Dear all,
Thank you for your always support and help. I am encountering an issue while solving an eigenvalue problem related to the computation of the discrete inf-sup constant for a saddle point problem. Specifically, I am solving the following system: Bt A^-1 B eigenvector = eigenvalue M eigenvector, where Bt A^-1 B is clearly symmetric and at least semi-positive definite, and M is a symmetric positive definite matrix. The problem arises when comparing the inf-sup constants obtained using Matlab and dealii (solving on the same mesh ): 1. Matlab: Using the "eig" solver, the inf-sup constant decays to zero as the mesh is refined. 2. deal.II: Using the "ArpackSolver", the inf-sup constant is bounded away from zero. The discrepancy is puzzling, as both methods should give consistent results. To illustrate this issue, I have attached a figure comparing the two results and included in this post the two data files obtained from Matlab and deal.II. I would greatly appreciate any insights or suggestions on what might be causing the discrepancy. Your help is highly appreciated, Najwa -- 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 on the web visit https://groups.google.com/d/msgid/dealii/d28832e9-9f39-4300-b8c6-ef0504f2bc38n%40googlegroups.com.
0.25 0.346321 0.125 0.215878 0.0625 0.167721 0.03125 0.153921
0.2500 0.0601 0.1250 0.0162 0.0625 0.0041 0.0312 0.0010
inf_sup.pdf
Description: Adobe PDF document