void GeneticAlgorithm::initPopulation() {
// #ifdef PARALLEL
// int argc = 0;
// char **argv = NULL;
// int myid; // id tego komputera
// int numprocs; // ilosc procesorow
// int namelen; // nazwa konkretnego procesora
// char processor_name[MPI_MAX_PROCESSOR_NAME];
// MPI_Init(&argc, &argv); // uzyskanie dostepu do srodowiska mpi
// MPI_Comm_size(MPI_COMM_WORLD, &numprocs); // definiujemy ilosc procesorow
// MPI_Comm_rank(MPI_COMM_WORLD, &myid); // definiujemy id tego procesora
// MPI_Get_processor_name(processor_name, &namelen); // pobieramy nazwe proc.
// if(myid == 0) {
// #endif
std::vector<Individual>::iterator it;
for(it = _population.begin(); it != _population.end(); ++it) {
it->init();
it->calculateFitness(_cp);
}
calculateStatistics();
// #ifdef PARALLEL
// }
// MPI_Finalize(); // obowiazkowe zamkniecie mpi
// #endif
}
Caller::Caller( const double poissonLambda, const int minDepth, const double leftStrandBias, const double rightStrandBias, const double readEndFraction, const int qCutoff, const char* tumorDirectoryPath, const char* outputDirectoryPath, const int usePoissonGermline, const int disableLvl5Filter)
{
this->poissonLambda = poissonLambda;
this->minDepth = minDepth;
this->strandBiasLeft = leftStrandBias;
this->strandBiasRight = rightStrandBias;
this->minQScore = qCutoff;
this->readEndFraction = readEndFraction;
this->usePoissonGermline = usePoissonGermline;
this->disableLvl5Filter = disableLvl5Filter;
( this->tumorDirectoryPath).assign( tumorDirectoryPath);
( this->outputDirectoryPath).assign( outputDirectoryPath);
// Setup output files
( this->outputPaths).push_back( this->outputDirectoryPath + "/calls_level1.sinvict");
( this->outputPaths).push_back( this->outputDirectoryPath + "/calls_level2.sinvict");
( this->outputPaths).push_back( this->outputDirectoryPath + "/calls_level3.sinvict");
( this->outputPaths).push_back( this->outputDirectoryPath + "/calls_level4.sinvict");
( this->outputPaths).push_back( this->outputDirectoryPath + "/calls_level5.sinvict");
( this->outputPaths).push_back( this->outputDirectoryPath + "/calls_level6.sinvict");
Common::getFilesInDir( this->tumorDirectoryPath, tumorPaths);
std::sort( tumorPaths.begin(), tumorPaths.end());
for( int i = 0; i < tumorPaths.size(); i++)
{
std::string nextTumorSamplePath = this->tumorDirectoryPath + "/" + tumorPaths[i];
loadEntries( nextTumorSamplePath);
}
calculateStatistics();
}
LGraph DTALikelihood::calculateStatisticsDirect(Particles &particles)
{
QVector<QVector3D> qp;
particles.appendToQVector3DList(qp);
LGraph graph;
calculateStatistics(qp, graph);
return graph;
}
void GeneticAlgorithm::nextGeneration() {
std::vector<Individual>::iterator it;
for(it = _population.begin(); it != _population.end(); ++it) {
it->_genotype.mutate(_mutationProbability);
it->calculateFitness(_cp);
}
calculateStatistics();
}
void ElevationProfileFloatItem::toggleZoomToViewport()
{
m_zoomToViewport = ! m_zoomToViewport;
calculateStatistics( m_eleData );
if ( ! m_zoomToViewport ) {
m_axisX.setRange( m_eleData.first().x(), m_eleData.last().x() );
m_axisY.setRange( qMin( m_minElevation, qreal( 0.0 ) ), m_maxElevation );
}
readSettings();
emit settingsChanged( nameId() );
}
void ElevationProfileFloatItem::handleDataUpdate(const GeoDataLineString &points, QList<QPointF> eleData)
{
m_eleData = eleData;
m_points = points;
calculateStatistics( m_eleData );
if ( m_eleData.length() >= 2 ) {
m_axisX.setRange( m_eleData.first().x(), m_eleData.last().x() );
m_axisY.setRange( qMin( m_minElevation, qreal( 0.0 ) ), m_maxElevation );
}
emit dataUpdated();
}
void ElevationProfileFloatItem::updateVisiblePoints()
{
if ( ! m_activeDataSource->isDataAvailable() || m_points.size() < 2 ) {
return;
}
// find the longest visible route section on screen
QList<QList<int> > routeSegments;
QList<int> currentRouteSegment;
for ( int i = 0; i < m_eleData.count(); i++ ) {
qreal lon = m_points[i].longitude(GeoDataCoordinates::Degree);
qreal lat = m_points[i].latitude (GeoDataCoordinates::Degree);
qreal x = 0;
qreal y = 0;
if ( m_marbleWidget->screenCoordinates(lon, lat, x, y) ) {
// on screen --> add point to list
currentRouteSegment.append(i);
} else {
// off screen --> start new list
if ( !currentRouteSegment.isEmpty() ) {
routeSegments.append( currentRouteSegment );
currentRouteSegment.clear();
}
}
}
routeSegments.append( currentRouteSegment ); // in case the route ends on screen
int maxLenght = 0;
foreach ( const QList<int> ¤tRouteSegment, routeSegments ) {
if ( currentRouteSegment.size() > maxLenght ) {
maxLenght = currentRouteSegment.size() ;
m_firstVisiblePoint = currentRouteSegment.first();
m_lastVisiblePoint = currentRouteSegment.last();
}
}
if ( m_firstVisiblePoint < 0 ) {
m_firstVisiblePoint = 0;
}
if ( m_lastVisiblePoint < 0 || m_lastVisiblePoint >= m_eleData.count() ) {
m_lastVisiblePoint = m_eleData.count() - 1;
}
// include setting range to statistics and test for m_zoomToViewport in calculateStatistics();
if ( m_zoomToViewport ) {
calculateStatistics( m_eleData );
m_axisX.setRange( m_eleData.value( m_firstVisiblePoint ).x(),
m_eleData.value( m_lastVisiblePoint ).x() );
m_axisY.setRange( m_minElevation, m_maxElevation );
}
return;
}
void DTALikelihood::calculateModel(Model *model)
{
m_modelParticles.clear();
m_originalParticles->calculateBoundingBox();
model->start();
for (Particle* pos : m_originalParticles->getParticles()) {
if (!model->isInVoid(pos->getPos())) {
m_modelParticles.append(pos->getPos());
}
}
model->stop();
calculateStatistics(m_modelParticles,m_modelData);
}
const SimulationResults& Simulation::run() {
//Stat* statistics = new Stat();
//statistics->setVerbose(true); // FIXME: remove!
//statistics->setSchedulerId(scheduler->getId());
LOG(0) << "######## Running Simulation with scheduler " << scheduler->getId();
LOG(0) << "\tAim are " << steps << " iterations";
initCounters();
for (size_t taskNum = 0; taskNum< (*taskset).size(); taskNum++) {
(*taskset)[taskNum]->start(0);
}
unsigned int time = 0;
bool schedSuccess = true;
for (time = 0; time < steps; time++) {
LOG(1) << "\n\nT : " << time;
doActivations(time, scheduler);
schedSuccess = doExecutions(time, scheduler/*, statistics*/);
if (!schedSuccess) {
LOG(0) << "Executions failed in regular time step " << time;
break;
}
}
if (schedSuccess) {
while (scheduler->hasPendingJobs()) {
bool bdisp = doExecutions(time, scheduler/*, statistics*/);
++time;
if (!bdisp) {
LOG(0) << "Executions failed in during cleanup in step " << time;
break;
}
}
}
LOG(0) << "Totally simulated time: " << time;
printExecStats(scheduler->getId());
printDelayCounters();
// Save the statistics in a other object
//storeStatistics(statistics);
calculateStatistics();
stats.simulatedTime = time;
stats.success = schedSuccess;
return stats;
}
bool MuSlashRhoLambdaES::startEvolutionRun( int mu,
int rho,
int lambda,
ESInitialize *init)
{
deletePopulation();
// set information struct to values
mInformation.currentGeneration = 0;
mInformation.mu = mu;
mInformation.rho = rho;
mInformation.lambda = lambda;
mInformation.numberOfLastSuccessfulIndividuals = 0;
if(mInformation.mu <= 0 || mInformation.rho <= 0 || mInformation.lambda <= 0){
return false;
}
mPopulation = init->createPopulation(mInformation.mu);
if(mPopulation.size() <= 0){
return false;
}
mInformation.numberOfStrategyParameters = mPopulation.first()->getStrategyParameters().size();
mInformation.numberOfObjectParameters = mPopulation.first()->getObjectParameters().size();
