CommonDataModel 0.3.6 changes how they handle 0-dimensional input - see release notes here.
This causes an error in RRTMGP when calling FT(Array(...)) and the inner array has dimension 0, here. This was discovered downstream in ClimaCoupler, e.g. in this build. This causes the test/all_sky.jl test to fail when using NCDatasets v0.14.
CommonDataModel 0.3.6 changes how they handle 0-dimensional input - see release notes here.
This causes an error in RRTMGP when calling
FT(Array(...))
and the inner array has dimension 0, here. This was discovered downstream in ClimaCoupler, e.g. in this build. This causes thetest/all_sky.jl
test to fail when using NCDatasets v0.14.