This is a small change to ignore the precip type plot at analysis. It is needed because the wgrib2 command (that converts the wind components to Earth-relative) removes the categorical precip fields at analysis, presumably because they are all zeroes. This causes pygraf to crash. This change corrects that. One caveat is that the --all_leads flag needs to be removed, since ignoring the precip type fields causes a different error at forecast hour 1.
Sorry, it occurred to me that this is not model-specific, and may cause problems with models other than HRRR_HI.
I'm putting in a hotfix for HRRR_HI, and closing this PR.
This is a small change to ignore the precip type plot at analysis. It is needed because the wgrib2 command (that converts the wind components to Earth-relative) removes the categorical precip fields at analysis, presumably because they are all zeroes. This causes pygraf to crash. This change corrects that. One caveat is that the --all_leads flag needs to be removed, since ignoring the precip type fields causes a different error at forecast hour 1.