how to get for multiple columns

Question

I have a data frame like this:

 Id   row      Date       BuyTime    SellPrice    App
 1    1      2017-10-30    94520       0          9:00:00
 1    2      2017-10-30    94538       0          9:00:00
 1    3      2017-10-30    94609       0          9:00:00
 1    4      2017-10-30    94615       0          9:00:00
 1    5      2017-10-30    94617       0          9:00:00
 1    1      2017-09-20    99100       99159      9:00:10
 1    2      2017-09-20    99102       99058      9:00:11
 1    3      2017-09-20    99103       99057      9:00:12
 1    4      2017-09-20    99104       99056      9:00:10
 1    5      2017-09-20    99105       99055      9:00:10
 1    1      2017-09-20    99100       99190      9:01:10
 1    2      2017-09-20    98099       99091      9:01:10
 1    3      2017-09-20    98098       99092      9:01:10
 1    4      2017-09-20    98097       99093      9:01:10
 1    5      2017-09-20    98096       99094      9:01:10
 2    1      2010-11-01    99890       100000     10:00:02
 2    2      2010-11-01    99899       100000     10:00:02
 2    3      2010-11-01    99901       99899      9:00:02
 2    4      2010-11-01    99920       99850      10:00:02
 2    5      2010-11-01    99933       99848      10:00:23

I want to change BuyTime format to time format (%H:%M:%S), and then create a new column and merge Date and BuyTime for all rows, for instance :

   Id     row     Date       BuyTime        SellPrice    App         TimeStamp
   1      1     2017-10-30   09:45:20        0           9:00:00     2017-10-30  09:45:20

score 1 · Accepted Answer · answered Nov 04 '17 at 13:26

Use str.zfill with str[] for select by positions of string and then to_timedelta:

s = df['BuyTime'].astype(str).str.zfill(6)
df['BuyTime'] = s.str[:2] + ':' + s.str[2:4] + ':' + s.str[4:]

df['TimeStamp'] = df['Date'] + pd.to_timedelta(df['BuyTime'])
#if not datetime column convert it
#df['TimeStamp'] = pd.to_datetime(df['Date']) + pd.to_timedelta(df['BuyTime'])
print (df)
    Id  row        Date   BuyTime  SellPrice       App           TimeStamp
0    1    1  2017-10-30  09:45:20          0   9:00:00 2017-10-30 09:45:20
1    1    2  2017-10-30  09:45:38          0   9:00:00 2017-10-30 09:45:38
2    1    3  2017-10-30  09:46:09          0   9:00:00 2017-10-30 09:46:09
3    1    4  2017-10-30  09:46:15          0   9:00:00 2017-10-30 09:46:15
4    1    5  2017-10-30  09:46:17          0   9:00:00 2017-10-30 09:46:17
5    1    1  2017-09-20  09:91:00      99159   9:00:10 2017-09-20 10:31:00
6    1    2  2017-09-20  09:91:02      99058   9:00:11 2017-09-20 10:31:02
7    1    3  2017-09-20  09:91:03      99057   9:00:12 2017-09-20 10:31:03
8    1    4  2017-09-20  09:91:04      99056   9:00:10 2017-09-20 10:31:04
9    1    5  2017-09-20  09:91:05      99055   9:00:10 2017-09-20 10:31:05
10   1    1  2017-09-20  09:91:00      99190   9:01:10 2017-09-20 10:31:00
11   1    2  2017-09-20  09:80:99      99091   9:01:10 2017-09-20 10:21:39
12   1    3  2017-09-20  09:80:98      99092   9:01:10 2017-09-20 10:21:38
13   1    4  2017-09-20  09:80:97      99093   9:01:10 2017-09-20 10:21:37
14   1    5  2017-09-20  09:80:96      99094   9:01:10 2017-09-20 10:21:36
15   2    1  2010-11-01  09:98:90     100000  10:00:02 2010-11-01 10:39:30
16   2    2  2010-11-01  09:98:99     100000  10:00:02 2010-11-01 10:39:39
17   2    3  2010-11-01  09:99:01      99899   9:00:02 2010-11-01 10:39:01
18   2    4  2010-11-01  09:99:20      99850  10:00:02 2010-11-01 10:39:20
19   2    5  2010-11-01  09:99:33      99848  10:00:23 2010-11-01 10:39:33

how to get for multiple columns

1 Answers1