使用python迭代日期范围

您好，我想扩展

StartDate

和

EndDate

所暗示的日期范围

import pandas as pd
import datetime
from pandas_datareader import data, wb
import csv

out= open("testfile.csv", "rb")
data = csv.reader(out)
data = [[row[0],row[1] + row[2],row[3] + row[4], row[5],row[6]] for row in data]
out.close()
print data

out=open("data.csv", "wb")
output = csv.writer(out)

for row in data:
    output.writerow(row)

out.close()

df = pd.read_csv('data.csv')
for DateDpt, DateAr in df.iteritems():
    df.DateDpt = pd.to_datetime(df.DateDpt, format='%Y-%m-%d')
    df.DateAr = pd.to_datetime(df.DateAr, format='%Y-%m-%d')

df['DateAr'] = [pd.to_datetime(x, format='%Y-%m-%d') for x in df['DateAr']]
df['DateDpt'] = [pd.to_datetime(x, format='%Y-%m-%d') for x in df['DateDpt']]

df['range'] = df['DateDpt']-df['DateAr']

print df

ID      ArCityArCountry      DptCityDptCountry    EndDate       StartDate  
1922    ParisFrance          NewYorkUnitedState   2008-03-10    2008-12-01  
1901    ParisFrance          LagosNigeria         2001-03-05    2001-02-02  
1922    ParisFrance          NewYorkUnitedState   2011-02-03    2008-12-01  
1002    ParisFrance          CaliforniaUnitedState 2003-03-04   2002-03-04  
1099    ParisFrance           BeijingChina         2011-02-03   2009-02-04  
1901    LosAngelesUnitedState ParisFrance          2001-03-05   2001-02-02 

输出：

ID    ArCityArCountry       DptCityDptCountry     EndDate  
1922  ParisFrance           NewYorkUnitedState    2008-03-10    
1002  LosAngelesUnitedState ForidaUnitedState     2008-03-10     
1901  ParisFrance           LagosNigeria          2001-03-05     
1922  ParisFrance           NewYorkUnitedState    2011-02-03     
1002  ParisFrance           CaliforniaUnitedState 2003-03-04     
1099  ParisFrance           BeijingChina          2011-02-03     
1901  LosAngelesUnitedState ParisFrance           2001-03-05   

StartDate     range    
2001-02-02 2593 days    
2008-12-01  266 days     
2001-02-02   31 days    
2008-12-01  794 days    
2002-03-04  365 days  
2009-02-04  729 days  
2001-02-02   31 days  

预期结果：

<> P> >考虑<代码> ROW1/<代码>，我们有2593天，我想要一个从<代码>开始日期>代码>的情况，即<代码> 2001—02-02/<代码>到<代码>结束日期<代码> IE代码> 2008—03-10<／代码>，请列出 这应该通过基于范围展开来遍历所有行，直到

StartDate

上的值与

EndDate

匹配

ID    ArCityArCountry        DptCityDptCountry    StartDate    EndDate  
1922    ParisFrance          NewYorkUnitedState   2004-03-10    2008-12-01  
1922    ParisFrance          NewYorkUnitedState   2004-03-11    2008-12-01 
1922    ParisFrance          NewYorkUnitedState   2004-03-12    2008-12-01 

在它到达结束日期之前，这意味着在这两个日期上，我都应该有类似StartDate=EndDate的东西，即两边都是2008-12-01。考虑到csv

1922    ParisFrance          NewYorkUnitedState   2008-12-01    2008-12-01

多谢各位

另一个问题是：

---

谢谢我还有一个问题。考虑到StartDate，我想创建一个JSON（但是，如果有两个日期相互匹配，则将使用其中一个日期，同时附加所有属性。让我举一个例子，

{“2001-02-02”={ParisFrance（ArCityArCountry）：1922纽约美国：1922}

如果我们遍历下一个csv，我们可能会有另一个2001-02-02。我们可以将其附加到初始StartDate，而不是创建它。但是，DPTCITYDPCountry可能不同，但是如果另一个ID与StartDate和DPTCITYDPCountry匹配，则会将其相加，即

{"2001-02-02" = { 
ParisFrance (ArCityArCountry): 1922, 2212  //these are IDs with same StartDate and ArCityArCountry   
NewYorkUnitedStates: 1922, 0029  //these are IDs with same StartDate and DptCityDptCountry}
}

