From: Jo Sugar on
Hi!

Question: I have a split plot design with "water" as main plot factor
and "clones" as subplot factor. I will also like to fit an AR(1) for
row and an ar(1)for plot. Is this code correct?

proc mixed data = user.wue (where=(cropid=1));
class block water clone;
model tch = water clone water*clone/ddfm=kr;
random block block*water;
repeated/type=sp(powa) (row plot) subject=intercept;
run;

Question:

I also have multiple "cropid" and these crops are correlated. I would
like to fit and AR(1) to the "cropid" too plus an AR(1) for rows and
plots so I have a 3 way R structure. What is the correct code for this
as I don't think this is correct:

proc mixed data = user.wue;
class cropid block water clone;
model tch = water clone water*clone/ddfm=kr;
random block block*water;
repeated/type=sp(powa) (cropid row plot) subject=intercept;
run;

Thanks

Jo