That makes sense as the usual numeric precision issue.   But I don't see how that could explain: a = 3; m = 1/a; n = 1/a; Assigning different values to M and N.  Can it? ... View more

You might try submitting to tech support.  I can't think of an explanation. It's interesting that you both have the same version (9.4M6) showing this odd behavior.  But I still have a hard time believing this is a weird bug.  ... View more

@FoxHope wrote: Hi,   Do you mind if I ask further clarification question? Because it seems that this goes beyond the precision issue. If 1/3 cannot be represented precisely in binary, shouldn't the value of c and d be BOTH off by the same amount and their difference should be zero regardless? Agree, what you are reporting doesn't make sense to me, and isn't the basic numeric precision issue.   Even if somehow these two statements could return different values (because of what, random noise in precision): m = 1/a; n = 1/a; I can't think of an explanation for how these two statements could return different values (accept again random noise in precision...) x = m-n; y = m-n; And yes, your addition of a PUT statement shouldn't change values.   Something odd is happening.  Did you quit SAS and restart a fresh session? ... View more

I can't replicate your results.  I ran on 9.4M7 (Windows) and 9.4M8 (Linux). ... View more

SAS Viya Copilot was recently released.  Perhaps it will satisfy this request? ... View more

I'm confused.  Are you saying you get different results from running the below two bits of code, one with double quotes and one with single quotes?  That would be very surprising.   *double quotes; proc sql noprint; select distinct brand into: brands separated by ', ' from sel_drugs where cont_cat="IUD (intrauterine device)/IUS (intrauterine system) without hormone (copper)" and Act_Ing="LEVONORGESTREL" ; %put &=brands; *single quotes; proc sql noprint; select distinct brand into: brands separated by ', ' from sel_drugs where cont_cat='IUD (intrauterine device)/IUS (intrauterine system) without hormone (copper)' and Act_Ing='LEVONORGESTREL' ; %put &=brands; Note when posting code, please click the "running man" insert SAS code button and paste your code into that window.  It formats the code as code rather than as part of the text. ... View more

Thanks Nancy! (and thanks Inka, I assume…) ... View more

You haven't shown the code that generates the report.  I assume it's inside a %stpbegin  block?   If you're using tagsets.excelxp the default is for titles and footnotes to go in the header/footer of the sheet.  Did you check there to see if they exist?   If you want to see the titles/footnotes in a cell in the spreadsheet, you can use the embedded_titles='yes' option. This is an option of tagsets.excelxp destination.   I think the way to do this would be to add to your excel block to create a global macro variable _odsoptions if upcase("&outtype") = "EXCEL" then do; rc = stpsrv_header('Content-type','application/vnd.ms-excel;'); rc = stpsrv_header('Content-disposition','attachment; filename=temp.xls'); call symput("odsdest","tagsets.excelxp"); call symput("_odsoptions", "embedded_titles='yes' embedded_footers='yes'"); *new line; end; This paper from the great Cynthia Zender explains the use of the global macro vars _odsoptions, _odsdstyle, etc: https://www.lexjansen.com/nesug/nesug07/ap/ap22.pdf   That said, tagsets.excelxp is pretty much deprecated, SAS introduced ODS EXCEL destination which will write .xlsx files 10 (?) years ago.  You might want to switch to ODS EXCEL.  But the general issue will likely be the same, you'll need to tinker with destination-specific options to get what you want. For comparison of tagsets.excelxp and excel destination, see this paper by the great Chevell Parker: https://support.sas.com/resources/papers/proceedings16/SAS5642-2016.pdf     hope that helps       ... View more

Thanks much @ShelleySessoms ! ... View more

Pick a subject value that you know is in both datasets, and check that PROC PRINT can find that value.  So if the subject ID is 333, then try: proc print data=AE.Log_AE; where subject='333'; run; proc print data=AE.Treatment; where subject='333'; run; Does that confirm that the subject exists in both datasets?   If so, please post the log from running: data want; merge AE.Log_AE (where=(subject='333') in=a) AE.Treatment (where=(subject='333') in=b) ; by subject; put (subject a b)(=) ; run; ... View more

Registration link: https://www.basug.org/event-details/basug-june-meeting-training-in-person ... View more

