Pandas merge: detecting where data came from

gistfile1.py

from pandas import DataFrame, merge

df = DataFrame(np.random.randn(10, 2), columns=["id", "sex"])
df2 = DataFrame(np.random.randn(10, 2), columns=["user_id", "name"])
df.id = range(10)
df2.user_id = range(3,13)

merge(df, df2, left_on="id", right_on="user_id",  how="outer")
"""
    id       sex  user_id      name
0    0 -0.254309      NaN       NaN
1    1 -0.363123      NaN       NaN
2    2 -0.408873      NaN       NaN
3    3 -1.209845        3  0.578440
4    4  0.952290        4 -1.336396
5    5 -0.091704        5  0.255794
6    6  0.984578        6 -0.469222
7    7 -0.694126        7  1.197256
8    8  0.369942        8 -0.656366
9    9  1.544090        9 -0.975548
10 NaN       NaN       10 -1.827958
11 NaN       NaN       11 -1.523407
12 NaN       NaN       12 -0.785032
"""

# since you know id came from the left and user_id came from the right,
# you know that if id is NaN then the data didn't exist in df and if
# user_id is NaN then the data didn't exist in df2