Struct BayesianNetworkState 

pub struct BayesianNetworkState {
    pub order: Vec<usize>,
    pub parents: Vec<Vec<usize>>,
    pub cpt: Vec<Vec<f32>>,
}

Fitted state for the BayesianNetwork model after one call to ProbabilityModel::fit.

On the prior path (prev = None) the network is edgeless: order is the natural order [0, 1, …, D-1], every parents[v] is empty, and every cpt[v] is the single-entry vector [init_prob].

Fields

order: Vec<usize>

Topological sampling order: a permutation of 0..D such that every node appears after all of its parents. Ancestral sampling walks this order so each gene’s parents are already drawn.

parents: Vec<Vec<usize>>

parents[node] is the node’s parent index set, kept sorted ascending. The sort defines the bit positions used by the CPT packing (see the module docs).

cpt: Vec<Vec<f32>>

Conditional probability tables: cpt[node][config] is P(node = 1 | parents = config), where config is the bit-packed parent configuration Σ_j bit(parent_j) << j over parents[node] in sorted order. Each inner vector has length 2^|parents[node]|.

Trait Implementations

impl Clone for BayesianNetworkState

fn clone(&self) -> BayesianNetworkState

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for BayesianNetworkState

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.