Brian Blaylock
Brian Blaylock
I looks to me that this policy change is only affecting ECMWF forecasts right now. Let me know if anyone sees any others.
Created a pull request with a quick and hacky fix. Needs some testing to find any issues and edge cases.
Closed by #493
Thanks for tracking this down. I don't use colab, so haven't ever tested there. Should an issue be opened with cfgrib?
yeah, this crash must just be in colab?? I can't replicate with a fresh install on linux (WSL). (I actually haven't ever used colab.) Create new env ```bash mamba env...
Thanks for the suggestion. Herbie is mainly built to get NWP data in GRIB/GRIB2 format. It would need to be refactored to access other formats like NetCDF. I'll keep this...
I see what the problem is. NAM uses submessages to pack the U and V wind together. ```python from herbie import Herbie H = Herbie("2022-12-16 12:00", model="nam", product="awphys", fxx=0) H.inventory("GRD:10...
@blaylockbk Note to self...add a note on this feature in the NAM tutorial notebook in the docs.
Can you provide a reproducible example that gives this error?
Oh, nice. Glad it's being worked on already. Didn't know about @DanielO efforts on this.