search_exhaustive.hpp

Go to the documentation of this file.

#ifndef FSTSEARCHEXHAUSTIVE_H
#define FSTSEARCHEXHAUSTIVE_H
 /* 
=========================================================================
Copyright and license:
  * FST4 software (with the exception of any bundled or externally linked
    third-party components) is Copyright (C) 2000-present Institute of
    Information Theory and Automation of the Czech Academy of Sciences (UTIA).
  * FST4 is free software: you can redistribute it and/or modify it under
    the terms of the GNU Affero General Public License as published by the
    Free Software Foundation, version 3. The full text of the license is in
    the file LICENSE in the distribution root and at
    https://www.gnu.org/licenses/agpl-3.0.html.
  * FST4 is distributed in the hope that it will be useful, but WITHOUT ANY
    WARRANTY; without even the implied warranty of MERCHANTABILITY or
    FITNESS FOR A PARTICULAR PURPOSE. See the GNU Affero General Public
    License for more details.
  * Closed-source use without the AGPL's copyleft obligations requires a
    commercial license from UTIA; in case of interest write to
    utia@utia.cas.cz. LICENSING.md in the distribution root explains both
    options, the contributor license agreement that every contribution
    requires, and the contacts (technical questions and pull requests:
    somol@utia.cas.cz, with "FST" in the subject line).
========================================================================== */
 
#include <memory>
#include <cassert>
#include <iostream>
#include <sstream>
#include <ctime>
#include "error.hpp"
#include "global.hpp"
#include "stopwatch.hpp"
#include "search.hpp"
#include "result_tracker.hpp"
#include "candidate_evaluator.hpp"
#include "candidate_evaluator_sequential.hpp"
 
/*============== Template parameter type naming conventions ==============
--------- Numeric types: -------------------------------------------------
DATATYPE - data sample values - usually real numbers (but may be integers
          in text processing etc.)
REALTYPE - must be real numbers - for representing intermediate results of 
          calculations like mean, covariance etc.
IDXTYPE - index values for enumeration of data samples - (nonnegative) integers, 
          extent depends on numbers of samples in data
DIMTYPE - index values for enumeration of features (dimensions), or classes (not 
          class sizes) - (nonnegative) integers, usually lower extent than IDXTYPE, 
          but be aware of expressions like _classes*_features*_features ! 
          in linearized representations of feature matrices for all classes
BINTYPE - feature selection marker type - represents ca. <10 different feature 
          states (selected, deselected, sel./desel. temporarily 1st nested loop, 2nd...)
RETURNTYPE - criterion value: real value, but may be extended in future to support 
          multiple values 
--------- Class types: ---------------------------------------------------
SUBSET       - class of class type Subset 
CLASSIFIER   - class implementing interface defined in abstract class Classifier 
EVALUATOR    - class implementing interface defined in abstract class Sequential_Step 
DISTANCE     - class implementing interface defined in abstract class Distance 
DATAACCESSOR - class implementing interface defined in abstract class Data_Accessor 
INTERVALCONTAINER - class of class type Data_Intervaller 
CONTAINER    - STL container of class type Data_Interval  
========================================================================== */
 
namespace FST {

 
template<class RETURNTYPE, typename DIMTYPE, class SUBSET, class CRITERION>
class Search_Exhaustive : public Search<RETURNTYPE,DIMTYPE,SUBSET,CRITERION>{
public:
        typedef Search<RETURNTYPE,DIMTYPE,SUBSET,CRITERION> parent; 
        typedef std::shared_ptr<CRITERION> PCriterion; 
        typedef std::shared_ptr<SUBSET> PSubset; 
        typedef Candidate_Evaluator<RETURNTYPE,DIMTYPE,SUBSET,CRITERION> EVALUATORENGINE; 
        typedef std::shared_ptr<EVALUATORENGINE> PEvaluatorEngine; 
        explicit Search_Exhaustive(PEvaluatorEngine engine=PEvaluatorEngine()):Search<RETURNTYPE,DIMTYPE,SUBSET,CRITERION>(),_engine(engine) {
                if(!_engine) _engine.reset(new Candidate_Evaluator_Sequential<RETURNTYPE,DIMTYPE,SUBSET,CRITERION>());
                notify("Search_Exhaustive constructor.");
        }
        virtual ~Search_Exhaustive() {notify("Search_Exhaustive destructor.");}
 
        // NOTE: crit is expected to be already initialize()d before use here
        virtual bool search(const DIMTYPE target_d, RETURNTYPE &result, const PSubset sub, const PCriterion crit, std::ostream& os=std::cout) override;
 
        virtual std::ostream& print(std::ostream& os) const override {
                os << "Exhaustive search  [";
                if(_engine->is_threaded()) os << "Search_Exhaustive(threads="<<_engine->get_max_threads()<<")";
                else os << "Search_Exhaustive()";
                if(parent::result_tracker_active()) os << " with " << *parent::_tracker;
                os<<"]"; return os;
        }
protected:
        PEvaluatorEngine _engine; 
};
 
template<class RETURNTYPE, typename DIMTYPE, class SUBSET, class CRITERION>
bool Search_Exhaustive<RETURNTYPE,DIMTYPE,SUBSET,CRITERION>::search(const DIMTYPE target_d, RETURNTYPE &result, const PSubset sub, const PCriterion crit, std::ostream& os) // returns found subset + criterion value
{
        StopWatch swatch;
        notify("Search_Exhaustive::search().");
        assert(sub);
        assert(crit);
        assert(_engine);
        const DIMTYPE n=sub->get_n();
        assert(target_d<=n);
 
        bool track=parent::result_tracker_active(); if(track) parent::_tracker->set_output_detail(parent::get_output_detail());
 
        if(parent::output_normal()) {
                msgbuf sos;
                sos << "---------------------------------------" << std::endl;
                sos << "Starting " << *this << std::endl;
                sos << "with Criterion: " << *crit << std::endl;
                if(target_d==0) sos << "Subset size to be optimized." << std::endl; else sos << "Target subset size set to: " << target_d << std::endl;
                sos << std::endl << std::flush;
                this->report(os,sos);
        }
 
        // stream all candidate subsets of each considered cardinality to the engine
        DIMTYPE t_to=n;
        DIMTYPE cardinality=1;
        if(target_d>0) {cardinality=target_d; t_to=target_d;}
        bool generate_first=true;
        bool generator_failed=false; // set if the subset object cannot start enumerating candidates of the current cardinality at all (e.g. cardinality>n); such a run yields no result and search() returns false
        PSubset gen_sub(new SUBSET(n));
        typename EVALUATORENGINE::Generator generator=[&gen_sub,&cardinality,&generate_first,&generator_failed,t_to](SUBSET &cand)->bool {
                for(;;) {
                        if(generate_first) {
                                if(cardinality>t_to) return false;
                                gen_sub->deselect_all();
                                if(!gen_sub->getFirstCandidateSubset(cardinality,false/*reverse*/)) {generator_failed=true; return false;} // no candidate subsets can be traversed for whatever reason
                                generate_first=false;
                                cand.stateless_copy(*gen_sub);
                                return true;
                        }
                        if(gen_sub->getNextCandidateSubset()) {cand.stateless_copy(*gen_sub); return true;}
                        ++cardinality; // all candidates of current cardinality generated - move on
                        generate_first=true;
                }
        };
 
        typename EVALUATORENGINE::BetterPredicate better=[](const RETURNTYPE val, const PSubset&, const RETURNTYPE bestval, const PSubset&)->bool {return val>bestval;};
 
        typename EVALUATORENGINE::BestCallback on_best;
        if(parent::output_normal()) {
                StopWatch *psw=&swatch;
                std::ostream *pos=&os;
                // both engines fire this on the master thread in generation order (tidx==-1),
                // so threaded output matches sequential output
                on_best=[this,psw,pos](const int /*tidx*/, const unsigned long long /*ordinal*/, const RETURNTYPE val, const PSubset &bsub) {
                        msgbuf sos;
                        sos << "new MAXCRIT="<<val<<", " << *bsub << std::endl << *psw << std::endl << std::endl << std::flush;
                        this->report_solution(*pos,val,*bsub,sos);
                };
        }
 
        bool found=_engine->evaluate_stream(generator, crit, track? parent::_tracker : typename parent::PResultTracker(), better, on_best, result, sub);
        if(generator_failed || !found) return false;
 
        if(parent::output_normal()) {
                msgbuf sos;
                sos << "......................................." << std::endl;
                if(_engine->is_threaded()) sos << "Search_Exhaustive_Threaded() search finished. " << swatch << std::endl;
                else sos << "Search_Exhaustive() search finished. " << swatch << std::endl;
                sos << "Search result: "<< std::endl << *sub << std::endl << "Criterion value: " << result << std::endl << std::endl << std::flush;
                this->report(os,sos);
        }
        return true;
}
 
} // namespace
#endif // FSTSEARCHEXHAUSTIVE_H ///:~