bayespy.inference.VB¶
- class bayespy.inference.VB(*nodes, tol=1e-05, autosave_filename=None, autosave_iterations=0, use_logging=False, user_data=None, callback=None)[source]¶
Variational Bayesian (VB) inference engine
- Parameters
- nodesnodes
Nodes that form the model. Must include all at least all stochastic nodes of the model.
- toldouble, optional
Convergence criterion. Tolerance for the relative change in the VB lower bound.
- autosave_filenamestring, optional
Filename for automatic saving
- autosave_iterationsint, optional
Iteration interval between each automatic saving
- callbackcallable, optional
Function which is called after each update iteration step
- __init__(*nodes, tol=1e-05, autosave_filename=None, autosave_iterations=0, use_logging=False, user_data=None, callback=None)[source]¶
Methods
__init__(*nodes[, tol, autosave_filename, ...])add(x1, x2[, scale])Add two vectors (in parameter format)
compute_lowerbound([ignore_masked])compute_lowerbound_terms(*nodes)dot(x1, x2)Computes dot products of given vectors (in parameter format)
get_gradients(*nodes[, euclidian])Computes gradients (both Riemannian and normal)
get_parameters(*nodes)Get parameters of the nodes
gradient_step(*nodes[, scale])Update nodes by taking a gradient ascent step
has_converged([tol])load(*nodes[, filename, nodes_only])load_user_data(filename)optimize(*nodes[, maxiter, verbose, method, ...])Optimize nodes using Riemannian conjugate gradient
pattern_search(*nodes[, collapsed, maxiter])Perform simple pattern search [4].
plot(*nodes, **kwargs)Plot the distribution of the given nodes (or all nodes)
plot_iteration_by_nodes([axes, diff])Plot the cost function per node during the iteration.
save(*nodes[, filename])set_annealing(annealing)Set deterministic annealing from range (0, 1].
set_autosave(filename[, iterations, nodes])set_callback(callback)set_parameters(x, *nodes)Set parameters of the nodes
update(*nodes[, repeat, plot, tol, verbose, ...])use_logging(use)Attributes