def get_rate():
#Default premium add is
total_premium_add = 15
posts = ''
plan_code = request.args.get('plancode')
plan_details = get_plan_name(plan_code)
plan_name = plan_details[0]
income_addition_rate = get_income_addition(session['income'])
_base_rate = session['base_rate'] + plan_details[1] + income_addition_rate
if (session['analyse_fb'] == True):
fb_text = get_fb_posts()
posts = fb_text
if (session['analyse_twitter'] == True):
twitter_posts = get_all_tweets(session['twitter_handle'])
twitter_text = u". ".join(twitter_posts)
posts = posts + '. ' + twitter_text
if(session['analyse_twitter'] == True or session['analyse_fb'] == True):
total_premium_add = get_deltas(posts)
else:
result_list = []
result_list.append({'attribute' : 'Alcohol', 'sentiment' : 'Neutral', 'relevance' : 0, 'delta' : 0, 'factor' : 0.5})
result_list.append({'attribute' : 'Drugs', 'sentiment' : 'Neutral', 'relevance' : 0, 'delta' : 0, 'factor' : 0.62} )
result_list.append({'attribute' : 'Smoking', 'sentiment' : 'Neutral', 'relevance' : 0, 'delta' : 0, 'factor' : 0.7} )
result_list.append({'attribute' : 'Lifestyle', 'sentiment' : 'Neutral', 'relevance' : 0, 'delta' : 0, 'factor' : 0.34})
result_list.append({'attribute' : 'Healthy', 'sentiment' : 'Neutral', 'relevance' : 0, 'delta' : 0, 'factor' : 1.0})
session['result_list'] = result_list
#Base rate should not be lowered
if(total_premium_add < 0):
total_premium_add = 0
#final_rate = session['base_rate'] + total_premium_add + income_addition_rate
return flask.render_template("viewPlanDetails.html", \
base_rate = _base_rate, delta = total_premium_add, \
result = session['result_list'], \
code = plan_code, name = plan_name)
p_l2 = model.predict(train_matrix)
print('Accuracy of train predictions:',
accuracy_score(train_labels, p_l2) * 100)
test_matrix, test_labels = extract_features(test_dir, vocab_size,
tweets_per_file)
# Predicting on test data
predicted_labels = model.predict(test_matrix)
print('Accuracy of test predictions:',
accuracy_score(test_labels, predicted_labels) * 100)
#making prediction on new data
answer = 1
while answer:
num_tweets = get_all_tweets()
to_predict_matrix, to_predict_tweets = extract_features(
predict_dir, vocab_size, num_tweets)
prediction_labels_new_tweets = model.predict(to_predict_matrix)
pos = 0
neg = 0
for label in prediction_labels_new_tweets:
if label == 0:
neg += 1
elif label == 4:
pos += 1
else:
print("Error")
print("Percentage of last", num_tweets, "tweets that were positive:",
pos / num_tweets)
print("Percentage of last", num_tweets, "tweets that were negative:",
try:
writer.writerow('{}{}{}'.format(result))
del(new_collect_1, result[0])
except Exception as e:
pass
pass
#_acct is a list with [tweet.id_str, tweet.created_at, tweet_content] format
# for index, sa in enumerate(second_acct):
# print("Start comparing tweet " + str(index) + " at " + '{%H:%M:%S}'.format(datetime.datetime.now()))
# results = helpers.hammingCompare(first_acct, sa[2])
# print("Finish comparing tweet " + str(index) + " at " + '{%H:%M:%S}'.format(datetime.datetime.now()))
# #write each row
# for result in results:
# try:
# writer.writerow('{}{}{}'.format(result))
# del(first_acct, result[0])
# except Exception as e:
# pass
if __name__ == '__main__':
second_acct = get_tweets.get_all_tweets(compareName.secondAccount)
first_acct = get_tweets.get_all_tweets(compareName.firstAccount)
if (not type(second_acct[0][2]) is bool) and (not type(first_acct[0][2]) is bool):
main_compare(first_acct, second_acct)
print("Please check compareName_tweets.csv for final result")
else:
print(str(second_acct[0][1]) + " " + str(first_acct[0][1]))