I am working with a collaborator using the same data, and we are sharing code via jupyter notebooks.
I noticed that although we start with the same data and process it using exactly the same code, our UMAP plots come out slightly different.
So I checked the package versions using sc.logging.print_versions() and found that although our scanpy and anndata versions are the same, our umap versions are not:
my versions: scanpy==1.4.6 anndata==0.7.1 umap==0.3.10
collaborator’s versions: scanpy==1.4.6 anndata==0.7.1 umap==0.4.2
I think that the different versions of umap could be causing the slight discrepancy in our results and would like to update umap on my end. I tried conda update --all but nothing changed. Is there a way to update umap within scanpy?
Thank you in advance for any help!