Discrete Choice Experiment - Canonical Correlations (HELP)

Hi,

I am planning a DCE and I have run into a problem of "canonical correlations" with the SAS macros.
Has anyone ever found this problem of canonical correlations with DCE?
I am wondering if this is just a consequence of our restrictions' macro or if we should be cautious with these canonical correlations...

Thanks a lot!!

The code we used is as follows:

%mktruns(3 3 3 2 3 2 2 2 2);

%macro resmac;
bad=0;

navail=(x7=1)+(x4=2);
if (navail=2) then bad=1;

navail=(x7=1)+(x5=2);
if (navail=2) then bad=1;

navail=(x7=1)+(x5=3);
if (navail=2) then bad=1;

navail=(x7=1)+(x9=2);
if (navail=2) then bad=1;

navail=(x9=2)+(x4=1);
if (navail=2) then bad=1;

navail=(x9=2)+(x5=1);
if (navail=2) then bad=1;

navail=(x9=2)+(x7=1);
if (navail=2) then bad=1;

%mend;

%mktex(3 3 3 2 3 2 2 2 2,
n=2592,
restrictions=resmac);

%mktlab(data=design, int=f1-f2 ,vars=x1-x9);

%choiceff(data=final,
model=class(x1-x9 / sta),
nsets=36,
maxiter=100,
flags=f1-f2,
options=relative,
beta=zero);

%mktblock(data=best,
nalts=2,
nblocks=4,
factors=x1-x9,
seed=472);

%mkteval (data=best)