.Builder

representing(Nda<?>) - Static method in class neureka.view.NdaAsString

A builder providing multiple different configuration options for building a NdaAsString instance in a fluent way.

reset() - Method in class neureka.backend.api.BackendContext

reset() - Method in interface neureka.backend.api.BackendExtension

This will indirectly be called through the Neureka.reset() method, which is responsible for resetting the library settings.

reset() - Method in class neureka.backend.ocl.CLBackend

reset() - Method in class neureka.backend.ocl.CLSettings

reset() - Method in class neureka.Neureka

This method will try to reload the "library_settings.groovy" script which will re-configure the library wide Neureka.Settings instance nested inside Neureka.

Reshape - Class in neureka.backend.main.operations.other

Reshape() - Constructor for class neureka.backend.main.operations.other.Reshape

reshape() - Method in class neureka.math.Functions

reshape(int...) - Method in interface neureka.Nda

Returns a nd-array with the same data and number of elements as this nd-array, but with the specified shape.

reshape(int...) - Method in interface neureka.Tensor

Returns a nd-array with the same data and number of elements as this nd-array, but with the specified shape.

resolve() - Method in class neureka.fluent.slicing.AxisSliceBuilder

restore(Tensor<Object>) - Method in class neureka.devices.file.FileDevice

restore(Tensor<Object>) - Method in class neureka.devices.host.CPU

restore(Tensor<Number>) - Method in class neureka.devices.opencl.OpenCLDevice

This method assumes that the passed tensor is stored on this device instance.

restore(Tensor<V>) - Method in interface neureka.devices.Storage

Result - Class in neureka.backend.api

An immutable wrapper for a tensor as a result of anb Execution as well as an ADActionSupplier for providing auto-differentiation support.

RMSProp<V extends java.lang.Number> - Class in neureka.optimization.implementations

Root Mean Squared Propagation, or RMSProp, is an extension of gradient descent and the AdaGrad version of gradient descent that uses a decaying average of partial gradients in the adaptation of the step size for each parameter.

RMSProp - Static variable in interface neureka.optimization.Optimizer

RMSPropFactory - Class in neureka.optimization.implementations

RMSPropFactory() - Constructor for class neureka.optimization.implementations.RMSPropFactory

rqsGradient() - Method in interface neureka.Tensor

This flag will indirectly trigger the activation of the autograd / auto-differentiation system of this library! If the flag is set to 'true' and the tensor is used for computation then it will also receive gradients when the Tensor.backward() method is being called on any descendant tensor within the computation graph.

run(Runnable) - Method in class neureka.backend.api.BackendContext.Runner

Use this method to supply a lambda which will be executed in the BackendContext which produced this very BackendContext.Runner instance.

run(ExecutionCall<D>) - Method in interface neureka.backend.api.ImplementationFor

This method is the entrypoint for a concrete implementation of the algorithm to which instances of this interface belong and the device on which this is implemented.

run(ExecutionCall<D>) - Method in class neureka.backend.api.template.implementations.AbstractImplementationFor

run(ExecutionCall<OpenCLDevice>) - Method in class neureka.backend.main.implementations.broadcast.CLScalarBroadcastAddition

run(ExecutionCall<OpenCLDevice>) - Method in class neureka.backend.main.implementations.broadcast.CLScalarBroadcastDivision

run(ExecutionCall<OpenCLDevice>) - Method in class neureka.backend.main.implementations.broadcast.CLScalarBroadcastIdentity

run(ExecutionCall<OpenCLDevice>) - Method in class neureka.backend.main.implementations.broadcast.CLScalarBroadcastModulo

run(ExecutionCall<OpenCLDevice>) - Method in class neureka.backend.main.implementations.broadcast.CLScalarBroadcastMultiplication

run(ExecutionCall<OpenCLDevice>) - Method in class neureka.backend.main.implementations.broadcast.CLScalarBroadcastPower

run(ExecutionCall<OpenCLDevice>) - Method in class neureka.backend.main.implementations.broadcast.CLScalarBroadcastSubtraction

run(ExecutionCall<CPU>) - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcast

run(ExecutionCall<CPU>) - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcast

run(ExecutionCall<CPU>) - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastAddition

run(ExecutionCall<CPU>) - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastMultiplication

run(ExecutionCall<CPU>) - Method in class neureka.backend.main.implementations.convolution.AbstractCPUConvolution

run(ExecutionCall<OpenCLDevice>) - Method in class neureka.backend.main.implementations.elementwise.CLRandomization

run(ExecutionCall<CPU>) - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWise

run(ExecutionCall<CPU>) - Method in class neureka.backend.main.implementations.elementwise.CPUElementwiseAssignFun

run(ExecutionCall<CPU>) - Method in class neureka.backend.main.implementations.elementwise.CPUElementwiseFunction

run(ExecutionCall<CPU>) - Method in class neureka.backend.main.implementations.elementwise.CPURandomization

run(ExecutionCall<OpenCLDevice>) - Method in class neureka.backend.main.implementations.linear.CLDot

run(ExecutionCall<CPU>) - Method in class neureka.backend.main.implementations.linear.CPUDot

run(ExecutionCall<CPU>) - Method in class neureka.backend.main.implementations.matmul.CPUMatMul

run(ExecutionCall<OpenCLDevice>) - Method in class neureka.backend.main.implementations.scalar.CLScalarFunction

run(ExecutionCall<CPU>) - Method in class neureka.backend.main.implementations.scalar.CPUScalarBroadcastFunction

run(ExecutionCall<CPU>) - Method in class neureka.backend.main.implementations.scalar.CPUScalarFunction

run(ExecutionCall<OpenCLDevice>) - Method in class neureka.backend.main.operations.linear.internal.opencl.CLGEMM

run(ExecutionCall<OpenCLDevice>) - Method in class neureka.backend.main.operations.linear.internal.opencl.CLReduce

run(ExecutionCall<OpenCLDevice>) - Method in class neureka.backend.main.operations.linear.internal.opencl.CLSum

run(Tensor<Float>, OpenCLDevice) - Static method in class neureka.backend.main.operations.linear.internal.opencl.CLSum

This method compiles and executes the kernel that will return the sum of the elements in the in tensor.

run(ExecutionCall<CPU>) - Method in class neureka.backend.main.operations.other.internal.CPUReduce

run(ExecutionCall<CPU>) - Method in class neureka.backend.main.operations.other.internal.CPUSum

runAndGet(Supplier<T>) - Method in class neureka.backend.api.BackendContext.Runner

Use this method to supply a lambda which will be executed in the BackendContext which produced this very BackendContext.Runner instance.

runner() - Method in class neureka.backend.api.BackendContext

A BackendContext.Runner wraps both the called context as well as the context of the caller in order to perform temporary context switching during the execution of lambdas passed to the BackendContext.Runner.

running(Operation) - Method in class neureka.backend.api.ExecutionCall.Builder

S

scalar(V) - Method in class neureka.fluent.building.NdaBuilder

scalar(V) - Method in interface neureka.fluent.building.states.WithShapeOrScalarOrVector

This method created and return a scalar Tensor instance which wraps the provided value.

scalar(V) - Method in interface neureka.fluent.building.states.WithShapeOrScalarOrVectorTensor

This method created and return a scalar Tensor instance which wraps the provided value.

ScalarAbsolute - Class in neureka.backend.main.implementations.fun

ScalarAbsolute() - Constructor for class neureka.backend.main.implementations.fun.ScalarAbsolute

ScalarAlgorithm - Class in neureka.backend.main.algorithms

ScalarAlgorithm() - Constructor for class neureka.backend.main.algorithms.ScalarAlgorithm

ScalarBroadcast - Class in neureka.backend.main.algorithms

ScalarBroadcast(ScalarFun) - Constructor for class neureka.backend.main.algorithms.ScalarBroadcast

ScalarCbrt - Class in neureka.backend.main.implementations.fun

ScalarCbrt() - Constructor for class neureka.backend.main.implementations.fun.ScalarCbrt

ScalarCosinus - Class in neureka.backend.main.implementations.fun

ScalarCosinus() - Constructor for class neureka.backend.main.implementations.fun.ScalarCosinus

ScalarExp - Class in neureka.backend.main.implementations.fun

ScalarExp() - Constructor for class neureka.backend.main.implementations.fun.ScalarExp

ScalarFun - Interface in neureka.backend.main.implementations.fun.api

ScalarGaSU - Class in neureka.backend.main.implementations.fun

The Self Gated ScalarSoftsign Unit is based on the ScalarSoftsign function (a computationally cheap non-exponential quasi ScalarTanh) making it a polynomially based version of the ScalarGaTU function which is itself based on the ScalarTanh function.

ScalarGaSU() - Constructor for class neureka.backend.main.implementations.fun.ScalarGaSU

ScalarGaTU - Class in neureka.backend.main.implementations.fun

The Self Gated ScalarTanh Unit is based on the ScalarTanh making it an exponentiation based version of the ScalarGaSU function which is itself based on the ScalarSoftsign function (a computationally cheap non-exponential quasi ScalarTanh).

ScalarGaTU() - Constructor for class neureka.backend.main.implementations.fun.ScalarGaTU

ScalarGaussian - Class in neureka.backend.main.implementations.fun

ScalarGaussian() - Constructor for class neureka.backend.main.implementations.fun.ScalarGaussian

ScalarGaussianFast - Class in neureka.backend.main.implementations.fun

ScalarGaussianFast() - Constructor for class neureka.backend.main.implementations.fun.ScalarGaussianFast

ScalarGeLU - Class in neureka.backend.main.implementations.fun

The GELU activation function is based on the standard Gaussian cumulative distribution function and is defined as x Φ( x ) and implemented as x * sigmoid(x * 1.702).

ScalarGeLU() - Constructor for class neureka.backend.main.implementations.fun.ScalarGeLU

ScalarIdentity - Class in neureka.backend.main.implementations.fun

ScalarIdentity() - Constructor for class neureka.backend.main.implementations.fun.ScalarIdentity

ScalarLog10 - Class in neureka.backend.main.implementations.fun

ScalarLog10() - Constructor for class neureka.backend.main.implementations.fun.ScalarLog10

ScalarLogarithm - Class in neureka.backend.main.implementations.fun

ScalarLogarithm() - Constructor for class neureka.backend.main.implementations.fun.ScalarLogarithm

ScalarQuadratic - Class in neureka.backend.main.implementations.fun

ScalarQuadratic() - Constructor for class neureka.backend.main.implementations.fun.ScalarQuadratic

ScalarReLU - Class in neureka.backend.main.implementations.fun

ScalarReLU() - Constructor for class neureka.backend.main.implementations.fun.ScalarReLU

ScalarSeLU - Class in neureka.backend.main.implementations.fun

The Scaled Exponential Linear Unit, or SELU, is an activation function that induces self-normalizing properties.

ScalarSeLU() - Constructor for class neureka.backend.main.implementations.fun.ScalarSeLU

ScalarSigmoid - Class in neureka.backend.main.implementations.fun

ScalarSigmoid() - Constructor for class neureka.backend.main.implementations.fun.ScalarSigmoid

ScalarSiLU - Class in neureka.backend.main.implementations.fun

The SiLu activation function, also known as the swish function, is defined as x * sigmoid(x).

ScalarSiLU() - Constructor for class neureka.backend.main.implementations.fun.ScalarSiLU

ScalarSinus - Class in neureka.backend.main.implementations.fun

ScalarSinus() - Constructor for class neureka.backend.main.implementations.fun.ScalarSinus

ScalarSoftplus - Class in neureka.backend.main.implementations.fun

SoftPlus is a smooth approximation to the ReLU function and can be used to constrain the output of a machine to always be positive.

ScalarSoftplus() - Constructor for class neureka.backend.main.implementations.fun.ScalarSoftplus

ScalarSoftsign - Class in neureka.backend.main.implementations.fun

The softsign function, defined as x / ( 1 + Math.abs( x ) ), is a computationally cheap 0 centered activation function which rescales the inputs between -1 and 1, very much like the ScalarTanh function.

ScalarSoftsign() - Constructor for class neureka.backend.main.implementations.fun.ScalarSoftsign

ScalarSqrt - Class in neureka.backend.main.implementations.fun

ScalarSqrt() - Constructor for class neureka.backend.main.implementations.fun.ScalarSqrt

ScalarSumAlgorithm - Class in neureka.backend.main.algorithms

ScalarSumAlgorithm() - Constructor for class neureka.backend.main.algorithms.ScalarSumAlgorithm

ScalarTanh - Class in neureka.backend.main.implementations.fun

ScalarTanh() - Constructor for class neureka.backend.main.implementations.fun.ScalarTanh

ScalarTanhFast - Class in neureka.backend.main.implementations.fun

ScalarTanhFast() - Constructor for class neureka.backend.main.implementations.fun.ScalarTanhFast

SELU - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

selu(double) - Static method in class neureka.backend.main.implementations.fun.ScalarSeLU

SeLU - Class in neureka.backend.main.operations.functions

The Scaled Exponential Linear Unit, or SELU, is an activation functions that induce self-normalizing properties.

SeLU() - Constructor for class neureka.backend.main.operations.functions.SeLU

selu() - Method in class neureka.math.Functions

The Scaled Exponential Linear Unit, or SELU, is an activation functions that induce self-normalizing properties.

sequential(int, CPU.RangeWorkload) - Method in class neureka.devices.host.CPU.JVMExecutor

This method will simply execute the provided CPU.RangeWorkload lambda sequentially with 0 as the start index and workloadSize as the exclusive range.

set(BackendExtension) - Method in class neureka.backend.api.BackendContext

Registers the provided BackendExtension instance which can then be accessed via BackendContext.find(Class).

set(Class<? extends Operation>, Class<? extends A>, Function<LoadingContext, ImplementationFor<D>>) - Method in interface neureka.backend.api.ini.ReceiveForDevice

set(Class<? extends DeviceAlgorithm>, Function<LoadingContext, ImplementationFor<D>>) - Method in interface neureka.backend.api.ini.ReceiveForOperation

set(T) - Method in class neureka.common.composition.AbstractComponentOwner

This methods stores the passed component inside the component collection of this class...

set(T) - Method in interface neureka.common.composition.ComponentOwner

Use this to set a component.

Set<V> - Interface in neureka.framing.fluent

set(int) - Method in interface neureka.framing.fluent.Set

set(V) - Method in interface neureka.MutateNda.Item

Set the value at the targeted position.

set(int[], T) - Method in interface neureka.MutateNda

Use this to place a single item at a particular position within this nd-array!

set(int, int, T) - Method in interface neureka.MutateNda

set(int, int, int, T) - Method in interface neureka.MutateNda

set(int, T) - Method in interface neureka.MutateNda

Individual entries for value items in this nd-array can be set via this method.

set(int[], T) - Method in interface neureka.MutateTensor

Use this to place a single item at a particular position within this nd-array!

set(int, int, T) - Method in interface neureka.MutateTensor

set(int, int, int, T) - Method in interface neureka.MutateTensor

set(int, T) - Method in interface neureka.MutateTensor

Individual entries for value items in this nd-array can be set via this method.

set(int, int) - Method in interface neureka.ndim.iterator.NDIterator

set(int[]) - Method in interface neureka.ndim.iterator.NDIterator

set(int, int) - Method in class neureka.ndim.iterator.types.permuted.Permuted2DCIterator

set(int[]) - Method in class neureka.ndim.iterator.types.permuted.Permuted2DCIterator

set(int, int) - Method in class neureka.ndim.iterator.types.permuted.Permuted3DCIterator

set(int[]) - Method in class neureka.ndim.iterator.types.permuted.Permuted3DCIterator

set(int, int) - Method in class neureka.ndim.iterator.types.simple.Simple1DCIterator

set(int[]) - Method in class neureka.ndim.iterator.types.simple.Simple1DCIterator

set(int, int) - Method in class neureka.ndim.iterator.types.simple.Simple2DCIterator

set(int[]) - Method in class neureka.ndim.iterator.types.simple.Simple2DCIterator

set(int, int) - Method in class neureka.ndim.iterator.types.simple.Simple3DCIterator

set(int[]) - Method in class neureka.ndim.iterator.types.simple.Simple3DCIterator

set(int, int) - Method in class neureka.ndim.iterator.types.sliced.Sliced1DCIterator

set(int[]) - Method in class neureka.ndim.iterator.types.sliced.Sliced1DCIterator

set(int, int) - Method in class neureka.ndim.iterator.types.sliced.Sliced2DCIterator

set(int[]) - Method in class neureka.ndim.iterator.types.sliced.Sliced2DCIterator

set(int, int) - Method in class neureka.ndim.iterator.types.sliced.Sliced3DCIterator

set(int[]) - Method in class neureka.ndim.iterator.types.sliced.Sliced3DCIterator

set(int, int) - Method in class neureka.ndim.iterator.types.sliced.SlicedNDIterator

set(int[]) - Method in class neureka.ndim.iterator.types.sliced.SlicedNDIterator

set(int, int) - Method in class neureka.ndim.iterator.types.virtual.VirtualNDIterator

set(int[]) - Method in class neureka.ndim.iterator.types.virtual.VirtualNDIterator

set(Neureka) - Static method in class neureka.Neureka

Neureka is a thread local singleton.

set(OptimizerFactory) - Method in interface neureka.Tensor

Configures an Optimizer for this tensor based on the given OptimizerFactory which will be used to create a new Optimizer instance specific to this tensor.

setAlgorithm(Class<T>, T) - Method in interface neureka.backend.api.Operation

Operation implementations embody a component system hosting unique Algorithm instances.

setAlgorithm(T) - Method in interface neureka.backend.api.Operation

setAlgorithm(Class<T>, T) - Method in class neureka.backend.api.template.operations.AbstractOperation

Operation implementations embody a component system hosting unique Algorithm instances.

setAutoConvertToFloat(boolean) - Method in class neureka.backend.ocl.CLSettings

setAutogradModeFor(ADSupportPredicate) - Method in class neureka.backend.api.template.algorithms.AbstractFunAlgorithm

A ADSupportPredicate lambda checks what kind of auto differentiation mode an Algorithm supports for a given ExecutionCall.

setAutogradModeFor(ADSupportPredicate) - Method in class neureka.backend.api.template.algorithms.AbstractFunDeviceAlgorithm

A ADSupportPredicate lambda checks what kind of auto differentiation mode an Algorithm supports for a given ExecutionCall.

setBackend(BackendContext) - Method in class neureka.Neureka

Use this method to attach a backend context (for operations) to this thread local library context.

setCallPreparation(ExecutionPreparation) - Method in class neureka.backend.api.template.algorithms.AbstractFunDeviceAlgorithm

An Algorithm will produce a Result when executing an ExecutionCall.

setCellSize(int) - Method in class neureka.view.NDPrintSettings

A cell size refers to the number of characters reserved to the String representation of a single element.

setData(Data<T>) - Method in interface neureka.MutateTensor

At the heart of every tensor is the Data object, which holds the actual data array, a sequence of values of the same type.

setDataAt(int, T) - Method in interface neureka.MutateTensor

A tensor ought to have some way to selectively modify its underlying data array.

setDefaultDataTypeClass(Class<?>) - Method in class neureka.Neureka.Settings.DType

The default data type is not relevant most of the time.

setDerivation(OperationBuilder.Derivation) - Method in class neureka.backend.api.template.operations.OperationBuilder

setExecution(Execution) - Method in class neureka.backend.api.template.algorithms.AbstractFunAlgorithm

setExecution(Execution) - Method in class neureka.backend.api.template.algorithms.AbstractFunDeviceAlgorithm

setGradientApplyRequested(boolean) - Method in interface neureka.Tensor

This flag works alongside two autograd features which can be enabled inside the library settings.

setHasDerivatives(boolean) - Method in class neureka.view.NDPrintSettings

setHasGradient(boolean) - Method in class neureka.view.NDPrintSettings

setHasRecursiveGraph(boolean) - Method in class neureka.view.NDPrintSettings

setHasShape(boolean) - Method in class neureka.view.NDPrintSettings

setHasSlimNumbers(boolean) - Method in class neureka.view.NDPrintSettings

setHasValue(boolean) - Method in class neureka.view.NDPrintSettings

setImplementationFor(Class<D>, I) - Method in interface neureka.backend.api.DeviceAlgorithm

Implementations of the DeviceAlgorithm interface ought to express a compositional design pattern.

setImplementationFor(Class<D>, E) - Method in class neureka.backend.api.template.algorithms.AbstractDeviceAlgorithm

setIndent(String) - Method in class neureka.view.NDPrintSettings

setIndex(int) - Method in interface neureka.framing.fluent.AxisFrame.Set

setIsApplyingGradientWhenRequested(boolean) - Method in class neureka.Neureka.Settings.AutoGrad

Gradients will only be applied if requested.

setIsApplyingGradientWhenTensorIsUsed(boolean) - Method in class neureka.Neureka.Settings.AutoGrad

Gradients will automatically be applied (or JITed) to tensors as soon as they are being used for calculation (GraphNode instantiation).

setIsAutoConvertingExternalDataToJVMTypes(boolean) - Method in class neureka.Neureka.Settings.DType

This flag will determine if foreign data types will be converted into the next best fit (in terms of bits) or if it should be converted into something that does not mess with the representation of the data.

setIsCellBound(boolean) - Method in class neureka.view.NDPrintSettings

setIsDeletingIntermediateTensors(boolean) - Method in class neureka.Neureka.Settings.Debug

Function instances will produce hidden intermediate results when executing an array of inputs.

setIsIntermediate(boolean) - Method in interface neureka.MutateTensor

Intermediate tensors are internal non-user tensors which may be eligible for deletion when further consumed by a Function.

setIsKeepingDerivativeTargetPayloads(boolean) - Method in class neureka.Neureka.Settings.Debug

Every derivative is calculated with respect to some graph node.

setIsLegacy(boolean) - Method in class neureka.view.NDPrintSettings

This flag determines the usage of bracket types, where "[1x3]:(1, 2, 3)" would be the legacy version of "(1x3):[1, 2, 3]".

setIsLocked(boolean) - Method in class neureka.Neureka.Settings

Can be used to lock or unlock the settings of the current thread-local Neureka instance.

setIsMultiline(boolean) - Method in class neureka.view.NDPrintSettings

setIsOnlyUsingDefaultNDConfiguration(boolean) - Method in class neureka.Neureka.Settings.NDim

Setting this flag determines which NDConfiguration implementations should be used for nd-arrays/tensors.

setIsPreventingInlineOperations(boolean) - Method in class neureka.Neureka.Settings.AutoGrad

Inline operations are operations where the data of a tensor passed into an operation is being modified.

setIsRetainingPendingErrorForJITProp(boolean) - Method in class neureka.Neureka.Settings.AutoGrad

This flag enables an optimization technique which only propagates error values to gradients if needed by a tensor (the tensor is used again) and otherwise accumulate them at divergent differentiation paths within the computation graph.
If the flag is set to true
then error values will accumulate at such junction nodes.

setIsScientific(boolean) - Method in class neureka.view.NDPrintSettings

setIsSuitableFor(SuitabilityPredicate) - Method in class neureka.backend.api.template.algorithms.AbstractFunAlgorithm

The SuitabilityPredicate received by this method checks if a given instance of an ExecutionCall is suitable to be executed in ImplementationFor instances residing in this Algorithm as components.

setIsSuitableFor(SuitabilityPredicate) - Method in class neureka.backend.api.template.algorithms.AbstractFunDeviceAlgorithm

The SuitabilityPredicate received by this method checks if a given instance of an ExecutionCall is suitable to be executed in ImplementationFor instances residing in this Algorithm as components.

setIsVirtual(boolean) - Method in interface neureka.MutateTensor

Virtualizing is the opposite to actualizing a tensor.

setItemAt(int, T) - Method in interface neureka.MutateNda

An NDArray implementation ought to have some way to selectively modify its underlying value.

setItemAt(int, T) - Method in interface neureka.MutateTensor

An NDArray implementation ought to have some way to selectively modify its underlying value.

setItems(Object) - Method in interface neureka.MutateNda

This method will receive an object an try to interpret it or its contents to be set as value for this nd-array.

setItems(Object) - Method in interface neureka.MutateTensor

This method will receive an object an try to interpret it or its contents to be set as value for this nd-array.

setNDConf(NDConfiguration) - Method in interface neureka.MutateTensor

This method sets the NDConfiguration of this NDArray.

setPostfix(String) - Method in class neureka.view.NDPrintSettings

setPrefix(String) - Method in class neureka.view.NDPrintSettings

setRowLimit(int) - Method in class neureka.view.NDPrintSettings

Very large tensors with a rank larger than 1 might take a lot of vertical space when converted to a String.

setRqsGradient(boolean) - Method in interface neureka.Tensor

Setting this flag to true will tell the autograd system to accumulate gradients at this tensor.

setSupplyADActionFor(ADActionSupplier) - Method in class neureka.backend.api.template.algorithms.AbstractFunDeviceAlgorithm

This method receives a ADActionSupplier which will supply ADAction instances which can perform backward and forward auto differentiation.

setter(At<Object, AxisFrame.Set<ValueType>>) - Method in class neureka.framing.fluent.AxisFrame.Builder

settings() - Method in class neureka.Neureka

settings(Object) - Method in class neureka.Neureka

This allows you to configure Neureka using a Groovy DSL.

