Scikit Learn warning for Logistic Regression

[gbartoli-zcs]

What's wrong?
When using a Logistic Regression classifier with a simple dataset (built-in "Zoo" in this case) in a "Test and Score" widget, the following warnings about "multi_class" option are displayed:

❯ python -m Orange.canvas
/home/bartoli/miniconda3/envs/orange3/lib/python3.10/site-packages/orangewidget/gui.py:2073: UserWarning: decorate
OWLogisticRegression.apply with @gui.deferred and then explicitly call apply.now or apply.deferred.
  warnings.warn(
/home/bartoli/miniconda3/envs/orange3/lib/python3.10/site-packages/sklearn/linear_model/_logistic.py:1247: FutureWarning:
'multi_class' was deprecated in version 1.5 and will be removed in 1.7. From then on, it will always use 'multinomial'.
Leave it to its default value to avoid this warning.

This is the workflow I'm using:

How can we reproduce the problem?
warning.ows.zip

What's your environment?

• Operating system: Linux Mint 22.1
• Orange version: 3.38.1
• How you installed Orange: Conda package

[markotoplak]

Thanks for reporting, this was very likely fixed in the master branch (see #6956).

[gbartoli-zcs]

Thanks for reporting, this was very likely fixed in the master branch (see #6956).

Can the master branch be installed in Conda? Following the standard installation, the same 3.38.1 version containing this warning is installed.

Thanks

[markotoplak]

For the master branch you'd have to clone the repository and install it as developer do. See the developing section on https://github.com/biolab/orange3.

But I think the best would be to just ignore the warning for now.