如何根据熊猫数据框中的月份计算不同列的 QTD 值？

Question

我想根据熊猫数据框中的月份计算不同列的 QTD 值。

代码：

data = {'month': ['April', 'May', 'June', 'July', 'August', 'September', 'October', 'November', 'December', 'January', 'February', 'March'],
    'kpi': ['sales', 'sales quantity', 'sales', 'sales', 'sales', 'sales', 'sales', 'sales quantity', 'sales', 'sales', 'sales', 'sales'],
    're': [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12],
    're3+9': [10, 20, 30, 40, 50, 60, 70, 80, 90, 10, 10, 20],
    're6+6': [5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60],
    're9+3': [2, 4, 6, 8, 10, 12, 14, 16, 20, 10, 10, 20],
    're_o' : [1, 1, 1, 11, 11, 11, 12, 12, 12, 13, 13, 13]
    }

# Create DataFrame
df = pd.DataFrame(data)
g = pd.to_datetime(df['month'], format='%B').dt.to_period('Q')

if (df['month'].isin(['April', 'May', 'June'])):
        df['Q-Total'] = df.groupby([g,'kpi'])['re'].cumsum()
elif (df['month'].isin(['July', 'August', 'September'])):
        df['Q-Total'] = df.groupby([g, 'kpi'])['re3+9'].cumsum()
elif (df['month'].isin(['October', 'November', 'December'])):
        df['Q-Total'] = df.groupby([g, 'kpi'])['re6+6'].cumsum()
elif (df['month'].isin(['January', 'February', 'March'])):
        df['Q-Total'] = df.groupby([g, 'kpi'])['re9+3'].cumsum()
else:
        print("zero")

我需要的输出如下：

       month             kpi   re re3+9 re6+6 re9+3  re_o  Q-Total
0       April           sales   1   10    5     2      1        1
1         May  sales quantity   2   20   10     4      1        2
2        June           sales   3   30   15     6      1        4
3        July           sales   4   40   20     8     11       40
4      August           sales   5   50   25    10     11       90
5   September           sales   6   60   30    12     11      150
6     October           sales   7   70   35    14     12       35
7    November  sales quantity   8   80   40    16     12       40
8    December           sales   9   90   45    20     12       80
9     January           sales  10   10   50    10     13       10
10   February           sales  11   10   55    10     13       20
11      March           sales  12   20   60    20     13       40

这里有四列名为 re,re3+9,re6+6,re9+3 用于获取累积和值。我想根据以下条件计算累积和：

1. 如果月份是 4 月、5 月和 6 月，则仅从列 re 中获取累积总和
2. 如果月份是 7 月、8 月和 9 月，则仅从 re3+9 中取累计和
3. 如果月份是 10 月、11 月和 12 月，则仅从 re6+6 中取累计和
4. 如果月份是一月、二月和三月，则只取re9+3的累计和

   但是当我运行代码时出现如下错误：

   Traceback (most recent call last):
   File "/home/a/p/s.py", line 54, in <module>
   if (df['month'].isin(['April', 'May', 'June'])):
   File "/home/a/anaconda3/envs/p/lib/python3.9/site-packages/pandas/core/generic.py", line 1527, in __nonzero__
   raise ValueError(
   ValueError: The truth value of a Series is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all().
   

   任何人都可以提出解决此问题的解决方案吗？

Answer 1

利用：

g = pd.to_datetime(df['month'], format='%B').dt.to_period('Q')

m1 = df['month'].isin(['April', 'May', 'June'])
m2 = df['month'].isin(['July', 'August', 'September'])
m3 = df['month'].isin(['October', 'November', 'December'])
m4 = df['month'].isin(['January', 'February', 'March'])


df['Q-Total'] = np.select([m1, m2, m3, m4], 
                          [df['re'], df['re3+9'], df['re6+6'], df['re9+3']], default=0)

df['Q-Total'] = df.groupby([g,'kpi'])['Q-Total'].cumsum()
print (df)
        month             kpi  re  re3+9  re6+6  re9+3  re_o  Q-Total
0       April           sales   1     10      5      2     1        1
1         May  sales quantity   2     20     10      4     1        2
2        June           sales   3     30     15      6     1        4
3        July           sales   4     40     20      8    11       40
4      August           sales   5     50     25     10    11       90
5   September           sales   6     60     30     12    11      150
6     October           sales   7     70     35     14    12       35
7    November  sales quantity   8     80     40     16    12       40
8    December           sales   9     90     45     20    12       80
9     January           sales  10     10     50     10    13       10
10   February           sales  11     10     55     10    13       20
11      March           sales  12     20     60     20    13       40

Answer 2

您可以使用：

quarters = {1: 're9+3', 2: 're', 3: 're3+9', 4: 're6+6'}

col = pd.to_datetime(df['month'], format='%B').dt.quarter.map(quarters)

idx, cols = pd.factorize(col)

df['Q-total'] = (
     pd.Series(df.reindex(cols, axis=1).to_numpy()[np.arange(len(df)), idx],
               index=df.index)
       .groupby(col).cumsum()
    )

输出：

        month             kpi  re  re3+9  re6+6  re9+3  re_o  Q-total
0       April           sales   1     10      5      2     1        1
1         May  sales quantity   2     20     10      4     1        3
2        June           sales   3     30     15      6     1        6
3        July           sales   4     40     20      8    11       40
4      August           sales   5     50     25     10    11       90
5   September           sales   6     60     30     12    11      150
6     October           sales   7     70     35     14    12       35
7    November  sales quantity   8     80     40     16    12       75
8    December           sales   9     90     45     20    12      120
9     January           sales  10     10     50     10    13       10
10   February           sales  11     10     55     10    13       20
11      March           sales  12     20     60     20    13       40