class EightBitBrain(object):
def __init__(self, dataset, inNodes, outNodes, hiddenNodes, classes):
self.__dataset = ClassificationDataSet(inNodes, classes-1)
for element in dataset:
self.addDatasetSample(self._binaryList(element[0]), element[1])
self.__dataset._convertToOneOfMany()
self.__network = buildNetwork(inNodes, hiddenNodes, self.__dataset.outdim, recurrent=True)
self.__trainer = BackpropTrainer(self.__network, learningrate = 0.01, momentum = 0.99, verbose = True)
self.__trainer.setData(self.__dataset)
def _binaryList(self, n):
return [int(c) for c in "{0:08b}".format(n)]
def addDatasetSample(self, argument, target):
self.__dataset.addSample(argument, target)
def train(self, epochs):
self.__trainer.trainEpochs(epochs)
def activate(self, information):
result = self.__network.activate(self._binaryList(information))
highest = (0,0)
for resultClass in range(len(result)):
if result[resultClass] > highest[0]:
highest = (result[resultClass], resultClass)
return highest[1]
class Suite(PyExperimentSuite):
def reset(self, params, repetition):
print params
self.nDimInput = 3
self.inputEncoder = PassThroughEncoder()
if params['output_encoding'] == None:
self.outputEncoder = PassThroughEncoder()
self.nDimOutput = 1
elif params['output_encoding'] == 'likelihood':
self.outputEncoder = ScalarBucketEncoder()
self.nDimOutput = self.outputEncoder.encoder.n
if (params['dataset'] == 'nyc_taxi' or
params['dataset'] == 'nyc_taxi_perturb_baseline'):
self.dataset = NYCTaxiDataset(params['dataset'])
else:
raise Exception("Dataset not found")
self.testCounter = 0
self.resets = []
self.iteration = 0
# initialize LSTM network
random.seed(6)
if params['output_encoding'] == None:
self.net = buildNetwork(self.nDimInput, params['num_cells'], self.nDimOutput,
hiddenclass=LSTMLayer, bias=True, outputbias=True, recurrent=True)
elif params['output_encoding'] == 'likelihood':
self.net = buildNetwork(self.nDimInput, params['num_cells'], self.nDimOutput,
hiddenclass=LSTMLayer, bias=True, outclass=SigmoidLayer, recurrent=True)
self.trainer = BackpropTrainer(self.net,
dataset=SequentialDataSet(self.nDimInput, self.nDimOutput),
learningrate=0.01,
momentum=0,
verbose=params['verbosity'] > 0)
(self.networkInput, self.targetPrediction, self.trueData) = \
self.dataset.generateSequence(
prediction_nstep=params['prediction_nstep'],
output_encoding=params['output_encoding'],
noise=params['noise'])
def window(self, data, params):
start = max(0, self.iteration - params['learning_window'])
return data[start:self.iteration]
def train(self, params, verbose=False):
if params['reset_every_training']:
if verbose:
print 'create lstm network'
random.seed(6)
if params['output_encoding'] == None:
self.net = buildNetwork(self.nDimInput, params['num_cells'], self.nDimOutput,
hiddenclass=LSTMLayer, bias=True, outputbias=True, recurrent=True)
elif params['output_encoding'] == 'likelihood':
self.net = buildNetwork(self.nDimInput, params['num_cells'], self.nDimOutput,
hiddenclass=LSTMLayer, bias=True, outclass=SigmoidLayer, recurrent=True)
self.net.reset()
ds = SequentialDataSet(self.nDimInput, self.nDimOutput)
networkInput = self.window(self.networkInput, params)
targetPrediction = self.window(self.targetPrediction, params)
# prepare a training data-set using the history
for i in xrange(len(networkInput)):
ds.addSample(self.inputEncoder.encode(networkInput[i]),
self.outputEncoder.encode(targetPrediction[i]))
if params['num_epochs'] > 1:
trainer = RPropMinusTrainer(self.net, dataset=ds, verbose=verbose)
if verbose:
print " train LSTM on ", len(ds), " records for ", params['num_epochs'], " epochs "
if len(networkInput) > 1:
trainer.trainEpochs(params['num_epochs'])
else:
self.trainer.setData(ds)
self.trainer.train()
# run through the training dataset to get the lstm network state right
self.net.reset()
for i in xrange(len(networkInput)):
self.net.activate(ds.getSample(i)[0])
def iterate(self, params, repetition, iteration, verbose=True):
self.iteration = iteration
if self.iteration >= len(self.networkInput):
return None
train = False
if iteration > params['compute_after']:
if iteration == params['train_at_iteration']:
train = True
if params['train_every_month']:
train = (self.dataset.sequence['time'][iteration].is_month_start and
self.dataset.sequence['time'][iteration].hour == 0 and
self.dataset.sequence['time'][iteration].minute == 0)
if params['train_every_week']:
train = (self.dataset.sequence['time'][iteration].dayofweek==0 and
self.dataset.sequence['time'][iteration].hour == 0 and
self.dataset.sequence['time'][iteration].minute == 0)
if params['online_training']:
train = True
if verbose:
print
print "iteration: ", iteration, " time: ", self.dataset.sequence['time'][iteration]
if train:
if verbose:
print " train at", iteration, " time: ", self.dataset.sequence['time'][iteration]
self.train(params, verbose)
if train:
# reset test counter after training
self.testCounter = params['test_for']
if self.testCounter == 0:
return None
else:
self.testCounter -= 1
symbol = self.networkInput[iteration]
output = self.net.activate(self.inputEncoder.encode(symbol))
if params['output_encoding'] == None:
predictions = self.dataset.reconstructSequence(output[0])
elif params['output_encoding'] == 'likelihood':
predictions = list(output/sum(output))
else:
predictions = None
if verbose:
print " test at :", iteration,
if iteration == params['perturb_after']:
if verbose:
print " perturb data and introduce new patterns"
(newNetworkInput, newTargetPrediction, newTrueData) = \
self.dataset.generateSequence(perturbed=True,
prediction_nstep=params['prediction_nstep'],
output_encoding=params['output_encoding'],
noise=params['noise'])
self.networkInput[iteration+1:] = newNetworkInput[iteration+1:]
self.targetPrediction[iteration+1:] = newTargetPrediction[iteration+1:]
self.trueData[iteration+1:] = newTrueData[iteration+1:]
return {"current": self.networkInput[iteration],
"reset": None,
"train": train,
"predictions": predictions,
"truth": self.trueData[iteration]}
class Suite(PyExperimentSuite):
def reset(self, params, repetition):
random.seed(params['seed'])
if params['encoding'] == 'basic':
self.encoder = BasicEncoder(params['encoding_num'])
elif params['encoding'] == 'distributed':
self.encoder = DistributedEncoder(params['encoding_num'],
maxValue=params['encoding_max'],
minValue=params['encoding_min'],
classifyWithRandom=params[
'classify_with_random'])
else:
raise Exception("Encoder not found")
if params['dataset'] == 'simple':
self.dataset = SimpleDataset()
elif params['dataset'] == 'reber':
self.dataset = ReberDataset(maxLength=params['max_length'])
elif params['dataset'] == 'high-order':
self.dataset = HighOrderDataset(numPredictions=params['num_predictions'],
seed=params['seed'])
else:
raise Exception("Dataset not found")
self.computeCounter = 0
self.history = []
self.resets = []
self.randoms = []
self.currentSequence = []
self.targetPrediction = []
self.replenishSequence(params, iteration=0)
self.net = buildNetwork(params['encoding_num'], params['num_cells'],
params['encoding_num'],
hiddenclass=LSTMLayer,
bias=True,
outputbias=params['output_bias'],
recurrent=True)
self.trainer = BackpropTrainer(self.net,
dataset=SequentialDataSet(params['encoding_num'], params['encoding_num']),
learningrate=0.01,
momentum=0,
verbose=params['verbosity'] > 0)
self.sequenceCounter = 0
def window(self, data, params):
start = max(0, len(data) - params['learning_window'])
return data[start:]
def train(self, params):
"""
Train LSTM network on buffered dataset history
After training, run LSTM on history[:-1] to get the state correct
:param params:
:return:
"""
if params['reset_every_training']:
n = params['encoding_num']
self.net = buildNetwork(n, params['num_cells'], n,
hiddenclass=LSTMLayer,
bias=True,
outputbias=params['output_bias'],
recurrent=True)
self.net.reset()
