Always specify explicitly that methods do not change an object in place but return a new one

[lars-reimann]

Users that are coming from other libraries like scikit-learn could easily forget that methods like fit do not work in place but return a new, fitted model/transformer.

We should always highlight this in the docstrings of these methods. This includes

• fit methods of TableTransformers
• fit methods of Classifiers
• fit methods of Regressors
• methods of Table that return a Table.

See https://stdlib.safe-ds.com/en/latest/development/guidelines/#document-api-elements for the required docstring format.

[zzril]

Some of the methods of Table that return a Table do actually modify the original table in-place. E.g. remove_columns(), which internally uses Pandas' drop method.
Is this intentional? If so, this issue needs further clarification on which methods' documentation needs to be updated.
If not, these methods need to be changed in order to actually return copies.

[lars-reimann]

As discussed, the pandas methods don't change data frames in place unless the inplace parameter is set to True. If data is really changed in place, that is not intended.

[lars-reimann]

🎉 This issue has been resolved in version 0.12.0 🎉

The release is available on:

• v0.12.0
• GitHub release

Your semantic-release bot 📦🚀