Skip to content

Conversation

@johntruckenbrodt
Copy link

Just stumbled across this. Using scipy.ndimage.uniform_filter should be more performant than the custom function cslc_utils.moving_window_mean. The results are nearly identical.

rubiedhillon and others added 30 commits February 7, 2023 13:50
Update nodejs in environment.yml, as earlier version (14.18.3) is not supported on the Mac M1.
@review-notebook-app
Copy link

Check out this pull request on  ReviewNB

See visual diffs & provide feedback on Jupyter Notebooks.


Powered by ReviewNB

scottstanie added a commit to scottstanie/dolphin that referenced this pull request Jan 31, 2024
See OPERA-Cal-Val/OPERA_Applications#35
`uniform_filter` seems to run faster for all input/filter sizes.
scottstanie added a commit to isce-framework/dolphin that referenced this pull request Jan 31, 2024
* Deprecate `moving_window_mean` in favor of `uniform_filter`

See OPERA-Cal-Val/OPERA_Applications#35
`uniform_filter` seems to run faster for all input/filter sizes.

* call `nearest` for mode, now there are not as obvious edge effects
@scottstanie
Copy link

Not sure if I thought the other numpy way was faster or what, but thanks for pointing this out @johntruckenbrodt !

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

9 participants