af_accum

The C interface for accum. More...

afError	af_accum_S (float *d_out, unsigned ndims, unsigned *dims, const float *d_in, unsigned batch, int dimension, int type, bool isupcast)
afError	af_accum_D (double *d_out, unsigned ndims, unsigned *dims, const double *d_in, unsigned batch, int dimension, int type)
	Double precision, inclusive sum. More...
afError	af_accum_I (int *d_out, unsigned ndims, unsigned *dims, const int *d_in, unsigned batch, int dimension, int type)
	32-bit signed integer, inclusive sum More...
afError	af_accum_U (unsigned *d_out, unsigned ndims, unsigned *dims, const unsigned *d_in, unsigned batch, int dimension, int type)
	32-bit unsigned integer, inclusive sum More...
afError	af_accum_C (cfloat *d_out, unsigned ndims, unsigned *dims, const cfloat *d_in, unsigned batch, int dimension, int type, bool isupcast)
	Complex, Single precision, inclusive sum. More...
afError	af_accum_Z (cdouble *d_out, unsigned ndims, unsigned *dims, const cdouble *d_in, unsigned batch, int dimension, int type)
	Complex, Double precision, inclusive sum. More...

Detailed Description

The C interface for accum.

Returns the accumulated sum of a set of elements

Function Documentation

afError af_accum_S	(	float *	d_out,
		unsigned	ndims,
		unsigned *	dims,
		const float *	d_in,
		unsigned	batch,
		int	dimension,
		int	type,
		bool	isupcast
	)

Parameters

	d_out	Output Matrix
	ndims	Number dimensions for the input
	dims	Dimensions of the input
	d_in	Input Matrix
	batch	Number of tiles in input matrix
	dimension	Dimension along which result is computed (0 along column, 1 along row)
	type	Type of accumulation (0 SUM, 1 PROD)
[in]	isupcast	Accumulate internally in double-precision if trueSingle precision, inclusive sum

afError af_accum_D	(	double *	d_out,
		unsigned	ndims,
		unsigned *	dims,
		const double *	d_in,
		unsigned	batch,
		int	dimension,
		int	type
	)

Double precision, inclusive sum.

afError af_accum_I	(	int *	d_out,
		unsigned	ndims,
		unsigned *	dims,
		const int *	d_in,
		unsigned	batch,
		int	dimension,
		int	type
	)

32-bit signed integer, inclusive sum

afError af_accum_U	(	unsigned *	d_out,
		unsigned	ndims,
		unsigned *	dims,
		const unsigned *	d_in,
		unsigned	batch,
		int	dimension,
		int	type
	)

32-bit unsigned integer, inclusive sum

afError af_accum_C	(	cfloat *	d_out,
		unsigned	ndims,
		unsigned *	dims,
		const cfloat *	d_in,
		unsigned	batch,
		int	dimension,
		int	type,
		bool	isupcast
	)

Complex, Single precision, inclusive sum.

afError af_accum_Z	(	cdouble *	d_out,
		unsigned	ndims,
		unsigned *	dims,
		const cdouble *	d_in,
		unsigned	batch,
		int	dimension,
		int	type
	)

Complex, Double precision, inclusive sum.