Add DataMet backend for Italian Radar data

[wolfidan]

Dear xradar people, this is a first try at adding a new backend to read the data from the Italian radars provided by Gianfranco Vulpiani that we plan to use at the open radar course at ERAD2024.

https://zenodo.org/records/4897697

I must say that I had some trouble understanding all the classes that are required to wrap it into xarray, but I tried my best and it seems to work, as I was able to read all 165 provided files. I have added some tests but they are somewhat simplistic at the moment and only test the basic reader not the wrapper to xarray (other tests seem to do the same things).

The DataMet files are not consistent in which sweep corresponds to which elevation: sometimes sweep 1 could be 16° other times 90°. I find it a bit weird but left it as is and didn't try to remap it. I wonder if this could be an issue when reading them into pyart?

Thank you for your time.
Next week I will be in vacation and might not answer immediately, sorry about that.

• Closes Add reader for italian radar data data provided by Gianfranco Vulpiano erad2024#7
• Tests added
• Changes are documented in history.md

[kmuehlbauer]

Thanks @wolfidan! The ruff and numpy2 issues should be resolved now by #177. I'll have a closer look here later.

[codecov]

Codecov Report

Attention: Patch coverage is 93.77593% with 15 lines in your changes missing coverage. Please review.

Project coverage is 91.70%. Comparing base (89d9a3e) to head (d12815d).
Report is 18 commits behind head on main.

Files with missing lines	Patch %	Lines
xradar/io/backends/datamet.py	93.72%	15 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #175      +/-   ##
==========================================
+ Coverage   91.58%   91.70%   +0.12%     
==========================================
  Files          21       22       +1     
  Lines        3991     4231     +240     
==========================================
+ Hits         3655     3880     +225     
- Misses        336      351      +15     

Flag	Coverage Δ
notebooktests	77.21% <24.06%> (-3.22%)	⬇️
unittests	89.83% <93.77%> (+0.23%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

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

[kmuehlbauer]

@wolfidan I've taken the chance to fix this up to green CI. Looks good so far. We might just copy/adapt code/tests from the other backends to finish this.

Thanks for this nice addition!

[kmuehlbauer]

This works for me with different elevations. Looks that this is a top down strategy.

import open_radar_data
import xarray as xr
fname = open_radar_data.DATASETS.fetch("H-000-VOL-ILMONTE-201907100700.tar.gz")
with xr.open_dataset(fname, engine="datamet", group="sweep_8") as ds:
    display(ds)
    ds = ds.set_coords("sweep_mode")
    ds = ds.wrl.georef.georeference()
    ds.DBTH.wrl.vis.plot()

[kmuehlbauer]

The DataMet files are not consistent in which sweep corresponds to which elevation: sometimes sweep 1 could be 16° other times 90°.

Maybe it's that the 90deg sweep is not measured in every volume?

[wolfidan]

That's amazing, thanks a lot @kmuehlbauer, I will add more tests the week after the next to increase codecov

[wolfidan]

@kmuehlbauer
All tests and codecov should be fine now. I still had an issue.

    ds = xr.open_dataset(datamet_file, engine="datamet")

doesn't work. I had to use

   ds = xr.open_dataset(datamet_file, engine=xradar.io.backends.DataMetBackendEntrypoint)

in the unit tests

Maybe you can easily fix that. I'm looking forward to see you do this, so I can do it next time.
Thanks a lot

[mgrover1]

Can you try adding back in the engine="datamet" now that the linting issues have been resolved?

[kmuehlbauer]

Not sure why it doesn't work for you. But you would need to install the package first before being able to use the entrypoints.

[wolfidan]

Thanks @mgrover1 @kmuehlbauer , I'll change it very soon and see if it passes the tests

[wolfidan]

Ok I changed according to your propositions.

[kmuehlbauer]

Great work @wolfidan!

[mgrover1]

Agreed - thanks so much for this contribution @wolfidan !!