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
Cholesky fails for sparse matrices #5373
Comments
I'm not aware of a cholesky decomposition implementation that works with sparse arrays. |
Update: Apologies this works if I set the following:
|
I believe the package |
You might want to engage with the github. com/pydata/sparse package, which
is probably where this sort of change would need to be made.
…On Sun, Sep 8, 2019, 12:30 PM TomMaullin ***@***.***> wrote:
I believe the package cvxoptdoes a sparse cholesky decomposition.
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<#5373?email_source=notifications&email_token=AACKZTFWLSBSWCZLXWA3FWDQIVHGRA5CNFSM4IUTZCJ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD6FXSRY#issuecomment-529234247>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AACKZTGLOV2UQFP7FTGK6X3QIVHGRANCNFSM4IUTZCJQ>
.
|
So if pydata/sparse implements a sparse Cholesky decomposition, what will dask need to do to use it? Right now we call |
If that happens (and it seems like it might be slightly far away) then I wonder if SciPy might be interested in adopting NEP-18 support. cc @rgommers |
We are at least interested in exploring it: http://scipy.github.io/devdocs/roadmap.html#support-for-distributed-arrays-and-gpu-arrays |
Hi, I don't know if it helps but https://scikit-sparse.readthedocs.io/en/latest/cholmod.html is an implementation of a Sparse Cholesky decomposition (cython). We use it quite a lot. If it could be natively implemented in dask, that would be very useful. |
If you are interested in taking this on @MickaelRigault a pull request would be welcome :) |
Hi I am trying to run a cholesky decomposition on a Sparse matrix like so:
But I get the following error:
This was run on a CPU in a google colab notebook; any help on resolving this would be greatly appreciated... ideally I'm hoping to find a solution that at least allows some of the computation to be done using sparse operations as time efficiency is a big factor in what I am working on (the above is a toy example... my actual work uses much larger matrices).
The text was updated successfully, but these errors were encountered: