search

KevinMenden / scaden

Deep Learning based cell composition analysis with Scaden.
https://scaden.readthedocs.io
MIT License
71 stars 26 forks source link

Error when using adata object #108

Open ThomasThaewel opened 3 years ago

ThomasThaewel commented 3 years ago

Hi, I was trying to run scaden with an adata object (dense matrix) and used a txt file with the samples as columns and genes as rows for the bulk data. However if I try to run scaden process .... I get the Error: IndexError: arrays used as indices must be of integer (or boolean) type. This does not really make sense to me as the Bulk data example also uses genes as "Index" and therefore is not using integers. Does someone have an idea how to solve this? Thank you

KevinMenden commented 3 years ago

Hi @PronatorT ,

sorry for the late reply. Quick question - how die you generate the adata object, with Scaden?

ThomasThaewel commented 3 years ago

No not with scaden itself, I used scanpy to generate the adata object. I was not aware of scaden being able to generate anndata objects.

KevinMenden commented 3 years ago

Okay, it's best to create the training data with Scaden as well. Have a look at the scaden simulate function. Scaden uses the anndata library internally - but it's just the safest way to get the formatting right :)