Closed ghtina closed 3 years ago
srajama1
commented
6 years ago No, there is no such limit (except 2B for int, a sequential direct solver may not be needed at this scale anyway). Klu is a simplicial (or non-supernodal) solver. In terms of performance, there are cases where it is very good and there are cases where Pardiso (which is a supernodal solver) is very good. However, you should not get zeroes back. What is this matrix ?
ghtina
commented
6 years ago Since the matrix is really large, I unfortunately couldn't show you. I used the Epetra_CrsMatrix and the entries are correctly inserted as I printed out the matrix. However, after calling the Solver and printing out the lhs, the whole vector was filled with zeros.
github-actions[bot]
commented
3 years ago This issue has had no activity for 365 days and is marked for closure. It will be closed after an additional 30 days of inactivity.
If you would like to keep this issue open please add a comment and/or remove the MARKED_FOR_CLOSURE label.
If this issue should be kept open even with no activity beyond the time limits you can add the label DO_NOT_AUTOCLOSE.
If it is ok for this issue to be closed, feel free to go ahead and close it. Please do not add any comments or change any labels or otherwise touch this issue unless your intention is to reset the inactivity counter for an additional year.
github-actions[bot]
commented
3 years ago This issue was closed due to inactivity for 395 days.
I had a matrix with a dimension over 2.5 Million and it works for Amesos with the Pardiso Solver, the number of nonzeros was about 200 Million). However, as I changed the Solver to the Klu solver, it didn't work anymore. The resulted vector was filled with zeros. So my question is why and has Amesos (or one of its Solver Library) a general constraint in either the dimension or number of nonzeros of the matrix?