def test_latent_node_boxes_latent_subgradient():
# same as above, now with elementary subgradients
# learn the "easy" 2x2 boxes dataset.
# a 2x2 box is placed randomly in a 4x4 grid
# we add a latent variable for each 2x2 patch
# that should make the model fairly simple
X, Y = toy.make_simple_2x2(seed=1)
latent_crf = LatentNodeCRF(n_labels=2, inference_method='lp',
n_hidden_states=2, n_features=1)
latent_svm = LatentSubgradientSSVM(model=latent_crf, max_iter=250, C=10,
verbose=10, learning_rate=0.1,
momentum=0)
G = [make_grid_edges(x) for x in X]
# make edges for hidden states:
edges = []
node_indices = np.arange(4 * 4).reshape(4, 4)
for i, (x, y) in enumerate(itertools.product([0, 2], repeat=2)):
for j in xrange(x, x + 2):
for k in xrange(y, y + 2):
edges.append([i + 4 * 4, node_indices[j, k]])
G = [np.vstack([make_grid_edges(x), edges]) for x in X]
# reshape / flatten x and y
X_flat = [x.reshape(-1, 1) for x in X]
Y_flat = [y.ravel() for y in Y]
X_ = zip(X_flat, G, [4 * 4 for x in X_flat])
latent_svm.fit(X_, Y_flat)
assert_equal(latent_svm.score(X_, Y_flat), 1)
def main():
X, Y = toy.generate_crosses(n_samples=20, noise=5, n_crosses=1,
total_size=8)
X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=.5)
n_labels = len(np.unique(Y_train))
crf = LatentGridCRF(n_labels=n_labels, n_states_per_label=[1, 2],
inference_method='lp')
#clf = LatentSSVM(model=crf, max_iter=500, C=1000., verbose=2,
#check_constraints=True, n_jobs=-1, break_on_bad=True,
#base_svm='1-slack', inference_cache=20, tol=.1)
clf = LatentSubgradientSSVM(
model=crf, max_iter=500, C=1000., verbose=2,
n_jobs=-1, learning_rate=0.1, show_loss_every=10)
clf.fit(X_train, Y_train)
#for X_, Y_, H, name in [[X_train, Y_train, clf.H_init_, "train"],
#[X_test, Y_test, [None] * len(X_test), "test"]]:
for X_, Y_, H, name in [[X_train, Y_train, [None] * len(X_test), "train"],
[X_test, Y_test, [None] * len(X_test), "test"]]:
Y_pred = clf.predict(X_)
i = 0
loss = 0
for x, y, h_init, y_pred in zip(X_, Y_, H, Y_pred):
loss += np.sum(y != y_pred)
fig, ax = plt.subplots(3, 2)
ax[0, 0].matshow(y, vmin=0, vmax=crf.n_labels - 1)
ax[0, 0].set_title("ground truth")
ax[0, 1].matshow(np.argmax(x, axis=-1),
vmin=0, vmax=crf.n_labels - 1)
ax[0, 1].set_title("unaries only")
if h_init is None:
ax[1, 0].set_visible(False)
else:
ax[1, 0].matshow(h_init, vmin=0, vmax=crf.n_states - 1)
ax[1, 0].set_title("latent initial")
ax[1, 1].matshow(crf.latent(x, y, clf.w),
vmin=0, vmax=crf.n_states - 1)
ax[1, 1].set_title("latent final")
ax[2, 0].matshow(crf.inference(x, clf.w),
vmin=0, vmax=crf.n_states - 1)
ax[2, 0].set_title("prediction latent")
ax[2, 1].matshow(y_pred,
vmin=0, vmax=crf.n_labels - 1)
ax[2, 1].set_title("prediction")
for a in ax.ravel():
a.set_xticks(())
a.set_yticks(())
fig.savefig("data_%s_%03d.png" % (name, i), bbox_inches="tight")
i += 1
print("loss %s set: %f" % (name, loss))
print(clf.w)
def test_directional_bars():
for inference_method in ['lp']:
X, Y = toy.generate_easy(n_samples=10, noise=5, box_size=2,
total_size=6, seed=1)
n_labels = 2
crf = LatentDirectionalGridCRF(n_labels=n_labels,
n_states_per_label=[1, 4],
inference_method=inference_method)
clf = LatentSubgradientSSVM(model=crf, max_iter=500, C=10. ** 5,
verbose=2)
clf.fit(X, Y)
Y_pred = clf.predict(X)
assert_array_equal(np.array(Y_pred), Y)
def test_directional_bars():
# this test is very fragile :-/
X, Y = toy.generate_easy(n_samples=20, noise=2, box_size=2,
total_size=6, seed=2)
n_labels = 2
crf = LatentDirectionalGridCRF(n_labels=n_labels,
n_states_per_label=[1, 4])
clf = LatentSubgradientSSVM(model=crf, max_iter=75, C=10.,
learning_rate=1, momentum=0,
decay_exponent=0.5, decay_t0=10)
clf.fit(X, Y)
Y_pred = clf.predict(X)
assert_array_equal(np.array(Y_pred), Y)
def test_objective():
