matrix-linear-algebra

Dense matrix math for Node with no dependencies. Matrices are plain arrays of rows (number[][]); vectors are flat number[]. Determinant, inverse and solve all share a single LU decomposition with partial pivoting.

Install

npm install matrix-linear-algebra

Usage

import { multiply, determinant, inverse, solve } from "matrix-linear-algebra";

multiply([[1, 2], [3, 4]], [[5, 6], [7, 8]]);
// [[19, 22], [43, 50]]

determinant([[6, 1, 1], [4, -2, 5], [2, 8, 7]]); // -306

// Solve  2x + y = 5,  x + 3y = 10
solve([[2, 1], [1, 3]], [5, 10]); // [1, 3]

inverse([[4, 7], [2, 6]]); // [[0.6, -0.7], [-0.2, 0.4]]

API

• shape(a) → [rows, cols]
• identity(n)
• transpose(a)
• add(a, b), sub(a, b), scale(a, k)
• multiply(a, b) — matrix product
• matVec(a, v) — matrix times a flat vector
• determinant(a) — returns exactly 0 for singular matrices
• solve(a, b) — solve A x = b
• inverse(a) — throws for singular matrices
• trace(a)
• equals(a, b, tol = 1e-9) — tolerant elementwise comparison

Numerics

Row operations use partial pivoting for stability. Determinant and inverse do not attempt symbolic exactness — expect floating-point rounding on ill-conditioned inputs.

License

Apache-2.0