tqdm.auto does not work fine in Atom hydrogen

[kamione]

Summary

tqdm.auto progress bar does not update if the codes are running in a Atom IDE with hydrogen

Additional details

• I run the NiMARE package which it uses tqdm.auto
• NiMARE version: 0.0.7
• I have reported this issue in the the NiMARE repo but this issue is generally caused by tqdm.auto in Atom Hydrogen environment [

What were you trying to do?

I tried to run:

ale = nimare.meta.cbma.ALE()
res = ale.fit(dset)

corr = nimare.correct.FWECorrector(method="montecarlo", n_iters=10, n_cores=-1)
cres = corr.transform(ale.results)

What did you expect to happen?

What actually happened?

This block of codes only produced:

INFO:nimare.correct:Using correction method implemented in Estimator: nimare.meta.cbma.ale.ALE.correct_fwe_montecarlo.
  0%|          | 0/10 [00:00<?, ?it/s]
INFO:nimare.meta.cbma.base:Using null distribution for voxel-level FWE correction.

The codes worked perfectly but the progress bar does not update if I work in a Atom IDE with hydrogen

Reproducing the bug

# This problem can be simply reproduced in a Atom IDE with hydrogen by:
import time
from tqdm.auto import tqdm

for _ in tqdm(range(10)):
    time.sleep(1)
    pass

[casperdcl]

Yes this is essentially an old issue with the design of the IPython kernel itself (see the accepted albeit relatively unpopular answer at https://stackoverflow.com/a/15412661/3896283 essentially stating "this shouldn't be possible by design.") I suppose ipywidgets itself should be providing widgets with better text representation functionality (#1098 (comment)).

You could replace tqdm.auto.tqdm with simply tqdm.tqdm, or perhaps try to implement one of the alternative work-arounds/hacks on that stackoverflow page.