Skip to content

anova.mlm for single model (one-way repeated measured anova)

1 message · Brian Ripley

#
On Sat, 12 Aug 2006, takahashi kohske wrote:

            
Yes, it is a bug.  The line

            comp <- object$effects[p1, ]

should be

            comp <- object$effects[p1, , drop=FALSE]

I am changing this in 2.3.1 patched and R-devel.