-
Notifications
You must be signed in to change notification settings - Fork 0
/
twitter_scrub_test.py
245 lines (221 loc) · 9.32 KB
/
twitter_scrub_test.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sun Aug 25 10:24:32 2019
@author: p
"""
import numpy as np
import pandas as pd
import tweepy
from datetime import datetime
import time
from vaderSentiment.vaderSentiment import SentimentIntensityAnalyzer
class twitter_scrubber():
def __init__(self):
self._twitter_api_tokens_df = pd.read_csv('Twitter_API_df_tokens.csv', index_col=0)
self._api = None
username_index = pd.MultiIndex(levels=[[],[], []], labels=[[],[], []], \
names=['username', 'created_at', 'tweet_id'])
self._user_tweets_df = pd.DataFrame(index=username_index, \
columns=['tweet'])
search_index = pd.MultiIndex(levels=[[],[], []], labels=[[],[], []], \
names=['symbol', 'created_at', 'tweet_id'])
self._user_tweets_df = pd.DataFrame(index=search_index, \
columns=['tweet', 'username'])
sentiment_index = pd.MultiIndex(levels=[[],[]], labels=[[],[]], \
names=['symbol', 'date'])
self._sentiment_tweets_df = pd.DataFrame(index=sentiment_index, \
columns = ['Total_User_Weights', 'Total_Tweet_Weights', \
'Bullish_Count', 'Bearish_Count', \
'Neutral_Count', 'Total_Count', \
'Unweighted_Sentiment', 'Total_Sentiment'])
self._analyzer = SentimentIntensityAnalyzer()
def set_api_app_keys(self, twitter_api_app):
"""
Set access keys and tokens to use the api app specified in call to method
"""
CONSUMER_KEY = self._twitter_api_tokens_df.loc[twitter_api_app, 'CONSUMER_KEY']
CONSUMER_SECRET = self._twitter_api_tokens_df.loc[twitter_api_app, 'CONSUMER_SECRET']
ACCESS_TOKEN = self._twitter_api_tokens_df.loc[twitter_api_app, 'ACCESS_KEY']
ACCESS_TOKEN_SECRET = self._twitter_api_tokens_df.loc[twitter_api_app, 'ACCESS_SECRET']
auth = tweepy.OAuthHandler(CONSUMER_KEY, CONSUMER_SECRET)
auth.set_access_token(ACCESS_TOKEN, ACCESS_TOKEN_SECRET)
self._api = tweepy.API(auth)
def _limit_handled(self, cursor):
while True:
try:
yield cursor.next()
except tweepy.RateLimitError:
time.sleep(2)
def update_tweets_username(self, username, count):
ct = 0
for tweet in self._limit_handled(tweepy.Cursor(self._api.user_timeline, id=username).items()):
if tweet.text.startswith('RT'):
continue
# Remove replies
elif tweet.text.startswith('@'):
continue
else:
self._user_tweets_df.loc[(username, tweet.created_at, tweet.id_str), 'tweet'] = tweet.text
ct += 1
if ct == count:
break
def update_tweets_symbol(self, symbol, date_since, count):
no_rt_keyword = "$" + symbol + " -filter:retweets"
new_tweets = self._limit_handled(tweepy.Cursor(self._api.search,
q=no_rt_keyword,
lang="en",
since=date_since).items(count))
outtweets = [[tweet.id_str, tweet.created_at, \
tweet.text, tweet.user.screen_name, \
tweet.retweet_count, tweet.favorite_count] for tweet in new_tweets]
for tweet in outtweets:
self._user_tweets_df.loc[(symbol, tweet[1], tweet[0]), ['tweet', 'username']] = tweet[2], tweet[3]
if not self._sentiment_tweets_df.index.isin([(symbol, str(tweet[1].date()))]).any():
self._sentiment_tweets_df.loc[(symbol, str(tweet[1].date())), :] = 0
self._sentiment_analyis_tweet(symbol, tweet)
def get_tweets_username(self, username):
return self._user_tweets_df.loc[[username], 'tweet']
def get_tweets_symbol(self, symbols):
return self._user_tweets_df.loc[symbols, :]
def get_all_sentiments(self):
return self._sentiment_tweets_df
def get_symbol_sentiments(self, symbol):
return self._sentiment_tweets_df.xs(symbol)
def _sentiment_analyis_tweet(self, symbol, tweet):
date = str(tweet[1].date())
tweet_text = tweet[2]
screenname = tweet[3]
num_retweets = tweet[4]
