MathGroup Archive 2008

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: FindMinimum constraints


Thanks for the quick response.  Join[] did indeed do the trick.

But adding this constraint introduced another problem.  By fixing
a[100] == 0.5*(a[0] + a[25], I (inadvertently) allowed FindMinimum to
choose solutions that seem to have discontinuous derivatives (i.e.,
there are really sharp points at a[100], a[0], and a[25]).  I guess I
should also somehow force the solution to be sufficiently 'smooth'.
Is there an easy way to do that?

Thanks again for all the help!

Dan

On Oct 21, 1:43 am, Jean-Marc Gulliet <jeanmarc.gull... at gmail.com>
wrote:
> dantimatter wrote:
> > I have a large and complicated model for which I'm using the
> > FindMinimum function to find a solution.  The solution consists of 10=
1
> > numbers a[i] (i goes from 0 to 100), all greater than zero, and the
> > search for each a[i] begins at 0.001:
>
> > sol = FindMinimum[{model, Table[a[i] > 0,{i, 0, 100}]}, Table[{a[i],
> > 0.001},{i, 0, 100}]];
>
> > The line of code that I'm showing here works great.  My problem now i=
s
> > that I'd like to add an additional constraint:  that a[100] ==
> > 0.5*(a[0]+a[25]).  I can't for the life of me figure out how to
> > include this added constraint in the line of code that already works.
> > I get errors when I use Append[] to the Table. I would appreciate any
> > thoughts on this.
>
> Using *Join* should do the trick. For instance,
>
> sol = FindMinimum[{model,
>     Join[{a[100] == 0.5*(a[0] + a[25])},
>      Table[a[i] > 0, {i, 0, 100}]]}, Table[{a[i], 0.001}, {i, 0, 10=
0}]]
>
> Regards,
> -- Jean-Marc



  • Prev by Date: Controlling the order of evaluations
  • Next by Date: Re: Speeding up a list construction
  • Previous by thread: Re: FindMinimum constraints
  • Next by thread: Re: Speeding up a list construction