- Sponsor
-
Notifications
You must be signed in to change notification settings - Fork 5.5k
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
cholfact for sparse matrix does not keep floating type #14810
Comments
See #14076 (comment). It's a limitation in the sparse factorization library CHOLMOD. Eventually, we'd like to have pure Julia implementations of the sparse factorizations, but for now there is not much we can do. |
Would it be worthwhile leaving this open (and renaming it) as an issue to create a pure Julia sparse factorization library? |
At the very least, we should document the behavior. As an issue, "pure Julia sparse linalg" is rather too broad. |
I've changed the title to reflect the documentation issue. |
Closes #14810 Also changes description of `LDL^T` to `LDL'`. The former is wrong for complex matrices.
- Document that they only work with double precision (Closes #14810) - Also changes description of `LDL^T` to `LDL'`. The former is wrong for complex matrices. - Remove incorrect description from `cholfact` as doing `LDL^T` from docstring
In Julia 0.4.3, and Julia 0.5.0-dev, calling cholfact on a sparse Float32 matrix promotes the Cholesky factorization to Float 64. On a dense matrix, it correctly gives a Float32 factorization.
Compare:
With:
The text was updated successfully, but these errors were encountered: