Struct PoissonRegressor 

pub struct PoissonRegressor<F> {
    pub alpha: F,
    pub max_iter: usize,
    pub tol: F,
    pub fit_intercept: bool,
    pub solver: Solver,
    pub warm_start: bool,
    pub coef_init: Option<(Array1<F>, F)>,
}

Poisson regressor — GLM with Poisson family and log link.

Suitable for modelling count data (y >= 0, integer-valued).

Type Parameters

• F: The floating-point type (f32 or f64).

Fields

alpha: F

L2 regularization strength.

max_iter: usize

Maximum number of IRLS iterations.

tol: F

Convergence tolerance.

fit_intercept: bool

Whether to fit an intercept.

solver: Solver

Optimization algorithm requested, mirroring sklearn’s solver parameter (glm.py:140-145, default "lbfgs").

ferrolearn fits via IRLS regardless of this value (R-DEV-7); the parameter is accepted for sklearn API parity (R-DEV-2) and the observable coef_ / intercept_ match sklearn for either value. See Solver.

warm_start: bool

Whether to warm-start from an explicit initial point, mirroring sklearn’s warm_start parameter (default false, glm.py:146, :576). See GLMRegressor::warm_start for the R-DEV-2 / R-DEV-7 rationale.

coef_init: Option<(Array1<F>, F)>

Explicit warm-start initial point (feature_coefficients, intercept) — the ferrolearn analog of sklearn reusing self.coef_ / self.intercept_ (R-DEV-7, glm.py:244-250). Consulted only when warm_start is true. Set via PoissonRegressor::with_coef_init.

Implementations

impl<F: Float + FromPrimitive> PoissonRegressor<F>

pub fn new() -> Self

Create a new PoissonRegressor with default settings.

Defaults: alpha = 1.0, max_iter = 100, tol = 1e-4, fit_intercept = true, solver = Solver::Lbfgs (sklearn default).

pub fn with_alpha(self, alpha: F) -> Self

Set the L2 regularization strength.

pub fn with_max_iter(self, max_iter: usize) -> Self

Set the maximum number of IRLS iterations.

pub fn with_tol(self, tol: F) -> Self

Set the convergence tolerance.

pub fn with_fit_intercept(self, fit_intercept: bool) -> Self

Set whether to fit an intercept.

pub fn with_solver(self, solver: Solver) -> Self

Set the optimization Solver, mirroring sklearn’s solver parameter (glm.py:140-145, default "lbfgs").

ferrolearn fits via IRLS regardless of the value (R-DEV-7); both values produce the same observable coef_ / intercept_ (sklearn API parity, R-DEV-2).

pub fn with_warm_start(self, warm_start: bool) -> Self

Enable or disable warm-starting, mirroring sklearn’s warm_start parameter (default false, glm.py:146, :576). See GLMRegressor::with_warm_start for the R-DEV-2 / R-DEV-7 rationale; the warm-start point is supplied explicitly via PoissonRegressor::with_coef_init.

pub fn with_coef_init(self, coef: Array1<F>, intercept: F) -> Self

Set the explicit warm-start initial point (feature_coefficients, intercept) — the ferrolearn analog of sklearn reusing self.coef_ / self.intercept_ (R-DEV-7, glm.py:244-250). Only consulted when warm_start is true; coef’s length must equal the number of features.

impl<F: Float + Send + Sync + ScalarOperand + FromPrimitive + 'static> PoissonRegressor<F>

pub fn fit_with_sample_weight( &self, x: &Array2<F>, y: &Array1<F>, sample_weight: &Array1<F>, ) -> Result<FittedGLMRegressor<F>, FerroError>

Fit the Poisson GLM via IRLS with per-sample weights sample_weight.

Mirrors sklearn’s PoissonRegressor.fit(X, y, sample_weight) (glm.py:170, :229-242). See GLMRegressor::fit_with_sample_weight for the weighting semantics; an all-ones weight vector reproduces Fit::fit exactly.

Errors

See GLMRegressor::fit_with_sample_weight.

Trait Implementations

impl<F: Clone> Clone for PoissonRegressor<F>

fn clone(&self) -> PoissonRegressor<F>

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl<F: Debug> Debug for PoissonRegressor<F>

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

Formats the value using the given formatter.

impl<F: Float + FromPrimitive> Default for PoissonRegressor<F>

fn default() -> Self

Returns the “default value” for a type.

impl<F: Float + Send + Sync + ScalarOperand + FromPrimitive + 'static> Fit<ArrayBase<OwnedRepr<F>, Dim<[usize; 2]>>, ArrayBase<OwnedRepr<F>, Dim<[usize; 1]>>> for PoissonRegressor<F>

fn fit( &self, x: &Array2<F>, y: &Array1<F>, ) -> Result<FittedGLMRegressor<F>, FerroError>

Fit the Poisson GLM via IRLS.

Errors

See GLMRegressor::fit.

type Fitted = FittedGLMRegressor<F>

The fitted model type returned by fit.

type Error = FerroError

The error type returned by fit.

impl<F> PipelineEstimator<F> for PoissonRegressor<F>

where F: Float + FromPrimitive + ScalarOperand + Send + Sync + 'static,

fn fit_pipeline( &self, x: &Array2<F>, y: &Array1<F>, ) -> Result<Box<dyn FittedPipelineEstimator<F>>, FerroError>

Fit this estimator on the given data.