Date:         Thu, 16 Jul 2009 09:47:55 -0700
Reply-To:     "Terjeson, Mark" <Mterjeson@RUSSELL.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         "Terjeson, Mark" <Mterjeson@RUSSELL.COM>
Subject:      Re: separating observations in a dataset
Comments: To: Randall Powers <powers_r@BLS.GOV>
In-Reply-To:  A<200907161617.n6GAlXS6007375@malibu.cc.uga.edu>
Content-Type: text/plain; charset="us-ascii"

Hi Randall,

Several ways to do this, here is one approach:


data sample(drop=i);
    length type $5;
    do i = 1 to 4;
        type = 'iter'; output;
    end;
    type = 'final'; output;
    do i = 1 to 107;
        type = 'iter'; output;
    end;
    type = 'final'; output;
    do i = 1 to 22;
        type = 'iter'; output;
    end;
    type = 'final'; output;
    do i = 1 to 111;
        type = 'iter'; output;
    end;
    type = 'final'; output;
run;

data result;
    set sample;
    retain run_number 1;
    retain counter 0;
    counter + 1;
    if type eq 'final' or counter eq 102 then
        do;
            desired_obs = 'Y'; * flag it ;
            output;
            counter = 0; * reset ;
            run_number + 1; * reset ;
        end;
    else output;
run;




Hope this is helpful.


Mark Terjeson
Investment Business Intelligence
Investment Management & Research
Russell Investments
253-439-2367


Russell
Global Leaders in Multi-Manager Investing





-----Original Message-----
From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU] On Behalf Of
Randall Powers
Sent: Thursday, July 16, 2009 9:18 AM
To: SAS-L@LISTSERV.UGA.EDU
Subject: separating observations in a dataset

Hello All,

There must be a way to do this!

I have a dataset with variable "type". The values for type are
either "iter" or "final". There are numerous (10,000+) observations.

The file is a collection of the parameters from a run of proc nlin
(probably not pertinent to my question). "Final" represents the final
value
of of a converging nlin program. "Iter" means that it is an interation.
The
program will either converge at or before 102 observations, or it will
stop
and thus the 102nd observation in the sequence is still "Iter".


What I want to do is output all of the "final"s to file A, and output of
the 102nd "iter" for those that don't reach a final value by that point,
to
file B. How do I do this?

Thus, an example of my starting dataset file would be:

iter (start of run #1)
iter
iter
iter
final
iter (start of run #2)
iter
.
.
.
.
(102nd) iter (and in this case 107th overall observation))
iter (this would be start of run #3)
iter
iter
iter
iter
iter
iter
final
iter (start of run #4...there are 1000 runs total)