# prepare training dataset
ds = SequentialDataSet(params['encoding_num'], params['encoding_num'])
history = self.window(self.history, params)
resets = self.window(self.resets, params)
for i in xrange(1, len(history)):
if not resets[i - 1]:
ds.addSample(self.encoder.encode(history[i - 1]),
self.encoder.encode(history[i]))
if resets[i]:
ds.newSequence()
print "Train LSTM network on buffered dataset of length ", len(history)
if params['num_epochs'] > 1:
trainer = RPropMinusTrainer(self.net,
dataset=ds,
verbose=params['verbosity'] > 0)
if len(history) > 1:
trainer.trainEpochs(params['num_epochs'])
# run network on buffered dataset after training to get the state right
self.net.reset()
for i in xrange(len(history) - 1):
symbol = history[i]
output = self.net.activate(self.encoder.encode(symbol))
self.encoder.classify(output, num=params['num_predictions'])
if resets[i]:
self.net.reset()
else:
self.trainer.setData(ds)
self.trainer.train()
# run network on buffered dataset after training to get the state right
self.net.reset()
for i in xrange(len(history) - 1):
symbol = history[i]
output = self.net.activate(self.encoder.encode(symbol))
self.encoder.classify(output, num=params['num_predictions'])
if resets[i]:
self.net.reset()
def killCells(self, killCellPercent):
"""
kill a fraction of LSTM cells from the network
:param killCellPercent:
:return:
"""
if killCellPercent <= 0:
return
inputLayer = self.net['in']
lstmLayer = self.net['hidden0']
numLSTMCell = lstmLayer.outdim
numDead = round(killCellPercent * numLSTMCell)
zombiePermutation = numpy.random.permutation(numLSTMCell)
deadCells = zombiePermutation[0:numDead]
# remove connections from input layer to dead LSTM cells
connectionInputToHidden = self.net.connections[inputLayer][0]
weightInputToHidden = reshape(connectionInputToHidden.params,
(connectionInputToHidden.outdim,
connectionInputToHidden.indim))
for cell in deadCells:
for dim in range(4):
weightInputToHidden[dim * numLSTMCell + cell, :] *= 0
newParams = reshape(weightInputToHidden,
(connectionInputToHidden.paramdim,))
self.net.connections[inputLayer][0]._setParameters(
newParams, connectionInputToHidden.owner)
# remove dead connections within LSTM layer
connectionHiddenToHidden = self.net.recurrentConns[0]
weightHiddenToHidden = reshape(connectionHiddenToHidden.params,
(connectionHiddenToHidden.outdim,
connectionHiddenToHidden.indim))
for cell in deadCells:
weightHiddenToHidden[:, cell] *= 0
newParams = reshape(weightHiddenToHidden,
(connectionHiddenToHidden.paramdim,))
self.net.recurrentConns[0]._setParameters(
newParams, connectionHiddenToHidden.owner)
# remove connections from dead LSTM cell to output layer
connectionHiddenToOutput = self.net.connections[lstmLayer][0]
weightHiddenToOutput = reshape(connectionHiddenToOutput.params,
(connectionHiddenToOutput.outdim,
connectionHiddenToOutput.indim))
for cell in deadCells:
weightHiddenToOutput[:, cell] *= 0
newParams = reshape(weightHiddenToOutput,
(connectionHiddenToOutput.paramdim,))
self.net.connections[lstmLayer][0]._setParameters(
newParams, connectionHiddenToOutput.owner)
def replenishSequence(self, params, iteration):
if iteration > params['perturb_after']:
sequence, target = self.dataset.generateSequence(params['seed']+iteration,
perturbed=True)
else:
sequence, target = self.dataset.generateSequence(params['seed']+iteration)
if (iteration > params['inject_noise_after'] and
iteration < params['stop_inject_noise_after']):
injectNoiseAt = random.randint(1, 3)
sequence[injectNoiseAt] = self.encoder.randomSymbol()
if params['separate_sequences_with'] == 'random':
sequence.append(self.encoder.randomSymbol(seed=params['seed']+iteration))
target.append(None)
if params['verbosity'] > 0:
print "Add sequence to buffer"
print "sequence: ", sequence
print "target: ", target
self.currentSequence += sequence
self.targetPrediction += target
def check_prediction(self, topPredictions, targets):
if targets is None:
correct = None
else:
if isinstance(targets, numbers.Number):
correct = targets in topPredictions
else:
correct = True
for prediction in topPredictions:
correct = correct and (prediction in targets)
return correct
def iterate(self, params, repetition, iteration):
currentElement = self.currentSequence.pop(0)
target = self.targetPrediction.pop(0)
# update buffered dataset
self.history.append(currentElement)
# whether there will be a reset signal after the current record
resetFlag = (len(self.currentSequence) == 0 and
params['separate_sequences_with'] == 'reset')
self.resets.append(resetFlag)
# whether there will be a random symbol after the current record
randomFlag = (len(self.currentSequence) == 1 and
params['separate_sequences_with'] == 'random')
self.randoms.append(randomFlag)
if len(self.currentSequence) == 0:
self.replenishSequence(params, iteration)
self.sequenceCounter += 1
# kill cells
killCell = False
if iteration == params['kill_cell_after']:
killCell = True
self.killCells(params['kill_cell_percent'])
# reset compute counter
if iteration > 0 and iteration % params['compute_every'] == 0:
self.computeCounter = params['compute_for']
if self.computeCounter == 0 or iteration < params['compute_after']:
computeLSTM = False
else:
computeLSTM = True
if computeLSTM:
self.computeCounter -= 1
train = (not params['compute_test_mode'] or
iteration % params['compute_every'] == 0)
if train:
if params['verbosity'] > 0:
print "Training LSTM at iteration {}".format(iteration)
self.train(params)
# run LSTM on the latest data record
output = self.net.activate(self.encoder.encode(currentElement))
if params['encoding'] == 'distributed':
predictions = self.encoder.classify(output, num=params['num_predictions'])
elif params['encoding'] == 'basic':
predictions = self.encoder.classify(output, num=params['num_predictions'])
correct = self.check_prediction(predictions, target)
if params['verbosity'] > 0:
print ("iteration: {0} \t"
"current: {1} \t"
"predictions: {2} \t"
"truth: {3} \t"
"correct: {4} \t").format(
iteration, currentElement, predictions, target, correct)
if self.resets[-1]:
if params['verbosity'] > 0:
print "Reset LSTM at iteration {}".format(iteration)
self.net.reset()
return {"iteration": iteration,
"current": currentElement,
"reset": self.resets[-1],
"random": self.randoms[-1],
"train": train,
"predictions": predictions,
"truth": target,
"killCell": killCell,
"sequenceCounter": self.sequenceCounter}
class Suite(PyExperimentSuite):
def reset(self, params, repetition):
random.seed(params['seed'])
if params['encoding'] == 'basic':
self.encoder = BasicEncoder(params['encoding_num'])
elif params['encoding'] == 'distributed':
self.encoder = DistributedEncoder(
params['encoding_num'],
maxValue=params['encoding_max'],
minValue=params['encoding_min'],
classifyWithRandom=params['classify_with_random'])
else:
raise Exception("Encoder not found")
if params['dataset'] == 'simple':
self.dataset = SimpleDataset()
elif params['dataset'] == 'reber':
self.dataset = ReberDataset(maxLength=params['max_length'])
elif params['dataset'] == 'high-order':
self.dataset = HighOrderDataset(
numPredictions=params['num_predictions'], seed=params['seed'])
else:
raise Exception("Dataset not found")
self.computeCounter = 0
self.history = []
self.resets = []
self.randoms = []
self.currentSequence = []
self.targetPrediction = []
self.replenishSequence(params, iteration=0)
self.net = buildNetwork(params['encoding_num'],
params['num_cells'],
params['encoding_num'],
hiddenclass=LSTMLayer,
bias=True,
outputbias=params['output_bias'],
recurrent=True)
self.trainer = BackpropTrainer(self.net,
dataset=SequentialDataSet(
params['encoding_num'],
params['encoding_num']),
learningrate=0.01,
momentum=0,
verbose=params['verbosity'] > 0)
self.sequenceCounter = 0
def window(self, data, params):
start = max(0, len(data) - params['learning_window'])
return data[start:]
def train(self, params):
"""
