Mathias Hauser
Mathias Hauser
For me the first line (`fn(da_fake.compute()).compute()`) already throws the error. What numpy version are you using?
No worries. This might be a numpy bug. This is a pure numpy repro: ```python import numpy as np otype = "datetime64[ns]" arr = np.array(['2024-01-01', '2024-01-02', '2024-01-03'], dtype='datetime64[ns]') np.vectorize(lambda x:...
Thanks, this is certainly not desired behavior. Similar to #7027.
I tested the examples again and the problem does no longer appear. So it seems this was fixed in gdal, most likely in version 3.1. So I think we can...
That would be super cool! Would need to decide if the `flag_values` are kept as well. I assume you would add this in `da.cf.groupby(groups)`? A short example (using regionmask -...
This still leaves `as_3D` though
Thanks for pointing that out!
Just found out about pydata/xarray#7695, and with the xindexes `air.air.weighted(mask).mean(["lat", "lon"])` raises an error so currently this is unfortunately not usable :disappointed:
Thanks for your interest and opening the issue! My understanding is that there is no explicit assumption on the coordinate system* but that it has to be same for the...
I opened #530 for the plotting aspect. Would it address this part of your issue? > [regionmask.Regions.mask](https://regionmask.readthedocs.io/en/stable/generated/regionmask.Regions.mask.html) also takes several arguments related specifically to the lat-lon coordinate system Good point...