Results/Plot_results_computation_time_and_model_sizes.R
In chstaerk/SubBoost: Subspace Boosting With Multivariable Base-learners
library("unikn")
#setwd("//epi-login/daten/CS/AdaSubBoost_Revision")
# Load results for low-dimensional setting
load("Sim_Toeplitz08_p20_n100_nSim500_Iter1000_fixedS0_4_autostop_25_07.RData")
results_low <- results
# Load results for high-dimensional setting (a)
load("Sim_Toeplitz08_p1000_n100_nSim500_Iter5000_fixedS0_10_autostop_25_07.RData")
results_high_a <- results
# Load results for high-dimensional setting (b)
load("Sim_Toeplitz08_p1000_n100_nSim500_Iter5000_randomS0_10_autostop_25_07.RData")
results_high_b <- results
# Load results for high-dimensional setting (c)
load("Sim_Toeplitz08_p1000_n1000_nSim500_Iter10000_fixedS0_100_autostop_choice_of_criterion_19_07.RData")
results_criterion <- results
load("Sim_Toeplitz08_p1000_n1000_nSim500_Iter10000_fixedS0_100_autostop_25_07.RData")
results_high_c <- results
take_element <- function(list,element1,element2,nSim){
#l = length(list)
l = nSim
vec = numeric(l)
for (i in 1:l) vec[i] = list[[i]][[element1]][[element2]]
return(vec)
}
take_element2 <- function(list,element1){
l = length(list)
vec = numeric(l)
for (i in 1:l) vec[i] = list[[i]][[element1]]
return(vec)
}
take_element3 <- function(list,element1,element2,element3,nSim){
l = nSim
vec = numeric(l)
for (i in 1:l) vec[i] = list[[i]][[element1]][[element2]][[element3]]
return(vec)
}
n_algos = length(results[[1]])
nSim = results$nSim
par(mar=c(4.7,3,4,1))
color_mean = c("black")
pal_col <- rev(seecol(pal_signal, n = 4, alpha = 0.67))
color_boxes = c(pal_col[1:4], rep("gray",6))
method_names = c("SubBoost", "RSubBoost", "AdaSubBoost", "L2Boost", "XGBoost", "TwinBoost", "StabSel",
"Lasso", "ENet", "ReLasso")
cex.size = 1.2
cex.lab.size = 1.1
cex.axis.size = 0.9 #0.75
par(cex.axis = cex.axis.size, cex.lab = cex.lab.size, cex.main = cex.size, las = 1)
par(font.axis = 2)
angle = 45
cex.names = 1
par(mfrow=c(2,2))
boxplot(take_element(results_low,1,5,nSim), take_element(results_low,2,5,nSim), take_element(results_low,3,5,nSim), take_element(results_low,4,5,nSim),
take_element(results_low,11,5,nSim),
take_element(results_low,5,5,nSim), take_element(results_low,6,5,nSim),
take_element(results_low,7,5,nSim), take_element(results_low,8,5,nSim), take_element(results_low,9,5,nSim),
xaxt="n", main ="Low-dimensional setting",
col = color_boxes,
outcol=color_boxes,outpch = 19)
text(1:length(method_names), par("usr")[3] - 0.75, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
color_boxes = c(pal_col[2:4], rep("gray",6))
method_names = c("RSubBoost", "AdaSubBoost", "L2Boost", "XGBoost", "TwinBoost", "StabSel",
"Lasso", "ENet", "ReLasso")
title(main = ("Computational time (in s)"), line = 0.4, cex.main = 1.1)
boxplot(take_element(results_high_a,1,5,nSim), take_element(results_high_a,2,5,nSim), take_element(results_high_a,3,5,nSim),
take_element(results_high_a,10,5,nSim),
take_element(results_high_a,4,5,nSim), take_element(results_high_a,5,5,nSim), take_element(results_high_a,6,5,nSim),
take_element(results_high_a,7,5,nSim), take_element(results_high_a,8,5,nSim),
xaxt="n", main ="Sparse high-dimensional setting (a)",
col = color_boxes,
outcol=color_boxes,outpch = 19)
text(1:length(method_names) , par("usr")[3] - 5, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Computational time (in s)"), line = 0.4, cex.main = 1.1)
boxplot(take_element(results_high_b,1,5,nSim), take_element(results_high_b,2,5,nSim), take_element(results_high_b,3,5,nSim),
take_element(results_high_b,10,5,nSim),
