|Date: ||Fri, 9 Mar 2001 08:22:08 -0500|
|Reply-To: ||John Jones <jonesj@PHARMARESEARCH.COM>|
|Sender: ||"SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>|
|From: ||John Jones <jonesj@PHARMARESEARCH.COM>|
|Subject: ||Re: Best way to display a PDF|
|Content-Type: ||text/plain; charset="us-ascii"; format=flowed|
A couple of ideas for you.
If you want a smooth line what about an xy Cartesian plot using proc GPLOT
together with the symbol statement join=spline option? Make as large a
number of intervals as you can and calculate the percent in each
interval. Then plot the mid point of each interval versus the percent to
generate the xy.
Along these lines you might also simulate your theoretical distribution and
overlay it on top of the empirical one. Try the estimated mean and
standard deviation with the normal or rannor functions (they're identical).
If you had the INSIGHT module you'd be set.
Good luck... John
At 07:50 AM 3/9/01 -0500, Jack Shoemaker wrote:
> We have BASE, STAT, and GRAPH. No QC. Looking for thoughts and
>suggestions on the best way to display an empirical PDF. There's always the
>histogram available in PROC GCHART. Wondering if there are any better ways
>to do this with the tools at hand.
>TIA - Jack
>Jack N Shoemaker / JShoemaker@Accordant.net
>Visit our patient communities at http://www.accordant.com or our corporate