`scipy.linalg._flapack:sgbtrf`

Summary

Compute the LU factorization of a banded matrix with partial pivoting (LAPACK sgbtrf).

ab : ndarray: Band storage of shape (2 * kl + ku + 1, n): a[i, j] is held at ab[ku + i - j, j].
kl : int: Number of subdiagonals.
ku : int: Number of superdiagonals.
m : int, optional: Number of rows to factor. Default is shape(ab, 1), a square matrix; a smaller value factors the leading m rows.
n : int, optional: Number of columns; must equal shape(ab, 1).
ldab : int, optional: Leading dimension; must equal shape(ab, 0).
overwrite_ab : int, optional: If nonzero, ab may be overwritten in place. Default is 0.

lu : ndarray: Combined L and U factors in band storage.
ipiv : ndarray: Pivot indices, 0-based, length min(m, n).
info : int: 0 on success; if negative, the -info-th argument had an illegal value; if positive, u[info-1, info-1] is exactly zero and the factor is singular.