Re: [Rd] problem with plot.lm?

2009-08-25 Thread Uwe Ligges
Hi, yes, this looks like a bug. Will investiagte. Best wishes, Uwe Ligges Benilton Carvalho wrote: I'm bringing this issue to r-devel as my idea of sending it to r-help appears to be wrong. The following is reproducible in R-patched and R-devel (also in older versions). An outlier is

[Rd] problem with plot.lm?

2009-08-21 Thread Benilton Carvalho
I'm bringing this issue to r-devel as my idea of sending it to r-help appears to be wrong. The following is reproducible in R-patched and R-devel (also in older versions). An outlier is inserted in group E and the last plot by plot.lm suggests that the point belongs to D. It appears to

[Rd] problem with plot.lm

2007-04-19 Thread Robert Kushler
There is a problem with the "Constant Leverage" plot. Try the following: y <- c(8,9,7,9,6,9,5,9) grp <- factor(rep(1:4,each=2),labels=c('A','B','C','D')) mod <- lm(y ~ grp) windows() par(mfrow=c(2,2)) plot(mod) # Plot 5 ("Constant Leverage") has X axis labels sorted # by the fitted value (as adv