groupby.transform failing in 0.13

[MichaelWS]

I have used groupby.transform to compute a number of statistics on data that is grouped by date. It seems to fail on indexing

All of my tests are failing for 0.13 but work in 0.12.

Here is one example:
the following three lines work fine in 0.12 and fails in 0.13.
print cik_df.head()
cik_df["portfolio_rank"] = cik_df.groupby("ex_date")["value"].transform(lambda x: x.rank(ascending=False))
print cik_df.head()

Here is the output: https://gist.github.com/MichaelWS/7991767

(I sent this to the pydata group but was asked to post it here by a more active contributor)

[MichaelWS]

Thanks

[jreback]

@MichaelWS ok...in master...pls give a test if you would

[MichaelWS]

Thank you for the quick fix. all my tests work.

[jreback]

gr8!

posting an issue here is likely to get more attention IMHO than the mailing list (because it sort of sticks around)...and IMHO conversation are MUCH easier (as is formatting)

my 2c

[MichaelWS]

Yes, I definitely agree and appreciate it.
Thank you