Reject underflowing sigma_spatial in bilateral() (#1390)#1392
Merged
Conversation
`bilateral()` validated `sigma_spatial > 0` only. At values below `sqrt(np.finfo(float64).tiny)` (~1.49e-154) the term `2 * sigma**2` underflows to zero or a denormal, so `1/(2*sigma**2)` is `+inf` (or raises ZeroDivisionError under numba) and `exp(-dist2 * inf) = 0 * inf = NaN` propagates through the whole output. Tighten the bound on `sigma_spatial` and `sigma_range` to `>= sqrt(tiny)` and update the docstring. Add three tests covering the underflow-NaN and underflow-ZeroDivision cases, the zero-sigma regression, and a realistic positive case.
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Fixes #1390.
bilateral()validatedsigma_spatial > 0only. Values small enough that2 * sigma**2underflows to a denormal or to zero made1/(2*sigma**2)either+inf(exp(-dist2 * inf) = 0 * inf = NaN, all-NaN output) or raisedZeroDivisionErrorunder numba. This is the MEDIUM-severity Cat 3 finding deferred from PR #1238.The bound on both
sigma_spatialandsigma_rangeis tightened to>= sqrt(np.finfo(float64).tiny)(~1.49e-154), the smallest value that keeps the squared term a normal float64. Realistic sigmas are unaffected; nonsense values now raiseValueErrornaming the parameter.The same threshold is applied to
sigma_rangefor symmetry, since the same underflow occurs ininv_2_sr.Test plan
pytest xrspatial/tests/test_bilateral.py-- 22 passed (19 prior + 3 new)test_bilateral_rejects_underflow_sigma_spatialcoverssigma_spatial=1e-200(wouldZeroDivisionError),1e-160(would all-NaN), andsigma_range=1e-200test_bilateral_rejects_zero_sigmaregresses the previous> 0rejection forsigma_spatial=0andsigma_range=0test_bilateral_accepts_normal_sigmaconfirmssigma_spatial=sigma_range=1.0still produces a finite result