apply method of DataFrameGroupBy object returns surprising result

[sadruddin]

In the example below, column "name" of groupby apply method has the same content as column "year". A similar column with a copy of the "name" values will remain untouched (as expected).

import pandas
import StringIO

csv = """"year","name","percent","sex"
1880,"John",0.081541,"boy"
1880,"William",0.080511,"boy"
1880,"James",0.050057,"boy"
1881,"Charles",0.045167,"boy"
1881,"George",0.043292,"boy"
"""

names = pandas.read_csv(StringIO.StringIO(csv))
names['name2'] = names['name']
grouped = names.groupby('year')

print 'Pandas version', pandas.__version__
print names
print grouped.apply(lambda x:x) 

OUTPUT:

Pandas version 0.7.2
   year     name   percent  sex    name2
0  1880     John  0.081541  boy     John
1  1880  William  0.080511  boy  William
2  1880    James  0.050057  boy    James
3  1881  Charles  0.045167  boy  Charles
4  1881   George  0.043292  boy   George
   year  name   percent  sex    name2
0  1880  1880  0.081541  boy     John
1  1880  1880  0.080511  boy  William
2  1880  1880  0.050057  boy    James
3  1881  1881  0.045167  boy  Charles
4  1881  1881  0.043292  boy   George

[wesm]

This has been fixed since 0.7.2, pushed a unit test. Will try to get the next release out in the next couple weeks