如何从包含 json 的文件创建新的 pandas dataframe 列？

Question

I have a dataset with metadata about voice calls

It looks like

Data columns (total 4 columns):
 #   Column       Non-Null Count  Dtype 
---  ------       --------------  ----- 
 0   phone        100 non-null    string
 1   group_id     100 non-null    int64 
 2   question_id  100 non-null    int64 
 3   result       100 non-null    bool 

I want to create column #4 that will contain some dialogue statistics, eg total_words (int64) , and data must be taken from exteral json file, containing speech-to-text recognition results

Is there any build-in pandas way to do that?

I've tested with pandas.read_json but getting module errors ( ValueError: DataFrame constructor not properly called! and TypeError: argument of type 'method' is not iterable )

I'm looking for something like

df['total_words'] = pd.read_json('file://localhost:8888/auido/' + df['phone'] + '.mp3.wstat.json')

I would be glad if someone will provide a working code example of solving similar issue

UPD: output of json file is like {"total_words": 74}

Answer 1

Try this:

df['total_words']=df['phone'].apply(lambda x: pd.read_json('file://localhost:8888/auido/' + x + '.mp3.wstat.json'))

Now each cell of total_words column contains another dataframe, you can access it using:

#for first row
df.iloc[0]["total_words"].head()