//calculate fitness values of population
bool ok = mFitnessEvaluation->doEvaluation(mPopulation,
mInformation);
if(ok == false){
return false;
}
qSort(mPopulation.begin(),
mPopulation.end(),
MuSlashRhoLambdaES::descendingOrder);
calculateStatistics();
return true;
}
std::vector<Genome> GenAlg::new_generation(std::vector<Genome> &last_generation)
{
generation_count++;
reset();
std::sort(last_generation.begin(), last_generation.end(), myComparer);
calculateStatistics();
std::vector<Genome> new_population;
//get elites
GetBest(my_params.num_elite, my_params.num_copies_of_elite, last_generation, new_population);
//genetic algorithm loop
while(new_population.size() < pop_size)
{
Genome mother = getRandomChromosome();
Genome father = getRandomChromosome();
std::vector<float> child1, child2;
crossover(mother.weights, father.weights, child1, child2);
mutate(child1);
mutate(child2);
new_population.push_back(Genome(child1, 0));
new_population.push_back(Genome(child2, 0));
}
population = new_population;
return population;
}
double ClassifierTester::StartTest(const LandmarkCollectionDataPtr& collection,
const std::string& savedTrainedDataFilePath,
const int numberOfIterations,
bool showSteps)
{
BOOST_ASSERT(savedTrainedDataFilePath.size() != 0);
BOOST_ASSERT(collection->CollectionSize() != 0);
m_collection = collection;
m_savedTrainedDataFilePath = savedTrainedDataFilePath;
m_showSteps = showSteps;
m_numberOfIterations = numberOfIterations;
m_detector = DetectorPtr(new Detector(m_savedTrainedDataFilePath));
int i = 0;
collection->EnumerateConstColectionWithCallback([&] (const ImageLandmarkDataPtr& landmarkData, const int index, bool* stop) {
// std::cout<<index<<std::endl;
const cv::Mat image = landmarkData->ImageSource();
const cv::Mat landmarks = processImage(image, landmarkData->LandmarksMat());
if (landmarks.empty())
{
i++;
// std::cout<<"Fail image "<<landmarkData->ImagePath()<<std::endl;
}
else
{
collectStatistics(landmarks, landmarkData->LandmarksMat());
}
});
std::cout<<"Numeber of fail images: "<<i<<std::endl;
return calculateStatistics();
}
std::shared_ptr<te::mem::DataSet> terrama2::services::analysis::core::createAggregationBuffer(
std::vector<uint32_t>& indexes, std::shared_ptr<ContextDataSeries> contextDataSeries, Buffer buffer,
StatisticOperation aggregationStatisticOperation,
const std::string& attribute)
{
std::shared_ptr<te::mem::DataSet> dsOut;
if(indexes.empty())
return dsOut;
// Creates memory dataset for buffer
te::da::DataSetType* dt = new te::da::DataSetType("buffer");
auto syncDs = contextDataSeries->series.syncDataSet;
auto sampleGeom = syncDs->getGeometry(0, contextDataSeries->geometryPos);
int geomSampleSrid = sampleGeom->getSRID();
te::gm::GeometryProperty* prop = new te::gm::GeometryProperty("geom", 0, te::gm::MultiPolygonType, true);
prop->setSRID(geomSampleSrid);
dt->add(prop);
te::dt::SimpleProperty* prop02 = new te::dt::SimpleProperty("attribute", te::dt::DOUBLE_TYPE, true);
dt->add(prop02);
dsOut.reset(new te::mem::DataSet(dt));
std::shared_ptr<te::gm::Envelope> box(syncDs->getExtent(contextDataSeries->geometryPos));
if(buffer.unit.empty())
buffer.unit = "m";
// Inserts each geometry in the rtree, if there is a conflict, it makes the union of the two geometries
te::sam::rtree::Index<OccurrenceAggregation*, 4> rtree;
for(size_t i = 0; i < indexes.size(); ++i)
{
auto geom = syncDs->getGeometry(indexes[i], contextDataSeries->geometryPos);
double distance = terrama2::core::convertDistanceUnit(buffer.distance, buffer.unit, "METER");
std::unique_ptr<te::gm::Geometry> tempGeom(dynamic_cast<te::gm::Geometry*>(geom.get()->clone()));
if(!tempGeom)
{
QString errMsg(QObject::tr("Invalid geometry in dataset: ").arg(contextDataSeries->series.dataSet->id));
throw terrama2::InvalidArgumentException() << ErrorDescription(errMsg);
}
int utmSrid = terrama2::core::getUTMSrid(tempGeom.get());
// Converts to UTM in order to create buffer in meters
if(tempGeom->getSRID() != utmSrid)
{
tempGeom->transform(utmSrid);
}
std::unique_ptr<te::gm::Geometry> aggBuffer(tempGeom->buffer(distance, 16, te::gm::CapButtType));
aggBuffer->setSRID(utmSrid);
// Converts buffer to DataSet SRID in order to compare with the occurrences in the rtree
aggBuffer->transform(geomSampleSrid);
std::vector<OccurrenceAggregation*> vec;
bool aggregated = false;
// Search for occurrences in the same area
rtree.search(*(aggBuffer->getMBR()), vec);
for(std::size_t t = 0; t < vec.size(); ++t)
{
OccurrenceAggregation* occurrenceAggregation = vec[t];
// If the an intersection is found, makes the union of the two geometries and mark the index.
if(aggBuffer->intersects(occurrenceAggregation->buffer.get()))
{
rtree.remove(*(occurrenceAggregation->buffer->getMBR()), occurrenceAggregation);
occurrenceAggregation->buffer.reset(aggBuffer->Union(occurrenceAggregation->buffer.get()));
occurrenceAggregation->indexes.push_back(i);
rtree.insert(*(occurrenceAggregation->buffer->getMBR()), occurrenceAggregation);
aggregated = true;
}
}
if(!aggregated)
{
OccurrenceAggregation* occurrenceAggregation = new OccurrenceAggregation();
occurrenceAggregation->buffer.reset(aggBuffer.release());
occurrenceAggregation->indexes.push_back(i);
rtree.insert(*(occurrenceAggregation->buffer->getMBR()), occurrenceAggregation);
}
}
// Fills the memory dataset with the geometries
std::vector<OccurrenceAggregation*> occurrenceAggVec;
rtree.search(*(box.get()), occurrenceAggVec);
int attributeType = -1;
if(aggregationStatisticOperation != StatisticOperation::COUNT)
{
auto property = contextDataSeries->series.teDataSetType->getProperty(attribute);
if(!property)
{
QString errMsg(QObject::tr("Invalid attribute: %1").arg(QString::fromStdString(attribute)));
throw terrama2::InvalidArgumentException() << ErrorDescription(errMsg);
}
attributeType = property->getType();
}
for(size_t i = 0; i < occurrenceAggVec.size(); i++)
{
OccurrenceAggregation* occurrenceAggregation = occurrenceAggVec[i];
OperatorCache cache;
std::vector<double> values;
values.reserve(occurrenceAggregation->indexes.size());
for(unsigned int j = 0; j < occurrenceAggregation->indexes.size(); ++j)
{
cache.count++;
if(aggregationStatisticOperation != StatisticOperation::COUNT)
{
if(attribute.empty())
{
QString errMsg(QObject::tr("Invalid attribute"));
throw terrama2::InvalidArgumentException() << ErrorDescription(errMsg);
}
double value = getValue(syncDs, attribute, occurrenceAggregation->indexes[j], attributeType);
values.push_back(value);
cache.sum += value;
if(value > cache.max)
cache.max = value;
if(value < cache.min)
cache.min = value;
}
}
calculateStatistics(values, cache);
auto item = new te::mem::DataSetItem(dsOut.get());
item->setGeometry(0, dynamic_cast<te::gm::Geometry*>(occurrenceAggregation->buffer->clone()));
item->setDouble(1, getOperationResult(cache, aggregationStatisticOperation));
dsOut->add(item);
}
return dsOut;
}
double
iAIDA::AIDA_Histogram_native::AIDA_Histogram2D::equivalentBinEntries() const
{
calculateStatistics();
return m_ebe;
}
double
iAIDA::AIDA_Histogram_native::AIDA_Histogram2D::sumExtraBinHeights() const
{
calculateStatistics();
return m_sumExtraBinHeights;
}
void ExperimentReport::generateLatexTables() {
latexDirectory_ = experimentBaseDirectory_ + "/" + latexDirectory_;
cout << "latex directory: " << latexDirectory_ << endl;
vector<double> *** data = new vector<double>**[indicatorList_.size()];
for (int indicator = 0; indicator < indicatorList_.size(); indicator++) {
