def readContestData(trainingSize=100, testSize=100):
rootdata = 'pacmandata'
rawTrainingData, trainingLabels = samples.loadPacmanData(rootdata + '/contest_training.pkl', trainingSize)
rawValidationData, validationLabels = samples.loadPacmanData(rootdata + '/contest_validation.pkl', testSize)
rawTestData, testLabels = samples.loadPacmanData(rootdata + '/contest_test.pkl', testSize)
trainingData = []
validationData = []
testData = []
return (trainingData, trainingLabels, validationData, validationLabels, rawTrainingData, rawValidationData, testData, testLabels, rawTestData)
def readSuicideData(trainingSize=100, testSize=100):
rootdata = "pacmandata"
rawTrainingData, trainingLabels = samples.loadPacmanData(rootdata + "/suicide_training.pkl", trainingSize)
rawValidationData, validationLabels = samples.loadPacmanData(rootdata + "/suicide_validation.pkl", testSize)
rawTestData, testLabels = samples.loadPacmanData(rootdata + "/suicide_test.pkl", testSize)
trainingData = []
validationData = []
testData = []
return (
trainingData,
trainingLabels,
validationData,
validationLabels,
rawTrainingData,
rawValidationData,
testData,
testLabels,
rawTestData,
)
def runClassifier(args, options):
featureFunction = args['featureFunction']
classifier = args['classifier']
printImage = args['printImage']
# Load data
numTraining = options.training
numTest = options.test
if(options.data=="pacman"):
agentToClone = args.get('agentToClone', None)
trainingData, validationData, testData = MAP_AGENT_TO_PATH_OF_SAVED_GAMES.get(agentToClone, (None, None, None))
trainingData = trainingData or args.get('trainingData', False) or MAP_AGENT_TO_PATH_OF_SAVED_GAMES['ContestAgent'][0]
validationData = validationData or args.get('validationData', False) or MAP_AGENT_TO_PATH_OF_SAVED_GAMES['ContestAgent'][1]
testData = testData or MAP_AGENT_TO_PATH_OF_SAVED_GAMES['ContestAgent'][2]
rawTrainingData, trainingLabels = samples.loadPacmanData(trainingData, numTraining)
rawValidationData, validationLabels = samples.loadPacmanData(validationData, numTest)
rawTestData, testLabels = samples.loadPacmanData(testData, numTest)
else:
rawTrainingData = samples.loadDataFile("digitdata/trainingimages", numTraining,DIGIT_DATUM_WIDTH,DIGIT_DATUM_HEIGHT)
trainingLabels = samples.loadLabelsFile("digitdata/traininglabels", numTraining)
rawValidationData = samples.loadDataFile("digitdata/validationimages", numTest,DIGIT_DATUM_WIDTH,DIGIT_DATUM_HEIGHT)
validationLabels = samples.loadLabelsFile("digitdata/validationlabels", numTest)
rawTestData = samples.loadDataFile("digitdata/testimages", numTest,DIGIT_DATUM_WIDTH,DIGIT_DATUM_HEIGHT)
testLabels = samples.loadLabelsFile("digitdata/testlabels", numTest)
# Extract features
print "Extracting features..."
trainingData = map(featureFunction, rawTrainingData)
validationData = map(featureFunction, rawValidationData)
testData = map(featureFunction, rawTestData)
# Conduct training and testing
print "Training..."
classifier.train(trainingData, trainingLabels, validationData, validationLabels)
print "Validating..."
guesses = classifier.classify(validationData)
correct = [guesses[i] == validationLabels[i] for i in range(len(validationLabels))].count(True)
print str(correct), ("correct out of " + str(len(validationLabels)) + " (%.1f%%).") % (100.0 * correct / len(validationLabels))
print "Testing..."
