def prediction():
symbol = request.get_data()
date = time.strftime('%Y-%m-%d')
stock = yahoo_finance.Share(symbol)
opening_price = stock.get_open()
closing_price = 0.0
twitter_sentiment = tweet_query(symbol)
headline_sentiment = nytimes_sentiment(symbol)
overall_sentiment_num = twitter_sentiment[0] + headline_sentiment[0]
overall_sentiment = 'Positive' if overall_sentiment_num > 505 else 'Negative'
pos_azure_sentiment = azure_req(date, symbol, 'Positive', float(opening_price) * 1.05)
neg_azure_sentiment = azure_req(date, symbol, 'Negative', float(opening_price) * .95)
pos_val = float(pos_azure_sentiment['Results']['output1']['value']['Values'][0][4])
neg_val = float(neg_azure_sentiment['Results']['output1']['value']['Values'][0][4])
print pos_val, neg_val
midpoint = abs((pos_val + neg_val) / 2)
if overall_sentiment == 'Positive':
closing_price += float(opening_price) + (((pos_val - midpoint) / midpoint) * float(opening_price))
else:
closing_price += float(opening_price) - (((midpoint - neg_val) / midpoint) * float(opening_price))
difference = closing_price - float(opening_price)
percent_diff = (difference / float(opening_price)) * 100
json_result = {
'success': True,
'opening_price' : opening_price,
'closing_price': closing_price,
'difference': str(percent_diff),
'sentiment': overall_sentiment
}
return jsonify(json_result)
def tweets():
symbol = request.get_data()
twitter_sentiment = tweet_query(symbol)
json_result = {
'success' : True,
'sentiment' : twitter_sentiment[2],
'num_pos' : str(twitter_sentiment[0]),
'num_neg' : str(twitter_sentiment[1]),
'percent_pos' : str((float(twitter_sentiment[0]) / (twitter_sentiment[0] + twitter_sentiment[1])) * 100),
'percent_neg' : str((float(twitter_sentiment[1]) / (twitter_sentiment[0] + twitter_sentiment[1])) * 100)
}
return jsonify(json_result)