Question about constraints

18 views
Skip to first unread message

yuesu jin

unread,
Aug 21, 2019, 11:58:01 PM8/21/19
to deal.II User Group
Hi all,
  I am learning step-7 which has the same class AffineConstraints class, in step6, we use the member function ::distribute_local_to_global to put the local matrix and rhs into global matrix and rhs, but in step-7, we still use system_matrix.add and then use the constraint.condense on system_matrix and system_rhs. Do those two procedure have the same effect?

-----------
Yuesu Jin,
Ph.D student,
University of Houston,
College of Natural Sciences and Mathematics,
Department of Earth and Atmospheric Sciences,
Houston, Texas 77204-5008


Daniel Arndt

unread,
Aug 22, 2019, 12:40:31 AM8/22/19
to deal.II User Group
Yuesu,


  I am learning step-7 which has the same class AffineConstraints class, in step6, we use the member function ::distribute_local_to_global to put the local matrix and rhs into global matrix and rhs, but in step-7, we still use system_matrix.add and then use the constraint.condense on system_matrix and system_rhs. Do those two procedure have the same effect?

You can find a discussion at https://www.dealii.org/9.0.1/doxygen/deal.II/group__constraints.html. In short: Yes, they have the same purpose, but different advantages and disadvantages. In general, we recommend using distribute_local_to_global over condensation.

Best,
Daniel

yuesu jin

unread,
Aug 22, 2019, 3:08:24 PM8/22/19
to deal.II User Group
Thank you very much, Daniel!! I appreciate your help!

--
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+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dealii/a9e938ad-06b6-49e1-9e99-91bafd4e8fc4%40googlegroups.com.


--
Reply all
Reply to author
Forward
0 new messages