def qda(train_features, train_labels, test_featues, test_labels): from modshogun import QDA, MulticlassAccuracy qda = QDA(train_features, train_labels) qda.train() train_output = qda.apply() test_output = qda.apply(test_features) evaluator = MulticlassAccuracy() print 'QDA training error is %.4f' % ((1-evaluator.evaluate(train_output, train_labels))*100) print 'QDA test error is %.4f' % ((1-evaluator.evaluate(test_output, test_labels))*100)
def qda(train_features, train_labels, test_featues, test_labels):
from modshogun import QDA, MulticlassAccuracy
qda = QDA(train_features, train_labels)
qda.train()
train_output = qda.apply()
test_output = qda.apply(test_features)
evaluator = MulticlassAccuracy()
print 'QDA training error is %.4f' % (
(1 - evaluator.evaluate(train_output, train_labels)) * 100)
print 'QDA test error is %.4f' % (
(1 - evaluator.evaluate(test_output, test_labels)) * 100)
def classifier_qda_modular (train_fname=traindat, test_fname=testdat, label_fname=label_traindat, tolerance=1e-4, store_covs=False): try: from modshogun import RealFeatures, MulticlassLabels, QDA, CSVFile except ImportError: return feats_train=RealFeatures(CSVFile(train_fname)) feats_test=RealFeatures(CSVFile(test_fname)) labels=MulticlassLabels(CSVFile(label_fname)) qda = QDA(feats_train, labels, tolerance, store_covs) qda.train() qda.apply(feats_test).get_labels() qda.set_features(feats_test) return qda, qda.apply().get_labels()
def classifier_qda_modular(train_fname=traindat,
test_fname=testdat,
label_fname=label_traindat,
tolerance=1e-4,
store_covs=False):
try:
from modshogun import RealFeatures, MulticlassLabels, QDA, CSVFile
except ImportError:
return
feats_train = RealFeatures(CSVFile(train_fname))
feats_test = RealFeatures(CSVFile(test_fname))
labels = MulticlassLabels(CSVFile(label_fname))
qda = QDA(feats_train, labels, tolerance, store_covs)
qda.train()
qda.apply(feats_test).get_labels()
qda.set_features(feats_test)
return qda, qda.apply().get_labels()
pylab.contour(xx, yy, Z, linewidths = 3, colors = 'k')
# Number of classes
M = 3
# Number of samples of each class
N = 300
# Dimension of the data
dim = 2
cols = ['blue', 'green', 'red']
fig = pylab.figure()
ax = fig.add_subplot(111)
pylab.title('Quadratic Discrimant Analysis')
X, y = gen_data()
labels = MulticlassLabels(y)
features = RealFeatures(X.T)
qda = QDA(features, labels, 1e-4, True)
qda.train()
ypred = qda.apply().get_labels()
plot_data(qda, X, y, ypred, ax)
for i in range(M):
plot_cov(ax, qda.get_mean(i), qda.get_cov(i), cols[i])
plot_regions(qda)
pylab.connect('key_press_event', util.quit)
pylab.show()
qda.set_labels(labels)
qda.train(features)
# compute output plot iso-lines
xs = np.array(np.concatenate([x_pos, x_neg]))
ys = np.array(np.concatenate([y_pos, y_neg]))
x1_max = max(1.2*xs)
x1_min = min(1.2*xs)
x2_max = max(1.2*ys)
x2_min = min(1.2*ys)
x1 = np.linspace(x1_min, x1_max, size)
x2 = np.linspace(x2_min, x2_max, size)
x, y = np.meshgrid(x1, x2)
dense = RealFeatures( np.array((np.ravel(x), np.ravel(y))) )
dense_labels = qda.apply(dense).get_labels()
z = dense_labels.reshape((size, size))
pcolor(x, y, z)
contour(x, y, z, linewidths = 1, colors = 'black', hold = True)
axis([x1_min, x1_max, x2_min, x2_max])
connect('key_press_event', util.quit)
show()
# Number of classes
M = 3
# Number of samples of each class
N = 300
# Dimension of the data
dim = 2
cols = ['blue', 'green', 'red']
fig = pylab.figure()
ax = fig.add_subplot(111)
pylab.title('Quadratic Discrimant Analysis')
X, y = gen_data()
labels = MulticlassLabels(y)
features = RealFeatures(X.T)
qda = QDA(features, labels, 1e-4, True)
qda.train()
ypred = qda.apply().get_labels()
plot_data(qda, X, y, ypred, ax)
for i in range(M):
plot_cov(ax, qda.get_mean(i), qda.get_cov(i), cols[i])
plot_regions(qda)
pylab.connect('key_press_event', util.quit)
pylab.show()
