How to repeat the stratified analysis in macro by the group variable?

[Cruise]

Dear SAS experts :

I'm trying to conduct a stratified analysis by age groups with 4 levels and use resulting ALL_MODEL data for a plotting purpose at the end. However, I get an error:

NOTE: The SAS System stopped processing this step because of errors.
WARNING: The data set WORK.MODEL0_AGE1 may be incomplete. When this step was stopped there were
0 observations and 1 variables.
NOTE: DATA statement used (Total process time):
real time 0.01 seconds
cpu time

I think the way I'm trying to loop %DO N=1 %TO 4 is wrong.

Any suggestions to correct the code below? Thanks for your time in advance.

%SYMDEL;
%macro models (PAR_MID,duration,AGEGRP);
%DO N=1 %TO 4;
ODS OUTPUT ParameterEstimates=&PAR_MID; 
PROC PHREG DATA = HAVE;
CLASS STAGE(REF='1')/PARAM=REF; 
MODEL duration*EXIT(0) = STAGE/RL EVENTCODE=1;
WHERE AGEGRP=&N;
%END;
RUN;
%LET MODEL=&PAR_MID;
DATA &PAR_MID; SET &PAR_MID;
LENGTH MODEL $15.; 
MODEL="&PAR_MID";
RUN;
%MEND MODELS;
%MODELS(MODEL0_AGE1,dur_MODEL01);
%MODELS(MODEL0_AGE2,dur_MODEL02);
%MODELS(MODEL0_AGE3,dur_MODEL03);
%MODELS(MODEL0_AGE4,dur_MODEL04);

%MODELS(MODEL5_AGE1,dur_MODEL5_1);
%MODELS(MODEL5_AGE2,dur_MODEL5_2);
%MODELS(MODEL5_AGE3,dur_MODEL5_3);
%MODELS(MODEL5_AGE4,dur_MODEL5_4);

%MODELS(MODEL25_AGE1,dur_MODEL25_1);
%MODELS(MODEL25_AGE2,dur_MODEL25_2);
%MODELS(MODEL25_AGE3,dur_MODEL25_3);
%MODELS(MODEL25_AGE4,dur_MODEL25_4);

%MODELS(MODEL50_AGE1,dur_MODEL50_1);
%MODELS(MODEL50_AGE2,dur_MODEL50_2);
%MODELS(MODEL50_AGE3,dur_MODEL50_3);
%MODELS(MODEL50_AGE4,dur_MODEL50_4);

%MODELS(MODEL75_AGE1,dur_MODEL75_1);
%MODELS(MODEL75_AGE2,dur_MODEL75_2);
%MODELS(MODEL75_AGE3,dur_MODEL75_3);
%MODELS(MODEL75_AGE4,dur_MODEL75_4);

%MODELS(MODEL100_AGE1,dur_MODEL100_1);
%MODELS(MODEL100_AGE2,dur_MODEL100_2);
%MODELS(MODEL100_AGE3,dur_MODEL100_3);
%MODELS(MODEL100_AGE4,dur_MODEL100_4);

DATA ALL_MODEL; SET MODEL:;
KEEP Estimate HRLowerCL HRUpperCL HazardRatio Label MODEL PERCENT;
PERCENT=SUBSTRN(MODEL,6,7);
RUN;

[ballardw · Accepted solution]

Is there some particular reason you don't want to use BY group processing to produce output for each age group?

[Cruise]

that's great idea. I should use BY.