v?copysign

Interface Definition

C interface:

void vscopysign(const int len, const float* src1, const float* src2, float* dst);

void vdcopysign(const int len, const double* src1, const double* src2, double* dst);

Parameters

Parameter	Type	Description	Input/Output
len	Integer	Number of elements in the input vector. If len ≤ 0, the system displays a message indicating that the value of len is invalid.	Input
src1	Single-precision floating-point type for vscopysign Double-precision floating-point type for vdcopysign	Input vector src1 with length len. If the pointer is null, the system prompts a null pointer error.	Input
src2	Single-precision floating-point type for vscopysign Double-precision floating-point type for vdcopysign	Input vector src2 with length len. If the pointer is null, the system prompts a null pointer error.	Input
dst	Single-precision floating-point type for vscopysign Double-precision floating-point type for vdcopysign	Output vector dst with length len. If the pointer is null, the system prompts a null pointer error.	Output

Return Value

Each operation value returns the result obtained by copying the sign bit of x2 to x1.
For special values, see the following table.
Input Value 1 (src1)

Sign Bit of the Input Value 2 (src2)

Output Value (dst)

±inf

0

+inf

±inf

1

-inf

nan

0

nan

nan

1

-nan

Input Value 1 (src1)	Sign Bit of the Input Value 2 (src2)	Output Value (dst)
±inf	0	+inf
±inf	1	-inf
nan	0	nan
nan	1	-nan

Dependency

C: "kvml.h"

Example

C interface:

    int i, len = 8; 
    float src1[8] = {-1.5f, -0.0f, 0.0f, 1.0f, 1.5f, -INFINITY, INFINITY, NAN};
    float src2[8] = {1.5f, 0.0f, +0.0f, -1.0f, -1.5f, +INFINITY, -INFINITY, -NAN}; 
    float* dst = (float*)malloc(sizeof(float) * len); 
    if (dst == NULL) { 
        printf("Malloc Failed!\n"); 
        return 0;
    }
    vscopysign(len, src1, src2, dst);

    /** 
     *  Output dst: 
     *     1.500000 0.000000 0.000000 -1.000000 -1.500000 inf -inf -nan
     */

Parent topic: Function Syntax