Closed mawc2019 closed 11 months ago
Merging #2599 (9c48402) into master (cd9a7eb) will decrease coverage by
0.09%
. Report is 1 commits behind head on master. The diff coverage is86.44%
.
:exclamation: Your organization is not using the GitHub App Integration. As a result you may experience degraded service beginning May 15th. Please install the Github App Integration for your organization. Read more.
@@ Coverage Diff @@
## master #2599 +/- ##
==========================================
- Coverage 73.90% 73.81% -0.09%
==========================================
Files 18 18
Lines 5293 5294 +1
==========================================
- Hits 3912 3908 -4
- Misses 1381 1386 +5
Files Changed | Coverage Δ | |
---|---|---|
python/adjoint/filters.py | 77.20% <86.44%> (-2.49%) |
:arrow_down: |
One suggestion would be to simply verify that defining a filter with the resolution tuple
(res, res)
produces the same result as passing in justres
.
I have added a different but more stringent test, in which one of the elements of the resolution
tuple is zero, which defines a 1d design pattern. Suppose a 1d design pattern is defined in the x direction. Whether this 1d design pattern is periodic in the y direction or not should make no difference in the filtered 1d patterns. Likewise, suppose a 1d design pattern is defined in the y direction. There should be no difference whether this 1d design pattern is periodic in the x direction. The new test function checks this identity based on lengthscale constraint functions.
This PR allows the
resolution
argument in filters.py to accept a tuple, so that different resolutions along different directions can be supported. This PR also refactors the code in filters.py.