SettingsLoader - Class in neureka.common.utility

This class is a helper class for Neureka instances (Thread local singletons).

SGD<V> - Class in neureka.optimization.implementations

Stochastic Gradient Descent is an iterative optimization technique that uses the gradient of a weight variable to adjust said variable, in order to reduce the error used to calculate said gradient.

SGD - Static variable in interface neureka.optimization.Optimizer

SGDFactory - Class in neureka.optimization.implementations

SGDFactory() - Constructor for class neureka.optimization.implementations.SGDFactory

shallowClone() - Method in interface neureka.Tensor

shallowCopy() - Method in interface neureka.Nda

This creates a copy where the underlying data is still the same.

shallowCopy() - Method in interface neureka.Tensor

This creates a copy where the underlying data is still the same.

shape() - Method in interface neureka.ndim.config.NDConfiguration

This method returns an array of axis sizes.

shape(int) - Method in interface neureka.ndim.config.NDConfiguration

This method receives an axis index and return the size of the axis.

shape() - Method in class neureka.ndim.config.types.permuted.Permuted1DConfiguration

This method returns an array of axis sizes.

shape(int) - Method in class neureka.ndim.config.types.permuted.Permuted1DConfiguration

This method receives an axis index and return the size of the axis.

shape() - Method in class neureka.ndim.config.types.permuted.Permuted2DConfiguration

This method returns an array of axis sizes.

shape(int) - Method in class neureka.ndim.config.types.permuted.Permuted2DConfiguration

This method receives an axis index and return the size of the axis.

shape() - Method in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

This method returns an array of axis sizes.

shape(int) - Method in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

This method receives an axis index and return the size of the axis.

shape() - Method in class neureka.ndim.config.types.permuted.PermutedNDConfiguration

This method returns an array of axis sizes.

shape(int) - Method in class neureka.ndim.config.types.permuted.PermutedNDConfiguration

This method receives an axis index and return the size of the axis.

shape() - Method in class neureka.ndim.config.types.simple.Simple0DConfiguration

This method returns an array of axis sizes.

shape(int) - Method in class neureka.ndim.config.types.simple.Simple0DConfiguration

This method receives an axis index and return the size of the axis.

shape() - Method in class neureka.ndim.config.types.simple.Simple1DConfiguration

This method returns an array of axis sizes.

shape(int) - Method in class neureka.ndim.config.types.simple.Simple1DConfiguration

This method receives an axis index and return the size of the axis.

shape() - Method in class neureka.ndim.config.types.simple.Simple2DConfiguration

This method returns an array of axis sizes.

shape(int) - Method in class neureka.ndim.config.types.simple.Simple2DConfiguration

This method receives an axis index and return the size of the axis.

shape() - Method in class neureka.ndim.config.types.simple.Simple3DConfiguration

This method returns an array of axis sizes.

shape(int) - Method in class neureka.ndim.config.types.simple.Simple3DConfiguration

This method receives an axis index and return the size of the axis.

shape() - Method in class neureka.ndim.config.types.simple.SimpleNDConfiguration

This method returns an array of axis sizes.

shape(int) - Method in class neureka.ndim.config.types.simple.SimpleNDConfiguration

This method receives an axis index and return the size of the axis.

shape() - Method in class neureka.ndim.config.types.sliced.Sliced0DConfiguration

This method returns an array of axis sizes.

shape(int) - Method in class neureka.ndim.config.types.sliced.Sliced0DConfiguration

This method receives an axis index and return the size of the axis.

shape() - Method in class neureka.ndim.config.types.sliced.Sliced1DConfiguration

This method returns an array of axis sizes.

shape(int) - Method in class neureka.ndim.config.types.sliced.Sliced1DConfiguration

This method receives an axis index and return the size of the axis.

shape() - Method in class neureka.ndim.config.types.sliced.Sliced2DConfiguration

This method returns an array of axis sizes.

shape(int) - Method in class neureka.ndim.config.types.sliced.Sliced2DConfiguration

This method receives an axis index and return the size of the axis.

shape() - Method in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

This method returns an array of axis sizes.

shape(int) - Method in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

This method receives an axis index and return the size of the axis.

shape() - Method in class neureka.ndim.config.types.sliced.SlicedNDConfiguration

This method returns an array of axis sizes.

shape(int) - Method in class neureka.ndim.config.types.sliced.SlicedNDConfiguration

This method receives an axis index and return the size of the axis.

shape() - Method in class neureka.ndim.config.types.views.SimpleReshapeView

This method returns an array of axis sizes.

shape(int) - Method in class neureka.ndim.config.types.views.SimpleReshapeView

This method receives an axis index and return the size of the axis.

shape() - Method in class neureka.ndim.config.types.views.virtual.VirtualNDConfiguration

This method returns an array of axis sizes.

shape(int) - Method in class neureka.ndim.config.types.views.virtual.VirtualNDConfiguration

This method receives an axis index and return the size of the axis.

shape(int) - Method in interface neureka.ndim.iterator.NDIterator

shape() - Method in interface neureka.ndim.iterator.NDIterator

shape(int) - Method in class neureka.ndim.iterator.types.sliced.SlicedNDIterator

shape() - Method in class neureka.ndim.iterator.types.sliced.SlicedNDIterator

shape(int) - Method in class neureka.ndim.iterator.types.virtual.VirtualNDIterator

shape() - Method in class neureka.ndim.iterator.types.virtual.VirtualNDIterator

shape() - Method in interface neureka.ndim.NDimensional

shape(int) - Method in interface neureka.ndim.NDimensional

This method receives an axis index and return the size of the targeted axis / dimension.

Shape - Interface in neureka

Basically a tuple of integers which is used to describe the shape of an array.

shaped(int...) - Static method in interface neureka.Nda

Returns a Collector that accumulates the input elements into a new Nda with the specified shape.

shaped(int...) - Static method in interface neureka.Tensor

Returns a Collector that accumulates the input elements into a new Tensor with the specified shape.

shaped(Shape) - Static method in interface neureka.Tensor

Returns a Collector that accumulates the input elements into a new Tensor with the specified shape.

shapeOfCon(int[], int[]) - Static method in class neureka.backend.main.operations.ConvUtil

shapeString(int[]) - Static method in class neureka.ndim.NDUtil

shortToBigInteger(short[]) - Static method in class neureka.common.utility.DataConverter.Utility

shortToByte(short[]) - Static method in class neureka.common.utility.DataConverter.Utility

shortToDouble(short[]) - Static method in class neureka.common.utility.DataConverter.Utility

shortToFloat(short[]) - Static method in class neureka.common.utility.DataConverter.Utility

shortToInt(short[]) - Static method in class neureka.common.utility.DataConverter.Utility

shortToLong(short[]) - Static method in class neureka.common.utility.DataConverter.Utility

sig(double) - Static method in class neureka.backend.main.implementations.fun.ScalarSigmoid

sig() - Method in interface neureka.Tensor

This method is a functionally identical to the following alternatives:

SIGMOID - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

Sigmoid - Class in neureka.backend.main.operations.functions

Sigmoid() - Constructor for class neureka.backend.main.operations.functions.Sigmoid

sigmoid() - Method in class neureka.math.Functions

sigmoid() - Method in interface neureka.Tensor

signed() - Method in class neureka.dtype.custom.F32

signed() - Method in class neureka.dtype.custom.F64

signed() - Method in class neureka.dtype.custom.I16

signed() - Method in class neureka.dtype.custom.I32

signed() - Method in class neureka.dtype.custom.I64

signed() - Method in class neureka.dtype.custom.I8

signed() - Method in class neureka.dtype.custom.UI16

signed() - Method in class neureka.dtype.custom.UI32

signed() - Method in class neureka.dtype.custom.UI64

signed() - Method in class neureka.dtype.custom.UI8

signed() - Method in interface neureka.dtype.NumericType

This boolean value tells if the data-type represented by concrete instances of implementations of this interface is signed!

SILU - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

silu(double) - Static method in class neureka.backend.main.implementations.fun.ScalarSiLU

SiLU - Class in neureka.backend.main.operations.functions

The SiLu activation function, also known as the swish function, is defined as x * sigmoid(x).

SiLU() - Constructor for class neureka.backend.main.operations.functions.SiLU

silu() - Method in class neureka.math.Functions

The SiLu activation function, also known as the swish function, is defined as x * sigmoid(x).

similarity(String, String) - Static method in class neureka.math.parsing.ParseUtil

This method estimates the similarity between 2 provided String instances.

Simple0DConfiguration - Class in neureka.ndim.config.types.simple

Simple1DCIterator - Class in neureka.ndim.iterator.types.simple

Simple1DCIterator(Simple1DConfiguration) - Constructor for class neureka.ndim.iterator.types.simple.Simple1DCIterator

Simple1DConfiguration - Class in neureka.ndim.config.types.simple

Simple1DConfiguration(int, int) - Constructor for class neureka.ndim.config.types.simple.Simple1DConfiguration

Simple2DCIterator - Class in neureka.ndim.iterator.types.simple

Simple2DCIterator(Simple2DConfiguration) - Constructor for class neureka.ndim.iterator.types.simple.Simple2DCIterator

Simple2DConfiguration - Class in neureka.ndim.config.types.simple

Simple2DConfiguration(int[], int[]) - Constructor for class neureka.ndim.config.types.simple.Simple2DConfiguration

Simple3DCIterator - Class in neureka.ndim.iterator.types.simple

Simple3DCIterator(Simple3DConfiguration) - Constructor for class neureka.ndim.iterator.types.simple.Simple3DCIterator

Simple3DConfiguration - Class in neureka.ndim.config.types.simple

Simple3DConfiguration(int[], int[]) - Constructor for class neureka.ndim.config.types.simple.Simple3DConfiguration

SimpleCLImplementation - Class in neureka.backend.main.implementations

SimpleCLImplementation(ImplementationFor<OpenCLDevice>, int, String, String) - Constructor for class neureka.backend.main.implementations.SimpleCLImplementation

SimpleNDConfiguration - Class in neureka.ndim.config.types.simple

SimpleNDConfiguration(int[], int[]) - Constructor for class neureka.ndim.config.types.simple.SimpleNDConfiguration

SimpleReshapeView - Class in neureka.ndim.config.types.views

SimpleReshapeView(int[], NDConfiguration) - Constructor for class neureka.ndim.config.types.views.SimpleReshapeView

sin() - Method in class neureka.math.Functions

sin() - Method in interface neureka.Tensor

This method is a functionally identical to the following alternatives:

singleFPConfig() - Method in class neureka.devices.opencl.OpenCLDevice

SINUS - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

Sinus - Class in neureka.backend.main.operations.functions

Sinus() - Constructor for class neureka.backend.main.operations.functions.Sinus

size() - Method in class neureka.autograd.GraphNode

This is the number of AD-actions stored inside this node.

size() - Method in class neureka.backend.api.BackendContext

size() - Method in class neureka.common.utility.Cache

size() - Method in interface neureka.ndim.config.NDConfiguration

size() - Method in interface neureka.ndim.NDimensional

size() - Method in interface neureka.Shape

sizeOfShape(int[]) - Static method in class neureka.ndim.config.NDConfiguration.Utility

Slice - Class in neureka.backend.main.operations.other

Slice() - Constructor for class neureka.backend.main.operations.other.Slice

slice(Object[], Tensor<ValType>) - Static method in class neureka.fluent.slicing.SmartSlicer

slice() - Method in interface neureka.Nda

This method returns a SliceBuilder instance exposing a simple builder API which enables the configuration of a slice of the current nd-array via method chaining.

slice(int, int) - Method in interface neureka.Shape

slice(int) - Method in interface neureka.Shape

slice() - Method in interface neureka.Tensor

This method returns a SliceBuilder instance exposing a simple builder API which enables the configuration of a slice of the current nd-array via method chaining.

