CODEWITHSUNDEEP
pythonpandasdictionary
Simon
Simon

Reputation: 631

Turn series of dictionaries into a DataFrame - Pandas

I am trying to convert a series of dictionaries into a dataframe

0      {'neg': 0.0, 'neu': 0.462, 'pos': 0.538}
1      {'neg': 0.0, 'neu': 0.609, 'pos': 0.391}
2      {'neg': 0.043, 'neu': 0.772, 'pos': 0.185}
3      {'neg': 0.035, 'neu': 0.765, 'pos': 0.2}
4      {'neg': 0.0, 'neu': 0.655, 'pos': 0.345}
5      {'neg': 0.0, 'neu': 0.631, 'pos': 0.369}

I want the resulting DataFrame to have each key be its own column.

neg   neu     pos
0.0.  0.462   0.538
0.0   0.609   0.391
..    ..      ..

How can I accomplish this with Pandas?

Upvotes: 11

Views: 7812

Answers (2)

user2285236
user2285236

Reputation:

Given your Series, ser

ser
Out: 
0      {'neg': 0.0, 'neu': 0.462, 'pos': 0.538}
1      {'neg': 0.0, 'neu': 0.609, 'pos': 0.391}
2    {'neg': 0.043, 'neu': 0.772, 'pos': 0.185}
3      {'neg': 0.035, 'neu': 0.765, 'pos': 0.2}
4      {'neg': 0.0, 'neu': 0.655, 'pos': 0.345}
5      {'neg': 0.0, 'neu': 0.631, 'pos': 0.369}

You can convert the Series to a list and call the DataFrame constructor:

pd.DataFrame(ser.tolist())
Out: 
     neg    neu    pos
0  0.000  0.462  0.538
1  0.000  0.609  0.391
2  0.043  0.772  0.185
3  0.035  0.765  0.200
4  0.000  0.655  0.345
5  0.000  0.631  0.369

Or you can apply the pd.Series constructor to each row. apply will be flexible and return a DataFrame since each row is a Series now.

ser.apply(pd.Series)
Out: 
     neg    neu    pos
0  0.000  0.462  0.538
1  0.000  0.609  0.391
2  0.043  0.772  0.185
3  0.035  0.765  0.200
4  0.000  0.655  0.345
5  0.000  0.631  0.369

Upvotes: 32

Alter
Alter

Reputation: 3464

There is probably a better way to do this... but this appears pretty easy with the structured data you have.

Otherwise look at this post for reforming the dictionary

import pandas as pd

a = [{'neg': 0.0, 'neu': 0.462, 'pos': 0.538},
{'neg': 0.0, 'neu': 0.609, 'pos': 0.391},
{'neg': 0.043, 'neu': 0.772, 'pos': 0.185},
{'neg': 0.035, 'neu': 0.765, 'pos': 0.2},
{'neg': 0.0, 'neu': 0.655, 'pos': 0.345},
{'neg': 0.0, 'neu': 0.631, 'pos': 0.369}]

b = dict()
for key in a[0].keys():
    b[key] = []
    for dic in a:
         b[key].append(dic[key])

pd.DataFrame(b)

enter image description here

Upvotes: 0

Related Questions