MIXED MODEL BY

[desireatem]

I have this data;

A    B  T    order

1    1   1      1

2    4    2       1

5    9    3      1

.....................

1   2     1      2

2     5    2     2

1     2    3     2

-----------------

1   2     1      3

2     5    2     3

1     2    3     3

I want to fit a mixed model such that, I will have a model for each order. That is , three mixed model output.

I tried 

PROC MIXED data=vert METHOD=ML COVTEST by order;
CLASS T;
MODEL A = B /SOLUTION;
RANDOM INTERCEPT T /SUB=SubjectID TYPE=UN G GCORR;
TITLE2 'Random trend model with group effects';
RUN:

But it didnt run. Please help.

[Rick_SAS · Accepted solution]

Put BY statement on a separate statement:

PROC MIXED data=vert METHOD=ML COVTEST;

by order;
CLASS T;

...