GridFire/src/network/lib/engine/engine_graph.cpp

#include "gridfire/engine/engine_graph.h"
#include "gridfire/reaction/reaction.h"
#include "gridfire/network.h"
#include "gridfire/screening/screening_types.h"
#include "gridfire/engine/procedures/priming.h"
#include "gridfire/partition/partition_ground.h"
#include "gridfire/engine/procedures/construction.h"

#include "fourdst/composition/species.h"
#include "fourdst/composition/atomicSpecies.h"

#include "quill/LogMacros.h"

#include <cstdint>
#include <iostream>
#include <set>
#include <stdexcept>
#include <string>
#include <string_view>
#include <unordered_map>
#include <utility>
#include <vector>
#include <fstream>
#include <ranges>

#include <boost/numeric/odeint.hpp>

#include "cppad/cppad.hpp"
#include "cppad/utility/sparse_rc.hpp"
#include "cppad/utility/sparse_rcv.hpp"


namespace gridfire {
    GraphEngine::GraphEngine(
        const fourdst::composition::Composition &composition,
        const BuildDepthType buildDepth
    ): GraphEngine(composition, partition::GroundStatePartitionFunction(), buildDepth) {}

    GraphEngine::GraphEngine(
        const fourdst::composition::Composition &composition,
        const partition::PartitionFunction& partitionFunction,
        const BuildDepthType buildDepth) :
    m_reactions(build_reaclib_nuclear_network(composition, buildDepth, false)),
    m_partitionFunction(partitionFunction.clone()),
    m_depth(buildDepth)
    {
        syncInternalMaps();
    }

    GraphEngine::GraphEngine(
        const reaction::LogicalReactionSet &reactions
    ) :
    m_reactions(reactions) {
        syncInternalMaps();
    }

    StepDerivatives<double> GraphEngine::calculateRHSAndEnergy(
        const std::vector<double> &Y,
        const double T9,
        const double rho
    ) const {
        if (m_usePrecomputation) {
            std::vector<double> bare_rates;
            std::vector<double> bare_reverse_rates;
            bare_rates.reserve(m_reactions.size());
            bare_reverse_rates.reserve(m_reactions.size());

            for (const auto& reaction: m_reactions) {
                bare_rates.push_back(reaction.calculate_rate(T9));
                bare_reverse_rates.push_back(calculateReverseRate(reaction, T9));
            }

            // --- The public facing interface can always use the precomputed version since taping is done internally ---
            return calculateAllDerivativesUsingPrecomputation(Y, bare_rates, bare_reverse_rates, T9, rho);
        } else {
            return calculateAllDerivatives<double>(Y, T9, rho);
        }
    }


    void GraphEngine::syncInternalMaps() {
        LOG_INFO(m_logger, "Synchronizing internal maps for REACLIB graph network (serif::network::GraphNetwork)...");
        collectNetworkSpecies();
        populateReactionIDMap();
        populateSpeciesToIndexMap();
        collectAtomicReverseRateAtomicBases();
        generateStoichiometryMatrix();
        reserveJacobianMatrix();
        recordADTape();

        const size_t n = m_rhsADFun.Domain();
        const size_t m = m_rhsADFun.Range();

        std::vector<bool> select_domain(n, true);
        std::vector<bool> select_range(m, true);

        m_rhsADFun.subgraph_sparsity(select_domain, select_range, false, m_full_jacobian_sparsity_pattern);
        m_jac_work.clear();

        precomputeNetwork();
        LOG_INFO(m_logger, "Internal maps synchronized. Network contains {} species and {} reactions.",
                 m_networkSpecies.size(), m_reactions.size());
    }

