UserWarnings in the documentation

[ArturoAmorQ]

Describe the issue linked to the documentation

Some UserWarning are present in the dev documentation and need to be fixed.
Here is a list:

• gaussian_process/plot_gpr_prior_posterior.html
• model_selection/plot_cv_indices.html DOC Removing warnings from plot CV examples #29072
• linear_model/plot_sgd_iris.html DOC Removing warnings from plot_sgd_iris example #29121
• linear_model/plot_logistic_multinomial.html DOC Removing warnings from plot_logistic_multinomial examples #29120
• tree/plot_iris_dtc.html DOC Removing warnings from plot_iris_dtc #29109
• svm/plot_svm_margin.html
• ensemble/plot_adaboost_twoclass.html

Contributors willing to address this issue, please fix one example per pull request. It is ok to fix other warnings or errors in a given example.

Thanks for your help!

Suggest a potential alternative/fix

No response

[ojoaomorais]

I'm working on plot_cv_indices.html !

[ojoaomorais]

Hello @ArturoAmorQ, my local build is showing different UserWarnings or, in some cases, showing nothing. Maybe it's the matplotlib package version? I changed from 3.9 to 3.3.4 (as seen here), but nothing changes. This is the command i use to build:

pip install --editable . \ --verbose --no-build-isolation \ --config-settings editable-verbose=true

My version:

Python dependencies:
sklearn: 1.6.dev0
pip: 24.0
setuptools: 69.5.1
numpy: 1.26.4
scipy: 1.13.0
Cython: 3.0.10
pandas: 2.2.2
matplotlib: 3.3.4
joblib: 1.4.2
threadpoolctl: 3.5.0
Built with OpenMP: True
threadpoolctl info:
user_api: blas
internal_api: openblas
num_threads: 2
prefix: libopenblas
filepath: /opt/anaconda3/envs/scikit-learn/lib/libopenblasp-r0.3.27.dylib
version: 0.3.27
threading_layer: openmp
architecture: Sandybridge
user_api: openmp
internal_api: openmp
num_threads: 4
prefix: libomp
filepath: /opt/anaconda3/envs/scikit-learn/lib/libomp.dylib
version: None

[Charlie-XIAO]

The matplolib version used in CI is 3.8.4 (not minimum, see this search link), though I don't think it's related.

[lucyleeow]

Note that it is possible to ignore warnings for all examples by adding to the doc/conf.py file. If we have a recurring warning, this may be nice and keeps our examples cleaner. Indeed we already ignore a warning:

scikit-learn/doc/conf.py

Line 783 in 1ccabef

warnings.filterwarnings(

[ojoaomorais]

I'm working on tree/plot_iris_dtc.html !!

[ojoaomorais]

Hello @ArturoAmorQ, my local build is showing different UserWarnings or, in some cases, showing nothing. Maybe it's the matplotlib package version? I changed from 3.9 to 3.3.4 (as seen here), but nothing changes. This is the command i use to build:

pip install --editable . \ --verbose --no-build-isolation \ --config-settings editable-verbose=true

My version:

Python dependencies: sklearn: 1.6.dev0 pip: 24.0 setuptools: 69.5.1 numpy: 1.26.4 scipy: 1.13.0 Cython: 3.0.10 pandas: 2.2.2 matplotlib: 3.3.4 joblib: 1.4.2 threadpoolctl: 3.5.0 Built with OpenMP: True threadpoolctl info: user_api: blas internal_api: openblas num_threads: 2 prefix: libopenblas filepath: /opt/anaconda3/envs/scikit-learn/lib/libopenblasp-r0.3.27.dylib version: 0.3.27 threading_layer: openmp architecture: Sandybridge user_api: openmp internal_api: openmp num_threads: 4 prefix: libomp filepath: /opt/anaconda3/envs/scikit-learn/lib/libomp.dylib version: None

FYI, it was missing the jupyterlite_sphinx lib. After install it i can see the warnings.

[ojoaomorais]

Working on plot_sgd_iris and plot_logistic_multinomial !