pandas groupby 分组取每组的前几行记录方法

import pandas as pd

df = pd.DataFrame({'class':['a','a','b','b','a','a','b','c','c'],'score':[3,5,6,7,8,9,10,11,14]})

df.sort_values(['class','score'],ascending=[1,0],inplace=True)

grouped = df.groupby(['class']).head(2)