search

jorisvandenbossche / pydata-amsterdam-pandas-sprint

1 stars 3 forks source link

Set `__module__` on top-level public objects #8

Open jorisvandenbossche opened 1 year ago

jorisvandenbossche commented 1 year ago

See https://github.com/pandas-dev/pandas/issues/55178 for context.

An initial PR has been done for DataFrame by @aimlnerd, see the changes that were needed for this here: https://github.com/pandas-dev/pandas/pull/55171/files

Next, separate PRs can be done for other top-level objects:

aimlnerd commented 1 year ago

take

jorisvandenbossche commented 8 months ago

Updated the list above with a task list that can be done as separate PRs

Amir-101 commented 1 month ago

Sent PR for modification, everything has been modified except for the Interval class. This class is defined as a cython function which is causing issues.

simonjayhawkins commented 2 weeks ago

@jorisvandenbossche

An initial PR has been done for DataFrame by @aimlnerd, see the changes that were needed for this here: https://github.com/pandas-dev/pandas/pull/55171/files

This has a milestone of 3.0.0 and I don't see any other PRs related to this.

so on main, we get say

>>> pd.DataFrame
<class 'pandas.DataFrame'>
>>> pd.Series
<class 'pandas.core.series.Series'>
>>> pd.Index
<class 'pandas.core.indexes.base.Index'>
>>>

I guess that we would want the changes for all the objects to happen in the same release? and that a release note could be helpful. Also, is this a breaking change?

jorisvandenbossche commented 2 weeks ago

It's indeed a breaking change (I don't expect much actual usage to depend on this, but tests / docs that rely on the output of this might depend on it), and indeed still needs a whatsnew note.

I guess that we would want the changes for all the objects to happen in the same release?

Ideally yes (I was planning to do a bunch of them just before 3.0rc if they don't get done by then)

I don't see any other PRs related to this.

There are two others (you can see them linked on the issue https://github.com/pandas-dev/pandas/issues/55178), but I have done a bad job following up on those PRs and getting them merged. Will do that now.

jorisvandenbossche commented 2 weeks ago

What is certainly not yet done is Series, and top-level functions like pd.concat and pd.date_range etc

simonjayhawkins commented 2 weeks ago

There are two others (you can see them linked on the issue pandas-dev/pandas#55178), but I have done a bad job following up on those PRs and getting them merged. Will do that now.

Thanks @jorisvandenbossche

missed that! simply searching for set_module didn't show the open PRS

jorisvandenbossche commented 2 weeks ago

I updated the top post in the issue https://github.com/pandas-dev/pandas/issues/55178 with a list of done PRs and remaining items

simonjayhawkins commented 2 weeks ago

Great 😃👍 thanks

On Fri, 8 Nov 2024, 08:24 Joris Van den Bossche, @.***> wrote:

I updated the top post in the issue pandas-dev/pandas#55178 https://github.com/pandas-dev/pandas/issues/55178 with a list of done PRs and remaining items

— Reply to this email directly, view it on GitHub https://github.com/jorisvandenbossche/pydata-amsterdam-pandas-sprint/issues/8#issuecomment-2464104842, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADEMUXLIAJ7XJKL62RUPETLZ7RYKPAVCNFSM6AAAAAA42XIQXSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDINRUGEYDIOBUGI . You are receiving this because you commented.Message ID: <jorisvandenbossche/pydata-amsterdam-pandas-sprint/issues/8/2464104842@ github.com>