Date:         Fri, 13 May 2005 14:16:33 -0400
Reply-To:     Jerry Davis <jwd@GRIFFIN.UGA.EDU>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Jerry Davis <jwd@GRIFFIN.UGA.EDU>
Subject:      Re: ODS statistical graphics
In-Reply-To:  <OFC35817EB.53FDB11B-ON88256FFF.0075AF18-88256FFF.0076B355@epamail.epa.gov>
Content-Type: text/plain; charset="US-ASCII"

"David L. Cassell" wrote:

> > I've been playing around with ODS statistical graphics, especially in
> > the new GLIMMIX procedure,
>
> Cool stuff, huh?

Yep.  GLIMMIX has some nice features (Oliver be praised).  I hope some
of the lsmeans stuff find their way to other procedures.  Even though
GLIMMIX is very robust and will solve lots of problems.

> Actually, I considered that to be a *feature*.  I would hate to spend
> the time to generate important graphics, only to turn around and find
> they had been 'cleaned up' with the trash.

I guess it's a matter of preference.  Once I get things the way I want,
I just save the program and move on to something else.  If it worked
once it will work again.

> In the ODS GRAPHICS statement, you can use the PERSIST option to
> control persistence of the data cache.  And you can use the RESET
> option to reset the image counter, so ODS will write over your
> previous graphics files.

A couple responders mentioned this option.  I haven't found much
documentation for ODS stat graphics, but did find the RESET option in a
SUGI paper (which I should have looked at before I posted).  I don't use
ODS much, if a client wants prettier output I wish them the best.  :-)

Jerry
--
Jerry Davis
Experimental Statistics
UGA, CAES, Griffin Campus