Ch 8. Tree Based Modeling
Irene Zurovec
library(tidyverse)## ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.2 ──
## ✔ ggplot2 3.4.0 ✔ purrr 1.0.1
## ✔ tibble 3.2.1 ✔ dplyr 1.1.1
## ✔ tidyr 1.3.0 ✔ stringr 1.5.0
## ✔ readr 2.1.3 ✔ forcats 0.5.2## Warning: package 'tibble' was built under R version 4.2.3## Warning: package 'dplyr' was built under R version 4.2.3## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()library(openintro)## Warning: package 'openintro' was built under R version 4.2.3## Loading required package: airports## Warning: package 'airports' was built under R version 4.2.3## Loading required package: cherryblossom## Warning: package 'cherryblossom' was built under R version 4.2.3## Loading required package: usdata## Warning: package 'usdata' was built under R version 4.2.3library(ISLR)## Warning: package 'ISLR' was built under R version 4.2.3library(tree)## Warning: package 'tree' was built under R version 4.2.3library(randomForest)## Warning: package 'randomForest' was built under R version 4.2.3## randomForest 4.7-1.1
## Type rfNews() to see new features/changes/bug fixes.
##
## Attaching package: 'randomForest'
##
## The following object is masked from 'package:dplyr':
##
## combine
##
## The following object is masked from 'package:ggplot2':
##
## marginlibrary(ISLR2)## Warning: package 'ISLR2' was built under R version 4.2.3##
## Attaching package: 'ISLR2'
##
## The following objects are masked from 'package:ISLR':
##
## Auto, CreditQuestion 3
Consider the Gini index, classification error, and entropy in a simple classification setting with two classes. Create a single plot that displays each of these quantities as a function of ˆpm1. The x-axis should display ˆpm1, ranging from 0 to 1, and the y-axis should display the value of the Gini index, classification error, and entropy. Hint: In a setting with two classes, pˆm1 = 1 − pˆm2. You could make this plot by hand, but it will be much easier to make in R
p=seq(0,1,0.01)
gini= p*(1-p)*2
class.entropy= -(p*log(p)+(1-p)*log(1-p))
class.error= 1-pmax(p,1-p)
plot(NA,NA,xlim=c(0,1),ylim=c(0,1),xlab='p',ylab='values')
lines(p,gini,col = 'blue')
lines(p,class.error,col='darkgrey')
lines(p,class.entropy,col='lightblue')
legend(x='topright',legend=c('gini','class error','cross entropy'),
col=c('blue','black','green'),lty=1,text.width = 0.22)
Question 8
In the lab, a classification tree was applied to the Carseats data set after converting Sales into a qualitative response variable. Now we will seek to predict Sales using regression trees and related approaches, treating the response as a quantitative variable.
a. Split the data set into a training set and a test set.
set.seed(1)
train <- sample(dim(Carseats)[1], dim(Carseats)[1]/2)
carseats.train <- Carseats[train, ]
carseats.test <- Carseats[-train, ]b. Fit a regression tree to the training set. Plot the tree, and interpret the results. What test MSE do you obtain?
carseats.tree = tree(Sales ~ ., data = carseats.train)
summary(carseats.tree)##
## Regression tree:
## tree(formula = Sales ~ ., data = carseats.train)
## Variables actually used in tree construction:
## [1] "ShelveLoc" "Price" "Age" "Advertising" "CompPrice"
## [6] "US"
## Number of terminal nodes: 18
## Residual mean deviance: 2.167 = 394.3 / 182
## Distribution of residuals:
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -3.88200 -0.88200 -0.08712 0.00000 0.89590 4.09900plot(carseats.tree)
text(carseats.tree)
tree.pred = predict(carseats.tree, carseats.test)
obs.sales = carseats.test$Sales
mean((tree.pred-obs.sales)^2)## [1] 4.922039The MSE appears to be 4.922039.
c. Use cross-validation in order to determine the optimal level of tree complexity. Does pruning the tree improve the test MSE?
