Use xarray dataset as internal data container#262
Conversation
…nd improving parsing of input
|
Much more to be done, but maybe good to merge soon to ease collaboration and minimize merge conflicts |
| return data | ||
|
|
||
|
|
||
| def _model2obs_interp( |
There was a problem hiding this comment.
The name of this function doesn't reveal that it also remove gaps, also not the docstring. Should gap removal be part of this function.
|
I tried to modify a few lines like this ds[name].attrs["quantity"] = quantity.to_dict()to ds[name].attrs["long_name"] = quantity.name
ds[name].attrs["units"] = quantity.unitTo be able to persist the dataset as NetCDF and be compatible with CF convention making easier to plot the data with xarray, but somehow managed to change more than necessary, making a lot of tests to fail. So I deleted those changes and only committed the tests that saves the data as NetCDF. |
|
x, y in track data are now coordinates and not data variables. |
I was wondering about this. Thanks |



Changing the internal data structure of observations from pd.DataFrame to xr.Dataset.
Some considerations:
Objectives of this PR: