fix LCS time handling for xarray>=2023.4.0

[CagtayFabry]

Some time interpolation code seems to break with the new xarray release (seemingly unrelated of pandas major version 1 or 2)

Changes

• add explicit handling of reference time for coordinates/orientation in LocalCoordinateSystem code
• set keep_attrs=True when applying xarray.apply_ufunc

[codecov[bot]]

Codecov Report

Merging #868 (5743a01) into master (9cb7f08) will not change coverage. The diff coverage is 100.00%.

@@           Coverage Diff           @@
##           master     #868   +/-   ##
=======================================
  Coverage   96.51%   96.51%           
=======================================
  Files          94       94           
  Lines        6283     6283           
=======================================
  Hits         6064     6064           
  Misses        219      219           

Impacted Files	Coverage Δ
weldx/transformations/local_cs.py	96.35% <ø> (ø)
weldx/util/xarray.py	96.15% <100.00%> (ø)

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[github-actions[bot]]

Test Results

2 189 tests  ±0   2 188 :heavy_check_mark: ±0   4m 3s :stopwatch: +2s        1 suites ±0          1 :zzz: ±0         1 files   ±0          0 :x: ±0 

Results for commit 5743a01c. ± Comparison against base commit 9cb7f08d.

:recycle: This comment has been updated with latest results.

[marscher]

Thanks! It'd be good to have a stack trace or failed report inside an issue, since the logs gets deleted after some time.

[CagtayFabry]

Thanks! It'd be good to have a stack trace or failed report inside an issue, since the logs gets deleted after some time.

The thing with these failures it the error itself is not that important, more the environment and installed packages. But we can paste those here

[marscher]

It is also strange, that the conda build also installs pandas 1.5.3, but the tests succeed.

[CagtayFabry]

It is also strange, that the conda build also installs pandas 1.5.3, but the tests succeed.

The problem seems to be related to xarray 2023.4.0, which is available on PyPi but not released on conda-forge yet

(maybe the error would also arise with pandas 2 but we cant test that against an older working xarray version)

[CagtayFabry]

I have tested this manually with old xarray=2022.9.0 and pandas=1.3.5 versions

The new code is more explicit in achieving the desired behavior. However much of the time related code should be reworked soon to reduce the compelxity