Skip to main content

Module glm

so_models

Module glm 

Source
Expand description

Generalized Linear Models (GLM)

This module implements Generalized Linear Models, which extend linear regression to response variables with non-normal distributions.

§Key Components

  1. Distribution Families: Gaussian, Binomial, Poisson, Gamma, Inverse Gaussian
  2. Link Functions: Identity, Logit, Log, Inverse, etc.
  3. IRLS Algorithm: Iteratively Reweighted Least Squares for estimation
  4. Model Diagnostics: Deviance, Pearson chi-squared, AIC, BIC

§Example Usage

use so_models::glm::{GLM, Family, Link};
use so_core::data::DataFrame;
use so_core::formula::Formula;

// Create a logistic regression model
let model = GLM::new()
    .family(Family::Binomial)
    .link(Link::Logit)
    .intercept(true);

// Example data would be needed to fit the model
// let formula = Formula::parse("y ~ x1 + x2").unwrap();
// let data = DataFrame::from_series(...).unwrap();
// let results = model.fit_formula(&formula, &data).unwrap();

§References

  • McCullagh, P., & Nelder, J. A. (1989). Generalized Linear Models.
  • R’s glm() function and statsmodels’ GLM implementation.

Re-exports§

pub use family::Family;
pub use model::GLM;
pub use model::GLMModelBuilder;
pub use results::GLMResults;

Modules§

family
Distribution families for Generalized Linear Models
model
Generalized Linear Model implementation using IRLS (Iteratively Reweighted Least Squares)
prelude
results
GLM results structure and diagnostics