迭代没有唯一值的python数据帧

Question

I'm having trouble with rearranging a dataframe in python, made from a CSV file to how I need it. The data in the dataframe is as such:

ID      VOLUME      DATETIME

900     2.36        11/01/2015 13:40
900     2.30        11/01/2015 13:40
900     2.18        11/01/2015 13:41
900     2.30        11/01/2015 13:41
901     1.88        07/01/2015 17:01
901     1.80        07/01/2015 17:01
901     1.73        07/01/2015 17:02
901     1.80        07/01/2015 17:02

I have tried all sorts to pivot the above to how I need it but due to the fields having no real unique values I can not do it. I have been thinking I need to use iterrows to get it how I need it but haven't been able to figure it out. This is how I'm looking to get the data:

    900↓    901↓

    2.36    1.88
    2.30    1.80
    2.18    1.73
    2.30    1.80

I am trying to display one column per item in the ID column but I'm really starting to bang my head against the wall on this one. Can I create a new dataframe as above or am I going about this the wrong way?

Answer 1

Solution for the case when you have ID's with different # of rows:

In [34]: df
Out[34]:
    ID  VOLUME          DATETIME
0  900    2.36  11/01/2015 13:40
1  900    2.30  11/01/2015 13:40
2  900    2.18  11/01/2015 13:41
3  900    2.30  11/01/2015 13:41
4  901    1.88  07/01/2015 17:01
5  901    1.80  07/01/2015 17:01
6  901    1.73  07/01/2015 17:02
7  901    1.80  07/01/2015 17:02
8  901    1.11  07/01/2015 17:03   # NOTE: i've intentionally added this row

In [35]: pd.DataFrame({k : pd.Series(v)
                       for k, v in df.groupby('ID').VOLUME.apply(list).to_dict().items()})
Out[35]:
    900   901
0  2.36  1.88
1  2.30  1.80
2  2.18  1.73
3  2.30  1.80
4   NaN  1.11

OLD answer:

try this:

In [12]: pd.DataFrame(df.groupby('ID').VOLUME.apply(list).to_dict())
Out[12]:
    900   901
0  2.36  1.88
1  2.30  1.80
2  2.18  1.73
3  2.30  1.80

or:

In [18]: pd.DataFrame(df.groupby('ID').VOLUME.apply(lambda x: x.values).to_dict())
Out[18]:
    900   901
0  2.36  1.88
1  2.30  1.80
2  2.18  1.73
3  2.30  1.80

NOTE: this will work if you have the same amount of rows for all your ID 's