Closed kmuehlbauer closed 6 months ago
Finally linting is good. Had outdated version on my machine :grimacing:
@mgrover1
This only reads the absolutely necessary metadata to get a hunch on sweeps/moments. Everything else is read when first requested by xarray (so it's lazy too). The reader uses numpy.memmap (like the sigmet reader). Will push the fixture in a minute...
Attention: Patch coverage is 94.62963%
with 29 lines
in your changes are missing coverage. Please review.
Project coverage is 91.30%. Comparing base (
a854715
) to head (63301e9
).
Files | Patch % | Lines |
---|---|---|
xradar/io/backends/nexrad_level2.py | 94.51% | 29 Missing :warning: |
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
@mgrover1 Somehow my rebase threw away to much. Finally this is working again. Maybe we can talk in the next days, if necessary?
yes! next week's radar community meeting?
Check out this pull request on
See visual diffs & provide feedback on Jupyter Notebooks.
Powered by ReviewNB
The docs and test coverage look great - we can continue to iterate + test through other PRs - what do you think about merging @kmuehlbauer ?
Let's wait after the meeting today. There are some issues which I want to show before we get this in.
I am putting together a notebook right now comparing the data fields from Py-ART + xradar here as well... which should help us understand some of the differences?
I am putting together a notebook right now comparing the data fields from Py-ART + xradar here as well... which should help us understand some of the differences?
Great idea! One major difference is that this PR does not mask the data. We would need to think how we want to handle masking in the future. But we can make first make a list with the current differences.
@mgrover1 If you think we should move forward from here, please go ahead and merge. I'll add several issues dedicated to different aspects/downsides of this implementation. Please also add any issues you find to the tracker, so that we can work on fixing enhancing.
Great!! Agreed 😄
@kmuehlbauer - should we close out #40 here and open new issues related to more specific improvements?
Or let's use this issue (#40) as a tracking issue. We could list all related PR/issues there? Can also pin that issue to the top. WDYT? I'm also good with closing #40.
Ohh I like that idea... that can be the issue we refer back to + scaffold out. We can rename it to NEXRAD Support, and refer to the different readers and improvements.
I created #160 to reflect this single reader, and modified #40 accordingly. Sound reasonable @kmuehlbauer ?
@mgrover1 Yes, sounds good.
history.md
This is a rewrite of #147 using parts of the structured reader of the iris/sigmet reader. This also enhances the ability of reading much more metadata from the nexrad level 2 archive.
Note: this only implements MSG_31 as of now. MSG_1 need to be included after adding such file to open-radar-data repo.
There is also relevant code over in https://github.com/jthielen/xradar/tree/nexrad-level2 by @jthielen, which should be added to this structured reader in a follow-up PR.