Python inference.get_installed函数代码示例

def test_edge_feature_latent_node_crf_no_latent():
    # no latent nodes

    # Test inference with different weights in different directions

    X, Y = generate_blocks_multinomial(noise=2, n_samples=1, seed=1, size_x=10)
    x, y = X[0], Y[0]
    n_states = x.shape[-1]

    edge_list = make_grid_edges(x, 4, return_lists=True)
    edges = np.vstack(edge_list)

    pw_horz = -1 * np.eye(n_states + 5)
    xx, yy = np.indices(pw_horz.shape)
    # linear ordering constraint horizontally
    pw_horz[xx > yy] = 1

    # high cost for unequal labels vertically
    pw_vert = -1 * np.eye(n_states + 5)
    pw_vert[xx != yy] = 1
    pw_vert *= 10

    # generate edge weights
    edge_weights_horizontal = np.repeat(pw_horz[np.newaxis, :, :],
                                        edge_list[0].shape[0], axis=0)
    edge_weights_vertical = np.repeat(pw_vert[np.newaxis, :, :],
                                      edge_list[1].shape[0], axis=0)
    edge_weights = np.vstack([edge_weights_horizontal, edge_weights_vertical])

    # do inference
    # pad x for hidden states...
    x_padded = -100 * np.ones((x.shape[0], x.shape[1], x.shape[2] + 5))
    x_padded[:, :, :x.shape[2]] = x
    res = lp_general_graph(-x_padded.reshape(-1, n_states + 5), edges,
                           edge_weights)

    edge_features = edge_list_to_features(edge_list)
    x = (x.reshape(-1, n_states), edges, edge_features, 0)
    y = y.ravel()

    for inference_method in get_installed(["lp"]):
        # same inference through CRF inferface
        crf = EdgeFeatureLatentNodeCRF(n_labels=3,
                                       inference_method=inference_method,
                                       n_edge_features=2, n_hidden_states=5)
        w = np.hstack([np.eye(3).ravel(), -pw_horz.ravel(), -pw_vert.ravel()])
        y_pred = crf.inference(x, w, relaxed=True)
        assert_array_almost_equal(res[0], y_pred[0].reshape(-1, n_states + 5),
                                  4)
        assert_array_almost_equal(res[1], y_pred[1], 4)
        assert_array_equal(y, np.argmax(y_pred[0], axis=-1))

    for inference_method in get_installed(["lp", "ad3", "qpbo"]):
        # again, this time discrete predictions only
        crf = EdgeFeatureLatentNodeCRF(n_labels=3,
                                       inference_method=inference_method,
                                       n_edge_features=2, n_hidden_states=5)
        w = np.hstack([np.eye(3).ravel(), -pw_horz.ravel(), -pw_vert.ravel()])
        y_pred = crf.inference(x, w, relaxed=False)
        assert_array_equal(y, y_pred)

def test_edge_type_graph_crf():
    # create two samples with different graphs
    # two states only, pairwise smoothing

    # all edges are of the first type. should do the same as GraphCRF
    # if we make w symmetric
    for inference_method in get_installed(['qpbo', 'lp', 'ad3', 'dai', 'ogm']):
        crf = EdgeTypeGraphCRF(n_states=2, inference_method=inference_method,
                               n_edge_types=1)
        assert_array_equal(crf.inference((x_1, [g_1]), w_sym), y_1)
        assert_array_equal(crf.inference((x_2, [g_2]), w_sym), y_2)

    # same, only with two edge types and no edges of second type
    w_sym_ = np.array([1, 0,    # unary
                      0, 1,
                      .22, 0,  # pairwise
                      0, .22,
                      2, -1,   # second edge type, doesn't exist
                      -1, 3])
    for inference_method in get_installed(['qpbo', 'lp', 'ad3', 'dai', 'ogm']):
        crf = EdgeTypeGraphCRF(n_states=2, inference_method=inference_method,
                               n_edge_types=2)
        assert_array_equal(crf.inference((x_1,
                                          [g_1, np.zeros((0, 2),
                                                         dtype=np.int)]),
                                         w_sym_), y_1)
        assert_array_equal(crf.inference((x_2, [g_2, np.zeros((0, 2),
                                                              dtype=np.int)]),
                                         w_sym_), y_2)

    print crf.get_pairwise_potentials((x_2, [g_2, np.zeros((0, 2),
                                                           dtype=np.int)]),
                                      w_sym_)

def test_switch_to_ad3():
    # test if switching between qpbo and ad3 works

    if not get_installed(['qpbo']) or not get_installed(['ad3']):
        return
    X, Y = toy.generate_blocks_multinomial(n_samples=5, noise=1.5,
                                           seed=0)
    crf = GridCRF(n_states=3, inference_method='qpbo')

