[R] nlminb() - how do I constrain the parameter vector properly?

[Prof J C Nash (U30A)]

This is one area where more internal communication between the objective 
function (inadmissible inputs) and optimizer (e.g., Quasi-Newton) is 
needed. This is NOT done at the moment in R, nor in most software. An 
area for R&D. In Nash and Walker-Smith (1987) we did some of this in 
BASIC back in mid-80s. Still trying to redo that for R, but it won't be 
done quickly due to the much bigger infrastructure of R.

The "trick" with using a large value or Inf (which sometimes causes
other errors) usually slows the optimization, whereas communicating that 
the objective is inadmissible in a line search can often be simply a 
shortening of the step size.

JN

On 13-10-21 06:00 AM, r-help-request at r-project.org wrote:
> Message: 34
> Date: Mon, 21 Oct 2013 05:56:45 -0400
> From: Mark Leeds<markleeds2 at gmail.com>
> To: Steven LeBlanc<oreslag at gmail.com>
> Cc:"r-help at R-project.org"  <r-help at r-project.org>
> Subject: Re: [R] nlminb() - how do I constrain the parameter vector
> 	properly?
> Message-ID:
> 	<CAHz+bWYEtvZjiCCauGVxstGcqmF6ENw0N3MMb7jfa3OkYkBTKA at mail.gmail.com>
> Content-Type: text/plain
>
> my mistake. since nlminb is minimizing, it should be +Inf  ( so that the
> likelihood
> is large ) as you pointed out. Note  that this approach is a heuristic and
> may  not work all the time.