CODEWITHSUNDEEP
pythonpandas
user8766186
user8766186

Reputation: 81

Pandas: Randomize letters in a column

I have a data frame that looks like this:

id1           | id2
----------------------------
ab51c-ee-1a   | cga--=%abd21

I am looking to randomize the letters only:

id1           | id2
----------------------------
ge51r-eq-1b   | olp--=%cqw21

I think I can do something like this:

newid1 = []
for index, row in df.iterrows():
    string = ''
    for i in row['id1']:
        if i.isalpha():
            string+=random.choice(string.letters)
        else:
            string+=i
    newcolumn.append(string)

But it doesn't seem very efficient. Is there a better way?

Upvotes: 2

Views: 1264

Answers (2)

Bharath M Shetty
Bharath M Shetty

Reputation: 30605

Lets use apply, with the power of str.replace to replace only alphabets using regex i.e 

import string 
import random

letters = list(string.ascii_lowercase)

def rand(stri):
    return random.choice(letters)

df.apply(lambda x : x.str.replace('[a-z]',rand))

Output :

           id1            id2
0  gp51e-id-1v      jvj--=%glw21

For one specific column use 

df['id1'].str.replace('[a-z]',rand)

Added by @antonvbr: For future reference, if we want to change upper and lower case we could do this:

letters = dict(u=list(string.ascii_uppercase),l=list(string.ascii_lowercase))

(df['id1'].str.replace('[a-z]',lambda x: random.choice(letters['l']))
          .str.replace('[A-Z]',lambda x: random.choice(letters['u'])))

Upvotes: 4

zipa
zipa

Reputation: 27899

How about this:

import pandas as pd
from string import ascii_lowercase as al
import random

df = pd.DataFrame({'id1': ['ab51c-ee-1a'],
                   'id2': ['cga--=%abd21']})

al = list(al)
df = df.applymap(lambda x: ''.join([random.choice(al) if i in al else i for i in list(x)]))

Upvotes: 1

Related Questions