[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: SQP doc patch on savannah
From: |
Rik |
Subject: |
Re: SQP doc patch on savannah |
Date: |
Tue, 02 Nov 2010 12:16:19 -0700 |
> We don't seem to be using the savannah patch tracker. However that
> doesn't mean that people aren't sending patches to this tracker. I've
> closed one as its been applied, asked for help on another, but the
> following patch
>
> https://savannah.gnu.org/patch/index.php?7207
>
> seems trivial. Should it be applied?
It is trivial, but it seems to be system dependent. The sqp documentation
currently says that the example problem takes 8 iterations and 10 function
evaluations. The patch changes this to 9 iterations and 9 function
evaluations. If I run it on my current machine I get 8 iterations and 9
function evaluations.
I suppose we could try to find an initial vector, x0, which always
converges in the same number of iterations and function evaluations, but
this seems like too much work.
--Rik