Date:         Tue, 27 Feb 2007 16:11:13 -0600
Reply-To:     Robin High <robinh@UNLSERVE.UNL.EDU>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Robin High <robinh@UNLSERVE.UNL.EDU>
Subject:      Re: RECODE MISSING
Comments: To: Ran S <raan67@YAHOO.COM>
In-Reply-To:  <200702272000.l1RGGN0x023948@mailgw.cc.uga.edu>
Content-Type: TEXT/PLAIN; charset=US-ASCII

On Tue, 27 Feb 2007, Ran S wrote:

> Hi,
>
> I am using an array to recode universally all the variables with missing
> values (.) to 90. How can I just recode only those variables starting with
> prefix X_?
>

No compelling reason was given why you want to recode missing values of x_
to 90 and is almost never a good idea to do so; it is much safer to code
these missing values you want to be represented as '90' to one of the
other SAS missing data values, such as .z -- and then, if for some reason
you absolutely must see missing data as 90, you can assign .z to be 90 (or
some other code) with a FORMAT , e.g. something like:

proc format;
value ms .z='90';
run;

* can apply logic already given to identify
  all variables that begin with x_ ;

DATA xyz;
do i=1 to 10;
  x_ = 88+ CEIL(4*(RANUNI(929)));
  IF x_ eq 90 then x_= .z;
  OUTPUT;
END;

PROC freq; TABLE x_ FORMAT x_: ms. ; RUN;

The FREQ Procedure

                               Cumulative    Cumulative
x_    Frequency     Percent     Frequency      Percent
-------------------------------------------------------
89           2       33.33             2        33.33
92           4       66.67             6       100.00

Frequency Missing = 4   <--- 4 missing data values still there


.. and here they are coded as .z and represented as '90'

PROC PRINT NOobs; FORMAT x_: ms. ; RUN;

     i    x_

     1    92
     2    89
     3    90
     4    92
     5    89
     6    90
     7    90
     8    92
     9    90
    10    92


Robin High