noamteyssier / adpbulk

pseudobulking on an AnnData object
MIT License
22 stars 1 forks source link

love the package, question about the multiple categories! #4

Open mjstrumillo opened 1 year ago

mjstrumillo commented 1 year ago

HI, package works great for one condition, but for two I was getting a bug - I worked around it by manually connecting the obs columns into one adata.obs["one_two"] = adata.obs["one"].astype(str) +"_" +adata.obs["two"].astype(str)

but maybe its something that can be done in the code if you need it? since I think that the dtype was somehow messing the multiple category function. thanks! MJ

noamteyssier commented 1 year ago

hey @mjstrumillo - I'm sorry I'm getting back so late on this, but could you provide a minimal reproducible example of where this fails?

i.e. can you give me an example anndata where it breaks down?

thanks!