I have a question about the parameter iparm (zero index, weighted matchings) of intel pardiso solver.
Actually the problem was in a solution of a non linear problem by Newton Raphson iteration. The used solver was Intel Pardiso for sparse non symmetric systems. The solution diverged, as in the topic
while the solver dgbsv (lapack) converged. The value of parameter was iparm = 1 (as in the examples.)
I don't remember if I tried iparm=0 or not (It's about a couple of months, I think I did). When I tried iparm = 2 it worked and Intel Pardiso and DGBSV solvers produced the same correct result. Actually I keep using iparm = 2, in difficult systems.
I cannot find the document "intel pardiso users guide" where this comment about iparm = 2 was written anymore. In the site and in all MKL documents the value of iparm is 0 or 1. Is it implemented the iparm = 2 ? or it has the same effect as iparm = 0?
In Pardiso's user guide https://pardiso-project.org/manual/manual.pdf they have a comment about iparm(13) = 2. (weighted matchings)
Thanks in advance,
That's correct, behavior of MKL pardiso is different in case of iparm is equal to 2 or 3 and sometimes results with these parameters better than 0 or 1. However the behavior currently is unstable so officially these parameters are unsupported and I recommend to not use them.