查找具有任何列列表中的所有字符串列表的数据框行

debugcn 发表于 Dev

发布

我列出了可能出现在三列中的任何三列的术语。我想在三列中的某处找到所有包含所有三个术语的行。

从...开始。。。

我有一个看起来像这样的数据框：

In [1]: data = [
    ['Missouri','Kansas City','Jackson County'],
    ['Boston','Massachusetts','Suffolk County'],
    ['Los Angeles','Los Angeles County','California'],
    ['Massachusetts','Suffolk County','Boston'],
    ['Philadelphia','Pennsylvania','Philadelphia County'],
    ['Kansas City','Missouri','Jackson County'],
    ['Kansas City','Missouri','Platte County']
]

In [2]: columns=['location0','location1','location2']

In [3]: df = pd.DataFrame(data,columns=columns)

In [4]: df

Out[4]: 
    location0       location1           location2
0   Missouri        Kansas City         Jackson County
1   Boston          Massachusetts       Suffolk County
2   Los Angeles     Los Angeles County  California
3   Massachusetts   Suffolk County      Boston
4   Philadelphia    Pennsylvania        Philadelphia County
5   Kansas City     Missouri            Jackson County
6   Kansas City     Missouri            Platte County

我想获得描述密苏里州杰克逊县堪萨斯城的数据框的所有行。

换句话说，我希望所有这三项均位于所有行中：['Kansas City'，'Jackson County'，'Missori']位于以下三列中的任何一列：['location0'，'location1'，'location2' ]：

目标

结果将如下所示：

    location0       location1       location2
0   Missouri        Kansas City     Jackson County
1   Kansas City     Missouri        Jackson County

尝试失败。。。

In [1]: look_for = ['Kansas City','Jackson County','Missouri']

In [2]: look_in = ['location0','location1','location2']

In [3]: result = df[col for col in look_in if df[col].isin(look_for)]

In [4]: result

Out [4]:   
File "<ipython-input-408-c42b1236af99>", line 3
result = df[col for col in look_in if df[col].isin(look_for)]
                  ^
SyntaxError: invalid syntax

在此先感谢您的帮助！

编辑：输入错误look_for

耶斯列尔

DataFrame.isin通过过滤后的值在过滤后的列中使用，并为True每行测试添加DataFrame.all：

result = df[df[look_in].isin(look_for).all(axis=1)]
print (result)

     location0    location1       location2
0     Missouri  Kansas City  Jackson County
5  Kansas City     Missouri  Jackson County

本文收集自互联网，转载请注明来源。

如有侵权，请联系[email protected] 删除。