scipy.linalg.lapack.sgebal

scipy.linalg.lapack.sgebal(a, scale=0, permute=0, overwrite_a=0) = <flapack function sgebal>

Balance a general matrix to improve eigenvalue accuracy (LAPACK sgebal).

Parameters:

andarray

Square matrix of shape (n, n).

scaleint, optional

If nonzero, rows and columns are scaled to make their norms comparable. Default is 0.

permuteint, optional

If nonzero, the matrix is permuted to isolate eigenvalues already in the corners. Default is 0.

overwrite_aint, optional

If nonzero, a may be overwritten in place. Default is 0.

Returns:

bandarray

The balanced matrix.

loint

First index of the balanced middle block, 0-based.

hiint

Last index of the balanced middle block, 0-based.

pivscalendarray

Permutation and scaling applied, packed as LAPACK returns them.

infoint

0 on success; if negative, the -info-th argument had an illegal value.