Closed
Description
In [2]: df = pd.DataFrame(0, columns=[], index=pd.MultiIndex.from_product([[], []]))
In [3]: df.loc['b', '2'] = 1
In [4]: df.loc['a', '3'] = 1
In [5]: df.sort_index()
Out[5]:
2 3
b 1.0 NaN
a NaN 1.0
In [6]: df.sort_index().index.is_monotonic
Out[6]: False
Am I missing anything obvious?!
(Notice that this is not related to any of the two axes being initialized as empty)