dimension_reduction jupyter kernel dies

[vjbaskar]

The jupyter kernel dies when I run dimension_reduction function with ~200,000 cells.
I am running stream==1.0.0 now. Is there any suggestion on how to run it or if I can skip this step and use the scanpy's dim reductions such as PCA?

Thanks.

[huidongchen]

The jupyter kernel dies when I run dimension_reduction function with ~200,000 cells. I am running stream==1.0.0 now. Is there any suggestion on how to run it or if I can skip this step and use the scanpy's dim reductions such as PCA?

Thanks.

Hi, 200k cells might be too large for the current implemented algorithm. In this case, you have the option to bypass this step by utilizing alternative dimension reduction methods like PCA and saving the result to the field adata.obsm['X_dr'].