Concatenate rows in python dataframe

Question

This question may be very basic, but I would like to concatenate three columns in a pandas DataFrame.
I would like to concatenate col1, col2 and col3 into col4. I know in R this could be done with the paste function quite easily.

df = pd.DataFrame({'col1': [2012, 2013, 2014], 'col2': 'q', 'col3': range(3)})

Edit: Code for clarity - I would like to generate col4 automatically:

x=pd.DataFrame()
x['col1'] = [2012,2013,2013]
x['col2'] = ['q', 'q', 'q']
x['col3'] = [1,2,3]
x['col4'] = ['2012q1', '2013q2', '2014q4']

piRSquared · Accepted Answer · 2017-08-23 19:32:46Z

4

Use pd.DataFrame.sum with axis=1 after converting to strings.
I use pd.DataFrame.assign to create a copy with the new column

df.assign(col4=df[['col1', 'col2', 'col3']].astype(str).sum(1))

   col1 col2  col3    col4
0  2012    q     1  2012q1
1  2013    q     2  2013q2
2  2014    q     3  2014q3

Or you can add a column inplace

df['col4'] = df[['col1', 'col2', 'col3']].astype(str).sum(1)
df

   col1 col2  col3    col4
0  2012    q     1  2012q1
1  2013    q     2  2013q2
2  2014    q     3  2014q3

If df only has the three columns, you can reduce code to

df.assign(col4=df.astype(str).sum(1))

If df has more than three columns but the three you want to concat are the first three

df.assign(col4=df.iloc[:, :3].astype(str).sum(1))

edited Aug 23, 2017 at 19:32

answered Aug 23, 2017 at 19:28

piRSquared

296k68 gold badges509 silver badges654 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

Vaishali Over a year ago

sum on strings:)

ktj1989 Over a year ago

This solution worked on the code that was provided but on my actual data set received a 'Wrong number of dimensions' error

piRSquared Over a year ago

That means you misrepresented your data. Also, I have no idea what your error means. You should post the entire error to provide more context.

C8H10N4O2 · Accepted Answer · 2017-08-23 20:19:03Z

2

To concatenate across all columns, it may be more convenient to write df.apply(..., axis=1), as in:

df['col4'] = df.apply(lambda x: "".join(x.astype(str)),axis=1)
df

#   col1 col2  col3    col4
#0  2012    q     1  2012q1
#1  2013    q     2  2013q2
#2  2014    q     3  2014q3

especially if you have a lot of columns and don't want to write them all out (as required by Kyle's answer).

edited Aug 23, 2017 at 20:19

answered Aug 23, 2017 at 20:12

C8H10N4O2

19.2k10 gold badges106 silver badges145 bronze badges

Comments

Kyle · Accepted Answer · 2017-08-23 19:40:15Z

1

df['col4'] = df.col1.astype(str) + df.col2 + df.col3.astype(str)

edited Aug 23, 2017 at 19:40

answered Aug 23, 2017 at 19:31

Kyle

2,9342 gold badges21 silver badges30 bronze badges

Collectives™ on Stack Overflow

Concatenate rows in python dataframe

3 Answers 3

3 Comments

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

3 Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related