Conditional statement within macro not working correctly

[Jolly1289]

Hello,

I am working on creating a stability analysis program using macros to assist with our redundant stability analysis that we perform.  For part of the program, I need to use a conditional statement to determine with model to run.  However, in my example code I am unable to get the conditional statement to correctly execute based on a p-value that I have save into a dataset.  The section of code that is giving me the issue is:

%macro Stability();
data stats;
set stats;
%if PROB > 0.25 %then %do;
proc glm data=stability;
model RP=time;
run;
%end;
%else %do;
proc glm data=stability;
class batch;
model RP=batch time batch*time;
run;
%end;
%mend Stability;
%Stability();

where "%if PROB > 0.25 %then %do;" is not leading to the use of the correct proc glm procedure being executed.  Basically, in this part of the code I am trying to automate the process of testing for poolability of data across batches or if the batches need to be kept separate. 

I have also attached an entire SAS code file for a full example.  Any help would be greatly appreciated.

Thank you,
Jeff S. O.

[PaigeMiller]

%if PROB > 0.25 %then %do;

This will never work, as to use %IF requires you test a macro variable against some constant. You can't test something named PROB as the macro facility will treat that as text and not a number. However this should work:

%if &PROB > 0.25 %then %do;

here I have used a macro variable &PROB — do you have such a macro variable? It's not obvious.

[Jolly1289]

Hello,

Thank you for your follow up.  PROB is a column in the dataset "stats" containing a numerical value.  Based on this numerical value would depend which proc glm statement I want to run.

Thank you,
Jeff S. O.

[ballardw · Accepted solution]

Do you really use this?

data stats;
set stats;

As shown a  complete waste of CPU cycles as you copy data with no changes into the same data set.

Note: If you are going to work with macro programming I strongly suggest that you always include a RUN; statement to end every procedure.

You state : " conditional statement to correctly execute based on a p-value that I have save into a dataset. ". I see no reference to any variable in a data set that might be used in that manner.

If you intend to use a value from from a data set as a comparison with macro code you first have to make a MACRO variable. That would use the function CALL SYMPUTX. And since your reference a p-value then you need to take control of converting that p_value to the text that a macro comparison will use.

I suspect that you need something like :

data _null_;
   set stats;
   call symputx('prob', put(pvaluevariablename,f4.2) );
run;


%if &prob. > 0.25 %then %do;

The format you use to convert the pvalue to text is up to you. However you want to be aware of probably rounding of the p-value AND how the comparison works.

[Jolly1289]

Thank you very much ballardw.  I completely forgot about the call symputx statements (picking up SAS first the first time in a few years).  I was thinking I could load the data sets within the macro and being able to pull directly from them without needing to do anything else.  Thank you very much!

[Astounding]

DATA steps contain a tool to make this easy.  You don't need CALL SYMPUT at all.  You don't need a macro definition.  To execute one set of code or another after your DATA step ends, here is the way to do it:

data _null_;
set stats;
if PROB > 0.25 then call execute(
'proc glm data=stability;
model RP=time;
run;');
else call execute(
'proc glm data=stability;
class batch;
model RP=batch time batch*time;
run;');
run;

If STATS contains multiple observations, you may need to add to the logic to make sure you execute this logic just once.