Error 'str' object has no attribute 'values' pops up when using json_normalize

Question

I`ve got the DataFrame with nested json inside the extra_data column

when I try to use pd.json_normalize(transactions['extra_data']) I get a 'str' object has no attribute 'values' error. Is there any way to solve this? The dataframe is taken straight from the database, so that`s not a csv file or something

please, check [ask], show [mre] and also check https://stackoverflow.com/q/20109391/4046632 — buran, Sep 22 '21 at 05:33
Most likely `extra_data` is `str`, not deserialized JSON. try `df['extra_data'] = df['extra_data'].apply(json.load)` — buran, Sep 22 '21 at 05:35
Thank you! The following code worked just fine `df['extra_data'] = df['extra_data'].apply(json.dumps) df['extra_data'] = df['extra_data'].apply(json.loads) df = pd.json_normalize(df['extra_data'])` — Alexey, Sep 22 '21 at 05:48
And I apologize for poor formatting, that`s my first question on stackoverflow — Alexey, Sep 22 '21 at 05:52

score 0 · Answer 1 · answered Sep 22 '21 at 06:09

0

Most likely values in extra_data column are str, not deserialized JSON. try

import json

df['extra_data']  = df['extra_data'].apply(json.load)
df = pd.json_normalize(df['extra_data'])

answered Sep 22 '21 at 06:09

buran

13,682
10
36
61

Error 'str' object has no attribute 'values' pops up when using json_normalize

1 Answers1