Closed jGaboardi closed 1 year ago
This is currently passing for Python 3.8 Ubuntu, before failing for Python 3.6 Ubuntu at test_ripley.py::test_primitives
with INTERNALERROR> RecursionError: maximum recursion depth exceeded
. Will try narrowing down the error source by restricting environments in unittests.yml
.
jGaboardi/pointpats
passes on:
Python/OS | macOS | Ubuntu | Windows |
---|---|---|---|
3.6 | ❌ | ❌ | ❌ |
3.7 | ✅ | ✅ | ✅ |
3.8 | ✅ | ✅ | ✅ |
Can confirm that the Python 3.6 test of No longer supporting Python 3.6test_ripley.py::test_primitives
fails for every OS on GHA and also macOS locally with:
After digging in here, it appears that No longer supporting Python 3.6pointpats.geometry.area(shape)
is the culprit, specifically line 38 (return area(numpy.asarray(shape))
), that is breaking in Python 3.6 tests. Not sure what the actually issue is here though. Any thoughts, @sjsrey, @ljwolf, @weikang9009, @shaohu?
Something that I should have checked first, but did not, was the current status of TravisCI builds for No longer supporting Python 3.6pointpats
... Seems that this has been a known issue for Python 3.6?
And now I see the actual commit that addresses this issue. This is an issue specific to pytest
?
FWIW, dramatically increasing the recursion limit (sys.setrecursionlimit(25000)
) does not solve the problem. Maybe the shape
parameter is not being registered?
The same tests also fail with nose
, so this is not a pytest
issue. Also, more nose
gives more detailed (albeit messy) feedback and seems to confirm the problem with the return area(numpy.asarray(shape))
recursion. There also appear to be errors with missing positional arguments.
Looking at & updating this PR again; now getting 1 error for test_centrography
, perhaps related to #69, but I couldn't pull up the job diagnostic because travis.org is no longer a thing (?).
@sjsrey maybe you could grant @ljwolf and I maintainership? @shaohu would that be OK with you?
@jGaboardi I see that you have already started working on this some time ago. We shall probably move some stuff from here to #87 or to update this once that one gets merged. This got out-of-date in the meantime.
I completely for got about this...
We should include .coveragerc
in the pytest call.
I'd love to get this merged and be done with it, but we still be enable codecov
, etc...
Merging #64 (54d0595) into master (6723b40) will decrease coverage by
10.68%
. The diff coverage isn/a
.
@@ Coverage Diff @@
## master #64 +/- ##
===========================================
- Coverage 63.15% 52.47% -10.68%
===========================================
Files 19 12 -7
Lines 2266 1841 -425
Branches 0 315 +315
===========================================
- Hits 1431 966 -465
+ Misses 835 819 -16
- Partials 0 56 +56
Impacted Files | Coverage Δ | |
---|---|---|
pointpats/window.py | 69.56% <0.00%> (-8.70%) |
:arrow_down: |
pointpats/distance_statistics.py | 75.00% <0.00%> (-8.42%) |
:arrow_down: |
pointpats/quadrat_statistics.py | 56.45% <0.00%> (-4.84%) |
:arrow_down: |
pointpats/spacetime.py | 89.37% <0.00%> (-3.75%) |
:arrow_down: |
pointpats/_deprecated_distance_statistics.py | 49.51% <0.00%> (-3.37%) |
:arrow_down: |
pointpats/pointpattern.py | 70.05% <0.00%> (-2.83%) |
:arrow_down: |
pointpats/geometry.py | 82.27% <0.00%> (-2.00%) |
:arrow_down: |
pointpats/process.py | 49.07% <0.00%> (-1.86%) |
:arrow_down: |
pointpats/random.py | 25.46% <0.00%> (-1.25%) |
:arrow_down: |
pointpats/centrography.py | 46.03% <0.00%> (-1.00%) |
:arrow_down: |
... and 8 more |
Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.
@jGaboardi merge when ready
Yeah, I guess we can merge for now and then update README for codecov, etc. later.
This PR addresses #63.
Still needs attention from a maintainer:
coveralls
CODECOV_TOKEN
inpysal/pointpats/Setting/Secrets