面向营销的 pandas 入门

使用 pandas 分析营销活动

Jill Rosok

Data Scientist

营销团队中的数据科学家做什么？

分析营销活动表现
将转化归因到各营销渠道
A/B 测试

pandas 是什么？

提供便于分析的表格型数据结构

pandas-logo

便捷导入/导出多种常见格式（如 CSV、TSV、Stata）
支持数据操作：连接数据集、按列分组聚合、选取列与行子集等。

使用 pandas 导入数据

import pandas as pd

marketing = pd.read_csv('marketing.csv')

检查数据

print(marketing.head())

      user_id date_served    channel          variant  conv  \
0  a100000029  2018-01-01  House Ads  personalization  True   
1  a100000030  2018-01-01  House Ads  personalization  True   
2  a100000031  2018-01-01  House Ads  personalization  True   
3  a100000032  2018-01-01  House Ads  personalization  True   
4  a100000033  2018-01-01  House Ads  personalization  True   

  language_displayed language_preferred    age_group  
0            English            English   0-18 years  
1            English            English  19-24 years  
2            English            English  24-30 years  
3            English            English  30-36 years  
4            English            English  36-45 years

汇总统计

print(marketing.describe())

           user_id date_served    channel  variant   conv    \
count         9882        9881       9882     9882   9882 
unique        7253          31          5        2      2 
top     a100000882  2018-01-15  House Ads  control  False 
freq             6         782       4682     4986   8883 

           language_displayed language_preferred     age_group
count                    9882               9882          9882
unique                      4                  4             7
top                   English            English   19-24 years
freq                     9695               9177          1650

缺失值与数据类型

print(marketing.info())

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 9996 entries, 0 to 9995
Data columns (total 12 columns):
#   Column                Non-Null Count  Dtype 
    ------                --------------  ----- 
0   user_id               9996 non-null   object
1   date_served           9980 non-null   object
    ...
9   date_subscribed       1815 non-null   object
10  date_canceled         568 non-null    object
11  subscribing_channel   1815 non-null   object
12  is_retained           1815 non-null   object
    dtypes: object(12)
memory usage: 937.2+ KB

Passons à la pratique !

使用 pandas 分析营销活动