fix: ensure statsforecast can run with PYTHONOPTIMIZE=2

[moagstar]

Fix for #976

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[jmoralez]

Thanks! Can you please export the notebooks (run nbdev_export)?

Alternatively you can just install the pre-commit hooks as described in our contributing guide and that'll export them.

[moagstar]

@jmoralez ah yes sorry I missed that step - I had some trouble with the pre-commit hooks, but I've managed to generate the files manually now.

[codspeed-hq]

CodSpeed Performance Report

Merging #977 will not alter performance

_{Comparing crunchr:fix/allow-pythonoptimize-2 (862b2ed) with main (52a5c9a)}

Summary

✅ 6 untouched benchmarks

[moagstar]

@jmoralez I see that some checks are failing.

nbdev_export.............................................................Failed
- hook id: nbdev_export
- files were modified by this hook
Error: Process completed with exit code 1.

When I ran nbdev locally it did generate some python files for notebooks I hadn't touched - but one of those generated files had some mypy errors:

python/statsforecast/distributed/multiprocess.py:35: error: Unexpected keyword argument "df" for "_StatsForecast"  [call-arg]
python/statsforecast/distributed/multiprocess.py:47: error: Unexpected keyword argument "df" for "_StatsForecast"  [call-arg]

So I just commit the python files for the two notebooks I had touched. I guess the fact that nbdev is generating more files is the reason for the CI failure?

Do you have some idea what might be happening here?

[jmoralez]

We use a specific version of nbdev

statsforecast/settings.ini

Line 24 in 52a5c9a

dev_requirements = black datasetsforecast fire nbdev==2.3.25 nbformat nbdev_plotly pandas[plot] pmdarima polars[numpy] pre-commit prophet pyarrow pybind11 pytest scikit-learn setuptools<70 supersmoother yfinance

so you have to set up the environment as described in the contributing guidelines, or I can fix the linting errors if you want.

[jmoralez]

Suggested change

"""Methods for Fit, Predict, Forecast (fast), Cross Validation and plotting"""

[jmoralez]

This should fix the linter error, I don't have permission to push to your branch.