def test_end_to_end_logistic_regression():
pos, neg = logistic.generate_well_separable(100, 0.50)
#graph_pos_neg(pos, neg)
X = logistic.vstack([pos, neg])
y = logistic.hstack([np.array([1] * len(pos)),
np.array([0] * len(neg)),])
data = logistic.generate_random_points(100,
center=np.array([2,2]),
scale=np.array([5,5]))
#theta = logistic.logistic_gradient_descent(X, y)
thetaC = logistic.fast_logistic_gradient_descent(X, y)
theta = thetaC
#assert np.allclose(theta, thetaC)
labels = logistic.label_data(data, theta, binarize=True)
assert len([l for l in labels if l == 0]) > 10
assert len([l for l in labels if l == 1]) > 10
labels = logistic.label_data(data, thetaC, binarize=True)
assert len([l for l in labels if l == 0]) > 10
assert len([l for l in labels if l == 1]) > 10
small_data = np.array([[-1, -1], [11, 11]])
labels2 = logistic.label_data(small_data, theta, binarize=True)
assert np.allclose([0, 1], labels2)
assert not np.allclose([1, 1], labels2)
labels2 = logistic.label_data(small_data, thetaC, binarize=True)
assert np.allclose([0, 1], labels2)
assert not np.allclose([1, 1], labels2)
ax.scatter(neg[:,0], neg[:,1], s=6, c='r', marker='o', lw=0)
delta = 0.01 * speed_multiple
x, y = np.arange(-8, 8, delta), np.arange(-10, 10, delta)
X, Y = np.meshgrid(x, y)
assert X.shape == Y.shape
shape = X.shape
data = np.hstack([X.flatten().reshape(-1, 1), Y.flatten().reshape(-1,1)])
assert data.shape[0] == (shape[0] * shape[1]) and data.shape[1] == 2
data = add_x2_y2(data)
scaled_data = scaler.transform(data)
# plot the LR on the true labels
labels = logistic.label_data(data, thetaTrue, normalizer=0.0, binarize=False)
labels.shape = shape
CS = pyplot.contour(X, Y, labels, [0.50,], colors='#0000FF')
CS.collections[0].set_label('LR True Labels')
labels = logistic.label_data(data, theta, normalizer=0.0, binarize=False)
labels.shape = shape
CS = pyplot.contour(X, Y, labels, [0.10,], colors='#AAAAFF')
CS.collections[0].set_label('LR Pos-only Labels')
labels = posonly.predict_proba(scaled_data)[:,1]
labels.shape = shape
CS = pyplot.contour(X, Y, labels, [0.50,], colors='#00FF00')
CS.collections[0].set_label('POLR Pos-only Labels')
print 'b: ', b
def predict_proba(self, X):
a = logistic.label_data(X, self.theta_, self.b_**2, binarize=False)
return np.vstack([1.0 - a, a]).T
def predict(self, X):
return logistic.label_data(X, self.theta_, self.b_**2, binarize=True)
x, y = np.arange(-8, 8, delta), np.arange(-10, 10, delta)
X, Y = np.meshgrid(x, y)
assert X.shape == Y.shape
shape = X.shape
data = np.hstack(
[X.flatten().reshape(-1, 1),
Y.flatten().reshape(-1, 1)])
assert data.shape[0] == (shape[0] * shape[1]) and data.shape[1] == 2
data = add_x2_y2(data)
scaled_data = scaler.transform(data)
# plot the LR on the true labels
labels = logistic.label_data(data,
thetaTrue,
normalizer=0.0,
binarize=False)
labels.shape = shape
CS = pyplot.contour(X, Y, labels, [
0.50,
], colors='#0000FF')
CS.collections[0].set_label('LR True Labels')
labels = logistic.label_data(data,
theta,
normalizer=0.0,
binarize=False)
labels.shape = shape
CS = pyplot.contour(X, Y, labels, [
0.10,
], colors='#AAAAFF')