This modification uses the dummy variable with character RESULT. Notice the changes in define statement for result and addition of define dummy. [pre] proc report data=pt_data nowd nocompletecols list; title 'Using ACROSS Variables'; by patient; column patient test result,day,date dummy; define patient / group; define test /group ; define day / across order=data ' '; define date / across order=data f=date9. ' '; define result / display 'Results'; define dummy / noprint; run; [/pre] I just looked at the threads you referenced and noticed the dummy thread creates a variable in the input data which is not necssary. PROC REPORT with treat dummy as calculated so all you need is noprint on the define. P.S. Message was edited by: data _null_; ... View more

It is because the decimal point is shifted by BEST format. I suggest using order=internal. [pre] define useord / order descending noprint order=internal; [/pre] ... View more

Do you also see [pre] Stderr output: File Not Found [/pre] in the log. Your program looks OK I suspect an error in the path. ... View more

This version address the variable order issue. The standardized versions of the variables are woven into the original variables immediately following the unstandardized version of the variables. This program could be made completly dynamic with a few more parameters say in a macro to specify input and output datasets, std variable list, and merge key. I suppose the merge key could be omitted if you wanted use One-to-One Merging. [pre] data score; length Student $ 9; input Student $ StudentNumber Section $ Test1 Test2 Final @@; format studentnumber z4.; datalines; Capalleti 0545 1 94 91 87 Dubose 1252 2 51 65 91 Engles 1167 1 95 97 97 Grant 1230 2 63 75 80 Krupski 2527 2 80 69 71 Lundsford 4860 1 92 40 86 McBane 0674 1 75 78 72 Mullen 6445 2 89 82 93 Nguyen 0886 1 79 76 80 Patel 9164 2 71 77 83 Si 4915 1 75 71 73 Tanaka 8534 2 87 73 76 ;;;; run; proc sort; by studentNumber; run; *** Get variable to score into a SAS data set; proc transpose data=score(obs=0) out=stdvars(index=(_name_)); var test1-test2; run; *** Get _ALL_ variable names into a SAS data set; proc transpose data=score(obs=0) out=allvars; var _all_; run; *** Weave all vars with STDvar names, to use in retain statement; data weaveSTDvars; length _name_ $32; set allvars(in=in1); output; set stdvars key=_name_/unique; if _iorc_ eq 0 then do; _name_ = cats('Std',_name_); output; end; else _error_=0; run; proc print; run; *** Gen some code RENAME option and varlist; %let RenameOpt=; %let varlist=; %let NewVariableOrder=; proc sql noprint; select catx('=',_name_,cats('Std',_name_)), _name_ into :RenameOpt separated by ' ', :VarList separated by ' ' from stdvars ; select _name_ into :newVariableOrder separated by ' ' from weaveSTDvars; quit; run; %put NOTE: RenameOpt=&RenameOpt; %put NOTE: Varlist=&varList; %put NOTE: NewVariableOrder=&NewVariableOrder; proc standard data = score ( keep = studentnumber &VarList rename = (&renameOpt) ) mean=75 std=5 out=stndtest; var Std:; run; data work.Combined; retain &NewVariableOrder; merge score stndtest; by studentNumber; run; proc print data=combined noobs round; title 'Standardized Test Scores for a College Course'; run; [/pre] ... View more

> Since your CALL SYMPUT is executed within a macro, > the macro variables created will be local. That is not exactly right. There are a few "depends on" that affect the action. Consider the following example. You can get the exact action you desire with SYMPUTX when you specify the symbol table option. CALL SYMPUTX(macro-variable, value <,symbol-table>); Having said that I don't think GLOBAL macro variables are the best choice here. I would think about keeping the data in a SAS data set would be better. I would need to see how these variables are being used to make a recomendation. [pre] 161 options nostimer; 162 %symdel a; 163 164 %macro test(arg); 165 data _null_; 166 call symput('A','What is my scope'); 167 run; 168 %put _global_; 169 %put _local_; 170 %mend; 171 172 %test 173 174 %symdel a; TEST ARG TEST A What is my scope WARNING: Attempt to delete macro variable A failed. Variable not found. 175 %macro test; 176 data _null_; 177 call symput('A','What is my scope'); 178 run; 179 %put _global_; 180 %put _local_; 181 %mend; 182 183 %test GLOBAL A What is my scope 184 185 186 %symdel a; 187 %global a; 188 %macro test(arg); 189 data _null_; 190 call symput('A','What is my scope'); 191 run; 192 %put _global_; 193 %put _local_; 194 %mend; 195 196 %test(); GLOBAL A What is my scope TEST ARG [/pre] ... View more

