slice in xarray gives error 'float' object cannot be interpreted as an integer

Question

I am trying to slice data by longtitude using xarray. The data is in a netcdf file I created from measurements I made.

The xarray.Dataset has the following attributes:

Dimensions:

(lat: 1321, lon: 1321)

Data variables:

(lon) float64 '8.413 8.411 8.409 ... 4.904 4.905'
(lat) float64 '47.4 47.4 47.41 ... 52.37 52.37'
(data) float64 ... #dimension: 1321

my code is:

import xarray as xr
obs = xr.open_dataset('data.nc')
obs=obs['data'].sel(lon=slice(4.905, 8.413))

The error I get is TypeError: 'float' object cannot be interpreted as an integer

I could not find out whether it is an error in my code, or an error in xarray. I would expect such an error using isel instead of sel. Could not find any solution on here or over at the xarray documentation.

Full error message:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-434-5b37e4c5d0c6> in <module>
----> 1 obs=obs['data'].sel(lon=slice(4.905, 8.413))

~/opt/anaconda3/lib/python3.7/site-packages/xarray/core/dataarray.py in sel(self, indexers, method, tolerance, drop, **indexers_kwargs)
   1059             method=method,
   1060             tolerance=tolerance,
-> 1061             **indexers_kwargs,
   1062         )
   1063         return self._from_temp_dataset(ds)

~/opt/anaconda3/lib/python3.7/site-packages/xarray/core/dataset.py in sel(self, indexers, method, tolerance, drop, **indexers_kwargs)
   2066             self, indexers=indexers, method=method, tolerance=tolerance
   2067         )
-> 2068         result = self.isel(indexers=pos_indexers, drop=drop)
   2069         return result._overwrite_indexes(new_indexes)
   2070 

~/opt/anaconda3/lib/python3.7/site-packages/xarray/core/dataset.py in isel(self, indexers, drop, **indexers_kwargs)
   1933             var_indexers = {k: v for k, v in indexers.items() if k in var_value.dims}
   1934             if var_indexers:
-> 1935                 var_value = var_value.isel(var_indexers)
   1936                 if drop and var_value.ndim == 0 and var_name in coord_names:
   1937                     coord_names.remove(var_name)

~/opt/anaconda3/lib/python3.7/site-packages/xarray/core/variable.py in isel(self, indexers, **indexers_kwargs)
   1058 
   1059         key = tuple(indexers.get(dim, slice(None)) for dim in self.dims)
-> 1060         return self[key]
   1061 
   1062     def squeeze(self, dim=None):

~/opt/anaconda3/lib/python3.7/site-packages/xarray/core/variable.py in __getitem__(self, key)
    701         array `x.values` directly.
    702         """
--> 703         dims, indexer, new_order = self._broadcast_indexes(key)
    704         data = as_indexable(self._data)[indexer]
    705         if new_order:

~/opt/anaconda3/lib/python3.7/site-packages/xarray/core/variable.py in _broadcast_indexes(self, key)
    540 
    541         if all(isinstance(k, BASIC_INDEXING_TYPES) for k in key):
--> 542             return self._broadcast_indexes_basic(key)
    543 
    544         self._validate_indexers(key)

~/opt/anaconda3/lib/python3.7/site-packages/xarray/core/variable.py in _broadcast_indexes_basic(self, key)
    568             dim for k, dim in zip(key, self.dims) if not isinstance(k, integer_types)
    569         )
--> 570         return dims, BasicIndexer(key), None
    571 
    572     def _validate_indexers(self, key):

~/opt/anaconda3/lib/python3.7/site-packages/xarray/core/indexing.py in __init__(self, key)
    369                 k = int(k)
    370             elif isinstance(k, slice):
--> 371                 k = as_integer_slice(k)
    372             else:
    373                 raise TypeError(

~/opt/anaconda3/lib/python3.7/site-packages/xarray/core/indexing.py in as_integer_slice(value)
    344 
    345 def as_integer_slice(value):
--> 346     start = as_integer_or_none(value.start)
    347     stop = as_integer_or_none(value.stop)
    348     step = as_integer_or_none(value.step)

~/opt/anaconda3/lib/python3.7/site-packages/xarray/core/indexing.py in as_integer_or_none(value)
    340 
    341 def as_integer_or_none(value):
--> 342     return None if value is None else operator.index(value)
    343 
    344

I want to select the entire data, because eventually I want to subtract the entire array from a bigger data base with a wider grid. This bigger data base is a NETCDF file as well. And for that one, I managed to slice the longitude with the exact same code I am trying on this smaller data set where I get the error. The only difference is, that the bigger NETCDF uses a float32 format. I don't suspect this could cause the error.

Any help is appreciated. Thank you.

Can you show the full repr of your dataset? (print(obs) should do it). — jhamman
– jhamman, Commented Jun 8, 2020 at 3:26

pwi · Accepted Answer · 2020-06-10 16:32:25Z

I think I found the problem. When I created the netcdf file for the observation, I made a mistake in the createDimension part, when I named the lon and lat data. Because of this, lat and lon showed up under 'Data variables' in the netcdf file, where they should show up under 'Coordinates'

wrong was something like:

#Specifying dimensions#
f.createDimension('longitude', len(lon_list))
f.createDimension('latitude', len(lat_list))
#Building variables
longitude = f.createVariable('lon', float, ('lon',), zlib=True)
latitude = f.createVariable('lat', float, ('lat',), zlib=True)  

data = f.createVariable('data', float, ('lat','lon'), zlib=True)

correct was:

#Specifying dimensions#
f.createDimension('lon', len(lon_list))
f.createDimension('lat', len(lat_list))
#Building variables
longitude = f.createVariable('lon', float, ('lon',), zlib=True)
latitude = f.createVariable('lat', float, ('lat',), zlib=True)  

data = f.createVariable('data', float, ('lat','lon'), zlib=True)

liakoyras · Accepted Answer · 2022-11-17 21:01:18Z

1

This is a little late but I just ran into a similar issue where I got a similar undecipherable error when trying to slice by a variable.

I think the problem is that if you are trying to slice by a variable that isn't a coordinate you get an error that isn't very informative.

data = data.assign_coords({"lat":data.lat,"lon":data.lon})

would have fixed this without rewriting the netcdf file.

edited Nov 17, 2022 at 21:01

liakoyras

1,20514 silver badges28 bronze badges

answered Nov 13, 2022 at 19:46

Axel

716 bronze badges

Collectives™ on Stack Overflow

slice in xarray gives error 'float' object cannot be interpreted as an integer

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related