Train LSTM network on buffered dataset history
After training, run LSTM on history[:-1] to get the state correct
:param params:
:return:
"""
if params['reset_every_training']:
n = params['encoding_num']
self.net = buildNetwork(n,
params['num_cells'],
n,
hiddenclass=LSTMLayer,
bias=True,
outputbias=params['output_bias'],
recurrent=True)
self.net.reset()
# prepare training dataset
ds = SequentialDataSet(params['encoding_num'], params['encoding_num'])
history = self.window(self.history, params)
resets = self.window(self.resets, params)
for i in xrange(1, len(history)):
if not resets[i - 1]:
ds.addSample(self.encoder.encode(history[i - 1]),
self.encoder.encode(history[i]))
if resets[i]:
ds.newSequence()
print "Train LSTM network on buffered dataset of length ", len(history)
if params['num_epochs'] > 1:
trainer = RPropMinusTrainer(self.net,
dataset=ds,
verbose=params['verbosity'] > 0)
if len(history) > 1:
trainer.trainEpochs(params['num_epochs'])
# run network on buffered dataset after training to get the state right
self.net.reset()
for i in xrange(len(history) - 1):
symbol = history[i]
output = self.net.activate(self.encoder.encode(symbol))
self.encoder.classify(output, num=params['num_predictions'])
if resets[i]:
self.net.reset()
else:
self.trainer.setData(ds)
self.trainer.train()
# run network on buffered dataset after training to get the state right
self.net.reset()
for i in xrange(len(history) - 1):
symbol = history[i]
output = self.net.activate(self.encoder.encode(symbol))
self.encoder.classify(output, num=params['num_predictions'])
if resets[i]:
self.net.reset()
def killCells(self, killCellPercent):
"""
kill a fraction of LSTM cells from the network
:param killCellPercent:
:return:
"""
if killCellPercent <= 0:
return
inputLayer = self.net['in']
lstmLayer = self.net['hidden0']
numLSTMCell = lstmLayer.outdim
numDead = round(killCellPercent * numLSTMCell)
zombiePermutation = numpy.random.permutation(numLSTMCell)
deadCells = zombiePermutation[0:numDead]
# remove connections from input layer to dead LSTM cells
connectionInputToHidden = self.net.connections[inputLayer][0]
weightInputToHidden = reshape(
connectionInputToHidden.params,
(connectionInputToHidden.outdim, connectionInputToHidden.indim))
for cell in deadCells:
for dim in range(4):
weightInputToHidden[dim * numLSTMCell + cell, :] *= 0
newParams = reshape(weightInputToHidden,
(connectionInputToHidden.paramdim, ))
self.net.connections[inputLayer][0]._setParameters(
newParams, connectionInputToHidden.owner)
# remove dead connections within LSTM layer
connectionHiddenToHidden = self.net.recurrentConns[0]
weightHiddenToHidden = reshape(
connectionHiddenToHidden.params,
(connectionHiddenToHidden.outdim, connectionHiddenToHidden.indim))
for cell in deadCells:
weightHiddenToHidden[:, cell] *= 0
newParams = reshape(weightHiddenToHidden,
(connectionHiddenToHidden.paramdim, ))
self.net.recurrentConns[0]._setParameters(
newParams, connectionHiddenToHidden.owner)
# remove connections from dead LSTM cell to output layer
connectionHiddenToOutput = self.net.connections[lstmLayer][0]
weightHiddenToOutput = reshape(
connectionHiddenToOutput.params,
(connectionHiddenToOutput.outdim, connectionHiddenToOutput.indim))
for cell in deadCells:
weightHiddenToOutput[:, cell] *= 0
newParams = reshape(weightHiddenToOutput,
(connectionHiddenToOutput.paramdim, ))
self.net.connections[lstmLayer][0]._setParameters(
newParams, connectionHiddenToOutput.owner)
def replenishSequence(self, params, iteration):
if iteration > params['perturb_after']:
sequence, target = self.dataset.generateSequence(params['seed'] +
iteration,
perturbed=True)
else:
sequence, target = self.dataset.generateSequence(params['seed'] +
iteration)
if (iteration > params['inject_noise_after']
and iteration < params['stop_inject_noise_after']):
injectNoiseAt = random.randint(1, 3)
sequence[injectNoiseAt] = self.encoder.randomSymbol()
if params['separate_sequences_with'] == 'random':
