pandas交集、差集、并集

      python3使用pandas进行数据的整合。有时候需要取交集、差集、并集。

 

 

# coding:utf-8

import nothing as my_project
import pandas as pd


def get_data():
    file_1 = my_project.project_root + '/resources/' + '1_uniq.txt'
    file_2 = my_project.project_root + '/resources/' + '2_uniq.txt'

    # 加一个额外列
    order_id_me = pd.read_table(file_1, names=['order_id'])
    order_id_me['v1'] = order_id_me.apply(lambda x: 1, axis=1)
    order_id_you = pd.read_table(file_2, names=['order_id'])
    order_id_you['v2'] = order_id_you.apply(lambda x: 2, axis=1)

    # 并集
    union_set = pd.merge(order_id_you, order_id_me, how='outer')
    # 差集
    difference_set = union_set[(union_set['v1'].isnull()) | (union_set['v2'].isnull())]
    # 交集
    intersection_set = union_set[(union_set['v1'].notnull()) & (union_set['v2'].notnull())]

 

     以上代码只针对某些特定的情况，主旨是灵活运用merge函数里的how参数。