SliceBuilder<V> - Class in neureka.fluent.slicing

This class is the heart of the slice builder API, collecting range configurations by exposing an API consisting of multiple interfaces which form a call state transition graph.

SliceBuilder(Tensor<V>) - Constructor for class neureka.fluent.slicing.SliceBuilder

An instance of a slice builder does not perform the actual slicing itself! Instead, it merely serves as a collector of slice configuration data.

sliceCount() - Method in interface neureka.Nda

This method returns the number of slices which have been created from this nd-array.

sliceCount() - Method in interface neureka.Tensor

This method returns the number of slices which have been created from this nd-array.

Sliced0DConfiguration - Class in neureka.ndim.config.types.sliced

Sliced0DConfiguration(int, int) - Constructor for class neureka.ndim.config.types.sliced.Sliced0DConfiguration

Sliced1DCIterator - Class in neureka.ndim.iterator.types.sliced

Sliced1DCIterator(Sliced1DConfiguration) - Constructor for class neureka.ndim.iterator.types.sliced.Sliced1DCIterator

Sliced1DConfiguration - Class in neureka.ndim.config.types.sliced

Sliced1DConfiguration(int, int, int, int, int) - Constructor for class neureka.ndim.config.types.sliced.Sliced1DConfiguration

Sliced2DCIterator - Class in neureka.ndim.iterator.types.sliced

Sliced2DCIterator(Sliced2DConfiguration) - Constructor for class neureka.ndim.iterator.types.sliced.Sliced2DCIterator

Sliced2DConfiguration - Class in neureka.ndim.config.types.sliced

Sliced2DConfiguration(int[], int[], int[], int[], int[]) - Constructor for class neureka.ndim.config.types.sliced.Sliced2DConfiguration

Sliced3DCIterator - Class in neureka.ndim.iterator.types.sliced

Sliced3DCIterator(Sliced3DConfiguration) - Constructor for class neureka.ndim.iterator.types.sliced.Sliced3DCIterator

Sliced3DConfiguration - Class in neureka.ndim.config.types.sliced

Sliced3DConfiguration(int[], int[], int[], int[], int[]) - Constructor for class neureka.ndim.config.types.sliced.Sliced3DConfiguration

SlicedNDConfiguration - Class in neureka.ndim.config.types.sliced

SlicedNDConfiguration(int[], int[], int[], int[], int[]) - Constructor for class neureka.ndim.config.types.sliced.SlicedNDConfiguration

SlicedNDIterator - Class in neureka.ndim.iterator.types.sliced

SlicedNDIterator(NDConfiguration) - Constructor for class neureka.ndim.iterator.types.sliced.SlicedNDIterator

SmartSlicer - Class in neureka.fluent.slicing

This class is responsible for receiving any input and trying to interpret it so that a slice can be formed.

SmartSlicer() - Constructor for class neureka.fluent.slicing.SmartSlicer

softmax() - Method in interface neureka.Tensor

softmax(int) - Method in interface neureka.Tensor

softmax(int...) - Method in interface neureka.Tensor

Calculates the softmax function along the specified axes.

SOFTPLUS - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

Softplus - Class in neureka.backend.main.operations.functions

SoftPlus is a smooth approximation to the ReLU function and can be used to constrain the output of a machine to always be positive.

Softplus() - Constructor for class neureka.backend.main.operations.functions.Softplus

softplus() - Method in class neureka.math.Functions

SoftPlus is a smooth approximation to the ReLU function and can be used to constrain the output of a machine to always be positive.

softplus() - Method in interface neureka.Tensor

This method is a functionally identical to the following alternatives:

SOFTSIGN - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

softsign(double) - Static method in class neureka.backend.main.implementations.fun.ScalarSoftsign

softsign(float) - Static method in class neureka.backend.main.implementations.fun.ScalarSoftsign

Softsign - Class in neureka.backend.main.operations.functions

The softsign function, defined as x / ( 1 + Math.abs( x ) ), is a computationally cheap 0 centered activation function which rescales the inputs between -1 and 1, very much like the Tanh function.

Softsign() - Constructor for class neureka.backend.main.operations.functions.Softsign

softsign() - Method in class neureka.math.Functions

The softsign function, defined as x / ( 1 + Math.abs( x ) ), is a computationally cheap 0 centered activation function which rescales the inputs between -1 and 1, very much like the Tanh function.

spaces(int) - Static method in class neureka.view.NdaAsString.Util

spread() - Method in interface neureka.ndim.config.NDConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread(int) - Method in interface neureka.ndim.config.NDConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread() - Method in class neureka.ndim.config.types.permuted.Permuted1DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread(int) - Method in class neureka.ndim.config.types.permuted.Permuted1DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread() - Method in class neureka.ndim.config.types.permuted.Permuted2DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread(int) - Method in class neureka.ndim.config.types.permuted.Permuted2DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread() - Method in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread(int) - Method in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread() - Method in class neureka.ndim.config.types.permuted.PermutedNDConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread(int) - Method in class neureka.ndim.config.types.permuted.PermutedNDConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread() - Method in class neureka.ndim.config.types.simple.Simple0DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread(int) - Method in class neureka.ndim.config.types.simple.Simple0DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread() - Method in class neureka.ndim.config.types.simple.Simple1DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread(int) - Method in class neureka.ndim.config.types.simple.Simple1DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread() - Method in class neureka.ndim.config.types.simple.Simple2DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread(int) - Method in class neureka.ndim.config.types.simple.Simple2DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread() - Method in class neureka.ndim.config.types.simple.Simple3DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread(int) - Method in class neureka.ndim.config.types.simple.Simple3DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread() - Method in class neureka.ndim.config.types.simple.SimpleNDConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread(int) - Method in class neureka.ndim.config.types.simple.SimpleNDConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread() - Method in class neureka.ndim.config.types.sliced.Sliced0DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread(int) - Method in class neureka.ndim.config.types.sliced.Sliced0DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread() - Method in class neureka.ndim.config.types.sliced.Sliced1DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread(int) - Method in class neureka.ndim.config.types.sliced.Sliced1DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread() - Method in class neureka.ndim.config.types.sliced.Sliced2DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread(int) - Method in class neureka.ndim.config.types.sliced.Sliced2DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread() - Method in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread(int) - Method in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread() - Method in class neureka.ndim.config.types.sliced.SlicedNDConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread(int) - Method in class neureka.ndim.config.types.sliced.SlicedNDConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread() - Method in class neureka.ndim.config.types.views.SimpleReshapeView

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread(int) - Method in class neureka.ndim.config.types.views.SimpleReshapeView

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread() - Method in class neureka.ndim.config.types.views.virtual.VirtualNDConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread(int) - Method in class neureka.ndim.config.types.views.virtual.VirtualNDConfiguration

The spread is the access step size of a slice within the n-dimensional data array of its parent tensor.

spread() - Method in interface neureka.ndim.NDimensional

SQRT - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

Sqrt - Class in neureka.backend.main.operations.functions

Sqrt() - Constructor for class neureka.backend.main.operations.functions.Sqrt

sqrt() - Method in class neureka.math.Functions

sqrt() - Method in interface neureka.Tensor

This method is a functionally identical to the following alternatives:

StaticKernelSource - Interface in neureka.devices.opencl

step(double) - Method in class neureka.fluent.building.NdaBuilder

Step<V> - Interface in neureka.fluent.building.states

This interface defines the last step in the call transition graph of the fluent builder API when building a Tensor instance populated based on the values within a defined range.

step(double) - Method in interface neureka.fluent.building.states.Step

This is the last step in the call transition graph of the fluent builder API when building a Tensor instance populated based on the values within a defined range.

step(double) - Method in interface neureka.fluent.building.states.StepForTensor

This is the last step in the call transition graph of the fluent builder API when building a Tensor instance populated based on the values within a defined range.

step(int) - Method in class neureka.fluent.slicing.AxisSliceBuilder

This method returns an instance of this very AxisSliceBuilder instance disguised by the AxisOrGet interface.

step(int) - Method in interface neureka.fluent.slicing.states.StepsOrAxisOrGet

This method allows one to specify a step size within the slice range previously specified for the currently sliced axis.

step(int) - Method in interface neureka.fluent.slicing.states.StepsOrAxisOrGetTensor

This method allows one to specify a step size within the slice range previously specified for the currently sliced axis.

StepForTensor<V> - Interface in neureka.fluent.building.states

StepsOrAxisOrGet<V> - Interface in neureka.fluent.slicing.states

This interface extends the AxisOrGet interface which provides the option to either continue slicing another axis or simply trigger the creation and return of a slice instance based on the already provided slice configuration.

StepsOrAxisOrGetTensor<V> - Interface in neureka.fluent.slicing.states

Storage<V> - Interface in neureka.devices

This is an abstract interface which simply describes "a thing that stores tensors".

store(Tensor<T>) - Method in class neureka.devices.AbstractDevice

Implementations of this method ought to store the data of the tensor in whatever formant suites the underlying implementation and or final type.

store(Tensor<T>) - Method in class neureka.devices.file.CSVHandle

store(Tensor<T>) - Method in class neureka.devices.file.FileDevice

Implementations of this method ought to store the data of the tensor in whatever formant suites the underlying implementation and or final type.

store(Tensor<T>, String) - Method in class neureka.devices.file.FileDevice

Stores the given tensor in the file system with the given filename.

store(Tensor<T>, String, Map<String, Object>) - Method in class neureka.devices.file.FileDevice

Stores the given tensor in the file system with the given filename.

store(Tensor<T>) - Method in class neureka.devices.file.IDXHandle

store(Tensor<T>) - Method in class neureka.devices.host.CPU

store(Tensor<T>) - Method in interface neureka.devices.Storage

Implementations of this method ought to store the data of the tensor in whatever formant suites the underlying implementation and or final type.

stream() - Method in interface neureka.Nda

stream() - Method in interface neureka.Shape

strides() - Method in interface neureka.ndim.config.NDConfiguration

The array returned by this method is used to translate an array of axis indices to a single ata array index.

strides(int) - Method in interface neureka.ndim.config.NDConfiguration

This method receives an axis index and returns the translation value for the targeted axis.

strides() - Method in class neureka.ndim.config.types.permuted.Permuted1DConfiguration

The array returned by this method is used to translate an array of axis indices to a single ata array index.

strides(int) - Method in class neureka.ndim.config.types.permuted.Permuted1DConfiguration

This method receives an axis index and returns the translation value for the targeted axis.

strides() - Method in class neureka.ndim.config.types.permuted.Permuted2DConfiguration

The array returned by this method is used to translate an array of axis indices to a single ata array index.

strides(int) - Method in class neureka.ndim.config.types.permuted.Permuted2DConfiguration

This method receives an axis index and returns the translation value for the targeted axis.

strides() - Method in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

The array returned by this method is used to translate an array of axis indices to a single ata array index.

strides(int) - Method in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

This method receives an axis index and returns the translation value for the targeted axis.

strides() - Method in class neureka.ndim.config.types.permuted.PermutedNDConfiguration

The array returned by this method is used to translate an array of axis indices to a single ata array index.

strides(int) - Method in class neureka.ndim.config.types.permuted.PermutedNDConfiguration

This method receives an axis index and returns the translation value for the targeted axis.

strides() - Method in class neureka.ndim.config.types.simple.Simple0DConfiguration

The array returned by this method is used to translate an array of axis indices to a single ata array index.

strides(int) - Method in class neureka.ndim.config.types.simple.Simple0DConfiguration

This method receives an axis index and returns the translation value for the targeted axis.

strides() - Method in class neureka.ndim.config.types.simple.Simple1DConfiguration

The array returned by this method is used to translate an array of axis indices to a single ata array index.

strides(int) - Method in class neureka.ndim.config.types.simple.Simple1DConfiguration

This method receives an axis index and returns the translation value for the targeted axis.

strides() - Method in class neureka.ndim.config.types.simple.Simple2DConfiguration

The array returned by this method is used to translate an array of axis indices to a single ata array index.

strides(int) - Method in class neureka.ndim.config.types.simple.Simple2DConfiguration

This method receives an axis index and returns the translation value for the targeted axis.

strides() - Method in class neureka.ndim.config.types.simple.Simple3DConfiguration

The array returned by this method is used to translate an array of axis indices to a single ata array index.

strides(int) - Method in class neureka.ndim.config.types.simple.Simple3DConfiguration

This method receives an axis index and returns the translation value for the targeted axis.

strides() - Method in class neureka.ndim.config.types.simple.SimpleNDConfiguration

The array returned by this method is used to translate an array of axis indices to a single ata array index.

strides(int) - Method in class neureka.ndim.config.types.simple.SimpleNDConfiguration

This method receives an axis index and returns the translation value for the targeted axis.

strides() - Method in class neureka.ndim.config.types.sliced.Sliced0DConfiguration

The array returned by this method is used to translate an array of axis indices to a single ata array index.

strides(int) - Method in class neureka.ndim.config.types.sliced.Sliced0DConfiguration

This method receives an axis index and returns the translation value for the targeted axis.

strides() - Method in class neureka.ndim.config.types.sliced.Sliced1DConfiguration

The array returned by this method is used to translate an array of axis indices to a single ata array index.

strides(int) - Method in class neureka.ndim.config.types.sliced.Sliced1DConfiguration

This method receives an axis index and returns the translation value for the targeted axis.

strides() - Method in class neureka.ndim.config.types.sliced.Sliced2DConfiguration

The array returned by this method is used to translate an array of axis indices to a single ata array index.

strides(int) - Method in class neureka.ndim.config.types.sliced.Sliced2DConfiguration

This method receives an axis index and returns the translation value for the targeted axis.

strides() - Method in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

The array returned by this method is used to translate an array of axis indices to a single ata array index.

strides(int) - Method in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

This method receives an axis index and returns the translation value for the targeted axis.

strides() - Method in class neureka.ndim.config.types.sliced.SlicedNDConfiguration

The array returned by this method is used to translate an array of axis indices to a single ata array index.

strides(int) - Method in class neureka.ndim.config.types.sliced.SlicedNDConfiguration

This method receives an axis index and returns the translation value for the targeted axis.

strides() - Method in class neureka.ndim.config.types.views.SimpleReshapeView

The array returned by this method is used to translate an array of axis indices to a single ata array index.

strides(int) - Method in class neureka.ndim.config.types.views.SimpleReshapeView

This method receives an axis index and returns the translation value for the targeted axis.

strides() - Method in class neureka.ndim.config.types.views.virtual.VirtualNDConfiguration

The array returned by this method is used to translate an array of axis indices to a single ata array index.

strides(int) - Method in class neureka.ndim.config.types.views.virtual.VirtualNDConfiguration

This method receives an axis index and returns the translation value for the targeted axis.

strides() - Method in interface neureka.ndim.NDimensional

stringifier(OperationBuilder.Stringifier) - Method in class neureka.backend.api.template.operations.OperationBuilder

stringify(String[]) - Method in interface neureka.backend.api.Operation

stringify(String[]) - Method in class neureka.backend.api.template.operations.AbstractOperation

stringify(String[]) - Method in interface neureka.backend.api.template.operations.OperationBuilder.Stringifier

stringify(String[]) - Method in class neureka.backend.main.operations.linear.XConvLeft

stringify(String[]) - Method in class neureka.backend.main.operations.linear.XConvRight

stringify(String[]) - Method in class neureka.backend.main.operations.other.AssignLeft

stringify(String[]) - Method in class neureka.backend.main.operations.other.Permute

submit(int, CPU.IndexedWorkload) - Method in class neureka.devices.host.concurrent.WorkScheduler.Divider

Subtraction - Class in neureka.backend.main.operations.operator

Subtraction() - Constructor for class neureka.backend.main.operations.operator.Subtraction

suitabilityIfValid(float) - Method in class neureka.backend.api.Call.Validator

SuitabilityPredicate - Interface in neureka.backend.api.fun

The SuitabilityPredicate checks if a given instance of an ExecutionCall is suitable to be executed in ImplementationFor residing in this Algorithm as components.

Sum - Class in neureka.backend.main.operations.other

Sum() - Constructor for class neureka.backend.main.operations.other.Sum

sum() - Method in class neureka.math.Functions

sum() - Method in interface neureka.Tensor

Calculate the sum value of all values within this tensor and returns it in the form of a scalar tensor.

sum(int) - Method in interface neureka.Tensor

Calculate the sum value of all values within this tensor along the specified axis and returns it in the form of a tensor.

sum(int...) - Method in interface neureka.Tensor

Calculate the sum value of all values within this tensor along the specified axes and returns it in the form of a tensor.

SumAlgorithm - Class in neureka.backend.main.algorithms

SumAlgorithm() - Constructor for class neureka.backend.main.algorithms.SumAlgorithm

Summation - Class in neureka.backend.main.operations.indexer

This type of operation belongs to the same species as the Product operation.

Summation() - Constructor for class neureka.backend.main.operations.indexer.Summation

supplyADActionFor(Function, ExecutionCall<? extends Device<?>>) - Method in interface neureka.backend.api.fun.ADActionSupplier

This method ought to return a new instance if the ADAction class responsible for performing automatic differentiation both for forward and backward mode differentiation.

supplyADActionFor(Function, ExecutionCall<? extends Device<?>>) - Method in class neureka.backend.api.template.algorithms.FallbackAlgorithm

supports(Class<T>) - Method in interface neureka.backend.api.Operation

supports(Class<T>) - Method in class neureka.backend.api.template.operations.AbstractOperation

supportsAlgorithm(Class<T>) - Method in interface neureka.backend.api.Operation

This method checks if this Operation contains an instance of the Algorithm implementation specified via its type class.

supportsAlgorithm(Class<T>) - Method in class neureka.backend.api.template.operations.AbstractOperation

This method checks if this Operation contains an instance of the Algorithm implementation specified via its type class.

T

T() - Method in interface neureka.Tensor

Creates and returns a new Tensor instance which is a transposed twin of this instance.
This is a shorter alternative to the functionally identical Tensor.getT() method.

TANH - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

tanh(double) - Static method in class neureka.backend.main.implementations.fun.ScalarTanh

tanh(float) - Static method in class neureka.backend.main.implementations.fun.ScalarTanh

Tanh - Class in neureka.backend.main.operations.functions

Tanh() - Constructor for class neureka.backend.main.operations.functions.Tanh

tanh() - Method in class neureka.math.Functions

tanh() - Method in interface neureka.Tensor

This method is a functionally identical to the following alternatives:

TANH_FAST - Static variable in interface neureka.backend.main.implementations.fun.api.ScalarFun

TanhFast - Class in neureka.backend.main.operations.functions

TanhFast() - Constructor for class neureka.backend.main.operations.functions.TanhFast

targetArrayType() - Method in class neureka.dtype.custom.F32

targetArrayType() - Method in class neureka.dtype.custom.F64

targetArrayType() - Method in class neureka.dtype.custom.I16

targetArrayType() - Method in class neureka.dtype.custom.I32

targetArrayType() - Method in class neureka.dtype.custom.I64

targetArrayType() - Method in class neureka.dtype.custom.I8

targetArrayType() - Method in class neureka.dtype.custom.UI16

targetArrayType() - Method in class neureka.dtype.custom.UI32

targetArrayType() - Method in class neureka.dtype.custom.UI64

targetArrayType() - Method in class neureka.dtype.custom.UI8

targetArrayType() - Method in interface neureka.dtype.NumericType

The target type is the targeted JVM data-type which can represent the holder type.

targetToForeignHolderBytes(Float) - Method in class neureka.dtype.custom.F32

targetToForeignHolderBytes(Double) - Method in class neureka.dtype.custom.F64

targetToForeignHolderBytes(Short) - Method in class neureka.dtype.custom.I16

targetToForeignHolderBytes(Integer) - Method in class neureka.dtype.custom.I32

targetToForeignHolderBytes(Long) - Method in class neureka.dtype.custom.I64

targetToForeignHolderBytes(Byte) - Method in class neureka.dtype.custom.I8

targetToForeignHolderBytes(Integer) - Method in class neureka.dtype.custom.UI16

targetToForeignHolderBytes(Long) - Method in class neureka.dtype.custom.UI32

targetToForeignHolderBytes(BigInteger) - Method in class neureka.dtype.custom.UI64

targetToForeignHolderBytes(Short) - Method in class neureka.dtype.custom.UI8

targetToForeignHolderBytes(TargetType) - Method in interface neureka.dtype.NumericType

targetType() - Method in class neureka.dtype.custom.F32

targetType() - Method in class neureka.dtype.custom.F64

targetType() - Method in class neureka.dtype.custom.I16

targetType() - Method in class neureka.dtype.custom.I32

targetType() - Method in class neureka.dtype.custom.I64

targetType() - Method in class neureka.dtype.custom.I8

targetType() - Method in class neureka.dtype.custom.UI16

targetType() - Method in class neureka.dtype.custom.UI32

targetType() - Method in class neureka.dtype.custom.UI64

targetType() - Method in class neureka.dtype.custom.UI8

targetType() - Method in interface neureka.dtype.NumericType

The target type is the targeted JVM data-type which can represent the holder type.

Tensor<V> - Interface in neureka

A Tensor is a mathematical concept and type of multidimensional data-structure with certain transformation properties.

Tensor.ImageType - Enum in neureka

Use this enum as argument for the Tensor.asImage(Tensor.ImageType) method to specify the type of image that should be returned.

tensors(Call.TensorsCondition) - Method in class neureka.backend.api.Call.Validator

TERRIBLE - Static variable in interface neureka.backend.api.fun.SuitabilityPredicate

THREAD_PREFIX - Static variable in class neureka.devices.host.CPU

threaded(int, CPU.RangeWorkload) - Method in class neureka.devices.host.CPU.JVMExecutor

This method slices the provided workload size into multiple ranges which can be executed in parallel.

threaded(int, CPU.IndexedWorkload) - Method in class neureka.devices.host.CPU.JVMExecutor

Executes the provided workload lambda across multiple threads where the provided worker lambda will receive the index/id of the current worker.

threaded(int, int, CPU.RangeWorkload) - Method in class neureka.devices.host.CPU.JVMExecutor

Takes the provided range and divides it into multithreaded workloads.

threads - Variable in class neureka.devices.host.machine.BasicMachine

threshold(int) - Method in class neureka.devices.host.concurrent.WorkScheduler.Divider

times(Tensor<V>) - Method in interface neureka.Tensor

This is a functionally identical synonym to the Tensor.multiply(Tensor) method.

times(V) - Method in interface neureka.Tensor

timesAssign(Tensor<T>) - Method in interface neureka.MutateTensor

timesAssign(T) - Method in interface neureka.MutateTensor

to(T) - Static method in class neureka.backend.api.Call

to(V) - Method in class neureka.fluent.building.NdaBuilder

To<V> - Interface in neureka.fluent.building.states

This step in the call transition graph of the fluent builder API is a followup call from the IterByOrIterFromOrAll.andFillFrom(Object) method which expects a range to be specified whose values will be used to populate the Tensor instance.

to(V) - Method in interface neureka.fluent.building.states.To

This step in the call transition graph of the fluent builder API is a followup call from the IterByOrIterFromOrAll.andFillFrom(Object) method which expects a range to be specified whose values will be used to populate the Tensor instance.

to(V) - Method in interface neureka.fluent.building.states.ToForTensor

This step in the call transition graph of the fluent builder API is a followup call from the IterByOrIterFromOrAll.andFillFrom(Object) method which expects a range to be specified whose values will be used to populate the Tensor instance.

to(int) - Method in class neureka.fluent.slicing.AxisSliceBuilder

This method returns an instance of this very AxisSliceBuilder instance disguised by the StepsOrAxisOrGet interface.

