Fix a test caused by a bug introduced in pandas 1.1.3.

[ueshin]

There is a bug introduced in pandas 1.1.3.

>>> pdf = pd.DataFrame({("x", "a"): [1, 2, 6, 4, 4, 6, 4, 3, 7], ("x", "b"): [4, 2, 7, 3, 3, 1, 1, 1, 2], ("y", "c"): [4, 2, 7, 3, None, 1, 1, 1, 2], ("z", "d"): list("abcdefght")}, index=[0, 1, 3, 5, 6, 8, 9, 9, 9])
>>> pdf
   x       y  z
   a  b    c  d
0  1  4  4.0  a
1  2  2  2.0  b
3  6  7  7.0  c
5  4  3  3.0  d
6  4  3  NaN  e
8  6  1  1.0  f
9  4  1  1.0  g
9  3  1  1.0  h
9  7  2  2.0  t
>>> pdf[("x", "a")].groupby(pdf[("x", "b")]).sum().sort_index()
Traceback (most recent call last):

...

TypeError: only integer scalar arrays can be converted to a scalar index

[ueshin]

I'd merge this now to unblock other PRs.