-
Notifications
You must be signed in to change notification settings - Fork 44
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
mode solver preconditioner improvement and turning off incidence matrix #1928
Conversation
3bdbc2a
to
5f11c6d
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The improved preconditioning makes sense to me. Very nice fix!
c185438
to
117c4ca
Compare
117c4ca
to
eec46bd
Compare
@@ -22,6 +22,8 @@ | |||
TOL_TENSORIAL = 1e-6 | |||
# shift target neff by this value, both rel and abs, whichever results in larger shift | |||
TARGET_SHIFT = 10 * fp_eps | |||
# Preconditioner: "Jacobi" or "Material" based | |||
PRECONDITIONER = "Material" |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Now I also include the previous preconditioning method (Jacobi) along with the new one. For non-magnetic material (mu=1 everywhere), the two are actually very similar. The new one performs better when both epsilon and mu can be huge.
Ax= lambda M x
, only eps^-1 and mu^-1 appear inA
andM
.