Cholesky and LDLT decompositions

RISCV_DSP_ATTRIBUTE riscv_status riscv_mat_cholesky_f16 (const riscv_matrix_instance_f16 *pSrc, riscv_matrix_instance_f16 *pDst)
RISCV_DSP_ATTRIBUTE riscv_status riscv_mat_cholesky_f32 (const riscv_matrix_instance_f32 *pSrc, riscv_matrix_instance_f32 *pDst)
RISCV_DSP_ATTRIBUTE riscv_status riscv_mat_cholesky_f64 (const riscv_matrix_instance_f64 *pSrc, riscv_matrix_instance_f64 *pDst)
RISCV_DSP_ATTRIBUTE riscv_status riscv_mat_ldlt_f32 (const riscv_matrix_instance_f32 *pSrc, riscv_matrix_instance_f32 *pl, riscv_matrix_instance_f32 *pd, uint16_t *pp)
RISCV_DSP_ATTRIBUTE riscv_status riscv_mat_ldlt_f64 (const riscv_matrix_instance_f64 *pSrc, riscv_matrix_instance_f64 *pl, riscv_matrix_instance_f64 *pd, uint16_t *pp)
group MatrixChol

Computes the Cholesky or LL^t decomposition of a matrix.

If the input matrix does not have a decomposition, then the algorithm terminates and returns error status RISCV_MATH_DECOMPOSITION_FAILURE.

Functions

RISCV_DSP_ATTRIBUTE riscv_status riscv_mat_cholesky_f16 (const riscv_matrix_instance_f16 *pSrc, riscv_matrix_instance_f16 *pDst)

Floating-point Cholesky decomposition of positive-definite matrix.

Floating-point Cholesky decomposition of Symmetric Positive Definite Matrix.

If the matrix is ill conditioned or only semi-definite, then it is better using the LDL^t decomposition. The decomposition of A is returning a lower triangular matrix U such that A = L L^t

The destination matrix should be set to 0 before calling the functions because the function may not overwrite all output elements.

Parameters

  • pSrc – [in] points to the instance of the input floating-point matrix structure.

  • pDst – [out] points to the instance of the output floating-point matrix structure.

Returns

The function returns RISCV_MATH_SIZE_MISMATCH, if the dimensions do not match.

Returns

execution status

  • RISCV_MATH_SUCCESS : Operation successful

  • RISCV_MATH_SIZE_MISMATCH : Matrix size check failed

  • RISCV_MATH_DECOMPOSITION_FAILURE : Input matrix cannot be decomposed

RISCV_DSP_ATTRIBUTE riscv_status riscv_mat_cholesky_f32 (const riscv_matrix_instance_f32 *pSrc, riscv_matrix_instance_f32 *pDst)

Floating-point Cholesky decomposition of positive-definite matrix.

Floating-point Cholesky decomposition of Symmetric Positive Definite Matrix.

If the matrix is ill conditioned or only semi-definite, then it is better using the LDL^t decomposition. The decomposition of A is returning a lower triangular matrix L such that A = L L^t

The destination matrix should be set to 0 before calling the functions because the function may not overwrite all output elements.

Parameters

  • pSrc – [in] points to the instance of the input floating-point matrix structure.

  • pDst – [out] points to the instance of the output floating-point matrix structure.

Returns

The function returns RISCV_MATH_SIZE_MISMATCH, if the dimensions do not match.

Returns

execution status

  • RISCV_MATH_SUCCESS : Operation successful

  • RISCV_MATH_SIZE_MISMATCH : Matrix size check failed

  • RISCV_MATH_DECOMPOSITION_FAILURE : Input matrix cannot be decomposed

RISCV_DSP_ATTRIBUTE riscv_status riscv_mat_cholesky_f64 (const riscv_matrix_instance_f64 *pSrc, riscv_matrix_instance_f64 *pDst)

Floating-point Cholesky decomposition of positive-definite matrix.

Floating-point Cholesky decomposition of Symmetric Positive Definite Matrix.

If the matrix is ill conditioned or only semi-definite, then it is better using the LDL^t decomposition. The decomposition of A is returning a lower triangular matrix L such that A = L L^t

The destination matrix should be set to 0 before calling the functions because the function may not overwrite all output elements.

Parameters

  • pSrc – [in] points to the instance of the input floating-point matrix structure.

  • pDst – [out] points to the instance of the output floating-point matrix structure.

Returns

The function returns RISCV_MATH_SIZE_MISMATCH, if the dimensions do not match.

Returns

execution status

  • RISCV_MATH_SUCCESS : Operation successful

  • RISCV_MATH_SIZE_MISMATCH : Matrix size check failed

  • RISCV_MATH_DECOMPOSITION_FAILURE : Input matrix cannot be decomposed

RISCV_DSP_ATTRIBUTE riscv_status riscv_mat_ldlt_f32 (const riscv_matrix_instance_f32 *pSrc, riscv_matrix_instance_f32 *pl, riscv_matrix_instance_f32 *pd, uint16_t *pp)

Floating-point LDL^t decomposition of positive semi-definite matrix.

Floating-point LDL decomposition of Symmetric Positive Semi-Definite Matrix.

Computes the LDL^t decomposition of a matrix A such that P A P^t = L D L^t.

Parameters

  • pSrc – [in] points to the instance of the input floating-point matrix structure.

  • pl – [out] points to the instance of the output floating-point triangular matrix structure.

  • pd – [out] points to the instance of the output floating-point diagonal matrix structure.

  • pp – [out] points to the instance of the output floating-point permutation vector.

Returns

The function returns RISCV_MATH_SIZE_MISMATCH, if the dimensions do not match.

Returns

execution status

  • RISCV_MATH_SUCCESS : Operation successful

  • RISCV_MATH_SIZE_MISMATCH : Matrix size check failed

  • RISCV_MATH_DECOMPOSITION_FAILURE : Input matrix cannot be decomposed

RISCV_DSP_ATTRIBUTE riscv_status riscv_mat_ldlt_f64 (const riscv_matrix_instance_f64 *pSrc, riscv_matrix_instance_f64 *pl, riscv_matrix_instance_f64 *pd, uint16_t *pp)

Floating-point LDL^t decomposition of positive semi-definite matrix.

Floating-point LDL decomposition of Symmetric Positive Semi-Definite Matrix.

Computes the LDL^t decomposition of a matrix A such that P A P^t = L D L^t.

Parameters

  • pSrc – [in] points to the instance of the input floating-point matrix structure.

  • pl – [out] points to the instance of the output floating-point triangular matrix structure.

  • pd – [out] points to the instance of the output floating-point diagonal matrix structure.

  • pp – [out] points to the instance of the output floating-point permutation vector.

Returns

The function returns RISCV_MATH_SIZE_MISMATCH, if the dimensions do not match.

Returns

execution status

  • RISCV_MATH_SUCCESS : Operation successful

  • RISCV_MATH_SIZE_MISMATCH : Matrix size check failed

  • RISCV_MATH_DECOMPOSITION_FAILURE : Input matrix cannot be decomposed