Pandas 条件和数据帧

我正试图根据列“性别”中的值，对熊猫数据框中的值进行聚合和求和。这是我正在处理的数据集的一个示例：

df_genders = pd.DataFrame({'Country': ['US','US','US','US','US','India','India','India','UK','UK','UK','UK'],
                          'Gender': ['Man','Woman', 'Non-Binary,Genderqueer', 'Non-Binary', 'Non-Binary,Genderqueer,Non-Conforming',
                                    'Man','Woman','Non-Binary','Man','Woman', 'Non-Binary,Genderqueer', 'Non-Binary,Genderqueer,Non-Conforming'],
                          'Count': [7996,915,11,34,153,3857,287,47,2566,272,72,99]})
df_genders

由于性别价值观不太一致，我想把它们归为一组，并对它们的数量进行合计，以便为每个国家得出男性、女性和非二元（非二元既不是“男性”也不是“女性”）的总和。 我无法编写条件分组和求和的代码，因此我的方法是找出每个国家的总数，然后从总数中减去男性+女性的总和，因此剩下非二进制的总和：

df_genders.groupby('Country')['Count'].sum() - df_genders[(df_genders['Gender']=='Man') | (df_genders['Gender']=='Woman')].groupby('Country')['Count'].sum()

您知道解决此问题的更好方法吗，或者通常知道执行条件聚合（group by和sum）的方法吗

谢谢大家!

您可以直接执行以下操作：

res = df_genders[~df_genders['Gender'].isin(('Man', 'Woman'))]['Count'].sum()
print(res)

输出

416

   Country                                 Gender  Count grouped-genders
0       US                                    Man   7996             Man
1       US                                  Woman    915           Woman
2       US                 Non-Binary,Genderqueer     11      Non-Binary
3       US                             Non-Binary     34      Non-Binary
4       US  Non-Binary,Genderqueer,Non-Conforming    153      Non-Binary
5    India                                    Man   3857             Man
6    India                                  Woman    287           Woman
7    India                             Non-Binary     47      Non-Binary
8       UK                                    Man   2566             Man
9       UK                                  Woman    272           Woman
10      UK                 Non-Binary,Genderqueer     72      Non-Binary
11      UK  Non-Binary,Genderqueer,Non-Conforming     99      Non-Binary

  grouped-genders  Count
0             Man  14419
1      Non-Binary    416
2           Woman   1474

但我认为，如果您创建一个包含您正在寻找的分类的新列，例如，一种方法：

df_genders['grouped-genders'] = df_genders['Gender'].map({ 'Man' : 'Man', 'Woman' : 'Woman' }).fillna('Non-Binary')
print(df_genders)

输出

416

   Country                                 Gender  Count grouped-genders
0       US                                    Man   7996             Man
1       US                                  Woman    915           Woman
2       US                 Non-Binary,Genderqueer     11      Non-Binary
3       US                             Non-Binary     34      Non-Binary
4       US  Non-Binary,Genderqueer,Non-Conforming    153      Non-Binary
5    India                                    Man   3857             Man
6    India                                  Woman    287           Woman
7    India                             Non-Binary     47      Non-Binary
8       UK                                    Man   2566             Man
9       UK                                  Woman    272           Woman
10      UK                 Non-Binary,Genderqueer     72      Non-Binary
11      UK  Non-Binary,Genderqueer,Non-Conforming     99      Non-Binary

  grouped-genders  Count
0             Man  14419
1      Non-Binary    416
2           Woman   1474

然后按新列分组以获得所有性别的计数：

res = df_genders.groupby('grouped-genders')['Count'].sum().reset_index()
print(res)

输出

416

   Country                                 Gender  Count grouped-genders
0       US                                    Man   7996             Man
1       US                                  Woman    915           Woman
2       US                 Non-Binary,Genderqueer     11      Non-Binary
3       US                             Non-Binary     34      Non-Binary
4       US  Non-Binary,Genderqueer,Non-Conforming    153      Non-Binary
5    India                                    Man   3857             Man
6    India                                  Woman    287           Woman
7    India                             Non-Binary     47      Non-Binary
8       UK                                    Man   2566             Man
9       UK                                  Woman    272           Woman
10      UK                 Non-Binary,Genderqueer     72      Non-Binary
11      UK  Non-Binary,Genderqueer,Non-Conforming     99      Non-Binary

  grouped-genders  Count
0             Man  14419
1      Non-Binary    416
2           Woman   1474

---

谢谢你，丹尼！我在groupby中还添加了“国家”一栏，结果正是我想要的：）