Core _io.concatenate() may fail due to case when one of the cubes is scalar - this fixes that
#1715
Conversation
Codecov Report
@@ Coverage Diff @@
## main #1715 +/- ##
==========================================
- Coverage 91.52% 91.47% -0.06%
==========================================
Files 206 206
Lines 11202 11204 +2
==========================================
- Hits 10253 10249 -4
- Misses 949 955 +6
Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here. |
|
hold yer horses, I'll improve the test coverage - @bouweandela why is our test coverage test passing even if the coverage is shrinking? Go home test, you drunk 🍺 |
|
The test coverage test tests that all the changed code is covered by tests. So if you delete some code that is covered by tests from a file with less than 100% coverage (such as the debug log call you removed), the relative amount of covered code decreases, while the changed code is fully covered because there isn't any. |
| cubes = iris.cube.CubeList([c1_delta, cubes[1]]) | ||
| logger.debug("Attempting concatenatenation of %s with %s", |
There was a problem hiding this comment.
Did you want to remove this?
There was a problem hiding this comment.
not at all, I was being silly - cheers for pointing it out, bud, will reinstate 👍
makes sense, cheers, Bouwe! V, go home, you're drunk 😁 |
|
cheers for the review @bouweandela 🍺 Your tests should all go fine now in your Dasky PR |
Description
Closes #1716
The introduction of scalar cubes in iris complicates a number of things, including our advanced concatenation mechanism - it is possible, as seen from the test case changed in this PR, that along the path of concatenation as we do it, one of the processed cubes that we use as intermediary laison to be scalar - then the whole show crumbles. I am not a fan of scalar cubes as they are defined in iris, but that's for another day. Anyway, this PR fixes that particular corner case, and adds a test case (converts the previous test case to a more meaningful one, that is).
All started from @bouweandela 's comment - cheers, Bouwe! 🍺
Before you get started
Checklist
It is the responsibility of the author to make sure the pull request is ready to review. The icons indicate whether the item will be subject to the 🛠 Technical or 🧪 Scientific review.
To help with the number pull requests: