KSLCore/ksl.controls.experiments/LinearModel

LinearModel

class LinearModel(val mainEffects: Set<String>, type: LinearModel.Type = Type.FirstOrder)

This class provides the ability to specify a linear model (for use in regression and design of experiments). This is only a string specification of the model. The terms are specified by the names of the factors.

Example: A model with 3 main effects, "A", "B", "C", with full model = "A B C AB AC BC ABC". // y = b0 + b1A + b2B + b3C + b12AB + b13AC + b23BC + b123ABC val m3 = LinearModel(setOf("A", "B", "C")) m3.specifyAllTerms()

The properties termsAsMap and termsAsList provide access to the specification.

Parameters

mainEffects

the names of the factors (regressors) in a set

type

the type of model to start with. The default is first order terms (main effects) as define by the provided set of main effects.

Constructors

LinearModel

constructor(mainEffects: Set<String>, type: LinearModel.Type = Type.FirstOrder)

Types

Companion

object Companion

Type

enum Type : Enum<LinearModel.Type>

Properties

intercept

var intercept: Boolean

By default, we assume an intercept term

mainEffects

val mainEffects: Set<String>

termsAsList

val termsAsList: List<List<String>>

Returns the terms of the model as a list of lists

termsAsMap

val termsAsMap: Map<String, List<String>>

Returns the terms of the model as a map. The key to the map is the string join of the elements in the associated list with the elements separated by '*'.

Functions

asString

fun asString(): String

cubic

fun cubic(name: String): LinearModel

Shorthand for adding a cubic term.

isValidTerm

fun isValidTerm(list: List<String>): Boolean

Every string in the list must be within the main effects set to be valid.

nWay

fun nWay(set: Set<String>): LinearModel

Shorthand for adding an n-way interaction term.

parseFromString

fun parseFromString(parseString: String): LinearModel

Assumes a parsable string and converts it to a list of terms for specifying the model from the string.

quadratic

fun quadratic(name: String): LinearModel

Shorthand for adding a quadratic term.

specify

fun specify(list: List<List<String>>): LinearModel

Repeatedly calls term() with the elements in the list to specify an entire model.

specifyAllTerms

fun specifyAllTerms(): LinearModel

Specifies a model with all terms (main effects, first order interactions, 2nd order interactions, etc.

term

fun term(list: List<String>): LinearModel

Add a term to the model. It is assumed that the list of strings represents a product term. Thus, listOf("A", "B") is the interaction term "A*B". The elements of the list must be valid single (main) effect term names.

threeWay

fun threeWay(name1: String, name2: String, name3: String): LinearModel

Shorthand for adding a three-way interaction term. The names must be different.

toString

open override fun toString(): String

twoWay

fun twoWay(name1: String, name2: String): LinearModel

Shorthand for adding a two-way interaction term. The names must be different.