Gradient

Gradient function. More...

Device pointer interface: Gradient along a dimension
Parameters: [out] d_dx Output gradient along the X dimension [out] d_dy Output gradient along the Y dimension [in] nx Number of rows in the input [in] ny Number of columns in the input [in] d_in The input the the gradient function [in] batch The number of input tiles being handled
afError	af_grad_B (bool *d_dx, bool *d_dy, unsigned nx, unsigned ny, const bool *d_in, unsigned batch)
	Boolean gradient function.
afError	af_grad_U (unsigned *d_dx, unsigned *d_dy, unsigned nx, unsigned ny, const unsigned *d_in, unsigned batch)
	Unsigne integer gradient function.
afError	af_grad_I (int *d_dx, int *d_dy, unsigned nx, unsigned ny, const int *d_in, unsigned batch)
	Integer gradient function.
afError	af_grad_S (float *d_dx, float *d_dy, unsigned nx, unsigned ny, const float *d_in, unsigned batch)
	Single precision, floating point gradient funciton.
afError	af_grad_D (double *d_dx, double *d_dy, unsigned nx, unsigned ny, const double *d_in, unsigned batch)
	Double precision, floating point gradient funciton.

---

Detailed Description

Gradient function.

---

Function Documentation

afError af_grad_B	(	bool *	d_dx,
		bool *	d_dy,
		unsigned	nx,
		unsigned	ny,
		const bool *	d_in,
		unsigned	batch
	)

Boolean gradient function.

afError af_grad_U	(	unsigned *	d_dx,
		unsigned *	d_dy,
		unsigned	nx,
		unsigned	ny,
		const unsigned *	d_in,
		unsigned	batch
	)

Unsigne integer gradient function.

afError af_grad_I	(	int *	d_dx,
		int *	d_dy,
		unsigned	nx,
		unsigned	ny,
		const int *	d_in,
		unsigned	batch
	)

Integer gradient function.

afError af_grad_S	(	float *	d_dx,
		float *	d_dy,
		unsigned	nx,
		unsigned	ny,
		const float *	d_in,
		unsigned	batch
	)

Single precision, floating point gradient funciton.

afError af_grad_D	(	double *	d_dx,
		double *	d_dy,
		unsigned	nx,
		unsigned	ny,
		const double *	d_in,
		unsigned	batch
	)

Double precision, floating point gradient funciton.