@Tom wrote in part: NOTE the %GLOBAL statement.  You cannot return macro variables from DOSUBL() into local macro variables. DOSUBL() can return side session global macro variables into main session local macro variables.  In this case, you can't use the %GLOBAL statement in the side session to create the macro variable, because it will throw (I think incorrectly): ERROR: Attempt to %GLOBAL a name (VARLIST) which exists in a local environment. But if you add an open-code %let statement inside the DOSUBL call, it will happily create a side session global macro variable, which will be returned to the main session local symbol table.   So code like: %macro list_vars_step(dataset,mvar=varlist); %if not %symexist(&mvar) %then %global &mvar; %let &mvar=; data _null_; if 0 then set &dataset; do while(1=1); length ___NAME___ $32 ; call vnext(___NAME___); if ___NAME___='___NAME___' then stop; if ' '=resolve('%let &mvar=&&&mvar '||___NAME___||';') then; end; stop; run; %mend list_vars_step; %macro list_vars(dataset); %local rc vars dsid i; %if %sysfunc(indexc(&dataset,())) %then %do; %local varlist; %let rc=%sysfunc(dosubl(%nrstr(%let varlist=;%list_vars_step(&dataset) %put INSIDE DOSUBL ; %put _user_ ; %put ; ))); /*added code here*/ %put AFTER DOSUBL ; %put _user_ ; %put ; /*added code here*/ %let vars=&varlist; %end; %else %do; %let dsid = %sysfunc(open(&dataset)); %do i=1 %to %sysfunc(attrn(&dsid, nvar)); %let vars=&vars %sysfunc(varname(&dsid,&i)); %end; %let dsid=%sysfunc(close(&dsid)); %end; &vars. %mend list_vars; %put %list_vars(sashelp.class(drop=sex)); Returns:   33 %mend list_vars; INSIDE DOSUBL GLOBAL VARLIST Name Age Height Weight LIST_VARS DATASET sashelp.class(drop=sex) LIST_VARS DSID LIST_VARS I LIST_VARS RC LIST_VARS VARLIST LIST_VARS VARS 34 35 %put %list_vars(sashelp.class(drop=sex)); AFTER DOSUBL LIST_VARS DATASET sashelp.class(drop=sex) LIST_VARS DSID LIST_VARS I LIST_VARS RC 0 LIST_VARS VARLIST Name Age Height Weight LIST_VARS VARS Name Age Height Weight The log is a bit out of order, but you can see that inside the DOSUBL block there is a global macro variable VARLIST,  the value of which is returned the local macro variable when DOSUBL completes.   ... View more

Today I would also use DOSUBL for this.  But before there was DOSUBL, I hit this problem.  I see it as a bug that that VARNAME function does not honor the KEEP/DROP option.  I've been told in that in SCL, it does honor KEEP/DROP.    Luckly, Søren Lassen posted some solutions to SAS-L.  Basic approach is to open the dataset twice, once with the keep/drop option, and once without.  With that, you can loop over the list of variables without the keep/drop applied, and use the VARNUM function to test if the variable exists in the dataset with the keep/drop applied.  Code like:   %macro ListVars(data); /*** Based on %VarList macro developed by Søren Lassen Posted to SAS-L December 2010 and May 4, 2015 Macro is complicated by fact that the varname function does not honor keep/drop option. Therefore need workaround to prevent er ror msg if user specifies keep/drop. The work around is to iterate over list of variables in dataset without keep/drop applied, and check if each variable is in the dataset with keep/drop applied. ***/ %local dsid /*id of &data, with any keep/drop options*/ dsid2 /*id of &data, without any keep/drop options*/ i var /*name of ith variable in &data*/ return rc ; %*If &data had a rename option, macro would not return the renamed variables; %*So user is not allowed to use rename option.; %if %sysfunc(find(&data,rename,i)) %then %do; %put ER%str()ROR: (%nrstr(%%)&sysmacroname) cannot include rename option on dataset.; %goto mexit; %end; %let dsid=%sysfunc(open(&data)); /*dataset with any keep/drop options*/ %let dsid2=%sysfunc(open(%scan(&data,1,%str(%( )))); /*dataset without keep/drop options*/ %do i=1 %to %sysfunc(attrn(&dsid2,nvars)); %let var=%sysfunc(varname(&dsid2,&i)); %if (%sysfunc(varnum(&dsid,&var))) %then %do; %if %superq(return) = %str() %then %let return=&var; %else %let return=&return &var; %end; %end; %let rc=%sysfunc(close(&dsid)); %let rc=%sysfunc(close(&dsid2)); &return %mexit: %mend ListVars; %put %listvars(sashelp.class) ; %put %listvars(sashelp.class(keep=_numeric_)) ; %put %listvars(sashelp.class(drop=_character_)) ; ... View more

See the approaches in this thread: https://communities.sas.com/t5/SAS-Programming/Print-the-same-SAS-Log-to-BOTH-the-log-window-and-external-file/m-p/966455/highlight/true#M376136   ... View more