// A data vector per problem
data[indicator] = new vector<double>*[problemList_.size()];
for (int problem = 0; problem < problemList_.size(); problem++) {
data[indicator][problem] = new vector<double>[algorithmNameList_.size()];
for (int algorithm = 0; algorithm < algorithmNameList_.size(); algorithm++) {
string directory = experimentBaseDirectory_;
directory += "/data/";
directory += "/" + algorithmNameList_[algorithm];
directory += "/" + problemList_[problem];
directory += "/" + indicatorList_[indicator];
// Read values from data files
std::ifstream in(directory.c_str());
if( !in ) {
cout << "Error trying to read quality indicator file: " << directory << endl;
exit(-1);
} // if
string aux;
while( getline(in, aux ) ) {
data[indicator][problem][algorithm].push_back(atof(aux.c_str()));
} // while
} // for
} // for
} // for
double *** mean;
double *** median;
double *** stdDeviation;
double *** iqr;
double *** max;
double *** min;
int *** numberOfValues;
map<string, double> statValues;
statValues["mean"] = 0.0;
statValues["median"] = 0.0;
statValues["stdDeviation"] = 0.0;
statValues["iqr"] = 0.0;
statValues["max"] = 0.0;
statValues["min"] = 0.0;
mean = new double**[indicatorList_.size()];
median = new double**[indicatorList_.size()];
stdDeviation = new double**[indicatorList_.size()];
iqr = new double**[indicatorList_.size()];
min = new double**[indicatorList_.size()];
max = new double**[indicatorList_.size()];
numberOfValues = new int**[indicatorList_.size()];
for (int indicator = 0; indicator < indicatorList_.size(); indicator++) {
// A data vector per problem
mean[indicator] = new double*[problemList_.size()];
median[indicator] = new double*[problemList_.size()];
stdDeviation[indicator] = new double*[problemList_.size()];
iqr[indicator] = new double*[problemList_.size()];
min[indicator] = new double*[problemList_.size()];
max[indicator] = new double*[problemList_.size()];
numberOfValues[indicator] = new int*[problemList_.size()];
for (int problem = 0; problem < problemList_.size(); problem++) {
mean[indicator][problem] = new double[algorithmNameList_.size()];
median[indicator][problem] = new double[algorithmNameList_.size()];
stdDeviation[indicator][problem] = new double[algorithmNameList_.size()];
iqr[indicator][problem] = new double[algorithmNameList_.size()];
min[indicator][problem] = new double[algorithmNameList_.size()];
max[indicator][problem] = new double[algorithmNameList_.size()];
numberOfValues[indicator][problem] = new int[algorithmNameList_.size()];
for (int algorithm = 0; algorithm < algorithmNameList_.size(); algorithm++) {
sort(data[indicator][problem][algorithm].begin(),
data[indicator][problem][algorithm].end());
string directory = experimentBaseDirectory_;
directory += "/" + algorithmNameList_[algorithm];
directory += "/" + problemList_[problem];
directory += "/" + indicatorList_[indicator];
//cout << "----" << directory << "-----" << endl;
//calculateStatistics(data[indicator][problem][algorithm], meanV, medianV, minV, maxV, stdDeviationV, iqrV);
calculateStatistics(data[indicator][problem][algorithm], &statValues);
/*
cout << "Mean: " << statValues["mean"] << endl;
cout << "Median : " << statValues["median"] << endl;
cout << "Std : " << statValues["stdDeviation"] << endl;
cout << "IQR : " << statValues["iqr"] << endl;
cout << "Min : " << statValues["min"] << endl;
cout << "Max : " << statValues["max"] << endl;
cout << "N_values: " << data[indicator][problem][algorithm].size() << endl;
*/
mean[indicator][problem][algorithm] = statValues["mean"];
median[indicator][problem][algorithm] = statValues["median"];
stdDeviation[indicator][problem][algorithm] = statValues["stdDeviation"];