首先是：

     ID        ArCityArCountry      DptCityDptCountry  EndDate StartDate
0  1922            ParisFrance     NewYorkUnitedState  3/10/08    2/2/01
1  1002  LosAngelesUnitedState      ForidaUnitedState  3/10/08   12/1/08
2  1901            ParisFrance           LagosNigeria   3/5/01    2/2/01
3  1922            ParisFrance     NewYorkUnitedState   2/3/11   12/1/08
4  1002            ParisFrance  CaliforniaUnitedState   3/4/03    3/4/02
5  1099            ParisFrance           BeijingChina   2/3/11    2/4/09
6  1901  LosAngelesUnitedState            ParisFrance   3/5/01    2/2/01

您可以按如下方式获得所需的输出：

df.EndDate = pd.to_datetime(df.EndDate)
df.StartDate = pd.to_datetime(df.StartDate)
df = df.set_index('StartDate')
new_df = pd.DataFrame()
for i, data in df.iterrows():
    data = data.to_frame().transpose()
    data = data.reindex(pd.date_range(start=data.index[0], end=data.EndDate[0])).fillna(method='ffill').reset_index().rename(columns={'index': 'StartDate'})
    new_df = pd.concat([new_df, data])

new_df = new_df[['ID', 'ArCityArCountry', 'DptCityDptCountry', 'StartDate', 'EndDate']]

