forked from vedantm314/text-analytics-for-reconnaissance
-
Notifications
You must be signed in to change notification settings - Fork 0
/
earthquake_data.py
209 lines (161 loc) · 8.94 KB
/
earthquake_data.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
import os
import time
import requests
import tweepy as tw
import pandas as pd
from datetime import date, timedelta
import configparser
from newsapi import NewsApiClient
class EarthquakeDataCollector:
config = configparser.ConfigParser()
config.read("config/config.ini")
VERBOSE = config.getboolean("default", "verbose")
DATA_DIR = config["default"]["data_dir"]
DURATION = config["default"]["duration"]
MIN_MAGNITUDE = int(config["default"]["min_magnitude"])
ALERT_LEVELS = config["default"]["alert_levels"].split(",")
DATE = config["default"]["date"]
MAX_TWEETS = int(config["default"]["max_tweets"])
MAX_PAGE = int(config["default"]["max_page"])
def __init__(self):
print("initialized")
def limit_handled(cursor):
"""Handle twitter rate limit."""
while True:
try:
yield cursor.next()
except tw.RateLimitError:
time.sleep(1 * 60)
def get_tweet(self, keys, search_words, date=str(date.today()), max_item=1000, geocode=None):
consumer_key, consumer_secret, access_token, access_token_secret = keys
auth = tw.OAuthHandler(consumer_key, consumer_secret)
auth.set_access_token(access_token, access_token_secret)
api = tw.API(auth, wait_on_rate_limit=True)
# Define the search term and date
search_words = search_words + " -filter:retweets"
# Collect tweets
tweets = tw.Cursor(api.search, q=search_words, geocode=geocode, since=date, tweet_mode='extended').items(10)
tweet_data = [[tweet.user.screen_name, tweet.user.location, tweet.created_at,
tweet.coordinates, tweet.full_text] for tweet in tweets]
tweet_df = pd.DataFrame(data=tweet_data, columns=['user', "location", "date", "coordinates", "content"])
return tweet_df
def get_news(self, api_key, news_search_words, from_date=str(date.today()), page=2, lang="en"):
newsapi = NewsApiClient(api_key=api_key)
# endpoint: /v2/everything
all_articles = newsapi.get_everything(q=news_search_words,
from_param=from_date,
language=lang,
sort_by='relevancy',
page=page) # each page is 20 articles
keywords = {"quake", "earthquake", "tremors"}
articles = []
for article in all_articles['articles']:
if keywords.intersection(article['title'].split()):
articles.append([article['source']['name'], article['title'], article['url'],
article['publishedAt'][:10], article['content']])
news = pd.DataFrame(data=articles, columns=['source_name', "title", "url", "date", "content"])
return news
def log_earthquakes(self, earthquakes, log_file):
events = []
for event in earthquakes:
properties = event["properties"]
coordinates = event["geometry"]["coordinates"]
rupture_time = time.strftime('%Y-%m-%d %H:%M:%S', time.localtime(properties["time"] / 1000))
country = properties["place"].split(", ")[-1] # get earthquake country
events.append([properties["place"], country, properties["mag"], rupture_time,
properties["felt"], properties["cdi"], properties["mmi"], properties["alert"],
coordinates[0], coordinates[1], coordinates[2], properties["url"], True])
log = pd.read_csv(log_file)
new_log = pd.DataFrame(events, columns=log.columns)
log = pd.concat([log, new_log], ignore_index=True)
log.drop_duplicates(subset=['place', 'country', 'magnitude', 'rupture_time']
, keep='first', inplace=True)
log.to_csv(log_file, index=False)
def runDailyUpdate(self):
keys_parser = configparser.ConfigParser()
keys_parser.read('config/keys.ini')
twitter_keys = [keys_parser['tweepyapi_key']['consumer_key'],
keys_parser['tweepyapi_key']['consumer_secret'],
keys_parser['tweepyapi_key']['access_token'],
keys_parser['tweepyapi_key']['access_token_secret']
]
