IMPORTANT NOTE: 

Due to changes in PROC HAPLOTYPE syntax "caco" option 
is unsupported in %happy for Version 9 (posted here).

If "caco" option is not specified (the default), the output 
data sets addscore, etc. are generated correctly, and the
haplotype scores they contain can be merged (by &ID) back into 
the original data set and used as covariates in a standard 
analysis (e.g. logistic regression). See manual and examples 
for more discussion.

Users of version 9 are recommended not to use the caco option.