| FazBrowse GitHub Viewer | Trending | | Home |
| Tools: [Download Repo ZIP] [Original HTTPS Page] |
Sorry, something went wrong.
There was a problem hiding this comment.
Looks good to me. Any reason bbox.update_from_data_xy is so slow?
Sorry, something went wrong.
|
Because it constructs general path objects and then iterates over it using overly general mechanisms, I guess. |
Sorry, something went wrong.
update_from_data_xy is much slower than needed; we can use plain numpy instead. For polylines, we can completely skip the bezier handling for even more speedup.
|
As a performance regression fix, should this be backported to 3.3.1? |
Sorry, something went wrong.
|
I think the regression has been there since 3.0.0, so not urgent, unless there is something more than the tight_bbox PR that added to the slowdown. |
Sorry, something went wrong.
|
Further discussion on #17974 says this was introduced in 3.3.0, so maybe we should backport this? |
Sorry, something went wrong.
…extents; special case polylines.
…995-on-v3.3.x Backport PR #17995 on branch v3.3.x (Avoid using Bbox machinery in Path.get_extents; special case polylines.)
| Back | FazBrowse Home | New Git URL |
update_from_data_xy is much slower than needed; we can use plain numpy
instead. (This speeds up #17974 another ~30-40% on top of #17994.)
For polylines, we can completely skip the bezier handling for even more
speedup. (This fixes the performance regression.)
PR Summary
PR Checklist