Calculate the Top N statistics from a dataframe series and the remaining info is grouped as 'others'

top_n_others.py

def top_n_dataframe(n,dataframe_field):
    
    top_n = n
    field_name = dataframe_field.name
    top  = dataframe_field.value_counts()[:top_n].to_frame().reset_index()

    new_row = pd.DataFrame(data = {
        'hits' : [ dataframe_field.value_counts()[top_n:].sum()],
        field_name : ['others'],
    })

    top.columns = [field_name, 'hits']
    top.set_index([field_name]).reset_index()
    top_result = pd.concat([top, new_row])
    
    # percentage field
    df = top_result.groupby(field_name).sum() 
    df=df.sort_values(by="hits", ascending=False)
    df['percent'] = df.transform(lambda x: (x/np.sum(x)*100)) 
    
    return (df)