Repository navigation
Conversation
Author
|
I found a compatibility regression while auditing this change: NaN detection rejected large Python integer nodata values. The updated commit preserves equality masking for those values and adds regressions for |
This file contains hidden or 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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Use
isnanwhen a NumPy raster's nodata value is NaN. Equality never marks these cells as missing, so nearest queries return NaN instead ofNone, and bilinear queries can return NaN for a valid sample beside a missing cell.The regression covers both interpolation modes and points outside the raster. A replay of the bundled DEM through arrays, read-only memory maps and GeoTIFFs matches an independent pixel-space reference after the fix; working point queries and polygon summaries remain unchanged. The installed wheel gives the same results.
All 128 tests pass with the older dependency profile. With current Affine/Shapely, two existing deprecation warnings require filtering; the unrestricted baseline has the same failures. Ruff checks and formatting pass for the changed files.
Large Python integer nodata values retain equality masking; regressions cover
2**64and10**20in both interpolation modes.