Date:         Thu, 16 Oct 1997 00:08:13 -0700
Reply-To:     Andrew James Llwellyn Cary <ajlcary@CARYCONSULTING.COM>
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         Andrew James Llwellyn Cary <ajlcary@CARYCONSULTING.COM>
Subject:      Re: PROC Transpose
Comments: To: RICHARD STEVENS <RSTEVENS@BCBSCT.COM>
Content-Type: text/plain; charset="us-ascii"

Here is two different solutions:
 
Data step;
 
DATA MCCCI.PDSTRAN;
        LENGTH VARNAME $8 MEMNAME $8;
        ARRAY V mbrnm1-mbrnm13;
        SET work.pdsin;
        DO OVER V;
               MEMNAME=V;
               CALLVNAME(V,VARNAME);
               OUTPUT;
        END;
        KEEP DSN MEMNAME VARNAME;
RUN;
 
The data doesn't even have to be sorted and it takes one pass. It also copes with more then two rows per
dataset name.
 
 
PROC TRANSPOSE SOLUTION
 
 
proc transpose data=work.pdsin out=work.junk;
  by dsname;
  var mbrnm1-mbrnm13;
 run;
proc transpose data=work.junk out=mccc1.pdstrn (where=(col1 ne ''));
  by dsname _NAME_;
  var COL1-COL2;
 run;
 
Two passes through after sorting. Hmmm.
 
 
 
 
On Wednesday, October 15, 1997 12:43 PM, RICHARD STEVENS[SMTP:RSTEVENS@BCBSCT.COM] wrote:
>Hi all,
>
>This probably has a simple solution, but I can't seem to see it.  The bottom
>line is I want the the results to have one COL1 not more than one.  I have
>tried this on MVS v6.09 and MS Windows 3.1, SAS 6.11 and get the
>same results.  The manual example shows that this is possible, but using
>the example, I don't get the results.  (SAS Procedures Guide v6 3rd ed
>p597).
>
>The files and code follow:
>
>SAS data -- mccc1.pds
>OBS    MBRNM1      MBRNM2      MBRNM3      MBRNM4      MBRNM5
>MBRNM6    . . .
>
>  1    @JOBX951    @JOB01X     @JOB02      CHCMEMRL    CLMX01
>CLM01
>  2    SEASONT1    SEASON01    SEASON02    SEASON03    SEASON99
> TEMP01
>
>
>MBRNM10     MBRNM11     MBRNM12     MBRNM13            DSNAME
>
>EMUC03      EMUC04      FORMAT      HEDISCL
>A438U11.CONTAIN.CODE
>                                                A438U11.CONTAIN.CODE
>
>I tried rearranging the variables to see if that helped, it didn't.
>proc sql;
>create table work.pdsin as
> select
>dsname,mbrnm1,mbrnm2,mbrnm3,mbrnm4,mbrnm5,mbrnm6,mbrnm7
>              ,mbrnm8,mbrnm9,mbrnm10,mbrnm11,mbrnm12,mbrnm13
> from mccc1.pds
>;
>
>proc transpose data=work.pdsin out=mccc1.pdstrn let;
> by dsname;
> var mbrnm1-mbrnm13;
>run;
>
>The output:  I would like to see the values under _NAME_ repeated with
>the values COL2, therefore no COL2 or greater in the out file.
>OBS         DSNAME          _NAME_   COL1      COL2
>
>  1  A438U11.CONTAIN.CODE   MBRNM1   @JOBX951  SEASONT1
>  2  A438U11.CONTAIN.CODE   MBRNM2   @JOB01X   SEASON01
>  3  A438U11.CONTAIN.CODE   MBRNM3   @JOB02    SEASON02
>  4  A438U11.CONTAIN.CODE   MBRNM4   CHCMEMRL  SEASON03
>  5  A438U11.CONTAIN.CODE   MBRNM5   CLMX01    SEASON99
>  6  A438U11.CONTAIN.CODE   MBRNM6   CLM01     TEMP01
>  7  A438U11.CONTAIN.CODE   MBRNM7   CONTENTS  TSODB2C
>  8  A438U11.CONTAIN.CODE   MBRNM8   EMUC01
>  9  A438U11.CONTAIN.CODE   MBRNM9   EMUC02
> 10  A438U11.CONTAIN.CODE   MBRNM10  EMUC03
> 11  A438U11.CONTAIN.CODE   MBRNM11  EMUC04
> 12  A438U11.CONTAIN.CODE   MBRNM12  FORMAT
> 13  A438U11.CONTAIN.CODE   MBRNM13  HEDISCL
>
>
>Thanks for you assistance.
>Richard
>
>
 
 
---------------------------------------
Andrew J. L. Cary
Senior Curmudgeon
Cary Consulting Services
http://www.caryconsulting.com