def main():
if len(sys.argv) != 5:
print "<python Init.py TrainingData.txt TrainingTarget.txt TestingData.txt TestingTarget.txt>"
sys.exit()
if not os.path.isfile(sys.argv[1]) or not os.path.isfile(
sys.argv[2]) or not os.path.isfile(
sys.argv[3]) or not os.path.isfile(sys.argv[4]):
print "File does not exist"
print "<python Init.py TrainingData.txt TrainingTarget.txt TestingData.txt TestingTarget.txt>"
sys.exit()
#calculating number of features
firstline = open(sys.argv[1]).readline()
first_example = [
numbers for numbers in firstline.split("\t") if numbers is not ""
]
if len(first_example) < 1:
print "There must be more than one feature in the data."
else:
num_features = len(first_example)
#reading from training file
training_features, training_targets, testing_features, testing_targets = FileReader.readFile(
sys.argv[1], sys.argv[2], sys.argv[3], sys.argv[4], num_features)
assert len(training_features) == len(training_targets)
assert len(testing_features) == len(testing_targets)
learning_rate = float(raw_input("What is the learning rate? "))
assert learning_rate > 0 and learning_rate < 1
num_iterations = input("How many iterations? ")
assert num_iterations > 0
numHidLayers = input("How many HIDDEN LAYERS? ")
assert isinstance(numHidLayers, int)
numNeuronsInLayer = num_features
numNeuronsOutLayer = len(testing_targets[0])
numNeuronsHidLayers = []
for i in range(0, numHidLayers):
numNeuronsHidLayers.append(
input("How many Neurons Hidden Layer " + str(i + 1) + " : "))
assert isinstance(numNeuronsHidLayers[i], int)
tic = time.time()
numNeurons = []
numNeurons.append(numNeuronsInLayer)
numNeurons.extend(numNeuronsHidLayers)
numNeurons.append(numNeuronsOutLayer)
allWeights = PerceptronLearning.initializeWeights(numNeurons)
allBiases = PerceptronLearning.initializeBiases(numNeurons)
for i in range(0, num_iterations):
predicted_targets = []
for j in range(0, len(training_features)):
allOutputs = PerceptronLearning.forwardProp(
training_features[j], allWeights, allBiases)
predicted_targets.append(allOutputs[-1])
allDeltas = PerceptronLearning.backwardProp(
training_targets[j], predicted_targets[j], allWeights)
allBiases = PerceptronLearning.updateBiases(
allDeltas, allBiases, learning_rate, len(training_features))
allWeights = PerceptronLearning.updateWeights(
allOutputs, allDeltas, allWeights, learning_rate,
len(training_features))
if i % 100 == 99:
error = PerceptronLearning.calculateError(training_targets,
predicted_targets)
print "Error at iteration", i, "is", np.squeeze(error)
predicted_test_targets_arrs = []
for i in range(0, len(testing_features)):
allOutputs = PerceptronLearning.forwardProp(testing_features[i],
allWeights, allBiases)
predicted_test_targets_arrs.append(allOutputs[-1])
testing_targets = Conversions.convertArrsToInts(testing_targets)
predicted_test_targets = Conversions.convertArrsToInts(
predicted_test_targets_arrs)
# print "original"
# print testing_targets
# print "predicted"
# print predicted_test_targets
assert len(testing_targets) == len(predicted_test_targets)
correct = 0
total = 0
for i in range(0, len(testing_targets)):
total += 1
if testing_targets[i] == predicted_test_targets[i]:
correct += 1
print "The accuracy is", correct / float(total)
toc = time.time()
print "Running Time :", (toc - tic)
return 0
