Struct PCAlgorithm 

pub struct PCAlgorithm {
    pub alpha: f64,
    pub max_cond_set: usize,
}

PC (Peter-Clark) algorithm for constraint-based structure learning.

Phase 1: Start with complete undirected graph; remove edges by testing conditional independence with increasing separator set sizes. Phase 2: Orient v-structures (colliders). Phase 3: Apply Meek orientation rules to avoid new cycles / v-structures.

Fields

alpha: f64

Significance level for conditional independence tests.

max_cond_set: usize

Maximum conditioning set size.

Implementations

impl PCAlgorithm

pub fn new(alpha: f64, max_cond_set: usize) -> Self

Create a new PCAlgorithm.

pub fn fit(&self, data: &[Vec<f64>]) -> Result<DAG, StatsError>

Learn the DAG from continuous data using Fisher’s z test.

pub fn conditional_independence_test( &self, data: &[Vec<f64>], x: usize, y: usize, z: &[usize], ) -> bool

Test conditional independence between x and y given z in the data.

Trait Implementations

impl Clone for PCAlgorithm

fn clone(&self) -> PCAlgorithm

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for PCAlgorithm

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

Formats the value using the given formatter.

impl Default for PCAlgorithm

fn default() -> Self

Returns the “default value” for a type.