return temp
trueIrrelevants = []
possibleRelevants = []
with open('cleaned_geo_tweets_Apr_12_to_22.csv') as csvfile:
tweetData = csv.DictReader(csvfile)
for tweet in tweetData:
if tweet['time'] != "":
# parse date/time into object
date = time.strptime(tweet['time'], tweet_time_fmt)
tweet['tweet_text'] = twc.cleanUpTweet(tweet['tweet_text'])
if date.tm_mday < 15:
trueIrrelevants.append(tweet['tweet_text'])
elif twc.tweetContainsKeyword(tweet['tweet_text']):
possibleRelevants.append(tweet['tweet_text'])
trueIrrelevants = randomSubset(trueIrrelevants)
possibleRelevants = randomSubset(possibleRelevants)
trueRelevants = []
for each in possibleRelevants:
print each
result = raw_input("Enter a r for relevant, i for irrelevant, n for neither (not English): ")
result = result.lower()
if result != '':
if result[0] == 'i':
trueIrrelevants.append(each)
elif result[0] == 'r':
# sentimentTweets[c] = []
# tweetList = []
# textList = []
with open('cleaned_geo_tweets_4_12_22.csv') as csvfile:
# reads first line of csv to determine keys for the tweet hash, tweets
# is an iterator through the list of tweet hashes the DictReader makes
tweets = csv.DictReader(csvfile)
# for all the tweets the reader finds
for tweetData in tweets:
# make sure its not a 'false tweet' from people using newlines in their tweet_text's
if tweetData['time'] != "":
# parse date/time into object
date = time.strptime(tweetData['time'], tweet_time_fmt)
if date.tm_mday == 15 and twc.tweetContainsKeyword(tweetData['tweet_text']):
#if date.tm_mday == 15:
if date.tm_hour == currentHour:
kwTweets.append(tweetData)
#tweetList.append(tweetData)
#textList.append(tweetData['tweet_text'])
if containsHandle(tweetData['tweet_text']):
infoTweets.append(tweetData)
elif date.tm_hour == currentHour + 1:
currentHour += 1
timeStr = getTimeString(currentHour)
# results = clssfr.classify(textList)
# for i in range(0, len(results)):
# sentimentTweets[cats[results[i]]].append(tweetList[i])
# for sentiment in sentimentTweets.keys():
with open('test_tweets_4_12_22.csv') as csvfile:
# reads first line of csv to determine keys for the tweet hash, tweets
# is an iterator through the list of tweet hashes the DictReader makes
tweets = csv.DictReader(csvfile)
# for all the tweets the reader finds
for tweetData in tweets:
# make sure its not a 'false tweet' from people using newlines in their tweet_text's
if tweetData['time'] != "":
# parse date/time into object
date = time.strptime(tweetData['time'], tweet_time_fmt)
#if date.tm_mday == 15 and twc.tweetContainsKeyword(tweetData['tweet_text']):
if date.tm_mday == 12:
count2 += 1
if date.tm_hour == currentHour:
if twc.tweetContainsKeyword(tweetData['tweet_text'].lower()):
kwTweets.append(tweetData)
if containsHandle(tweetData['tweet_text']):
infoTweets.append(tweetData)
tweetList.append(tweetData)
textList.append(tweetData['tweet_text'])
elif date.tm_hour == currentHour + 1:
currentHour += 1
timeStr = getTimeString(currentHour)
results = relClssfr.classify(textList)
for i in range(0, len(results)):
if results[i] == 0:
relTweets.append(tweetList[i])
if containsHandle(textList[i]):
relInfo.append(tweetList[i])
import twittercriteria as twc
import matplotlib.pyplot as plt
# author: Hayden Fuss
handles = {}
senders = {}
with open('cleaned_geo_tweets_4_12_22.csv') as csvfile:
twitterData = csv.DictReader(csvfile)
for tweet in twitterData:
if tweet['time'] != "":
date = twc.getTweetDate(tweet['time'])
if date.tm_mday > 15 or (date.tm_mday == 15 and date.tm_hour >= 14):
if twc.tweetContainsKeyword(tweet['tweet_text'].lower()):
if not tweet['sender_name'] in senders.keys():
senders[tweet['sender_name']] = 1
else:
senders[tweet['sender_name']] += 1
results = twc.getHandlesFromTweet(tweet['tweet_text'])
for r in results:
handle = r.strip("@").lower()
if not handle in handles.keys():
handles[handle] = {'senders':[], 'count':1}
handles[handle]['senders'].append(tweet['sender_name'])
else:
if not tweet['sender_name'] in handles[handle]['senders']:
handles[handle]['count'] += 1
handles[handle]['senders'].append(tweet['sender_name'])