    // --- Network Graph Construction Methods ---
    void GraphEngine::collectNetworkSpecies() {
        m_networkSpecies.clear();
        m_networkSpeciesMap.clear();

        std::set<std::string_view> uniqueSpeciesNames;

        for (const auto& reaction: m_reactions) {
            for (const auto& reactant: reaction.reactants()) {
                uniqueSpeciesNames.insert(reactant.name());
            }
            for (const auto& product: reaction.products()) {
                uniqueSpeciesNames.insert(product.name());
            }
        }

        for (const auto& name: uniqueSpeciesNames) {
            auto it = fourdst::atomic::species.find(std::string(name));
            if (it != fourdst::atomic::species.end()) {
                m_networkSpecies.push_back(it->second);
                m_networkSpeciesMap.insert({name, it->second});
            } else {
                LOG_ERROR(m_logger, "Species '{}' not found in global atomic species database.", name);
                m_logger->flush_log();
                throw std::runtime_error("Species not found in global atomic species database: " + std::string(name));
            }
        }

    }

    void GraphEngine::populateReactionIDMap() {
        LOG_TRACE_L1(m_logger, "Populating reaction ID map for REACLIB graph network (serif::network::GraphNetwork)...");
        m_reactionIDMap.clear();
        for (auto& reaction: m_reactions) {
            m_reactionIDMap.emplace(reaction.id(), &reaction);
        }
        LOG_TRACE_L1(m_logger, "Populated {} reactions in the reaction ID map.", m_reactionIDMap.size());
    }

    void GraphEngine::populateSpeciesToIndexMap() {
        m_speciesToIndexMap.clear();
        for (size_t i = 0; i < m_networkSpecies.size(); ++i) {
            m_speciesToIndexMap.insert({m_networkSpecies[i], i});
        }
    }

    void GraphEngine::reserveJacobianMatrix() {
        // The implementation of this function (and others) constrains this nuclear network to a constant temperature and density during
        // each evaluation.
        size_t numSpecies = m_networkSpecies.size();
        m_jacobianMatrix.clear();
        m_jacobianMatrix.resize(numSpecies, numSpecies, false); // Sparse matrix, no initial values
        LOG_TRACE_L2(m_logger, "Jacobian matrix resized to {} rows and {} columns.",
                 m_jacobianMatrix.size1(), m_jacobianMatrix.size2());
    }

    // --- Basic Accessors and Queries ---
    const std::vector<fourdst::atomic::Species>& GraphEngine::getNetworkSpecies() const {
        // Returns a constant reference to the vector of unique species in the network.
        LOG_TRACE_L3(m_logger, "Providing access to network species vector. Size: {}.", m_networkSpecies.size());
        return m_networkSpecies;
    }

    const reaction::LogicalReactionSet& GraphEngine::getNetworkReactions() const {
        // Returns a constant reference to the set of reactions in the network.
        LOG_TRACE_L3(m_logger, "Providing access to network reactions set. Size: {}.", m_reactions.size());
        return m_reactions;
    }

    bool GraphEngine::involvesSpecies(const fourdst::atomic::Species& species) const {
        // Checks if a given species is present in the network's species map for efficient lookup.
        const bool found = m_networkSpeciesMap.contains(species.name());
        LOG_DEBUG(m_logger, "Checking if species '{}' is involved in the network: {}.", species.name(), found ? "Yes" : "No");
        return found;
    }

    // --- Validation Methods ---
    bool GraphEngine::validateConservation() const {
        LOG_TRACE_L1(m_logger, "Validating mass (A) and charge (Z) conservation across all reactions in the network.");

        for (const auto& reaction : m_reactions) {
            uint64_t totalReactantA = 0;
            uint64_t totalReactantZ = 0;
            uint64_t totalProductA = 0;
            uint64_t totalProductZ = 0;

            // Calculate total A and Z for reactants
            for (const auto& reactant : reaction.reactants()) {
                auto it = m_networkSpeciesMap.find(reactant.name());
                if (it != m_networkSpeciesMap.end()) {
                    totalReactantA += it->second.a();
                    totalReactantZ += it->second.z();
                } else {
                    // This scenario indicates a severe data integrity issue:
                    // a reactant is part of a reaction but not in the network's species map.
                    LOG_ERROR(m_logger, "CRITICAL ERROR: Reactant species '{}' in reaction '{}' not found in network species map during conservation validation.",
                             reactant.name(), reaction.id());
                    return false;
                }
            }

