自己先顶一个，不要沉

Za是random effect的design matrix

那如果是换一个公式，这个design matrix该如何设置，换个说法就是这个design matrix是怎么得来的呢

楼主最好详细的说说这段程序是在干什么，这段code也不能运行，很难进一步帮你。

这是完整的程序编程：
%let n=2; /*n=number of tress used for calibration*/
Data example;
Input d h ddom hdom weight;
/*weight =242.87/(area of the subplot), in the case of unequal selection probabilities; weight=1, in other case*/
Hhat=1.3+(hdom-1.3)*(exp(-7.914*d**(-1.444+0.02259*(hdom-1.3)))/exp(7.914*ddom**(-1.444+0.02259*(hdom-1.3))));
Cards;
9 11 18 18 3.092
9.5 12 18 18 3.092
10 . 18 18 3.092
13 . 18 18 0.7731
15 . 18 18 0.7731
18 . 18 18 0.7731
;
Run;
Data calibration;
Set example;
Res=h-hhat;
B0=-7.914*d**(-1.444+0.02259*(hdom-1.3));/*auxiliary variable*/
B1=-7.914*ddom**(-1.444+0.02259*(hdom-1.3));/*auxiliary variable*/
Za1=(hdom-1.3)*exp(b0)*(b0*log(d)-b1*log(ddom))/exp(b1);
Za2=(hdom-1.3)**2*exp(b0)*(b0*log(d)-b1*log(ddom))/exp(b1);
W=(1/weight)**0.5;
Where h is not null;
Run;
Proc iml;
Use calibration;
Read all var {za1 za2} into z;
Read all var {res} into res;
D={0.1512 -0.004188, -0.004188 0.0001244};
Read all var {w} into w;
G=w#i(&n);
R=2.561*g*i(&n)*g;
B=d*z`*inv(z*d*z`+r)*res;
Print b;
就是求混合模型中两个随机效应参数u和v的值，