news_key = keys_parser['newsapi_key']['api_key']
# Query USGS and get social media and news data
today = date.today()
print("Today is: ", today)
# Yesterday date
yesterday = today - timedelta(days=67)
print("Query date: ", yesterday)
log_file = self.runDailyUSGSQuery(yesterday)
self.runTweetAndNewsCollection(log_file, yesterday, twitter_keys, news_key)
def runDailyUSGSQuery(self, query_date):
# Create earthquake log if not existed
log_file = os.path.join(os.getcwd(), self.DATA_DIR, "earthquakes_log.csv")
log_path = os.path.join(os.getcwd(), self.DATA_DIR)
if (not os.path.exists(log_path)):
os.mkdir(log_path)
if not os.path.isfile(log_file):
log = pd.DataFrame(columns=['place', 'country', 'magnitude', 'rupture_time',
'felt', 'cdi', 'mmi', 'alert',
'longitude', 'latitude', 'depth', 'url', 'collect_data'])
log.to_csv(log_file, index=False)
# Query earthquake with magnitude >= 5 and alert level yellow, orange, red
for alert in self.ALERT_LEVELS:
earthquakes = self.get_usgs(self.MIN_MAGNITUDE, alert, query_date)
if len(earthquakes) > 0:
self.log_earthquakes(earthquakes, log_file)
return log_file
def runTweetAndNewsCollection(self, log_file, query_date, twitter_keys, news_key):
log = pd.read_csv(log_file)
# Change collect_data flag to False for events longer than DURATION
duration = pd.to_datetime(query_date.today()) - pd.to_datetime(log.rupture_time)
log.loc[duration > pd.to_timedelta(self.DURATION), "collect_data"] = False
# Collect social media and news data for recent earthquakes within DURATION
events = log[log.collect_data == True]
earthquake_tweet = {}
earthquake_news = {}
for rupture_time, country in zip(events.rupture_time, events.country):
rupture_date = rupture_time.split()[0]
# Collect tweets
twitter_search_words = country + "+earthquake"
tweet_df = self.get_tweet(twitter_keys, twitter_search_words, query_date, self.MAX_TWEETS)
earthquake_tweet[(rupture_date, country)] = tweet_df
# Collect news
news_search_words = '+("earthquake" AND "{}")'.format(country)
news_df = self.get_news(news_key, news_search_words, query_date, self.MAX_PAGE)
earthquake_news[(rupture_date, country)] = news_df
# Store tweet and news to csv file
for (rupture_date, country) in earthquake_tweet.keys():
tweet_file = os.path.join(os.getcwd(), self.DATA_DIR, "twitter",
"{}_{}_earthquake.csv".format(rupture_date, country))
news_file = os.path.join(os.getcwd(), self.DATA_DIR, "news",
"{}_{}_earthquake.csv".format(rupture_date, country))
# Create new csv if file not exited; otherwise append to existing file
if not os.path.isfile(tweet_file):
# os.mkdir(os.path.join(os.getcwd(), DATA_DIR, "twitter"))
earthquake_tweet[(rupture_date, country)].to_csv(tweet_file, index=False)
else:
current_df = pd.read_csv(tweet_file)
new_df = earthquake_tweet[(rupture_date, country)]
tweet_df = pd.concat([current_df, new_df])
tweet_df.to_csv(tweet_file, index=False)
if not os.path.isfile(news_file):
# os.mkdir(os.path.join(os.getcwd(), DATA_DIR, "news"))
earthquake_news[(rupture_date, country)].to_csv(news_file, index=False)
else:
current_df = pd.read_csv(news_file)
new_df = earthquake_news[(rupture_date, country)]
news_df = pd.concat([current_df, new_df])
news_df.to_csv(news_file, index=False)
def get_usgs(self, min_magnitude, alert, date):
parameters = {"minmagnitude": min_magnitude,
"alertlevel": alert,
"starttime": date}
URL = "https://earthquake.usgs.gov/fdsnws/event/1/query?format=geojson&"
# Format url parameters: https://earthquake.usgs.gov/fdsnws/event/1/#parameters
url_params = "&".join(["{}={}".format(key, value) for key, value in parameters.items()])
URL += url_params
response = requests.get(URL)
earthquakes = response.json()['features']
if self.VERBOSE: print("{0} {1} earthquake today.".format(len(earthquakes), alert))
return earthquakes