[R] ordinal logistic model with pre-defined coefficients

[Prof Brian Ripley]

On Fri, 7 Nov 2008, kazys.o.simutis at aib.ie wrote:

> Hi,
>
> I'm trying to fit a proportional ordinal logistic model using function
> polr() (package MASS).
>
> Is there a way to fix certain betas in the regression (e.g. function
> arima() allows this by defining fixed )
> Maybe there is another function than polr() which allows that?

See ?offset, the standard tool in regression-like fitting in R (and 
described in the book MASS supports).

> Thanks
> Kazys

-- 
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