将有序字典转换为 Python Pandas 中的新列

Question

In my dataframe, some of the columns are in OrderedDictionary format. How could I convert them into new columns (without knowing which columns contain OrderedDictionary and the elements in OrderedDictionary)

Example column:

inventors
[OrderedDict([('@sequence', '001'), ('@app-type', 'applicant'), ('@designation', 'us-only'), ('addressbook', OrderedDict([('last-name', 'Nahm'), ('first-name', 'Seung Hoon'), ('address', OrderedDict([('city', 'Daejeon'), ('country', 'KR')]))])), ('residence', OrderedDict([('country', 'KR')]))]), OrderedDict([('@sequence', '002'), ('@app-type', 'applicant'), ('@designation', 'us-only'), ('addressbook', OrderedDict([('last-name', 'Jang'), ('first-name', 'Hoon Sik'), ('address', OrderedDict([('city', 'Daegu'), ('country', 'KR')]))])), ('residence', OrderedDict([('country', 'KR')]))])]

I want to convert it to the following dataframe (did not write all of the columns):

@sequence1  @app_type1   @designation1 @last_name1 @first_name1 ....
001         applicant    us_only        Nahm       Seung Hoon

In this example, last_name and first_name is coming from another nested dictionary. And in the data, I don't know which columns contain OrderedDictonary, for the sake of the simplicty, I just included one column from the dataset which is inventors

Answer 1

Have you considered the pandas from_dict() function?

# Create example dict
import collections
inventors = collections.OrderedDict()
inventors['@sequence'] =  "001"
inventors['@app-type'] =  "applicant"
inventors['@designation'] =  "us-only"
inventors['last-name'] =  "Nahm"
inventors['first-name'] =  "Seung Hoon"

# Import pandas to use from_dict function
import pandas as pd

# Use from_dict() function; include orient='index' for now to avoid index error
df = pd.DataFrame.from_dict(inventors, orient='index')

# Transpose for final output
df.T