Skip to content

Conversation

glemaitre
Copy link
Member

@glemaitre glemaitre commented Nov 9, 2021

PyWavelets 1.2.0 requires numpy >= 1.17 that is not has a known bug with pandas <= 1.2.
Pinning the version of PyWavelets would be enough here.

@glemaitre glemaitre marked this pull request as draft November 9, 2021 13:57
@glemaitre
Copy link
Member Author

OK. So the thing that is fishy here is:

@jorisvandenbossche mentioned that it could be something linked with the latest release of setuptools. I will downgrade to investigate this.

@glemaitre
Copy link
Member Author

Actually, it might only be some pip printing misleading. PyWavelets have been updated from 1.1.1 to 1.2.0 which probably change their minimal NumPy dependency.

@glemaitre
Copy link
Member Author

I opened pypa/pip#10651 to report the misleading log given by pip

@glemaitre glemaitre marked this pull request as ready for review November 10, 2021 13:52
@glemaitre glemaitre changed the title MNT check doc_min_build failure MNT pin PyWavelet in doc-min builds Nov 10, 2021
@glemaitre
Copy link
Member Author

pinging @adrinjalali @thomasjpfan

Comment on lines +148 to +149
wget https://github.com/conda-forge/miniforge/releases/latest/download/Mambaforge-Linux-x86_64.sh \
-O miniconda.sh
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

O_o do we want to switch to mamba? why not miniforge?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It was for debugging reasons first. Then, I assume that mamba will be faster than conda to resolve the dependencies required here.

# NumPy version required. If PyWavelets 1.2+ is installed, it would require
# NumPy 1.17+ that trigger a bug with Pandas 0.25:
# https://github.com/numpy/numpy/issues/18355#issuecomment-774610226
pip install PyWavelets==1.1.1
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

we need a comment here, or in our min dependency file, which would remind us to remove this once we bump our min numpy requirement.

Copy link
Member

@adrinjalali adrinjalali left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM, and the circleCI passes.

Copy link
Member

@thomasjpfan thomasjpfan left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM to unblock other PRs

For the future, I would prefer to remove the skimage dependency from our examples and docs.

@thomasjpfan thomasjpfan merged commit bbb81e9 into scikit-learn:main Nov 12, 2021
glemaitre added a commit to glemaitre/scikit-learn that referenced this pull request Nov 22, 2021
glemaitre added a commit to glemaitre/scikit-learn that referenced this pull request Nov 29, 2021
samronsin pushed a commit to samronsin/scikit-learn that referenced this pull request Nov 30, 2021
glemaitre added a commit to glemaitre/scikit-learn that referenced this pull request Dec 24, 2021
glemaitre added a commit that referenced this pull request Dec 25, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants