I thought that you were touching other functionalities, but that's ok. I don't think that anyone has used that fuzzy methods. At least not seriously.
On Wed Apr 3 21:21:39 2013, Daniel Russel wrote: > It has a bunch of arguments having to do with trying to approximate assignments and do fuzzy merges and things like that. > > Specifically keeping > - const Subset &first_subset > - AssignmentContainer* first, > - const Subset &second_subset, > - AssignmentContainer* second, > - const SubsetFilterTablesTemp &filters, > - AssignmentContainer* ret, > > These arguments have to do with attempts to do fuzzy merging of assignments: > - double max_error=0, > - ParticleStatesTable *pst=nullptr, > - const statistics::Metrics &metrics = statistics::Metrics(), > > This argument has to do with limiting the number of states generated, but the same effect can be gotten by using a CappedAssignmentContainer: > - unsigned int max_states = std::numeric_limits<int>::max()); > > > > On Apr 3, 2013, at 8:08 PM, Javier Velazquez-Muriel javi@salilab.org wrote: > >> >> Hi, >> >> I've used that function for em2D, but I don't think that the filters were necessary. Dave and I used it too for the pepdock and pepdock2 modules. What failed experiments are you referring to? >> >> Thanks >> Javi >> >> On Wed Apr 3 14:32:40 2013, Daniel Russel wrote: >>> IMP::domino::load_merged_assignments has a whole bunch of extra parameters that we added as part of various (to my knowledge) failed experiments. I'd like to simplify it to >>> >>> IMPDOMINOEXPORT void load_merged_assignments(const Subset &first_subset, >>> AssignmentContainer* first, >>> const Subset &second_subset, >>> AssignmentContainer* second, >>> const SubsetFilterTablesTemp >>> &filters, >>> AssignmentContainer* ret); >>> >>> Let me know if any of the experiments were actually working for you and so should be kept around. Thanks. >>> _______________________________________________ >>> 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