## thymus weight data 
y<-c(25,26,31,24,27,20,27,19,25,32,35,27,23,
     31,34,41,32,39,36,43,41,30,39,38,44,43,32,
     44,35,26,39,32,25,21,31,27,41,
     44,41,38,39,34,45,55,48)
     
grp<-rep(c(1,2,3,4),c(13,14,10,8))

## table 4.2
cbind(c(1,2,3,4), c(13,14,10,8),
 round(c(mean(y[grp==1]),mean(y[grp==2]),mean(y[grp==3]),mean(y[grp==4])),1),
 round(sqrt(c(var(y[grp==1]),var(y[grp==2]),var(y[grp==3]),var(y[grp==4]))),2))

cas<-rep(c(0,1,0,1),c(13,14,10,8))
adr<-rep(c(0,1),c(27,18))

cbind(y,cas,adr)  # data

fit<-lm(y~cas+adr+cas:adr)  ## same as lm(y~cas*adr)
> summary(fit)

Coefficients:
            Estimate Std. Error t value Pr(>|t|)    
(Intercept)  26.2308     1.6000  16.395  < 2e-16 ***
cas          11.1264     2.2219   5.008 1.09e-05 ***
adr           5.8692     2.4265   2.419   0.0201 *  
cas:adr      -0.2264     3.5249  -0.064   0.9491    

Residual standard error: 5.769 on 41 degrees of freedom
Multiple R-Squared: 0.5438,     Adjusted R-squared: 0.5104 
F-statistic: 16.29 on 3 and 41 DF,  p-value: 4.022e-07 

> anova(fit)
Analysis of Variance Table

Response: y
          Df  Sum Sq Mean Sq F value    Pr(>F)    
cas        1 1269.75 1269.75 38.1551 2.423e-07 ***
adr        1  356.67  356.67 10.7178   0.00216 ** 
cas:adr    1    0.14    0.14  0.0041   0.94911       
Residuals 41 1364.42   33.28                         
## slightly diff than table 4.3, type III Sum Sq?

> fit1<-lm(y~factor(grp))
> anova(fit1)
Analysis of Variance Table

Response: y
            Df  Sum Sq Mean Sq F value    Pr(>F)    
factor(grp)  3 1626.56  542.19  16.292 4.022e-07 ***
Residuals   41 1364.42   33.28                      

> 1269.75+356.67+0.14
[1] 1626.56