CODEWITHSUNDEEP
pythonpandascsv
sshah98
sshah98

Reputation: 352

Replace part of string with a different column's value pandas + python

The Problem

I have a CSV file full of data like this 

 rowid       filepath              curr_time    cameraid    old_cameraid
    0        1/1_20180625_234436.jpg    2018-06-25 23:44:36 1       19
    1        1/1_20180626_005104.jpg    2018-06-26 00:51:04 1       19
    2        1/1_20180626_015735.jpg    2018-06-26 01:57:35 1       19
    3        1/1_20180626_030430.jpg    2018-06-26 03:04:30 1       19
    ...
    2605    2/2_20180622_064322.jpg     2018-06-22 06:43:22 2       64
    2606    2/2_20180622_074326.jpg     2018-06-22 07:43:26 2       64
    2607    2/2_20180622_084332.jpg     2018-06-22 08:43:32 2       64

What I want to do

I want to use the old_cameraid string value to replace the filepath number:

rowid   filepath    curr_time   cameraid    old_cameraid
0       19/19_20180625_234436.jpg       2018-06-25 23:44:36 1    19
...
2605    64/64_20180622_064322.jpg       2018-06-22 06:43:22 2       64

What I've tried

I've tried using 

df.apply(lambda x: x['filepath'].replace('a',x['b']), axis=1)

from this This StackOverflow question but I'm not replacing the 'a', but instead a part of the string in the filepath column. Therefore, this wasn't working and I'm not sure.

I also apologize for the formatting of the rows. I tried to search how to properly format rows in a stackoverflow question but could not find the correct information.

Upvotes: 1

Views: 97

Answers (1)

Kenan
Kenan

Reputation: 14094

You can add them all together, i assumed all filepath have the same syntax

1/1_20180625_234436.jpg 

df['filepath'] = df['old_cameraid'].astype(str) + '/' + df['filepath'].astype(str).map(lambda x: x.split('/')[-1])

Upvotes: 2

Related Questions