# test that LatentSubgradientSSVM does the same as SubgradientSVM,
# in particular that it has the same loss, if there are no latent states.
X, Y = toy.generate_blocks_multinomial(n_samples=10)
n_labels = 3
crfl = LatentGridCRF(n_labels=n_labels, n_states_per_label=1)
clfl = LatentSubgradientSSVM(model=crfl, max_iter=50, C=10.,
learning_rate=0.001, momentum=0.98,
decay_exponent=0)
clfl.w = np.zeros(crfl.size_psi) # this disables random init
clfl.fit(X, Y)
crf = GridCRF(n_states=n_labels)
clf = SubgradientSSVM(model=crf, max_iter=50, C=10.,
learning_rate=0.001, momentum=0.98, decay_exponent=0)
clf.fit(X, Y)
assert_array_almost_equal(clf.w, clfl.w)
assert_array_equal(clf.predict(X), Y)
assert_almost_equal(clf.objective_curve_[-1], clfl.objective_curve_[-1])
def test_with_crosses():
# very simple dataset. k-means init is perfect
for n_states_per_label in [2, [1, 2]]:
# test with 2 states for both foreground and background,
# as well as with single background state
#for inference_method in ['ad3', 'qpbo', 'lp']:
for inference_method in ['lp']:
X, Y = toy.generate_crosses(n_samples=10, noise=5, n_crosses=1,
total_size=8)
n_labels = 2
crf = LatentGridCRF(n_labels=n_labels,
n_states_per_label=n_states_per_label,
inference_method=inference_method)
clf = LatentSubgradientSSVM(model=crf, max_iter=250, C=10. ** 5,
verbose=20, learning_rate=0.0001,
show_loss_every=10, momentum=0.98,
decay_exponent=0)
clf.fit(X, Y)
Y_pred = clf.predict(X)
assert_array_equal(np.array(Y_pred), Y)
def test_latent_node_boxes_latent_subgradient():
# same as above, now with elementary subgradients
X, Y = toy.make_simple_2x2(seed=1)
latent_crf = LatentNodeCRF(n_labels=2, n_hidden_states=2, n_features=1)
latent_svm = LatentSubgradientSSVM(model=latent_crf, max_iter=250, C=10,
learning_rate=0.1, momentum=0)
G = [make_grid_edges(x) for x in X]
edges = make_edges_2x2()
G = [np.vstack([make_grid_edges(x), edges]) for x in X]
# reshape / flatten x and y
X_flat = [x.reshape(-1, 1) for x in X]
Y_flat = [y.ravel() for y in Y]
X_ = zip(X_flat, G, [4 * 4 for x in X_flat])
latent_svm.fit(X_, Y_flat)
assert_equal(latent_svm.score(X_, Y_flat), 1)
X_train_, X_test_, X_train, X_test, y_train, y_test, y_org_train, y_org_test =\
train_test_split(X_, X, Y, y_org, test_size=.5)
# first, do it with a standard CRF / SVM
pbl = GraphCRF(n_features=64, n_states=2, inference_method='lp')
svm = StructuredSVM(pbl, verbose=1, check_constraints=True, C=1000, n_jobs=1,
batch_size=-1)
svm.fit(X_train_, y_train)
y_pred = np.vstack(svm.predict(X_test_))
print("Score with pystruct crf svm: %f" % np.mean(y_pred == y_test))
print(svm.score(X_train_, y_train))
print(svm.score(X_test_, y_test))
# now with latent CRF SVM
latent_pbl = LatentGraphCRF(n_features=64, n_labels=2, n_states_per_label=5,
inference_method='dai')
latent_svm = LatentSubgradientSSVM(model=latent_pbl, max_iter=5000, C=1,
verbose=2, n_jobs=1, learning_rate=0.1,
show_loss_every=10, momentum=0.0,
decay_exponent=0.5)
#latent_svm = LatentSSVM(latent_pbl, verbose=2, check_constraints=True, C=100,
#n_jobs=1, batch_size=-1, tol=.1, latent_iter=2)
latent_svm.fit(X_train_, y_train)
print(latent_svm.score(X_train_, y_train))
print(latent_svm.score(X_test_, y_test))
h_pred = np.hstack(latent_svm.predict_latent(X_test_))
print("Latent class counts: %s" % repr(np.bincount(h_pred)))
G = [make_grid_edges(x) for x in X]
asdf = zip(X_flat, G)
svm.fit(asdf, Y_flat)
plot_boxes(svm.predict(asdf))
print("Training score multiclass svm CRF: %f" % svm.score(asdf, Y_flat))
# using one latent variable for each 2x2 rectangle
latent_crf = LatentNodeCRF(n_labels=2, n_features=1, inference_method='lp',
n_hidden_states=2)
#latent_svm = LatentSSVM(model=latent_crf, max_iter=200, C=10, verbose=10,
#check_constraints=True, break_on_bad=True, n_jobs=1,
#latent_iter=10, base_svm='subgradient', tol=-1,
#inactive_window=0, learning_rate=0.01, momentum=0)
latent_svm = LatentSubgradientSSVM(model=latent_crf, max_iter=200, C=100,
verbose=1, n_jobs=1, show_loss_every=10,
learning_rate=0.01, momentum=0)
# make edges for hidden states:
edges = []
node_indices = np.arange(4 * 4).reshape(4, 4)
for i, (x, y) in enumerate(itertools.product([0, 2], repeat=2)):
for j in xrange(x, x + 2):
for k in xrange(y, y + 2):
edges.append([i + 4 * 4, node_indices[j, k]])
G = [np.vstack([make_grid_edges(x), edges]) for x in X]
#G = [make_grid_edges(x) for x in X]
#H_init = [np.hstack([y.ravel(), 2 + y[1: -1, 1: -1].ravel()]) for y in Y]
H_init = [np.hstack([y.ravel(), np.random.randint(2, 4, size=2 * 2)]) for y in