If ALL of your variables are character variables, and NONE have a name that is longer than 31 characters, and you want to change ALL of them to be numeric variables according to your stated recode statement, and you want to keep the same variable names, then you could use a data _null_ step with arrays to write and run call execute statements to accomplish the task. If all of the above is true, then the following code can accomplish the task: /*create some example data*/ data j1;   length this that something_else $12;   input (this that something_else) (& $);   cards; 1  y  n na  n/a  5 x  $$"ER", 9903  no ; /*create and run some call execute statements*/ data _null_;   set j1 (obs=1);   array all(*) _char_;   length exline $200;   call execute('data j2 (drop=_:); set j1 (rename=(');   exline=catt(vname(all(1)),'=_',vname(all(1)));   call execute(exline);   do i=2 to dim(all);     exline=cat(' ',strip(vname(all(i))),'=_',strip(vname(all(i))));     call execute(exline);   end;   call execute('));');   do i=1 to dim(all);     exline=catt('if missing(_',vname(all(i)),') or upcase(_',     vname(all(i)),       ') in ("NO", "N", "NA", " ",',       "'$$",'"ER", 9903',"') then do;",vname(all(i)),'=0;end;else do;',       vname(all(i)),'= 1; end;');     call execute(exline);   end;   call execute('run;'); run; ... View more

At windows (destktop) by default SAS is, should be, allowed to used all memory present in the desktop. SAS(R) 9.4 Companion for Windows, Third Edition  The only place it can be changed as at startup (sas-config or parameter sas-script) At servers it is usual to have limits set to allow it as multi-user processing with a lot of processes. A server can have 256Gb or 2Tb of internal memory 64 cores where your desktop is coming with 8 of may be 16 using 8 cores. Sometimes with VDI your Windows session is stripped to less than 3Gb with a single core. This is according the IT dogma that users do not run heavy processes. With sas you can see your memory settings and usage. There are more memory options like SAS(R) 9.4 Companion for Windows, Third Edition (maxmemquery) setting a s size of allocation blocks. proc options group=memory   Base SAS(R) 9.4 Procedures Guide, Third Edition will give your sas settings. What is your hardware configuration? You SAS version ... View more

If you change the order of the datasets this warning will desapear. Also if you change the format before the "set" with format var BEST12. it will be solved too. ... View more

Sorry, I cannot read that code.  Please use indentations, new lines etc.  If you expect 13307 records then your last step should presumably be: proc sql;      create table J1 as      select     A.*,                    B. the additional variables here      from       H82 A      left join   H111 B      on           A.CUSIP=B.CUSIP      and        A.PFOYR=B.PFOYR; quit; ... View more

Add the statement ODS output ParameterEstimates=myEstimates; to your proc reg code. The dataset myEstimates will be created with all the parameter estimates table info for every year. Look for the section ODS table names in the Details section of proc reg documentation. Every table seen in the output window can be written to a dataset this way. PG ... View more

If I understood what you mean. data class; set sashelp.class; run; proc sort data=class;by sex;run; proc summary data=class ; by sex; var weight; output out=stat(drop=_:) mean=mean std=std; run; data want; merge class stat; by sex; select;   when(weight(mean+std)) pfo=3;   otherwise pfo=2; end; run; Xia Keshan ... View more

data have; input name $ count; datalines; a          1 b          0 c          0 d          1 e          1 ; proc tabulate data=have; var count; table count,sum; run; ... View more

The value of "pfo" depends on the variables pfoyr and index, right? So for different years, different index-values need to be checked. If those assumptions are right, try this: define an informat for each year set pfo using that format Example code: proc format;    invalue Index1991Fmt       low -< -1.22 = 1       -1.22 - 2.48 = 2       2.48 <- high = 3    ; quit; data f2;      set f1;      length pfo 8 FmtName $ 32;      drop FmtName;           FmtName = cats('Index', pfoyr, 'Fmt.');      pfo = inputn(index,FmtName); run; ... View more

From your post   you want to delete rows with those values in siccd. OK just insert the word NOT  into your code before the IN( Data r8; set r7; if siccd NOT in (131,170, 195, 204); run; The other is a range not an IN( list    so that is a subsetting IF condition like if 2100 LE siccd LE 2199 then delete ; ... View more

Or just move the ARRAY statement to before the LENGTH statement that defines SCORE. ... View more