[Rd] Bug #7829 reproducible example (rejected by bug tracker)

kwright at eskimo.com kwright at eskimo.com
Tue May 3 15:17:49 CEST 2005


I submitted a bug report (#7829) that is in the bug tracking system, but
bounced off the email list for some reason.  Since then, I've come up with
a reproducible example:

Oats <- Oats[-1,]
Oats$nitro <- as.factor(Oats$nitro)
Oats.aov <- aov(yield~Block+Variety+nitro+Variety:nitro,data=Oats)
model.tables(Oats.aov,"means",cterms=c("nitro","Variety:nitro"))

Gives the following results.  The bug is in the number of replications for
the interaction term.  See the bug tracking system for more details.


Tables of means
Grand mean

103.8732

 nitro
        0  0.2   0.4 0.6
    78.74 98.5 113.8 123
rep 17.00 18.0  18.0  18

 Variety:nitro
             nitro
Variety       0      0.2    0.4    0.6
  Golden Rain  79.53  98.03 114.20 124.37
  rep          79.53  98.03 114.20 124.37
  Marvellous   86.20 108.03 116.70 126.37
  rep          86.20 108.03 116.70 126.37
  Victory      69.77  89.20 110.37 118.03
  rep          69.77  89.20 110.37 118.03



More information about the R-devel mailing list