Struct Likelihood 

pub struct Likelihood { /* private fields */ }

A resolved collection of likelihood terms sharing one parameter layout.

Implementations

impl Likelihood

pub fn cross_section( self: &Arc<Likelihood>, term_name: impl Into<String>, generated_mc: Dataset, luminosity: f64, parameters: Vec<f64>, ) -> Result<CrossSection, LikelihoodError>

Prepares a central-value cross-section analysis from a shared likelihood.

Errors

Returns an error for invalid inputs or likelihood preparation failure.

pub fn cross_section_with_ensemble( self: &Arc<Likelihood>, term_name: impl Into<String>, generated_mc: Dataset, luminosity: f64, parameters: Vec<f64>, ensemble: Ensemble, ) -> Result<CrossSection, LikelihoodError>

Prepares an ensemble-backed cross-section analysis.

Errors

Returns an error for invalid inputs, mismatched draws, or preparation failure.

impl Likelihood

pub fn new<T>( terms: impl IntoIterator<Item = T>, ) -> Result<Likelihood, LikelihoodError>

where T: LikelihoodTerm + 'static,

Construct a likelihood from terms of one concrete type.

The terms are boxed internally, so the common case does not require a manual LikelihoodTerm::boxed call.

Errors

Returns LikelihoodError when term names or parameter definitions conflict, or a term cannot be resolved.

pub fn new_boxed( terms: impl IntoIterator<Item = Box<dyn LikelihoodTerm>>, ) -> Result<Likelihood, LikelihoodError>

Construct a likelihood containing heterogeneous, already boxed terms.

Errors

Returns LikelihoodError when term names or parameter definitions conflict, or a term cannot be resolved.

pub fn with_execution<T>( terms: impl IntoIterator<Item = T>, execution: &Execution, ) -> Result<Likelihood, LikelihoodError>

where T: LikelihoodTerm + 'static,

Constructs a likelihood with an explicit execution context.

Errors

Returns LikelihoodError when term names or parameter definitions conflict, or a term cannot be resolved for execution.

pub fn with_execution_boxed( terms: impl IntoIterator<Item = Box<dyn LikelihoodTerm>>, execution: &Execution, ) -> Result<Likelihood, LikelihoodError>

Construct a heterogeneous likelihood using a borrowed execution setup.

Errors

Returns LikelihoodError when term names or parameter definitions conflict, or a term cannot be resolved for execution.

pub fn params(&self) -> &ParamLayout

Returns the global parameter layout.

pub fn default_params(&self) -> Vec<f64>

Return deterministic initial values in the order expected by Self::nll.

pub fn params_with(&self, value: impl FnMut(&Parameter) -> f64) -> Vec<f64>

Generate one value for each free parameter in objective-vector order.

pub fn sample_initial(&self, seed: u64) -> Vec<f64>

Sample uniform initial ranges while preserving fixed and point-initialized parameters.

pub fn bootstrap(&self, seed: u64) -> Result<Likelihood, LikelihoodError>

Rebuilds this likelihood with Poisson-bootstrapped observed datasets.

Errors

Returns LikelihoodError when a term cannot be bootstrap-cloned or the rebuilt likelihood cannot be prepared.

pub fn terms(&self) -> &[Box<dyn LikelihoodTerm>]

Returns the resolved likelihood terms.

pub fn execution(&self) -> &Execution

Returns the execution context used by the likelihood.

pub fn diagnostics(&self) -> LikelihoodDiagnostics

Returns a snapshot of preparation and objective-evaluation diagnostics.

pub fn nll<'a>( &self, parameters: impl Into<Parameters<'a>>, ) -> Result<f64, LikelihoodError>

Evaluate the objective from free values in Self::params order.

Errors

Returns LikelihoodError when parameters have the wrong layout or length, violate bounds, or a term cannot be evaluated.

pub fn nll_with_gradient<'a>( &self, parameters: impl Into<Parameters<'a>>, ) -> Result<LikelihoodEvaluation, LikelihoodError>

Evaluates the objective and gradient from free or resolved parameter values.

Errors

Returns LikelihoodError when parameters are invalid or a term’s value or gradient cannot be evaluated.

pub fn stochastic_nll_with_gradient( &self, free_parameters: &[f64], fraction: f64, seed: u64, ) -> Result<LikelihoodEvaluation, LikelihoodError>

Evaluates an unbiased stochastic objective and gradient.

Errors

Returns LikelihoodError when fraction is outside (0, 1], parameters are invalid, or stochastic term evaluation fails.

pub fn cross_section_integrals( &self, term_name: &str, generated_mc: &Dataset, ) -> Result<CrossSectionIntegrals, LikelihoodError>

Prepares accepted and generated Monte Carlo integrals for an intensity term.

Errors

Returns LikelihoodError when term_name is missing or not an intensity term, or dataset preparation fails.

pub fn cross_section_integrals_with_tags<'a>( &self, term_name: &str, generated_mc: &Dataset, tags: impl IntoIterator<Item = &'a str>, ) -> Result<CrossSectionIntegrals, LikelihoodError>

Prepares tag-narrowed accepted and generated Monte Carlo integrals.

The selected tags define the numerator contribution. Cross sections retain the full accepted-model normalization, matching Self::projection.

Errors

Returns LikelihoodError when term_name is missing or not an intensity term, graph projection fails, or dataset preparation fails.

pub fn intensity_datasets( &self, term_name: &str, ) -> Result<(&Dataset, &Dataset), LikelihoodError>

Returns the observed and accepted Monte Carlo sources for an intensity term.

Errors

Returns LikelihoodError when term_name is missing or is not an intensity term.

pub fn projection<'a>( &self, term_name: &str, generated_mc: &Dataset, tags: impl IntoIterator<Item = &'a str>, ) -> Result<LikelihoodProjection, LikelihoodError>

Projects an intensity term onto selected model tags over generated Monte Carlo.

Errors

Returns LikelihoodError when term_name is missing or not an intensity term, graph projection fails, or dataset preparation fails.

Trait Implementations

impl Debug for Likelihood

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

Formats the value using the given formatter.

impl Objective for Likelihood

fn parameter_layout(&self) -> &ParamLayout

Returns the stable parameter layout used by objective vectors.

fn value(&self, free_parameters: &[f64]) -> Result<f64, LikelihoodError>

Evaluates the objective at the supplied free-parameter vector.

fn value_gradient( &self, free_parameters: &[f64], ) -> Result<LikelihoodEvaluation, LikelihoodError>

Evaluates the objective and its gradient.

impl StochasticObjective for Likelihood

fn stochastic_value_gradient( &self, free_parameters: &[f64], fraction: f64, seed: u64, ) -> Result<LikelihoodEvaluation, LikelihoodError>

Evaluates an unbiased stochastic objective and gradient on a deterministic batch.