def test_mds123(save_data=False):
Y1 = np.load('examples/123/true1.npy')
D1 = distances.coord2dist(Y1)
Y2 = np.load('examples/123/true2.npy')
D2 = distances.coord2dist(Y2)
Y3 = np.load('examples/123/true3.npy')
D3 = distances.coord2dist(Y3)
D = [D1, D2, D3]
Q = np.load('examples/123/params.npy')
proj = perspective.Proj()
proj.set_params_list(params_list=Q)
mv = Multiview(D, persp=proj)
mv.setup_visualization('mds')
mv.initialize_X(verbose=1)
mv.optimize_X(algorithm='cdm', learning_rate=0.005, iterations=300)
mv.figureX()
mv.figureY()
plt.show()
if save_data:
np.save('examples/123/computed123.npy', mv.X)
np.save('examples/123/computed1.npy', mv.Y[0])
np.save('examples/123/computed2.npy', mv.Y[1])
np.save('examples/123/computed3.npy', mv.Y[2])
def noise_all(N=100):
noise_levels = [0.001,0.01,0.07,0.15,0.4]
stress = []
X = misc.disk(N,dim=3)
proj = perspective.Proj(dimX=2,dimY=2)
proj.set_params_list(special='identity',number=3)
Y = proj.project(X)
D = distances.compute(Y)
for noise in noise_levels:
D_noisy = distances.add_noise(D,noise)
mv = Multiview(D_noisy,persp=proj)
mv.setup_visualization(visualization='mds')
mv.initialize_X(verbose=1)
mv.optimize_X(algorithm='gd',learning_rate=1,max_iters=300,
verbose=1)
stress.append(mv.cost)
fig = plt.figure()
plt.semilogx(noise_levels,stress)
plt.show()
def noisy(N=100):
noise_levels = [0.0001, 0.001, 0.01, 0.1, 0.5]
stress = []
X = misc.disk(N, dim=3)
proj = perspective.Proj()
proj.set_params_list(special='standard')
Y = proj.project(X)
D = distances.compute(Y)
for noise in noise_levels:
D_noisy = distances.add_noise(D, noise)
stress_best = []
for i in range(3):
mv = Multiview(D_noisy, persp=proj, verbose=1)
mv.setup_visualization(visualization='mds')
mv.initialize_X()
mv.optimize_X(algorithm='agd')
stress_best.append(mv.normalized_cost)
stress.append(min(stress_best))
fig = plt.figure()
plt.loglog(noise_levels, stress, linestyle='--', marker='o')
plt.title('Normalized total stress')
plt.xlabel('noise level')
plt.ylabel('total stress')
plt.show()