Skip to content

doc build failures and new warnings on main #3800

@Remi-Gau

Description

@Remi-Gau

Some new warnings seem related to the new sklearn release.

Note: doc warnings are flagged as error during doc build

See
https://github.com/nilearn/nilearn/actions/runs/5441381451/jobs/9895271361#step:33:35688

Attached log
33_Build docs.txt

The culprits seems to be:

2023-07-03T08:27:09.5655542Z /usr/share/miniconda3/envs/testenv/lib/python3.9/site-packages/nilearn/connectome/connectivity_matrices.py:docstring of sklearn.utils._metadata_requests._MetadataRequester.get_metadata_routing:11: WARNING: py:class reference target not found: utils.metadata_routing.MetadataRequest
2023-07-03T08:27:09.5704758Z /usr/share/miniconda3/envs/testenv/lib/python3.9/site-packages/nilearn/connectome/connectivity_matrices.py:docstring of sklearn.utils._metadata_requests.RequestMethod.__get__.<locals>.func:26: WARNING: py:class reference target not found: pipeline.Pipeline

Seems related to this: https://scikit-learn.org/stable/whats_new/v1.3.html#experimental-under-development

EDIT: suggestion to rely on intersphinx does not seem to apply as this is already the case for nilearn.


Listing some of the other most frequent warnings below but are not causing a failure:

1

2023-07-03T06:42:25.5886498Z /usr/share/miniconda3/envs/testenv/lib/python3.9/site-packages/sklearn/svm/_classes.py:32: FutureWarning: The default value of `dual` will change from `True` to `'auto'` in 1.5. Set the value of `dual` explicitly to suppress the warning.

Possibly related to

https://scikit-learn.org/stable/whats_new/v1.3.html#sklearn-svm

2

2023-07-03T07:27:55.7437431Z /usr/share/miniconda3/envs/testenv/lib/python3.9/site-packages/sklearn/linear_model/_base.py:121: FutureWarning:
2023-07-03T07:27:55.7442173Z 
2023-07-03T07:27:55.7446221Z 'normalize' was deprecated in version 1.2 and will be removed in 1.4. Please leave the normalize parameter to its default value to silence this warning. The default behavior of this estimator is to not do any normalization. If normalization is needed please use sklearn.preprocessing.StandardScaler instead.

3

EDIT: See #3719

2023-07-03T07:44:10.4039818Z /usr/share/miniconda3/envs/testenv/lib/python3.9/site-packages/nilearn/connectome/connectivity_matrices.py:495: FutureWarning:
2023-07-03T07:44:10.4040149Z 
2023-07-03T07:44:10.4041028Z The default strategy for standardize is currently 'zscore' which incorrectly uses population std to calculate sample zscores. The new strategy 'zscore_sample' corrects this behavior by using the sample std. In release 0.13, the default strategy will be replaced by the new strategy and the 'zscore' option will be removed. Please use 'zscore_sample' instead.

Metadata

Metadata

Assignees

No one assigned

    Labels

    Documentationfor documentation related questions or requestsMaintenanceThis issue is related to maintenance work.

    Type

    No type

    Projects

    Status

    Done

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions