pureErrorAnova {alr3} | R Documentation |
For a linear model object, finds the sum of squares for lack of fit and the sum of squares for pure error. These are added to the standard anova table to give a test for lack of fit. If there is no pure error, then the regular anova table is returned.
### This is a generic function.
pureErrorAnova(mod)
## S3 method for class 'lm'
pureErrorAnova(mod)
### Methods for other than models for type lm have not been defined.
mod |
an object of type |
For regression models with one predictor, say y ~ x
, this method fits
y ~ x + factor(x)
and prints the anova table. With more than one predictor, it
computes a random linear combination L
of the terms in the mean function
and then gives the anova table for update(mod, ~.+factor(L))
.
Returns an analsis of variance table.
Sanford Weisberg, sandy@stat.umn.edu
Weisberg, S. (2005). Applied Linear Regression, third edition, New York: Wiley, Chapter 5.
x <- c(1,1,1,2,3,3,4,4,4,4)
y <- c(2.55,2.75,2.57,2.40,4.19,4.70,3.81,4.87,2.93,4.52)
m1 <- lm(y~x)
anova(m1) # ignore pure error
pureErrorAnova(m1) # include pure error
head(forbes)
m2 <- lm(Lpres~Temp, data=forbes)
pureErrorAnova(m2) # function does nothing because there is no pure error