Errors on pycharm completion

[Hoeze]

Code Sample, a copy-pastable example if possible

# execute:
import numpy as np
import xarray as xr

sample_data = np.random.uniform(size=[2,2000,10000])
x = xr.Dataset({"sample_data": (sample_data.shape, sample_data)})

x2 = x["sample_data"]

# now type by hand:
x2.

Problem description

I'm not completely sure if it's a xarray problem, but each time I enter [some dataset]. (note the point) inside PyCharm's python console, I get a python exception instead of some autocompletion.

Expected Output

Output of xr.show_versions()

/usr/lib/python3.6/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`. from ._conv import register_converters as _register_converters

INSTALLED VERSIONS

commit: None
python: 3.6.5.final.0
python-bits: 64
OS: Linux
OS-release: 4.16.9-1-ARCH
machine: x86_64
processor:
byteorder: little
LC_ALL: None
LANG: de_DE.UTF-8
LOCALE: de_DE.UTF-8

xarray: 0.10.4
pandas: 0.22.0
numpy: 1.14.3
scipy: 1.0.1
netCDF4: None
h5netcdf: 0.5.1
h5py: 2.7.1
Nio: None
zarr: None
bottleneck: 1.2.1
cyordereddict: None
dask: None
distributed: None
matplotlib: 2.2.2
cartopy: None
seaborn: None
setuptools: 39.2.0
pip: 10.0.1
conda: None
pytest: None
IPython: 6.3.1
sphinx: None

[shoyer]

Thanks for the report. This is definitely a bug, calling dir() on an xarray object should never raise an error:

import numpy as np
import xarray as xr

sample_data = np.random.uniform(size=[2,2000,10000])
x = xr.Dataset({"sample_data": (sample_data.shape, sample_data)})
x2 = x["sample_data"]
dir(x2)

Results in:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-2-39db0ce27e1d> in <module>()
----> 1 dir(x2)

/usr/local/lib/python3.6/dist-packages/xarray/core/common.py in __dir__(self)
    195         """
    196         extra_attrs = [item
--> 197                        for sublist in self._attr_sources
    198                        for item in sublist
    199                        if isinstance(item, basestring)]

/usr/local/lib/python3.6/dist-packages/xarray/core/dataarray.py in _attr_sources(self)
    494     def _attr_sources(self):
    495         """List of places to look-up items for attribute-style access"""
--> 496         return self._item_sources + [self.attrs]
    497 
    498     @property

/usr/local/lib/python3.6/dist-packages/xarray/core/dataarray.py in _item_sources(self)
    499     def _item_sources(self):
    500         """List of places to look-up items for key-completion"""
--> 501         return [self.coords, {d: self[d] for d in self.dims},
    502                 LevelCoordinatesSource(self)]
    503 

/usr/local/lib/python3.6/dist-packages/xarray/core/dataarray.py in <dictcomp>(.0)
    499     def _item_sources(self):
    500         """List of places to look-up items for key-completion"""
--> 501         return [self.coords, {d: self[d] for d in self.dims},
    502                 LevelCoordinatesSource(self)]
    503 

/usr/local/lib/python3.6/dist-packages/xarray/core/dataarray.py in __getitem__(self, key)
    471         else:
    472             # xarray-style array indexing
--> 473             return self.isel(**self._item_key_to_dict(key))
    474 
    475     def __setitem__(self, key, value):

TypeError: isel() keywords must be strings

[shoyer]

The direct cause for this issue appears to be using integers as dimension names:

>>> x
<xarray.Dataset>
Dimensions:      (2: 2, 2000: 2000, 10000: 10000)
Dimensions without coordinates: 2, 2000, 10000
Data variables:
    sample_data  (2, 2000, 10000) float64 0.8346 0.7321 0.00948 0.04605 ...

We do try to support arbitrary hashable values as dimensions names for most operations, but you will have better luck in general using strings.