RLT Package Testing Classification Functions and Features
Ruoqing Zhu
Last Updated: July 17, 2025
Test-Cla.RmdInstall and Load Package
Install and load the GitHub version of the RLT package. Do not use the CRAN version.
# install.packages("devtools")
# devtools::install_github("teazrq/RLT")
library(RLT)
## RLT and Random Forests v4.2.6
## pre-release at github.com/teazrq/RLTLoad other packages used in this guide.
library(randomForest)
library(randomForestSRC)
library(ranger)
library(parallel)Benchmark Against Existing Packages
We generate a dataset with 1000 observations and 400 variables, with 200 continuous variables and 200 categorical ones with three categories.
# Set seed for reproducibility
set.seed(1)
# Define data size
trainn <- 800
testn <- 1000
n <- trainn + testn
p <- 30
# Generate continuous variables (X1) and categorical variables (X2)
X1 <- matrix(rnorm(n * p / 2), n, p / 2)
#X2 <- matrix(rnorm(n * p / 2), n, p / 2)
X2 <- matrix(as.integer(runif(n * p / 2) * 10), n, p / 2)
# Combine continuous and categorical variables into a data frame (X)
X <- data.frame(X1, X2)
# Convert the second half of the columns in X to factors
X[, (p / 2 + 1):p] <- lapply(X[, (p / 2 + 1):p], as.factor)
# Generate outcomes (y)
logit <- function(x) exp(x) / (1 + exp(x))
# y <- as.factor(rbinom(n, 1, prob = logit(1 + rowSums(X[, 1:5]) + 2 * (X[, p / 2 + 1] %in% c(1, 3)) + rnorm(n))) + 2)
y <- as.factor(rbinom(n, 1, prob = logit(1 + 1*X[, 2] + 3*(X[, p] %in% c(1, 3, 5, 7)))) + 2)
# Set tuning parameters
ntrees <- 1000
ncores <- 10
nmin <- 20
mtry <- p/2
samplereplace <- TRUE
sampleprob <- 0.75
rule <- "best"
nsplit <- ifelse(rule == "best", 0, 3)
importance <- TRUE
# Split data into training and testing sets
trainX <- X[1:trainn, ]
trainY <- y[1:trainn]
testX <- X[(trainn + 1):(trainn + testn), ]
testY <- y[(trainn + 1):(trainn + testn)]
# recording results
metric = data.frame(matrix(NA, 5, 6))
rownames(metric) = c("RLT", "randomForestSRC", "randomForest", "ranger", "ranger fast")
colnames(metric) = c("fit.time", "pred.time", "oob.error",
"pred.error", "obj.size", "ave.tree.size")
# using RLT package
start_time <- Sys.time()
RLTfit <- RLT(trainX, trainY, model = "classification",
ntrees = ntrees, mtry = mtry, nmin = nmin,
resample.prob = sampleprob, split.gen = rule,
resample.replace = samplereplace,
nsplit = nsplit, importance = importance,
param.control = list("alpha" = 0),
ncores = ncores, verbose = TRUE)