    ssvm = NSlackSSVM(crf, max_iter=10000)

    ssvm_with_switch = NSlackSSVM(crf, max_iter=10000, switch_to=('ad3'))
    ssvm.fit(X, Y)
    ssvm_with_switch.fit(X, Y)
    assert_equal(ssvm_with_switch.model.inference_method, 'ad3')
    # we check that the dual is higher with ad3 inference
    # as it might use the relaxation, that is pretty much guraranteed
    assert_greater(ssvm_with_switch.objective_curve_[-1],
                   ssvm.objective_curve_[-1])
    print(ssvm_with_switch.objective_curve_[-1], ssvm.objective_curve_[-1])

    # test that convergence also results in switch
    ssvm_with_switch = NSlackSSVM(crf, max_iter=10000, switch_to=('ad3'),
                                  tol=10)
    ssvm_with_switch.fit(X, Y)
    assert_equal(ssvm_with_switch.model.inference_method, 'ad3')

def test_switch_to_ad3():
    # smoketest only
    # test if switching between qpbo and ad3 works inside latent svm
    # use less perfect initialization

    if not get_installed(['qpbo']) or not get_installed(['ad3']):
        return
    X, Y = generate_crosses(n_samples=20, noise=5, n_crosses=1, total_size=8)
    X_test, Y_test = X[10:], Y[10:]
    X, Y = X[:10], Y[:10]

    crf = LatentGridCRF(n_states_per_label=2,
                        inference_method='qpbo')
    crf.initialize(X, Y)
    H_init = crf.init_latent(X, Y)

    np.random.seed(0)
    mask = np.random.uniform(size=H_init.shape) > .7
    H_init[mask] = 2 * (H_init[mask] / 2)

    base_ssvm = OneSlackSSVM(crf, inactive_threshold=1e-8, cache_tol=.0001,
                             inference_cache=50, max_iter=10000,
                             switch_to=('ad3', {'branch_and_bound': True}),
                             C=10. ** 3)
    clf = LatentSSVM(base_ssvm)

    clf.fit(X, Y, H_init=H_init)
    assert_equal(clf.model.inference_method[0], 'ad3')

    Y_pred = clf.predict(X)

    assert_array_equal(np.array(Y_pred), Y)
    # test that score is not always 1
    assert_true(.98 < clf.score(X_test, Y_test) < 1)

def test_inference():
    # Test inference with different weights in different directions

    X, Y = toy.generate_blocks_multinomial(noise=2, n_samples=1, seed=1)
    x, y = X[0], Y[0]
    n_states = x.shape[-1]

    edge_list = make_grid_edges(x, 4, return_lists=True)
    edges = np.vstack(edge_list)

    pw_horz = -1 * np.eye(n_states)
    xx, yy = np.indices(pw_horz.shape)
    # linear ordering constraint horizontally
    pw_horz[xx > yy] = 1

    # high cost for unequal labels vertically
    pw_vert = -1 * np.eye(n_states)
    pw_vert[xx != yy] = 1
    pw_vert *= 10

    # generate edge weights
    edge_weights_horizontal = np.repeat(pw_horz[np.newaxis, :, :],
                                        edge_list[0].shape[0], axis=0)
    edge_weights_vertical = np.repeat(pw_vert[np.newaxis, :, :],
                                      edge_list[1].shape[0], axis=0)
    edge_weights = np.vstack([edge_weights_horizontal, edge_weights_vertical])

    # do inference
    res = lp_general_graph(-x.reshape(-1, n_states), edges, edge_weights)

    edge_features = edge_list_to_features(edge_list)
    x = (x.reshape(-1, n_states), edges, edge_features)
    y = y.ravel()

    for inference_method in get_installed(["lp", "ad3"]):
        # same inference through CRF inferface
        crf = EdgeFeatureGraphCRF(n_states=3,
                                  inference_method=inference_method,
                                  n_edge_features=2)
        w = np.hstack([np.eye(3).ravel(), -pw_horz.ravel(), -pw_vert.ravel()])
        y_pred = crf.inference(x, w, relaxed=True)
        if isinstance(y_pred, tuple):
            # ad3 produces an integer result if it found the exact solution
            assert_array_almost_equal(res[1], y_pred[1])
            assert_array_almost_equal(res[0], y_pred[0].reshape(-1, n_states))
            assert_array_equal(y, np.argmax(y_pred[0], axis=-1))

