Complex Magnitude

void riscv_cmplx_mag_f16(const float16_t *pSrc, float16_t *pDst, uint32_t numSamples)
void riscv_cmplx_mag_f32(const float32_t *pSrc, float32_t *pDst, uint32_t numSamples)
void riscv_cmplx_mag_f64(const float64_t *pSrc, float64_t *pDst, uint32_t numSamples)
void riscv_cmplx_mag_fast_q15(const q15_t *pSrc, q15_t *pDst, uint32_t numSamples)
void riscv_cmplx_mag_q15(const q15_t *pSrc, q15_t *pDst, uint32_t numSamples)
void riscv_cmplx_mag_q31(const q31_t *pSrc, q31_t *pDst, uint32_t numSamples)
group cmplx_mag

Computes the magnitude of the elements of a complex data vector.

The pSrc points to the source data and pDst points to the where the result should be written. numSamples specifies the number of complex samples in the input array and the data is stored in an interleaved fashion (real, imag, real, imag, …). The input array has a total of 2*numSamples values; the output array has a total of numSamples values.

The underlying algorithm is used:

There are separate functions for floating-point, Q15, and Q31 data types.

Functions

void riscv_cmplx_mag_f16(const float16_t *pSrc, float16_t *pDst, uint32_t numSamples)

Floating-point complex magnitude.

Parameters

  • pSrc – [in] points to input vector

  • pDst – [out] points to output vector

  • numSamples – [in] number of samples in each vector

Returns

none

void riscv_cmplx_mag_f32(const float32_t *pSrc, float32_t *pDst, uint32_t numSamples)

Floating-point complex magnitude.

Parameters

  • pSrc – [in] points to input vector

  • pDst – [out] points to output vector

  • numSamples – [in] number of samples in each vector

Returns

none

void riscv_cmplx_mag_f64(const float64_t *pSrc, float64_t *pDst, uint32_t numSamples)

Floating-point complex magnitude.

Parameters

  • pSrc – [in] points to input vector

  • pDst – [out] points to output vector

  • numSamples – [in] number of samples in each vector

Returns

none

void riscv_cmplx_mag_fast_q15(const q15_t *pSrc, q15_t *pDst, uint32_t numSamples)

Q15 complex magnitude.

Scaling and Overflow Behavior

The function implements 1.15 by 1.15 multiplications and finally output is converted into 2.14 format. Fast functions are less accurate. This function will tend to clamp to 0 the too small values. So sqrt(x*x) = x will not always be true.

Parameters

  • pSrc – [in] points to input vector

  • pDst – [out] points to output vector

  • numSamples – [in] number of samples in each vector

Returns

none

void riscv_cmplx_mag_q15(const q15_t *pSrc, q15_t *pDst, uint32_t numSamples)

Q15 complex magnitude.

Scaling and Overflow Behavior

The function implements 1.15 by 1.15 multiplications and finally output is converted into 2.14 format.

Parameters

  • pSrc – [in] points to input vector

  • pDst – [out] points to output vector

  • numSamples – [in] number of samples in each vector

Returns

none

void riscv_cmplx_mag_q31(const q31_t *pSrc, q31_t *pDst, uint32_t numSamples)

Q31 complex magnitude.

Scaling and Overflow Behavior

The function implements 1.31 by 1.31 multiplications and finally output is converted into 2.30 format. Input down scaling is not required.

Parameters

  • pSrc – [in] points to input vector

  • pDst – [out] points to output vector

  • numSamples – [in] number of samples in each vector

Returns

none