###################################################################################33
# allTweets_file = open('tweets_negative.txt', 'r')
# allTweetsList = allTweets_file.readlines()
# i = 0
# for tweet in allTweetsList:
# sentence = Sentence(tweet)
# classifier.predict(sentence)
# if(sentence.labels[0].value == 'POSITIVE'):
# score = sentence.labels[0].score
# else:
# score = 0 - sentence.labels[0].score
# allTweetsFeatures[i].append(score)
# i += 1
for tweet in posTweetsFeatures:
print(cosineSimilarity(tweet, positiveFeatures))
#######################################################################################3
#negTweets_afteraugment = augment('tweets_negative.txt', 'tweets_augment_neg.txt')
#negTweets_afteraugment = 'tweets_augment_neg.txt'
#negTweets_afterlemma = lemm(negTweets_afteraugment, 'tweets_augment_lemma_neg.txt')
#print('6')
#negTweetsFeatures = np.genfromtxt(fasttext(negTweets_afterlemma))
negTweetsFeatures = np.genfromtxt(fasttext('tweets_negative.txt'))
negTweetsFeatures = negTweetsFeatures.tolist()
for tweet in negTweetsFeatures:
print(cosineSimilarity(tweet, positiveFeatures))
aux_emotion_embeddings = get_emotion_embeddings('currentIterTweets.txt')
i = 0
for tweetFeature in aux_emotion_embeddings:
for val in tweetFeature:
currentIterFeat_withemotion[i].append(val)
i+=1
iterationLabels = []
iterationTweets = []
iterationTweets_withemotion = []
positiveFeat = []
negativeFeat = []
count = 0
for tweetFeature in currentIterFeat:
meancosineSim = cosineSimilarity(tweetFeature, positiveTrainingFeatures)
#print(meancosineSim)
#print(tweetFeature)
#print(meancosineSim)
if(meancosineSim > minSimilarityThreshold):
iterationLabels.append(1.0)
iterationTweets.append(tweetFeature)
iterationTweets_withemotion.append(currentIterFeat_withemotion[count])
positiveFeat.append(tweetFeature)
logger.info(str(1) + ' ' + str(meancosineSim) + ' ' + currentIterTweets[count])
#print('yes')
#print(iterationLabels)
elif meancosineSim > minSimilarityThreshold - similarityWindowSize:
#print('pass')
logger.info(str(-1) + ' ' + str(meancosineSim) +' ' + currentIterTweets[count])
pass
fout.close()
#log and write to auxTweets
#auxiliaryDataFile = fasttext('auxTweets.txt')
#currentIterFeat = np.genfromtxt(auxiliaryDataFile)
currentIterFeat = bert('auxTweets.txt', document_embeddings)
iterationLabels = []
iterationTweets = []
#iterationTweets_bert = []
positiveFeat = []
negativeFeat = []
count = 0
for tweetFeature in currentIterFeat:
poscosineSim = cosineSimilarity(tweetFeature, positiveTrainingFeatures)
negcosineSim = cosineSimilarity(tweetFeature, negativeTrainingFeatures)
#print(meancosineSim)
#print(tweetFeature)
#print(meancosineSim)
# if abs(poscosineSim - negcosineSim) < 0.05:
# #print('pass')
# logger.info(str(-1) + ' ' + str(meancosineSim) +' ' + currentIterTweets[count])
# pass
if poscosineSim > negcosineSim:
iterationLabels.append(1.0)
iterationTweets.append(tweetFeature)
#iterationTweets_bert.append(currentIterFeat_bert[count])
positiveFeat.append(tweetFeature)
logger.info(str(1) + ' ' + str(poscosineSim) + ' ' + str(negcosineSim) + ' ' + currentIterTweets[count])
#print('yes')