Question

我有这个数据框：

word, string1, string2
SQL, SQL is good, Programming
Java, Programming, Java is good
C#, Programming, Programming

如果我的列字值出现在列string1上，则我有一个给出布尔值的列：

data['res'] = data.apply(lambda x: x.word in x.string1
                               if (x.string1 == x.string1) and (x.word == x.word)
                               else False)

但是我想看看列字值是否出现在string1列或string2中？像这样：

data['res'] = data.apply(lambda x: x.word in x.string1
                               if (x.string1 == x.string1) and (x.word == x.word)
                               else (x.word in x.string2
                                    if (x.string2 == x.string2) and (x.word == x.word))axis=1)
                    else False)

我想要的是：

word, string1, string2, res
SQL, SQL is good, Programming, True
Java, Programming, Java is good, True
C#, Programming, Programming, False

这可能吗？

谢谢！

Answer 1

您需要检查第1列中的字符串是否存在于其他任何列中，且any()的第1轴位于上方：

df.apply(lambda x:x.str.contains(x.word),axis=1).iloc[:,1:].any(axis=1)

0     True
1     True
2    False

完整代码：

df=df.assign(res=df.apply(lambda x:x.str.contains(x.word),axis=1).iloc[:,1:].any(axis=1))

   word       string1        string2    res
0   SQL   SQL is good    Programming   True
1  Java   Programming   Java is good   True
2    C#   Programming    Programming  False

Answer 2

最简单的方法是同时连接两列并添加另一个过滤器：

data['res'] = data.apply(lambda x: x.word in x.string1 + x.string2
                               if (x.string1 == x.string1 ) and 
                                  (x.word == x.word) and 
                                  (x.string1 == x.string1)
                               else False, axis=1)

Python-在DataFrame Apply函数上添加OR运算符

2 个答案: