# 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
}
# 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:
