af_sum

the C interface for sum More...

All Elements
Parameters [out] count Scalar and on host [in] numel The number of elements in the vector. [in] mat The input stream
afError	af_sum_vector_S (float *h_dst, unsigned numel, const float *d_src)
	sum The value of a single precision vector More...
afError	af_sum_vector_B (float *h_dst, unsigned numel, const bool *d_src)
	sum The value of a boolean vector More...
afError	af_sum_vector_I (int *h_dst, unsigned numel, const int *d_src)
	sum The value of a 32-bit signed integer vector More...
afError	af_sum_vector_U (unsigned *h_dst, unsigned numel, const unsigned *d_src)
	sum The value of a 32-bit unsigned integer vector More...
afError	af_sum_vector_C (cfloat *h_dst, unsigned numel, const cfloat *d_src)
	sum The value of a single precision, complex vector More...
afError	af_sum_vector_D (double *h_dst, unsigned numel, const double *d_src)
	sum The value of a double precision vector More...
afError	af_sum_vector_Z (cdouble *h_dst, unsigned numel, const cdouble *d_src)
	sum The value of a double precision, complex vector More...

Along a Dimension
Parameters [out] count vector and on device [in] numel The number of elements in the vector. [in] mat The input stream [in] batch The number of input tiles.
afError	af_sum_S (float *d_dst, unsigned ndims, const unsigned *dims, const float *d_src, int dim)
	sum the value along a dimension in single precision matrix More...
afError	af_sum_B (float *d_dst, unsigned ndims, const unsigned *dims, const bool *d_src, int dim)
	sum the value along a dimension in boolean matrix More...
afError	af_sum_I (int *d_dst, unsigned ndims, const unsigned *dims, const int *d_src, int dim)
	sum the value along a dimension in 32-bit signed integer matrix More...
afError	af_sum_U (unsigned *d_dst, unsigned ndims, const unsigned *dims, const unsigned *d_src, int dim)
	sum the value along a dimension in 32-bit unsigned integer matrix More...
afError	af_sum_C (cfloat *d_dst, unsigned ndims, const unsigned *dims, const cfloat *d_src, int dim)
	sum the value along a dimension in single precision, complex matrix More...
afError	af_sum_D (double *d_dst, unsigned ndims, const unsigned *dims, const double *d_src, int dim)
	sum the value along a dimension in double precision matrix More...
afError	af_sum_Z (cdouble *d_dst, unsigned ndims, const unsigned *dims, const cdouble *d_src, int dim)
	sum the value along a dimension in double precision, complex matrix More...

Detailed Description

the C interface for sum

Sum is a reduction operation that computes the arithmetic
sum of the elements in an array.

The first sum function sums elements along a specific dimension
of the array. The defualt dimension is the first non-singleton
dimension.

array a = constant(1,5,5);
//a = 1 1 1 1 1
//  = 1 1 1 1 1
//  = 1 1 1 1 1
//  = 1 1 1 1 1
//  = 1 1 1 1 1

b = sum(a); //b = [5]
            //    [5]
            //    [5]
            //    [5]
            //    [5]

The second form of sum is templated, and returns the sum of all array elements to the host.

float hx[] = { -3, -5, 2, -9 };
array x1(2,2, hx); // x1 = -3  2
                   //      -5 -9
printf("%g\n", sum<float>(x1)); // -15

Function Documentation

afError af_sum_vector_S	(	float *	h_dst,
		unsigned	numel,
		const float *	d_src
	)

sum The value of a single precision vector

afError af_sum_vector_B	(	float *	h_dst,
		unsigned	numel,
		const bool *	d_src
	)

sum The value of a boolean vector

afError af_sum_vector_I	(	int *	h_dst,
		unsigned	numel,
		const int *	d_src
	)

sum The value of a 32-bit signed integer vector

afError af_sum_vector_U	(	unsigned *	h_dst,
		unsigned	numel,
		const unsigned *	d_src
	)

sum The value of a 32-bit unsigned integer vector

afError af_sum_vector_C	(	cfloat *	h_dst,
		unsigned	numel,
		const cfloat *	d_src
	)

sum The value of a single precision, complex vector

afError af_sum_vector_D	(	double *	h_dst,
		unsigned	numel,
		const double *	d_src
	)

sum The value of a double precision vector

afError af_sum_vector_Z	(	cdouble *	h_dst,
		unsigned	numel,
		const cdouble *	d_src
	)

sum The value of a double precision, complex vector

afError af_sum_S	(	float *	d_dst,
		unsigned	ndims,
		const unsigned *	dims,
		const float *	d_src,
		int	dim
	)

sum the value along a dimension in single precision matrix

afError af_sum_B	(	float *	d_dst,
		unsigned	ndims,
		const unsigned *	dims,
		const bool *	d_src,
		int	dim
	)

sum the value along a dimension in boolean matrix

afError af_sum_I	(	int *	d_dst,
		unsigned	ndims,
		const unsigned *	dims,
		const int *	d_src,
		int	dim
	)

sum the value along a dimension in 32-bit signed integer matrix

afError af_sum_U	(	unsigned *	d_dst,
		unsigned	ndims,
		const unsigned *	dims,
		const unsigned *	d_src,
		int	dim
	)

sum the value along a dimension in 32-bit unsigned integer matrix

afError af_sum_C	(	cfloat *	d_dst,
		unsigned	ndims,
		const unsigned *	dims,
		const cfloat *	d_src,
		int	dim
	)

sum the value along a dimension in single precision, complex matrix

afError af_sum_D	(	double *	d_dst,
		unsigned	ndims,
		const unsigned *	dims,
		const double *	d_src,
		int	dim
	)

sum the value along a dimension in double precision matrix

afError af_sum_Z	(	cdouble *	d_dst,
		unsigned	ndims,
		const unsigned *	dims,
		const cdouble *	d_src,
		int	dim
	)

sum the value along a dimension in double precision, complex matrix