nlmixed and classes

#1
Hi,
I was wondering if anybody can help me out here.

I am working with some growth data from different lines and ***es and I'm not sure if it is possible but I want to get these class variables in my proc nlmixed to estimate the differences between all the classes.
I know that you need to use glmmod to get dummy variables but after that I don't know how to get it in my model.

ods output DesignPoints=dsn;
ods exclude Designpoints;proc glmmod data=f;
class ***e line;
model age=weight ***e*line;
run;


The model I made just to check differences between gender is below (female=-1 and male=1) with a Gompertz curve. How to tranform or use all the classes at the same time, or is it just not possible??

*one random variable, a0 b0 and b1 random;

proc nlmixed data=f;* gconv=0.000000000000001;
parms
a0=3.5
a0gen=0.12
b0=-3
b0gen=-0.02
b1=-0.01
b1gen=0.0001
s2u1=0.17
s2e=0.73;
a0ind=a0+a0gen*gender + u1;
*individual a0;
b0ind=b0+b0gen*gender;
b1ind=b1+b1gen*gender;

*additional parameters;
a0males=a0+a0gen;
a0females=a0-a0gen;
a0gender=a0+a0gen*gender;

*predictions;

Gind=a0ind*exp(exp(b0ind)*(exp(b1ind*age)-1)/b1ind);
Vind=Gind*exp(b0ind)*exp(b1ind*age);
Ggender=a0gender*exp(exp(b0)*(exp(b1*age)-1)/b1);

model weight~normal(a0ind*exp(exp(b0ind)*(exp(b1ind*age)-1)/b1ind),s2e);
random u1 ~normal(0,s2u1) subject=id;

predict Gind out=predGind;
id Vind Ggender;

estimate 'a0' a0;
estimate 'b0' b0;
estimate 'b1' b1;
estimate 'a0males' a0males;
estimate 'a0females' a0females;
estimate 'a0males -a0females' a0males-a0females;

ods output AdditionalEstimates=addest;

title 'Model with 1 random effect u1 and gender effects b0gen b1gen';

run;

Thanks for any help!!
Naomi Duijvesteijn