If any of the existing docstrings don't have examples, take an example from one of these answers and add it. Likewise, if there's a really good example from these answers which isn't present in an existing docstring, please add it.

please comment letting others know which example you will be adding. Please only take 1-2 examples at a time.

Overload interpolate

missing link to multiindex in https://pandas.pydata.org/pandas-docs/dev/user_guide/10min.html#stack

lots of missing links in https://pandas.pydata.org/pandas-docs/dev/user_guide/dsintro.html too

remove pd_array from tests

Currently, lots of tests use pd_array, e.g. pandas/tests/arithmetic/common.py .
It would be better to import array from pandas and use that

Update black to new version, make sure comments are preserved

Fix regression with numpy ufuncs and DataFrames with out keyword

See pandas-dev/pandas#40662

continue work on https://github.com/pandas-dev/pandas/pull/39207

Overload Series.drop_duplicates

add future imports to files

xref pandas-dev/pandas#41901

Move grep checks to pre-commit

Check the pygrep hooks in https://github.com/pandas-dev/pandas/blob/master/.pre-commit-config.yaml

Reorganise pre-commit-config

Currently, it's structured quite randomly.

Let's instead have:

third-party hooks
local hooks

, both sorted alphabetically

Add tests to pandas-dev-flaker

Currently, pandas has a lot of code quality checks. I've recently started work on writing a flake8 plugin, which unifies the existing checks and which will make its way into pandas once ready (hopefully within 1-2 weeks).

The repository is here: https://github.com/MarcoGorelli/pandas-dev-flaker

The issues to work on are there, in https://github.com/MarcoGorelli/pandas-dev-flaker/issues

If you'd like to work on an issue there, please leave a comment saying that you're working on it

Extend coverage of inconsistent-namespace-usage

See pandas-dev/pandas#39992

There are still some folders which need doing

Migrate grep checks from ci/code_checks.sh to pre-commit

ci/code_checks.sh contains lots of grep-based checks. These should be moved to pre-commit, as it comes with a load of benefits.

The following should be moved to the id unwanted-patterns:

    MSG='Check for use of exec' ; echo $MSG
    invgrep -R --include="*.py*" -E "[^a-zA-Z0-9_]exec\(" pandas
    RET=$(($RET + $?)) ; echo $MSG "DONE"

    MSG='Check for use of builtin filter function' ; echo $MSG
    invgrep -R --include="*.py" -P '(?<!def)[\(\s]filter\(' pandas
    RET=$(($RET + $?)) ; echo $MSG "DONE"

    MSG='Check for deprecated messages without sphinx directive' ; echo $MSG
    invgrep -R --include="*.py" --include="*.pyx" -E "(DEPRECATED|DEPRECATE|Deprecated)(:|,|\.)" pandas
    RET=$(($RET + $?)) ; echo $MSG "DONE"

    MSG='Check for use of {foo!r} instead of {repr(foo)}' ; echo $MSG
    invgrep -R --include=*.{py,pyx} '!r}' pandas
    RET=$(($RET + $?)) ; echo $MSG "DONE"
    echo $MSG "DONE"

The following to id unwanted-patterns-in-tests:

    MSG='Check for pytest warns' ; echo $MSG
    invgrep -r -E --include '*.py' 'pytest\.warns' pandas/tests/
    RET=$(($RET + $?)) ; echo $MSG "DONE"

    MSG='Check for invalid testing' ; echo $MSG
    invgrep -r -E --include '*.py' --exclude testing.py '(numpy|np)(\.testing|\.array_equal)' pandas/tests/
    RET=$(($RET + $?)) ; echo $MSG "DONE"

    MSG='Check that unittest.mock is not used (pytest builtin monkeypatch fixture should be used instead)' ; echo $MSG
    invgrep -r -E --include '*.py' '(unittest(\.| import )mock|mock\.Mock\(\)|mock\.patch)' pandas/tests/
    RET=$(($RET + $?)) ; echo $MSG "DONE"

All that needs copying is the regex pattern, along with a comment explaining what it does. You should also run pre-commit run unwanted-patterns --all-files and make sure it still passes

Add RT05 and PR10 to validate_docstrings

should just be able to add them to

https://github.com/pandas-dev/pandas/blob/c3ab0c99264be1c7907525ea4846f3f024057a28/ci/code_checks.sh#L81-L82

Getting started

Welcome!

Welcome to the 6th edition of "Let's contribute to pandas"! The format will be as follows:

5-10 minutes: brief introduction, overview of issues you could choose to work on
rest of the session: your chance to work on issues and ask questions!

Here are some things you could choose to work on:

any issue from this repository
good first issues (as above)
any other issue from the pandas-dev/pandas issue tracker (not recommended for first-timers): can be challenging and time-consuming

Remember - it's OK to push work that isn't perfect or finished, or that you're not 100% sure of!

Please visit and use it throughout the session to track your progress: https://docs.google.com/spreadsheets/d/1k3KTsL6x57K_qUFp9IqUI2_YQgyXKZ2gGgON6vTnbOE/edit?usp=sharing

In general, please refer to contributing guide for pandas-dev/pandas: https://pandas.pydata.org/pandas-docs/dev/development/contributing.html

clarify Series.combine_first

Docs are unclear:

https://pandas.pydata.org/pandas-docs/dev/reference/api/pandas.DataFrame.combine_first.html?highlight=combine_first#pandas.DataFrame.combine_first

https://pandas.pydata.org/pandas-docs/dev/reference/api/pandas.Series.combine_first.html?highlight=combine_first#pandas.Series.combine_first

DataFrame is alright, but Series is unclear. Can the docs be shared?

add `regex` argument to rsplit

see pandas-dev/pandas#46488

add example to reorder levels

https://pandas.pydata.org/pandas-docs/dev/reference/api/pandas.DataFrame.reorder_levels.html?highlight=reorder_levels#pandas.DataFrame.reorder_levels

No examples! Can we add at least one?

clarify reindex

pandas-dev/pandas#33888

User complained it's unclear, but didn't do anything about it. Check closed PR pandas-dev/pandas#34584 for starting material

Overload Categorical functions with inplace

Overload Series.drop

add reverse rolling window example

i.e. revive pandas-dev/pandas#39091

Hasn't been worked on for ~2 months, so OK to take over

Move np.array_equal check to pre-commit

This check is currently here:

https://github.com/pandas-dev/pandas/blob/9bfa67cf028deec509d4e07138f6659bd2b342fe/ci/code_checks.sh#L83-L86

it should be moved to the bottom of

https://github.com/pandas-dev/pandas/blob/822db7a53fdcf8d860aaa8b51da4b767b3f56fad/.pre-commit-config.yaml#L136-L156

Task here is:

copy the regular expression (numpy|np)(\.testing|\.array_equal) into .pre-commit-config.yaml
remove this check from ci/code_checks.sh
git add, git commit, git push 🚀

Add test for fixed regression in concat with categorical datetime data

pandas-dev/pandas#39443 was a regression that in the meantime is fixed. But, we still want to add a test for it, to ensure it keeps working in the future.

A test for this can be added to pandas/tests/reshape/concat/test_categorical.py

Function overloads

The issue is to use typing.overload to make the return type of some functions more precise.

An introductory video on typing.overload: https://youtu.be/rY9NZ-tXiDQ

A (work-in-progress) blog post on this topic: https://m-e-gorelli.medium.com/making-sense-of-typing-overload-437e6deecade

In the following methods all return something like Optional[FrameOrSeries]. We should overload them, so that if inplace=True they return None and if inplace=False they return FrameOrSeries

https://github.com/pandas-dev/pandas/blob/master/pandas/core/series.py

drop_duplicates
drop

https://github.com/pandas-dev/pandas/blob/master/pandas/core/generic.py

interpolate
bfill
ffill
https://github.com/pandas-dev/pandas/blob/master/pandas/core/frame.py
drop_duplicates

From the pandas codebase, see set_axis and reset_index for examples

use better example in pd.lookup replacement

pandas-dev/pandas#39171 (comment)

In the user guide, let's use

idx, cols = pd.factorize(df['col'])
df.reindex(cols, axis=1).to_numpy()[np.arange(len(df)), idx]

instead of the current workaround

https://pandas.pydata.org/pandas-docs/dev/user_guide/indexing.html#looking-up-values-by-index-column-labels

specify fewer errors in validate docstrings

We currently have a check which runs scripts/validate_docstrings.py, bu only with the following error codes:

--errors=GL03,GL04,GL05,GL06,GL07,GL09,GL10,SS02,SS04,SS05,PR03,PR04,PR05,PR10,EX04,RT01,RT04,RT05,SA02,SA03

Some error codes currently excluded are:

SS03
PR08
SA04
GL01
EX01
SS03
SS06

If you just run python scripts/validate_docstrings.py, you will see these and plenty more.

So, to take on this issue, you should:

Pick an extra error code to take on
Run validate_docstrings.py with that error code, for example:

python scripts/validate_docstrings.py --errors=SS01

Pick 5-10 failing files, and fix them so that they pass this check!

Please comment with:

which error code you will work on
which files you will fix

Update str.cat documentation

issue: pandas-dev/pandas#35556

open PR, but was closed: pandas-dev/pandas#35784

Fixup mypy errors

pandas-dev/pandas#37715

Lots of places in code where we have # type: ignore - any help would fixing these would be greatly appreciated

Add YD01 to validate_docstrings codes

and fix related errors:

/Users/marcogorelli/pandas-dev/pandas/core/groupby/groupby.py:753:YD01:pandas.core.resample.Resampler.iter:No Yields section found
/Users/marcogorelli/pandas-dev/pandas/core/groupby/groupby.py:753:YD01:pandas.core.groupby.GroupBy.iter:No Yields section found
/Users/marcogorelli/pandas-dev/pandas/core/groupby/groupby.py:2364:YD01:pandas.core.groupby.GroupBy.ohlc:No Yields section found
/Users/marcogorelli/pandas-dev/pandas/core/frame.py:1345:YD01:pandas.DataFrame.itertuples:No Yields section found
/Users/marcogorelli/pandas-dev/pandas/core/frame.py:9552:YD01:pandas.DataFrame.round:No Yields section found

marcogorelli / pyladieslondon-sprints Goto Github PK

pyladieslondon-sprints's Introduction

pyladieslondon-sprints's People

Stargazers

Watchers

pyladieslondon-sprints's Issues

Welcome!

Recommend Projects

Recommend Topics

Recommend Org