## Classification Random Forest ...
## ---------- Parameters Summary ----------
## (N, P) = (800, 30)
## # of trees = 1000
## (mtry, nmin) = (15, 20)
## split generate = Best
## sampling = 0.75 w/ replace
## (Obs, Var) weights = (No, No)
## importance = permute
## reinforcement = No
## ----------------------------------------
## Do not have 10 cores, use maximum 4 cores.
metric[1, 1] = difftime(Sys.time(), start_time, units = "secs")
start_time <- Sys.time()
RLTPred <- predict(RLTfit, testX, ncores = ncores)
metric[1, 2] = difftime(Sys.time(), start_time, units = "secs")
metric[1, 3] = mean(RLTfit$Prediction != trainY)
metric[1, 4] = mean(RLTPred$Prediction != testY)
metric[1, 5] = object.size(RLTfit)
metric[1, 6] = mean(unlist(lapply(RLTfit$FittedForest$SplitVar, length)))
# use randomForestSRC
options(rf.cores = ncores)
start_time <- Sys.time()
rsffit <- rfsrc(y ~ ., data = data.frame(trainX, "y"= trainY),
ntree = ntrees, nodesize = nmin/2, mtry = mtry,
samptype = ifelse(samplereplace == TRUE, "swor", "swr"),
nsplit = nsplit, sampsize = trainn*sampleprob,
importance = ifelse(importance, "permute", "none"))
metric[2, 1] = difftime(Sys.time(), start_time, units = "secs")
start_time <- Sys.time()
rsfpred = predict(rsffit, data.frame(testX))
metric[2, 2] = difftime(Sys.time(), start_time, units = "secs")
metric[2, 3] = mean(apply(rsffit$predicted.oob, 1, which.max) + 1 != trainY)
metric[2, 4] = mean(rsfpred$class != testY)
metric[2, 5] = object.size(rsffit)
metric[2, 6] = rsffit$forest$totalNodeCount / rsffit$ntree
# use randomForest
start_time <- Sys.time()
rf.fit <- randomForest(trainX, trainY, ntree = ntrees,
mtry = mtry, nodesize = nmin,
replace = samplereplace,
sampsize = trainn*sampleprob,
importance = importance)
metric[3, 1] = difftime(Sys.time(), start_time, units = "secs")
start_time <- Sys.time()
rf.pred <- predict(rf.fit, testX)
metric[3, 2] = difftime(Sys.time(), start_time, units = "secs")
metric[3, 3] = mean(rf.fit$predicted != trainY)
metric[3, 4] = mean(rf.pred != testY)
metric[3, 5] = object.size(rf.fit)
metric[3, 6] = mean(colSums(rf.fit$forest$nodestatus != 0))
# use ranger
start_time <- Sys.time()
rangerfit <- ranger(y ~ ., data = data.frame(trainX, "y"= trainY),
num.trees = ntrees, min.node.size = nmin,
mtry = mtry, num.threads = ncores,
replace = samplereplace,
sample.fraction = sampleprob,
importance = "permutation",
respect.unordered.factors = "partition")
## Growing trees.. Progress: 60%. Estimated remaining time: 21 seconds.
metric[4, 1] = difftime(Sys.time(), start_time, units = "secs")
start_time <- Sys.time()
rangerpred = predict(rangerfit, data.frame(testX))
metric[4, 2] = difftime(Sys.time(), start_time, units = "secs")
metric[4, 3] = mean(rangerfit$predictions != trainY)
metric[4, 4] = mean(rangerpred$predictions != testY)
metric[4, 5] = object.size(rangerfit)
metric[4, 6] = mean(unlist(lapply(rangerfit$forest$split.varIDs, length)))
# use ranger without partitioning
start_time <- Sys.time()
rangerfast <- ranger(y ~ ., data = data.frame(trainX, "y"= trainY),
num.trees = ntrees, min.node.size = nmin,
mtry = mtry, num.threads = ncores,
replace = samplereplace,
sample.fraction = sampleprob,
importance = "permutation")
metric[5, 1] = difftime(Sys.time(), start_time, units = "secs")
start_time <- Sys.time()
rangerpred = predict(rangerfast, data.frame(testX))
metric[5, 2] = difftime(Sys.time(), start_time, units = "secs")
metric[5, 3] = mean(rangerfast$predictions != trainY)
metric[5, 4] = mean(rangerpred$predictions != testY)
metric[5, 5] = object.size(rangerfast)
metric[5, 6] = mean(unlist(lapply(rangerfast$forest$split.varIDs, length)))
# performance summary
metric
## fit.time pred.time oob.error pred.error obj.size ave.tree.size
## RLT 3.8877819 0.04142427 0.18375 0.182 4432112 58.618
## randomForestSRC 42.5205061 0.17455602 0.18250 0.171 12907088 71.766
## randomForest 5.0521028 0.04990530 0.18000 0.176 2122056 50.650
## ranger 52.2684577 0.09444880 0.19250 0.183 2182560 58.514
## ranger fast 0.6899104 0.08232188 0.20125 0.201 2754232 76.386Print a Single Tree
You can use the get.one.tree() function to peek into a
single tree.
