Error loading brain9k dataset
Closed this issue · 1 comments
emdann commented
I get an error loading the brain 9k datasets, all other datasets work fine.
mudatasets.load('brain9k_multiome')
---------------------------------------------------------------------------
NameError Traceback (most recent call last)
Input In [4], in <cell line: 2>()
1 import os
----> 2 mudatasets.load('brain9k_multiome')
File ~/my-conda-envs/scverse-tutorial/lib/python3.10/site-packages/mudatasets/core.py:219, in load(dataset, data_dir, full, files, version, with_info, backed, chunk_size)
217 custom_loader = getattr(dset, "load", None)
218 if callable(custom_loader):
--> 219 mdata = dset.load(data_dir=data_dir)
220 else:
221 warn("There seems to be no file with accepted extension to load (h5mu, h5ad, h5). There is no custom loader either.")
File ~/my-conda-envs/scverse-tutorial/lib/python3.10/site-packages/mudatasets/datasets/brain9k_multiome.py:150, in Brain9kMultiome.load(self, data_dir)
148 counts = {"rna": "GSE162170_multiome_rna_counts.tsv.gz", "atac": "GSE162170_multiome_atac_counts.tsv.gz"}
149 for m, fname in counts.items():
--> 150 fpath = os.path.join(data_dir, fname)
151 modalities[m] = read_counts(fpath, sep="\t", obs_in_rows=False)
153 mdata = MuData(modalities)
NameError: name 'os' is not defined
Importing os doesn't help.
gtca commented
Thanks for noticing. This datasets hasn't actually been made available as part of the v0.0.1
(though it was displayed in the list of datasets), now it's available in v0.0.2
.