668  *************************************************
669  * file = manip_if.sas                           *
670  * The SAS program in this file creates new      *
671  * variables using IF-THEN statements and several*
672  * several assignment statements.                *
673  * The program assumes that the data set "htwt"  *
674  * has been created.                             *
675  *************************************************;
676
677  options linesize=min;
678
679  *------------------------------------------------*
680  * Create 2 grouping formats, one character, one  *
681  * numeric.  Also create 2 labelling formats,     *
682  * both of which must be character because the    *
683  * grouping format's new values are character.    *
684  * All 4 formats are used later in the program.   *
685  *------------------------------------------------*;
686
687  /* The values on the right of "=" have been
688     enclosed in quotes so that when they are used
689     to create new variables the new variables will
690     be character (and take up less space) */
691
692  proc format;
693    value $namef    /* requires "$" because values
694                       on left of "=" are character */
695                 'Elizabeth','David','James' = '1'
696                                      other  = '0';
NOTE: Format $NAMEF is already on the library.
NOTE: Format $NAMEF has been output.
697
698    value agefmt    /* does not require "$" because
699                       values on left are numeric  */
700                 0-29 = '1'
701                30-39 = '2'
702                40-49 = '3'
703              50-high = '4';
NOTE: Format AGEFMT is already on the library.
NOTE: Format AGEFMT has been output.
704
705    value $namel     '1' = '3 names'
706                     '0' = 'all other names';
NOTE: Format $NAMEL is already on the library.
NOTE: Format $NAMEL has been output.
707
708    value namel     1 = '3 names'
709                    0 = 'all other names';
NOTE: Format NAMEL is already on the library.
NOTE: Format NAMEL has been output.
710
711    value $agelbl      '1' = '1: 0 to 29 yrs'
712                      '2' = '2: 30 to 39 yrs'
713                      '3' = '3: 40 to 49 yrs'
714                      '4' = '4: 50+ years old';
NOTE: Format $AGELBL is already on the library.
NOTE: Format $AGELBL has been output.
715  run;

NOTE: PROCEDURE FORMAT used:
      real time           0.10 seconds


716
717  ******************************************
718  * Create a new temporary SAS data set,   *
719  * same name, to add new variables        *
720  ******************************************;
721
722  data htwt;
723    set htwt;
724
725  *------------------------------------------------*
726  * 1. Create dichotomous variables by referencing *
727  *    one category of values of one variable.     *
728  *------------------------------------------------*;
729
730           /* The original variable is "age" and the new
731              variables are "age2grp" and "age2grpx".  Both
732          new variables have identical values ("1" and "0")
733              except the 1st approach results in a character
734              variable and the 2nd approach results in a
735              numeric value. */
736
737  if age<50 then age2grp='1';   /* using IF/THEN  */
738    else age2grp='0';
739
740  age2grpx=(age<50);            /* assignment statement */
741
742  *------------------------------------------------*
743  * 2. Create dichotomous variable by referencing  *
744  *    multiple values of one variable.            *
745  *------------------------------------------------*;
746
747           /* The original variable is "name" and the new
748              variables are "newname" and "newnamex".  All 3
749          new variables have identical values ("1" and "0")
750              which differ re numeric vs character */
751
752  if name in ('Elizabeth','David','James')
753     then newname='1';
754     else newname='0';      /* new variable is character
755                          - use character format to label*/
756
757                            /* new variable is numeric
758                          - use numeric format to label */
759  newnamex=(name in ('Elizabeth','David','James'));
760
761  newnamey=put(name,$namef.);/* new variable is character */
762
763        /* If the series of values were numeric, no quotes
764            would be used, e.g., if region in (1,5,9)... */
765
766  *-----------------------------------------------------*
767  * 3. Create 2 multi-value variables called "agegroup" *
768  *    and "agegrpx", referencing ranges of values of   *
769  *    one variable ("age").                            *
770  *-----------------------------------------------------*;
771
772  if 0<=age<=29 then agegroup='1';
773  else if 30<=age<=39 then agegroup='2';
774  else if 40<=age<=49 then agegroup='3';
775  else if age>49 then agegroup='4';
776
777  agegrpx=put(age,agefmt.);     /* use the agefmt format */
778
779                         /* label one of the new variables */
780  label agegrpx = 'Age grouped into 4 categories';
781
782  run;

NOTE: The data set WORK.HTWT has 18 observations and 12
      variables.
NOTE: DATA statement used:
      real time           0.27 seconds


783
784
785  proc freq data=htwt;
786    tables age * age2grp * age2grpx /list missing;
787    tables name * newname * newnamex * newnamey/list missing;
788    tables age * agegroup * agegrpx /list missing;
789
790         /* add labels to the values of the new variables */
791    format newname newnamey $namel. newnamex namel.
792           agegroup agegrpx $agelbl.;
793  title1 'The height/weight data set';
794  title2 'Check new variables against original variables';
795  run;

NOTE: PROCEDURE FREQ used:
      real time           0.59 seconds


796
797  proc contents data=htwt;
798  title2;          /* remove 2nd title for remaining procs */
799  run;

NOTE: PROCEDURE CONTENTS used:
      real time           0.16 seconds


800
801  proc print data=htwt (obs=10);
802  run;

NOTE: PROCEDURE PRINT used:
      real time           0.04 seconds