take_element(results_high_b,4,5,nSim), take_element(results_high_b,5,5,nSim), take_element(results_high_b,6,5,nSim),
take_element(results_high_b,7,5,nSim), take_element(results_high_b,8,5,nSim),
xaxt="n", main ="Sparse high-dimensional setting (b)",
col = color_boxes,
outcol=color_boxes,outpch = 19)
text(1:length(method_names) , par("usr")[3] - 5, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Computational time (in s)"), line = 0.4, cex.main = 1.1)
cex.names = 0.9
color_boxes = c(pal_col[2:3], pal_col[2:4], rep("gray",6))
method_names = c("RSubBoost \n AIC", "AdaSubBoost \n AIC", "RSubBoost \n EBIC1", "AdaSubBoost \n EBIC1",
"L2Boost", "XGBoost", "TwinBoost", "StabSel",
"Lasso", "ENet", "ReLasso")
boxplot(take_element3(results_criterion,1,1,5,nSim), take_element3(results_criterion,2,1,5,nSim),
take_element3(results_criterion,1,3,5,nSim), take_element3(results_criterion,2,3,5,nSim),
take_element(results_high_c,3,5,nSim),
take_element(results_high_c,10,5,nSim),
take_element(results_high_c,4,5,nSim), take_element(results_high_c,5,5,nSim), take_element(results_high_c,6,5,nSim),
take_element(results_high_c,7,5,nSim), take_element(results_high_c,8,5,nSim),
xaxt="n", main ="Non-sparse high-dimensional setting",
col = color_boxes,
outcol=color_boxes,outpch = 19)
text(1:length(method_names) , par("usr")[3] - 50, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Computational time (in s)"), line = 0.4, cex.main = 1.1)
#########################################################################
##### Comp. time for biomedical data applications
#########################################################################
load("LOOCV_Bodyfat_CV_25_07.RData")
results_bodyfat <- LOOCV_results
load("LOOCV_Diabetes_CV_25_07.RData")
results_diabetes <- LOOCV_results
load("LOOCV_Riboflavin_CV_25_07.RData")
results_riboflavin <- LOOCV_results
load("LOOCV_PCR_CV_25_07.RData")
results_PCR <- LOOCV_results
par(mar=c(4.7,3,4,1))
color_mean = c("black")
pal_col <- rev(seecol(pal_signal, n = 4, alpha = 0.67))
color_boxes = c(pal_col[1:4], rep("gray",6))
method_names = c("SubBoost", "RSubBoost", "AdaSubBoost", "L2Boost", "XGBoost", "TwinBoost", "StabSel",
"Lasso", "ENet", "ReLasso")
cex.size = 1.2
cex.lab.size = 1.1
cex.axis.size = 1 #0.75
par(cex.axis = cex.axis.size, cex.lab = cex.lab.size, cex.main = cex.size, las = 1)
par(font.axis = 2)
angle = 45
cex.names = 1
par(mfrow=c(2,2))
boxplot(take_element2(results_bodyfat,"timeSubBoost"),
take_element2(results_bodyfat,"timeRSubBoost"),
take_element2(results_bodyfat,"timeAdaSubBoost"),
take_element2(results_bodyfat,"timeBoosting"),
take_element2(results_bodyfat,"timeXGBoostCD"),
take_element2(results_bodyfat,"timeTwinBoost"),
take_element2(results_bodyfat,"timeStability"),
take_element2(results_bodyfat,"timeLasso"),
take_element2(results_bodyfat,"timeENet"),
take_element2(results_bodyfat,"timeReLasso"),
xaxt="n", main ="Bodyfat data (n=71, p=9)",
col = color_boxes,
outcol=color_boxes,outpch = 19)
text(1:length(method_names), par("usr")[3] - 0.5, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Computational time (in s)"), line = 0.4, cex.main = 1.1)
boxplot(take_element2(results_diabetes,"timeSubBoost"),
take_element2(results_diabetes,"timeRSubBoost"),
take_element2(results_diabetes,"timeAdaSubBoost"),
take_element2(results_diabetes,"timeBoosting"),
take_element2(results_diabetes,"timeXGBoostCD"),
take_element2(results_diabetes,"timeTwinBoost"),
take_element2(results_diabetes,"timeStability"),
take_element2(results_diabetes,"timeLasso"),
take_element2(results_diabetes,"timeENet"),
