熊猫总计数不同

如何：

>>> df         date  duration user_id0  2013-04-01        30    00011  2013-04-01        15    00012  2013-04-01        20    00023  2013-04-02        15    00024  2013-04-02        30    0002>>> df.groupby("date").agg({"duration": np.sum, "user_id": pd.Series.nunique}) duration  user_iddate   2013-04-01        65        22013-04-02        45        1>>> df.groupby("date").agg({"duration": np.sum, "user_id": lambda x: x.nunique()}) duration  user_iddate   2013-04-01        65        22013-04-02        45        1