pystruct.models.LatentGraphCRF¶

class pystruct.models.LatentGraphCRF(n_labels=None, n_features=None, n_states_per_label=2, inference_method=None)[source]¶

CRF with latent states for variables.

This is also called “hidden dynamics CRF”. For each output variable there is an additional variable which can encode additional states and interactions.

Parameters:

Parameters:	n_labels : int Number of states of output variables. n_featues : int or None (default=None). Number of input features per input variable. `None` means it is equal to `n_labels`. n_states_per_label : int or list (default=2) Number of latent states associated with each observable state. Can be either an integer, which means the same number of hidden states will be used for all observable states, or a list of integers of length `n_labels`. inference_method : string, default=”ad3” Function to call to do inference and loss-augmented inference. Possible values are: ‘max-product’ for max-product belief propagation. Recommended for chains an trees. Loopy belief propagatin in case of a general graph. ‘lp’ for Linear Programming relaxation using cvxopt. ‘ad3’ for AD3 dual decomposition. ‘qpbo’ for QPBO + alpha expansion. ‘ogm’ for OpenGM inference algorithms.

n_labels : int

Number of states of output variables.

n_featues : int or None (default=None).

Number of input features per input variable. None means it is equal to n_labels.

n_states_per_label : int or list (default=2)

Number of latent states associated with each observable state. Can be either an integer, which means the same number of hidden states will be used for all observable states, or a list of integers of length n_labels.

inference_method : string, default=”ad3”

Function to call to do inference and loss-augmented inference. Possible values are:

‘max-product’ for max-product belief propagation.

Recommended for chains an trees. Loopy belief propagatin in case of a general graph.

‘lp’ for Linear Programming relaxation using cvxopt.

‘ad3’ for AD3 dual decomposition.

‘qpbo’ for QPBO + alpha expansion.

‘ogm’ for OpenGM inference algorithms.

Methods

`base_loss`(y, y_hat)
`batch_inference`(X, w[, relaxed])
`batch_joint_feature`(X, Y[, Y_true])
`batch_loss`(Y, Y_hat)
`batch_loss_augmented_inference`(X, Y, w[, ...])
`continuous_loss`(y, y_hat)
`inference`(x, w[, relaxed, return_energy])	Inference for x using parameters w.
`init_latent`(X, Y)
`initialize`(X, Y)
`joint_feature`(x, y)	Feature vector associated with instance (x, y).
`label_from_latent`(h)
`latent`(x, y, w)
`loss`(h, h_hat)
`loss_augmented_inference`(x, h, w[, relaxed, ...])
`max_loss`(y)

__init__(n_labels=None, n_features=None, n_states_per_label=2, inference_method=None)[source]¶

inference(x, w, relaxed=False, return_energy=False)¶

Inference for x using parameters w.

Finds (approximately) armin_y np.dot(w, joint_feature(x, y)) using self.inference_method.

Parameters:

Parameters:	x : tuple Instance of a graph with unary evidence. x=(unaries, edges) unaries are an nd-array of shape (n_nodes, n_states), edges are an nd-array of shape (n_edges, 2) w : ndarray, shape=(size_joint_feature,) Parameters for the CRF energy function. relaxed : bool, default=False Whether relaxed inference should be performed. Only meaningful if inference method is ‘lp’ or ‘ad3’. By default fractional solutions are rounded. If relaxed=True, fractional solutions are returned directly. return_energy : bool, default=False Whether to return the energy of the solution (x, y) that was found.
Returns:	y_pred : ndarray or tuple By default an inter ndarray of shape=(width, height) of variable assignments for x is returned. If `relaxed=True` and inference_method is `lp` or `ad3`, a tuple (unary_marginals, pairwise_marginals) containing the relaxed inference result is returned. unary marginals is an array of shape (width, height, n_states), pairwise_marginals is an array of shape (n_states, n_states) of accumulated pairwise marginals.

x : tuple

Instance of a graph with unary evidence. x=(unaries, edges) unaries are an nd-array of shape (n_nodes, n_states), edges are an nd-array of shape (n_edges, 2)

w : ndarray, shape=(size_joint_feature,)

Parameters for the CRF energy function.

relaxed : bool, default=False

Whether relaxed inference should be performed. Only meaningful if inference method is ‘lp’ or ‘ad3’. By default fractional solutions are rounded. If relaxed=True, fractional solutions are returned directly.

return_energy : bool, default=False

Whether to return the energy of the solution (x, y) that was found.

Returns:

y_pred : ndarray or tuple

By default an inter ndarray of shape=(width, height) of variable assignments for x is returned. If relaxed=True and inference_method is lp or ad3, a tuple (unary_marginals, pairwise_marginals) containing the relaxed inference result is returned. unary marginals is an array of shape (width, height, n_states), pairwise_marginals is an array of shape (n_states, n_states) of accumulated pairwise marginals.

joint_feature(x, y)¶

Feature vector associated with instance (x, y).

Feature representation joint_feature, such that the energy of the configuration (x, y) and a weight vector w is given by np.dot(w, joint_feature(x, y)).

Parameters:

Parameters:	x : tuple Unary evidence. y : ndarray or tuple Either y is an integral ndarray, giving a complete labeling for x. Or it is the result of a linear programming relaxation. In this case, `y=(unary_marginals, pariwise_marginals)`.
Returns:	p : ndarray, shape (size_joint_feature,) Feature vector associated with state (x, y).

x : tuple

Unary evidence.

y : ndarray or tuple

Either y is an integral ndarray, giving a complete labeling for x. Or it is the result of a linear programming relaxation. In this case, y=(unary_marginals, pariwise_marginals).

Returns:

p : ndarray, shape (size_joint_feature,)

Feature vector associated with state (x, y).