take_element2(results_diabetes,"timeReLasso"),
xaxt="n", main ="Diabetes data (n=442, p=10)",
col = color_boxes,
outcol=color_boxes,outpch = 19)
text(1:length(method_names), par("usr")[3] - 0.5, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Computational time (in s)"), line = 0.4, cex.main = 1.1)
method_names = c("RSubBoost", "AdaSubBoost", "L2Boost", "XGBoost", "TwinBoost", "StabSel",
"Lasso", "ENet", "ReLasso")
color_boxes = c(pal_col[2:4], rep("gray",6))
boxplot(take_element2(results_riboflavin,"timeRSubBoost"),
take_element2(results_riboflavin,"timeAdaSubBoost"),
take_element2(results_riboflavin,"timeBoosting"),
take_element2(results_riboflavin,"timeXGBoostCD"),
take_element2(results_riboflavin,"timeTwinBoost"),
take_element2(results_riboflavin,"timeStability"),
take_element2(results_riboflavin,"timeLasso"),
take_element2(results_riboflavin,"timeENet"),
take_element2(results_riboflavin,"timeReLasso"),
xaxt="n", main ="Riboflavin data (n=71, p=4088)",
col = color_boxes,
outcol=color_boxes,outpch = 19)
text(1:length(method_names), par("usr")[3] - 5, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Computational time (in s)"), line = 0.4, cex.main = 1.1)
boxplot(take_element2(results_PCR,"timeRSubBoost"),
take_element2(results_PCR,"timeAdaSubBoost"),
take_element2(results_PCR,"timeBoosting"),
take_element2(results_PCR,"timeXGBoostCD"),
take_element2(results_PCR,"timeTwinBoost"),
take_element2(results_PCR,"timeStability"),
take_element2(results_PCR,"timeLasso"),
take_element2(results_PCR,"timeENet"),
take_element2(results_PCR,"timeReLasso"),
xaxt="n", main ="PCR data (n=60, p=22,575)",
col = color_boxes,
outcol=color_boxes,outpch = 19)
text(1:length(method_names), par("usr")[3] - 15, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Computational time (in s)"), line = 0.4, cex.main = 1.1)
####################################################################################
### Model sizes
####################################################################################
SubBoost_sizes_low <- take_element(results_low,1,1,nSim) + (4 - take_element(results_low,1,2,nSim))
RSubBoost_sizes_low <- take_element(results_low,2,1,nSim) + (4 - take_element(results_low,2,2,nSim))
RSubBoost_sizes_high_a <- take_element(results_high_a,1,1,nSim) + (10 - take_element(results_high_a,1,2,nSim))
RSubBoost_sizes_high_b <- take_element(results_high_b,1,1,nSim) + (10 - take_element(results_high_b,1,2,nSim))
RSubBoost_sizes_high_c <- take_element(results_high_c,1,1,nSim) + (100 - take_element(results_high_c,1,2,nSim))
AdaSubBoost_sizes_low <- take_element(results_low,3,1,nSim) + (4 - take_element(results_low,3,2,nSim))
AdaSubBoost_sizes_high_a <- take_element(results_high_a,2,1,nSim) + (10 - take_element(results_high_a,2,2,nSim))
AdaSubBoost_sizes_high_b <- take_element(results_high_b,2,1,nSim) + (10 - take_element(results_high_b,2,2,nSim))
AdaSubBoost_sizes_high_c <- take_element(results_high_c,2,1,nSim) + (100 - take_element(results_high_c,2,2,nSim))
S0_sizes_low <- take_element(results_low,10,1,nSim) + (4 - take_element(results_low,10,2,nSim))
S0_sizes_high_a <-take_element(results_high_a,9,1,nSim) + (10 - take_element(results_high_a,9,2,nSim))
S0_sizes_high_b <-take_element(results_high_b,9,1,nSim) + (10 - take_element(results_high_b,9,2,nSim))
S0_sizes_high_c <-take_element(results_high_c,9,1,nSim) + (100 - take_element(results_high_c,9,2,nSim))
par(mar=c(4.7,3,4,1))
cex.size = 1.2
cex.lab.size = 1.1
cex.axis.size = 1
par(cex.axis = cex.axis.size, cex.lab = cex.lab.size, cex.main = cex.size, las = 1)
par(font.axis = 2)
angle = 45
