def cycleTesting2Layer(stock,
length,
lrs,
activations,
nodes,
epoch,
batches,
saveName,
saveBase,
reshape=False):
# Obtain the stock data
(train_d, train_t, test_d,
test_t) = sf.getInputData(stock, length, reshape)
result = 0
bestResult = 0
count = 1
totalRuns = len(lrs) * len(activations) * len(nodes) * len(batches) * len(
nodes)
startTime = time.time()
for lr in lrs:
for func in activations:
for depth in nodes:
for depth1 in nodes:
for mbs in batches:
func1 = func
sf.printProgress2(startTime, count, totalRuns, lr,
func, func1, depth, depth1, mbs,
length, result, bestResult)
count += 1
model = tflearn.DNN(
ann_two_level(length, (depth, depth1), lr,
(func, func1)))
model.fit(train_d,
train_t,
n_epoch=epoch,
shuffle=False,
validation_set=(test_d, test_t),
show_metric=False,
batch_size=mbs)
result = test_model2(model, length, test_d, test_t)
if result > bestResult:
bestResult = result
bestLr = lr
bestFunc = func
bestFunc1 = func1
bestDepth = depth
bestDepth1 = depth1
bestMBS = mbs
sf.writeFile2Level(length, epoch, bestResult,
bestLr, bestFunc, bestFunc1,
bestDepth, bestDepth1, bestMBS,
saveName, saveBase)
model.save(saveBase + 'nets/ann/twoLayer/' +
saveName + '.pck')
tf.reset_default_graph()
def confirmTestingAccuracy(stock, netStyle, netName, layers):
# Load network parameters.
(netStyle, memory, lr, funcs,
depths) = sf.load("results/" + netStyle + "/" + sf.intToString(layers) +
"Level/" + netName + ".pck")
# Obtain the data.
data = sf.getInputData(stock, memory, netStyle == 'conv')
# Build the network and make a prediction.
network = getPredictionNet(netName, netStyle, layers, memory, lr, funcs,
depths)
accuracy = test_model2(network, memory, data[2], data[3],
netStyle == 'conv')
tf.reset_default_graph()
return accuracy
def train_net(stock,
length,
netStyle,
epoch,
mbs,
lr,
funcs,
depths,
saveName,
saveBase,
reshape=False):
# Make the network build
build = getTrainingBuild(netStyle, length, lr, depths, funcs)
# Obtain the stock data
(train_d, train_t, test_d,
test_t) = sf.getInputData(stock, length, reshape)
# Build and train the network
model = tflearn.DNN(build)
model.fit(train_d,
train_t,
n_epoch=epoch,
shuffle=False,
validation_set=(test_d, test_t),
show_metric=True,
batch_size=mbs)
tf.reset_default_graph()
# Save the network and other training files.
model.save('nets/' + netStyle + '/' + sf.intToString(len(funcs)) +
'Layer/' + saveName + '.pck')
sf.chooseFileWriter(length, netStyle, epoch,
test_model2(model, length, test_d, test_t, reshape),
lr, funcs, depths, mbs, saveName, saveBase)