org.nd4j.linalg.fft

Class FFT

java.lang.Object

org.nd4j.linalg.fft.FFT

public class FFT
extends Object

FFT and IFFT

Author:

Adam Gibson

Constructor Summary

Constructors

Constructor and Description
FFT()

Method Summary

Modifier and Type	Method and Description
static IComplexNDArray	fft(IComplexNDArray inputC) 1d discrete fourier op, note that this will throw an exception if the passed in input isn't a vector.
static IComplexNDArray	fft(IComplexNDArray inputC, int numElements) 1d discrete fourier op, note that this will throw an exception if the passed in input isn't a vector.
static IComplexNDArray	fft(IComplexNDArray inputC, int numElements, int dimension) 1d discrete fourier op, note that this will throw an exception if the passed in input isn't a vector.
static IComplexNDArray	fft(INDArray input) 1d discrete fourier op, note that this will throw an exception if the passed in input isn't a vector.
static IComplexNDArray	fft(INDArray transform, int numElements) FFT along a particular dimension
static IComplexNDArray	fft(INDArray transform, int numElements, int dimension) FFT along a particular dimension
static IComplexNDArray	fftn(IComplexNDArray transform) FFT on the whole array (n is equal the first dimension shape)
static IComplexNDArray	fftn(IComplexNDArray transform, int dimension, int numElements) Performs FFT along the first non singleton dimension of op.
static IComplexNDArray	fftn(INDArray transform) FFT on the whole array (n is equal the first dimension shape)
static IComplexNDArray	fftn(INDArray transform, int dimension, int numElements) Computes the fft along the first non singleton dimension of op when it is a matrix
static IComplexNDArray	ifft(IComplexNDArray inputC) 1d discrete fourier op, note that this will throw an exception if the passed in input isn't a vector.
static IComplexNDArray	ifft(IComplexNDArray inputC, int numElements, int dimension) 1d discrete fourier op, note that this will throw an exception if the passed in input isn't a vector.
static IComplexNDArray	ifft(INDArray transform, int numElements) FFT along a particular dimension
static IComplexNDArray	ifft(INDArray transform, int numElements, int dimension) IFFT along a particular dimension
static IComplexNDArray	ifftn(IComplexNDArray transform)
static IComplexNDArray	ifftn(IComplexNDArray transform, int dimension, int numElements) ND IFFT
static IComplexNDArray	ifftn(INDArray transform, int dimension, int numElements) ND IFFT, computes along the first on singleton dimension of op
static IComplexNDArray	irfft(IComplexNDArray arr)
static IComplexNDArray	irfft(IComplexNDArray arr, int dimension)
static IComplexNDArray	irfftn(IComplexNDArray arr)
static IComplexNDArray	rawfft(IComplexNDArray transform, int n, int dimension) Underlying fft algorithm
static IComplexNDArray	rawfftn(IComplexNDArray transform, int[] shape, int[] axes)
static IComplexNDArray	rawifft(IComplexNDArray transform, int dimension)
static IComplexNDArray	rawifft(IComplexNDArray transform, int n, int dimension)
static IComplexNDArray	rawifftn(IComplexNDArray transform, int[] shape, int[] axes)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

FFT

public FFT()

Method Detail

fft

public static IComplexNDArray fft(INDArray transform,
                                  int numElements)

FFT along a particular dimension

Parameters:

transform - the ndarray to op

numElements - the desired number of elements in each fft

Returns:

the ffted output

fft

public static IComplexNDArray fft(IComplexNDArray inputC)

1d discrete fourier op, note that this will throw an exception if the passed in input isn't a vector. See matlab's fft2 for more information

Parameters:

inputC - the input to op

Returns:

the the discrete fourier op of the passed in input

fft

public static IComplexNDArray fft(INDArray input)

1d discrete fourier op, note that this will throw an exception if the passed in input isn't a vector. See matlab's fft2 for more information

Parameters:

input - the input to op

Returns:

the the discrete fourier op of the passed in input

fft

public static IComplexNDArray fft(INDArray transform,
                                  int numElements,
                                  int dimension)

FFT along a particular dimension

Parameters:

transform - the ndarray to op

numElements - the desired number of elements in each fft

Returns:

the ffted output

fft

public static IComplexNDArray fft(IComplexNDArray inputC,
                                  int numElements)

1d discrete fourier op, note that this will throw an exception if the passed in input isn't a vector. See matlab's fft2 for more information

