Do Loop to macro variables from character to numeric

[changxuosu]

Hi SAS community,

I have a question that needs your wisdom.

I want to execute a macro and feed the macro with the following three variables: start, end, and incre.

All these three variables take two decimal points.

As shown below

%macro loop(start=,end=,incre=);

%local weight;

%do weight = &start %to &end %by &incre;

XXXXXXXXXXXXXX;

XXXXXXXXXXXXXX (macro code);

XXXXXXXXXXXXXX;

%end;

%mend;

%macro(start= 0.5 ,end=0.99 ,incre=0.1 );

But, there is an error msg when I execute the code above, the error msg says the following:

A character operand was found in the %EVAL function or %IF condition where a numeric operand is required. The condition was:

.5

I think it's because the values I feed into start, end, and incre variables are taken as character, rather than numeric

What are the solutions to this problem?

Thank you all

[Astounding]

Macro %DO loops do not permit any characters, even decimal points.  You would have to switch to:

%loop (start=50, end=99, incr=10)

Then within the logic of the macro, divide these values by 100 as needed, to restore the proper logic.

[ballardw]

You might more completely describe what this code is supposed to do. Most of the time I when I see multiple variables with decimal values what is being done more appropriately belongs in a data step.

[Kurt_Bremser]

Create your macro for a single instance, and use a data step loop with call execute().

[changxuosu]

thanks a lot! I studied execute() it's a very useful tool. First time to use it. Thanks for introducing me to this useful tool!

[Tom · Accepted solution]

I think it's because the values I feed into start, end, and incre variables are taken as character, rather than numeric

No.  Macro variables are ALWAYS character. 

The reason it doesn't work is because the iterative %DO loop only works for INTEGERS.  The implicit %EVAL() function call used in %DO var= ... (and in %IF and %UNTIL and %WHILE)  can only do arithmetic on strings that look like integers.

If you want to make your own looping construct for floating point values then build it using %SYSEVALF().

%let weight=&start;
%do %while (%sysevalf(&weight <= &end));
   ....
   %let weight=%sysevalf(&weight + &incre);
%end;

[changxuosu]

thanks a lot Tom. It works perfect!