scikits.umfpack.umfpack.UmfpackContext.lu

UmfpackContext.lu(mtx)

Perform LU decomposition.

For a given matrix A, the decomposition satisfies:

LU = PRAQ        when do_recip is true
LU = P(R^-1)AQ   when do_recip is false
Parameters:

mtx : scipy.sparse.csc_matrix or scipy.sparse.csr_matrix

Input.

Returns:

L : csr_matrix

Lower triangular m-by-min(m,n) CSR matrix

U : csc_matrix

Upper triangular min(m,n)-by-n CSC matrix

P : ndarray

Vector of row permutations

Q : ndarray

Vector of column permutations

R : ndarray

Vector of diagonal row scalings

do_recip : bool

Whether R is R^-1 or R