What happened: I was combining two Datasets using combine_first and to my surprise the dtype of one of the DataArrays in the merged Dataset was changed (from bool to float64).

What you expected to happen: No change in dtype.

Minimal Complete Verifiable Example:

```python

import xarray as xr  ds = xr.Dataset(coords={'abc': list('abc')})  ds['x'] = ('abc', [1., 2., 3.]) ds['y'] = ('abc', [-1., -2., -3.])  ds['t'] = ('abc', [True, False, True]) ds <xarray.Dataset> Dimensions: (abc: 3) Coordinates: * abc (abc) <U1 'a' 'b' 'c' Data variables: x (abc) float64 1.0 2.0 3.0 y (abc) float64 -1.0 -2.0 -3.0 t (abc) bool True False True  xy4b = ds[['x', 'y']].sel(abc=~ds.t) * 10 xy4b.combine_first(ds) Out[14]: <xarray.Dataset> Dimensions: (abc: 3) Coordinates: * abc (abc) object 'a' 'b' 'c' Data variables: x (abc) float64 1.0 20.0 3.0 y (abc) float64 -1.0 -20.0 -3.0 t (abc) float64 1.0 0.0 1.0 ```

Anything else we need to know?: No.

Environment:

Code Sample

```python

da = xr.DataArray([0., 1., 2.]) db = da.stack(z=da.dims) db <xarray.DataArray (z: 3)> array([0., 1., 2.]) Coordinates: * z (z) int64 0 1 2 db.unstack('z') Traceback (most recent call last):

File "<ipython-input-127-ec66ebede9c5>", line 1, in <module> db.unstack('z')

File "/usr/lib64/python3.6/site-packages/xarray/core/dataarray.py", line 1248, in unstack ds = self._to_temp_dataset().unstack(dim)

File "/usr/lib64/python3.6/site-packages/xarray/core/dataset.py", line 2315, in unstack raise ValueError('cannot unstack a dimension that does not have '

ValueError: cannot unstack a dimension that does not have a MultiIndex ```

Problem description

stack is not consistently creating a MultiIndex, but just renames a dimension in case there is just the single dimension to be stacked. This breaks code where the (number of) dimensions are (is) not known in advance.

Expected Output

stack should create a MultiIndex in all cases and a stack/unstack sequence should always return the original DataArray.

Output of xr.show_versions()

