DataArray.to_series() creates a MultiIndex even for 1D data

[shoyer]

xref #1542

This appears when using the development version of pandas:

In [7]: import pandas as pd

In [8]: pd.Series(range(5))
Out[8]:
0    0
1    1
2    2
3    3
4    4
dtype: int64

In [9]: pd.Series(range(5)).to_xarray().to_series()
Out[9]:
index
0        0
1        1
2        2
3        3
4        4
dtype: int64

In [10]: pd.Series(range(5)).to_xarray().to_series().index
Out[10]:
MultiIndex(levels=[[0, 1, 2, 3, 4]],
           labels=[[0, 1, 2, 3, 4]],
           names=['index'])

I think the problem boils down to these lines:

xarray/xarray/core/coordinates.py

Lines 52 to 58 in 78ca20a

	def to_index(self, ordered_dims=None):
	"""Convert all index coordinates into a :py:class:`pandas.MultiIndex`
	"""
	if ordered_dims is None:
	ordered_dims = self.dims
	indexes = [self._data.get_index(k) for k in ordered_dims]
	return pd.MultiIndex.from_product(indexes, names=list(ordered_dims))

We really want a normal Index for the coordinates from 1D data, not a MultiIndex. MultiIndex.from_product used to return different types depending on how it was called, but in pandas-dev/pandas#17236 it was fixed to always return a MultiIndex.