我想对两个条件之间的连续行值求和。这是我的资料框：

df = pd.DataFrame({'A': ["yes","no","no","no","no","yes","yes","no","no","no","yes"],'B':["no","no","no","no","yes","yes","no","no","no","yes","yes"],'C':[2,5,1,4,6,13,7,8,3,9,1]},index=[0, 1, 2, 3,4,5,6,7,8,9,10])

事实上，当A="yes"和时B="no"，我想开始添加行值直到A="no"和B="yes"。我想得到以下结果：

df1 = pd.DataFrame({'A': ["yes","yes"],'B':["no","no"],'C':[12,18]},index=[0, 6])

uj5u.com热心网友回复：

您可以使用布尔掩码两次来获得所需的结果。

在第一个掩码中，您过滤需要查找其总和的行。在第二个中，您找到总和。

mask = df.apply(lambda x: 1 if (x['A']=='yes')&(x['B']=='no') else (-1 if (x['A']=='no')&(x['B']=='yes') else 0), axis=1).cumsum().astype(bool)
mask2 = df[mask].apply(lambda x: 1 if (x['A']=='yes') & (x['B']=='no') else 0, axis=1)
out = df[mask][mask2.astype(bool)]
out['C'] = df[mask].groupby(mask2.cumsum())['C'].sum().to_numpy()

输出：

     A   B   C
0  yes  no  12
6  yes  no  18