def linearRegressionObject():
AnalysisObjectFactory.initialFactory()
types = AnalysisObjectFactory.createObject("LinearRegressionObject","hillary clinton",["2/17/2016","2/18/2016","2/19/2016","2/20/2016","2/21/2016","2/22/2016","2/23/2016"],("GDELT",[1,2,3,7,4,3,10]),("Tweet",[20,10,15,20,5,17,21]))
(text,data)= types.linearRegressionAnalysis()
print text
print data
return 0
def regressionObject():
fake_tweets = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9])
#change these values to plot different points
fake_polls = np.array([3, 2, 5, 0, 4, 10, 7, 0, 5])
AnalysisObjectFactory.initialFactory()
types = AnalysisObjectFactory.createObject("RegressionObject", "Bernie", "Dates I guess", fake_tweets, fake_polls)
types.Interpolate()
def main():
dic = QueryingDavid.candidateCountRangeHourlyDict("trump","2016041100","2016041700")
ordered_dic = collections.OrderedDict(sorted(dic.items()))
dates =[]
count =[]
for item in ordered_dic:
dates.append(item)
count.append(float(ordered_dic[item]))
AnalysisObjectFactory.initialFactory()
types = AnalysisObjectFactory.createObject("TimeSeriesObject","trump","Twitter",dates,count)
text= types.timeSeriesAnalysis()
return 0
def main():
candidate = sys.argv[1]
start = sys.argv[2]
end = sys.argv[3]
tweets, dates = getTweets(candidate, start, end)
gdelt = getGDELT(candidate, start, end)
print 'Tweets: ' + str(tweets)
print 'GDELT: ' + str(gdelt)
types = AnalysisObjectFactory.createObject("RegressionObject", candidate, dates, gdelt, tweets)
def main():
buildModel(300000)
AnalysisObjectFactory.initialFactory()
list_of_userids = []
list_of_list_of_tweets = []
list_of_candidates = ["trump"]
for candidate in list_of_candidates:
list_of_userids = []
list_of_list_of_tweets = []
final_result = session.execute("select * from topTweeters where candidate_name = '" + candidate + "'")
for user in final_result:
list_of_userids.append(str(getattr(user, "user_id")))
list_of_list_of_tweets.append(getattr(user, "list_of_tweets"))
types = AnalysisObjectFactory.createObject(
"TweetAnalWord2Vec", candidate, ["3/6/2016", "4/5/2016"], list_of_userids, list_of_list_of_tweets
)
# types = AnalysisObjectFactory.createObject("TweetAnalWord2Vec","trump",["2/17/2016","2/18/2016","2/19/2016","2/20/2016","2/21/2016","2/22/2016","2/23/2016"],['123','234'],[["Trump is a motherfucking Genius! Ban all Muslims is the best thing ever!","RT @kateloving: KING: Bill Clinton almost sounds like Donald Trump with jab https://t.co/5a2ze5YTc4","This is an exercise in taking Donald Trump at his word"," @GlobeIdeas editor @katiekings... https://t.co/vEa1mhKVDz via @jorgeramosnews"],["The TRUMP Fight Song [unofficial] - TRUMP 2016 https://t.co/9AnOeQV83V via @YouTube","Cheats abd liars in GOPe - time to write in Trump! #StopTheSteal help Trump #BurnDownTheGOP https://t.co/3z8qY8IW5Q","Trump nomination up to 200 people","Trump nomination up to 200 people"]])
types.sentimentAnalysis()
return 0
def plotGDELT(candidate, start, end):
GDELT_counts, dates = GetDailyGDELTCounts.getGDELTCounts(candidate, start, end)
#change Nones to 0's
i = 0
for count in GDELT_counts:
if count == None:
GDELT_counts[i] = 0
i += 1
#normalize counts
types = AnalysisObjectFactory.createObject("GraphAnalysisNormalized", candidate, "twitter", 'dates go here', GDELT_counts)
GDELT_counts = types.getlist_para_two()
print 'Normalized: ' + str(GDELT_counts) + '\n'
#generate x-axis
x = list(range(0, len(GDELT_counts)))
line, = plt.plot(GDELT_counts, 'b')
global legend
legend[0].append(line)
legend[1].append('GDELT')
return GDELT_counts
def plotTwitter(candidate, start, end):
twitter_counts, dates = GetHourlyTweetsDayRange.hourlyTweets(candidate, start, end)
#change Nones to 0's
i = 0
for count in twitter_counts:
if count == None:
twitter_counts[i] = 0
i += 1
#turn hourly counts into daily counts
daily_counts = []
i = 0
while i < len(twitter_counts):
daily_counts.append(sum(twitter_counts[i:i+23]))
i += 24
#normalize counts
types = AnalysisObjectFactory.createObject("GraphAnalysisNormalized", candidate, "twitter", 'dates go here', daily_counts)
daily_counts = types.getlist_para_two()
print 'Normalized: ' + str(daily_counts) + '\n'
num_counts = len(daily_counts)
#generate x-axis
x = list(range(0, len(daily_counts)))
line, = plt.plot(daily_counts, 'r')
plt.xticks(x, dates, rotation = 'vertical')
plt.xlim(0, num_counts)
plt.gcf().subplots_adjust(bottom=0.20)
plt.title('Daily tweet counts and article counts for ' + candidate)
global legend
legend[0].append(line)
legend[1].append('Tweets')
return daily_counts
from AnalysisObjectFactory import AnalysisObjectFactory
import numpy as np
from sklearn import linear_model
import matplotlib as mpl
mpl.use('Agg')
import matplotlib.pyplot as plt
import pandas as pd
import GetHourlyTweetsDayRange
from RegressionAnalysis import RegressionObject
import GetDailyGDELTCounts
import sys
AnalysisObjectFactory.initialFactory()
def getTweets(candidate, start, end):
#get the tweets
twitter_counts, dates = GetHourlyTweetsDayRange.hourlyTweets(candidate, start, end)
#change Nones to 0's
i = 0
for count in twitter_counts:
if count == None:
twitter_counts[i] = 0
i += 1
#turn hourly counts into daily counts
daily_counts = []
i = 0
while i < len(twitter_counts):
daily_counts.append(sum(twitter_counts[i:i+23]))
i += 24
def timeSeriesObject():
AnalysisObjectFactory.initialFactory()
types = AnalysisObjectFactory.createObject("TimeSeriesObject", "Trump", "Date", [1, 2, 3, 4, 5, 6, 7, 8], [10, 8, 4, 6, 7, 5, 9, 2])
print types.getlist_para_two()
def graphAnalysis():
AnalysisObjectFactory.initialFactory()
types = AnalysisObjectFactory.createObject("GraphAnalysisNormalized","hilary clinton","gdelt",["2/23/2016","2/22/2016"],[15,23])
print types.getlist_para_two()
def classificationAnalysis():
AnalysisObjectFactory.initialFactory()
types = AnalysisObjectFactory.createObject("ClassificationAnalysis","2/23/2016","gdelt",["hilary clinton","donaldtrump","ted cruz"],[9,8,2])
print types.getlist_para_two()