carseats.cv <- cv.tree(carseats.tree, FUN = prune.tree)
par(mfrow = c(1, 2))
plot(carseats.cv$size, carseats.cv$dev, type = "b")
plot(carseats.cv$k, carseats.cv$dev, type = "b")
pruned.carseats = prune.tree(carseats.tree, best = 13)
par(mfrow = c(1, 1))
plot(pruned.carseats)
text(pruned.carseats, pretty = 0)
prune.pred <- predict(pruned.carseats, carseats.test)
prune.mse <- mean((carseats.test$Sales - prune.pred)^2)
prune.mse## [1] 4.96547The MSE for the pruned data appears to be 4.96547.
d. Use the bagging approach in order to analyze this data. What test MSE do you obtain? Use the importance() function to determine which variables are most important.
set.seed(1)
bag.car <- randomForest(Sales ~ ., data = carseats.train, mtry = 10, importance = TRUE)
bag.car##
## Call:
## randomForest(formula = Sales ~ ., data = carseats.train, mtry = 10, importance = TRUE)
## Type of random forest: regression
## Number of trees: 500
## No. of variables tried at each split: 10
##
## Mean of squared residuals: 2.889221
## % Var explained: 63.26bag.pred <- predict(bag.car, carseats.test)
bag.err <- mean((carseats.test$Sales - bag.pred)^2)
bag.err## [1] 2.605253importance(bag.car)## %IncMSE IncNodePurity
## CompPrice 24.8888481 170.182937
## Income 4.7121131 91.264880
## Advertising 12.7692401 97.164338
## Population -1.8074075 58.244596
## Price 56.3326252 502.903407
## ShelveLoc 48.8886689 380.032715
## Age 17.7275460 157.846774
## Education 0.5962186 44.598731
## Urban 0.1728373 9.822082
## US 4.2172102 18.073863Based on the importance output we see that Price, ShelveLoc, and CompPrice are the most important predictors. When using the bagging approach, it lowers the MSE to 2.605253, which is better than the other ones previously used.
e. Use random forests to analyze this data. What test MSE do you obtain? Use the importance() function to determine which variables are most important. Describe the effect of m, the number of variables considered at each split, on the error rate obtained.
rf.car <- randomForest(Sales ~ ., data = carseats.train, mtry = 5, importance = TRUE)
rf.pred <- predict(rf.car, carseats.test)
rf.err <- mean((carseats.test$Sales - rf.pred)^2)
rf.err## [1] 2.710806importance(rf.car)## %IncMSE IncNodePurity
## CompPrice 18.3350929 163.77398
## Income 4.7546303 115.94202
## Advertising 10.4995404 98.99790
## Population -0.8865995 78.91872
## Price 45.2942175 451.27503
## ShelveLoc 40.8250417 333.45728
## Age 13.6552105 165.03566
## Education 0.9386481 56.40921
## Urban -0.8796633 11.21732
## US 6.1388918 25.33755We can see again that Price, ShelveLoc, and CompPrice are the most important predictors. The new MSE is 2.711455 which is slightly better than the bagging model.
f. Now analyze the data using BART, and report your results.
