Date:         Wed, 23 Feb 2005 10:08:14 -0500
Reply-To:     Art@DrKendall.org
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Art Kendall <Art@DRKENDALL.ORG>
Organization: Social Research Consultants
Subject:      Re: deleting single values
Comments: To: Lorenz Dobler <doblerl@uni-muenster.de>
In-Reply-To:  <000501c51994$7ac0d840$3c82b080@UPBC06>
Content-type: text/plain; charset=ISO-8859-1; format=flowed

Usually it is good practice to distinguish user-missing from system-missing.
User-missing means that you have decided that a particular value is
missing for a reason that you know about.
System-missing is when the computer either cannot resolve the input into
a meaningful result, or that a transformation cannot produce a
meaningful result.
"a31" or "O" (the letter oh) or "L" cannot be read as legitimate numbers.
SQRT of a missing or negative negative number, etc.

One method, useful if the legitimate numbers are in a limited range. For
IQ 55 to 145  is legitimate.
You can specify 3 specific missing values, or a range and one specific
value.
if IQ gt 145 IQ= IQ * -1.
Missing values IQ (lo thru 54).

Or
MIssing values IQ (156 thru hi).
value labels IQ
  990 'test interrupted'
  991 'probably faking low'
  992 'little English ability'
  993 'results misplaced'
  994 'pattern in options chosen'
  995 'behavior prevented valid test'
etc.

recode city (11,22,14, 8, 40 = -1)(else=copy) into newcity..
missing values city (lo thru 0, 99).
value labels
 99 'refused to tell'
 -1 'code given not valid'.

If you really want to assign system-missing
recode city (11,22,14, 8, 40 = sysmis )(else=copy) into newcity.

Art
Art@DrKendall.org
Social Research Consultants
University Park, MD  USA
(301) 864-5570





Lorenz Dobler wrote:

>Hello list,
>
>
>
>I want to delete single values from a numeric variable => convert selected
>single values to system missing values. How can I do that via syntax ??
>
>
>
>Hope you can help
>
>
>
>Regards
>
>
>
>Lenz
>
>
>
>Dr. Lorenz Dobler
>
>Umweltprobenbank des Bundes - Teilbank Humanproben
>
>Domagkstrasse 11
>
>48149 Münster
>
>Tel.: 0251/8352307
>
>FAX: 0251/55524
>
>email:  <mailto:doblerl@uni-muenster.de> doblerl@uni-muenster.de
>
>
>
>
>
>
>
>