Bug in tokenizer.py: tokenize_loom

#565
by yqsu95 - opened

Small bug in tokenize_loom function

Hi Geneformer devleopers, thank you for the great model in Geneformer, I have been trying to use the newest version of the package and came across a bug when running the example tutorial for this line in tokenizer.py

line 503: file_tokenized_cells, file_cell_metadata, file_tokenized_counts = tokenize_file_fn(file_path, file_format=file_format)
ValueError: not enough values to unpack (expected 3, got 2)

It would seem that the return of tokenize_file_fn should be a tuple of length 3 variables, the tokenize_loom function still returns 2 variables, while tokenize_anndata was already changed to return 3 variables, this would cause an error if loom was chosen as the format for tokenization input.

Thank you for pointing this out. There was a fix pushed - please reach out if any further issues.

ctheodoris changed discussion status to closed

Sign up or log in to comment