df.squeeze() doesn't support axis parameter

[kernc]

Code Sample, a copy-pastable example if possible

>>> df = pd.DataFrame([1], columns=['A'])

>>> df 
   A
0  1

>>> df.squeeze()
1

>>> type(df.squeeze())
numpy.int64

>>> df.squeeze(axis=1)  # Only want to squeeze a single dimension, i.e. into a series
...
ValueError: the 'axis' parameter is not supported in the pandas implementation of squeeze()

Problem description

np.squeeze supports axis parameter and this comment in the source implies it should eventually be implemented.

Expected Output

>>> df.squeeze(axis=1)  # Squeeze a single dimension, i.e. into a series
0    1
Name: A, dtype: int64

>>> df.squeeze(axis=0)
A    1
Name: 0, dtype: int64

Output of pd.show_versions()

pandas: 0.19.0+416.ge1390cd

[jreback]

@kernc how is this actually useful though? numpy needs this because of n-dim shrinking to a lower dim. pandas generally has 2-dim max so this is not so important.

[kernc]

Indeed, as the example shows, it is only useful to squeeze a 1×1 frame into a series instead of a scalar, or to keep 1×N frame a frame. 😃