print(i)
cat = Cat("kitty")
cat.add_trick("meow!!")
dog = Dog("puppy")
dog.add_trick("ruff")
print(cat.get_name(), "abc")
print(dog.get_name() + " " + str(dog.get_tricks()))
i = 0
print " 10 mod 3 =" + str(10 / 3)
dataSet = readDataSets('/home/grant/Downloads/phd_data/AMC2017-2018.csv')
print dataSet.nextBatch(3)
print dataSet.nextBatch(3)
tradingData = readDataSets(
'/home/grant/Downloads/phd_data/AMC2000-2012.csv')
batch_xs, batch_ys = tradingData.nextLstmBatch(100)
print "nextLstmBatch"
print batch_xs
print batch_ys
model.compile(loss='mean_absolute_error',
optimizer=myOptimizer,
metrics=['accuracy', mean_pred, earning_pred])
return model
stocks = ['AMC', 'ANZ', 'BHP', 'CBA', 'NAB', 'GPT', 'CIM', 'CCL']
priceIncrease = 1.005
for stock in stocks:
print stock
tradingData = readDataSets(
'/home/grant/Downloads/phd_data/' + stock + '2000-2012.csv',
priceIncrease)
epochSize = 60
batchSize = 50
batch_xs, batch_ys = tradingData.nextBatch(batchSize * epochSize)
batch_xs = numpy.reshape(batch_xs, [epochSize, batchSize, 100])
batch_ys = numpy.reshape(batch_ys, [epochSize, batchSize, 1])
#print batch_ys;
model = buildModel()
history = model.fit(batch_xs,
batch_ys,
def main(_):
#AMC, ANZ, BHP, CBA, NAB, GPT, CIM, CCL
stock = "CBA"
# Import data
tradingData = readDataSets('/home/grant/Downloads/phd_data/'+stock+'2000-2012.csv')
# Create the model
x = tf.placeholder(tf.float32, [None, 100], "tradingPrices")
W = tf.Variable(tf.zeros([100, 1]), name='priceWeight')
b = tf.Variable(tf.zeros([1]), name='bias')
y = tf.sigmoid(tf.matmul(x, W) + b)
# Define loss and optimizer
y_ = tf.placeholder(tf.float32, [None,1], "probHigherPrice")
# The raw formulation of cross-entropy,
#
# tf.reduce_mean(-tf.reduce_sum(y_ * tf.log(tf.nn.softmax(y)),
# reduction_indices=[1]))
#
# can be numerically unstable.
#
# So here we use tf.losses.sparse_softmax_cross_entropy on the raw
# outputs of 'y', and then average across the batch.
#cross_entropy = tf.losses.mean_squared_error(y_, y)
#cross_entropy = tf.losses.softmax_cross_entropy(y_,y);
cross_entropy = tf.reduce_sum(tf.abs(tf.subtract(y_, y)))
train_step = tf.train.GradientDescentOptimizer(0.5).minimize(cross_entropy)
correct_prediction = tf.equal( tf.cast(tf.round(y), tf.int32), tf.cast(tf.round(y_), tf.int32))
accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
tf.summary.scalar('accuracy', accuracy)
tf.summary.scalar('cross_entropy', cross_entropy)
merged = tf.summary.merge_all()
with tf.Session() as sess:
train_writer = tf.summary.FileWriter('/tmp/train', sess.graph)
tf.global_variables_initializer().run()
# Train
for i in range(60):
batchSize = 50;
batch_xs, batch_ys = tradingData.nextBatch(batchSize)
batch_xs = numpy.reshape(batch_xs, [batchSize, 100])
batch_ys = numpy.reshape(batch_ys, [batchSize, 1])
sess.run(train_step, feed_dict={x: batch_xs, y_: batch_ys})
summary, entropy = sess.run([merged,cross_entropy], feed_dict={x: batch_xs, y_: batch_ys})
#print(entropy)
#print (sess.run( cross_entropy , feed_dict={x: batch_xs, y_: batch_ys}))
train_writer.add_summary(summary, i)
# Test trained model
train_writer.close()
print(W.eval())
testData = readDataSets('/home/grant/Downloads/phd_data/'+stock+'2013-2017.csv');
test_xs, test_ys = testData.nextBatch(200)
test_xs = numpy.reshape(test_xs, [200, 100])
test_ys = numpy.reshape(test_ys, [200, 1])
#print (test_xs)
#print(sess.run(tf.round(test_ys)))
print(sess.run(tf.abs(y), feed_dict={x: test_xs,y_:test_ys}))