What are the ways of repeating a SAS program many times, by specifying inputs & outputs each time?

[AnOrangeTree]

Hi all, 

I am a new user of SAS who has only worked with R previously. My goal is to generate many sets of large simulations. I have written some codes for running one set of simulations, which are shown below. My questions are:

1. What is the best way to repeat these codes for x number of times automatically, and

2. how to specify a range of inputs and outputs for each time the codes are excuted? (e.g. importing files named 'design_matrix_1.csv', 'design_matrix_2.csv', ...., 'design_matrix_10.csv', and exporting the results as 'result1_1.csv', ..., 'result1_10.csv', respectively)

One way I can think of is to create a macro, and manually specify the input and output files for each scenario.

%let N = 10000;

* import design matrix
PROC IMPORT DATAFILE='/folders/myfolders/design_matrix_1.csv' replace
	OUT=design;
	GETNAMES=YES;
RUN;

* extract information from design matrix
proc sql noprint;
select max(Clus) into : nclus from design;
quit;

* generate N copies of the design matrix
data design_rep;
   do i = 1 to &N;
      do j = 1 to n;
         set design nobs=n point=j;
         output;
         end;
      end;
   stop;
   run;

* generate random numbers to simulate datasets
data sim;
set design_rep ;
    by i Clus;
	retain clus_int;
	if first.Clus then do;
	clus_int = rand('normal', 0, sigma_clus_lp);
	end;
error = rand("Normal", 0, &sigma);
y = Tx* &Tx_effect + period + clus_int + error;
run;

ods exclude all;
ods noresults;

* analyze data
PROC glimmix data=sim;
by i;
CLASS Clus ;
MODEL y = Tx Period /S  DDFM=KR;
RANDOM intercept / subject = Clus ;
NLOPTIONS TECHNIQUE=DBLDOG;
ods output Tests3 = result1 ParameterEstimates = result2;
RUN;

ods exclude none;
ods results;

PROC EXPORT
DATA=result1
replace
OUTFILE="/folders/myfolders/result_1.csv"
DBMS=CSV;

PROC EXPORT
DATA=result2
replace
OUTFILE="/folders/myfolders/result_2.csv"
DBMS=CSV;

data reject (keep = Reject); 
   set result1; 
   where Effect = 'Tx';
   Reject = (ProbF < 0.05); 
run; 

proc summary data = reject;
var Reject;
output out = power(drop = _FREQ_ _TYPE_ ) mean = ;
run;

PROC PRINT DATA=power;
run;

Many thanks!

[Reeza]

Macros, Call Execute or DOBSUBL are the usual options. 

UCLA introductory tutorial on macro variables and macros

https://stats.idre.ucla.edu/sas/seminars/sas-macros-introduction/

Tutorial on converting a working program to a macro

This method is pretty robust and helps prevent errors and makes it much easier to debug your code. Obviously biased, because I wrote it 🙂 https://github.com/statgeek/SAS-Tutorials/blob/master/Turning%20a%20program%20into%20a%20macro.md

Macro Appendix with examples:

https://communities.sas.com/t5/SAS-Communities-Library/SAS-9-4-Macro-Language-Reference-Has-a-New-Ap...

---

@AnOrangeTree wrote:

Hi all, 

I am a new user of SAS who has only worked with R previously. My goal is to generate many sets of large simulations. I have written some codes for running one set of simulations, which are shown below. My questions are:

1. What is the best way to repeat these codes for x number of times automatically, and

2. how to specify a range of inputs and outputs for each time the codes are excuted? (e.g. importing files named 'design_matrix_1.csv', 'design_matrix_2.csv', ...., 'design_matrix_10.csv', and exporting the results as 'result1_1.csv', ..., 'result1_10.csv', respectively)

One way I can think of is to create a macro, and manually specify the input and output files for each scenario.

%let N = 10000;

* import design matrix
PROC IMPORT DATAFILE='/folders/myfolders/design_matrix_1.csv' replace
	OUT=design;
	GETNAMES=YES;
RUN;

* extract information from design matrix
proc sql noprint;
select max(Clus) into : nclus from design;
quit;

* generate N copies of the design matrix
data design_rep;
   do i = 1 to &N;
      do j = 1 to n;
         set design nobs=n point=j;
         output;
         end;
      end;
   stop;
   run;

* generate random numbers to simulate datasets
data sim;
set design_rep ;
    by i Clus;
	retain clus_int;
	if first.Clus then do;
	clus_int = rand('normal', 0, sigma_clus_lp);
	end;
error = rand("Normal", 0, &sigma);
y = Tx* &Tx_effect + period + clus_int + error;
run;

ods exclude all;
ods noresults;

* analyze data
PROC glimmix data=sim;
by i;
CLASS Clus ;
MODEL y = Tx Period /S  DDFM=KR;
RANDOM intercept / subject = Clus ;
NLOPTIONS TECHNIQUE=DBLDOG;
ods output Tests3 = result1 ParameterEstimates = result2;
RUN;

ods exclude none;
ods results;

PROC EXPORT
DATA=result1
replace
OUTFILE="/folders/myfolders/result_1.csv"
DBMS=CSV;

PROC EXPORT
DATA=result2
replace
OUTFILE="/folders/myfolders/result_2.csv"
DBMS=CSV;

data reject (keep = Reject); 
   set result1; 
   where Effect = 'Tx';
   Reject = (ProbF < 0.05); 
run; 

proc summary data = reject;
var Reject;
output out = power(drop = _FREQ_ _TYPE_ ) mean = ;
run;

PROC PRINT DATA=power;
run;

Many thanks!

---

[AnOrangeTree]

Thanks for sharing these resources! They're really helpful.