Skip to content

PERF: Use fastpath for accessing option value in internals #49771

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 3 commits into from
Dec 8, 2022
Merged

PERF: Use fastpath for accessing option value in internals #49771

Changes from 1 commit
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
400b33a
PERF: Use fastpath for accessing option value in internals
jorisvandenbossche Nov 18, 2022
0e76c1b
Merge remote-tracking branch 'upstream/main' into perf-cow-check-inte…
jorisvandenbossche Nov 28, 2022
97b1a76
add test
jorisvandenbossche Nov 28, 2022
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
7 changes: 5 additions & 2 deletions pandas/core/internals/managers.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,7 +15,7 @@

import numpy as np

from pandas._config import get_option
from pandas._config.config import _global_config
Copy link
Member

Choose a reason for hiding this comment

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

is this much faster than using pandas._config.config.options?

Copy link
Member Author

Choose a reason for hiding this comment

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

It's in any case faster, since _global_config is just a plain python dict (which is quite fast), while the options object is a custom class mimicking attribute access:

In [32]: %timeit _global_config["mode"]["copy_on_write"]
52.6 ns ± 1.85 ns per loop (mean ± std. dev. of 7 runs, 10,000,000 loops each)

In [36]: %timeit pd._config.config.options.mode.copy_on_write
2.78 µs ± 22 ns per loop (mean ± std. dev. of 7 runs, 100,000 loops each)

Strangely it seems that this is even slower than the pd.get_option("mode.copy_on_write"):

In [37]: %timeit pd.get_option("mode.copy_on_write")
1.38 µs ± 14.6 ns per loop (mean ± std. dev. of 7 runs, 1,000,000 loops each)

Copy link
Member Author

@jorisvandenbossche jorisvandenbossche Nov 18, 2022 •
edited
Loading

Choose a reason for hiding this comment

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

If we really want to avoid a private import, this is also an option:

In [39]: %timeit pd.options.d["mode"]["copy_on_write"]
85.9 ns ± 2.39 ns per loop (mean ± std. dev. of 7 runs, 10,000,000 loops each)

The options DictWrapper class seems to have a "public" d attribute exposing the underlying dict (_global_config)

Copy link
Member

Choose a reason for hiding this comment

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

not worth tying ourself in knots over, thanks for checking


from pandas._libs import (
algos as libalgos,
Expand Down Expand Up @@ -2377,5 +2377,8 @@ def _preprocess_slice_or_indexer(
return "fancy", indexer, len(indexer)


_mode_options = _global_config["mode"]


def _using_copy_on_write():
return get_option("mode.copy_on_write")
return _mode_options["copy_on_write"]