Skip to content

Scala library for fitting linear and generalised linear statistical models

License

Notifications You must be signed in to change notification settings

darrenjw/scala-glm

Repository files navigation

scala-glm

pages-build-deployment

Scala library for regression modelling

Regression modelling, including fitting linear and generalised linear statistical models, diagnosing fit and making predictions. Diagnostic plots. Also, flexible regression modelling using orthogonal polynomials, cosine series, B-splines, etc. This library builds on top of Breeze, and is aimed primarily at people already using Scala and Breeze who want more statistical functionality.

Installation/Getting started

This library is simplest to use with sbt. You should install sbt before attempting to use this library.

To use the pre-built binary, add the following lines to your build.sbt:

libraryDependencies += "com.github.darrenjw" %% "scala-glm" % "0.9"

The current stable release is "0.9". It is cross-built for Scala 3, Scala 2.12 and 2.13 and published to the central repository. Version 0.3 was the final release for Scala 2.11.

There is a giter8 template for scala-glm, so using recent versions of sbt you can create a minimal scala-glm project template with:

sbt new darrenjw/scala-glm.g8

If you just want to try out the library without setting up any kind of project, you can do so with a session like:

$ sbt "-Dsbt.version=1.10.1"
> set scalaVersion := "3.3.4"
> set libraryDependencies += "com.github.darrenjw" %% "scala-glm" % "0.9"
> console
scala> import scalaglm.*

Alternatively, if you use scala-cli, just add

//> using scala 3.3.4
//> using dep com.github.darrenjw::scala-glm:0.9

to the top of your script. A minimal but complete runnable example script for scala-cli is given below:

//> using scala 3.3.4
//> using dep com.github.darrenjw::scala-glm:0.9

import scalaglm.Pca
import breeze.linalg.*

@main def sglm() =
  val X = DenseMatrix((1.0,1.5),(1.5,2.0),(2.0,1.5))
  val pca = Pca(X, List("V1","V2"))
  pca.summary

See below for documentation links.

This library has a dependence on Breeze, so if you have a dependence on scala-glm you don't need to add an additional dependence on Breeze. Some familiarity with Breeze is assumed for effective use of this library.

Latest snapshot

If you want to use the latest snapshot, add the following to your build.sbt:

libraryDependencies += "com.github.darrenjw" %% "scala-glm" % "1.0-SNAPSHOT"
resolvers += "Sonatype Snapshots" at "https://oss.sonatype.org/content/repositories/snapshots/"

Building from source

If building from source, running sbt console from this directory should give a Scala REPL with a dependence on the library. Running sbt test will run all tests (but note that an R installation is required for many of the tests, which cross-check results against R). Running sbt doc will generate ScalaDoc API documentation.

Documentation

  • QuickStart Guide - start with this to get a feeling for what the library can do
  • Flexible regression - guide to flexible and non-parametric regression, using orthogonal polynomials, cosine series, and B-spline basis functions
  • The examples subdirectory of this repo contains more interesting, self-contained runnable examples
  • API documentation - ScalaDoc (for the most recent snapshot)
  • For anyone not very familiar with Scala or Breeze, it may be worth working through my Scala for Statistical Computing short course. This library originated from example code prepared for that course.

Author

This library is Copyright (C) 2017-2024 Darren J Wilkinson, but released as open source software under an Apache 2.0 license.

About

Scala library for fitting linear and generalised linear statistical models

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published