Pandas.DataFrame实现分组、排序并且为分组插入排名

1. 示例数据（各班级学生得分）

import pandas as pd
data_dict = {"name": 
             ["Rose", "Jack", "Tom", "Kyle", "Jery", "Adam", "Bill", "David", "Denny", "Evan"],
            "class": [1, 2, 3, 1, 2, 3, 1, 2, 3, 1],
            "score": [88, 92, 38, 98, 22, 65, 45, 53, 97, 100]}
df = pd.DataFrame(data=data_dict)
df

2. 按班级分组

df = df.groupby('class', sort=False)\
        .apply(lambda x:x.sort_values("score", ascending=False))\
        .reset_index(drop=True)
df

3. 给各分组班级增加排名列

df["rank"] = None
# 标识班级
flag = df.loc[0].values[1]
rank = 0
for i in range(len(df)):
    temp = df.loc[i].values[1]
    if (temp == flag).all():
        # 同一班级
        rank += 1
    else:
        # 不同班级，重新计算排名
        flag = temp
        rank = 1
    df.loc[i, "rank"] = rank
df

完！