    for inference_method in get_installed(["lp", "ad3", "qpbo"]):
        # again, this time discrete predictions only
        crf = EdgeFeatureGraphCRF(n_states=3,
                                  inference_method=inference_method,
                                  n_edge_features=2)
        w = np.hstack([np.eye(3).ravel(), -pw_horz.ravel(), -pw_vert.ravel()])
        y_pred = crf.inference(x, w, relaxed=False)
        assert_array_equal(y, y_pred)

def test_switch_to_ad3():
    # smoketest only
    # test if switching between qpbo and ad3 works inside latent svm
    # use less perfect initialization

    if not get_installed(["qpbo"]) or not get_installed(["ad3"]):
        return
    X, Y = generate_crosses(n_samples=20, noise=5, n_crosses=1, total_size=8)
    X_test, Y_test = X[10:], Y[10:]
    X, Y = X[:10], Y[:10]

    crf = LatentGridCRF(n_states_per_label=2, inference_method="qpbo")
    crf.initialize(X, Y)
    H_init = crf.init_latent(X, Y)

    np.random.seed(0)
    mask = np.random.uniform(size=H_init.shape) > 0.7
    H_init[mask] = 2 * (H_init[mask] / 2)

    base_ssvm = OneSlackSSVM(
        crf,
        inactive_threshold=1e-8,
        cache_tol=0.0001,
        inference_cache=50,
        max_iter=10000,
        switch_to=("ad3", {"branch_and_bound": True}),
        C=10.0 ** 3,
    )
    clf = LatentSSVM(base_ssvm)

    # evil hackery to get rid of ad3 output
    try:
        devnull = open("/dev/null", "w")
        oldstdout_fno = os.dup(sys.stdout.fileno())
        os.dup2(devnull.fileno(), 1)
        replaced_stdout = True
    except:
        replaced_stdout = False

    clf.fit(X, Y, H_init=H_init)

    if replaced_stdout:
        os.dup2(oldstdout_fno, 1)
    assert_equal(clf.model.inference_method[0], "ad3")

    Y_pred = clf.predict(X)

    assert_array_equal(np.array(Y_pred), Y)
    # test that score is not always 1
    assert_true(0.98 < clf.score(X_test, Y_test) < 1)

def test_psi_continuous():
    # FIXME
    # first make perfect prediction, including pairwise part
    X, Y = toy.generate_blocks_multinomial(noise=2, n_samples=1, seed=1)
    x, y = X[0], Y[0]
    n_states = x.shape[-1]
    edge_list = make_grid_edges(x, 4, return_lists=True)
    edges = np.vstack(edge_list)
    edge_features = edge_list_to_features(edge_list)
    x = (x.reshape(-1, 3), edges, edge_features)
    y = y.ravel()

    pw_horz = -1 * np.eye(n_states)
    xx, yy = np.indices(pw_horz.shape)
    # linear ordering constraint horizontally
    pw_horz[xx > yy] = 1

    # high cost for unequal labels vertically
    pw_vert = -1 * np.eye(n_states)
    pw_vert[xx != yy] = 1
    pw_vert *= 10

    # create crf, assemble weight, make prediction
    for inference_method in get_installed(["lp", "ad3"]):
        crf = EdgeFeatureGraphCRF(n_states=3,
                                  inference_method=inference_method,
                                  n_edge_features=2)
        w = np.hstack([np.eye(3).ravel(), -pw_horz.ravel(), -pw_vert.ravel()])
        y_pred = crf.inference(x, w, relaxed=True)

        # compute psi for prediction
        psi_y = crf.psi(x, y_pred)
        assert_equal(psi_y.shape, (crf.size_psi,))

def test_chain():
    # test LP, AD3, AD3-BB and JT on a chain.
    # they should all be exact
    rnd = np.random.RandomState(0)
    algorithms = get_installed([('ad3', {'branch_and_bound': False}),
                                ('ad3', {'branch_and_bound': True}),
                                ('ogm', {'alg': 'dyn'}),
                                ('ogm', {'alg': 'dd'}),
                                ('ogm', {'alg': 'trw'})])
    n_states = 3
    n_nodes = 10

    for i in xrange(10):
        forward = np.c_[np.arange(n_nodes - 1), np.arange(1, n_nodes)]
        backward = np.c_[np.arange(1, n_nodes), np.arange(n_nodes - 1)]
        unary_potentials = rnd.normal(size=(n_nodes, n_states))
        pairwise_potentials = rnd.normal(size=(n_states, n_states))
        # test that reversing edges is same as transposing pairwise potentials
        y_forward = inference_dispatch(unary_potentials, pairwise_potentials,
                                       forward, 'lp')
        y_backward = inference_dispatch(unary_potentials,
                                        pairwise_potentials.T, backward, 'lp')
        assert_array_equal(y_forward, y_backward)
        for chain in [forward, backward]:
            y_lp = inference_dispatch(unary_potentials, pairwise_potentials,
                                      chain, 'lp')
            for alg in algorithms:
                if chain is backward and alg[0] == 'ogm':
                    # ogm needs sorted indices
                    continue
                y = inference_dispatch(unary_potentials, pairwise_potentials,
                                       chain, alg)
                assert_array_equal(y, y_lp)

def test_energy():
    # make sure that energy as computed by ssvm is the same as by lp
    np.random.seed(0)
    for inference_method in get_installed(["lp", "ad3"]):
        found_fractional = False
        crf = DirectionalGridCRF(n_states=3, n_features=3,
                                 inference_method=inference_method)
        while not found_fractional:
            x = np.random.normal(size=(7, 8, 3))
            unary_params = np.random.normal(size=(3, 3))
            pw1 = np.random.normal(size=(3, 3))
            pw2 = np.random.normal(size=(3, 3))
            w = np.hstack([unary_params.ravel(), pw1.ravel(), pw2.ravel()])
            res, energy = crf.inference(x, w, relaxed=True, return_energy=True)
            found_fractional = np.any(np.max(res[0], axis=-1) != 1)

            joint_feature = crf.joint_feature(x, res)
            energy_svm = np.dot(joint_feature, w)

            assert_almost_equal(energy, -energy_svm)
            if not found_fractional:
                # exact discrete labels, test non-relaxed version
                res, energy = crf.inference(x, w, relaxed=False,
                                            return_energy=True)
                joint_feature = crf.joint_feature(x, res)
                energy_svm = np.dot(joint_feature, w)

                assert_almost_equal(energy, -energy_svm)

def test_binary_blocks_cutting_plane():
    #testing cutting plane ssvm on easy binary dataset
    # generate graphs explicitly for each example
    for inference_method in get_installed(["lp", "qpbo", "ad3", 'ogm']):
        X, Y = generate_blocks(n_samples=3)
        crf = GraphCRF(inference_method=inference_method)
        clf = NSlackSSVM(model=crf, max_iter=20, C=100, check_constraints=True,
                         break_on_bad=False, n_jobs=1)
        x1, x2, x3 = X
        y1, y2, y3 = Y
        n_states = len(np.unique(Y))
        # delete some rows to make it more fun
        x1, y1 = x1[:, :-1], y1[:, :-1]
        x2, y2 = x2[:-1], y2[:-1]
        # generate graphs
        X_ = [x1, x2, x3]
        G = [make_grid_edges(x) for x in X_]

        # reshape / flatten x and y
        X_ = [x.reshape(-1, n_states) for x in X_]
        Y = [y.ravel() for y in [y1, y2, y3]]

        X = list(zip(X_, G))

        clf.fit(X, Y)
        Y_pred = clf.predict(X)
        for y, y_pred in zip(Y, Y_pred):
            assert_array_equal(y, y_pred)

def test_joint_feature_continuous():
    # FIXME
    # first make perfect prediction, including pairwise part
    X, Y = generate_blocks_multinomial(noise=2, n_samples=1, seed=1)
    x, y = X[0], Y[0]
    n_states = x.shape[-1]

    pw_horz = -1 * np.eye(n_states)
    xx, yy = np.indices(pw_horz.shape)
    # linear ordering constraint horizontally
    pw_horz[xx > yy] = 1

    # high cost for unequal labels vertically
    pw_vert = -1 * np.eye(n_states)
    pw_vert[xx != yy] = 1
    pw_vert *= 10

    # create crf, assemble weight, make prediction
    for inference_method in get_installed(["lp", "ad3"]):
        crf = DirectionalGridCRF(inference_method=inference_method)
        crf.initialize(X, Y)
        w = np.hstack([np.eye(3).ravel(), -pw_horz.ravel(), -pw_vert.ravel()])
        y_pred = crf.inference(x, w, relaxed=True)

