How to split a pandas column with a list of dicts into separate columns for each key

user_12461913 0
Viewed 2151

I'm analyzing Political Advertisements from Facebook, which is a dataset released here, by ProPublica.

Here's what I mean. I have an entire column of targets that I want to analyze but it's formatted in a very inaccessible way for someone of my skill level.

This is from just 1 cell: [{"target": "NAge", "segment": "21 and older"}, {"target": "MinAge", "segment": "21"}, {"target": "Retargeting", "segment": "people who may be similar to their customers"}, {"target": "Region", "segment": "the United States"}]

and another: [{"target": "NAge", "segment": "18 and older"}, {"target": "Location Type", "segment": "HOME"}, {"target": "Interest", "segment": "Hispanic culture"}, {"target": "Interest", "segment": "Republican Party (United States)"}, {"target": "Location Granularity", "segment": "country"}, {"target": "Country", "segment": "the United States"}, {"target": "MinAge", "segment": 18}]

What I need to do is to separate every "target" item to become the column label with each of its corresponding "segment" to be a possible value within that column.

Or, is the solution to create a function to call each dictionary key within each row to count frequency?

json-normalize pandas python
1 Answers
  • The columns are lists of dicts.
    • Each dict in the list can be moved to a separate column by using pandas.explode().
    • Convert the column of dicts to a dataframe where the keys are column headers and the values are observations, by using pandas.json_normalize(), .join() this back to df.
  • Use .drop() to remove the unneeded column.
  • If the column contains list of dicts that are strings (e.g. "[{key: value}]"), refer to this solution in Splitting dictionary/list inside a Pandas Column into Separate Columns, and use:
    • df.col2 = df.col2.apply(literal_eval), with from ast import literal_eval.
import pandas as pd

# create sample dataframe
df = pd.DataFrame({'col1': ['x', 'y'], 'col2': [[{"target": "NAge", "segment": "21 and older"}, {"target": "MinAge", "segment": "21"}, {"target": "Retargeting", "segment": "people who may be similar to their customers"}, {"target": "Region", "segment": "the United States"}], [{"target": "NAge", "segment": "18 and older"}, {"target": "Location Type", "segment": "HOME"}, {"target": "Interest", "segment": "Hispanic culture"}, {"target": "Interest", "segment": "Republican Party (United States)"}, {"target": "Location Granularity", "segment": "country"}, {"target": "Country", "segment": "the United States"}, {"target": "MinAge", "segment": 18}]]})

# display(df)
  col1                                                                                                                                                                                                                                                                                                                                                                                 col2
0    x                                                                                                                                                   [{'target': 'NAge', 'segment': '21 and older'}, {'target': 'MinAge', 'segment': '21'}, {'target': 'Retargeting', 'segment': 'people who may be similar to their customers'}, {'target': 'Region', 'segment': 'the United States'}]
1    y  [{'target': 'NAge', 'segment': '18 and older'}, {'target': 'Location Type', 'segment': 'HOME'}, {'target': 'Interest', 'segment': 'Hispanic culture'}, {'target': 'Interest', 'segment': 'Republican Party (United States)'}, {'target': 'Location Granularity', 'segment': 'country'}, {'target': 'Country', 'segment': 'the United States'}, {'target': 'MinAge', 'segment': 18}]

# use explode to give each dict in a list a separate row
df = df.explode('col2', ignore_index=True)

# normalize the column of dicts, join back to the remaining dataframe columns, and drop the unneeded column
df = df.join(pd.json_normalize(df.col2)).drop(columns=['col2'])

display(df)

   col1                target                                       segment
0     x                  NAge                                  21 and older
1     x                MinAge                                            21
2     x           Retargeting  people who may be similar to their customers
3     x                Region                             the United States
4     y                  NAge                                  18 and older
5     y         Location Type                                          HOME
6     y              Interest                              Hispanic culture
7     y              Interest              Republican Party (United States)
8     y  Location Granularity                                       country
9     y               Country                             the United States
10    y                MinAge                                            18

Get count

  • If the goal is to get the count for each 'target' and associated 'segment'
counts = df.groupby(['target', 'segment']).count()

Updated

  • This update is implemented for the full file
import pandas as pd
from ast import literal_eval

# load the file
df = pd.read_csv('en-US.csv')

# replace NaNs with '[]', otherwise literal_eval will error
df.targets = df.targets.fillna('[]')

# replace null with None, otherwise literal_eval will error
df.targets = df.targets.str.replace('null', 'None')

# convert the strings to lists of dicts
df.targets = df.targets.apply(literal_eval)

# use explode to give each dict in a list a separate row
df = df.explode('targets', ignore_index=True)

# fillna with {} is required for json_normalize
df.targets = df.targets.fillna({i: {} for i in df.index})

# normalize the column of dicts, join back to the remaining dataframe columns, and drop the unneeded column
normalized = pd.json_normalize(df.targets)

# get the counts
counts = normalized.groupby(['target', 'segment']).segment.count().reset_index(name='counts')
user_77588040
Related