library(BART)## Warning: package 'BART' was built under R version 4.2.3## Loading required package: nlme##
## Attaching package: 'nlme'## The following object is masked from 'package:dplyr':
##
## collapse## Loading required package: nnet## Loading required package: survival##
## Attaching package: 'survival'## The following object is masked from 'package:openintro':
##
## transplantx=Carseats[,1:11]
y=Carseats[,"Sales"]
xtrain=x[train,]
ytrain=y[train]
xtest=x[-train,]
ytest=y[-train]
set.seed(5)
bart.fit=gbart(xtrain,ytrain,x.test=xtest)## *****Calling gbart: type=1
## *****Data:
## data:n,p,np: 200, 15, 200
## y1,yn: 2.781850, 1.091850
## x1,x[n*p]: 10.360000, 1.000000
## xp1,xp[np*p]: 11.220000, 1.000000
## *****Number of Trees: 200
## *****Number of Cut Points: 100 ... 1
## *****burn,nd,thin: 100,1000,1
## *****Prior:beta,alpha,tau,nu,lambda,offset: 2,0.95,0.273474,3,4.01382e-30,7.57815
## *****sigma: 0.000000
## *****w (weights): 1.000000 ... 1.000000
## *****Dirichlet:sparse,theta,omega,a,b,rho,augment: 0,0,1,0.5,1,15,0
## *****printevery: 100
##
## MCMC
## done 0 (out of 1100)
## done 100 (out of 1100)
## done 200 (out of 1100)
## done 300 (out of 1100)
## done 400 (out of 1100)
## done 500 (out of 1100)
## done 600 (out of 1100)
## done 700 (out of 1100)
## done 800 (out of 1100)
## done 900 (out of 1100)
## done 1000 (out of 1100)
## time: 3s
## trcnt,tecnt: 1000,1000yhat.bart=bart.fit$yhat.test.mean
mean((ytest-yhat.bart)^2)## [1] 0.1100149It seems that using BART, our test MSE was .1101.
Question 9
This problem involves the OJ data set which is part of the ISLR2 package.
a. Create a training set containing a random sample of 800 observations, and a test set containing the remaining observations.
set.seed(1)
sample.train <- sample(dim(OJ)[1], 800)
oj.train <- OJ[sample.train, ]
oj.test <- OJ[-sample.train, ]b. Fit a tree to the training data, with Purchase as the response and the other variables as predictors. Use the summary() function to produce summary statistics about the tree, and describe the results obtained. What is the training error rate? How many terminal nodes does the tree have?
oj.tree <- tree(Purchase~., data= oj.train)
summary(oj.tree)##
## Classification tree:
## tree(formula = Purchase ~ ., data = oj.train)
## Variables actually used in tree construction:
## [1] "LoyalCH" "PriceDiff" "SpecialCH" "ListPriceDiff"
## [5] "PctDiscMM"
## Number of terminal nodes: 9
## Residual mean deviance: 0.7432 = 587.8 / 791
## Misclassification error rate: 0.1588 = 127 / 800The training error rate is 0.1588 (or 15.88%) for the tree. The tree has 9 terminal nodes.
c. Type in the name of the tree object in order to get a detailed text output. Pick one of the terminal nodes, and interpret the information displayed.
oj.tree ## node), split, n, deviance, yval, (yprob)
## * denotes terminal node
##
## 1) root 800 1073.00 CH ( 0.60625 0.39375 )
## 2) LoyalCH < 0.5036 365 441.60 MM ( 0.29315 0.70685 )
## 4) LoyalCH < 0.280875 177 140.50 MM ( 0.13559 0.86441 )
## 8) LoyalCH < 0.0356415 59 10.14 MM ( 0.01695 0.98305 ) *
## 9) LoyalCH > 0.0356415 118 116.40 MM ( 0.19492 0.80508 ) *
## 5) LoyalCH > 0.280875 188 258.00 MM ( 0.44149 0.55851 )
## 10) PriceDiff < 0.05 79 84.79 MM ( 0.22785 0.77215 )
## 20) SpecialCH < 0.5 64 51.98 MM ( 0.14062 0.85938 ) *
## 21) SpecialCH > 0.5 15 20.19 CH ( 0.60000 0.40000 ) *
## 11) PriceDiff > 0.05 109 147.00 CH ( 0.59633 0.40367 ) *
## 3) LoyalCH > 0.5036 435 337.90 CH ( 0.86897 0.13103 )
## 6) LoyalCH < 0.764572 174 201.00 CH ( 0.73563 0.26437 )
## 12) ListPriceDiff < 0.235 72 99.81 MM ( 0.50000 0.50000 )
## 24) PctDiscMM < 0.196196 55 73.14 CH ( 0.61818 0.38182 ) *
## 25) PctDiscMM > 0.196196 17 12.32 MM ( 0.11765 0.88235 ) *
## 13) ListPriceDiff > 0.235 102 65.43 CH ( 0.90196 0.09804 ) *
## 7) LoyalCH > 0.764572 261 91.20 CH ( 0.95785 0.04215 ) *I’m looking at Node 2, where LoyalCH is a terminal node with a split of 0.5036. There are 365 observations, the final prediction for this is MM. There is a yprobability of ( 0.29315 0.70685 ), meaning about 29% of the observations in the branch have a CH value and the other 71% have a MM value.
