# tickers = ['ibe_en', 'ibe_es', 'vod_en', 'vod_es', 'san_en', 'san_es', 'tef_en'] fromDate = datetime.date(2009, 05, 01) toDate = datetime.date(2010, 03, 31) tickers = ['aapl', 'goog', 'amzn'] mongoClient = pymongo.MongoClient() db_twitter = mongoClient.twitter db_ftt = mongoClient.ftt for ticker in tickers: print ticker.upper() stock_data = downloadData.getData(ticker, fromDate, toDate) for key in stock_data.keys(): document = {} date = datetime.datetime(key.year, key.month, key.day, 0, 0, 0) emotion = { 'anger': 0, 'happiness': 0, 'fear': 0, 'disgust': 0, 'sadness': 0, 'surprise': 0 }
import downloadData client = pymongo.MongoClient() tickers = ['aapl', 'goog', 'amzn', 'ibe_en', 'ibe_es', 'san_en', 'san_es', 'vod_en', 'vod_es', 'tef_en'] for ticker in tickers: print ticker if (ticker == 'aapl') or (ticker == 'goog') or (ticker == 'amzn'): fromDate = datetime.datetime(2009, 05, 01) toDate = datetime.datetime(2010, 03, 31) else: fromDate = datetime.datetime(2013, 12, 12) toDate = datetime.datetime(2014, 03, 13) vix_col = downloadData.getData("%5EVXN", fromDate, toDate) # print vix_col db = client.ftt date = fromDate while (date <= toDate): post = db[ticker].find({'date': date}) if not datetime.date(date.year, date.month, date.day) in vix_col.keys(): print 'not_date' date = date + datetime.timedelta(days=1) continue if post.count() == 0: date = date + datetime.timedelta(days=1)
import tensorflow as tf import numpy as np import random import downloadData from sklearn.model_selection import train_test_split #load the data and get the dictionary size trX, trY, dictionarySize = downloadData.getData() #split data into train and test X_train, X_test, y_train, y_test = train_test_split(trX, trY, test_size=0.10) #init weights def init_weights(shape): return tf.Variable(tf.random_normal(shape, stddev=0.01)) #the mlp def model(X, w_h1, w_h2, w_h3, w_h4, w_o): h = tf.nn.relu(tf.matmul(X, w_h1)) # layer 1 h2 = tf.nn.relu(tf.matmul(h, w_h2)) # layer 2 h3 = tf.nn.relu(tf.matmul(h2, w_h3)) # layer 3 h4 = tf.nn.relu(tf.matmul(h3, w_h4)) # layer 4 return tf.matmul(h4, w_o) # relabel train and test trX, teX,trY, teY = X_train, X_test, y_train, y_test #variables initilizaitons
tickers = [ 'aapl', 'goog', 'amzn', 'ibe_en', 'ibe_es', 'san_en', 'san_es', 'vod_en', 'vod_es', 'tef_en' ] for ticker in tickers: print ticker if (ticker == 'aapl') or (ticker == 'goog') or (ticker == 'amzn'): fromDate = datetime.datetime(2009, 05, 01) toDate = datetime.datetime(2010, 03, 31) else: fromDate = datetime.datetime(2013, 12, 12) toDate = datetime.datetime(2014, 03, 13) vix_col = downloadData.getData("%5EVXN", fromDate, toDate) # print vix_col db = client.ftt date = fromDate while (date <= toDate): post = db[ticker].find({'date': date}) if not datetime.date(date.year, date.month, date.day) in vix_col.keys(): print 'not_date' date = date + datetime.timedelta(days=1) continue if post.count() == 0: