DOC: to_pickle arguments for compression have a mistake (easy-fix)

[ianozsvald]

Location of the documentation

https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.to_pickle.html

Documentation problem

compression{‘infer’, ‘gzip’, ‘bz2’, ‘zip’, ‘xz’, None}, default ‘infer’ notes that gzip is a valid extension, this is wrong, the extension checked for is gz. The documentation for to_csv is correct with ... detect compression mode from the following extensions: ‘.gz’, ‘.bz2’, ‘.zip’ or ‘.xz’ and should act as a guide: https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.to_csv.html

Suggested fix for documentation

compression{‘infer’, ‘gz’, ‘bz2’, ‘zip’, ‘xz’, None}, default ‘infer’ to match the code implementation: https://github.com/pandas-dev/pandas/blob/v1.0.5/pandas/io/common.py#L223

There may be an argument to extending the _compression_to_extension code to accept both gzip and gzfor gzip compression, whoever reads this might decide this is worth escalating. For me making sure we fix the docs in line with the code is the priority.

[rhshadrach]

It's a bit confusing because gzip is the one compression type where the mode and the extension differ. The option compression mode is gzip. On the other hand, when mode=infer, the extension gz is looked for.

This is how to_csv works too:

Compression mode may be any of the following possible values: {‘infer’, ‘gzip’, ‘bz2’, ‘zip’, ‘xz’, None}. If compression mode is ‘infer’ and path_or_buf is path-like, then detect compression mode from the following extensions: ‘.gz’, ‘.bz2’, ‘.zip’ or ‘.xz’. (otherwise no compression). If dict given and mode is ‘zip’ or inferred as ‘zip’, other entries passed as additional compression options.

It seems to me having these details would be useful in the to_pickle docs as well.

[ianozsvald]

Thanks to @jreback for picking this up 👍