Skip to content
Prev 43182 / 398513 Next

Re-Post: Combining Factors in model.matrix

On Sat, 24 Jan 2004 02:31:19 -0500, you wrote:

            
I didn't see the earlier posting, but nothing seems to be wrong with
this one.
You seem to want something like
t12 t22
1   0   0
2   0   1
3   1   0
4   1   1

(i.e. leave the intercept in the model, but delete it from the
result).  This doesn't give the exact encoding you asked for; the
"contrasts" options might be able to fix this (see ?contr.poly, or
maybe ?contrasts, and experiment a bit.)
I don't know.

Duncan Murdoch