Date:         Thu, 12 Apr 2012 00:52:27 -0400
Reply-To:     Søren Lassen <s.lassen@POST.TELE.DK>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Søren Lassen <s.lassen@POST.TELE.DK>
Subject:      Re: Missing
Comments: To: Val Krem <valkrem@YAHOO.COM>
Content-Type: text/plain; charset=ISO-8859-1

Val,
Generally, SAS excludes missing values from statistics calculations.
The problem is that SAS does not see -9 as a missing value.

You will have to change you data before executing PROC MEANS. As your
data set is big, it may be worth the trouble to do the change using
a view (SQL or datastep) instead of a temporary table, e.g.

data v_temp/view=v_temp;
  set have;
  array x x1-x5;
  do over x;
    if x=-9 then x=.;
    end;
run;

proc means data=v_temp;
  var x1-x5;
/* etc. */

Using a view can save both disk space and processing time (as the large
table is only read once).

Regards,
Søren

On Tue, 10 Apr 2012 14:40:17 -0700, Val Krem <valkrem@YAHOO.COM> wrote:

>Hi all,


I� big data set containing several variable.� Some of the variables have
missing observation� represented by -9.

I want to calculate means of each variable by excluding the the missing
observations

I used proc means; var� x1 x2 x3 x4 x5;

Does proc means have option that exclude the missing observation?


Thanks� in advance