To<V> - Interface in neureka.fluent.slicing.states

This is the second part for defining the slice range of a specified axis within the call transition graph exposed by the slice builder API.

to(int) - Method in interface neureka.fluent.slicing.states.To

This is the second part for defining the slice range of a specified axis within the call transition graph exposed by the slice fluent builder API.

to(int) - Method in interface neureka.fluent.slicing.states.ToForTensor

This is the second part for defining the slice range of a specified axis within the call transition graph exposed by the slice fluent builder API.

to(Device<?>) - Method in interface neureka.Tensor

This method takes a Device and tries to migrate the contents of this Tensor instance to that Device!

to(String) - Method in interface neureka.Tensor

toByteArray(Function<Integer, Number>) - Method in class neureka.common.utility.DataConverter.ForTensor

toDoubleArray(Function<Integer, Number>) - Method in class neureka.common.utility.DataConverter.ForTensor

toFloatArray(Function<Integer, Number>) - Method in class neureka.common.utility.DataConverter.ForTensor

ToForTensor<V> - Interface in neureka.fluent.building.states

ToForTensor<V> - Interface in neureka.fluent.slicing.states

toIntArray(Function<Integer, Number>) - Method in class neureka.common.utility.DataConverter.ForTensor

toIntArray() - Method in interface neureka.Shape

toLayout(Tensor<?>, NDConfiguration.Layout) - Static method in class neureka.backend.main.operations.other.ReLayout

toLayout(NDConfiguration.Layout) - Method in interface neureka.MutateTensor

This method allows you to modify the data-layout of this AbstractNda.

toLongArray(Function<Integer, Number>) - Method in class neureka.common.utility.DataConverter.ForTensor

toObjectArray(Function<Integer, Object>) - Method in class neureka.common.utility.DataConverter.ForTensor

toOptional() - Method in interface neureka.Nda.Item

Converts this item into an optional value.

toShortArray(Function<Integer, Number>) - Method in class neureka.common.utility.DataConverter.ForTensor

toString() - Method in class neureka.autograd.GraphNode

toString(GraphNode.Print) - Method in class neureka.autograd.GraphNode

toString() - Method in class neureka.autograd.JITProp

toString() - Method in class neureka.backend.api.BackendContext

toString() - Method in class neureka.backend.api.ExecutionCall

toString() - Method in class neureka.backend.api.LazyRef

toString() - Method in class neureka.backend.api.template.algorithms.AbstractDeviceAlgorithm

toString() - Method in class neureka.backend.api.template.operations.AbstractOperation

toString() - Method in class neureka.backend.ocl.CLBackend

toString() - Method in class neureka.devices.file.FileDevice

toString() - Method in class neureka.devices.host.CPU

toString() - Method in class neureka.devices.host.machine.BasicMachine

toString() - Method in class neureka.devices.host.machine.ConcreteMachine

toString() - Method in class neureka.devices.host.machine.Hardware

toString() - Method in class neureka.devices.opencl.OpenCLDevice

toString() - Method in class neureka.devices.opencl.OpenCLPlatform

toString() - Method in class neureka.dtype.DataType

toString() - Method in class neureka.framing.NDFrame

toString() - Method in class neureka.framing.Relation

toString() - Method in class neureka.math.args.Arg

toString() - Method in interface neureka.math.Function

Turns this function into a string representation which can be used to reconstruct this function or combine it with other function strings to parse entirely new functions...

toString() - Method in class neureka.math.FunctionCache

toString() - Method in class neureka.math.Functions

toString() - Method in class neureka.math.implementations.FunctionConstant

toString() - Method in class neureka.math.implementations.FunctionInput

toString() - Method in class neureka.math.implementations.FunctionNode

toString() - Method in class neureka.math.implementations.FunctionVariable

toString(NDPrintSettings) - Method in interface neureka.Nda

Use this to turn this nd-array into a String instance based on the provided NDPrintSettings instance, which allows you to configure things like the number of chars per entry, delimiters, the number of items per line, etc.

toString(Consumer<NDPrintSettings>) - Method in interface neureka.Nda

This allows you to provide a lambda which configures how this nd-array should be converted to String instances.

toString() - Method in interface neureka.Nda

This method returns a String representation of this nd-array.

toString() - Method in class neureka.ndim.config.AbstractNDC

toString() - Method in class neureka.Neureka.Settings.AutoGrad

toString() - Method in class neureka.Neureka.Settings.Debug

toString() - Method in class neureka.Neureka.Settings.DType

toString() - Method in class neureka.Neureka.Settings.NDim

toString() - Method in class neureka.Neureka.Settings

toString() - Method in class neureka.Neureka.Settings.View

toString() - Method in class neureka.Neureka

toString(String) - Method in interface neureka.Tensor

toString(NDPrintSettings) - Method in interface neureka.Tensor

Use this to turn this nd-array into a String instance based on the provided NDPrintSettings instance, which allows you to configure things like the number of chars per entry, delimiters, the number of items per line, etc.

toString(Consumer<NDPrintSettings>) - Method in interface neureka.Tensor

This allows you to provide a lambda which configures how this nd-array should be converted to String instances.

toString() - Method in class neureka.view.NdaAsString

toTarget(Float) - Method in class neureka.dtype.custom.F32

toTarget(Double) - Method in class neureka.dtype.custom.F64

toTarget(Short) - Method in class neureka.dtype.custom.I16

toTarget(Integer) - Method in class neureka.dtype.custom.I32

toTarget(Long) - Method in class neureka.dtype.custom.I64

toTarget(Byte) - Method in class neureka.dtype.custom.I8

toTarget(Short) - Method in class neureka.dtype.custom.UI16

toTarget(Integer) - Method in class neureka.dtype.custom.UI32

toTarget(Long) - Method in class neureka.dtype.custom.UI64

toTarget(Byte) - Method in class neureka.dtype.custom.UI8

toTarget(HolderType) - Method in interface neureka.dtype.NumericType

toType(Class<V>) - Method in interface neureka.MutateNda

This method is an inline operation which changes the underlying data of this tensor.

toType(Class<V>) - Method in interface neureka.MutateTensor

This method is an inline operation which changes the underlying data of this tensor.

transpose(Tensor<T>) - Static method in class neureka.backend.main.algorithms.Util

transpose(int, int) - Method in interface neureka.Nda

Returns a view of the original tensor input the targeted axes are swapped / transposed.

transpose(int, int) - Method in interface neureka.Tensor

Returns a view of the original tensor input the targeted axes are swapped / transposed.

transpose2D() - Method in class neureka.math.Functions

tryGroovyClosureOn(Object, Object) - Static method in class neureka.common.utility.SettingsLoader

This method makes it possible to configure the library via a Groovy DSL!

tryGroovyScriptsOn(Neureka, Consumer<String>) - Static method in class neureka.common.utility.SettingsLoader

type() - Method in class neureka.autograd.GraphNode

TYPE - Static variable in class neureka.backend.main.implementations.broadcast.CLScalarBroadcast

type() - Method in interface neureka.common.composition.Component.OwnerChangeRequest

This method will return one of the following states: Component.IsBeing.ADDED, Component.IsBeing.REMOVED, Component.IsBeing.REPLACED, Component.IsBeing.UPDATED

type() - Method in class neureka.devices.opencl.OpenCLDevice

type() - Method in class neureka.devices.ReferenceCounter.ChangeEvent

typeClassImplements(Class<?>) - Method in class neureka.dtype.DataType

U

UI16 - Class in neureka.dtype.custom

UI16() - Constructor for class neureka.dtype.custom.UI16

UI32 - Class in neureka.dtype.custom

UI32() - Constructor for class neureka.dtype.custom.UI32

UI64 - Class in neureka.dtype.custom

UI64() - Constructor for class neureka.dtype.custom.UI64

UI8 - Class in neureka.dtype.custom

UI8() - Constructor for class neureka.dtype.custom.UI8

units - Variable in class neureka.devices.host.machine.CommonMachine

The number of top level (L3 or L2) cache units.

unpackAndCorrect(String) - Static method in class neureka.math.parsing.ParseUtil

UNSUITABLE - Static variable in interface neureka.backend.api.fun.SuitabilityPredicate

upcast(Class<U>) - Method in interface neureka.MutateTensor

Use this to do a runtime checked upcast of the type parameter of the tensor.

update(Component.OwnerChangeRequest<Tensor<V>>) - Method in class neureka.autograd.GraphNode

update(Component.OwnerChangeRequest<Extensions>) - Method in class neureka.backend.ocl.CLBackend

Updating the CLContext will cause the list of existing OpenCLPlatform instances to be cleared and refilled with completely new OpenCLPlatform instances.

update(Component.OwnerChangeRequest<O>) - Method in interface neureka.common.composition.Component

Components are not the slaves of their owners.

update(Component.OwnerChangeRequest<Tensor<V>>) - Method in class neureka.devices.AbstractDevice

A Device is a component of a tensor.

update(Component.OwnerChangeRequest<Tensor<Object>>) - Method in class neureka.devices.file.FileDevice

update(Component.OwnerChangeRequest<Tensor<Object>>) - Method in class neureka.devices.host.CPU

This method is part of the component system built into the Tensor class.

update(Component.OwnerChangeRequest<Tensor<Number>>) - Method in class neureka.devices.opencl.OpenCLDevice

update(Component.OwnerChangeRequest<Tensor<V>>) - Method in class neureka.framing.NDFrame

update(Component.OwnerChangeRequest<Tensor<V>>) - Method in class neureka.framing.Relation

update(Component.OwnerChangeRequest<Args>) - Method in class neureka.math.args.Arg

update(Component.OwnerChangeRequest<Tensor<V>>) - Method in interface neureka.Tensor

Important : Components of type Tensor are simply gradients! Currently, this method is used only to catch illegal arguments which is for example the case when trying to attach a gradient with a different shape...

usages() - Method in interface neureka.Data

This method returns the number of times this data object is currently in use by a nd-array, meaning that the number of usages is also the number of nd-arrays which are currently referencing this data object.

usages() - Method in class neureka.devices.AbstractDeviceData

usesAD() - Method in class neureka.autograd.GraphNode

This gradient node is involved in auto-differentiation.

usesForwardAD() - Method in class neureka.autograd.GraphNode

This node propagates forward.

usesReverseAD() - Method in class neureka.autograd.GraphNode

This node propagates _backward.

