Hi Wei,

Here is one method I like to use:

* create sample dataset to work with ; data sample; name ='Barney '; hometown = 'Graniteville '; amount = 101; flag = 'A'; output; name ='Betty '; hometown = 'ChrystalSprings'; amount = 102; flag = '1'; output; name ='BamBam '; hometown = 'Bedrock '; amount = 103; flag = 'B'; output; name ='BamBam '; hometown = 'Bedrock '; amount = 203; flag = 'B'; output; name ='BamBam '; hometown = 'Bedrock '; amount = 303; flag = 'B'; output; name ='Wilma '; hometown = 'Bedrock '; amount = 104; flag = '2'; output; name ='Fred '; hometown = 'Slab City '; amount = 105; flag = 'C'; output; name ='Pebbles '; hometown = 'Bedrock '; amount = 106; flag = '3'; output; name ='Pebbles '; hometown = 'Bedrock '; amount = 206; flag = '3'; output; name ='Dino '; hometown = 'Bedrock '; amount = 107; flag = 'D'; output; name ='Mr.Slate'; hometown = 'Marblemount '; amount = 108; flag = '4'; output; run;

proc sort data=sample; by name; run;

data singles multiples; set sample; by name; if first.name and last.name then output singles; else output multiples; run;

Hope this is helpful.

Mark Terjeson Senior Programmer Analyst, IM&R Russell Investment Group

Russell Global Leaders in Multi-Manager Investing

On Thu, 8 Sep 2005 17:58:11 -0700, wei yi <wy78712@YAHOO.COM> wrote:

>Hello, All >I have a dataset with some duplicate records, I know >to use nodupkey to exclude the duplicate records. >But how I could keep all these duplicate records in a >file? > >Thanks, >Wei > >__________________________________________________ >Do You Yahoo!? >Tired of spam? Yahoo! Mail has the best spam protection around >http://mail.yahoo.com