guesses = classifier.classify(testData)
correct = [guesses[i] == testLabels[i] for i in range(len(testLabels))].count(True)
print str(correct), ("correct out of " + str(len(testLabels)) + " (%.1f%%).") % (100.0 * correct / len(testLabels))
analysis(classifier, guesses, testLabels, testData, rawTestData, printImage)
# do odds ratio computation if specified at command line
if((options.odds) & (options.classifier == "naiveBayes" or (options.classifier == "nb")) ):
label1, label2 = options.label1, options.label2
features_odds = classifier.findHighOddsFeatures(label1,label2)
if(options.classifier == "naiveBayes" or options.classifier == "nb"):
string3 = "=== Features with highest odd ratio of label %d over label %d ===" % (label1, label2)
else:
string3 = "=== Features for which weight(label %d)-weight(label %d) is biggest ===" % (label1, label2)
print string3
printImage(features_odds)
if((options.weights) & (options.classifier == "perceptron")):
for l in classifier.legalLabels:
features_weights = classifier.findHighWeightFeatures(l)
print ("=== Features with high weight for label %d ==="%l)
printImage(features_weights)
def runClassifier(args, options):
featureFunction = args['featureFunction']
classifier = args['classifier']
printImage = args['printImage']
# Load data
numTraining = options.training
numTest = options.test
if (options.data == "pacman"):
agentToClone = args.get('agentToClone', None)
trainingData, validationData, testData = MAP_AGENT_TO_PATH_OF_SAVED_GAMES.get(
agentToClone, (None, None, None))
trainingData = trainingData or args.get(
'trainingData',
False) or MAP_AGENT_TO_PATH_OF_SAVED_GAMES['ContestAgent'][0]
validationData = validationData or args.get(
'validationData',
False) or MAP_AGENT_TO_PATH_OF_SAVED_GAMES['ContestAgent'][1]
testData = testData or MAP_AGENT_TO_PATH_OF_SAVED_GAMES[
'ContestAgent'][2]
rawTrainingData, trainingLabels = samples.loadPacmanData(
trainingData, numTraining)
rawValidationData, validationLabels = samples.loadPacmanData(
validationData, numTest)
rawTestData, testLabels = samples.loadPacmanData(testData, numTest)
else:
rawTrainingData = samples.loadDataFile("digitdata/trainingimages",
numTraining, DIGIT_DATUM_WIDTH,
DIGIT_DATUM_HEIGHT)
trainingLabels = samples.loadLabelsFile("digitdata/traininglabels",
numTraining)
rawValidationData = samples.loadDataFile("digitdata/validationimages",
numTest, DIGIT_DATUM_WIDTH,
DIGIT_DATUM_HEIGHT)
validationLabels = samples.loadLabelsFile("digitdata/validationlabels",
numTest)
rawTestData = samples.loadDataFile("digitdata/testimages", numTest,
DIGIT_DATUM_WIDTH,
DIGIT_DATUM_HEIGHT)
testLabels = samples.loadLabelsFile("digitdata/testlabels", numTest)
# Extract features
print "Extracting features..."
trainingData = map(featureFunction, rawTrainingData)
validationData = map(featureFunction, rawValidationData)
testData = map(featureFunction, rawTestData)
# Conduct training and testing
print "Training..."
classifier.train(trainingData, trainingLabels, validationData,
validationLabels)
print "Validating..."
guesses = classifier.classify(validationData)
correct = [
guesses[i] == validationLabels[i] for i in range(len(validationLabels))
].count(True)
print str(correct), ("correct out of " + str(len(validationLabels)) +
" (%.1f%%).") % (100.0 * correct /
len(validationLabels))
print "Testing..."
guesses = classifier.classify(testData)
correct = [guesses[i] == testLabels[i]
for i in range(len(testLabels))].count(True)
print str(correct), ("correct out of " + str(len(testLabels)) +
" (%.1f%%).") % (100.0 * correct / len(testLabels))
analysis(classifier, guesses, testLabels, testData, rawTestData,
printImage)
# do odds ratio computation if specified at command line
if ((options.odds) & (options.classifier == "naiveBayes" or
(options.classifier == "nb"))):
label1, label2 = options.label1, options.label2
features_odds = classifier.findHighOddsFeatures(label1, label2)
if (options.classifier == "naiveBayes" or options.classifier == "nb"):
string3 = "=== Features with highest odd ratio of label %d over label %d ===" % (
label1, label2)
else:
string3 = "=== Features for which weight(label %d)-weight(label %d) is biggest ===" % (
label1, label2)
print string3
printImage(features_odds)
if ((options.weights) & (options.classifier == "perceptron")):
for l in classifier.legalLabels:
features_weights = classifier.findHighWeightFeatures(l)
print("=== Features with high weight for label %d ===" % l)
printImage(features_weights)
