Date:         Wed, 13 Jun 2001 11:54:54 -0400
Reply-To:     ahutson@BIOSTAT.UFL.EDU
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Alan Hutson <ahutson@BIOSTAT.UFL.EDU>
Organization: University of Florida
Subject:      Re: weibull, log normal, exp, gamma?
Comments: To: "P.J. Kononoff" <pjk168@PSU.EDU>
Content-Type: text/plain; charset=us-ascii

I would suggest first  setting the maxiter option to something other
than default and see what happens

Best
Alan

"P.J. Kononoff" wrote:

> Hello all I have a question regarding curve fitting. A few years ago  I
> wrote some code which was supposed determine what distributions some data
> followed. I wrote the code using version 6.12 and have just tried to run it
> using version 8.1. I am receiving info in my log that the data "does not
> converge." This is peculiar to me as I did managed to have at least some of
> the distributions to converge.
>
> So I am wondering: if the data is indeed not fitting or I need different
> code for version 8.1.
>
> Can anyone help? Thanks,
> Sincerely Paul Kononoff
>
> /*************************************************************************
> Experiment: WPF-9
> Author: P.J. Kononoff
> Date: June 13, 2001
>
> This is a file of  which a sample of
> corn sialge and a sample of alfalfa haylage were sieved using the PSPS
> and data was fit to five distrbution types
> 1. Linear
> 2. Weibull
> 3. Exponential
> 4. Log Normal
> 5. Gamma
> *************************************************************************/
>
> options ls= 68 ps=70;
>
> data one;
> input size pus;
> cards;
>
> 19      97
> 12.7    91
> 6.3     52
> 3.96    16
> 1.17    4
> ;
>
> title 'model 1: linear model';
> proc reg data = one;
>         model pus = size /p;
>         run;
>
> title 'model 2 - weibull';
> proc nlin data = one;
> parms alpha=0.01 beta=0.05;
> model pus = 1 - exp (-(size/beta)**alpha);
> output out=out p=p;
> run;
>
> title 'model 3 expoential';
> proc nlin data = one;
> parms lambda = 0.1;
> model pus = 1 - exp (-lambda*size);
> output out=out p=p;
> run;