      ID        ArCityArCountry   DptCityDptCountry  StartDate    EndDate
0   1922            ParisFrance  NewYorkUnitedState 2001-02-02 2008-03-10
1   1922            ParisFrance  NewYorkUnitedState 2001-02-03 2008-03-10
2   1922            ParisFrance  NewYorkUnitedState 2001-02-04 2008-03-10
3   1922            ParisFrance  NewYorkUnitedState 2001-02-05 2008-03-10
4   1922            ParisFrance  NewYorkUnitedState 2001-02-06 2008-03-10
5   1922            ParisFrance  NewYorkUnitedState 2001-02-07 2008-03-10
6   1922            ParisFrance  NewYorkUnitedState 2001-02-08 2008-03-10
7   1922            ParisFrance  NewYorkUnitedState 2001-02-09 2008-03-10
8   1922            ParisFrance  NewYorkUnitedState 2001-02-10 2008-03-10
9   1922            ParisFrance  NewYorkUnitedState 2001-02-11 2008-03-10
10  1922            ParisFrance  NewYorkUnitedState 2001-02-12 2008-03-10
11  1922            ParisFrance  NewYorkUnitedState 2001-02-13 2008-03-10
12  1922            ParisFrance  NewYorkUnitedState 2001-02-14 2008-03-10
13  1922            ParisFrance  NewYorkUnitedState 2001-02-15 2008-03-10
14  1922            ParisFrance  NewYorkUnitedState 2001-02-16 2008-03-10
15  1922            ParisFrance  NewYorkUnitedState 2001-02-17 2008-03-10
16  1922            ParisFrance  NewYorkUnitedState 2001-02-18 2008-03-10
17  1922            ParisFrance  NewYorkUnitedState 2001-02-19 2008-03-10
18  1922            ParisFrance  NewYorkUnitedState 2001-02-20 2008-03-10
19  1922            ParisFrance  NewYorkUnitedState 2001-02-21 2008-03-10
20  1922            ParisFrance  NewYorkUnitedState 2001-02-22 2008-03-10
21  1922            ParisFrance  NewYorkUnitedState 2001-02-23 2008-03-10
22  1922            ParisFrance  NewYorkUnitedState 2001-02-24 2008-03-10
23  1922            ParisFrance  NewYorkUnitedState 2001-02-25 2008-03-10
24  1922            ParisFrance  NewYorkUnitedState 2001-02-26 2008-03-10
25  1922            ParisFrance  NewYorkUnitedState 2001-02-27 2008-03-10
26  1922            ParisFrance  NewYorkUnitedState 2001-02-28 2008-03-10
27  1922            ParisFrance  NewYorkUnitedState 2001-03-01 2008-03-10
28  1922            ParisFrance  NewYorkUnitedState 2001-03-02 2008-03-10
29  1922            ParisFrance  NewYorkUnitedState 2001-03-03 2008-03-10
..   ...                    ...                 ...        ...        ...
2   1901  LosAngelesUnitedState         ParisFrance 2001-02-04 2001-03-05
3   1901  LosAngelesUnitedState         ParisFrance 2001-02-05 2001-03-05
4   1901  LosAngelesUnitedState         ParisFrance 2001-02-06 2001-03-05
5   1901  LosAngelesUnitedState         ParisFrance 2001-02-07 2001-03-05
6   1901  LosAngelesUnitedState         ParisFrance 2001-02-08 2001-03-05
7   1901  LosAngelesUnitedState         ParisFrance 2001-02-09 2001-03-05
8   1901  LosAngelesUnitedState         ParisFrance 2001-02-10 2001-03-05
9   1901  LosAngelesUnitedState         ParisFrance 2001-02-11 2001-03-05
10  1901  LosAngelesUnitedState         ParisFrance 2001-02-12 2001-03-05
11  1901  LosAngelesUnitedState         ParisFrance 2001-02-13 2001-03-05
12  1901  LosAngelesUnitedState         ParisFrance 2001-02-14 2001-03-05
13  1901  LosAngelesUnitedState         ParisFrance 2001-02-15 2001-03-05
14  1901  LosAngelesUnitedState         ParisFrance 2001-02-16 2001-03-05
15  1901  LosAngelesUnitedState         ParisFrance 2001-02-17 2001-03-05
16  1901  LosAngelesUnitedState         ParisFrance 2001-02-18 2001-03-05
17  1901  LosAngelesUnitedState         ParisFrance 2001-02-19 2001-03-05
18  1901  LosAngelesUnitedState         ParisFrance 2001-02-20 2001-03-05
19  1901  LosAngelesUnitedState         ParisFrance 2001-02-21 2001-03-05
20  1901  LosAngelesUnitedState         ParisFrance 2001-02-22 2001-03-05
21  1901  LosAngelesUnitedState         ParisFrance 2001-02-23 2001-03-05
22  1901  LosAngelesUnitedState         ParisFrance 2001-02-24 2001-03-05
23  1901  LosAngelesUnitedState         ParisFrance 2001-02-25 2001-03-05
24  1901  LosAngelesUnitedState         ParisFrance 2001-02-26 2001-03-05
25  1901  LosAngelesUnitedState         ParisFrance 2001-02-27 2001-03-05
26  1901  LosAngelesUnitedState         ParisFrance 2001-02-28 2001-03-05
27  1901  LosAngelesUnitedState         ParisFrance 2001-03-01 2001-03-05
28  1901  LosAngelesUnitedState         ParisFrance 2001-03-02 2001-03-05
29  1901  LosAngelesUnitedState         ParisFrance 2001-03-03 2001-03-05
30  1901  LosAngelesUnitedState         ParisFrance 2001-03-04 2001-03-05
31  1901  LosAngelesUnitedState         ParisFrance 2001-03-05 2001-03-05

---

这只是将数据帧初始化为我可以看到的数据帧：

cols = ['ID', 'ArCityArCountry', 'DptCityDptCountry', 'EndDate', 'StartDate']
df = pd.DataFrame(dict(ID=[1922, 1002, 1901, 1922, 1002, 1099, 1902],
                       ArCityArCountry=['ParisFrance',
                                      'LosAngelesUnitedStates',
                                      'ParisFrance',
                                      'ParisFrance',
                                      'ParisFrance',
                                      'ParisFrance',
                                      'LosAngelesUnitedStates'],
                       DptCityDptCountry=['NewYorkUnitedStates',
                                          'FloridaUnitedStates',
                                          'LagosNigeria',
                                          'NewYorkUnitedStates',
                                          'CaliforniaUnitedStates',
                                          'BeijingChina',
                                          'ParisFrance'],
                       EndDate=pd.to_datetime(['3/10/08',
                                               '3/10/08',
                                               '3/5/01',
                                               '2/3/11',
                                               '3/4/03',
                                               '2/3/11',
                                               '3/5/01']),
                       StartDate=pd.to_datetime(['2/2/01',
                                                 '12/1/08',
                                                 '2/2/01',
                                                 '12/1/08',
                                                 '3/4/02',
                                                 '2/4/09',
                                                 '2/2/01'])))[cols]

