Converting Array Of Arrays Into Flattened Dataframe

Got a pandas dataframe with the below structure 0 [{'review_id': 4873356, 'rating': '5.0'}, {'review_id': 4973356, 'rating': '4.0'}] 1 [{'review_id': 4635892, 'rating': '5.0'

Solution 1:

You wind up getting an array of lists of dicts so need:

import pandas as pd
pd.DataFrame([x foryin df1.values forxin y])

  rating  review_id
05.0487335614.0497335625.0463589233.04645839

Or if willing to use itertools:

from itertools import chain

pd.DataFrame(chain.from_iterable(df1.values.ravel()))

Solution 2:

1st unnesting , then re build your dataframe (assuming you have columns name 0)

pd.DataFrame(unnesting(df,[0])[0].values.tolist())
Out[61]: 
  rating  review_id
05.0487335614.0497335625.0463589233.04645839

---

defunnesting(df, explode):
    idx=df.index.repeat(df[explode[0]].str.len())
    df1=pd.concat([pd.DataFrame({x:np.concatenate(df[x].values)} )for x in explode],axis=1)
    df1.index=idx
    return df1.join(df.drop(explode,1),how='left')