Re: optimset/optimget

[Jaroslav Hajek]

On Mon, Jan 26, 2009 at 8:29 PM, John W. Eaton <address@hidden> wrote:
> I see you checked in this change:
>
>  http://hg.savannah.gnu.org/hgweb/octave/rev/ffc9e9737507
>
> Without something like this, how do you propose to handle calls like
>
>  optimset ("fsolve")
>  optimset ("leasqr")
>
> which are supposed to return the set of options relevant to the named
> function.  I think the solution of putting this information in a
> separate file is better than embedding it in the optimset function
> because it allows anyone to extend optimset to handle their new
> optimization function which requires uses new/extra options.
>

Sorry, I didn't realize that the function was actually used - I
grepped through the sources and found no use (for reasons now
obvious), so I supposed it was some old relic.
Maybe we can copy Matlab's way to get the defaults, i.e. calling
fzero('defaults') yields the default options for fzero etc?
In any case, the default options list needs to be updated.

> Also, whether or not you decide to remove the __*_defopts__ functions,
> I think we should make a change like this so that optimset and
> optimget are not case sensitive.

OK, right, they shouldn't be.

cheers

-- 
RNDr. Jaroslav Hajek
computing expert
Aeronautical Research and Test Institute (VZLU)
Prague, Czech Republic
url: www.highegg.matfyz.cz