Example: Wage data
library(ISLR)
library(ggplot2)
library(caret)## Loading required package: latticedata(Wage)
Wage<- subset(Wage,select=-c(logwage))
##Becasue logwage is the variable we wanna predict, we delete it
summary(Wage)## year age sex maritl
## Min. :2003 Min. :18.00 1. Male :3000 1. Never Married: 648
## 1st Qu.:2004 1st Qu.:33.75 2. Female: 0 2. Married :2074
## Median :2006 Median :42.00 3. Widowed : 19
## Mean :2006 Mean :42.41 4. Divorced : 204
## 3rd Qu.:2008 3rd Qu.:51.00 5. Separated : 55
## Max. :2009 Max. :80.00
##
## race education region
## 1. White:2480 1. < HS Grad :268 2. Middle Atlantic :3000
## 2. Black: 293 2. HS Grad :971 1. New England : 0
## 3. Asian: 190 3. Some College :650 3. East North Central: 0
## 4. Other: 37 4. College Grad :685 4. West North Central: 0
## 5. Advanced Degree:426 5. South Atlantic : 0
## 6. East South Central: 0
## (Other) : 0
## jobclass health health_ins
## 1. Industrial :1544 1. <=Good : 858 1. Yes:2083
## 2. Information:1456 2. >=Very Good:2142 2. No : 917
##
##
##
##
##
## wage
## Min. : 20.09
## 1st Qu.: 85.38
## Median :104.92
## Mean :111.70
## 3rd Qu.:128.68
## Max. :318.34
## Split the data
inTrain<- createDataPartition(y=Wage$wage,p=0.7,list=FALSE)
training<- Wage[inTrain,]
testing<- Wage[-inTrain,]
dim(training)## [1] 2102 11dim(testing)## [1] 898 11Feature Plot
featurePlot(x=training[,c("age","education","jobclass")],y=training$wage,plot="pairs")
#Plot age versus wage colour by jobclass
qplot(age,wage,colour=jobclass,data=training)
qplot(age,wage,colour=education,data=training)
#Fit a linear model
modFit<- train(wage~age+jobclass+education,method="lm",data=training)
finMod<- modFit$finalModel
print(modFit)## Linear Regression
##
## 2102 samples
## 10 predictor
##
## No pre-processing
## Resampling: Bootstrapped (25 reps)
##
## Summary of sample sizes: 2102, 2102, 2102, 2102, 2102, 2102, ...
##
## Resampling results
##
## RMSE Rsquared RMSE SD Rsquared SD
## 35.83034 0.2598788 1.611693 0.02021685
##
## Diagnostics
plot(finMod,1,pch=19,cex=0.5,col="#00000010")
#The line is mostly close to zero, so the residual is small. But the outfilier on the top, they might be expalin by some other predictorsColor by variables not used in the model
qplot(finMod$fitted,finMod$residuals,colour=race,data=training)
#Plot by index
plot(finMod$residuals,pch=19)
#Predicted versus truth in test set
pred<- predict(modFit,testing)
qplot(wage,pred,colour=year,data=testing) #Idealy, it should be a 45 degree straight line
#Also, another evaluation menthod, RMSD, the result is not good
sqrt(sum((predict(modFit,testing)-testing$logwage)^2))## [1] 0If you want to use all covariates
modFitAll<- train(wage~., data=training,method="lm")## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleading## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleadingpred<- predict(modFitAll, testing)## Warning in predict.lm(modelFit, newdata): prediction from a rank-deficient
## fit may be misleadingqplot(wage,pred,data=testing)