Behavior change in indexing using ix in 0.17.0

[ankurankan]

I am getting an error when indexing by comparing tuples:

In [13]: df = pd.DataFrame([[('A', 1), ('B', 0)], [('A', 0), ('B', 1)], [('A', 1), ('B', 0)]], columns=['A', 'B'])

In [14]: df
Out[14]: 
        A       B
0  (A, 1)  (B, 0)
1  (A, 0)  (B, 1)
2  (A, 1)  (B, 0)

In [15]: df[df.A == ('A', 1)]
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-15-2663526e3e28> in <module>()
----> 1 df[df.A == ('A', 1)]

/home/ankur/anaconda3/envs/py3/lib/python3.4/site-packages/pandas/core/ops.py in wrapper(self, other, axis)
    724     op_desc = _op_descriptions[op_name]
    725     if op_desc['reversed']:
--> 726         equiv = 'other ' + op_desc['op'] + ' series'
    727     else:
    728         equiv = 'series ' + op_desc['op'] + ' other'

/home/ankur/anaconda3/envs/py3/lib/python3.4/site-packages/pandas/core/ops.py in na_op(x, y)
    642             if isinstance(y, list):
    643                 y = lib.list_to_object_array(y)
--> 644 
    645             if isinstance(y, (np.ndarray, pd.Series)):
    646                 if (x.dtype == np.bool_ and

pandas/lib.pyx in pandas.lib.vec_compare (pandas/lib.c:14361)()

ValueError: Arrays were different lengths: 3 vs 2

But this used to work in 0.16.2, so not sure if this is expected or a bug.

[jreback]

this is a dupe of: #11339

and fixed for 0.17.1

note that storing tuples inside a column is quite inefficient, and not recommended.