Date:         Sun, 16 Sep 2007 08:39:09 -0400
Reply-To:     Arthur Tabachneck <art297@NETSCAPE.NET>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Arthur Tabachneck <art297@NETSCAPE.NET>
Subject:      Re: Calculate percentages of a subsample
Comments: To: Haigang Zhou <hzhou3@UNLSERVE.UNL.EDU>

Haigang,

I apparently misunderstood what you're asking for, as my results are
slightly different than those shown in your example.

Regardless, there are numerous ways to get what you want, including:

data temp;
  merge one (rename=(value=cutoff)) two;
  by id;
  if value gt cutoff then test="higher";
    else test="lower";
run;

proc freq data=temp;
  tables test/out=want ( drop=count where=(test eq "higher"));
  by id;
run;

I'd use proc freq so that you can easily control whether you want missing
values included in the calculations and/or overall percentages

HTH,
Art
---------
On Sun, 16 Sep 2007 06:35:42 -0400, Haigang Zhou <hzhou3@UNLSERVE.UNL.EDU>
wrote:

>I want to calculate the percentages of values in dataset2 that are larger
>than the pre-specied value in dataset1 grouped by id.
>
>data one;
>     input value id;
>  cards;
>  -.1  150
>  2    1150
>  ;
>data two;
>     input value id;
>  cards;
>-2.02301487 150
>1.160065383 150
>0.134215614 150
>1.87326498 150
>1.214724878 150
>0.278588803 1150
>2.090436399 1150
>0.192193627 1150
>1.771033228 1150
>2.009589141 1150
>;
>run;
>
>I want to have a dataset contains the following result.
>
>ID   percent
>--------------
>150   80%
>1150  60%
>
>Any help is appreciated.