saberbud/q3DMASC
1
0
Fork 0
mirror of https://github.com/dgirardeau/q3DMASC.git synced 2026-08-30 09:00:50 +08:00
Code Issues Packages Projects Releases Wiki Activity

confusion matrix added to the training step

Browse Source 
Add the confusion matrix to the training step
Delete confusion matrices properly after the training step
This commit is contained in:
Paul Leroy
2023-01-18 23:57:59 +01:00
parent 83e08660ee
commit f104fd5b17
6 changed files with 30 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files
q3DMASCClassifier.cpp
+10 -2
View File
@@ -38,6 +38,7 @@
#include <QProgressDialog>
#include <QtConcurrent>
#include "qTrain3DMASCDialog.h"
#include "confusionmatrix.h"
using namespace masc;
@@ -274,7 +275,7 @@ bool Classifier::classify( const Feature::Source::Set& featureSources,
QCoreApplication::processEvents();
}
ConfusionMatrix *confusionMatrix = new ConfusionMatrix(*classifSFBackup, *classificationSF);
ConfusionMatrix *confusionMatrix = new ConfusionMatrix(*classifSFBackup, *classificationSF, parentWidget);
return success;
}
@@ -283,6 +284,7 @@ bool Classifier::evaluate(const Feature::Source::Set& featureSources,
ccPointCloud* testCloud,
AccuracyMetrics& metrics,
QString& errorMessage,
Train3DMASCDialog& train3DMASCDialog,
CCCoreLib::ReferenceCloud* testSubset/*=nullptr=*/,
QString outputSFName/*=QString()*/,
QWidget* parentWidget/*=nullptr*/)
@@ -394,7 +396,10 @@ bool Classifier::evaluate(const Feature::Source::Set& featureSources,
}
}
//estimate the efficiency of the classifier
std::vector<ScalarType> actualClass(testSampleCount);
std::vector<ScalarType> predictectedClass(testSampleCount);
{
metrics.sampleCount = testSampleCount;
metrics.goodGuess = 0;
@@ -412,6 +417,8 @@ bool Classifier::evaluate(const Feature::Source::Set& featureSources,
float fPredictedClass = m_rtrees->predict(test_data.row(i), cv::noArray(), cv::ml::DTrees::PREDICT_MAX_VOTE);
int iPredictedClass = static_cast<int>(fPredictedClass);
actualClass.at(i) = iClass;
predictectedClass.at(i) = iPredictedClass;
if (iPredictedClass == iClass)
{
++metrics.goodGuess;
@@ -434,7 +441,8 @@ bool Classifier::evaluate(const Feature::Source::Set& featureSources,
metrics.ratio = static_cast<float>(metrics.goodGuess) / metrics.sampleCount;
}
ConfusionMatrix *confusionMatrix = new ConfusionMatrix(*classifSF, *outputSF);
std::unique_ptr<ConfusionMatrix> confusionMatrix(new ConfusionMatrix(actualClass, predictectedClass));
train3DMASCDialog.deleteLaterConfusionMatrix(confusionMatrix);
return true;
}