iqr[indicator][problem][algorithm] = statValues["iqr"];
min[indicator][problem][algorithm] = statValues["min"];
max[indicator][problem][algorithm] = statValues["max"];
numberOfValues[indicator][problem][algorithm] = data[indicator][problem][algorithm].size();
}
}
}
if (FileUtils::existsPath(latexDirectory_.c_str()) != 1) {
if (FileUtils::createDirectory(latexDirectory_) == -1) {
cout << "Error creating directory: " << latexDirectory_ << endl;
exit(-1);
} // if
cout << "Creating " << latexDirectory_ << " directory" << endl;
} // if
cout << "Experiment name: " << experimentName_ << endl;
string latexFile = latexDirectory_ + "/" + experimentName_ + ".tex";
printHeaderLatexCommands(latexFile);
for (int i = 0; i < indicatorList_.size(); i++) {
printMeanStdDev(latexFile, i, mean, stdDeviation);
printMedianIQR(latexFile, i, median, iqr);
} // for
printEndLatexCommands(latexFile);
// Free memory:
for (int indicator = 0; indicator < indicatorList_.size(); indicator++) {
for (int problem = 0; problem < problemList_.size(); problem++) {
delete [] data[indicator][problem];
delete [] mean[indicator][problem];
delete [] median[indicator][problem];
delete [] stdDeviation[indicator][problem];
delete [] iqr[indicator][problem];
delete [] max[indicator][problem];
delete [] min[indicator][problem];
delete [] numberOfValues[indicator][problem];
}
delete [] data[indicator];
delete [] mean[indicator];
delete [] median[indicator];
delete [] stdDeviation[indicator];
delete [] iqr[indicator];
delete [] max[indicator];
delete [] min[indicator];
delete [] numberOfValues[indicator];
}
delete [] data;
delete [] mean;
delete [] median;
delete [] stdDeviation;
delete [] iqr;
delete [] max;
delete [] min;
delete [] numberOfValues;
} // generateLatexTables
double
iAIDA::AIDA_Histogram_native::AIDA_Histogram2D::maxBinHeight() const
{
calculateStatistics();
return m_maxHeight;
}
double
iAIDA::AIDA_Histogram_native::AIDA_Histogram2D::meanX() const
{
calculateStatistics();
return m_meanX;
}
int main() {
bool endSim = false;
Randomize();
while (true) {
double invalidElectResult = 0;
int voters;
double spread;
double votingError;
int trialRuns;
Vector<int> elections;
while (true) {
cout << "Enter number of voters (or 0 to end simulation): ";
voters = GetInteger();
if (voters == 0) {
endSim = true;
break;
}
else if (voters < 0)
cout << "Enter a positive integer." << endl;
else
break;
}
if (endSim) break;
while (true) {
cout << "Enter percentage spread between candidates (0 - 1.0): ";
spread = GetReal();
if (spread >=0 && spread <= 1.0)
break;
else
cout << "Spread must be between 0 and 1.0" << endl;
}
while (true) {
cout << "Enter voting error percentage: (0 - 1.0): ";
votingError = GetReal();
if (votingError >=0 && votingError <= 1.0)
break;
else
cout << "Voting error must be between 0 and 1.0" << endl;
}
while (true) {
cout << "Enter number of times to run the trial: ";
trialRuns = GetInteger();
if (trialRuns >= 1)
break;
else
cout << "Enter a positive integer" << endl;
}
for (int i = 0; i < trialRuns; i++) {
invalidElectResult = RunElectionTrial(voters, spread, votingError);
cout << "Trial " << i + 1 << ": ";
cout << "Chance of an invalid election result after 500 trials = " << invalidElectResult << "%" << endl;
elections.add(invalidElectResult);
}
scoreStatsT scores = calculateStatistics(elections);
cout << "Scores: MIN = " << scores.min << endl;
cout << "Scores: MAX = " << scores.max << endl;
cout << "Scores: AVG = " << scores.avg << endl;
cout << "Scores: NUM = " << scores.num << endl;
cout << "********************************************************" << endl << endl;
}
return 0;
}
double
iAIDA::AIDA_Histogram_native::AIDA_Histogram2D::rmsY() const
{
calculateStatistics();
return m_rmsY;
}
bool MuSlashRhoLambdaES::doNextGeneration()
{
/*
QFile file ("evolution.log");
file.open(QIODevice::WriteOnly | QIODevice::Append);
QTextStream out(&file);
out << "\n\n--------------------------\nNew generation:\n";
*/
bool ok = true;
QList<ESIndividual*> offsprings;
QList<double> bestParentsFitness;
// generate offsprings
for(int i = 0; i < mInformation.lambda; i++)
{
// select parents which should be used to generate current offspring
QList<ESIndividual*> curParents = mMarriage->doMarriage(getPopulation(),
mInformation.rho,
mInformation);
// check if an error occured
if(curParents.size() <= 0){
return false;
}
// get fitness of the best parent
qSort(curParents.begin(),
curParents.end(),
MuSlashRhoLambdaES::descendingOrder);
bestParentsFitness.append(curParents.first()->getFitness());
/*
out << "\tParents:\n";
for(int parentIndex = 0; parentIndex < curParents.size(); parentIndex++)
{
out << "\t\t" << curParents.at(parentIndex)->toString() << "\n";
}
*/
// recombine offspring strategy parameters from parents
QList<OptimizationDouble> recombinedStratParas
= mStrategyParametersRecombination->doRecombination(
curParents,
ESRecombination::STRATEGY,
mInformation);
// check if an error occured
if(recombinedStratParas.size() <= 0){
return false;
}
// recombine offspring object parameters from parents
QList<OptimizationDouble> recombinedObjParas
= mObjectParametersRecombination->doRecombination(
curParents,
ESRecombination::OBJECT,
mInformation);
// check if an error occured
if(recombinedObjParas.size() <= 0){
return false;
}
// mutate the offspring strategy parameters
QList<OptimizationDouble> mutatedStratParas = mStrategyParametersMutation->doMutation(
recombinedStratParas,
mInformation);
// check if an error occured
if(mutatedStratParas.size() <= 0){
return false;
}
// mutate the offspring object parameters with help of is mutated strategy parameters
QList<OptimizationDouble> mutatedObjParas = mObjectParametersMutation->doMutation(
recombinedObjParas,
mutatedStratParas,
mInformation);
// check if an error occured
if(mutatedObjParas.size() <= 0){
return false;
}
// create new Individual and add it to offspring list
offsprings.append( new ESIndividual(mutatedObjParas, mutatedStratParas, 0.0) );
/*
out << "\t--> Offspring:\n";
out << "\t\t" << offsprings.last()->toString();
out << "\n\n";
*/
}
//calculate fitness values of offsprings
ok &= mFitnessEvaluation->doEvaluation(offsprings,
mInformation);
if(ok == false){
return false;
}
// calculate sum of offsprings which have a better fitness then their best parent
mInformation.numberOfLastSuccessfulIndividuals = 0;
for(int i = 0; i < offsprings.size(); i++)
{
if(offsprings.at(i)->getFitness() > bestParentsFitness.at(i)){
mInformation.numberOfLastSuccessfulIndividuals++;
}
}
bestParentsFitness.clear();
// sort offspring accordingly their fitness
qSort(offsprings.begin(),
offsprings.end(),
MuSlashRhoLambdaES::descendingOrder);
// get list with new population
QList<ESIndividual*> selectionList = mSelection->doSelection(mPopulation,
offsprings,
mInformation.mu,
mInformation);
// check if an error occured
if(selectionList.size() <= 0){
return false;
}
// delete old parents which are not in the selection
deletePopulation(selectionList);
// use selection as new population
mPopulation = selectionList;
// update statisticvalues in mInformation
calculateStatistics();
mInformation.currentGeneration++;
/*
file.flush();
file.close();
*/
return ok;
}
const Column::ColumnStatistics& Column::statistics() {
if (!statisticsAvailable())
calculateStatistics();
return m_column_private->statistics;
}
int
iAIDA::AIDA_Histogram_native::AIDA_Histogram2D::extraEntries() const
{
calculateStatistics();
return m_extraEntries;
}