然后，我使用set_index将除1列以外的所有列推入索引。这将留下一列作为序列返回。然后使用apply并返回一个序列，该序列在扩展的日期集合上为每行编制索引（Series of Series=DataFrame）.因此，对于数据框中的7行中的每一行，我都会在扩展的日期范围内获得一个系列索引。然后，这就是巧妙的堆叠、命名和重置索引

# Use idx to clean up the set_index call
idx = ['ID', 'ArCityArCountry', 'DptCityDptCountry', 'EndDate']

def f(x):
    # x will be an element of a series with the values of the columns specified in idx
    # as the index value which is stored in the name attribute.
    # x.name[-1] is the last element of the name attribute which is the
    # EndDate.  This corresponds to the last element of the idx list above
    date_index = pd.Index(pd.date_range(x.StartDate, x.name[-1])

    # I return a named series so the 'Date' becomes a column name
    return pd.Series(x.StartDate, index=date_index, name='Date'))

temp = df.set_index(idx).apply(f, axis=1)
# I didn't have to wrap temp.stack() in a series but doing so allows me
# to name it and have that show up as a column name
final = pd.Series(temp.stack(), name='StartDate').reset_index()

结果如下所示（为了美观起见，我删除了StartDate和EndDate）

标准进口 将表复制到剪贴板 将剪贴板导入数据帧 将日期列设置为datetime64数据类型 由于ID在给定数据集中不唯一，因此创建密钥 将StartDate设置为reindex连接数据的索引 数据转换的核心 从这里进一步清理桌子…但这是期望的结果 资料来源：

---

您想将第一行扩展到2593行吗？这些行的内容应该是什么？您如何处理明显重叠的日期范围？您可以发布一个您预期输出的插图吗？是的，先生…我会这样做的，先生，Stefan Jansen您是问题的解决方案，但我不明白单击o是什么意思请看下面我的答案。先生，我补充了另一个问题。ThanksIt非常有效，但并不完全符合我的要求。我认为Stefan已经提出了解决方案。谢谢先生，我非常感谢。Hanks。我还有一个问题

print final[idx[:-1] + ['Date']]

