在两个单独列上的熊猫数据框中进行聚合

Question

I am trying to do aggregation on fields cat1, cat2, cat3 on the following DataFrame. I need to count the number of trials and the number of unique subjects in each group. The code below does find the number of trials correct but the number of subject is not correct.

mydata = pd.DataFrame(np.array([
         ['Adam', 1L, 1L, 1L, 1L],
         ['Adam', 2L, 1L, 2L, 1L],
         ['Adam', 3L, 2L, 2L, 3L],
         ['Adam', 1L, 1L, 1L, 1L],
         ['Adam', 2L, 1L, 1L, 2L],
         ['Adam', 3L, 1L, 2L, 1L],
         ['Bob', 1L, 1L, 2L, 3L],
         ['Bob', 2L, 1L, 2L, 3L],
         ['Bob', 3L, 1L, 1L, 1L],
         ['Bob', 1L, 1L, 2L, 3L],
         ['Bob', 2L, 2L, 2L, 3L],
         ['Bob', 3L, 1L, 3L, 1L]], dtype=object),
         columns = ['ID','trial','cat1','cat2','cat3']
        )

grouped = mydata.groupby(['cat1', 'cat2', 'cat3']).agg(['count'])

grouped.reset_index()

Result:

   cat1  cat2  cat3     ID  trial
                     count  count
0     1     1     1      3      3
1     1     1     2      1      1
2     1     2     1      2      2
3     1     2     3      3      3
4     1     3     1      1      1
5     2     2     3      2      2

The result that I am expecting is :

   cat1  cat2  cat3  trial  ID
0     1     1     1      3   2
1     1     1     2      1   1
2     1     2     1      2   1
3     1     2     3      3   1
4     1     3     1      1   1
5     2     2     3      2   2

Answer 1

You could aggregate on ID with pd.Series.nunique and get count from trail

In [215]: (mydata.groupby(['cat1', 'cat2', 'cat3'])
                .agg({'ID': pd.Series.nunique, 'trial': 'count'})
                .reset_index())
Out[215]:
   cat1  cat2  cat3  trial  ID
0     1     1     1      3   2
1     1     1     2      1   1
2     1     2     1      2   1
3     1     2     3      3   1
4     1     3     1      1   1
5     2     2     3      2   2