Roopali24
commented
3 years ago Turns out there was a formatting error in my SC count data, the genes name was subsetted. Sorry to have posted this issue. Unable to delete it so closing the issue. Thank you.
Hi,
So I am trying to run stereoscope on a random dataset created using the Hippocampus data from your example, just to check if I am doing things correctly (hence the low epochs and small data). My ST data just has 100 genes (columns) and 100 spots(rows) and I created the SC data files i.e. the counts and the labels. The counts have 100 genes and 500 cells and labels have 500 cells and 10 cell types. However when I run stereoscope:
stereoscope run --sc_cnt test_sc_count.tsv --sc_labels test_sc_label.tsv -scb 256 -sce 50 --st_cnt test_X1.tsv -ste 50 -o /Users/roopalisingh/stereoscope/Synthetic/res1 --gpu -lr 0.1
It ends with the following error: [ERROR] : single cell count and annotation data did not match. Exiting. SC data GENES : 100 SC data CELLS : 0 SC data TYPES : 0 ValueError: batch_size should be a positive integer value, but got batch_size=0
But when I run using your SC files and my ST data:
stereoscope run --sc_cnt validcnt --sc_labels validmta -scb 256 -sce 50 --st_cnt test_X1.tsv -ste 50 -o /Users/roopalisingh/stereoscope/Synthetic/res1 --gpu -lr 0.1
This runs fine. I am curious what am I doing wrong in my SC data format. It looks similar to your files. Could you please help? Thank you in advance.