        ID         ArCityArCountry    DptCityDptCountry       Date
0     1922             ParisFrance  NewYorkUnitedStates 2001-02-02
1     1922             ParisFrance  NewYorkUnitedStates 2001-02-03
2     1922             ParisFrance  NewYorkUnitedStates 2001-02-04
3     1922             ParisFrance  NewYorkUnitedStates 2001-02-05
4     1922             ParisFrance  NewYorkUnitedStates 2001-02-06
5     1922             ParisFrance  NewYorkUnitedStates 2001-02-07
6     1922             ParisFrance  NewYorkUnitedStates 2001-02-08
7     1922             ParisFrance  NewYorkUnitedStates 2001-02-09
8     1922             ParisFrance  NewYorkUnitedStates 2001-02-10
9     1922             ParisFrance  NewYorkUnitedStates 2001-02-11
10    1922             ParisFrance  NewYorkUnitedStates 2001-02-12
11    1922             ParisFrance  NewYorkUnitedStates 2001-02-13
12    1922             ParisFrance  NewYorkUnitedStates 2001-02-14
13    1922             ParisFrance  NewYorkUnitedStates 2001-02-15
14    1922             ParisFrance  NewYorkUnitedStates 2001-02-16
15    1922             ParisFrance  NewYorkUnitedStates 2001-02-17
16    1922             ParisFrance  NewYorkUnitedStates 2001-02-18
17    1922             ParisFrance  NewYorkUnitedStates 2001-02-19
18    1922             ParisFrance  NewYorkUnitedStates 2001-02-20
19    1922             ParisFrance  NewYorkUnitedStates 2001-02-21
20    1922             ParisFrance  NewYorkUnitedStates 2001-02-22
21    1922             ParisFrance  NewYorkUnitedStates 2001-02-23
22    1922             ParisFrance  NewYorkUnitedStates 2001-02-24
23    1922             ParisFrance  NewYorkUnitedStates 2001-02-25
24    1922             ParisFrance  NewYorkUnitedStates 2001-02-26
25    1922             ParisFrance  NewYorkUnitedStates 2001-02-27
26    1922             ParisFrance  NewYorkUnitedStates 2001-02-28
27    1922             ParisFrance  NewYorkUnitedStates 2001-03-01
28    1922             ParisFrance  NewYorkUnitedStates 2001-03-02
29    1922             ParisFrance  NewYorkUnitedStates 2001-03-03
...    ...                     ...                  ...        ...
4519  1901  LosAngelesUnitedStates          ParisFrance 2001-02-04
4520  1901  LosAngelesUnitedStates          ParisFrance 2001-02-05
4521  1901  LosAngelesUnitedStates          ParisFrance 2001-02-06
4522  1901  LosAngelesUnitedStates          ParisFrance 2001-02-07
4523  1901  LosAngelesUnitedStates          ParisFrance 2001-02-08
4524  1901  LosAngelesUnitedStates          ParisFrance 2001-02-09
4525  1901  LosAngelesUnitedStates          ParisFrance 2001-02-10
4526  1901  LosAngelesUnitedStates          ParisFrance 2001-02-11
4527  1901  LosAngelesUnitedStates          ParisFrance 2001-02-12
4528  1901  LosAngelesUnitedStates          ParisFrance 2001-02-13
4529  1901  LosAngelesUnitedStates          ParisFrance 2001-02-14
4530  1901  LosAngelesUnitedStates          ParisFrance 2001-02-15
4531  1901  LosAngelesUnitedStates          ParisFrance 2001-02-16
4532  1901  LosAngelesUnitedStates          ParisFrance 2001-02-17
4533  1901  LosAngelesUnitedStates          ParisFrance 2001-02-18
4534  1901  LosAngelesUnitedStates          ParisFrance 2001-02-19
4535  1901  LosAngelesUnitedStates          ParisFrance 2001-02-20
4536  1901  LosAngelesUnitedStates          ParisFrance 2001-02-21
4537  1901  LosAngelesUnitedStates          ParisFrance 2001-02-22
4538  1901  LosAngelesUnitedStates          ParisFrance 2001-02-23
4539  1901  LosAngelesUnitedStates          ParisFrance 2001-02-24
4540  1901  LosAngelesUnitedStates          ParisFrance 2001-02-25
4541  1901  LosAngelesUnitedStates          ParisFrance 2001-02-26
4542  1901  LosAngelesUnitedStates          ParisFrance 2001-02-27
4543  1901  LosAngelesUnitedStates          ParisFrance 2001-02-28
4544  1901  LosAngelesUnitedStates          ParisFrance 2001-03-01
4545  1901  LosAngelesUnitedStates          ParisFrance 2001-03-02
4546  1901  LosAngelesUnitedStates          ParisFrance 2001-03-03
4547  1901  LosAngelesUnitedStates          ParisFrance 2001-03-04
4548  1901  LosAngelesUnitedStates          ParisFrance 2001-03-05

[4549 rows x 4 columns]

import pandas as pd
import numpy as np

     ID        ArCityArCountry      DptCityDptCountry  EndDate StartDate
0  1922            ParisFrance     NewYorkUnitedState  3/10/08    2/2/01
1  1002  LosAngelesUnitedState      ForidaUnitedState  3/10/08   12/1/08
2  1901            ParisFrance           LagosNigeria   3/5/01    2/2/01
3  1922            ParisFrance     NewYorkUnitedState   2/3/11   12/1/08
4  1002            ParisFrance  CaliforniaUnitedState   3/4/03    3/4/02
5  1099            ParisFrance           BeijingChina   2/3/11    2/4/09
6  1901  LosAngelesUnitedState            ParisFrance   3/5/01    2/2/01

df = pd.read_clipboard()

df['StartDate'] = pd.to_datetime(df['StartDate'])
df['EndDate'] = pd.to_datetime(df['EndDate'])

df['Unique_ID'] = df.index

df.set_index('StartDate', inplace=True)

def reindex_by_date(df):
    dates = pd.date_range(df.index.min(), df['EndDate'].min())
    return df.reindex(dates).ffill()

df = df.groupby('Unique_ID').apply(reindex_by_date)