            // Calculate total A and Z for products
            for (const auto& product : reaction.products()) {
                auto it = m_networkSpeciesMap.find(product.name());
                if (it != m_networkSpeciesMap.end()) {
                    totalProductA += it->second.a();
                    totalProductZ += it->second.z();
                } else {
                    // Similar critical error for product species
                    LOG_ERROR(m_logger, "CRITICAL ERROR: Product species '{}' in reaction '{}' not found in network species map during conservation validation.",
                             product.name(), reaction.id());
                    return false;
                }
            }

            // Compare totals for conservation
            if (totalReactantA != totalProductA) {
                LOG_ERROR(m_logger, "Mass number (A) not conserved for reaction '{}': Reactants A={} vs Products A={}.",
                         reaction.id(), totalReactantA, totalProductA);
                return false;
            }
            if (totalReactantZ != totalProductZ) {
                LOG_ERROR(m_logger, "Atomic number (Z) not conserved for reaction '{}': Reactants Z={} vs Products Z={}.",
                         reaction.id(), totalReactantZ, totalProductZ);
                return false;
            }
        }

        LOG_TRACE_L1(m_logger, "Mass (A) and charge (Z) conservation validated successfully for all reactions.");
        return true; // All reactions passed the conservation check
    }

    void GraphEngine::validateComposition(const fourdst::composition::Composition &composition, double culling, double T9) {
        // Check if the requested network has already been cached.
        // PERF: Rebuilding this should be pretty fast but it may be a good point of optimization in the future.
        const reaction::LogicalReactionSet validationReactionSet = build_reaclib_nuclear_network(composition, false);
        // TODO: need some more robust method here to
        //       A. Build the basic network from the composition's species with non zero mass fractions.
        //       B. rebuild a new composition from the reaction set's reactants + products (with the mass fractions from the things that are only products set to 0)
        //       C. Rebuild the reaction set from the new composition
        //       D. Cull reactions where all reactants have mass fractions below the culling threshold.
        //       E. Be careful about maintaining caching through all of this


        // This allows for dynamic network modification while retaining caching for networks which are very similar.
        if (validationReactionSet != m_reactions) {
            LOG_DEBUG(m_logger, "Reaction set not cached. Rebuilding the reaction set for T9={} and culling={}.", T9, culling);
            m_reactions = validationReactionSet;
            syncInternalMaps(); // Re-sync internal maps after updating reactions. Note this will also retrace the AD tape.
        }
    }

    double GraphEngine::calculateReverseRate(
        const reaction::Reaction &reaction,
        const double T9
    ) const {
        if (!m_useReverseReactions) {
            LOG_TRACE_L2(m_logger, "Reverse reactions are disabled. Returning 0.0 for reverse rate of reaction '{}'.", reaction.id());
            return 0.0; // If reverse reactions are not used, return 0.0
        }
        const double expFactor = std::exp(-reaction.qValue() / (m_constants.kB * T9));
        if (s_debug) {
            std::cout << "\texpFactor = exp(-qValue/(kB * T9))\n";
            std::cout << "\texpFactor: " << expFactor << " for reaction: " << reaction.peName() << std::endl;
            std::cout << "\tQ-value: " << reaction.qValue() << " at T9: " << T9 << std::endl;
            std::cout << "\tT9: " << T9 << std::endl;
            std::cout << "\tkB * T9: " << m_constants.kB * T9 << std::endl;
            std::cout << "\tqValue/(kB * T9): " << reaction.qValue() / (m_constants.kB * T9) << std::endl;
        }
        double reverseRate = 0.0;
        const double forwardRate = reaction.calculate_rate(T9);

        if (reaction.reactants().size() == 2 && reaction.products().size() == 2) {
            reverseRate = calculateReverseRateTwoBody(reaction, T9, forwardRate, expFactor);
        } else {
            LOG_WARNING_LIMIT_EVERY_N(1000000, m_logger, "Reverse rate calculation for reactions with more than two reactants or products is not implemented (reaction id {}).", reaction.peName());
        }
        LOG_TRACE_L2(m_logger, "Calculated reverse rate for reaction '{}': {:.3E} at T9={:.3E}.", reaction.id(), reverseRate, T9);
        return reverseRate;
    }

