Regression: Regression

In simdat.base: Simulating data in R

Description

Constructs a linear Regression model.

Usage

Regression(DV, family,predictors,predictorModels,N,mu,sigma,nu,tau,beta)

Arguments

DV	either a list or a RandomVariable

.

family	a gamlss.family object
predictors	either a data.frame or VariableList containing the predictors in the design.
predictorModels	(optional) a list with models to simulate random predictors.
N	a vector with the number of observations in each cell of the design.
mu	a vector with the values of mu in each cell of the design.
sigma	a vector with the values of sigma in each cell of the design.
nu	a vector with the values of nu in each cell of the design.
tau	a vector with the values of tau in each cell of the design.
beta	a vector or matrix with the slopes of the covariates in each cell of the design.

Details

The Regression function constructs a (multiple) regression model. The workhorse function is GLM_create, which should not be called directly.

Value

An object of class Regression (directly extending the GLM class).

Author(s)

Maarten Speekenbrink

Examples

mod <- Regression(predictors=VariableList(list(RandomIntervalVariable(numeric(10),digits=8,min=-Inf,max=Inf,name="X"))),predictorModels=ModelList(list(NormalModel(mean=0,sd=1))),N=10,mu=0,sigma=1,nu=NULL,tau=NULL,beta=1,DV=list(name="Y",min=-Inf,max=Inf,digits=8),family=NO())
mod <- simulate(mod)

# and a random interval covariate
x <- VariableList(list(
  RandomIntervalVariable(numeric(1),name="X1",min=-5,max=5,digits=2),
  RandomIntervalVariable(numeric(1),name="X2",min=-9,max=11,digits=4)
  ))
# and a random interval dependent variable
d <- RandomIntervalVariable(numeric(1),name="Y",min=-Inf,max=Inf,digits=2)
mod2 <- Regression(predictors=x,
  predictorModels=ModelList(list(UniformModel(),NormalModel(mean=1,sd=2))),
  N=10000,mu=2,sigma=3,beta=matrix(c(1,2),ncol=2),
  DV=d,family=NO())
mod2 <- simulate(mod2)
summary(lm(Y~X1+X2,data=getData(simulate(mod2))))

simdat.base documentation built on May 2, 2019, 5:53 p.m.