grad

Calculate the gradients of the input. More...

Functions
AFAPI void	grad (array &dx, array &dy, const array &in)
	C++ Interface for calculating the gradients. More...
AFAPI af_err	af_gradient (af_array *dx, af_array *dy, const af_array in)
	C Interface for calculating the gradients. More...

Detailed Description

Calculate the gradients of the input.

dx is the gradient along the 1st dimension of in. dy is the gradient along the 2nd dimension of in.

The gradients along the first and second dimensions are calculated simultaneously

array in = randu(dim4(5, 3));
array dx, dy;
grad(dx, dy, in);

// in [5 3 1 1]
//  0.0000     0.2190     0.3835
//  0.1315     0.0470     0.5194
//  0.7556     0.6789     0.8310
//  0.4587     0.6793     0.0346
//  0.5328     0.9347     0.0535
//
// dx [5 3 1 1]
//  0.1315    -0.1719     0.1359
//  0.3778     0.2300     0.2237
//  0.1636     0.3161    -0.2424
// -0.1114     0.1279    -0.3888
//  0.0741     0.2554     0.0189
//
// dy [5 3 1 1]
//  0.2190     0.1917     0.1645
// -0.0845     0.1939     0.4724
// -0.0767     0.0377     0.1521
//  0.2206    -0.2120    -0.6447
//  0.4019    -0.2397    -0.8812

Function Documentation

AFAPI af_err af_gradient	(	af_array *	dx,
		af_array *	dy,
		const af_array	in
	)

C Interface for calculating the gradients.

Parameters

[out]	dx	the gradient along first dimension
[out]	dy	the gradient along second dimension
[in]	in	is the input array

Returns

AF_SUCCESS if the color transformation is successful, otherwise an appropriate error code is returned.

AFAPI void af::grad	(	array &	dx,
		array &	dy,
		const array &	in
	)

C++ Interface for calculating the gradients.

Parameters

[out]	dx	the gradient along first dimension
[out]	dy	the gradient along second dimension
[in]	in	is the input array

Examples:

computer_vision/harris.cpp, machine_learning/deep_belief_net.cpp, and machine_learning/neural_network.cpp.