fix test

[emanuel-schmid]

with pull request #63 one of the datasets got filtered out in TCTracks().read_ibtracs_netcdf(rovider='usa', year_range=(1993, 1994),basin='EP', estimate_missing=True):

Frozen(OrderedDict([('time_step', <xarray.Variable (time: 13)>
array([10800.00004045, 10800.        , 10800.        , 10800.        ,
       10800.        , 10800.        , 10800.        , 10800.        ,
       10800.        , 10800.        , 10800.        , 10800.        ,
       10800.        ])), ('radius_max_wind', <xarray.Variable (time: 13)>
array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.], dtype=float32)), ('radius_oci', <xarray.Variable (time: 13)>
array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.], dtype=float32)), ('max_sustained_wind', <xarray.Variable (time: 13)>
array([-1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.],
      dtype=float32)), ('central_pressure', <xarray.Variable (time: 13)>
array([-1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1., -1.],
      dtype=float32)), ('environmental_pressure', <xarray.Variable (time: 13)>
array([1005., 1005., 1005., 1005., 1005., 1005., 1005., 1005., 1005.,
       1005., 1005., 1010., 1010.])), ('time', <xarray.IndexVariable 'time' (time: 13)>
array(['1994-09-29T00:00:00.000000000', '1994-09-29T03:00:00.000000000',
       '1994-09-29T06:00:00.000000000', '1994-09-29T09:00:00.000000000',
       '1994-09-29T12:00:00.000000000', '1994-09-29T15:00:00.000000000',
       '1994-09-29T18:00:00.000000000', '1994-09-29T21:00:00.000000000',
       '1994-09-30T00:00:00.000000000', '1994-09-30T03:00:00.000000000',
       '1994-09-30T06:00:00.000000000', '1994-09-30T09:00:00.000000000',
       '1994-09-30T12:00:00.000000000'], dtype='datetime64[ns]')), ('lat', <xarray.Variable (time: 13)>
array([39.5     , 39.745296, 40.      , 40.275566, 40.5     , 40.498066,
       40.5     , 40.728046, 41.000004, 40.92378 , 41.      , 41.62489 ,
       42.5     ], dtype=float32)), ('lon', <xarray.Variable (time: 13)>
array([159.     , 160.34946, 162.     , 163.51527, 165.     , 166.505  ,
       168.     , 169.1625 , 171.     , 174.29718, 178.5    , 183.1068 ,
       188.     ], dtype=float32))]))

Accept this pull request if it makes sense to omit that dataset.
If the dataset should not be filtered out - we have an issue.

[bguillod]

@emanuel-schmid do you have the sid or id_no for that track?

[bguillod]

I've found it it's 1994272N40159 (PAT), which doesn't contain any data from provider USA, not even lon/lat. Only provider 'tokyo' gives data for that one. In my view it's correct to filter that out if provider='usa' but I leave it to @tovogt to confirm.
BTW as a side note, reading lat/lon from the provider would have filtered out this track earlier on (see also issue [64]).

[emanuel-schmid]

@bguillod Thanks a lot for checking!