cex.names = 1.1
method_names <- c("S[0]", "SubBoost", "RSubBoost", "AdaSubBoost")
par(mfrow=c(1,4))
color_boxes = c("gray", pal_col[1:3])
boxplot(S0_sizes_low, SubBoost_sizes_low, RSubBoost_sizes_low, AdaSubBoost_sizes_low,
ylim=c(0, 20),
col = color_boxes, outcol=color_boxes,
xaxt="n",
main = "Low-dimensional setting")
text(1:length(method_names), par("usr")[3] - 0.2, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Model size (number of variables)"), line = 0.4, cex.main = 1.1)
abline(h=4, lty=1, col="red")
method_names <- c("S[0]", "RSubBoost", "AdaSubBoost")
color_boxes = c("gray", pal_col[2:3])
boxplot(S0_sizes_high_a, RSubBoost_sizes_high_a, AdaSubBoost_sizes_high_a,
ylim=c(0, 20),
col = color_boxes, outcol=color_boxes,
xaxt="n",
main = "Sparse high-dimensional setting (a)")
text(1:length(method_names), par("usr")[3] - 0.2, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Model size (number of variables)"), line = 0.4, cex.main = 1.1)
abline(h=10, lty=1, col="red")
boxplot(S0_sizes_high_b, RSubBoost_sizes_high_b, AdaSubBoost_sizes_high_b,
ylim=c(0, 20),
col = color_boxes, outcol=color_boxes,
xaxt="n",
main = "Sparse high-dimensional setting (b)")
text(1:length(method_names), par("usr")[3] - 0.2, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Model size (number of variables)"), line = 0.4, cex.main = 1.1)
abline(h=10, lty=1, col="red")
boxplot(S0_sizes_high_c, RSubBoost_sizes_high_c, AdaSubBoost_sizes_high_c,
ylim=c(0, 200),
col = color_boxes, outcol=color_boxes,
xaxt="n",
main = "Non-sparse high-dimensional setting")
text(1:length(method_names), par("usr")[3] - 2, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Model size (number of variables)"), line = 0.4, cex.main = 1.1)
abline(h=100, lty=1, col="red")
chstaerk/SubBoost documentation built on Dec. 19, 2021, 4:06 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
library("unikn")
#setwd("//epi-login/daten/CS/AdaSubBoost_Revision")
# Load results for low-dimensional setting
load("Sim_Toeplitz08_p20_n100_nSim500_Iter1000_fixedS0_4_autostop_25_07.RData")
results_low <- results
# Load results for high-dimensional setting (a)
load("Sim_Toeplitz08_p1000_n100_nSim500_Iter5000_fixedS0_10_autostop_25_07.RData")
results_high_a <- results
# Load results for high-dimensional setting (b)
load("Sim_Toeplitz08_p1000_n100_nSim500_Iter5000_randomS0_10_autostop_25_07.RData")
results_high_b <- results
# Load results for high-dimensional setting (c)
load("Sim_Toeplitz08_p1000_n1000_nSim500_Iter10000_fixedS0_100_autostop_choice_of_criterion_19_07.RData")
results_criterion <- results
load("Sim_Toeplitz08_p1000_n1000_nSim500_Iter10000_fixedS0_100_autostop_25_07.RData")
results_high_c <- results
take_element <- function(list,element1,element2,nSim){
#l = length(list)
l = nSim
vec = numeric(l)
for (i in 1:l) vec[i] = list[[i]][[element1]][[element2]]
return(vec)
}
take_element2 <- function(list,element1){
l = length(list)
vec = numeric(l)
for (i in 1:l) vec[i] = list[[i]][[element1]]
return(vec)
}
take_element3 <- function(list,element1,element2,element3,nSim){
l = nSim
vec = numeric(l)
for (i in 1:l) vec[i] = list[[i]][[element1]][[element2]][[element3]]
return(vec)
}
n_algos = length(results[[1]])
nSim = results$nSim
par(mar=c(4.7,3,4,1))
color_mean = c("black")
pal_col <- rev(seecol(pal_signal, n = 4, alpha = 0.67))
color_boxes = c(pal_col[1:4], rep("gray",6))
method_names = c("SubBoost", "RSubBoost", "AdaSubBoost", "L2Boost", "XGBoost", "TwinBoost", "StabSel",
"Lasso", "ENet", "ReLasso")
cex.size = 1.2
cex.lab.size = 1.1
cex.axis.size = 0.9 #0.75
