Rename TSDataset.index -> TSDataset.timestamps, TSDataset.add_columns_from_pandas -> TSDataset.add_features_from_pandas, TSDataset.update_columns_from_pandas -> TSDataset.update_features_from_pandas

[ArtemLiA]

Before submitting (must do checklist)

• Did you read the contribution guide?
• Did you update the docs? We use Numpy format for all the methods and classes.
• Did you write any new necessary tests?
• Did you update the CHANGELOG?

Proposed Changes

Rename TSDataset.index to TSDataset.timestamps and make it read-only:

• Rename TSDataset.index to TSDataset.timestamps
• Make it return self.df.index.copy() (make read-only)
• Fix all places where ts.index was used and replace them with ts.timestamps

Rename TSDataset.add_columns_from_pandas to TSDataset.add_features_from_pandas:

• Rename TSDataset.add_columns_from_pandas to TSDataset.add_features_from_pandas
• Replace all usage of old name
• Fix tests

Rename TSDataset.update_columns_from_pandas to TSDataset.update_features_from_pandas:

• Rename TSDataset.update_columns_from_pandas to TSDataset.update_features_from_pandas
• Replace all usage of old name
• Fix tests and fixtures

Closing issues

Closes #564 #567

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[ArtemLiA]

Hi! I have reproduced failed test test_ensembles/test_stacking_ensemble::test_make_features. The traceback looks like this:

  File "/home/artem_liakhov/Projects/etna/tests/test_ensembles/conftest.py", line 214, in targets
    [
  File "/home/artem_liakhov/Projects/etna/tests/test_ensembles/conftest.py", line 215, in <listcomp>
    example_tsds[forecasts_ts[0].timestamps.min() : forecasts_ts[0].timestamps.max(), segment, "target"]
  File "/home/artem_liakhov/Projects/etna/.venv/lib/python3.10/site-packages/pandas/core/generic.py", line 5902, in __getattr__
    return object.__getattribute__(self, name)
AttributeError: 'DataFrame' object has no attribute 'timestamps'. Did you mean: 'to_timestamp'?

but according to the type annotation forecast_ts has type List[TSDataset] and the fixture of the same name returns the List[TSDataset]. Should this be a bug? Do you have any idea which test might be having problems and how to fix it?

[d-a-bunin]

Hi! It is probaby the same problem as in #563. Look at the comments. The probable solution can be seen here.

[ArtemLiA]

Thanks! I found the reason. Error was in typing annotation of forecasts_ts fixtures. It returns and should return List[pd.DataFrame] instead of List[TSDataset].

I fixed annotation and renamed forecasts_ts to forecasts_df so that name doesn't mislead anyone in the future.

[codecov]

Codecov Report

Attention: Patch coverage is 95.34884% with 2 lines in your changes missing coverage. Please review.

Project coverage is 90.59%. Comparing base (307f635) to head (354ad22).
Report is 1 commits behind head on master.

Files with missing lines	Patch %	Lines
etna/analysis/eda/plots.py	0.00%	2 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##           master     #593   +/-   ##
=======================================
  Coverage   90.59%   90.59%           
=======================================
  Files         261      261           
  Lines       18061    18061           
=======================================
  Hits        16363    16363           
  Misses       1698     1698           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[d-a-bunin]

Looks good!