PERF: Additional performance updates #1209
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The __hash__() calls add some overhead for the relatively few number items that we can inline instead.
greglucas
commented
Dec 20, 2022
greglucas
force-pushed
the
perf-updates
branch
from
964e6eb to
58395b1
Compare
|
This is great! How much does it improve the performance? |
|
I go from about 1.5 us -> 1.1 us / iteration of transform_point (rebased on your branch). Maybe see the results of calling it 2,000,000 times to get the cumulative. Here are the figures below relative to the original one included above, although I'm not sure we really get that much of a speedup. After the first commit: After the second commit: |
|
Every little speedup counts. I think that this is a worthwhile change. Thanks for working on this 👍 |
This moves the internal PROJ error message to Cython for faster get/clear from the C side. The Python side can still access it, it is just no longer stored as a class attribute. This provides a performance benefit primarily within the Transformer functions.
greglucas
force-pushed
the
perf-updates
branch
from
58395b1 to
76f7c5f
Compare
snowman2
reviewed
Dec 21, 2022
|
Thanks @greglucas 👍 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This is whittling down the last two easy to find performance bottlenecks using snakeviz. See the image below.
Commit 1: Removes the hash calls to enum from the dict access. It is small enough and only used in one place that we can inline it.
Commit 2: (far right purple box is
ProjError.clear()) Removes the class attribute of the internal_proj_error and moves that into Cython to take care of all the error messages without needing to access Python objects while within C.