尝试使用 groupby() 函数但不断收到“类型错误：列表索引必须是整数或切片，而不是 str”

Question

edit: i added the "columns" in front of the "event_id" because i kept getting "keyerror: "event_id" when i tried the code with just 'event_id' alone before. i took the "columns" code off now, but am still getting the same keyerror; i checked the "event_id" and it is recognized as a column by python....any suggestions??

i'm trying to aggregate df_userpolice into categories according to "event_id" and then aggregate all the other numbers for each event_id (average follower count for each event_id, etc.); then i need to merge it with the smaller dataframe df_eventpolice. i've changed every row in event_id into integer in excel, but it's still not working for some reason, this is my code:

import pandas as pd
import dateutil
df_userpolice = pd.read_csv(filepath_or_buffer='userpolice.csv', error_bad_lines=False)
df_eventpolice = pd.read_csv(filepath_or_buffer='eventpolice.csv', index_col = 0)
columns = ['event_id', 'city_indiv', 'post_id_indiv', 'content_indiv', 'content_media', 'is_same_event', 'post_id_media', 'prov_code', 'date_indiv', 'geolocation', 'issue_type_indiv', 'followers_count', 'fan_count', 'gender', 'status_count', 'issue_type_words_indiv',  'action_form_indiv', 'action_form_words_indiv', 'username', 'city_media', 'uid', 'verified', 'self_description', 'verified_type', 'refined', 'date_media', 'issue_type_media', 'issue_type_words_media', 'action_form_media', 'action_form_words_media']
print(df_userpolice)

for row in df_userpolice:
  print(row)
for row in df_eventpolice:
  print(row)

df_userpolice.groupby['event_id'].groups.keys() <------this is where the error happens

but i keep getting the following error:

---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
<ipython-input-38-75f25f3b87eb> in <module>()
     15 # Drop NA values, listing the converted columns explicitly
     16 #   so NA values in other columns aren't dropped
---> 17 df.dropna(subset = ['event_id'])
     18 
     19 

/usr/local/lib/python3.6/dist-packages/pandas/core/frame.py in dropna(self, axis, how, thresh, subset, inplace)
   4746             check = indices == -1
   4747             if check.any():
-> 4748                 raise KeyError(list(np.compress(check, subset)))
   4749             agg_obj = self.take(indices, axis=agg_axis)
   4750 

KeyError: ['event_id']

i really don't know where i'm going wrong.

Answer 1

The problem is with "columns['event_id']".

Here the columns type is list and the list items can be accessed by its indexes but not like columns['event_id'].

I hope your intention here is to create a dict "columns", but the way you declared the "columns" is type list.

If you want the "columns" to be a dict, just try something like this.

columns = {'city_indiv': 'city_name', 'post_id_indiv': 'post_id', 'content_indiv':'content_of_indiv', 'content_media':'content_of_media'}.

Then you can perform the key value operations on the columns as it is a dict type.

Answer 2

使用df_userpolice.groupby("event_id").size()获取每个event_id的行数