You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
While playing with ECMWF files, such as the ones you can get here, converted to netcdf (after some dimension renaming), I discovered that xcube serve does not manage to open the cube if it has a global coverage with a 90 to -90 latitude range.
Simply truncating it a bit (eg. x = x.sel(lat=slice(89, -89))) is a workaround.
The issue here is not that I must crop the dataset at the poles, but that xserve fails silently with it.
Additional context
xcube v1.2.0
The text was updated successfully, but these errors were encountered:
@mdespriee thanks for reporting! There is another workaround which does not include modifying the dataset: When using a configuration file for the server, you can pass a parameter specifying the boundingbox:
BoundingBox: [-180, -90, 180, 90]
Describe the bug
While playing with ECMWF files, such as the ones you can get here, converted to netcdf (after some dimension renaming), I discovered that
xcube serve
does not manage to open the cube if it has a global coverage with a 90 to -90 latitude range.Simply truncating it a bit (eg.
x = x.sel(lat=slice(89, -89))
) is a workaround.The issue here is not that I must crop the dataset at the poles, but that xserve fails silently with it.
Additional context
xcube v1.2.0
The text was updated successfully, but these errors were encountered: