gam-model-api 0.3.126

Model and custom-family trait API and contracts for the gam penalized-likelihood engine
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37

//! The `CustomFamily` trait itself plus the evaluation result structs it returns
//! (`FamilyEvaluation`, joint-gradient/batched-term carriers) and the eval-scope /
//! outer-eval-context enums that parameterize trait calls.

pub mod families {
    pub mod custom_family {
        pub mod family_trait;
        pub mod joint_newton_defaults;
        pub mod options;
        pub mod psi_design;

        pub use family_trait::*;
        pub use options::*;
        pub use psi_design::*;
    }
}

// `joint_penalty` and `outer_subsample` are neutral low-layer primitives that
// live in `gam-problem` (below the `CustomFamily` trait). Re-export the modules
// and their public types so existing `gam_model_api::{joint_penalty,
// outer_subsample}::*` and `crate::{OuterScoreSubsample, JointPenaltyBundle, …}`
// paths stay stable without duplicating the sources.
pub use gam_problem::{joint_penalty, outer_subsample};

pub use families::custom_family::*;
pub use gam_problem::joint_penalty::{JointPenaltyBundle, JointPenaltyError, JointPenaltySpec};
pub use gam_problem::outer_subsample::{OuterScoreSubsample, RowSet, WeightedOuterRow};

#[derive(Clone, Copy, Debug, PartialEq, Eq)]
pub enum OuterEvalOrder {
    /// Compute only the objective value.
    Value,
    /// Compute value and gradient only.
    ValueAndGradient,
    /// Compute value, gradient, and analytic Hessian when available.
    ValueGradientHessian,
}