Codecov Report

❌ Patch coverage is 71.42857% with 4 lines in your changes missing coverage. Please review.
✅ Project coverage is 74.46%. Comparing base (36a3938) to head (fcfd99e).
⚠️ Report is 26 commits behind head on develop.

@@             Coverage Diff             @@
##           develop     #589      +/-   ##
===========================================
+ Coverage    74.33%   74.46%   +0.13%     
===========================================
  Files           26       26              
  Lines         7878     7887       +9     
  Branches      1633     1637       +4     
===========================================
+ Hits          5856     5873      +17     
+ Misses        1541     1532       -9     
- Partials       481      482       +1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

Wrote a test in C. I suppose this is covered by CI. Am I correct, @LecrisUT?

I found a critical bug in python interface.

This lattice has to be transposed. I think this as well, @lan496?

Should I be better to make another PR for this fix?

I found a critical bug in python interface.

I did find some inconsistencies, where the transpose is done on the Cpp side. Don't remember which ones these are, but I will try to consolidate them in #579 and do them all in the cpp side.

Edit: Checked the cpp side. It's not transposed there either :/

Don't remember which ones these are,

In python, crystal structure is given in this format, but for get_spacegroup_type_from_symmetry, only lattice (basis vectors) is given, so this transpose has to be handled in that function without relying on _expand_cell function.

How about making another PR for the fix of the python interface? If this way, we can first merge this PR after approval, then I will make the fix and its PR.

How about making another PR for the fix of the python interface? If this way, we can first merge this PR after approval, then I will make the fix and its PR.

Of course. Do you think you can find a test where it breaks? Usually I make the test commit first and run the CI to confirm it actually breaks/covers the thing we are fixing (in this case I confirmed locally by reverting the commit).

Do you think you can find a test where it breaks? Usually I make the test commit first and run the CI to confirm it actually breaks/covers the thing we are fixing (in this case I confirmed locally by reverting the commit).

I will try in this way.

I found a critical bug in python interface.

@atztogo I think it is better to separate PR

Do you think you can find a test where it breaks?

Possibly, get_spacegroup_type_from_symmetry with a hexagonal lattice will return an unexpected result