Bad gridlengthXDirection when reading gaussian grid

[alcoat]

Input files could be CSFR V2 or any CFS forecast or may be any files with a gaussian grid

import grib2io
file = "flxf2024011700.01.2024011700.grb2"
g = grib2io.open(file)
msg = g.select(shortName='UGRD')[0]
print(msg.gridlengthXDirection)

• Problem description: the result is 938.0 !
• Expected output : 0.938
• Plateform: Linux / conda
• Versions:
  • Python 3.11.7
  • grib2io 2.1.2

This issue is leading to an error in the xarray backend as the dimension of longitude is not correct using the msg.grid() method.

[EricEngle-NOAA]

Thanks for reporting. I can confirm the issue. I have identified the issue and a solution is working. This also led me to discover a missing attribute for messages with Gaussian grids and is a fairly important one, the number of parallels.

v2.1.3 coming in the next few hours with this fix.

[EricEngle-NOAA]

v2.1.3 released with the fix for this. It is now available on PyPI and will be available on conda-forge in a few hours.