BoostA2DE.cc

// ***************************************************************
// SPDX-FileCopyrightText: Copyright 2024 Ricardo Montañana Gómez
// SPDX-FileType: SOURCE
// SPDX-License-Identifier: MIT
// ***************************************************************
 
#include <limits.h>
#include <tuple>
#include <folding.hpp>
#include "BoostA2DE.h"
 
namespace bayesnet {
 
    BoostA2DE::BoostA2DE(bool predict_voting) : Boost(predict_voting)
    {
    }
    std::vector<int> BoostA2DE::initializeModels(const Smoothing_t smoothing)
    {
        torch::Tensor weights_ = torch::full({ m }, 1.0 / m, torch::kFloat64);
        std::vector<int> featuresSelected = featureSelection(weights_);
        if (featuresSelected.size() < 2) {
            notes.push_back("No features selected in initialization");
            status = ERROR;
            return std::vector<int>();
        }
        for (int i = 0; i < featuresSelected.size() - 1; i++) {
            for (int j = i + 1; j < featuresSelected.size(); j++) {
                auto parents = { featuresSelected[i], featuresSelected[j] };
                std::unique_ptr<Classifier> model = std::make_unique<SPnDE>(parents);
                model->fit(dataset, features, className, states, weights_, smoothing);
                models.push_back(std::move(model));
                significanceModels.push_back(1.0); // They will be updated later in trainModel
                n_models++;
            }
        }
        notes.push_back("Used features in initialization: " + std::to_string(featuresSelected.size()) + " of " + std::to_string(features.size()) + " with " + select_features_algorithm);
        return featuresSelected;
    }
    void BoostA2DE::trainModel(const torch::Tensor& weights, const Smoothing_t smoothing)
    {
        //
        // Logging setup
        //
        // loguru::set_thread_name("BoostA2DE");
        // loguru::g_stderr_verbosity = loguru::Verbosity_OFF;
        // loguru::add_file("boostA2DE.log", loguru::Truncate, loguru::Verbosity_MAX);
 
        // Algorithm based on the adaboost algorithm for classification
        // as explained in Ensemble methods (Zhi-Hua Zhou, 2012)
        fitted = true;
        double alpha_t = 0;
        torch::Tensor weights_ = torch::full({ m }, 1.0 / m, torch::kFloat64);
        bool finished = false;
        std::vector<int> featuresUsed;
        if (selectFeatures) {
            featuresUsed = initializeModels(smoothing);
            if (featuresUsed.size() == 0) {
                return;
            }
            auto ypred = predict(X_train);
            std::tie(weights_, alpha_t, finished) = update_weights(y_train, ypred, weights_);
            // Update significance of the models
            for (int i = 0; i < n_models; ++i) {
                significanceModels[i] = alpha_t;
            }
            if (finished) {
                return;
            }
        }
        int numItemsPack = 0; // The counter of the models inserted in the current pack
        // Variables to control the accuracy finish condition
        double priorAccuracy = 0.0;
        double improvement = 1.0;
        double convergence_threshold = 1e-4;
        int tolerance = 0; // number of times the accuracy is lower than the convergence_threshold
        // Step 0: Set the finish condition
        // epsilon sub t > 0.5 => inverse the weights policy
        // validation error is not decreasing
        // run out of features
        bool ascending = order_algorithm == Orders.ASC;
        std::mt19937 g{ 173 };
        std::vector<std::pair<int, int>> pairSelection;
        while (!finished) {
            // Step 1: Build ranking with mutual information
            pairSelection = metrics.SelectKPairs(weights_, featuresUsed, ascending, 0); // Get all the pairs sorted
            if (order_algorithm == Orders.RAND) {
                std::shuffle(pairSelection.begin(), pairSelection.end(), g);
            }
            int k = bisection ? pow(2, tolerance) : 1;
            int counter = 0; // The model counter of the current pack
            // VLOG_SCOPE_F(1, "counter=%d k=%d featureSelection.size: %zu", counter, k, featureSelection.size());
            while (counter++ < k && pairSelection.size() > 0) {
                auto feature_pair = pairSelection[0];
                pairSelection.erase(pairSelection.begin());
                std::unique_ptr<Classifier> model;
                model = std::make_unique<SPnDE>(std::vector<int>({ feature_pair.first, feature_pair.second }));
                model->fit(dataset, features, className, states, weights_, smoothing);
                alpha_t = 0.0;
                if (!block_update) {
                    auto ypred = model->predict(X_train);
                    // Step 3.1: Compute the classifier amout of say
                    std::tie(weights_, alpha_t, finished) = update_weights(y_train, ypred, weights_);
                }
                // Step 3.4: Store classifier and its accuracy to weigh its future vote
                numItemsPack++;
                models.push_back(std::move(model));
                significanceModels.push_back(alpha_t);
                n_models++;
                // VLOG_SCOPE_F(2, "numItemsPack: %d n_models: %d featuresUsed: %zu", numItemsPack, n_models, featuresUsed.size());
            }
            if (block_update) {
                std::tie(weights_, alpha_t, finished) = update_weights_block(k, y_train, weights_);
            }
            if (convergence && !finished) {
                auto y_val_predict = predict(X_test);
                double accuracy = (y_val_predict == y_test).sum().item<double>() / (double)y_test.size(0);
                if (priorAccuracy == 0) {
                    priorAccuracy = accuracy;
                } else {
                    improvement = accuracy - priorAccuracy;
                }
                if (improvement < convergence_threshold) {
                    // VLOG_SCOPE_F(3, "  (improvement<threshold) tolerance: %d numItemsPack: %d improvement: %f prior: %f current: %f", tolerance, numItemsPack, improvement, priorAccuracy, accuracy);
                    tolerance++;
                } else {
                    // VLOG_SCOPE_F(3, "* (improvement>=threshold) Reset. tolerance: %d numItemsPack: %d improvement: %f prior: %f current: %f", tolerance, numItemsPack, improvement, priorAccuracy, accuracy);
                    tolerance = 0; // Reset the counter if the model performs better
                    numItemsPack = 0;
                }
                if (convergence_best) {
                    // Keep the best accuracy until now as the prior accuracy
                    priorAccuracy = std::max(accuracy, priorAccuracy);
                } else {
                    // Keep the last accuray obtained as the prior accuracy
                    priorAccuracy = accuracy;
                }
            }
            // VLOG_SCOPE_F(1, "tolerance: %d featuresUsed.size: %zu features.size: %zu", tolerance, featuresUsed.size(), features.size());
            finished = finished || tolerance > maxTolerance || pairSelection.size() == 0;
        }
        if (tolerance > maxTolerance) {
            if (numItemsPack < n_models) {
                notes.push_back("Convergence threshold reached & " + std::to_string(numItemsPack) + " models eliminated");
                // VLOG_SCOPE_F(4, "Convergence threshold reached & %d models eliminated of %d", numItemsPack, n_models);
                for (int i = 0; i < numItemsPack; ++i) {
                    significanceModels.pop_back();
                    models.pop_back();
                    n_models--;
                }
            } else {
                notes.push_back("Convergence threshold reached & 0 models eliminated");
                // VLOG_SCOPE_F(4, "Convergence threshold reached & 0 models eliminated n_models=%d numItemsPack=%d", n_models, numItemsPack);
            }
        }
        if (pairSelection.size() > 0) {
            notes.push_back("Pairs not used in train: " + std::to_string(pairSelection.size()));
            status = WARNING;
        }
        notes.push_back("Number of models: " + std::to_string(n_models));
    }
    std::vector<std::string> BoostA2DE::graph(const std::string& title) const
    {
        return Ensemble::graph(title);
    }
}