Struct BestDoseProblem 

pub struct BestDoseProblem<E: Equation> { /* private fields */ }

A dose-optimization problem over a parameter distribution.

A BestDoseProblem pairs a model with a distribution over its parameters — the support points (Theta) and their probability Weights. The distribution is supplied by the caller: it may be a population fit, or a patient-specific posterior computed with NCNPAG/NPMAP from past data.

Call optimize to solve for the doses that best hit a target profile.

Example

let problem = BestDoseProblem::new(eq, theta, weights)?;
let result = problem.optimize(
    target,
    Target::Concentration,
    DoseRange::new(0.0, 300.0),
    0.5,
    BestDoseOptions::default(),
)?;
let optimal_subject = result.subject();
let cost = result.cost();

Implementations

impl<E: Equation> BestDoseProblem<E>

pub fn new(eq: E, theta: Theta, weights: Weights) -> Result<Self>

Creates a dose-optimization problem from a model and a parameter distribution.

Returns an error if the number of weights does not match the number of support points in theta.

pub fn optimize( &self, target: Subject, target_type: Target, dose_range: DoseRange, bias: f64, options: BestDoseOptions, ) -> Result<BestDoseResult>

Solves for the optimal doses that hit target.

Doses in the target subject with amount 0.0 are optimized; non-zero doses stay fixed. bias (λ ∈ [0, 1]) trades off minimizing the expected squared error over the distribution (λ = 0) against minimizing the error of the weighted-mean prediction (λ = 1).

Returns the optimal dosing Subject and its cost. Predictions can be generated afterwards by simulating the returned subject.

pub fn theta(&self) -> &Theta

The support points of the distribution.

pub fn weights(&self) -> &Weights

The probability weights of the distribution.

Trait Implementations

impl<E: Clone + Equation> Clone for BestDoseProblem<E>

fn clone(&self) -> BestDoseProblem<E>

Returns a duplicate of the value.

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

Performs copy-assignment from source.