Struct MultinomialLogitRowProgram 

pub struct MultinomialLogitRowProgram<'row> { /* private fields */ }

Production gam_math::jet_tower::RowProgram for one reference-coded multinomial-logit row.

Active-class logits are the M primaries and class M is the implicit reference with logit zero. The generic row NLL is the mechanical tower oracle for the retained normalized-softmax/Fisher lowerings in this module; production parity tests invoke this type directly rather than restating its expression under cfg(test).

Implementations

impl<'row> MultinomialLogitRowProgram<'row>

pub fn new( eta: &'row [f64], response: &'row [f64], weight: f64, ) -> Result<Self, String>

Construct one validated row. eta contains the active-class logits and response contains the complete simplex row, including the implicit reference class in its last slot.

Trait Implementations

impl<'row> Clone for MultinomialLogitRowProgram<'row>

fn clone(&self) -> MultinomialLogitRowProgram<'row>

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl<'row> Copy for MultinomialLogitRowProgram<'row>

impl<'row> Debug for MultinomialLogitRowProgram<'row>

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

Formats the value using the given formatter.

impl<const M: usize> RowProgram<M> for MultinomialLogitRowProgram<'_>

fn n_rows(&self) -> usize

Number of observations the program covers.

fn primaries(&self, row: usize) -> Result<[f64; M], String>

Current primary-scalar values for row (where to seed the scalar).

fn eval<S: JetScalar<M>>(&self, row: usize, p: &[S; M]) -> Result<S, String>

The row NLL evaluated on a generic jet scalar. p[a] arrives pre-seeded (base value + per-scalar nilpotent directions) by the caller; the body uses ONLY crate::jet_scalar::JetScalar ops and per-row data (response, censoring, offsets) entering as constants.