Code Sample

```python

import xarray as xr

test = xr.Dataset(coords={'xy': ['x', 'y']})

test <xarray.Dataset> Dimensions: (xy: 2) Coordinates: * xy (xy) <U1 'x' 'y' # NOTE '<U1' dtype Data variables: empty

test['test'] = xr.DataArray(np.array([0, 0]), dims=['xy'])

test <xarray.Dataset> Dimensions: (xy: 2) Coordinates: * xy (xy) object 'x' 'y' # NOTE 'object' dtype Data variables: test (xy) int64 0 0 ```

Problem description

The coordinate dtype changes from <U1 to object.

Expected Output

The coordinate dtype should not change.

Output of xr.show_versions()

Context

Consider a netCDF4 file with a group structure. For example, the following toy: ```python import netCDF4 as nc

netCDF4 file

f = nc.Dataset('simple_hierarchy.nc', 'w')

coordinates in root

f.createDimension('x', 3) f.createVariable('x', 'f4', ('x',), fill_value=False) f['x'][:] = [1.1, 2.2, 3.3] f.createDimension('y', 2) f.createVariable('y', 'f4', ('y',), fill_value=False) f['y'][:] = [-0.9, -1.8]

variables in root

f.createVariable('u', 'i1', (), fill_value=False) f.createVariable('v', 'u1', ('x','y'), fill_value=False)

group

f.createGroup('g') g = f['g']

new/modified coordinates in g

g.createDimension('y', 3) g.createVariable('y', 'f4', ('y',), fill_value=False) g['y'][:] = [-0.9, -1.8, -2.7]

variable in g

g.createVariable('w', 'u1', ('x', 'y'), fill_value=False) f.close() ```

Current behavior

1. It is currently a hassle to get a DataArray from variable in a group with multiple non-coordinate variables: ```python >>> xr.open_dataarray('simple_hierarchy.nc') … ValueError: Given file dataset contains more than one data variable. Please read with xarray.open_dataset and then select the variable you want. >>> xr.open_dataarray('simple_hierarchy.nc', group='v') xr.open_dataarray('simple_hierarchy.nc', group='v') … OSError: [Errno group not found: v] 'v' >>> xr.open_dataarray('simple_hierarchy.nc', drop_variables='u') <xarray.DataArray 'v' (x: 3, y: 2)> array([[120, 219], [178, 172], [ 9, 127]], dtype=uint8) Coordinates:

   • x (x) float32 1.1 2.2 3.3
   • y (y) float32 -0.9 -1.8 ```

2. Also, coordinates defined at a group level closer tot the root are not taken into account: ```python >>> xr.open_dataarray('simple_hierarchy.nc', group='g') <xarray.DataArray 'w' (x: 3, y: 3)> array([[216, 219, 178], [172, 9, 127], [ 0, 0, 64]], dtype=uint8) Coordinates:

   • y (y) float32 -0.9 -1.8 -2.7 Dimensions without coordinates: x ``` So the DataArray is not loaded correctly, as part of its defining coordinates are missing.

Suggested behavior

1. Add a variable kwarg in the open_dataarray method: ```python >>> xr.open_dataarray('simple_hierarchy.nc', variable='v') <xarray.DataArray 'v' (x: 3, y: 2)> array([[120, 219], [178, 172], [ 9, 127]], dtype=uint8) Coordinates:

   • x (x) float32 1.1 2.2 3.3
   • y (y) float32 -0.9 -1.8 ```

2. Have the function that loads variables go up the group hierarchy to see if some coordinate arrays can be found for dimensions lacking them within this group: ```python >>> xr.open_dataarray('simple_hierarchy.nc', group='g') <xarray.DataArray 'w' (x: 3, y: 3)> array([[216, 219, 178], [172, 9, 127], [ 0, 0, 64]], dtype=uint8) Coordinates:

   • x (x) float32 1.1 2.2 3.3
   • y (y) float32 -0.9 -1.8 -2.7 ``` I guess care needs to be taken as well upon writing to netCDF, to make sure no spurious dimension/coordinate definitions are added.

Version

xarray 0.9.6

MCVE Code Sample

```python

da = xr.DataArray([1., 2.], coords=[('dim', [.5, 1.])]) da.coords['nondim'] = ('dim', [0., 1.]) da <xarray.DataArray (dim: 2)> array([1., 2.]) Coordinates: * dim (dim) float64 0.5 1.0 nondim (dim) float64 0.0 1.0 da.interp(dim=.75) <xarray.DataArray ()> array(1.5) Coordinates: nondim float64 0.5 dim float64 0.75 da.interp(nondim=.5) Traceback (most recent call last):

File "<ipython-input-192-e3df34cff90f>", line 1, in <module> da.interp(nondim=.5)

File "/usr/lib64/python3.6/site-packages/xarray/core/dataarray.py", line 951, in interp **coords_kwargs)

File "/usr/lib64/python3.6/site-packages/xarray/core/dataset.py", line 1860, in interp indexers = OrderedDict(self._validate_indexers(coords))

File "/usr/lib64/python3.6/site-packages/xarray/core/dataset.py", line 1316, in _validate_indexers raise ValueError("dimensions %r do not exist" % invalid)

ValueError: dimensions ['nondim'] do not exist ```

Expected Output

The same interpolation as for 'dim'.

Problem Description

Apparently, xarray currently cannot interpolate using non-dimension coordinates.

Output of xr.show_versions()