Struct NonParametricResult 

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

Contains the results of a nonparametric estimation, including the final parameter

Implementations

impl<E: Equation> NonParametricResult<E>

pub fn cycles(&self) -> usize

pub fn objf(&self) -> f64

pub fn converged(&self) -> bool

pub fn get_theta(&self) -> &Theta

pub fn prior(&self) -> &Theta

The prior distribution (Theta) that seeded the algorithm.

This is the initial set of support points, as opposed to the optimized solution returned by get_theta.

pub fn data(&self) -> &Data

pub fn equation(&self) -> &E

pub fn cycle_log(&self) -> &CycleLog

pub fn chain<A>(self, algorithm: A) -> Result<NonParametricResult<E>>

where A: Algorithm<E, NonParametric, Output = NonParametricResult<E>>, E: EquationMetadataSource,

Chain this result into a new fit with a different algorithm.

Uses the optimized theta as the prior for the new run. Keeps the same equation, data, and error models. Consumes self.

Example

let result = problem
    .fit_with(NpagConfig::new().max_cycles(100))?
    .chain(NpodConfig::new().max_cycles(50))?;

pub fn psi(&self) -> &Psi

pub fn weights(&self) -> &Weights

pub fn error_models(&self) -> &AssayErrorModels

pub fn posterior(&self) -> Result<Posterior>

Compute the posterior probabilities on demand from Psi and the Weights. This is a cheap matrix operation and is intentionally not cached on the result.

pub fn predictions(&self, idelta: f64, tad: f64) -> Result<NPPredictions>

Compute predictions on demand. Nothing is cached on the result; callers that need the predictions repeatedly should hold on to the returned value themselves.

pub fn write_theta(&self, path: &Path) -> Result<()>

pub fn write_posterior(&self, path: &Path) -> Result<()>

pub fn write_covariates(&self, path: &Path) -> Result<()>

pub fn write_cycles(&self, path: &Path) -> Result<()>

Write the cycle log to a CSV file readable by Pmetrics.

pub fn write_predictions( &self, path: &Path, idelta: f64, tad: f64, ) -> Result<()>

Compute and write the population and posterior predictions to a CSV file readable by Pmetrics.

idelta is the interval used to densify the prediction grid and tad is the additional time after the last event to simulate.

pub fn write_json(&self, path: &Path, idelta: f64, tad: f64) -> Result<()>

Serialize the complete result to a single JSON file.

This includes the data, error models, prior, optimized support points, likelihoods, weights, objective function, status, cycle log, posterior probabilities, and predictions. idelta and tad control the density of the embedded predictions (see predictions).

pub fn write_outputs( &self, directory: impl AsRef<Path>, idelta: f64, tad: f64, ) -> Result<()>

Write the full set of result artifacts to directory.

The directory is created if it does not exist, and the following files are produced (matching the names expected by Pmetrics):

• theta.csv — optimized support points with weights
• posterior.csv — per-subject posterior probabilities
• pred.csv — population and posterior predictions
• covs.csv — interpolated covariates
• cycles.csv — cycle log
• result.json — the complete result as JSON

The posterior and predictions are computed once and shared across the CSV and JSON outputs. idelta and tad control the density of the predictions (see predictions).

Trait Implementations

impl<E: Debug + Equation> Debug for NonParametricResult<E>

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

Formats the value using the given formatter.

impl<E: Equation> FitResult for NonParametricResult<E>

fn objf(&self) -> f64

fn converged(&self) -> bool

fn summary(&self) -> FitSummary

fn population_summary(&self) -> PopulationSummary

fn individual_summaries(&self) -> Vec<IndividualSummary>