man, 24 05 2010 kl. 22:16 +0200, skrev Alois Schlögl:
> I've committed the changes already in r7326, so others (e.g. Michael)
> can see and test against them. The diff is attached.
These changes look good to me; thanks for doing this.
A few minor questions:
> -while niter++ <= maxout && nev(1) < maxev
> +while niter <= maxout
> + niter += 1;
> + if nev(1) < maxev, break; end;
I don't understand this change. I guess you moved the ++ operation to
help 'oct2mat' (which is fine by me), but why didn't you just write
while (niter <= maxout && nev (1) < maxev)
niter ++;
?
> - c = 1 + sum ((x-y)(:)'*z*((x-y)(:)));
> + c = 1 + sum (vec(x-y)'*z*(vec(x-y)));
What is the purpose of this (and similar) change?
> Index: main/optim/inst/minimize.m
> ===================================================================
> --- main/optim/inst/minimize.m (revision 7325)
> +++ main/optim/inst/minimize.m (revision 7326)
> @@ -134,23 +134,7 @@
> "isz", nan);
>
> if nargin == 3 # Accomodation to struct and list
> optional
> - va_arg_cnt = 1; # args
> - tmp = nth (varargin, va_arg_cnt++);
> -
> - if isstruct (tmp)
> - opls = list ();
> - for [v,k] = tmp # Treat separately unary and binary opts
> - if findstr ([" ",k," "],op0)
> - opls = append (opls, k);
> - else
> - opls = append (opls, k, v);
> - end
> - end
> - elseif is_list (tmp)
> - opls = tmp;
> - else
> - opls = list (tmp);
> - end
> + opls = varargin{1};
I am most certainly missing the something here, but this change just
looks odd to me. Can all of the above really be reduced to one line of
code?
Søren
------------------------------------------------------------------------------
_______________________________________________
Octave-dev mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/octave-dev
