is the one using the XYZ decorator has any efficiency advantage ? On Aug 7, 2009, at 11:00 AM, Daniel Russel wrote:
> > On Aug 7, 2009, at 10:32 AM, Dina Schneidman wrote: > >> wouldn't it be easier to have transform function after all? :) > > we basically do :-) > > std::foreach(ps.begin(), ps.end(), SingletonFunctor(new > core::Transform(t))) > > That said, personally, I would just go for: > > for (Particles::iterator it = ps.begin(); it != ps.end();it++){ > core::XYZ d(*it); > d.set_coordinates(t.transform(d.get_coordinates()); > } > > or > > std::foreach(ps.begin(), ps.end(), > boost::bind(&Transformation3D::transform, &t, > boost::lambda::placeholders::_1)) > > ok, maybe not that last one :-) > > > > >> >> On Thu, Aug 6, 2009 at 11:20 PM, Daniel Russeldrussel@gmail.com >> wrote: >>> Side effect of enforcing that IMP::Objects aren't allocated on the >>> stack >>> (since they are ref counted). >>> >>> >>> On Aug 6, 2009, at 11:04 PM, Keren Lasker wrote: >>> >>>> Daniel - >>>> why is it better using: >>>> IMP_NEW(core::Transform, tsm,(t)); >>>> for(Particles::iterator it = ps.begin(); it != ps.end();it++) { >>>> tsm->apply(*it); >>>> } >>>> instead of: >>>> std::for_each(ps.begin(),ps.end(),SingletonFunctor(new >>>> IMP::core::Transform(t))); >>>> >>>> I saw you changed it in domino . >>>> >>>> thanks, >>>> Keren. >>>> _______________________________________________ >>>> IMP-dev mailing list >>>> IMP-dev@salilab.org >>>> https://salilab.org/mailman/listinfo/imp-dev >>> >>> _______________________________________________ >>> IMP-dev mailing list >>> IMP-dev@salilab.org >>> https://salilab.org/mailman/listinfo/imp-dev >>> >> _______________________________________________ >> IMP-dev mailing list >> IMP-dev@salilab.org >> https://salilab.org/mailman/listinfo/imp-dev > > _______________________________________________ > IMP-dev mailing list > IMP-dev@salilab.org > https://salilab.org/mailman/listinfo/imp-dev