On 2/12/25 09:21, Yann Jobic wrote:

I want to reuse the factorization of a PETSC/MUMPS solve for different
right-hand side vectors. I found the following discussion:

https://nam10.safelinks.protection.outlook.com/? url=https%3A%2F%2Fgroups.google.com%2Fg%2Fdealii%2Fc%2FSw0wHKEFuiA&data=05%7C02%7CWolfgang.Bangerth%40colostate.edu%7Ccc9f132e5dd34347116708dd4b815658%7Cafb58802ff7a4bb1ab21367ff2ecfc8b%7C0%7C0%7C638749741055859417%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=tYsPCfUrZbq7pQEyTSZ6ld7S8ykwLbGWsiZ%2Fre43UzE%3D&reserved=0

and the issue: https://nam10.safelinks.protection.outlook.com/? url=https%3A%2F%2Fgithub.com%2Fdealii%2Fdealii%2Fissues%2F2869&data=05%7C02%7CWolfgang.Bangerth%40colostate.edu%7Ccc9f132e5dd34347116708dd4b815658%7Cafb58802ff7a4bb1ab21367ff2ecfc8b%7C0%7C0%7C638749741055880166%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=LArEuAs1qSPsEZSrU2jGkLyfH%2BgyQVQOBAQ8qy8Osdo%3D&reserved=0

However, it's quite old. Is there a way to do it properly ? For example,
an option as keep the factorized matrices outside of the solve method ?

Yann -- I don't think there is any specific news I can share (though good job finding old discussions on the topic!). The interface of SparseDirectMUMPS does not currently provide a way to separate the factorization step from the solve step.

I suspect that it would not be very difficult to actually change the interface in such a way that it looks more like the one of SparseDirectUMFPACK that does exactly that. It's just that nobody so far has found the time or motivation to do that. But if you wanted to, that would of course be fantastic and we'd be excited to see a patch (and help you with it)!

Best
 Wolfgang

--
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see 
https://groups.google.com/d/forum/dealii?hl=en
--- You received this message because you are subscribed to the Google Groups "deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to dealii+unsubscr...@googlegroups.com.
To view this discussion visit 
https://groups.google.com/d/msgid/dealii/f3a8b05d-309b-440b-83f6-0f87a6cd22d6%40colostate.edu.

Reply via email to