需要编写函数以在 pivot table 中找到总数的百分比

Question

例如这个table:

list_1=[['A',10,42,12,64],
        ['B',24,11,62,95],
       ['C',14,78,20,112]]
labels=['Class','Amount_1','Amount_2','Amount_3','Total']
df=pd.DataFrame(list_1,columns=labels)
df

Class   Amount_1    Amount_2    Amount_3    Total
0   A   10            42         12          64
1   B   24            11         62          95
2   C   14            78         20          112

我需要写函数来得到这个table（总额率）：


Class    Amount_1   Amount_2    Amount_3            
A       0.156250    0.656250    0.187500
B       0.252632    0.115789    0.652632
C       0.125000    0.696429    0.178571

Answer 1

IIUC

df.update(df.loc[:,df.columns.str.contains('Amount')].div(df.Total,0))
df
Out[41]: 
  Class  Amount_1  Amount_2  Amount_3  Total
0     A  0.156250  0.656250  0.187500     64
1     B  0.252632  0.115789  0.652632     95
2     C  0.125000  0.696429  0.178571    112

Answer 2

尝试：

from sklearn.preprocessing import normalize

df[["Amount_1", "Amount_2", "Amount_3"]]=normalize(df[["Amount_1", "Amount_2", "Amount_3"]], axis=1, norm="l1")

输出：

  Class  Amount_1  Amount_2  Amount_3  Total
0     A  0.156250  0.656250  0.187500     64
1     B  0.247423  0.113402  0.639175     95
2     C  0.125000  0.696429  0.178571    112

需要编写函数以在 pivot table 中找到总数的百分比

need writing function to find percentage from total in pivot table

python

analytics

numpy

pivot-table

pandas