num_likes = tweet[5]
sent_score, sentiment = self._parse_tweet(tweet_text)
user_weights = self._calc_user_weight(screenname)
tweet_weights = self._calc_tweet_weight(num_retweets, num_likes)
total_weight = user_weights+tweet_weights
self._sentiment_tweets_df.loc[(symbol, date), \
['Total_User_Weights', 'Total_Tweet_Weights', \
sentiment, "Total_Count", \
'Unweighted_Sentiment', 'Total_Sentiment']] += [user_weights, tweet_weights, \
1, 1, sent_score, sent_score*total_weight]
def _calc_tweet_weight(self, num_rt, num_lks):
retweet_wt = self._calc_retweets_weight(num_rt)
likes_wt = self._calc_likes_weight(num_lks)
tweet_wts = retweet_wt+likes_wt
return tweet_wts/10
def _calc_user_weight(self, username):
user_info = self._api.get_user(id=username)
followers_wt = self._calc_followers_weight(user_info.followers_count)
friends_wt = self._calc_friends_weight(user_info.friends_count)
favourites_wt = self._calc_favourites_weight(user_info.favourites_count)
verified_wt = self._calc_verified_weight(user_info.verified)
statuses_wt = self._calc_statuses_weight(user_info.statuses_count)
user_wts = followers_wt+friends_wt+favourites_wt+verified_wt+statuses_wt
return user_wts/10
def _parse_tweet(self, tweet):
vs = self._analyzer.polarity_scores(tweet)
if vs['compound'] > 0.5:
return vs['compound'], 'Bullish_Count'
elif vs['compound'] < -0.5:
return vs['compound'], 'Bearish_Count'
else:
return 0, "Neutral_Count"
def _calc_followers_weight(self, num_followers):
if num_followers < 0:
weight = 0
elif num_followers < 100:
weight = 0.2
elif num_followers < 500:
weight = 0.4
elif num_followers < 2000:
weight = 0.6
elif num_followers < 10000:
weight = 0.8
else:
weight = 1.0
return weight*3
def _calc_friends_weight(self, num_friends):
if num_friends < 0:
weight = 0
elif num_friends < 50:
weight = 0.2
elif num_friends < 100:
weight = 0.4
elif num_friends < 500:
weight = 0.6
elif num_friends < 1000:
weight = 0.8
else:
weight = 1.0
return weight
def _calc_statuses_weight(self, num_statuses):
if num_statuses < 0:
weight = 0
elif num_statuses < 50:
weight = 0.2
elif num_statuses < 100:
weight = 0.4
elif num_statuses < 500:
weight = 0.6
elif num_statuses < 1000:
weight = 0.8
else:
weight = 1.0
return weight*2
def _calc_favourites_weight(self, num_favourites):
if num_favourites < 0:
weight = 0
elif num_favourites < 50:
weight = 0.2
elif num_favourites < 100:
weight = 0.4
elif num_favourites < 500:
weight = 0.6
elif num_favourites < 1000:
weight = 0.8
else:
weight = 1.0
return weight
def _calc_verified_weight(self, verified_bool):
weight = 3*float(verified_bool)
return weight
def _calc_retweets_weight(self, num_retweets):
if num_retweets < 0:
weight = 0
elif num_retweets < 100:
weight = 0.2
elif num_retweets < 500:
weight = 0.4
elif num_retweets < 2000:
weight = 0.6
elif num_retweets < 10000:
weight = 0.8
else:
weight = 1.0
return weight*5
def _calc_likes_weight(self, num_likes):
if num_likes < 0:
weight = 0
elif num_likes < 100:
weight = 0.2
elif num_likes < 500:
weight = 0.4
elif num_likes < 2000:
weight = 0.6
elif num_likes < 10000:
weight = 0.8
else:
weight = 1.0
return weight*5
if __name__ == '__main__':
tws = twitter_scrubber()
tws.set_api_app_keys('tweepy_test')
tws.update_tweets_username('@realDonaldTrump', 10)
username_tweets = tws.get_tweets_username('@realDonaldTrump')
symbols = ["AAPL", "GE", "FB"]
date_from = "2019-08-20"
for symb in symbols:
tws.update_tweets_symbol(symb, date_from, 10)
symbol_tweets = tws.get_tweets_symbol(symbols)
all_sentiment = tws.get_all_sentiments()
ge_sentiment = tws.get_symbol_sentiments("GE")
aapl_sentiment = tws.get_symbol_sentiments("AAPL")