-
Chris currently uses steepest descent when doing energy minimisation, but with lots of local moves. Want to do group moves, but which groups? -> community detection!
-
`mfem::GMRESSolver` and `mfem::CGSolver` measure the residual norm after applying the preconditioner, while `mfem::NewtonSolver` (which has no notion of a preconditioner) measure the residual norm dir…
-
### Description of the enhancement or error report
I have a SPD system and want to take advantage by using the CG solver. However, the solve is noticeable slower than with GMRES and the convergence s…
hr87 updated
5 months ago
-
Report from Dave Eberly about problems with the oriented box `SolveCubic` helper function.
> The SolveCubic is called to compute the eigenvalues of a real-valued symmetric 3x3 matrix. Such matrices A…
-
When running the target `time_stepping` or `fluid_multigrid` one more than one processor, with `USE_PRECOND = TRUE`, the relative residual decreases very, very slowly. The solver doesn't converge to t…
-
Since PR #11 , a proof-of-concept for a SparseBLAS interface with an implementation generated by ALP is introduced. This by far, however, does not support all standard functions. As adding new functio…
-
Sometimes, during the minimization, the eigenvalues do not converge, causing an error that halts the code.
This seems to happen especially with the new line-minimization feature, and in systems wit…
-
@itsMeDavidV
I was thinking about starting this item list for requesting future enhancements to the RVS icons for your consideration. Just thinking about all the icon state possibilities, but rea…
-
-
Follows from #24. The keywords to start searching are "Strohner-type" and "Chan-type".
S.A. Goreinov mentioned at ELS 12 conference that they had very positive experience with such preconditioners …