[Octave-bug-tracker] [bug #56232] Octave crash when inverting an empty sparse matrix.

[Marco Caliari]

Follow-up Comment #9, bug #56232 (project octave):

It is not C-C to be marked as triangular. SparseMatrix::inverse is called in
dSparse.cc. The branch if (! mattype.ishermitian ()) is entered,
octave::math::sparse_lu<SparseMatrix> fact is computed and, instead of
failing, InvL and InvU are tried to be computed. Notice that if you replace C
with


C = complex (sparse (ones (2)));


fact is computed, but then you get the nicer message


error: inv: sparse_lu: symbolic factorization failed


from sparse_lu. I don't know if it is related with bug #53390 (it is currently
not safe to use UMFPACK to factorize into LU a sparse matrix without column
reordering).

@David: nice to hear you again.




    _______________________________________________________

Reply to this item at:

  <https://savannah.gnu.org/bugs/?56232>

_______________________________________________
  Message sent via Savannah
  https://savannah.gnu.org/