Series/Dataframe to_csv inconsistency when path=None [0.14.1]

[ischwabacher]

It looks like Series.to_csv and DataFrame.to_csv differ in their handling of path=None:

In [1]: import pandas as pd

In [2]: print pd.Series([1, 1, 1], index=[1, 2, 3]).to_csv(None)
None

In [3]: print pd.DataFrame([1, 1, 1], index=[1, 2, 3]).to_csv(None)
,0
1,1
2,1
3,1

I haven't gone digging yet but I expect these two methods share enough code that this is a trivial fix.

[onesandzeroes]

The behaviour of DataFrame.to_csv() is explicitly mentioned in the docstring, so I think it's best to change the series method to reflect that. Putting together a PR now.

[jreback]

closed by #8220