projecting PC dataset to ASF grid w/ different time dims
e-marshall opened this issue · 3 comments
In dataset_comparison.ipynb
handling different spatial resolutions section:
fix issue that arises when trying to mask out the NaN values in the ASF dataset in the PC dataset
code looks like:
pc_mask = xr.where(asf_short.notnull(), pc_downsample,np.nan)
end up with a 4-d array with (x,y, acq_date, time) <-- acq date is the time dim of asf_subset
figure out best way to organize dims
Put some steps I've been trying in a notebook here, didn't have time to write out more but will update it and keep working on it later, just thought it was easier to work in this notebook than in the middle of the other one
Right you're running into automatic broadcasting. This is a key concept when working with multip[el arrays.
You'll want to rename acq_date to time so that it's consistent.
Thanks! I had tried that but wasn't sure if I was on the right track/ hadn't realized that the time of the time and acq_date dims didn't match exactly so had problems. Think its working now!
closed on daa81a7