Try this technique: data detail ; input tapefile larry more curly ; cards ; 3 31 32 33 1 11 12 13 2 21 22 23 ; proc sort data=detail ; by tapefile ; run ; proc transpose data=detail out=long ; by tapefile ; var larry more curly ; run ; proc univariate data=long noprint ; by tapefile ; output out=res mean=mean pctlpts = 50 to 100 by 10 pctlpre=p ; run ; proc print data=res label ; label p50="50th Percentile" p60="60th Percentile" p70="70th Percentile" p80="80th Percentile" p90="90th Percentile" p100="100th Percentile" ; run ; Torri92 wrote: Hi everyone, My goal is to create a large data table that has descriptive statistics accompanying each observation. I am loading in a large file for which each observation has its own mean, standard deviation and percentiles. The variable that I am using to calculate all of this is named 'tapefile'. Basically what I am trying to do is create a large data table that neatly displays these statistics. I have been trying to accomplish this by saving the proc univariate command as a variable and then printing. I am new to sas so I am not entirely sure if this is even possible. Below is a sample of the data table that I seek to achieve. Observation Mean Std Dev 50th Percentile 60th Percentile 70th Percentile 80th Percentile 90th Percentile 100th Percentile 1 2 3 4 Here is a sample of my code as well: options ls=176 ps=44; libname master 'library location'; filename input0 'file location'; data detail;         infile input0 lrecl=605 recfm=f;                         input (long list of ebcdic variables); run; proc sort data=detail;         by tapefile; run; proc univariate data=detail noprint;         var tapefile;         output out=res mean=mean pctlpts = 50 to 100 by 10 pctlpre=p; run; proc print data=detail label;         var tapefile tab_guca_mod_1 res;         label         p50="50th Percentile"         p60="60th Percentile"         p70="70th Percentile"         p80="80th Percentile"         p90="90th Percentile"         p100="100th Percentile"; run; Thank you for any help you may be able to provide; I greatly appreciate it. ... View more

1. Try    proc sql _method ; 2. There are two completely separate usages of DISTINCT. One operates to remove redundant arguments to a summary function. The other eliminates redundant rows from a result set. 3. Automatic "remerging" of summary data with original detail is an extension to standard SQL. It is supported by SAS but not by all vendors. 4. You can name competing vendors. ... View more

data have ; input custID code ; cards ;         1      5         1      6         1      10         2      3            2      3         2      3 ; proc sql ; create view all as select range(code) as not_same , *   from have   group by custID   order by custID , code ; quit ; data same not_same ; set all ; if not_same then output not_same ; else             output     same ; drop not_same ; run ; ... View more

I meant of course to say Tech Support (not Report). TS has responded. They say this is a bug and that it will be fixed in Version 9.4. A SAS Note will be posted once analysis of the bug has been completed. From my own testing I perceive that a named view used in place of the inline view does not avoid the problem, but using a table to hold the intermediate result is a workaround. Howard Schreier wrote: We are still lacking any explanation, so I submitted a streamlined version of the problem to SAS Tech Report. I will pass along what I learn. PGStats wrote: I'm still scratching my head over this one. I don't understand why there in no remerging in the second version. I hope someone can explain this feature PG ... View more

and questions which have already been posted should not be re-posted with a different subject. ... View more

Maybe this is satisfactory: proc sort data=majors ; by name degree major ; run ; proc print data=majors ; by name degree ; id name degree ; run ; ... View more

There is an "almost" solution which will allow the logic to remain in SQL. See http://www.sascommunity.org/wiki/CREATE_statement_(SQL)_can_generate_only_one_output rsva wrote: Hi, I am trying to use Proc Sql to create 2 tables at a time. Below is my sample code in data step. Any suggestions would help. Data Match NoMatch;      Merge One (in=a) Two (in=b);      By id;      if a*b then output Match;      else if a=1 and b ne 1 then output NoMatch;      else delete; Run; Thanks. ... View more

Are you using Version 9.2 or later? medanesh wrote: I'm trying to use INTCK with a custom calendar. For that I have to set the INTERVALDS option of SAS. For some reason my SAS keeps giving me the Error that it doesn't recognize the option. Here is the exact line I'm submitting and the error I'm getting. options intervalds=(BankingDays=BankDayDS); ERROR 13-12: Unrecognized SAS option name, INTERVALDS. By the way even when I go to Tools>Options>System>Input Control>Data Processing (this is where the INTERVALDS option should be), I don't see it. Any help is appreciated. m ... View more

We are still lacking any explanation, so I submitted a streamlined version of the problem to SAS Tech Report. I will pass along what I learn. PGStats wrote: I'm still scratching my head over this one. I don't understand why there in no remerging in the second version. I hope someone can explain this feature PG ... View more