[Octave-bug-tracker] [bug #40743] num2str fails to handle sparse matrix with mixed Inf/NaN

[Rik]

Update of bug #40743 (project octave):

                  Status:                    None => Confirmed              

    _______________________________________________________

Follow-up Comment #1:

I think the failure is in the max() operator for sparse matrices.  It should
not bother to check the dimensions of null matrices.

Some sample code:


## Works with Full matrices
max (zeros (1,0), ones (1,1))
ans = [](1x0)
## Fails with Sparse matrices
max (sparse(zeros (1,0)), sparse(ones (1,1)))
error: max: matrix size mismatch
## Works partially with mixed full/sparse matrices
max (sparse(zeros (1,0)), ones (1,1))
ans = Compressed Column Sparse (rows = 1, cols = 0, nnz = 0)
max (zeros (1,0), sparse(ones (1,1)))
error: max: nonconformant arguments (op1 is 1x0, op2 is 1x1)



    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/bugs/?40743>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.gnu.org/