CODEWITHSUNDEEP
pythonjsonpandasjson-normalize
Krystof
Krystof

Reputation: 315

Converting nested JSON to flattened Pandas Dataframe

I know there has been already several questions on this topic, but I can't find the answer.

I have a nested JSON, let's say:

data = {'items': [{'block': 1,
   'log_events': [{'decoded': {'name': 'Transfer'},
     'params': ['param0', 'param1', 'param2']}]},
  {'block': 2,
   'log_events': [{'decoded': {'name': 'SafeReceived'},
     'params': ['param3', 'param4', 'param5']}]}]}

I tried:

pd.json_normalize(
    data=data['items'],
    record_path=['log_events'],
    meta=['block']
)

But I would like to "flatten" both on the 'log_events' and on the 'params'. How can I do that ?

Upvotes: 0

Views: 40

Answers (1)

jezrael
jezrael

Reputation: 862491

IIUC use:

df = pd.json_normalize(
    data=data['items'],
    record_path=['log_events'],
    meta=['block']
).explode('params').reset_index(drop=True)
print (df)
   params  decoded.name block
0  param0      Transfer     1
1  param1      Transfer     1
2  param2      Transfer     1
3  param3  SafeReceived     2
4  param4  SafeReceived     2
5  param5  SafeReceived     2

Upvotes: 1

Related Questions