tidy netcdf with xarray about drb-gw-hw-model-prep HOT 4 CLOSED

def subset_nc_to_comids(nc_file, comids):
    comids = [int(c) for c in comids]

    ds = xr.open_dataset(nc_file, decode_times = True)

    # filter out comids that are not in climate drivers (should only be 4781767)
    comids = np.array(comids)
    comids_in_climate = comids[np.isin(comids, ds.COMID.values)]
    comids_not_in_climate = comids[~np.isin(comids, ds.COMID.values)]
    print(comids_not_in_climate)

    # We know of one COMID that has no catchment and so should be included
    # in `comids_not_in_climate` if passed through in `comids`. Use assert 
    # statement to make sure we are aware of any others. COMIDs within 
    # `comids_not_in_climate` will not have matched climate data. 
    if len(comids_not_in_climate) > 0 :
        assert list(comids_not_in_climate) == [4781767]
    ds_comids = ds.sel(COMID=comids_in_climate)
    # [Lauren] we have been using a function written by Jeff Sadler for the DRB
    # PGDL-DO project to process the xarray object to a ~tiday data frame. Below
    # I've replaced ds_to_dataframe_faster(ds_comids) with a more generic function
    # to speed up the run time. See this issue for further details: 
    # https://github.com/USGS-R/drb-gw-hw-model-prep/issues/44.
    ds_comids_df = ds_comids.to_dataframe().reset_index()
    return ds_comids_df

> tar_meta() %>% filter(name == "p2_met_data_nhd_mainstem_reaches") %>% pull(seconds)/60
[1] 33.42533

> tar_meta() %>% filter(name == "p2_met_data_nhd_mainstem_reaches") %>% pull(seconds)/60
[1] 29.52967
>