50 lines
1.9 KiB
C++
50 lines
1.9 KiB
C++
// ***************************************************************
|
|
// SPDX-FileCopyrightText: Copyright 2024 Ricardo Montañana Gómez
|
|
// SPDX-FileType: SOURCE
|
|
// SPDX-License-Identifier: MIT
|
|
// ***************************************************************
|
|
|
|
#include "bayesnet/utils/bayesnetUtils.h"
|
|
#include "FCBF.h"
|
|
namespace bayesnet {
|
|
|
|
FCBF::FCBF(const torch::Tensor& samples, const std::vector<std::string>& features, const std::string& className, const int maxFeatures, const int classNumStates, const torch::Tensor& weights, const double threshold) :
|
|
FeatureSelect(samples, features, className, maxFeatures, classNumStates, weights), threshold(threshold)
|
|
{
|
|
if (threshold < 1e-7) {
|
|
throw std::invalid_argument("Threshold cannot be less than 1e-7");
|
|
}
|
|
}
|
|
void FCBF::fit()
|
|
{
|
|
initialize();
|
|
computeSuLabels();
|
|
auto featureOrder = argsort(suLabels); // sort descending order
|
|
auto featureOrderCopy = featureOrder;
|
|
for (const auto& feature : featureOrder) {
|
|
// Don't self compare
|
|
featureOrderCopy.erase(featureOrderCopy.begin());
|
|
if (suLabels.at(feature) == 0.0) {
|
|
// The feature has been removed from the list
|
|
continue;
|
|
}
|
|
if (suLabels.at(feature) < threshold) {
|
|
break;
|
|
}
|
|
// Remove redundant features
|
|
for (const auto& featureCopy : featureOrderCopy) {
|
|
double value = computeSuFeatures(feature, featureCopy);
|
|
if (value >= suLabels.at(featureCopy)) {
|
|
// Remove feature from list
|
|
suLabels[featureCopy] = 0.0;
|
|
}
|
|
}
|
|
selectedFeatures.push_back(feature);
|
|
selectedScores.push_back(suLabels[feature]);
|
|
if (selectedFeatures.size() == maxFeatures) {
|
|
break;
|
|
}
|
|
}
|
|
fitted = true;
|
|
}
|
|
} |