omb: accept dataframe

[gauteh]

@jerabaul29 opinions?

Use case:

#!/usr/bin/env python

import xarray as xr
import trajan as ta
import pandas as pd
import coloredlogs
import plz

coloredlogs.install('debug')

buoys = ['03', '10', '16', '17', '19', '20']

ds = [
    pd.read_csv(f'{b}.csv',
                names=['user', *ta.readers.omb.EXPECTED_COLUMNS, 'extra1'],
                header=0) for b in buoys
]
ds = pd.concat(ds)
ds = ta.readers.read_omb_csv(ds.drop('user', axis=1)).traj.assign_cf_attrs(
    creator_name='Gaute Hope',
    creator_email='[email protected]',
    title='ISMAR OMB deployment 2024',
    summary=' ',
    creator_institution='Norwegian Meteorological Institute')
ds.to_netcdf('2024-ismar-omb.nc')
print(ds)

[jerabaul29]

Looks good to me :) . I usually have a single csv that I read directly, but perfectly fine to me to support what you push if you use this workflow yourself :) .

[gauteh]

Yeah, my problem is that I need to filter out only some of the buoys from the CSVs.

[jerabaul29]

I think I would personally have "simply" done filtering of the CSV into a "reduced CSV" as a pre-processing step (typically I just take the CSV, and use my text editor to grep and delete the buoys entries I do not want, and then save it under a new name; maybe a bit "ugly", but works well enough ^^ ), and then parsed this "reduced CSV" as usual, but of course if you want to integrate this sort of functionality directly in trajan this is good to me :) .

[gauteh]

I need an automated setup. Now I grep when downloading (in a script).