Struct JointPoissonResult 

pub struct JointPoissonResult {

    pub deviance: f64,
    pub deviance_per_dof: f64,
    pub n_data: usize,
    pub n_active: usize,
    pub n_free: usize,
    pub gn_iterations: usize,
    pub polish_iterations: usize,
    pub gn_converged: bool,
    pub polish_converged: bool,
    pub polish_improved: bool,
    pub params: Vec<f64>,
    pub covariance: Option<FlatMatrix>,
    pub uncertainties: Option<Vec<f64>>,
}

Outcome of joint_poisson_fit.

Fields

deviance: f64

Final deviance D at the fitted parameters.

deviance_per_dof: f64

D / (n − k). Primary GOF statistic per memo 35 §P1.2.

n_data: usize

Number of data bins on the configured grid (n). This is the total bin count; when a fit-energy-range mask is in effect, the count of bins that actually contributed to the cost function is reported separately as Self::n_active.

n_active: usize

Number of active data bins — equal to n_data when no mask is set, or the count of true entries in the objective’s active_mask otherwise. The deviance / dof ratio uses (n_active − n_free) so reduced deviance is unbiased when a fit-energy-range mask is in effect (SAMMY EMIN/EMAX semantics, #514).

n_free: usize

Number of free parameters (k).

gn_iterations: usize

Iterations performed in the damped-Fisher stage.

polish_iterations: usize

Iterations performed by the Nelder-Mead polish stage (0 if disabled).

gn_converged: bool

true when the stage-1 (damped Fisher) optimizer met its tol_d and tol_param criteria before hitting max_iter.

polish_converged: bool

true when the Nelder-Mead polish met xatol and fatol before max_iter (always false if enable_polish == false).

polish_improved: bool

true when the polish step lowered the deviance below the stage-1 best value. Useful diagnostic — if polish improved D materially, stage 1 likely stalled.

params: Vec<f64>

Final parameter values (all parameters, including fixed).

covariance: Option<FlatMatrix>

Inverse Fisher covariance of free parameters (n_free × n_free), computed at the final θ. None if the Fisher matrix was singular or compute_covariance == false.

uncertainties: Option<Vec<f64>>

√diag(covariance) for each free parameter, in free-index order.

Trait Implementations

impl Clone for JointPoissonResult

fn clone(&self) -> JointPoissonResult

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for JointPoissonResult

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

Formats the value using the given formatter.