par(cex.axis = cex.axis.size, cex.lab = cex.lab.size, cex.main = cex.size, las = 1)
par(font.axis = 2)
angle = 45
cex.names = 1
par(mfrow=c(2,2))
boxplot(take_element(results_low,1,5,nSim), take_element(results_low,2,5,nSim), take_element(results_low,3,5,nSim), take_element(results_low,4,5,nSim),
take_element(results_low,11,5,nSim),
take_element(results_low,5,5,nSim), take_element(results_low,6,5,nSim),
take_element(results_low,7,5,nSim), take_element(results_low,8,5,nSim), take_element(results_low,9,5,nSim),
xaxt="n", main ="Low-dimensional setting",
col = color_boxes,
outcol=color_boxes,outpch = 19)
text(1:length(method_names), par("usr")[3] - 0.75, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
color_boxes = c(pal_col[2:4], rep("gray",6))
method_names = c("RSubBoost", "AdaSubBoost", "L2Boost", "XGBoost", "TwinBoost", "StabSel",
"Lasso", "ENet", "ReLasso")
title(main = ("Computational time (in s)"), line = 0.4, cex.main = 1.1)
boxplot(take_element(results_high_a,1,5,nSim), take_element(results_high_a,2,5,nSim), take_element(results_high_a,3,5,nSim),
take_element(results_high_a,10,5,nSim),
take_element(results_high_a,4,5,nSim), take_element(results_high_a,5,5,nSim), take_element(results_high_a,6,5,nSim),
take_element(results_high_a,7,5,nSim), take_element(results_high_a,8,5,nSim),
xaxt="n", main ="Sparse high-dimensional setting (a)",
col = color_boxes,
outcol=color_boxes,outpch = 19)
text(1:length(method_names) , par("usr")[3] - 5, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Computational time (in s)"), line = 0.4, cex.main = 1.1)
boxplot(take_element(results_high_b,1,5,nSim), take_element(results_high_b,2,5,nSim), take_element(results_high_b,3,5,nSim),
take_element(results_high_b,10,5,nSim),
take_element(results_high_b,4,5,nSim), take_element(results_high_b,5,5,nSim), take_element(results_high_b,6,5,nSim),
take_element(results_high_b,7,5,nSim), take_element(results_high_b,8,5,nSim),
xaxt="n", main ="Sparse high-dimensional setting (b)",
col = color_boxes,
outcol=color_boxes,outpch = 19)
text(1:length(method_names) , par("usr")[3] - 5, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Computational time (in s)"), line = 0.4, cex.main = 1.1)
cex.names = 0.9
color_boxes = c(pal_col[2:3], pal_col[2:4], rep("gray",6))
method_names = c("RSubBoost \n AIC", "AdaSubBoost \n AIC", "RSubBoost \n EBIC1", "AdaSubBoost \n EBIC1",
"L2Boost", "XGBoost", "TwinBoost", "StabSel",
"Lasso", "ENet", "ReLasso")
boxplot(take_element3(results_criterion,1,1,5,nSim), take_element3(results_criterion,2,1,5,nSim),
take_element3(results_criterion,1,3,5,nSim), take_element3(results_criterion,2,3,5,nSim),
take_element(results_high_c,3,5,nSim),
take_element(results_high_c,10,5,nSim),
take_element(results_high_c,4,5,nSim), take_element(results_high_c,5,5,nSim), take_element(results_high_c,6,5,nSim),
take_element(results_high_c,7,5,nSim), take_element(results_high_c,8,5,nSim),
xaxt="n", main ="Non-sparse high-dimensional setting",
col = color_boxes,
outcol=color_boxes,outpch = 19)
text(1:length(method_names) , par("usr")[3] - 50, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Computational time (in s)"), line = 0.4, cex.main = 1.1)
#########################################################################
##### Comp. time for biomedical data applications
#########################################################################
load("LOOCV_Bodyfat_CV_25_07.RData")
results_bodyfat <- LOOCV_results
load("LOOCV_Diabetes_CV_25_07.RData")
results_diabetes <- LOOCV_results
load("LOOCV_Riboflavin_CV_25_07.RData")
results_riboflavin <- LOOCV_results