sequence.append(
self.encoder.randomSymbol(seed=params['seed'] + iteration))
target.append(None)
if params['verbosity'] > 0:
print "Add sequence to buffer"
print "sequence: ", sequence
print "target: ", target
self.currentSequence += sequence
self.targetPrediction += target
def check_prediction(self, topPredictions, targets):
if targets is None:
correct = None
else:
if isinstance(targets, numbers.Number):
correct = targets in topPredictions
else:
correct = True
for prediction in topPredictions:
correct = correct and (prediction in targets)
return correct
def iterate(self, params, repetition, iteration):
currentElement = self.currentSequence.pop(0)
target = self.targetPrediction.pop(0)
# update buffered dataset
self.history.append(currentElement)
# whether there will be a reset signal after the current record
resetFlag = (len(self.currentSequence) == 0
and params['separate_sequences_with'] == 'reset')
self.resets.append(resetFlag)
# whether there will be a random symbol after the current record
randomFlag = (len(self.currentSequence) == 1
and params['separate_sequences_with'] == 'random')
self.randoms.append(randomFlag)
if len(self.currentSequence) == 0:
self.replenishSequence(params, iteration)
self.sequenceCounter += 1
# kill cells
killCell = False
if iteration == params['kill_cell_after']:
killCell = True
self.killCells(params['kill_cell_percent'])
# reset compute counter
if iteration > 0 and iteration % params['compute_every'] == 0:
self.computeCounter = params['compute_for']
if self.computeCounter == 0 or iteration < params['compute_after']:
computeLSTM = False
else:
computeLSTM = True
if computeLSTM:
self.computeCounter -= 1
train = (not params['compute_test_mode']
or iteration % params['compute_every'] == 0)
if train:
if params['verbosity'] > 0:
print "Training LSTM at iteration {}".format(iteration)
self.train(params)
# run LSTM on the latest data record
output = self.net.activate(self.encoder.encode(currentElement))
if params['encoding'] == 'distributed':
predictions = self.encoder.classify(
output, num=params['num_predictions'])
elif params['encoding'] == 'basic':
predictions = self.encoder.classify(
output, num=params['num_predictions'])
correct = self.check_prediction(predictions, target)
if params['verbosity'] > 0:
print("iteration: {0} \t"
"current: {1} \t"
"predictions: {2} \t"
"truth: {3} \t"
"correct: {4} \t").format(iteration, currentElement,
predictions, target, correct)
if self.resets[-1]:
if params['verbosity'] > 0:
print "Reset LSTM at iteration {}".format(iteration)
self.net.reset()
return {
"iteration": iteration,
"current": currentElement,
"reset": self.resets[-1],
"random": self.randoms[-1],
"train": train,
"predictions": predictions,
"truth": target,
"killCell": killCell,
"sequenceCounter": self.sequenceCounter
}
class Suite(PyExperimentSuite):
def reset(self, params, repetition):
print params
self.nDimInput = 3
self.inputEncoder = PassThroughEncoder()
if params['output_encoding'] == None:
self.outputEncoder = PassThroughEncoder()
self.nDimOutput = 1
elif params['output_encoding'] == 'likelihood':
self.outputEncoder = ScalarBucketEncoder()
self.nDimOutput = self.outputEncoder.encoder.n
if (params['dataset'] == 'nyc_taxi'
or params['dataset'] == 'nyc_taxi_perturb_baseline'):
self.dataset = NYCTaxiDataset(params['dataset'])
else:
raise Exception("Dataset not found")
self.testCounter = 0
self.resets = []
self.iteration = 0
# initialize LSTM network
random.seed(6)
if params['output_encoding'] == None:
self.net = buildNetwork(self.nDimInput,
params['num_cells'],
self.nDimOutput,
hiddenclass=LSTMLayer,
bias=True,
outputbias=True,
recurrent=True)
elif params['output_encoding'] == 'likelihood':
self.net = buildNetwork(self.nDimInput,
params['num_cells'],
self.nDimOutput,
hiddenclass=LSTMLayer,
bias=True,
outclass=SigmoidLayer,
recurrent=True)
self.trainer = BackpropTrainer(self.net,
dataset=SequentialDataSet(
