Mona -

To add a couple cents to the great answers you already have - what you think are missing values are likely to be non-printing characters. You may want to process non-writable characters out of your variable values - I show two methods - the COMPRESS will remove them, the TRANSLATE will replace them with blanks. I suggest the TRANSLATE because it will keep the spacing of the original values. If you then want to reduce the multiple spaces to single spaces the COMPBL function will work: missingvar=COMPBL(missingvar)

data _null_;

missingvar='00'x; /*looks like a "missing" but is really a ASCII null character*/

put missingvar=; if missing(missingvar) then put 'Really Missing'; else put 'Not Really Missing';

/*missingvar=COMPRESS(missingvar,,'wk');*/ missingvar=TRANSLATE(missingvar,' ',compress(collate(0,,256),,'w'));

put missingvar=; if missing(missingvar) then put 'Really Missing'; else put 'Not Really Missing';

run;

hope that's useful -

Paul Choate DDS Data Extraction (916) 654-2160

-----Original Message----- From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU] On Behalf Of Mona Sent: Monday, May 04, 2009 11:49 AM To: SAS-L@LISTSERV.UGA.EDU Subject: Weird Character String

Hi All,

I have a dataset where this one variable has a missing value for a lot of observations. The problem I having is when I use the missing function or just a plain if to filter out records with missing value of that variable, nothing gets filtered. I do see a lot of them with missing values but somehow it is not picking. Is there any function to resolve this ?

Any help appreciated.