Before starting on using the macro processor to generate some SAS code you need to be clear about what SAS you need to generate.   Can you explain a little more about what code you are trying to create?   For example you mentioned that you were able to get this code fragment to work. data SAMP ; set VAMP; where sampdate ge &Qstart. ; run; That will work when value of the macro variable QSTART is text that will make the syntax of the WHERE statement valid.  So if SAMPDATE is numeric variable with DATE values then any text that looks to the SAS compiler like a valid date value should work.  So you could have QSTART set to things like: %let qstart="01JUL2025"d; %let qstart='01-jul-2025'd; %let qstart=mdy(7,1,2025); %let qstart=intnx('qtr',date(),0,'b'); %let qstart=23923; The first two are date literals.  The next two are expression that SAS will evaluate to a date value. The last one is the actual number that SAS uses to represent the first of July in the year 2025.   Note that to the macro processor all of them are just text strings. ... View more

If the value of &mynote is a lon string of letters and numbers without quotation marks around them, then you don't want quotation marks around the value on your %IF statement.  The macro language does not use quotation marks to indicate text values, because every value is a text value in the macro language.  Similarly, you don't need quotation marks on your %PUT statements.  Try:   %macro server; %if &mynode = longstringofletters_numbers %then %put TEST Server; %else %put PROD Server; %mend; ... View more

@Jumboshrimps wrote: Running this code generates a data set of all variables and an additional column "i", with the value "1" for each ob.  This is where I got stuck in a previous attempt, I don't have the code from that job, so I can't remember how I resolved it.                                                                                                                                                                                               Also, where in the code is it including only those with a value of "1", and/or excluding nulls? Should I add an IF f[i] IS NOT NULL THEN                                if f[i] then fields=catx(',',fields,vname(f[i]));   Thanx.   If I has 1 after that DO loop then there weren't any variables in the array!  That because a normal DO loop like :  DO I=1 to 5; .... END; will stop when I becomes larger then the upper bound.  So if I ends you as 1 then there were zero variables in the array.  You should see a message in the log to that effect. WARNING: Defining an array with zero elements. To your question the answer is NO.    First because the syntax is wrong. IF f[i] IS NOT NULL THEN ... Will not work in SAS code.  If you do need to test if something is MISSING (what SQL means by the keyword NULL) in SAS code then use the MISSING() function.   Second because the test is already covered by the existing test when the variables are NUMERIC. if f[i] then fields=catx(',',fields,vname(f[i])); SAS will treat a number that is missing or zero as FALSE and any other value as TRUE.   Now if the variables are CHARACTER then you will want to change the test in the IF statement to reflect that fact.  So perhaps something like if left(f[i]) = '1' then fields=catx(',',fields,vname(f[i])); Because otherwise SAS will first try to convert the character value into a number and then test if the number means TRUE or FALSE.  You will get a message in the LOG when SAS does that type of automatic type conversion. 1759 data test; 1760 array x [2] $2 ; 1761 do i=1 to dim(x); 1762 if x[i] then put x[i]= 'TRUE'; 1763 else put x[i]= 'FALSE'; 1764 end; 1765 run; NOTE: Character values have been converted to numeric values at the places given by: (Line):(Column). 1762:8 x1= FALSE x2= FALSE NOTE: The data set WORK.TEST has 1 observations and 3 variables.     ... View more

THX.   I see your point. There is a format library with a $YES/NO format.  I'll try that for now, and see if any complains. ... View more

This is a classic example of where a long data table (diagnoses in rows) is way better than your current wide table (diagnoses in columns) and will make your processing a lot easier. If you want to try this than please provide us with some sample data using the DATA step DATALINES statement. ... View more

Much better!!!!   I had to use real dates for SAS to convert them? Who knew?   Thanx.   ... View more

%PUT is a macro statement. The macro processor is a preprocessor which does its work while code is compiled, long before the code is actually executed. Macro code can therefore also not use data step variables, as these come into exustence long after the macro processor did its work. What you see from your %PUT is a remnant from an earlier try where you used %LET. Your code does not set the macro variable &dstring anywhere, so on its own it would result in a log message telling you that an apparent symbolic reference could not be resolved.   To store values from a data step into a macro variable, you need the CALL SYMPUTX routine. To display the newly created macro variable, the %PUT must be entered after the RUN which terminates the data step. ... View more

The code you posted would not set TRACKER to "memname" unless that was the value of the MEMNAME variable. But the %PUT statement you have would not show you whether or not the value was set because you ran it BEFORE the value was set.  Macro statements like %PUT execute while the code is being analyzed and so before the data step has a chance to run. Do not use the the old call symput() function. Use the modern CALL SYMPUTX() function and avoid storing the trailing spaces from the MEMNAME variable into the macro variable TRACKER. Also there's no need to make a dataset named NULL if the goal is just to create a macro variable.  Use the keyword _NULL_ instead so no dataset is made. %let tracker=BEFORE DATA STEP RUNS; data _null_; set work.tracker_tbl; call symputx('tracker', memname); run; %put &=tracker;   ... View more

data _null_; Format last_mnth yymmn6.; format lst_mnth $6.; format char_mnth $6.; last_month = intnx('month',today(),-1,'E'); call symput("last_mnth", put (last_month, YYMMN6.)); %put &=last_mnth.; run; The %put-statement won't work as expected, because it is resolved before the data step is executed. From the second program execution you will see the value of the prior execution. Moving the %put statement after run; solves the problem. ... View more

@Jumboshrimps wrote:   At this point, I cannot say with 100% certainty where this is going.  It is a bunch of Enterprise Guide 7.13 legacy code made by someone who know how to drag-and-drop components only.  My task is to try to refactor the 10,000 lines of incomprehensible, undocumented, uncommented code into SAS Studio code that can be loaded back into the Warehouse on an automated basis.   The  proc transpose code module show above, (that one that works) replaced 80 lines of code.   You must have some idea about how this output is used. You inherited something that produces a report, or an analysis, or an output. You did not inherit something that does complicated transposing and then stops there and does nothing with the transposed data.   If you don't know how this is being used, then I suggest that's the first step you need to tackle.   And in any event, I will stick with my opinion that putting dates into variable names is a poor practice, and much easier methods of dealing with this data are available if you do not transpose. ... View more

@Jumboshrimps wrote: Thanx for all your help. There are TWO HUNDRED lines of case statements as below (1 to 200), assigning a numeric value, (in the example below,  8), to any records where the prescription string contains  8 U or 8U or 8 u or 8u: Which with UPCASE of the values would be reduced to   8 U or 8U   Regular expressions, which in SAS use functions whose names start with PRX or call routines with similar names are used to search strings for patterns. Patterns can be any digit followed by one or more specific characters. PRXMATCH finds the position in a string for example. Which position could be fed to one of the other functions to extract the "word" before that position. But it will help to have some examples to test what specific code works best with your data. I am not a guru with the regular expression syntax but this is what would likely best replace 200 lines of (garbage-like) code. And likely run in a data step instead of SQL and may be quicker. ... View more