CODEWITHSUNDEEP
pythonjsonpython-3.xpandasdataframe
dm2111
dm2111

Reputation: 145

Create df column as json of row

I have the below code, that iterates through the df and creates an additional column new with a dict of the other columns. Is there a better way to achieve this without using iterrows? My actual dataset is much larger and iterating through rows is not performant.

Code

import pandas as pd
import json

data = {'Name':['Tom', 'Nick', 'Jim', 'John'],
        'Age':[20, 21, 35, 11]}
 
df = pd.DataFrame(data)

for i, row in df.iterrows():
    df.loc[i, 'new'] = json.dumps(row.to_dict())

print(df)

Output

  Name  Age                          new
0   Tom   20   {"Name": "Tom", "Age": 20}
1  Nick   21  {"Name": "Nick", "Age": 21}
2   Jim   35   {"Name": "Jim", "Age": 35}
3  John   11  {"Name": "John", "Age": 11}

Upvotes: 1

Views: 42

Answers (1)

anky
anky

Reputation: 75080

You can try df.to_dict with df.join

out = df.join(pd.Series(df.to_dict('records'),index=df.index,name='new'))

print(out)

   Name  Age                          new
0   Tom   20   {'Name': 'Tom', 'Age': 20}
1  Nick   21  {'Name': 'Nick', 'Age': 21}
2   Jim   35   {'Name': 'Jim', 'Age': 35}
3  John   11  {'Name': 'John', 'Age': 11}

Upvotes: 2

Related Questions