DOC: don't teach infer_datetime_format in user_guide #50334

ShashwatAgrawal20 · 2022-12-18T05:05:36Z

edited the file doc/user_guide/io.rst.

Just removed the infer_datetime_format from the docs.

It's my first pull request & it's not fully complete, so can you guide me on what else changes should be made.

phofl · 2022-12-18T10:32:08Z

doc/source/user_guide/io.rst

@@ -272,9 +272,6 @@ parse_dates : boolean or list of ints or names or list of lists or dict, default

  .. note::
     A fast-path exists for iso8601-formatted dates.
-infer_datetime_format : boolean, default ``False``


Hi, we can't simply remove it. We want to add a deprecation warning before removing it when we enforce the actual deprecation

Yea, I asked for it in the issue but no one replied

@phofl should this be Ok 71466b6

Hey - this parameter is now a no-op, so I'd say that it's correct to remove it from the user guide

It still shows in the docstring as deprecated, but I don't think there's any need to teach it in the user guide

Thoughts @phofl ?

I would prefer mentioning that it is a no op but to keep it in here with a deprecation warning anyway. This should increase visibility, which is helpful with deprecations ( I think). Also keeps us consistent with what we are doing for other parameters in read_csv

Sure but does a no-op need this much visibility? I think it looks confusing to beginners to read the user guide and find "you can use this parameter to speed things up. Though actually, never mind, it's deprecated, don't use it"

Ok for docstring, it's just for the user guide that I'm questioning the practice

I'll check tomorrow what's been done for other deprecations anyway, and what other libraries do

And sorry @ShashwatAgrawal20 for conflicting instructions here ☺

ShashwatAgrawal20 · 2022-12-19T15:42:32Z

@MarcoGorelli any updates?

MarcoGorelli

How about:

we mark it as deprecated in the list of parameters (as was done for squeeze in the example I linked)
we remove it from the section that teaches about what to try performance-wise, as it's no longer relevant (the deprecation is about how the parameter will be removed, the behaviour has already changed)

MarcoGorelli · 2022-12-19T18:17:12Z

doc/source/user_guide/io.rst

+  .. note::
+   The ``infer_datetime_format`` is deprecated and will be removed in the future versions. 
 infer_datetime_format : boolean, default ``False``


check here for how to mark as deprecated

pandas/doc/source/user_guide/io.rst

Lines 157 to 162 in eb226bd

squeeze : boolean, default ``False``

If the parsed data only contains one column then return a ``Series``.

.. deprecated:: 1.4.0

Append ``.squeeze("columns")`` to the call to ``{func_name}`` to squeeze

the data.

MarcoGorelli · 2022-12-19T18:17:26Z

doc/source/user_guide/io.rst

-1. Try to infer the format using ``infer_datetime_format=True`` (see section below).
+1. Try to infer the format using ``infer_datetime_format=True`` (this is deprecated and will be removed in the future versions.).


I'd remove it from here

MarcoGorelli · 2022-12-19T18:17:55Z

doc/source/user_guide/io.rst

-    24458940 Oct 10 06:44 test_table_compress.hdf
+    24458940 Oct 10 06:44 test_table_compress.hdf


not sure what's happened with the line ending - can you run the pre-commit checks on the files you've modified?

phofl · 2022-12-19T19:18:30Z

Yeah sorry if I wasn't clear enough. That's exactly what I wanted to do

MarcoGorelli

thanks for updating

could you set a descriptive title please?

MarcoGorelli · 2022-12-20T14:33:47Z

doc/source/user_guide/io.rst

+  .. deprecated:: 1.5.2
+   ``infer_datetime_format`` be deprecated (as a strict version of it will become the default);
+    an easy workaround for non-strict parsing be clearly documented.
+
+   Currently, the only way to ensure consistent parsing is by explicitly passing


could you copy the deprecation message from https://pandas.pydata.org/docs/dev/reference/api/pandas.to_datetime.html please

@MarcoGorelli
are you talking about

should I be directly using it as it is?

MarcoGorelli

thanks @ShashwatAgrawal20

ShashwatAgrawal20 · 2022-12-21T16:15:19Z

@MarcoGorelli thanks!
Had a great experience working with you and learned a lot.

ShashwatAgrawal20 marked this pull request as draft December 18, 2022 05:05

phofl reviewed Dec 18, 2022

View reviewed changes

ShashwatAgrawal20 force-pushed the doc-date-parser branch from e1214b4 to 96594f1 Compare December 18, 2022 14:22

ShashwatAgrawal20 requested a review from phofl December 18, 2022 14:46

MarcoGorelli suggested changes Dec 19, 2022

View reviewed changes

Shashwat added 2 commits December 20, 2022 17:20

docs: infer_datetime_format() deprecated

c1c1c63

Merge remote-tracking branch 'upstream/main' into doc-date-parser

d30ed08

ShashwatAgrawal20 force-pushed the doc-date-parser branch from bb23c43 to d30ed08 Compare December 20, 2022 11:52

ShashwatAgrawal20 requested review from MarcoGorelli and removed request for phofl December 20, 2022 14:09

MarcoGorelli suggested changes Dec 20, 2022

View reviewed changes

Shashwat added 2 commits December 20, 2022 22:17

Merge branch 'main' into doc-date-parser

056642e

docs: changes in deprecation msg

49c7b24

ShashwatAgrawal20 requested a review from MarcoGorelli December 20, 2022 16:52

MarcoGorelli changed the title ~~docs: user_guide io~~ DOC: don't teach infer_datetime_format in user_guide Dec 20, 2022

MarcoGorelli added the Docs label Dec 20, 2022

MarcoGorelli added this to the 2.0 milestone Dec 20, 2022

Merge branch 'main' into doc-date-parser

e50682c

MarcoGorelli marked this pull request as ready for review December 21, 2022 12:04

MarcoGorelli approved these changes Dec 21, 2022

View reviewed changes

MarcoGorelli merged commit 8469f74 into pandas-dev:main Dec 21, 2022

ShashwatAgrawal20 deleted the doc-date-parser branch May 11, 2024 10:17

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

DOC: don't teach infer_datetime_format in user_guide #50334

DOC: don't teach infer_datetime_format in user_guide #50334

ShashwatAgrawal20 commented Dec 18, 2022 •

edited by MarcoGorelli

Loading

phofl Dec 18, 2022

ShashwatAgrawal20 Dec 18, 2022

ShashwatAgrawal20 Dec 18, 2022

MarcoGorelli Dec 18, 2022

phofl Dec 18, 2022

MarcoGorelli Dec 18, 2022

ShashwatAgrawal20 commented Dec 19, 2022

MarcoGorelli left a comment •

edited

Loading

MarcoGorelli Dec 19, 2022

MarcoGorelli Dec 19, 2022 •

edited

Loading

MarcoGorelli Dec 19, 2022

phofl commented Dec 19, 2022

MarcoGorelli left a comment

MarcoGorelli Dec 20, 2022

ShashwatAgrawal20 Dec 20, 2022

MarcoGorelli Dec 20, 2022

ShashwatAgrawal20 Dec 20, 2022

MarcoGorelli Dec 20, 2022

MarcoGorelli left a comment

ShashwatAgrawal20 commented Dec 21, 2022

	squeeze : boolean, default ``False``
	If the parsed data only contains one column then return a ``Series``.

	.. deprecated:: 1.4.0
	Append ``.squeeze("columns")`` to the call to ``{func_name}`` to squeeze
	the data.

		1. Try to infer the format using ``infer_datetime_format=True`` (see section below).
		1. Try to infer the format using ``infer_datetime_format=True`` (this is deprecated and will be removed in the future versions.).

		24458940 Oct 10 06:44 test_table_compress.hdf
		24458940 Oct 10 06:44 test_table_compress.hdf

DOC: don't teach infer_datetime_format in user_guide #50334

DOC: don't teach infer_datetime_format in user_guide #50334

Conversation

ShashwatAgrawal20 commented Dec 18, 2022 • edited by MarcoGorelli Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ShashwatAgrawal20 commented Dec 19, 2022

MarcoGorelli left a comment • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

MarcoGorelli Dec 19, 2022 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

phofl commented Dec 19, 2022

MarcoGorelli left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

MarcoGorelli left a comment

Choose a reason for hiding this comment

ShashwatAgrawal20 commented Dec 21, 2022

ShashwatAgrawal20 commented Dec 18, 2022 •

edited by MarcoGorelli

Loading

MarcoGorelli left a comment •

edited

Loading

MarcoGorelli Dec 19, 2022 •

edited

Loading