orca_predict 'NoneType' object is not iterable

Hi Dr. Zhou,

Thanks for your efforts on this user-friendly tool.

However, when I tried to run the examples by using python, I met some problems.

I've configured orca environment as instructed, and then tried to do predition:

import orca_predict 
orca_predict.load_resources(models=['32M', '256M'])
from orca_predict import *
outputs = process_region('chr9', 110404000, 111404000, hg38, window_radius=16000000, file='./chr17_110404000_111404000')

And then I got a warning and a error:

/root/miniconda3/envs/orca_env/lib/python3.9/site-packages/cooltools/lib/numutils.py:1376: RuntimeWarning: invalid value encountered in divide
  val_cur = ar_cur / armask_cur

TypeError                                 Traceback (most recent call last)
Cell In[22], line 1
----> 1 outputs = process_region('chr9', 110404000, 111404000, hg38, window_radius=16000000, file='./chr17_110404000_111404000')

File /mnt/e/WSL/orca/orca_predict.py:1157, in process_region(mchr, mstart, mend, genome, file, custom_models, target, show_genes, show_tracks, window_radius, padding_chr, model_labels, use_cuda)
   1153         genomeplot_256Mb(
   1154             outputs_ref, show_coordinates=True, model_labels=model_labels, file=file + ".256m.pdf",
   1155         )
   1156     else:
-> 1157         genomeplot(
   1158             outputs_ref,
   1159             show_genes=show_genes,
   1160             show_tracks=show_tracks,
   1161             show_coordinates=True,
   1162             model_labels=model_labels,
   1163             file=file + ".pdf",
   1164         )
   1165 return outputs_ref

File /mnt/e/WSL/orca/orca_utils.py:160, in genomeplot(output, show_genes, show_tracks, show_coordinates, unscaled, file, cmap, unscaled_cmap, colorbar, maskpred, vmin, vmax, model_labels)
    158 if output["experiments"] is not None:
    159     current_axis = 0
--> 160     for label in model_labels:
    161         for suffix in [" Pred", " Obs"]:
    162             all_axes[current_axis, 0].set_ylabel(
    163                 label + suffix,
    164                 labelpad=20,
   (...)
    169                 va="center",
    170             )

TypeError: 'NoneType' object is not iterable

Firstly I thought it might caused by specific cooltools version. Results above are based on cooltools 0.5.4. Then I updated cooltools to newest version (0.6.1), the same error occured with different warnings:

/root/miniconda3/envs/orca_env/lib/python3.9/site-packages/cooltools/lib/numutils.py:1376: RuntimeWarning: invalid value encountered in divide
  val_cur = ar_cur / armask_cur
/root/miniconda3/envs/orca_env/lib/python3.9/site-packages/cooltools/lib/numutils.py:1376: RuntimeWarning: invalid value encountered in divide
  val_cur = ar_cur / armask_cur
/root/miniconda3/envs/orca_env/lib/python3.9/site-packages/torch/utils/checkpoint.py:429: UserWarning: torch.utils.checkpoint: please pass in use_reentrant=True or use_reentrant=False explicitly. The default value of use_reentrant will be updated to be False in the future. To maintain current behavior, pass use_reentrant=True. It is recommended that you use use_reentrant=False. Refer to docs for more details on the differences between the two variants.
  warnings.warn(
/mnt/e/WSL/orca/orca_predict.py:406: RuntimeWarning: Mean of empty slice
  np.nanmean(
/mnt/e/WSL/orca/orca_predict.py:405: RuntimeWarning: Mean of empty slice
  target_r = np.nanmean(

Then I noticed the error is correlated with model_labels, which seems cannot be set as None? So I tried to set model_lables = ["test"] and this time it worked.

Anyway, the differences in packages versions still could lead to various problems. So could you please specifiy the package versions in config files?

Thank you!

jzhoulab / orca

orca_predict 'NoneType' object is not iterable #8