Struct Mediation 

pub struct Mediation {
    pub outcome_endog: Array1<f64>,
    pub outcome_exog: Array2<f64>,
    pub mediator_endog: Array1<f64>,
    pub mediator_exog: Array2<f64>,
    pub exp_pos_outcome: usize,
    pub exp_pos_mediator: usize,
    pub med_pos_outcome: usize,
}

Specification of a linear mediation analysis (no formulas).

outcome_endog / outcome_exog define the outcome regression Y ~ X_o (which must include both the mediator and the exposure as columns), and mediator_endog / mediator_exog the mediator regression M ~ X_m. The *_pos fields give the column positions of the exposure in each design and of the mediator in the outcome design, exactly as in the reference’s positional API (Mediation(outcome, mediator, [exp_pos_outcome, exp_pos_mediator], med_pos_outcome)).

Fields

outcome_endog: Array1<f64>

Outcome model response vector.

outcome_exog: Array2<f64>

Outcome model design matrix (includes exposure and mediator columns).

mediator_endog: Array1<f64>

Mediator model response vector.

mediator_exog: Array2<f64>

Mediator model design matrix (includes the exposure column).

exp_pos_outcome: usize

Column position of the exposure in the outcome design.

exp_pos_mediator: usize

Column position of the exposure in the mediator design.

med_pos_outcome: usize

Column position of the mediator in the outcome design.

Implementations

impl Mediation

pub fn fit(&self) -> MediationResults

Fit both linear models and return the deterministic mediation point estimates. Mirrors the deterministic limit of the reference Mediation.fit for linear outcome and mediator models.

Trait Implementations

impl Clone for Mediation

fn clone(&self) -> Mediation

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for Mediation

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

Formats the value using the given formatter.