def runClassifier(args, options):
featureFunction = args['featureFunction']
classifier = args['classifier']
printImage = args['printImage']
# Load data
numTraining = options.training
numTest = options.test
if (options.data == "pacman"):
agentToClone = args.get('agentToClone', None)
trainingData, validationData, testData = MAP_AGENT_TO_PATH_OF_SAVED_GAMES.get(
agentToClone, (None, None, None))
trainingData = trainingData or args.get(
'trainingData',
False) or MAP_AGENT_TO_PATH_OF_SAVED_GAMES['ContestAgent'][0]
validationData = validationData or args.get(
'validationData',
False) or MAP_AGENT_TO_PATH_OF_SAVED_GAMES['ContestAgent'][1]
testData = testData or MAP_AGENT_TO_PATH_OF_SAVED_GAMES[
'ContestAgent'][2]
rawTrainingData, trainingLabels = samples.loadPacmanData(
trainingData, numTraining)
rawValidationData, validationLabels = samples.loadPacmanData(
validationData, numTest)
rawTestData, testLabels = samples.loadPacmanData(testData, numTest)
else:
rawTrainingData = samples.loadDataFile("digitdata/trainingimages",
numTraining, DIGIT_DATUM_WIDTH,
DIGIT_DATUM_HEIGHT)
trainingLabels = samples.loadLabelsFile("digitdata/traininglabels",
numTraining)
rawValidationData = samples.loadDataFile("digitdata/validationimages",
numTest, DIGIT_DATUM_WIDTH,
DIGIT_DATUM_HEIGHT)
validationLabels = samples.loadLabelsFile("digitdata/validationlabels",
numTest)
rawTestData = samples.loadDataFile("digitdata/testimages", numTest,
DIGIT_DATUM_WIDTH,
DIGIT_DATUM_HEIGHT)
testLabels = samples.loadLabelsFile("digitdata/testlabels", numTest)
# Extract features
print "Extracting features..."
trainingData = map(featureFunction, rawTrainingData)
validationData = map(featureFunction, rawValidationData)
testData = map(featureFunction, rawTestData)
# Conduct training and testing
print "Training..."
classifier.train(trainingData, trainingLabels, validationData,
validationLabels)
print "Validating..."
guesses = classifier.classify(validationData)
correct = [
guesses[i] == validationLabels[i] for i in range(len(validationLabels))
].count(True)
print str(correct), ("correct out of " + str(len(validationLabels)) +
" (%.1f%%).") % (100.0 * correct /
len(validationLabels))
print "Testing..."
guesses = classifier.classify(testData)
correct = [guesses[i] == testLabels[i]
for i in range(len(testLabels))].count(True)
print str(correct), ("correct out of " + str(len(testLabels)) +
" (%.1f%%).") % (100.0 * correct / len(testLabels))
def runClassifier(args, options):
classifier = args['classifier']
# Load data
if options.data == "pacman":
agentToClone = args.get('agentToClone', None)
trainingData, validationData, testData = MAP_AGENT_TO_PATH_OF_SAVED_GAMES.get(
agentToClone, (None, None, None))
trainingData = trainingData or args.get(
'trainingData',
False) or MAP_AGENT_TO_PATH_OF_SAVED_GAMES['ContestAgent'][0]
validationData = validationData or args.get(
'validationData',
False) or MAP_AGENT_TO_PATH_OF_SAVED_GAMES['ContestAgent'][1]
testData = testData or MAP_AGENT_TO_PATH_OF_SAVED_GAMES[
'ContestAgent'][2]
trainingData, trainingLabels = samples.loadPacmanData(
trainingData, options.training)
validationData, validationLabels = samples.loadPacmanData(
validationData, options.validation)
testData, testLabels = samples.loadPacmanData(testData, None)
elif options.data == "digits":
if options.training is None:
options.training = 2000
if options.validation is None:
options.validation = 1000
numTest = 1000
trainingData = samples.loadDataFile("digitdata/trainingimages",
options.training,
DIGIT_DATUM_WIDTH,
DIGIT_DATUM_HEIGHT)
trainingLabels = samples.loadLabelsFile("digitdata/traininglabels",
options.training)
validationData = samples.loadDataFile("digitdata/validationimages",
options.validation,
DIGIT_DATUM_WIDTH,
DIGIT_DATUM_HEIGHT)
validationLabels = samples.loadLabelsFile("digitdata/validationlabels",
options.validation)
testData = samples.loadDataFile("digitdata/testimages", numTest,
DIGIT_DATUM_WIDTH, DIGIT_DATUM_HEIGHT)
testLabels = samples.loadLabelsFile("digitdata/testlabels", numTest)
else:
raise ValueError('unrecognized dataset %r' % options.data)
# Conduct training and testing
print "Training..."
classifier.train(trainingData, trainingLabels, validationData,
validationLabels)
print "Testing..."
guesses = classifier.classify(testData)
correct = [guesses[i] == testLabels[i]
for i in range(len(testLabels))].count(True)
print str(correct), ("correct out of " + str(len(testLabels)) +
" (%.1f%%).") % (100.0 * correct / len(testLabels))