BNCausalInference

causal_hub::inference

Trait BNCausalInference 

Source 
pub trait BNCausalInference<T>
where
    T: BN,
{
    // Required method
    fn cace_estimate(
        &self,
        x: &Set<usize>,
        y: &Set<usize>,
        z: &Set<usize>,
    ) -> Option<T::CPD>;

    // Provided method
    fn ace_estimate(&self, x: &Set<usize>, y: &Set<usize>) -> Option<T::CPD> { ... }
}
Expand description

A trait for causal inference with Bayesian Networks.

Required Methods§

Source

fn cace_estimate( &self, x: &Set<usize>, y: &Set<usize>, z: &Set<usize>, ) -> Option<T::CPD>

Estimate the conditional average causal effect of X on Y given Z as E(Y | do(X), Z).

§Arguments
  • x - The cause variables.
  • y - The effect variables.
  • z - The conditioning variables.
§Panics
  • If X is empty.
  • If Y is empty.
  • If X and Y are not disjoint.
  • If X and Z are not disjoint.
  • If Y and Z are not disjoint.
§Returns

The estimated conditional average causal effect of X on Y given Z.

Provided Methods§

Source

fn ace_estimate(&self, x: &Set<usize>, y: &Set<usize>) -> Option<T::CPD>

Estimate the average causal effect of X on Y as E(Y | do(X)).

§Arguments
  • x - The cause variables.
  • y - The effect variables.
§Panics
  • If X is empty.
  • If Y is empty.
  • If X and Y are not disjoint.
§Returns

The estimated average causal effect of X on Y.

Implementors§