pasteur.extras.synth.privbayes.implementation

Description

Functions

add_multiple_to_pset(s, x, h)	Given parent set s, adds attributes x with heights h.
add_to_pset(s, x, h)	Given parent set s, adds attribute x with height h.
calc_entropy(req, mar, info)	Calculates the entropy for the provided data.
calc_mutual_info(req, mar, info)	Calculates mutual information I(X,P) for the provided data using log2.
calc_noisy_marginals(oracle, nodes, ...[, ...])	Calculates the marginals and adds laplacian noise with scale noise_scale.
calc_r_function(req, mar, info)	Calculates the R(X,P) function for the provided data.
calculate_attr_combinations(table, attr)
calculate_attrs_combinations(table, attrs)
get_attrs(ds_attrs, sel)
greedy_bayes(oracle, ds_attrs, n, e1, e2, ...)	Performs the greedy bayes algorithm for variable domain data.
maximal_parents(domains, tau)	Given a set V containing hierarchical attributes (by int) and a tau score that is divided by the size of the domain, return a set of all possible combinations of attributes, such that if t > 1 there isn't an attribute that can be indexed in a higher level
print_tree(attrs, nodes, e1, e2, theta, t[, ...])
sample_rows(idx, attrs, hist, nodes, marginals)
sens_entropy(n)	Provides the sensitivity for the entropy function for a given dataset size (n).
sens_mutual_info(n)	Provides the the log2 sensitivity of the mutual information function for a given dataset size (n).
sens_r_function(n)	Provides the the R function sensitivity for a given dataset size (n).
to_str(a)

Classes

Node(attr, value, p, domain, partial)