Paris clustering does not converge

[KSimi7]

Hi!

Thanks for the amazing package! I am using scarf to subsample a dataset of 0.4 million cells. I was able to create a zarr datset from h5ad and perform all the required steps before clustering. When I ran clustering I left the code running for a day it was still not done. I ran it again today and after keeping the cell running for 10~12 hours it is still not done. I also noticed that the kernel status as unknown on this notebook. I tried reducing the number of clusters that did not help either. Is there any other way I could use the subsampling or speed up the clustering process?

Thanks for your help!

[parashardhapola]

@KSimi7 There might be an upstream issue with the sknetwork package.
Have you tried Leiden clustering?
ds.run_leiden_clustering