Skip to content

DOC: Fix #24268 by updating description for keep in Series.nlargest #25358

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 11 commits into from
Mar 5, 2019
Merged

DOC: Fix #24268 by updating description for keep in Series.nlargest #25358

Changes from 3 commits
Commits
Show all changes
11 commits
Select commit Hold shift + click to select a range
2ebe8ae
DOC: Fix #24268 by updating description for keep
bharatr21 Feb 18, 2019
ab33dc1
Fix PEP8 issues and a typo
bharatr21 Feb 18, 2019
794f370
Fix PEP8 issues
bharatr21 Feb 18, 2019
28aa45b
Update the description of keep
bharatr21 Feb 19, 2019
c9de2f6
Fix PEP8 issues again
bharatr21 Feb 19, 2019
62ccfed
Fix trailing whitespaces
bharatr21 Feb 19, 2019
8506a6e
Fix some docstring issues (as produced by tests)
bharatr21 Feb 19, 2019
fdb6ba7
Take suggestions from @TomAugspurger and @WillAyd to improve readability
Feb 27, 2019
310682a
Rewording docs according to @WillAyd 's suggestion
bharatr21 Mar 1, 2019
5c56fad
Merge remote-tracking branch 'upstream/master'
bharatr21 Mar 2, 2019
9a8367d
Removed the confusing lines as suggested by @jreback
bharatr21 Mar 3, 2019
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 9 additions & 5 deletions pandas/core/series.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3100,8 +3100,10 @@ def nlargest(self, n=5, keep='first'):
When there are duplicate values that cannot all fit in a
Series of `n` elements:

- ``first`` : take the first occurrences based on the index order
- ``last`` : take the last occurrences based on the index order
- ``first`` : return the first `n` occurrences in the
Copy link
Contributor

@TomAugspurger TomAugspurger Feb 18, 2019 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is "given index order" the correct description? I interpreted
#24268 (comment) as a modification on the result. So below the descriptions of first / last / all, I would note something like

The `keep` parameter determines which ones to keep when there are duplicates.
Regardless of `keep`, the result will be sorted by the row label.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I will include this, but when keep="last", the row labels are sorted in reverse order even in the examples in #24268
And I feel "given index order" is correct because it is "given" by the user as input, if it is not an implicit RangeIndex

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I understand what you are writing here but I find the readability rather difficult. Any reason we don't use the same terminology in the unique docstring?

Uniques are returned in order of appearance

^ obviously accounting for reversal with the last argument

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@WillAyd What do you suggest me to do? If it's changing the unique docstring, I want to leave this as-is, because I still want to somehow emphasize the fact that index order is reversed when keep="last"

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

My suggested wording is Return the first n occurrences in order of appearance for first and then Return the last n occurrences in reverse order of appearance

given index order.
- ``last`` : return the last `n` occurrences in the
reverse of the given index order.
- ``all`` : keep all occurrences. This can result in a Series of
size larger than `n`.

Expand Down Expand Up @@ -3196,8 +3198,10 @@ def nsmallest(self, n=5, keep='first'):
When there are duplicate values that cannot all fit in a
Series of `n` elements:

- ``first`` : take the first occurrences based on the index order
- ``last`` : take the last occurrences based on the index order
- ``first`` : return the first `n` occurrences in the
given index order.
- ``last`` : return the last `n` occurrences in the
reverse of the given index order.
- ``all`` : keep all occurrences. This can result in a Series of
size larger than `n`.

Expand Down Expand Up @@ -3238,7 +3242,7 @@ def nsmallest(self, n=5, keep='first'):
Monserat 5200
dtype: int64

The `n` largest elements where ``n=5`` by default.
The `n` smallest elements where ``n=5`` by default.

>>> s.nsmallest()
Monserat 5200
Expand Down