Pandas按日期比较数值

Question

我有一个类似于

Date, Name, Something
2020-04-01,John,10
2020-04-01,Ivory,5
2020-04-01,Sam,3
2020-04-02,John,5
2020-04-02,Ivory,2
2020-04-02,Sam,1
2020-04-02,John,20
2020-04-02,Ivory,3
2020-04-02,Sam,9

和熊猫一起工作，我想通过名字比较昨天的值，和今天的结果来得到增加，减少。轮班（3）不起作用，因为每天的姓名数量不同。 我该怎么做呢。我想加上以前的号码。我试过了

df['old_data'] = df[(work['Name'] == df['Name']) & (df['Date'] ==
                                                                 (df['Date'] - pd.Timedelta(days = 1))
                                                                 )]['Something']

但它不起作用

Answer 1

试试这个：

import pandas as pd


df = pd.DataFrame({'Date': {0:'2020-04-01', 1:'2020-04-01', 2:'2020-04-01', 
                             3:'2020-04-02', 4:'2020-04-02', 5:'2020-04-02', 
                             6:'2020-04-03', 7:'2020-04-03', 8:'2020-04-03'},
                    'Name': {0:'John', 1:'Ivory', 2:'Sam', 
                             3:'John', 4:'Ivory', 5:'Sam', 
                             6:'John', 7:'Ivory', 8:'Sam'},
                    'Something': {0:10, 1:5, 2:3, 
                                  3:5, 4:2, 5:1, 
                                  6:20, 7:3, 8:9}})

df['diff'] = df.groupby('Name')['Something'].diff()

df.dropna()

我更改了您的数据，将最后三个日期更改为“2020-04-03”