Is there a special reason of not using numpy in AffineTransformer class? #2935
Replies: 2 comments
-
Linked pull request: #2936 |
Beta Was this translation helpful? Give feedback.
0 replies
-
My pull request (#2936) was merged into main by @sgillies. Thank you @sgillies and @underchemist for reviewing! |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
When I look at AffineTransformer class in transform.py, I see that a collection of points are transformed via a loop where in each step transformer is called one at a time. This kind of implementation is very slow when there are a million pair of points (which is the situation in my case).
My question is that, is there a specific reason why numpy's matrix vector multiplication have not been used in this implementation. I've reimplemented with numpy and the speed gain is enormous. Before I submit a PR, I just want to make sure that numpy-type implementation is welcome by the rasterio community.
Thanks
Existing class:
Beta Was this translation helpful? Give feedback.
All reactions