metpy_1.5.1+ds-2_source.changes ACCEPTED into unstable

[Sebastiaan Couwenberg]

On 8/2/23 05:29, Sebastiaan Couwenberg wrote:
> Testing migration is blocked by autopkgtest failure on arm64:
> 
> 167s =================================== FAILURES 
> ===================================
> 167s ___________________________ test_moist_lapse_failure 
> ___________________________
> 167s
> 167s     @pytest.mark.xfail(sys.platform == 'win32', reason='solve_ivp() 
> does not error on Windows')
> 167s     @pytest.mark.xfail(packaging.version.parse(scipy.__version__) < 
> packaging.version.parse('1.7'),
> 167s                        reason='solve_ivp() does not error on Scipy 
> < 1.7')
> 167s     def test_moist_lapse_failure():
> 167s         """Test moist_lapse under conditions that cause the ODE 
> solver to fail."""
> 167s         p = np.logspace(3, -1, 10) * units.hPa
> 167s >       with pytest.raises(ValueError) as exc:
> 167s E       Failed: DID NOT RAISE <class 'ValueError'>
> 167s
> 167s tests/calc/test_thermo.py:206: Failed
> 
> https://ci.debian.net/data/autopkgtest/testing/arm64/m/metpy/36313527/log.gz

I've added a patch to use xfail for this test on arm64 like it does on 
Windows and with old SciPy.

Do you think we should upstream that, or dig deeper why it doesn't raise 
an exception on arm64?

Kind Regards,

Bas

-- 
  GPG Key ID: 4096R/6750F10AE88D4AF1
Fingerprint: 8182 DE41 7056 408D 6146  50D1 6750 F10A E88D 4AF1