load("LOOCV_PCR_CV_25_07.RData")
results_PCR <- LOOCV_results
par(mar=c(4.7,3,4,1))
color_mean = c("black")
pal_col <- rev(seecol(pal_signal, n = 4, alpha = 0.67))
color_boxes = c(pal_col[1:4], rep("gray",6))
method_names = c("SubBoost", "RSubBoost", "AdaSubBoost", "L2Boost", "XGBoost", "TwinBoost", "StabSel",
"Lasso", "ENet", "ReLasso")
cex.size = 1.2
cex.lab.size = 1.1
cex.axis.size = 1 #0.75
par(cex.axis = cex.axis.size, cex.lab = cex.lab.size, cex.main = cex.size, las = 1)
par(font.axis = 2)
angle = 45
cex.names = 1
par(mfrow=c(2,2))
boxplot(take_element2(results_bodyfat,"timeSubBoost"),
take_element2(results_bodyfat,"timeRSubBoost"),
take_element2(results_bodyfat,"timeAdaSubBoost"),
take_element2(results_bodyfat,"timeBoosting"),
take_element2(results_bodyfat,"timeXGBoostCD"),
take_element2(results_bodyfat,"timeTwinBoost"),
take_element2(results_bodyfat,"timeStability"),
take_element2(results_bodyfat,"timeLasso"),
take_element2(results_bodyfat,"timeENet"),
take_element2(results_bodyfat,"timeReLasso"),
xaxt="n", main ="Bodyfat data (n=71, p=9)",
col = color_boxes,
outcol=color_boxes,outpch = 19)
text(1:length(method_names), par("usr")[3] - 0.5, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Computational time (in s)"), line = 0.4, cex.main = 1.1)
boxplot(take_element2(results_diabetes,"timeSubBoost"),
take_element2(results_diabetes,"timeRSubBoost"),
take_element2(results_diabetes,"timeAdaSubBoost"),
take_element2(results_diabetes,"timeBoosting"),
take_element2(results_diabetes,"timeXGBoostCD"),
take_element2(results_diabetes,"timeTwinBoost"),
take_element2(results_diabetes,"timeStability"),
take_element2(results_diabetes,"timeLasso"),
take_element2(results_diabetes,"timeENet"),
take_element2(results_diabetes,"timeReLasso"),
xaxt="n", main ="Diabetes data (n=442, p=10)",
col = color_boxes,
outcol=color_boxes,outpch = 19)
text(1:length(method_names), par("usr")[3] - 0.5, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Computational time (in s)"), line = 0.4, cex.main = 1.1)
method_names = c("RSubBoost", "AdaSubBoost", "L2Boost", "XGBoost", "TwinBoost", "StabSel",
"Lasso", "ENet", "ReLasso")
color_boxes = c(pal_col[2:4], rep("gray",6))
boxplot(take_element2(results_riboflavin,"timeRSubBoost"),
take_element2(results_riboflavin,"timeAdaSubBoost"),
take_element2(results_riboflavin,"timeBoosting"),
take_element2(results_riboflavin,"timeXGBoostCD"),
take_element2(results_riboflavin,"timeTwinBoost"),
take_element2(results_riboflavin,"timeStability"),
take_element2(results_riboflavin,"timeLasso"),
take_element2(results_riboflavin,"timeENet"),
take_element2(results_riboflavin,"timeReLasso"),
xaxt="n", main ="Riboflavin data (n=71, p=4088)",
col = color_boxes,
outcol=color_boxes,outpch = 19)
text(1:length(method_names), par("usr")[3] - 5, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Computational time (in s)"), line = 0.4, cex.main = 1.1)
boxplot(take_element2(results_PCR,"timeRSubBoost"),
take_element2(results_PCR,"timeAdaSubBoost"),
take_element2(results_PCR,"timeBoosting"),
take_element2(results_PCR,"timeXGBoostCD"),
take_element2(results_PCR,"timeTwinBoost"),
take_element2(results_PCR,"timeStability"),
take_element2(results_PCR,"timeLasso"),
take_element2(results_PCR,"timeENet"),
take_element2(results_PCR,"timeReLasso"),
xaxt="n", main ="PCR data (n=60, p=22,575)",
col = color_boxes,
outcol=color_boxes,outpch = 19)
text(1:length(method_names), par("usr")[3] - 15, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Computational time (in s)"), line = 0.4, cex.main = 1.1)
