fre, 12 03 2010 kl. 12:33 +0100, skrev Olaf Till: > On Thu, Mar 11, 2010 at 12:51:51PM -0800, Søren Hauberg wrote: > > ... > > In general it seems to me like 'optim' doesn't blend in as well with > > more recent versions of Octave as it really should. The main problem > > seems to be that both Octave and 'optim' comes with an implementation of > > 'fminbnd'. I don't know which is better, > > I also don't know, but the packages 'fminbnd' seems incomplete --- it > does not honour "options", and a comment says "this will not work for > symmetric funcs" for the whole method. I speculate Jaroslav thought it > inadequate to start from and so wrote it new instead ... I'd just say > take his.
Does anybody disagree with this? I haven't really seen any replies here. If nobody replies in the next couple of days, I will remove 'fminbnd' from 'optim'. > >but I don't like the > > duplication. > > What should be done if a package should work with different Octave > versions, one of which replaces a package function written in m-code? > Conditional install of this package function, even if it were > possible, would not work if one wants to work with both of these > different Octave versions (since m-code is installed under > version-independent paths). I think the best way to deal with this is to have namespaces for packages. This isn't supported at the moment by Octave, so I think we should just try not to create functions that have the same name as those in Octave core. Søren ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ Octave-dev mailing list Octave-dev@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/octave-dev
