> ## sec. 10.5.1 Leaf Sping Experiment
> # 
> data<-read.table("http://www.stat.ucla.edu/~hqxu/stat225/data/springrobust.dat")
> data
  V1 V2 V3 V4   V5   V6   V7   V8   V9  V10
1 -1  1  1 -1 7.78 7.78 7.81 7.50 7.25 7.12
2  1  1  1  1 8.15 8.18 7.88 7.88 7.88 7.44
3 -1 -1  1  1 7.50 7.56 7.50 7.50 7.56 7.50
4  1 -1  1 -1 7.59 7.56 7.75 7.63 7.75 7.56
5 -1  1 -1  1 7.94 8.00 7.88 7.32 7.44 7.44
6  1  1 -1 -1 7.69 8.09 8.06 7.56 7.69 7.62
7 -1 -1 -1 -1 7.56 7.62 7.44 7.18 7.18 7.25
8  1 -1 -1  1 7.56 7.81 7.69 7.81 7.50 7.59

> names(data)[1:4] <- LETTERS[2:5]
> names(data)
 [1] "B"   "C"   "D"   "E"   "V5"  "V6"  "V7"  "V8"  "V9"  "V10"
> 
> # responses for location and dispersion models
> ybar <- apply(data[,5:10], 1, mean)
> lns2 <- log( apply(data[,5:10], 1, var) )
> 
> # make names be available
> attach(data)  
> 
> # we need the halfnormal function in halfnormal.R 
> source("http://www.stat.ucla.edu/~hqxu/stat225/R/halfnormal.R")
> 
> # we need to know the aliasing among factorial effects
> # I=BCDE
> # we can entertain 4 ME's + 3 2fi's
> 
> # Fig 10.3 half-normal plot for location
> g<-lm(ybar~B+C+D+E+B:C+B:D+C:D)  # BC=DE, BD=CE, BE=CD
> summary(g)

Call:
lm(formula = ybar ~ B + C + D + E + B:C + B:D + C:D)

Residuals:
ALL 8 residuals are 0: no residual degrees of freedom!

Coefficients:
             Estimate Std. Error t value Pr(>|t|)
(Intercept)  7.636042                            
B            0.110625                            
C            0.088125                            
D            0.014375                            
E            0.051875                            
B:C          0.008542                            
B:D          0.009792                            
C:D         -0.017708                            

Residual standard error: NaN on 0 degrees of freedom
Multiple R-Squared:     1,	Adjusted R-squared:   NaN 
F-statistic:   NaN on 7 and 0 DF,  p-value:   NaN 

> halfnormal(2*g$coef[-1], label=T, type="n") # remove intercept
> 
> # Fig 10.3 half-normal plot for dispersion
> g<-lm(lns2~B+C+D+E+B:C+B:D+C:D)  # BC=DE, BD=CE, BE=CD
> summary(g)

Call:
lm(formula = lns2 ~ B + C + D + E + B:C + B:D + C:D)

Residuals:
ALL 8 residuals are 0: no residual degrees of freedom!

Coefficients:
            Estimate Std. Error t value Pr(>|t|)
(Intercept) -3.68862                            
B            0.06749                            
C            1.09010                            
D           -0.52218                            
E           -0.32493                            
B:C         -0.26294                            
B:D          0.39974                            
C:D          0.59257                            

Residual standard error: NaN on 0 degrees of freedom
Multiple R-Squared:     1,	Adjusted R-squared:   NaN 
F-statistic:   NaN on 7 and 0 DF,  p-value:   NaN 

> halfnormal(2*g$coef[-1], label=T, type="n") # remove intercept
>