RandomizedSearchCV: warnings cannot be suppressed

Description

I would like to suppress the warnings thrown by sklearn. This can be done while I was working on a single classifier, however, it failed on SearchCV task. I am wondering if this is designed for purpose and if there are any workarounds.

Steps/Code to Reproduce

File 1 (No warnings)

import warnings
import sklearn
from sklearn.linear_model import LogisticRegression
from sklearn.model_selection import RandomizedSearchCV
import pandas as pd
import numpy as np


x = np.random.normal(0, 1, (100, 100))
y = [0] * 50 + [1] * 50

if __name__ == "__main__":
    warnings.filterwarnings("ignore")
    clf = LogisticRegression()
    clf.fit(x, y)
    clf.predict(x)
    print("done")

Run the code from terminal

File 2 (Warnings still exist)

import warnings
import sklearn
from sklearn.linear_model import LogisticRegression
from sklearn.model_selection import RandomizedSearchCV
import pandas as pd
import numpy as np


x = np.random.normal(0, 1, (100, 100))
y = [0] * 50 + [1] * 50

if __name__ == "__main__":
    warnings.filterwarnings("ignore")
    clf = LogisticRegression()
    clf = RandomizedSearchCV(clf, {"C": [1,2]}, n_jobs=-1)
    clf.fit(x, y)
    clf.predict(x)
    print("done")

Expected Results

should see done from terminal. While the second file gave me multiple FutureWarning

Actual Results

As above

Versions

appnope==0.1.0
backcall==0.1.0
certifi==2018.11.29
Click==7.0
cycler==0.10.0
decorator==4.3.0
feather-format==0.4.0
ipykernel==5.1.0
ipython==7.2.0
ipython-genutils==0.2.0
jedi==0.13.2
jupyter-client==5.2.4
jupyter-core==4.4.0
kiwisolver==1.0.1
matplotlib==3.0.2
mkl-fft==1.0.6
mkl-random==1.0.2
numpy==1.15.4
pandas==0.23.4
parso==0.3.1
pexpect==4.6.0
pickleshare==0.7.5
prompt-toolkit==2.0.7
ptyprocess==0.6.0
pyarrow==0.11.1
Pygments==2.3.1
pyparsing==2.3.0
python-dateutil==2.7.5
pytz==2018.7
pyzmq==17.1.2
scikit-learn==0.20.2
scipy==1.1.0
six==1.12.0
tornado==5.1.1
traitlets==4.3.2
wcwidth==0.1.7
xlrd==1.2.0

Tested on 3.6.8 with MacOS 10.14

[dmnapolitano]

Per @tomMoral 's comment, if you don’t want to completely disable warnings for everything, this “resolves” the issue with Python 3.10, sci-kit learn 1.0.2 on Ubuntu 20.04:

from sklearn.exceptions import ConvergenceWarning
from sklearn.utils import parallel_backend
...

gs = GridSearchCV(...)
with parallel_backend("multiprocessing"):
  with warnings.catch_warnings():
    warnings.filterwarnings("ignore", category=ConvergenceWarning, module="sklearn")
    gs.fit(...)

Hope this helps someone 😄

[JoElfner]

Any news on this? This reduces the usability of RandomizedSearchCV, since it hides useful information… Setting PYTHONWARNINGS works, but seems like a quite clumsy workaround…