Error with sklearn.model_selection.cross_val_score

[carlossierra311]

Hi CodeApp team.

When I run this sample code:

# correct data preparation for model evaluation with k-fold cross-validation
from numpy import mean
from numpy import std
from sklearn.datasets import make_classification
from sklearn.model_selection import cross_val_score
from sklearn.model_selection import RepeatedStratifiedKFold
from sklearn.preprocessing import MinMaxScaler
from sklearn.linear_model import LogisticRegression
from sklearn.pipeline import Pipeline
# define dataset
X, y = make_classification(n_samples=1000, n_features=20, n_informative=15, n_redundant=5, random_state=7)
# define the pipeline
steps = list()
steps.append(('scaler', MinMaxScaler()))
steps.append(('model', LogisticRegression()))
pipeline = Pipeline(steps=steps)
# define the evaluation procedure
cv = RepeatedStratifiedKFold(n_splits=10, n_repeats=3, random_state=1)
# evaluate the model using cross-validation
scores = cross_val_score(pipeline, X, y, scoring='accuracy', cv=cv, n_jobs=-1)
# report performance
print('Accuracy: %.3f (%.3f)' % (mean(scores)*100, std(scores)*100))

I get this error message:

../compiled/joblib/externals/loky/backend/resource_tracker.py:120: UserWarning: resource_tracker: process died unexpectedly, relaunching.  Some folders/sempahores might leak.
Traceback (most recent call last):
  File "/private/var/mobile/Containers/Shared/AppGroup/AE37A578-E67A-4738-880C-DB1A18573B68/File Provider Storage/Machine Learning/Data Preparation for Machine Learning/code/chapter_04/05_cv_eval_correct.py", line 20, in <module>
    scores = cross_val_score(pipeline, X, y, scoring='accuracy', cv=cv, n_jobs=-1)
  File "../../../downloadable-site-packages/compiled/sklearn/utils/validation.py", line 60, in inner_f
  File "../../../downloadable-site-packages/compiled/sklearn/model_selection/_validation.py", line 440, in cross_val_score
  File "../../../downloadable-site-packages/compiled/sklearn/utils/validation.py", line 60, in inner_f
  File "../../../downloadable-site-packages/compiled/sklearn/model_selection/_validation.py", line 246, in cross_validate
  File "../compiled/joblib/parallel.py", line 966, in __call__
  File "../compiled/joblib/parallel.py", line 733, in _initialize_backend
  File "../compiled/joblib/_parallel_backends.py", line 494, in configure
  File "../compiled/joblib/executor.py", line 20, in get_memmapping_executor
  File "../compiled/joblib/executor.py", line 42, in get_memmapping_executor
  File "../compiled/joblib/_memmapping_reducer.py", line 531, in __init__
  File "../compiled/joblib/_memmapping_reducer.py", line 535, in set_current_context
  File "../compiled/joblib/_memmapping_reducer.py", line 560, in register_new_context
  File "../compiled/joblib/_memmapping_reducer.py", line 590, in register_folder_finalizer
  File "../compiled/joblib/externals/loky/backend/resource_tracker.py", line 191, in register
  File "../compiled/joblib/externals/loky/backend/resource_tracker.py", line 209, in _send
BrokenPipeError: [Errno 32] Broken pipe
>>> 

Any ideas on how to solve it?

Thanks

[bummoblizard]

Hi. Sklearn isn’t currently supported by the local runtime because it requires a module that is not yet ported to iOS. I’m unsure how you reach that far though. Can you share how you installed sklearn?

[carlossierra311]

I tried:
pip install sklearn

But that did fail. I also tried using the built-in PyPi installation, but this is what happens:

Finally, when running another python script, I think I saw Pyto trying to install some (¿older?) versions of sklearn automatically, but it got in some sort of loop, so I just closed the app.

[bummoblizard]

After a quick look it seems like Pyto has built in support for sklearn. I'm afraid it isn't something Code App supports currently.

You should probably run the code inside Pyto and perhaps ask Pyto's author regarding this.

[carlossierra311]

I'm really sorry for the mixup 😳😳😳.
I've been jumping between the two apps to see which one lets me run these samples, and obviously ended up sending the message to the wrong developer.
Please excuse me.