get.one.tree(RLTfit, 1)
## Tree #1 in the fitted classification forest:
## SplitVar SplitValue LeftNode RightNode NodeWeight Prob.of..2 Prob.of..3
## 1 X2 -0.6926111 2 3 600 0.193333333 0.8066667
## 2 X4.1 (F) 448.0000000 4 5 146 0.390410959 0.6095890
## 3 X15.1 (F) 852.0000000 26 27 454 0.129955947 0.8700441
## 4 X8.1 (F) 306.0000000 6 7 98 0.510204082 0.4897959
## 5 X3 0.6296932 20 21 48 0.145833333 0.8541667
## 6 X15.1 (F) 682.0000000 8 9 57 0.666666667 0.3333333
## 7 X10.1 (F) 442.0000000 14 15 41 0.292682927 0.7073171
## 8 <NA> NA NA NA 12 0.166666667 0.8333333
## 9 X3 1.0694546 10 11 45 0.800000000 0.2000000
## 10 X14.1 (F) 528.0000000 12 13 40 0.875000000 0.1250000
## 11 <NA> NA NA NA 5 0.200000000 0.8000000
## 12 <NA> NA NA NA 34 1.000000000 0.0000000
## 13 <NA> NA NA NA 6 0.166666667 0.8333333
## 14 <NA> NA NA NA 12 0.833333333 0.1666667
## 15 X9 0.9688454 16 17 29 0.068965517 0.9310345
## 16 X3.1 (F) 512.0000000 18 19 28 0.035714286 0.9642857
## 17 <NA> NA NA NA 1 1.000000000 0.0000000
## 18 <NA> NA NA NA 27 0.000000000 1.0000000
## 19 <NA> NA NA NA 1 1.000000000 0.0000000
## 20 X3.1 (F) 32.0000000 22 23 40 0.050000000 0.9500000
## 21 <NA> NA NA NA 8 0.625000000 0.3750000
## 22 X13.1 (F) 256.0000000 24 25 39 0.025641026 0.9743590
## 23 <NA> NA NA NA 1 1.000000000 0.0000000
## 24 <NA> NA NA NA 38 0.000000000 1.0000000
## 25 <NA> NA NA NA 1 1.000000000 0.0000000
## 26 X4.1 (F) 778.0000000 28 29 227 0.215859031 0.7841410
## 27 X3 2.0473906 52 53 227 0.044052863 0.9559471
## 28 X1.1 (F) 892.0000000 30 31 145 0.296551724 0.7034483
## 29 X11.1 (F) 2.0000000 46 47 82 0.073170732 0.9268293
## 30 X4 0.3217051 32 33 38 0.552631579 0.4473684
## 31 X13.1 (F) 256.0000000 36 37 107 0.205607477 0.7943925
## 32 X7.1 (F) 902.0000000 34 35 27 0.777777778 0.2222222
## 33 <NA> NA NA NA 11 0.000000000 1.0000000
## 34 <NA> NA NA NA 7 0.285714286 0.7142857
## 35 <NA> NA NA NA 20 0.950000000 0.0500000
## 36 X5 0.2664197 38 39 96 0.145833333 0.8541667
## 37 <NA> NA NA NA 11 0.727272727 0.2727273
## 38 X7.1 (F) 1018.0000000 40 41 67 0.059701493 0.9402985
## 39 X5 0.9116686 44 45 29 0.344827586 0.6551724
## 40 <NA> NA NA NA 10 0.300000000 0.7000000
## 41 X12 1.1907859 42 43 57 0.017543860 0.9824561
## 42 <NA> NA NA NA 54 0.000000000 1.0000000
## 43 <NA> NA NA NA 3 0.333333333 0.6666667
## 44 <NA> NA NA NA 13 0.692307692 0.3076923
## 45 <NA> NA NA NA 16 0.062500000 0.9375000
## 46 X11 -2.4272374 48 49 72 0.027777778 0.9722222
## 47 <NA> NA NA NA 10 0.400000000 0.6000000
## 48 <NA> NA NA NA 1 1.000000000 0.0000000
## 49 X7.1 (F) 512.0000000 50 51 71 0.014084507 0.9859155
## 50 <NA> NA NA NA 66 0.000000000 1.0000000
## 51 <NA> NA NA NA 5 0.200000000 0.8000000
## 52 X7 1.2252040 54 55 222 0.031531532 0.9684685
## 53 <NA> NA NA NA 5 0.600000000 0.4000000
## 54 X11 2.3314162 56 57 192 0.010416667 0.9895833
## 55 X7 1.4388455 60 61 30 0.166666667 0.8333333
## 56 X11 -2.0074093 58 59 190 0.005263158 0.9947368
## 57 <NA> NA NA NA 2 0.500000000 0.5000000
## 58 <NA> NA NA NA 9 0.111111111 0.8888889
## 59 <NA> NA NA NA 181 0.000000000 1.0000000
## 60 <NA> NA NA NA 7 0.714285714 0.2857143
## 61 <NA> NA NA NA 23 0.000000000 1.0000000