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

[Marco Caliari]

Update of bug #56232 (project octave):

                  Status:                   Fixed => Patch Submitted        

    _______________________________________________________

Follow-up Comment #60:

@Rik, comment #58: I agree to keep the behavior of inv(zeros(2)). Then,
diagonal matrices are the most prominent example of sparse matrices (although
they are displayed in octave as full), so I would treat the singular case as
for sparse matrices. I fixed the test, hence.

@Kai, comment #59: so mattype.ishermitian() really means "is the matrix
hermitian AND nonsingular", good to know. Then, I removed the check of rcond
in that branch.

I attach my fourth patch, thanks for your feedback.

(file #46913)
    _______________________________________________________

Additional Item Attachment:

File name: sparse-singular4.cset          Size:9 KB
    <https://savannah.gnu.org/file/sparse-singular4.cset?file_id=46913>



    _______________________________________________________

Reply to this item at:

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

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