如何筛选pandas dataframe后对列执行计算并回填至原数据框

方法1：映射字典+向量化操作（推荐，扩展性强、性能高） #

先把Group和Class的组合与对应除数做成映射字典，直接批量计算回填即可，后续新增规则只要更新字典就行，不用改逻辑：

import pandas as pd

# 构造除数映射：key为(Group, Class)元组，value为对应除数
div_map = {
    ('a', 'x'): 2,
    ('a', 'y'): 3,
    ('b', 'x'): 4,
    ('b', 'y'): 5
}

# 批量匹配每行的除数，计算后直接回填到Data列
df['Data'] = df['Data'] / df.set_index(['Group', 'Class']).index.map(div_map.get)

方法2：np.select条件匹配（适合规则少的场景） #

如果规则不会频繁新增，直接写条件匹配更直观：

import numpy as np

conditions = [
    (df['Group'] == 'a') & (df['Class'] == 'x'),
    (df['Group'] == 'a') & (df['Class'] == 'y'),
    (df['Group'] == 'b') & (df['Class'] == 'x'),
    (df['Group'] == 'b') & (df['Class'] == 'y')
]
# 顺序和上面的条件一一对应
div_list = [2, 3, 4, 5]

df['Data'] = df['Data'] / np.select(conditions, div_list)

内容的提问来源于stack exchange，提问作者user14207664