Util - Class in neureka.backend.main.algorithms

Util() - Constructor for class neureka.backend.main.algorithms.Util

Util() - Constructor for class neureka.view.NdaAsString.Util

Utility() - Constructor for class neureka.common.utility.DataConverter.Utility

Utility() - Constructor for class neureka.ndim.config.NDConfiguration.Utility

utility() - Method in class neureka.Neureka

Utility() - Constructor for class neureka.Neureka.Utility

V

validate() - Method in class neureka.backend.api.Call

Validator() - Constructor for class neureka.backend.api.Call.Validator

valOf(Class<T>) - Method in class neureka.math.args.Args

valOfOr(Class<T>, V) - Method in class neureka.math.args.Args

value() - Method in class neureka.math.implementations.FunctionConstant

valueOf(String) - Static method in enum neureka.autograd.GraphNode.Print

Returns the enum constant of this type with the specified name.

valueOf(String) - Static method in enum neureka.backend.api.AutoDiffMode

Returns the enum constant of this type with the specified name.

valueOf(String) - Static method in enum neureka.backend.main.operations.linear.internal.opencl.CLReduce.Type

Returns the enum constant of this type with the specified name.

valueOf(String) - Static method in enum neureka.backend.main.operations.other.internal.CPUReduce.Type

Returns the enum constant of this type with the specified name.

valueOf(String) - Static method in enum neureka.common.composition.Component.IsBeing

Returns the enum constant of this type with the specified name.

valueOf(String) - Static method in enum neureka.devices.host.concurrent.Parallelism

Returns the enum constant of this type with the specified name.

valueOf(String) - Static method in enum neureka.devices.opencl.OpenCLDevice.Type

Returns the enum constant of this type with the specified name.

valueOf(String) - Static method in enum neureka.devices.opencl.utility.Messages.Tips

Returns the enum constant of this type with the specified name.

valueOf(String) - Static method in enum neureka.devices.ReferenceCounter.ChangeType

Returns the enum constant of this type with the specified name.

valueOf(String) - Static method in enum neureka.ndim.config.NDConfiguration.Layout

Returns the enum constant of this type with the specified name.

valueOf(String) - Static method in enum neureka.ndim.config.NDTrait

Returns the enum constant of this type with the specified name.

valueOf(String) - Static method in enum neureka.ndim.iterator.NDIterator.NonVirtual

Returns the enum constant of this type with the specified name.

valueOf(String) - Static method in enum neureka.Tensor.ImageType

Returns the enum constant of this type with the specified name.

values() - Static method in enum neureka.autograd.GraphNode.Print

Returns an array containing the constants of this enum type, in the order they are declared.

values() - Static method in enum neureka.backend.api.AutoDiffMode

Returns an array containing the constants of this enum type, in the order they are declared.

values() - Static method in enum neureka.backend.main.operations.linear.internal.opencl.CLReduce.Type

Returns an array containing the constants of this enum type, in the order they are declared.

values() - Static method in enum neureka.backend.main.operations.other.internal.CPUReduce.Type

Returns an array containing the constants of this enum type, in the order they are declared.

values() - Static method in enum neureka.common.composition.Component.IsBeing

Returns an array containing the constants of this enum type, in the order they are declared.

values() - Static method in enum neureka.devices.host.concurrent.Parallelism

Returns an array containing the constants of this enum type, in the order they are declared.

values() - Static method in enum neureka.devices.opencl.OpenCLDevice.Type

Returns an array containing the constants of this enum type, in the order they are declared.

values() - Static method in enum neureka.devices.opencl.utility.Messages.Tips

Returns an array containing the constants of this enum type, in the order they are declared.

values() - Static method in enum neureka.devices.ReferenceCounter.ChangeType

Returns an array containing the constants of this enum type, in the order they are declared.

values() - Static method in enum neureka.ndim.config.NDConfiguration.Layout

Returns an array containing the constants of this enum type, in the order they are declared.

values() - Static method in enum neureka.ndim.config.NDTrait

Returns an array containing the constants of this enum type, in the order they are declared.

values() - Static method in enum neureka.ndim.iterator.NDIterator.NonVirtual

Returns an array containing the constants of this enum type, in the order they are declared.

values() - Static method in enum neureka.Tensor.ImageType

Returns an array containing the constants of this enum type, in the order they are declared.

vector(Object[]) - Method in class neureka.fluent.building.NdaBuilder

vector(V...) - Method in interface neureka.fluent.building.states.WithShapeOrScalarOrVector

This method creates and returns a vector Tensor instance which wraps the provided values.

vector(List<V>) - Method in interface neureka.fluent.building.states.WithShapeOrScalarOrVector

This method creates and returns a vector Tensor instance which wraps the provided values.

vector(Iterable<V>) - Method in interface neureka.fluent.building.states.WithShapeOrScalarOrVector

This method creates and returns a vector Tensor instance which wraps the provided values.

vector(V...) - Method in interface neureka.fluent.building.states.WithShapeOrScalarOrVectorTensor

This method creates and returns a vector Tensor instance which wraps the provided values.

vector(List<V>) - Method in interface neureka.fluent.building.states.WithShapeOrScalarOrVectorTensor

This method creates and returns a vector Tensor instance which wraps the provided values.

vector(Iterable<V>) - Method in interface neureka.fluent.building.states.WithShapeOrScalarOrVectorTensor

This method creates and returns a vector Tensor instance which wraps the provided values.

vendor() - Method in class neureka.devices.opencl.OpenCLDevice

version() - Method in class neureka.devices.opencl.OpenCLDevice

version() - Static method in class neureka.Neureka

VERY_GOOD - Static variable in interface neureka.backend.api.fun.SuitabilityPredicate

view() - Method in class neureka.Neureka.Settings

view(Object) - Method in class neureka.Neureka.Settings

virtualize() - Method in interface neureka.devices.Device.Access

virtualize() - Method in class neureka.devices.host.machine.Hardware

VirtualNDConfiguration - Class in neureka.ndim.config.types.views.virtual

VirtualNDConfigurations represent tensors which are filled homogeneously with a single value exclusively, like for example a tensor filled with only zeros.

VirtualNDIterator - Class in neureka.ndim.iterator.types.virtual

VirtualNDIterator(VirtualNDConfiguration) - Constructor for class neureka.ndim.iterator.types.virtual.VirtualNDIterator

W

with(Tensor<N>...) - Method in class neureka.backend.api.Call.Builder

with(F) - Method in interface neureka.backend.main.algorithms.internal.WithForward

With<ValueType,TargetType> - Interface in neureka.framing.fluent

with(ValueType) - Method in interface neureka.framing.fluent.With

with(Arg<?>...) - Method in interface neureka.math.Function

Use this to call this Function alongside with some additional meta-arguments which will be passed to the underlying Operation(s).

with(Args) - Method in interface neureka.math.Function

Use this to call this Function alongside with some additional meta-arguments which will be passed to the underlying Operation(s).

with(NDPrintSettings) - Method in class neureka.view.NDPrintSettings

with(String) - Method in class neureka.view.NDPrintSettings

withADAction(ADAction) - Method in class neureka.backend.api.Result

withAddedInputAt(int, Tensor<?>) - Method in class neureka.backend.api.ExecutionCall

withArgs(Arg<?>...) - Method in class neureka.backend.api.ExecutionCall

Use this to produce a clone with a new set of meta arguments.

withArity(int) - Static method in class neureka.backend.main.implementations.CPUImplementation

withAutoDiff(ADActionSupplier) - Method in class neureka.backend.api.Result

withAxesLabels(List<List<Object>>) - Method in class neureka.framing.NDFrame

withConfig(NDPrintSettings) - Method in interface neureka.view.NdaAsString.Builder

withConfig(String) - Method in interface neureka.view.NdaAsString.Builder

withDecayRate(double) - Method in class neureka.optimization.implementations.MomentumFactory

withDecayRate(double) - Method in class neureka.optimization.implementations.RMSPropFactory

WithForward<F> - Interface in neureka.backend.main.algorithms.internal

withInputAt(int, Tensor<?>) - Method in class neureka.backend.api.ExecutionCall

withInputs(Tensor<?>...) - Method in class neureka.backend.api.ExecutionCall

Use this to produce a clone with a new array of input tensors.

withLabel(String) - Method in class neureka.framing.NDFrame

withLabel(String) - Method in interface neureka.Nda

withLabel(String) - Method in interface neureka.Tensor

withLabels(String[]...) - Method in interface neureka.Nda

This method receives a nested String array which ought to contain a label for the index of this nd-array.

withLabels(List<List<Object>>) - Method in interface neureka.Nda

This method receives a nested String list which ought to contain a label for the index of this nd-array.

withLabels(Map<Object, List<Object>>) - Method in interface neureka.Nda

This method provides the ability to label not only the indices of the shape of this nd-array, but also the dimension of the shape.

withLabels(String[]...) - Method in interface neureka.Tensor

This method receives a nested String array which ought to contain a label for the index of this nd-array.

withLabels(List<List<Object>>) - Method in interface neureka.Tensor

This method receives a nested String list which ought to contain a label for the index of this nd-array.

withLabels(Map<Object, List<Object>>) - Method in interface neureka.Tensor

This method provides the ability to label not only the indices of the shape of this nd-array, but also the dimension of the shape.

withLearningRate(double) - Method in class neureka.optimization.implementations.AdaGradFactory

withLearningRate(double) - Method in class neureka.optimization.implementations.ADAMFactory

withLearningRate(double) - Method in class neureka.optimization.implementations.MomentumFactory

withLearningRate(double) - Method in class neureka.optimization.implementations.RMSPropFactory

withLearningRate(double) - Method in class neureka.optimization.implementations.SGDFactory

withName(String) - Static method in interface neureka.backend.api.Algorithm

This is a factory method for creating a new instance of this FunAlgorithm class.

withName(String) - Static method in interface neureka.backend.api.DeviceAlgorithm

This is a factory method for creating a new instance of this FunDeviceAlgorithm class.

withOperation(Operation) - Method in class neureka.backend.api.ExecutionCall

withRemovedInputAt(int) - Method in class neureka.backend.api.ExecutionCall

withShape(int...) - Method in class neureka.fluent.building.NdaBuilder

withShape(int...) - Method in interface neureka.fluent.building.states.WithShapeOrScalarOrVector

Define a tensor shape by passing an array of int values to this method, which represent the shape of the Tensor that should be built.

withShape(List<N>) - Method in interface neureka.fluent.building.states.WithShapeOrScalarOrVector

Define a tensor shape by passing a list of numbers to this method, which represent the shape of the Tensor that should be built.

withShape(int...) - Method in interface neureka.fluent.building.states.WithShapeOrScalarOrVectorTensor

Define a tensor shape by passing an array of int values to this method, which represent the shape of the Tensor that should be built.

withShape(List<N>) - Method in interface neureka.fluent.building.states.WithShapeOrScalarOrVectorTensor

Define a tensor shape by passing a list of numbers to this method, which represent the shape of the Tensor that should be built.

WithShapeOrScalarOrVector<V> - Interface in neureka.fluent.building.states

WithShapeOrScalarOrVectorOnDevice<V> - Interface in neureka.fluent.building.states

WithShapeOrScalarOrVectorTensor<V> - Interface in neureka.fluent.building.states

withTime(long) - Method in class neureka.optimization.implementations.ADAMFactory

WorkScheduler - Class in neureka.devices.host.concurrent

An API for registering workloads which will be divided into smaller workloads so that they can be executed efficiently by a thread pool...

WorkScheduler() - Constructor for class neureka.devices.host.concurrent.WorkScheduler

WorkScheduler.Divider - Class in neureka.devices.host.concurrent

Divides workloads until they can be processed efficiently and then submits them to a thread pool for execution...

write(V) - Method in interface neureka.devices.Device.Access

Use this to write a single scalar item into the accessed tensor at one or more positions within the tensor.

writeDataTo(DataOutput, Iterator<TargetType>) - Method in interface neureka.dtype.NumericType

This method writes all the target type elements returned by the provided iterator and write them into the provided "DataOutput" stream as bytes.

writeFrom(Object, int) - Method in interface neureka.devices.Device.Access

Use this to write data from an array into the accessed tensor.

writeFrom(Object) - Method in interface neureka.devices.Device.Access

Use this method to write data to the provided tensor, given that the tensor is already stored on this device!

X

XConvLeft - Class in neureka.backend.main.operations.linear

XConvLeft() - Constructor for class neureka.backend.main.operations.linear.XConvLeft

XConvRight - Class in neureka.backend.main.operations.linear

XConvRight() - Constructor for class neureka.backend.main.operations.linear.XConvRight

xor(Tensor<V>) - Method in interface neureka.Tensor

This method is a functionally identical synonym to the Tensor.power(Tensor) method.

xor(double) - Method in interface neureka.Tensor

This method is a functionally identical synonym to the Tensor.power(Tensor) method.

