Re: [R] force glm estimates to be nonnegative

From: Prof Brian Ripley <ripley_at_stats.ox.ac.uk>
Date: Fri, 25 Apr 2008 05:53:10 +0100 (BST)

On Fri, 25 Apr 2008, Michele Christina Itten wrote:

> Is there a way to force certain formula parameters to be nonnegative?
>
> What I want to do is to estimate student capacity over time, namely by
>
>> capacity ~ Student + Student:Day
>
> I add this formula to a glm call and obtain negative learning slope estimates (Student:Day) in some cases.
> However, I don't want to allow for that. In such a case, glm should solve
>

>> capacity ~ Student
>
> and I will set the slope to zero.
>
> Is there an option to constrain a parameter set to nonnegative values?

Not in glm(). Easily if you fit the model by optim(), and there is example code for binomial GLMs in MASS chapter 16 (and the scripts in its package).

>
>
>
>
> Many thanks in advance!
>
> Michele Itten,
> BS. Carnegie Mellon
>
> ______________________________________________
> R-help_at_r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>

-- 
Brian D. Ripley,                  ripley_at_stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

______________________________________________
R-help_at_r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Received on Fri 25 Apr 2008 - 05:01:41 GMT

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.2.0, at Fri 25 Apr 2008 - 06:30:31 GMT.

Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-help. Please read the posting guide before posting to the list.

list of date sections of archive