Count matrix: getting the celltypes

[acycliq]

Hi!

Thanks for developing this format. I am trying to get the expession matrix (rows the genes, columns the cell type) but havent figured out how to do this.

import loompy
import numpy as np
import pandas as pd

filename = "l1_hippocampus.agg.loom"
ds = loompy.connect(filename)

I can get the Genes from the row attributes: ds.ra.Gene however I am puzzled about the cell types. Is this meant to be under ds.ca.AutoAnnotation? I get a list of strings (length 74) like: ['@OL,COP', '@OL,OLIG', '@IEG,@OL,@VGLUT1,@XIN,DG-IGC,HC-CA1,HC-CA2,OLIG',....], some are duplicates. How can I get the cell tpes please?

Thanks