_

_actualize(Tensor<?>) - Method in class neureka.devices.AbstractDevice

_actualize(Tensor<?>) - Method in class neureka.devices.host.CPU

_actualize(Tensor<?>) - Method in class neureka.devices.opencl.OpenCLDevice

_approveExecutionOf(Tensor<?>[], int, Operation) - Method in class neureka.devices.AbstractDevice

This method is the internal approval routine called by its public counterpart and implemented by classes extending this very abstract class.

_approveExecutionOf(Tensor<?>[], int, Operation) - Method in class neureka.devices.host.CPU

_approveExecutionOf(Tensor<?>[], int, Operation) - Method in class neureka.devices.opencl.OpenCLDevice

_arguments - Variable in class neureka.backend.api.Call

Meta arguments which are usually specific to certain operations.

_arity - Variable in class neureka.backend.api.template.operations.AbstractOperation

Arity is the number of arguments or operands that this function or operation takes.

_cacheArray(int[]) - Static method in class neureka.ndim.config.AbstractNDC

This method receives an int array and returns an int array which can either be the one provided or an array found in the global int array cache residing inside this class.

_cached(T) - Static method in class neureka.ndim.config.AbstractNDC

_cleaning(Object, Runnable) - Method in class neureka.devices.AbstractDevice

_dataRef - Variable in class neureka.devices.AbstractDeviceData

_dataType - Variable in class neureka.devices.AbstractDeviceData

_dataTypeOf(Object) - Method in class neureka.devices.AbstractDevice

_dataTypeOf(Object) - Method in class neureka.devices.host.CPU

_dataTypeOf(Object) - Method in class neureka.devices.opencl.OpenCLDevice

_deleteComponents() - Method in class neureka.common.composition.AbstractComponentOwner

This method deletes the array of components of this component owner by nulling the array variable field.

_device - Variable in class neureka.backend.api.Call

This field references the device on which this ExecutionCall should be executed.

_function - Variable in class neureka.backend.api.template.operations.AbstractOperation

An operation may have two ways in which it can describe itself as String within a Function AST.

_getDeriveAt0() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcast

_getDeriveAt0() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastAddition

_getDeriveAt0() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastDivision

_getDeriveAt0() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastModulo

_getDeriveAt0() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastMultiplication

_getDeriveAt0() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastPower

_getDeriveAt0() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastSubtraction

_getDeriveAt0() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastSummation

_getDeriveAt0() - Method in class neureka.backend.main.implementations.broadcast.CPUScalaBroadcastPower

_getDeriveAt0() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcast

_getDeriveAt0() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastAddition

_getDeriveAt0() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastDivision

_getDeriveAt0() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastIdentity

_getDeriveAt0() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastModulo

_getDeriveAt0() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastMultiplication

_getDeriveAt0() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastSubtraction

_getDeriveAt0() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWise

_getDeriveAt0() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWiseAddition

_getDeriveAt0() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWiseDivision

_getDeriveAt0() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWiseModulo

_getDeriveAt0() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWiseMultiplication

_getDeriveAt0() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWisePower

_getDeriveAt0() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWiseSubtraction

_getDeriveAt1() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcast

_getDeriveAt1() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastAddition

_getDeriveAt1() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastDivision

_getDeriveAt1() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastModulo

_getDeriveAt1() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastMultiplication

_getDeriveAt1() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastPower

_getDeriveAt1() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastSubtraction

_getDeriveAt1() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastSummation

_getDeriveAt1() - Method in class neureka.backend.main.implementations.broadcast.CPUScalaBroadcastPower

_getDeriveAt1() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcast

_getDeriveAt1() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastAddition

_getDeriveAt1() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastDivision

_getDeriveAt1() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastIdentity

_getDeriveAt1() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastModulo

_getDeriveAt1() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastMultiplication

_getDeriveAt1() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastSubtraction

_getDeriveAt1() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWise

_getDeriveAt1() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWiseAddition

_getDeriveAt1() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWiseDivision

_getDeriveAt1() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWiseModulo

_getDeriveAt1() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWiseMultiplication

_getDeriveAt1() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWisePower

_getDeriveAt1() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWiseSubtraction

_getFun() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcast

_getFun() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastAddition

_getFun() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastDivision

_getFun() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastModulo

_getFun() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastMultiplication

_getFun() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastPower

_getFun() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastSubtraction

_getFun() - Method in class neureka.backend.main.implementations.broadcast.CPUBroadcastSummation

_getFun() - Method in class neureka.backend.main.implementations.broadcast.CPUScalaBroadcastPower

_getFun() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcast

_getFun() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastAddition

_getFun() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastDivision

_getFun() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastIdentity

_getFun() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastModulo

_getFun() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastMultiplication

_getFun() - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcastSubtraction

_getFun() - Method in class neureka.backend.main.implementations.convolution.AbstractCPUConvolution

_getFun() - Method in class neureka.backend.main.implementations.convolution.CPUConvolution

_getFun() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWise

_getFun() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWiseAddition

_getFun() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWiseDivision

_getFun() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWiseModulo

_getFun() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWiseMultiplication

_getFun() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWisePower

_getFun() - Method in class neureka.backend.main.implementations.elementwise.CPUBiElementWiseSubtraction

_implementations - Variable in class neureka.backend.api.template.algorithms.AbstractDeviceAlgorithm

_inputs - Variable in class neureka.backend.api.Call

The tensor arguments from which an operation will either read or to which it will write.

_isDifferentiable - Variable in class neureka.backend.api.template.operations.AbstractOperation

Certain operations are not differentiable, meaning they cannot participate in neither forward nor reverse mode differentiation.

_isIndexer - Variable in class neureka.backend.api.template.operations.AbstractOperation

This flag determines if this operation is auto-indexing passed input arguments.

_isInline - Variable in class neureka.backend.api.template.operations.AbstractOperation

Inline operations are operations which change the state of the arguments passed to them.

_isOperator - Variable in class neureka.backend.api.template.operations.AbstractOperation

_loadData() - Method in class neureka.devices.file.CSVHandle

_loadData() - Method in class neureka.devices.file.IDXHandle

_log - Variable in class neureka.devices.AbstractDevice

_numberOfDataObjects - Variable in class neureka.devices.AbstractBaseDevice

_numberOfTensors - Variable in class neureka.devices.AbstractBaseDevice

_operator - Variable in class neureka.backend.api.template.operations.AbstractOperation

An operation may have two ways in which it can describe itself as String within a Function AST.

_owner - Variable in class neureka.devices.AbstractDeviceData

_prepareForExecution(ExecutionCall<? extends Device<?>>) - Static method in class neureka.backend.api.template.algorithms.AbstractDeviceAlgorithm

_readAll(Tensor<T>, boolean) - Method in class neureka.devices.AbstractDevice

_readAll(Tensor<T>, boolean) - Method in class neureka.devices.host.CPU

_readAll(Tensor<T>, boolean) - Method in class neureka.devices.opencl.OpenCLDevice

_readArray(Tensor<T>, Class<A>, int, int) - Method in class neureka.devices.AbstractDevice

_readArray(Tensor<T>, Class<A>, int, int) - Method in class neureka.devices.host.CPU

_readArray(Tensor<T>, Class<A>, int, int) - Method in class neureka.devices.opencl.OpenCLDevice

_readItem(Tensor<T>, int) - Method in class neureka.devices.AbstractDevice

_readItem(Tensor<T>, int) - Method in class neureka.devices.host.CPU

_readItem(Tensor<T>, int) - Method in class neureka.devices.opencl.OpenCLDevice

_refCounter - Variable in class neureka.devices.AbstractDeviceData

_removeOrReject(T) - Method in class neureka.backend.api.Extensions

_removeOrReject(T) - Method in class neureka.common.composition.AbstractComponentOwner

An implementation of this method checks if the passed component should be removed from the component collection of this class or its removal should be "rejected".

_removeOrReject(T) - Method in class neureka.math.args.Args

_set(Component<T>) - Method in class neureka.common.composition.AbstractComponentOwner

_setOrReject(T) - Method in class neureka.backend.api.Extensions

_setOrReject(T) - Method in class neureka.common.composition.AbstractComponentOwner

This abstract method ought to be implemented further down the inheritance hierarchy where it's responsibility makes more sense, namely : An implementation of this method checks if the passed component should be added or "rejected" to the component collection of this class.

_setOrReject(T) - Method in class neureka.math.args.Args

_shape - Variable in class neureka.ndim.config.types.permuted.Permuted1DConfiguration

The shape of the NDArray.

_shape - Variable in class neureka.ndim.config.types.simple.Simple1DConfiguration

The shape of the NDArray.

_shape - Variable in class neureka.ndim.config.types.sliced.Sliced1DConfiguration

The shape of the NDArray.

_shape1 - Variable in class neureka.ndim.config.types.permuted.Permuted2DConfiguration

The shape of the NDArray.

_shape1 - Variable in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

The shape of the NDArray.

_shape1 - Variable in class neureka.ndim.config.types.simple.Simple2DConfiguration

The shape of the NDArray.

_shape1 - Variable in class neureka.ndim.config.types.simple.Simple3DConfiguration

The shape of the NDArray.

_shape1 - Variable in class neureka.ndim.config.types.sliced.Sliced2DConfiguration

The shape of the NDArray.

_shape1 - Variable in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

The shape of the NDArray.

_shape2 - Variable in class neureka.ndim.config.types.permuted.Permuted2DConfiguration

_shape2 - Variable in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

_shape2 - Variable in class neureka.ndim.config.types.simple.Simple2DConfiguration

_shape2 - Variable in class neureka.ndim.config.types.simple.Simple3DConfiguration

_shape2 - Variable in class neureka.ndim.config.types.sliced.Sliced2DConfiguration

_shape2 - Variable in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

_shape3 - Variable in class neureka.ndim.config.types.permuted.Permuted3DConfiguration

_shape3 - Variable in class neureka.ndim.config.types.simple.Simple3DConfiguration

_shape3 - Variable in class neureka.ndim.config.types.sliced.Sliced3DConfiguration

_simpleReshape(int[], NDConfiguration) - Static method in class neureka.ndim.config.AbstractNDC

_sizeOccupiedBy(Tensor<T>) - Method in class neureka.devices.AbstractDevice

_sizeOccupiedBy(Tensor<T>) - Method in class neureka.devices.host.CPU

_sizeOccupiedBy(Tensor<T>) - Method in class neureka.devices.opencl.OpenCLDevice

_swap(Tensor<T>, Tensor<T>) - Method in class neureka.devices.AbstractDevice

This method is used internally mostly and should not be used in most cases.

_swap(Tensor<T>, Tensor<T>) - Method in class neureka.devices.host.CPU

_swap(Tensor<T>, Tensor<T>) - Method in class neureka.devices.opencl.OpenCLDevice

_this() - Method in class neureka.common.composition.AbstractComponentOwner

_transferFrom(AbstractComponentOwner<C>) - Method in class neureka.common.composition.AbstractComponentOwner

A component owner might need to exchange components.

_virtualize(Tensor<?>) - Method in class neureka.devices.AbstractDevice

_virtualize(Tensor<?>) - Method in class neureka.devices.host.CPU

_virtualize(Tensor<?>) - Method in class neureka.devices.opencl.OpenCLDevice

_work(int, int) - Method in class neureka.devices.host.concurrent.WorkScheduler

_workloadFor(ExecutionCall<CPU>) - Method in class neureka.backend.main.implementations.broadcast.CPUScalarBroadcast

_writeArray(Tensor<T>, Object, int, int, int) - Method in class neureka.devices.AbstractDevice

_writeArray(Tensor<T>, Object, int, int, int) - Method in class neureka.devices.host.CPU

_writeArray(Tensor<T>, Object, int, int, int) - Method in class neureka.devices.opencl.OpenCLDevice

_writeItem(Tensor<T>, T, int, int) - Method in class neureka.devices.AbstractDevice

_writeItem(Tensor<T>, T, int, int) - Method in class neureka.devices.host.CPU

_writeItem(Tensor<T>, T, int, int) - Method in class neureka.devices.opencl.OpenCLDevice

A B C D E F G H I J K L M N O P Q R S T U V W X _