d. Create a plot of the tree, and interpret the results.
plot(oj.tree)
text(oj.tree, pretty = 0)
The tree tells us that LoyalCH is the most important variable in the tree. It is the most important predictor of the top 3 nodes.
e. Predict the response on the test data, and produce a confusion matrix comparing the test labels to the predicted test labels. What is the test error rate?
oj.pred <- predict(oj.tree, oj.test, type = "class")
table(oj.test$Purchase, oj.pred)## oj.pred
## CH MM
## CH 160 8
## MM 38 64(38+8)/270 ## [1] 0.1703704The test error rate is .1703704
f. Apply the cv.tree() function to the training set in order to determine the optimal tree size.
cv.oj <- cv.tree(oj.tree, FUN = prune.tree)
cv.oj## $size
## [1] 9 8 7 6 5 4 3 2 1
##
## $dev
## [1] 685.6493 698.8799 702.8083 702.8083 714.1093 725.4734 780.2099
## [8] 790.0301 1074.2062
##
## $k
## [1] -Inf 12.62207 13.94616 14.35384 26.21539 35.74964 43.07317
## [8] 45.67120 293.15784
##
## $method
## [1] "deviance"
##
## attr(,"class")
## [1] "prune" "tree.sequence"g. Produce a plot with tree size on the x-axis and cross-validated classification error rate on the y-axis.
plot(cv.oj$size, cv.oj$dev, type = "b", xlab = "Tree Size", ylab = "Cross-Validation Classification Error")
h. Which tree size corresponds to the lowest cross-validated classification error rate?
According to the plot, a tree size of 6 appears to have the lowest cross-validation classification error rate.
i. Produce a pruned tree corresponding to the optimal tree size obtained using cross-validation. If cross-validation does not lead to selection of a pruned tree, then create a pruned tree with five terminal nodes.
oj.prune <- prune.misclass(oj.tree, best = 5)
plot(oj.prune)
text(oj.prune, pretty = 0)
j. Compare the training error rates between the pruned and unpruned trees. Which is higher?
summary(oj.tree) ##
## Classification tree:
## tree(formula = Purchase ~ ., data = oj.train)
## Variables actually used in tree construction:
## [1] "LoyalCH" "PriceDiff" "SpecialCH" "ListPriceDiff"
## [5] "PctDiscMM"
## Number of terminal nodes: 9
## Residual mean deviance: 0.7432 = 587.8 / 791
## Misclassification error rate: 0.1588 = 127 / 800summary(oj.prune) ##
## Classification tree:
## snip.tree(tree = oj.tree, nodes = c(4L, 10L))
## Variables actually used in tree construction:
## [1] "LoyalCH" "PriceDiff" "ListPriceDiff" "PctDiscMM"
## Number of terminal nodes: 7
## Residual mean deviance: 0.7748 = 614.4 / 793
## Misclassification error rate: 0.1625 = 130 / 800The pruned tree appears to have a higher test error rate at 0.7748 vs 0.7432.
k. Compare the test error rates between the pruned and unpruned trees. Which is higher?
prune.pred <- predict(oj.prune, oj.test, type = "class")
table(prune.pred, oj.test$Purchase)##
## prune.pred CH MM
## CH 160 36
## MM 8 661-(160+66)/270## [1] 0.162963The pruned tree increased the test error as opposed to the unpruned model.