> I don't think that DOPEN or any other directory SAS > function will do the job. You would need to combine DOPEN with DNUM DREAD and perhaps DINFO but it can be done with relative ease. I prefer the INFILE PIPE method you eluded to but find that on unix the FIND command is much better than LS as FIND provides many more options with regard to file info and format of returned output. Also also on UNIX always point directly to the command as /usr/bin/ls or where ever because an alias can cause unpleasant results. ... View more

this address the following issues 1) variable list of unknown length 2) create new variable with prefix STD 3) combine with original(requires merge key) It does not create the variable order that was achieved with proc sql but that could be address if desired. [Pre] data score; length Student $ 9; input Student $ StudentNumber Section $ Test1 Test2 Final @@; format studentnumber z4.; datalines; Capalleti 0545 1 94 91 87 Dubose 1252 2 51 65 91 Engles 1167 1 95 97 97 Grant 1230 2 63 75 80 Krupski 2527 2 80 69 71 Lundsford 4860 1 92 40 86 McBane 0674 1 75 78 72 Mullen 6445 2 89 82 93 Nguyen 0886 1 79 76 80 Patel 9164 2 71 77 83 Si 4915 1 75 71 73 Tanaka 8534 2 87 73 76 ;;;; run; proc sort; by studentNumber; run; *** Get variables list into SAS data set; proc transpose data=score(obs=0) out=varlist; var test1-test2; run; *** Gen some code RENAME option and varlist; %let RenameOpt=; proc sql; select catx('=',_name_,cats('Std',_name_)), _name_ into :RenameOpt separated by ' ', :VarList separated by ' ' from varlist ; quit; run; %put NOTE: RenameOpt=&RenameOpt; %put NOTE: Varlist=&varList; proc standard data = score ( keep = studentnumber &VarList rename = (&renameOpt) ) mean=75 std=5 out=stndtest; var Std:; run; data work.Combined; merge score stndtest; by studentNumber; run; proc print data=combined noobs round; title 'Standardized Test Scores for a College Course'; run; [/pre] ... View more

I would change year = 9999 to missing then use MISSING DESCENDING on class statement. [pre] data test; input id:$2. date:monyy.; if year(date) eq 9999 then date = .; format date monyy.; cards; 01 sep9999 01 nov2008 02 sep2008 02 sep2007 ;;;; run; proc print; run; proc format; value nadate . = 'N/A' other = [monyy7.]; run; proc tabulate; class date / missing descending; tables date,all; format date nadate.; run; [/pre] ... View more

I have EXECL 2003 and get the proper result with when using the updated TAGSET. With this really old verion NOTE: This is the Excel XP tagset (SAS 9.1.3, v1.28, 08/29/05). TAGATTR seems to be ignored. Upgrading to NOTE: This is the Excel XP tagset (SAS 9.1.3, v1.86, 04/15/08). I get the correct result. ... View more

I can tell you the problem but I don't know what happened to your program probably a macro variable with null value. In the INFILE statement the first parameter is positional and is FILEREF or Physical file. In your code that is missing and TRUNCOVER infile is now the first parameter. I think SAS then says "truncover is too long for a fileref then it must be a physical file but I can't find" error message. > ERROR: Physical file does not exist, > /usr/local/sas/***/Lev1/SASNew/TRUNCOVER. > INFILE > TRUNCOVER > DSD > %IF &IsMVS=FALSE %THEN %DO; > LRECL=32767 ... View more

This may not be what you want but I think it produces the output you desire. Seems to me the only problem is the TOTAL which I don't believe PROC REPORT will produce using BREAK statements, but if you create a total ID group you can get the desired output. [pre] proc plan noprint seed=469844687; factors id=3 ordered year=3 ordered y1 = 1 of 100 y2 = 1 of 200 ; output out=plan id cvals=('A' 'B' 'C') year nvals=(2006 2007 2008); run; proc print; run; data planV / view=planV; set plan; output; ID = 'T'; output; run; proc report nowd list headline; columns id year y1 y2 y3; define id / group; define year / group; compute y3; y3 = sum(y1.sum,y2.sum); endcomp; run; [/pre] ... View more

You can avoid the use of QUOTING functions in this case with the use of parenthesis(). Notice the change in your code where &SB is used in the macro, I removed the parenthesis because they are now part of the value. If you had a situation where the parenthiesis were not needed they could be removed with %qsubstr(%superq(sb),2,%length(%superq(sb))-2); [pre] 433 %let sc = '1224','1612','9603','9919','1920'; 434 %let FACTOR = ?; **Where is this?; 435 436 %macro sb_metrics(sb=('9999'), sb_txt_shrt=none, 436! sb_txt_lng=none); 437 %put _LOCAL_; 438 %put , count(distinct case when C.itm_sbrnd_cd in &sb 438! then A.lbi_indv_id else . end)&factor as 438! &sb_txt_shrt._indv_cnt 439 /* more code here but same concept */. 440 &factor as &sb_txt_shrt._indv_cnt 441 ; 442 %mend sb_metrics; 443 444 %sb_metrics(sb=(&sc)) SB_METRICS SB_TXT_SHRT none SB_METRICS SB_TXT_LNG none SB_METRICS SB ('1224','1612','9603','9919','1920') , count(distinct case when C.itm_sbrnd_cd in ('1224','1612','9603','9919','1920') then A.lbi_indv_id else . end)? as none_indv_cnt . ? as none_indv_cnt [pre] ... View more

Disclaimer: I'm not an expert. This problem is resolved with %UNQUOTE. Sometimes automatic unquoting does not work and UNQUOTEing is needed. Notice that the data step produces two data sets WORK.TEST and WORK.A. No error but not the intended result. Then in PROC PRINT data = TESTA causes another error. There are no spaces rather two tokens. ... View more

Use redirection to send STDERR to STDOUT. cmd args 2>&1 The syntax may be wrong you will need to check. Then in your data step read any lines returned when the command is executed and proceed as needed. By default any STDERR output from commands executed with PIPE should be showing up in the SASLOG. ... View more

This is a common problem solved with %UNQUOTE. systask command %unquote(%str(%')&BatchIt%str(%')); ... View more