MixRF: Mixed Random Forest

In randel/MixRF: A Random-Forest-Based Approach for Imputing Clustered Incomplete Data

Description

The function to fit a random forest with random effects.

Usage

MixRF(Y, X, random, data, initialRandomEffects = 0, ErrorTolerance = 0.001, 
    MaxIterations = 1000)

Arguments

Y	The outcome variable.
X	A data frame or matrix contains the predictors.
random	A string in lme4 format indicates the random effect model.
data	The data set as a data frame.
initialRandomEffects	The initial values for random effects.
ErrorTolerance	The tolerance for log-likelihood.
MaxIterations	The maximum iteration times.

Value

A list contains the random forest ($forest), mixed model ($MixedModel), and random effects ($RandomEffects). See the example below for the usage.

Examples

data(sleepstudy)

tmp = MixRF(Y = sleepstudy$Reaction, X = as.data.frame(sleepstudy$Days), 
    random = "(Days|Subject)", data = sleepstudy, initialRandomEffects = 0, 
    ErrorTolerance = 0.01, MaxIterations = 100)

# tmp$forest

# tmp$MixedModel

# tmp$RandomEffects

randel/MixRF documentation built on May 26, 2019, 10:59 p.m.