self.nDimInput, self.nDimOutput),
learningrate=0.01,
momentum=0,
verbose=params['verbosity'] > 0)
(self.networkInput, self.targetPrediction, self.trueData) = \
self.dataset.generateSequence(
prediction_nstep=params['prediction_nstep'],
output_encoding=params['output_encoding'],
noise=params['noise'])
def window(self, data, params):
start = max(0, self.iteration - params['learning_window'])
return data[start:self.iteration]
def train(self, params, verbose=False):
if params['reset_every_training']:
if verbose:
print 'create lstm network'
random.seed(6)
if params['output_encoding'] == None:
self.net = buildNetwork(self.nDimInput,
params['num_cells'],
self.nDimOutput,
hiddenclass=LSTMLayer,
bias=True,
outputbias=True,
recurrent=True)
elif params['output_encoding'] == 'likelihood':
self.net = buildNetwork(self.nDimInput,
params['num_cells'],
self.nDimOutput,
hiddenclass=LSTMLayer,
bias=True,
outclass=SigmoidLayer,
recurrent=True)
self.net.reset()
ds = SequentialDataSet(self.nDimInput, self.nDimOutput)
networkInput = self.window(self.networkInput, params)
targetPrediction = self.window(self.targetPrediction, params)
# prepare a training data-set using the history
for i in xrange(len(networkInput)):
ds.addSample(self.inputEncoder.encode(networkInput[i]),
self.outputEncoder.encode(targetPrediction[i]))
if params['num_epochs'] > 1:
trainer = RPropMinusTrainer(self.net, dataset=ds, verbose=verbose)
if verbose:
print " train LSTM on ", len(
ds), " records for ", params['num_epochs'], " epochs "
if len(networkInput) > 1:
trainer.trainEpochs(params['num_epochs'])
else:
self.trainer.setData(ds)
self.trainer.train()
# run through the training dataset to get the lstm network state right
self.net.reset()
for i in xrange(len(networkInput)):
self.net.activate(ds.getSample(i)[0])
def iterate(self, params, repetition, iteration, verbose=True):
self.iteration = iteration
if self.iteration >= len(self.networkInput):
return None
train = False
if iteration > params['compute_after']:
if iteration == params['train_at_iteration']:
train = True
if params['train_every_month']:
train = (
self.dataset.sequence['time'][iteration].is_month_start
and self.dataset.sequence['time'][iteration].hour == 0
and self.dataset.sequence['time'][iteration].minute == 0)
if params['train_every_week']:
train = (
self.dataset.sequence['time'][iteration].dayofweek == 0
and self.dataset.sequence['time'][iteration].hour == 0
and self.dataset.sequence['time'][iteration].minute == 0)
if params['online_training']:
train = True
if verbose:
print
print "iteration: ", iteration, " time: ", self.dataset.sequence[
'time'][iteration]
if train:
if verbose:
print " train at", iteration, " time: ", self.dataset.sequence[
'time'][iteration]
self.train(params, verbose)
if train:
# reset test counter after training
self.testCounter = params['test_for']
if self.testCounter == 0:
return None
else:
self.testCounter -= 1
symbol = self.networkInput[iteration]
output = self.net.activate(self.inputEncoder.encode(symbol))
if params['output_encoding'] == None:
predictions = self.dataset.reconstructSequence(output[0])
elif params['output_encoding'] == 'likelihood':
predictions = list(output / sum(output))
else:
predictions = None
if verbose:
print " test at :", iteration,
if iteration == params['perturb_after']:
if verbose:
print " perturb data and introduce new patterns"
(newNetworkInput, newTargetPrediction, newTrueData) = \
self.dataset.generateSequence(perturbed=True,
prediction_nstep=params['prediction_nstep'],
output_encoding=params['output_encoding'],
noise=params['noise'])
self.networkInput[iteration + 1:] = newNetworkInput[iteration + 1:]
self.targetPrediction[iteration +
1:] = newTargetPrediction[iteration + 1:]
self.trueData[iteration + 1:] = newTrueData[iteration + 1:]
return {
"current": self.networkInput[iteration],
"reset": None,
"train": train,
"predictions": predictions,
"truth": self.trueData[iteration]
}
