美好的一天，我想要扩展StartDate和EndDate隐含的日期范围。
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
期待：
让我们考虑row1，我们有2593天，我想要的情况从StartDate即2001-02-02到EndDate即2008-03-10 ，被列出
这应该通过基于范围扩展来遍历所有行，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
直到它到达EndDate的那个日期，这意味着在这两个日期我应该有像StartDate = EndDate，即双方的2008-12-01。考虑到csv
1922 ParisFrance NewYorkUnitedState 2008-12-01 2008-12-01
非常感谢
另一个问题：
谢谢..我有另一个问题。我想创建一个JSON，考虑StartDate（但是，只要有两个日期相互匹配，就会使用其中一个日期，同时附加所有属性。让我举一个例子，
{ "2001-02-02" = { ParisFrance (ArCityArCountry): 1922 NewYorkUnitedStates: 1922} }
如果我们遍历下行csv我们可能会有另一个2001-02-02。我们可以将它附加到初始StartDate，而不是创建它。但是，DptCityDptCountry可能不同，但如果另一个ID与StartDate和DptCityDptCountry匹配，它将被加起来，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}
}
答案 0 :(得分：2)
从：
开始 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
答案 1 :(得分：1)
这只是初始化DataFrame，我可以看到你有：
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列之外的所有列推送到索引中。这留下一列作为系列返回。然后使用apply并返回在每行的扩展日期集上编入索引的系列（Series of Series = DataFrame）。因此，对于DataFrame中的每一行，我得到一个在扩展日期范围内编入索引的系列。然后它只是巧妙的堆叠，命名和reset_index。
# 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（为了审美目的，我删除了StartDate和EndDate）
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]
答案 2 :(得分：0)
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)