####################################################################################
### Model sizes
####################################################################################
SubBoost_sizes_low <- take_element(results_low,1,1,nSim) + (4 - take_element(results_low,1,2,nSim))
RSubBoost_sizes_low <- take_element(results_low,2,1,nSim) + (4 - take_element(results_low,2,2,nSim))
RSubBoost_sizes_high_a <- take_element(results_high_a,1,1,nSim) + (10 - take_element(results_high_a,1,2,nSim))
RSubBoost_sizes_high_b <- take_element(results_high_b,1,1,nSim) + (10 - take_element(results_high_b,1,2,nSim))
RSubBoost_sizes_high_c <- take_element(results_high_c,1,1,nSim) + (100 - take_element(results_high_c,1,2,nSim))
AdaSubBoost_sizes_low <- take_element(results_low,3,1,nSim) + (4 - take_element(results_low,3,2,nSim))
AdaSubBoost_sizes_high_a <- take_element(results_high_a,2,1,nSim) + (10 - take_element(results_high_a,2,2,nSim))
AdaSubBoost_sizes_high_b <- take_element(results_high_b,2,1,nSim) + (10 - take_element(results_high_b,2,2,nSim))
AdaSubBoost_sizes_high_c <- take_element(results_high_c,2,1,nSim) + (100 - take_element(results_high_c,2,2,nSim))
S0_sizes_low <- take_element(results_low,10,1,nSim) + (4 - take_element(results_low,10,2,nSim))
S0_sizes_high_a <-take_element(results_high_a,9,1,nSim) + (10 - take_element(results_high_a,9,2,nSim))
S0_sizes_high_b <-take_element(results_high_b,9,1,nSim) + (10 - take_element(results_high_b,9,2,nSim))
S0_sizes_high_c <-take_element(results_high_c,9,1,nSim) + (100 - take_element(results_high_c,9,2,nSim))
par(mar=c(4.7,3,4,1))
cex.size = 1.2
cex.lab.size = 1.1
cex.axis.size = 1
par(cex.axis = cex.axis.size, cex.lab = cex.lab.size, cex.main = cex.size, las = 1)
par(font.axis = 2)
angle = 45
cex.names = 1.1
method_names <- c("S[0]", "SubBoost", "RSubBoost", "AdaSubBoost")
par(mfrow=c(1,4))
color_boxes = c("gray", pal_col[1:3])
boxplot(S0_sizes_low, SubBoost_sizes_low, RSubBoost_sizes_low, AdaSubBoost_sizes_low,
ylim=c(0, 20),
col = color_boxes, outcol=color_boxes,
xaxt="n",
main = "Low-dimensional setting")
text(1:length(method_names), par("usr")[3] - 0.2, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Model size (number of variables)"), line = 0.4, cex.main = 1.1)
abline(h=4, lty=1, col="red")
method_names <- c("S[0]", "RSubBoost", "AdaSubBoost")
color_boxes = c("gray", pal_col[2:3])
boxplot(S0_sizes_high_a, RSubBoost_sizes_high_a, AdaSubBoost_sizes_high_a,
ylim=c(0, 20),
col = color_boxes, outcol=color_boxes,
xaxt="n",
main = "Sparse high-dimensional setting (a)")
text(1:length(method_names), par("usr")[3] - 0.2, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Model size (number of variables)"), line = 0.4, cex.main = 1.1)
abline(h=10, lty=1, col="red")
boxplot(S0_sizes_high_b, RSubBoost_sizes_high_b, AdaSubBoost_sizes_high_b,
ylim=c(0, 20),
col = color_boxes, outcol=color_boxes,
xaxt="n",
main = "Sparse high-dimensional setting (b)")
text(1:length(method_names), par("usr")[3] - 0.2, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Model size (number of variables)"), line = 0.4, cex.main = 1.1)
abline(h=10, lty=1, col="red")
boxplot(S0_sizes_high_c, RSubBoost_sizes_high_c, AdaSubBoost_sizes_high_c,
ylim=c(0, 200),
col = color_boxes, outcol=color_boxes,
xaxt="n",
main = "Non-sparse high-dimensional setting")
text(1:length(method_names), par("usr")[3] - 2, labels = method_names, srt = angle, adj = 1, xpd = TRUE, cex=cex.names)
title(main = ("Model size (number of variables)"), line = 0.4, cex.main = 1.1)
abline(h=100, lty=1, col="red")
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.