Parameters:

inputC - the input to op

Returns:

the the discrete fourier op of the passed in input

fft

public static IComplexNDArray fft(IComplexNDArray inputC,
                                  int numElements,
                                  int dimension)

1d discrete fourier op, note that this will throw an exception if the passed in input isn't a vector. See matlab's fft2 for more information

Parameters:

inputC - the input to op

Returns:

the the discrete fourier op of the passed in input

ifft

public static IComplexNDArray ifft(INDArray transform,
                                   int numElements,
                                   int dimension)

IFFT along a particular dimension

Parameters:

transform - the ndarray to op

numElements - the desired number of elements in each fft

dimension - the dimension to do fft along

Returns:

the iffted output

ifft

public static IComplexNDArray ifft(IComplexNDArray inputC)

1d discrete fourier op, note that this will throw an exception if the passed in input isn't a vector. See matlab's fft2 for more information

Parameters:

inputC - the input to op

Returns:

the the discrete fourier op of the passed in input

ifft

public static IComplexNDArray ifft(INDArray transform,
                                   int numElements)

FFT along a particular dimension

Parameters:

transform - the ndarray to op

numElements - the desired number of elements in each fft

Returns:

the ffted output

ifft

public static IComplexNDArray ifft(IComplexNDArray inputC,
                                   int numElements,
                                   int dimension)

1d discrete fourier op, note that this will throw an exception if the passed in input isn't a vector. See matlab's fft2 for more information

Parameters:

inputC - the input to op

Returns:

the the discrete fourier op of the passed in input

ifftn

public static IComplexNDArray ifftn(INDArray transform,
                                    int dimension,
                                    int numElements)

ND IFFT, computes along the first on singleton dimension of op

Parameters:

transform - the ndarray to op

dimension - the dimension to iterate along

numElements - the desired number of elements in each fft

Returns:

the reverse ifft of the passed in array

irfftn

public static IComplexNDArray irfftn(IComplexNDArray arr)

irfft

public static IComplexNDArray irfft(IComplexNDArray arr,
                                    int dimension)

Parameters:

arr -

dimension -

Returns:

irfft

public static IComplexNDArray irfft(IComplexNDArray arr)

Parameters:

arr -

Returns:

ifftn

public static IComplexNDArray ifftn(IComplexNDArray transform,
                                    int dimension,
                                    int numElements)

ND IFFT

Parameters:

transform - the ndarray to op

dimension - the dimension to iterate along

numElements - the desired number of elements in each fft

Returns:

the transformed array

fftn

public static IComplexNDArray fftn(IComplexNDArray transform,
                                   int dimension,
                                   int numElements)

Performs FFT along the first non singleton dimension of op. This means

Parameters:

transform - the ndarray to op

dimension - the dimension to iterate along

numElements - the desired number of elements in each fft along each dimension from each slice (note: each slice)

Returns:

the transformed array

fftn

public static IComplexNDArray fftn(INDArray transform,
                                   int dimension,
                                   int numElements)

Computes the fft along the first non singleton dimension of op when it is a matrix

Parameters:

transform - the ndarray to op

dimension - the dimension to do fft along

numElements - the desired number of elements in each fft

Returns:

the fft of the specified ndarray

fftn

public static IComplexNDArray fftn(INDArray transform)

FFT on the whole array (n is equal the first dimension shape)

Parameters:

transform - the matrix to op

Returns:

the ffted array

fftn

public static IComplexNDArray fftn(IComplexNDArray transform)

FFT on the whole array (n is equal the first dimension shape)

Parameters:

transform - the matrix to op

Returns:

the ffted array

ifftn

public static IComplexNDArray ifftn(IComplexNDArray transform)

rawifftn

public static IComplexNDArray rawifftn(IComplexNDArray transform,
                                       int[] shape,
                                       int[] axes)

rawfftn

public static IComplexNDArray rawfftn(IComplexNDArray transform,
                                      int[] shape,
                                      int[] axes)

rawfft

public static IComplexNDArray rawfft(IComplexNDArray transform,
                                     int n,
                                     int dimension)

Underlying fft algorithm

Parameters:

transform - the ndarray to op

n - the desired number of elements

dimension - the dimension to do fft along

Returns:

the transformed ndarray

rawifft

public static IComplexNDArray rawifft(IComplexNDArray transform,
                                      int n,
                                      int dimension)

rawifft

public static IComplexNDArray rawifft(IComplexNDArray transform,
                                      int dimension)