多索引切片比较

Question

我正在使用来自以下位置的婴儿姓名数据集： https://raw.githubusercontent.com/hadley/data-baby-names/master/baby-names.csv

其中的数据看起来像：

"year","name","percent","sex"
1880,"John",0.081541,"boy"
1880,"William",0.080511,"boy"
1880,"James",0.050057,"boy"
1880,"Charles",0.045167,"boy"
1880,"George",0.043292,"boy"
1880,"Frank",0.02738,"boy"
1880,"Joseph",0.022229,"boy"

我将所有名字组合在一起，并将男孩和女孩的百分比相加：

data1.groupby(['name','sex'])[['percent']].sum()

这会创建一个多索引数据框：

   Name    Sex  Percent

   Aaron   boy  0.292292
           girl 0.000805
   Abagail girl 0.001326
   Abbie   boy  0.000092
           girl 0.022804

对于每个名字，我想在一个新的数据框中返回更高百分比的性别：

   Name    Sex  Percent

   Aaron   boy  0.292292
   Abagail girl 0.001326
   Abbie   girl 0.022804

我一直在查看multi-index documentation，但无法弄清楚这一点。任何帮助表示赞赏。

Answer 1

您可以在groupby sum 之后使用groupby tail

s=df.groupby(['name','sex'])[['percent']].sum()
s.sort_values('Percent').groupby(level=0).tail(1)