htree (version 1.0.0)

quantile_hrf: Quantiles for historical random forest

Description

Quantiles for historical random forest

Usage

quantile_hrf(object, x=NULL,time=NULL,id=NULL,
	ntrees=NULL,prob=seq(.1,.9,length=10))

Arguments

object

The return list from running hrf.

x

a data frame or matrix containing new data. If NULL then training data in object is used.

time

observation times.

id

subject identifiers

ntrees

number of trees to use in prediction.

prob

Probabilities at which quantiles are to be computed

Value

Returns a matrix of quantiles, with columns corresponding values of prob.

Details

Meinhausen (2006) showed how a random forest model could be used to estimate conditional quantiles. quantile_hrf implements this technique for a random forest with historical regression trees.

References

Meinhausen (2006) 'Quantile regression forests' Journal of Machine Learning Research.

See Also

hrf

Examples

# NOT RUN {
# }
# NOT RUN {



# -------------------- ##
# Boston Housing data 
# -------------------- ##
library(mlbench)
library(randomForest)

data(BostonHousing)
dat=as.data.frame(na.omit(BostonHousing))


## omitting arguments time/id assumes rows are iid 
h=hrf(x=dat,yindx="medv",ntrees=500)

## get deciles on training data 
quant=quantile_hrf(h,prob=seq(.1,.9,length=10))


# ------------------------------------------------ ##
# Simulated data from model with random intercept 
# ------------------------------------------------ ##

p=5;sigma_e=.5;sigma_a=.5;v=rep(1,p);n=500;pnoise=2
random_intercept=rnorm(n,sd=sigma_a)
random_intercept=as.numeric(matrix(random_intercept,nrow=p,ncol=n,byrow=TRUE))
x=random_intercept+rnorm(n*p,sd=sigma_e)
id=sort(rep(1:n,p))
time<-rep(1:p,n)
znoise=matrix(rnorm(n*p*pnoise),ncol=pnoise)
xx=cbind(time,x,znoise)

# fit historical random forest
hb=hrf(x=xx,time=time,id=id,yindx=2,ntrees=100,mtry=4,nsamp=5)

# get deciles on training data 
quant=quantile_hrf(hb,prob=seq(.1,.9,length=10))
# }