Package deepboof.misc

Class TensorFactory_F32

java.lang.Object

deepboof.misc.TensorFactory_F32

public class TensorFactory_F32
extends Object

Various functions for unit tests

Constructor Summary

Constructors

Constructor	Description
TensorFactory_F32()

Method Summary

Modifier and Type	Method	Description
static Tensor_F32	random(Random rand, boolean subTensor, int... shape)	Creates a random tensor with the specified shape and values from -1 to 1
static Tensor_F32	randomMM(Random rand, boolean subTensor, float min, float max, int... shape)	Creates a random tensor with the specified shape and value range
static List<Tensor_F32>	randomMM(Random rand, boolean subTensor, float min, float max, List<int[]> shapes)	Creates a random tensor with the specified shape and value range
static void	randomMM(Random rand, float min, float max, Tensor_F32 tensor)	Fills the tensor with random numbers selected from a uniform distribution.
static Tensor_F32	zeros(Random rand, int... shape)	Generate a zeros tensor with the option for a sub-tensor

Methods inherited from class java.lang.Object

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

Constructor Details

TensorFactory_F32

public TensorFactory_F32()

Method Details

zeros

public static Tensor_F32 zeros(Random rand,
 int... shape)

Generate a zeros tensor with the option for a sub-tensor

Parameters:

rand - If you wish to generate a sub-matrix pass in this RNG and it will randomly offset the data. null for regular tensor

shape - Shape of the tensor

Returns:

tensor

random

public static Tensor_F32 random(Random rand,
 boolean subTensor,
 int... shape)

Creates a random tensor with the specified shape and values from -1 to 1

Parameters:

rand - Random number generator

subTensor - Should it be a sub-tensor or not?

shape - Shape of the tensor

Returns:

The random tensor

randomMM

public static Tensor_F32 randomMM(Random rand,
 boolean subTensor,
 float min,
 float max,
 int... shape)

Creates a random tensor with the specified shape and value range

Parameters:

rand - Random number generator

subTensor - Should it be a sub-tensor or not?

min - Minimum value of each element

max - Maximum value of each element

shape - Shape of the tensor

Returns:

The random tensor

randomMM

public static List<Tensor_F32> randomMM(Random rand,
 boolean subTensor,
 float min,
 float max,
 List<int[]> shapes)

Creates a random tensor with the specified shape and value range

Parameters:

rand - Random number generator

subTensor - Should it be a sub-tensor or not?

min - Minimum value of each element

max - Maximum value of each element

shapes - Shapes of the tensors

Returns:

The random tensor

randomMM

public static void randomMM(Random rand,
 float min,
 float max,
 Tensor_F32 tensor)

Fills the tensor with random numbers selected from a uniform distribution.

Parameters:

rand - Random number generator

min - min value, inclusive

max - max value, inclusive

tensor - Tensor that is to be filled.