        # compute joint_feature for prediction
        joint_feature_y = crf.joint_feature(x, y_pred)
        assert_equal(joint_feature_y.shape, (crf.size_joint_feature,))

def test_chain():
    # test LP, AD3, AD3-BB and JT on a chain.
    # they should all be exact
    rnd = np.random.RandomState(0)
    algorithms = get_installed([('ad3', {'branch_and_bound':False}),
                                ('ad3', {'branch_and_bound':True}),
                                ('dai', {'alg':'jt'})])
    for i in xrange(10):
        forward = np.c_[np.arange(9), np.arange(1, 10)]
        backward = np.c_[np.arange(1, 10), np.arange(9)]
        unary_potentials = rnd.normal(size=(10, 3))
        pairwise_potentials = rnd.normal(size=(3, 3))
        # test that reversing edges is same as transposing pairwise potentials
        y_forward = inference_dispatch(unary_potentials, pairwise_potentials,
                                       forward, 'lp')
        y_backward = inference_dispatch(unary_potentials,
                                        pairwise_potentials.T, backward, 'lp')
        assert_array_equal(y_forward, y_backward)
        for chain in [forward, backward]:
            y_lp = inference_dispatch(unary_potentials, pairwise_potentials,
                                      chain, 'lp')
            for alg in algorithms:
                print(alg)
                y = inference_dispatch(unary_potentials, pairwise_potentials,
                                       chain, alg)
                assert_array_equal(y, y_lp)

def test_binary_grid_unaries():
    # test handling on unaries for binary grid CRFs
    for ds in binary:
        X, Y = ds(n_samples=1)
        x, y = X[0], Y[0]
        for inference_method in get_installed():
            #NOTE: ad3+ fails because it requires a different data structure
            if inference_method == 'ad3+': continue            
            crf = GridCRF(inference_method=inference_method)
            crf.initialize(X, Y)
            w_unaries_only = np.zeros(7)
            w_unaries_only[:4] = np.eye(2).ravel()
            # test that inference with unaries only is the
            # same as argmax
            inf_unaries = crf.inference(x, w_unaries_only)

            assert_array_equal(inf_unaries, np.argmax(x, axis=2),
                               "Wrong unary inference for %s"
                               % inference_method)
            assert(np.mean(inf_unaries == y) > 0.5)

            # check that the right thing happens on noise-free data
            X, Y = ds(n_samples=1, noise=0)
            inf_unaries = crf.inference(X[0], w_unaries_only)
            assert_array_equal(inf_unaries, Y[0],
                               "Wrong unary result for %s"
                               % inference_method)

def test_energy_continuous():
    # make sure that energy as computed by ssvm is the same as by lp
    np.random.seed(0)
    for inference_method in get_installed(["lp", "ad3"]):
        found_fractional = False
        crf = EdgeFeatureGraphCRF(n_states=3,
                                  inference_method=inference_method,
                                  n_edge_features=2)
        while not found_fractional:
            x = np.random.normal(size=(7, 8, 3))
            edge_list = make_grid_edges(x, 4, return_lists=True)
            edges = np.vstack(edge_list)
            edge_features = edge_list_to_features(edge_list)
            x = (x.reshape(-1, 3), edges, edge_features)

            unary_params = np.random.normal(size=(3, 3))
            pw1 = np.random.normal(size=(3, 3))
            pw2 = np.random.normal(size=(3, 3))
            w = np.hstack([unary_params.ravel(), pw1.ravel(), pw2.ravel()])
            res, energy = crf.inference(x, w, relaxed=True, return_energy=True)
            found_fractional = np.any(np.max(res[0], axis=-1) != 1)

            psi = crf.psi(x, res)
            energy_svm = np.dot(psi, w)

            assert_almost_equal(energy, -energy_svm)

def test_one_slack_constraint_caching():
    # testing cutting plane ssvm on easy multinomial dataset
    X, Y = generate_blocks_multinomial(n_samples=10, noise=0.5, seed=0,
                                       size_x=9)
    n_labels = len(np.unique(Y))
    exact_inference = get_installed([('ad3', {'branch_and_bound': True}), "lp"])[0]
    crf = GridCRF(n_states=n_labels, inference_method=exact_inference)
    clf = OneSlackSSVM(model=crf, max_iter=150, C=1,
                       check_constraints=True, break_on_bad=True,
                       inference_cache=50, inactive_window=0)
    clf.fit(X, Y)
    Y_pred = clf.predict(X)
    assert_array_equal(Y, Y_pred)
    assert_equal(len(clf.inference_cache_), len(X))
    # there should be 13 constraints, which are less than the 94 iterations
    # that are done
    # check that we didn't change the behavior of how we construct the cache
    constraints_per_sample = [len(cache) for cache in clf.inference_cache_]
    if exact_inference == "lp":
        assert_equal(len(clf.inference_cache_[0]), 18)
        assert_equal(np.max(constraints_per_sample), 18)
        assert_equal(np.min(constraints_per_sample), 18)
    else:
        assert_equal(len(clf.inference_cache_[0]), 13)
        assert_equal(np.max(constraints_per_sample), 20)
        assert_equal(np.min(constraints_per_sample), 11)