Proc Candisc and RBD or other experimental designs

[rodjunper]

Hi.

I'm trying to do a Canonical Discriminant Analysis using data from a Randomized Block Design (RBD).
I would like to know how can I take in account, in the PROC CANDISC, that data were generated in a RBD?

Thank you very much.

--
Rodrigo

[SteveDenham]

I don't believe it can be done in SAS easily.  The CLASS statement in CANDISC takes only a single variable as an argument, and I don't see any way to incorporate the design.  The closest I can think of is to look at PROC PLS, and try a model that has the RBD factors on the right hand side of the MODEL statement, and the responses for each of the categories on the left hand side, and try for a prinipal components regression.

Steve Denham