How does the `tl.get_c_niche` function work?

[atharva-bhagwat]

Can you please provide some insights on how tl.get_c_niche works? Thank you!

[wheat-whq]

tl.get_c_niche calculates the proportion of cell types in the neighborhood of each cell through the spatial network, and then clusters them. There is a bug in the version installed by pip. It has been fixed in github and will be updated in pip later.

[atharva-bhagwat]

One bug i found is that there is a mismatch between C_niche values in adata.uns['SOAPy'] and adata.obs. Is it fixed now?