Module matrix_kernels 

SIMD-accelerated matrix kernels operating on MatrixData.

All functions operate directly on MatrixData (row-major, SIMD-aligned) to avoid the overhead of nested-array extraction in the old matrix.rs.

Functions

matrix_add

Element-wise matrix addition: C = A + B (same dimensions).

matrix_determinant

Matrix determinant via LU decomposition (partial pivoting).

matrix_element_mul

Element-wise (Hadamard) multiplication: C[i,j] = A[i,j] * B[i,j].

matrix_inverse

Matrix inverse via Gauss-Jordan elimination. Only works for square matrices.

matrix_matmul

Matrix multiplication: C = A * B. A is (m x k), B is (k x n), result is (m x n).

matrix_matvec

Matrix-vector multiplication: y = A * v. A is (m x n), v has length n, result has length m.

matrix_scale

Scalar multiplication: C = A * scalar.

matrix_sub

Element-wise matrix subtraction: C = A - B (same dimensions).

matrix_trace

Matrix trace: sum of diagonal elements.

matrix_transpose

Matrix transpose: B = A^T.