Small area error - specific to one ERA5 pixel?

[Claire Michailovsky]

I ran into an issue when running defaults of v3L2- chunking could not be done because x, y were not found.

x,y were removed as dimensions through .squeeze in aligner.py - this means x,y were coordinates but not dimensions. Perhaps an issue because I used a small area with only one ERA5 pixel, removing .squeeze from aligner.py and compositer.py worked.

Note: running this produced a properly opened output in QGIS also on my QGIS. Testing creating inputs for other cases without .squeeze - to see if my QGIS would react better, taking very long, not a good idea.

‌

[bert.coerver]

can you add the log file?

[Claire Michailovsky]

  <div class="preview-container wiki-content"><!-- loaded via ajax --></div>
  <div class="mask"></div>
</div>

</div> </form>

[Claire Michailovsky]

Hi Bert, the log doesn’t say much - I reran this morning with the original code (including squeeze) to give you a cleaner version of the log. Here is a screenshot of the error:

‌

[Claire Michailovsky]

Bounding box in case:

bb = [33.255, 14.275, 33.285, 14.295]

[bert.coerver]

thanks, normally the log should include the error?

[Claire Michailovsky]

It does not for some reason though - unsure why.

[bert.coerver]

fixed in https://bitbucket.org/cioapps/pywapor/commits/70f101f7385f1bb407be9b1cfa3e7def60e4f6d3

and error not showing in log fixed in https://bitbucket.org/cioapps/pywapor/commits/fc57514713341d88d684e4b435cd697c3cfc113a