search

aertslab / pycisTopic

pycisTopic is a Python module to simultaneously identify cell states and cis-regulatory topics from single cell epigenomics data.
Other
58 stars 12 forks source link

Error while running run_cgs_models_mallet() #164

Closed skoturan closed 2 months ago

skoturan commented 2 months ago

Hi, I'm running topic modelling function in cistopic in the new SCENIC+ with this:

os.environ['MALLET_MEMORY'] = '200G' from pycisTopic.lda_models import run_cgs_models_mallet mallet_path="/filepath/Mallet-202108/bin/mallet"

models=run_cgs_models_mallet(cistopic_obj, n_topics=[2, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50], n_cpu=12, n_iter=500, random_state=555, alpha=50, alpha_by_topic=True, eta=0.1, eta_by_topic=False, tmp_path="tmpdir", save_path="none", mallet_path=mallet_path, )

I keep getting this error:

2024-09-09 11:33:00,119 cisTopic INFO Formatting input to corpus 2024-09-09 11:33:00,124 cisTopic INFO Running model with 2 topics 2024-09-09 11:33:00,125 LDAMalletWrapper INFO Serializing temporary corpus to filepath/corpus.txt Unexpected exception formatting exception. Falling back to standard exception . . . AttributeError: 'Source' object has no attribute 'asttext'

Could you please help to resolve this?

Thanks

SeppeDeWinter commented 2 months ago

Hi @skoturan

This seems to be an issue with mallet. Which version do you have?

All the best,

Seppe

skoturan commented 2 months ago

Hi, thanks for replying!

It was a permission issue for the mallet file- I executed chmod and it's running now!