and(t,r)
and
a(t,r)
overlay (add) r times the design matrix for model
term t to the existing design matrix.
Specifically, if the model up to this point has p effects
and t
has a effects,
the a columns of the design matrix for t are multiplied by the scalar r
(default value 1.0) and added to the last a of the p columns already defined.
The overlaid term must agree in size with the term it overlays.
This can be used to force a correlation of 1 between two terms as in a diallel analysis
male and(female)
assuming the i th male is the same individual
as the i th female.