CODEWITHSUNDEEP
scikit-learnsuppress-warnings
Valeria
Valeria

Reputation: 1220

Silence UndefinedMetricWarning

I have really a lot of UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. resulting from running a randomized search pipeline with cross-validation. I know what is causing this behavior and suggested setting to 0.0 score is currently fine with me, so I want to just silence this warning for now.

I tried:

warnings.filterwarnings('ignore')

and

from sklearn.exceptions import UndefinedMetricWarning
warnings.filterwarnings('ignore', category=UndefinedMetricWarning)

But I am still getting these warnings, even though other answers on StackOverflow suggested that they should be suppressed by these lines (and actually it worked for me some time ago in a notebook).

The warnings.filterwarnings(...) line is located directly under import statements, and the warnings are from in one of the nested functions.

Upvotes: 2

Views: 2535

Answers (2)

crimper
crimper

Reputation: 31

For me, the solution in the question works:

from sklearn.exceptions import UndefinedMetricWarning
import warnings
warnings.filterwarnings("ignore", category=UndefinedMetricWarning)

also see: How to disable Python warnings?

I am using this:

Python 3.8.10 (default, Nov 22 2023, 10:22:35) 
>>> print('The scikit-learn version is {}.'.format(sklearn.__version__))
The scikit-learn version is 1.0.1.

Upvotes: 0

seralouk
seralouk

Reputation: 33147

Use the following:

from sklearn.exceptions import UndefinedMetricWarning

def warn(*args, **kwargs):
    pass
import warnings
warnings.warn = warn

# more code here...
# more code here...

Upvotes: 0

Related Questions