Pandas how to sort the connected string within every row?

Question

The sample data is :

import pandas as pd
import numpy as np
d=pd.DataFrame({'lender':['tony','wood','tony','tidy'],
                'borrower':['wood','tony','wood','tony']})

I want to connect lender and borrower and most importantly sort them within each row. In a word, I want get P:

The sort principle is the same as the default sorted function. For example:

sorted(['tony','wood'])
Out[221]: ['tony', 'wood']

sorted(['wood','tony'])
Out[222]: ['tony', 'wood']

apply is preferred unless you have to use for loop. Please read it clearly before you duplicate my question!!!

Are you sure my question is duplicate as the one you mentioned??? — adafdwwf, Mar 14 '19 at 08:07
So you make my question as duplicate, but there is not any answer. Can you cancel you mark??? Or I need to ask this question again!!! — adafdwwf, Mar 14 '19 at 08:10

score 2 · Accepted Answer · answered Mar 14 '19 at 08:17

One solution is use apply with sorted and join per each row:

d['p'] = d[['lender','borrower']].apply(lambda x: '_'.join(sorted(x)), axis=1)
print (d)
  lender borrower          p
0   tony     wood  tony_wood
1   wood     tony  tony_wood
2   tony     wood  tony_wood
3   tidy     tony  tidy_tony

Or use numpy.sort with DataFrame constructor is performance is important:

d1 = pd.DataFrame(np.sort(d[['lender','borrower']], axis=1))
d['p'] = d1[0] + '_' + d1[1]
print (d)
  lender borrower          p
0   tony     wood  tony_wood
1   wood     tony  tony_wood
2   tony     wood  tony_wood
3   tidy     tony  tidy_tony

Pandas how to sort the connected string within every row?

1 Answers1