    double GraphEngine::calculateReverseRateTwoBody(
        const reaction::Reaction &reaction,
        const double T9,
        const double forwardRate,
        const double expFactor
    ) const {
        std::vector<double> reactantPartitionFunctions;
        std::vector<double> productPartitionFunctions;

        reactantPartitionFunctions.reserve(reaction.reactants().size());
        productPartitionFunctions.reserve(reaction.products().size());

        std::unordered_map<fourdst::atomic::Species, int> reactantMultiplicity;
        std::unordered_map<fourdst::atomic::Species, int> productMultiplicity;

        reactantMultiplicity.reserve(reaction.reactants().size());
        productMultiplicity.reserve(reaction.products().size());

        for (const auto& reactant : reaction.reactants()) {
            reactantMultiplicity[reactant] += 1;
        }
        for (const auto& product : reaction.products()) {
            productMultiplicity[product] += 1;
        }
        double reactantSymmetryFactor = 1.0;
        double productSymmetryFactor = 1.0;
        for (const auto& count : reactantMultiplicity | std::views::values) {
            reactantSymmetryFactor *= std::tgamma(count + 1);
        }
        for (const auto& count : productMultiplicity | std::views::values) {
            productSymmetryFactor *= std::tgamma(count + 1);
        }
        const double symmetryFactor = reactantSymmetryFactor / productSymmetryFactor;

        // Accumulate mass terms
        auto mass_op = [](double acc, const auto& species) { return acc * species.a(); };
        const double massNumerator = std::accumulate(
            reaction.reactants().begin(),
            reaction.reactants().end(),
            1.0,
            mass_op
        );
        const double massDenominator = std::accumulate(
            reaction.products().begin(),
            reaction.products().end(),
            1.0,
            mass_op
        );

        // Accumulate partition functions
        auto pf_op = [&](double acc, const auto& species) {
            return acc * m_partitionFunction->evaluate(species.z(), species.a(), T9);
        };
        const double partitionFunctionNumerator = std::accumulate(
            reaction.reactants().begin(),
            reaction.reactants().end(),
            1.0,
            pf_op
        );
        const double partitionFunctionDenominator = std::accumulate(
            reaction.products().begin(),
            reaction.products().end(),
            1.0,
            pf_op
        );

        const double CT = std::pow(massNumerator/massDenominator, 1.5) *
            (partitionFunctionNumerator/partitionFunctionDenominator);

        double reverseRate = forwardRate * symmetryFactor * CT * expFactor;
        if (!std::isfinite(reverseRate)) {
            return 0.0; // If the reverse rate is not finite, return 0.0
        }
        return reverseRate; // Return the calculated reverse rate

    }

    double GraphEngine::calculateReverseRateTwoBodyDerivative(
        const reaction::Reaction &reaction,
        const double T9,
        const double reverseRate
    ) const {
        if (!m_useReverseReactions) {
            LOG_TRACE_L2(m_logger, "Reverse reactions are disabled. Returning 0.0 for reverse rate derivative of reaction '{}'.", reaction.id());
            return 0.0; // If reverse reactions are not used, return 0.0
        }
        const double d_log_kFwd = reaction.calculate_forward_rate_log_derivative(T9);

        auto log_deriv_pf_op = [&](double acc, const auto& species) {
            const double g = m_partitionFunction->evaluate(species.z(), species.a(), T9);
            const double dg_dT = m_partitionFunction->evaluateDerivative(species.z(), species.a(), T9);
            return (g == 0.0) ? acc : acc + (dg_dT / g);
        };

        const double reactant_log_derivative_sum = std::accumulate(
            reaction.reactants().begin(),
            reaction.reactants().end(),
            0.0,
            log_deriv_pf_op
        );

        const double product_log_derivative_sum = std::accumulate(
            reaction.products().begin(),
            reaction.products().end(),
            0.0,
            log_deriv_pf_op
        );

        const double d_log_C = reactant_log_derivative_sum - product_log_derivative_sum;

        const double d_log_exp = reaction.qValue() / (m_constants.kB * T9 * T9);

        const double log_total_derivative = d_log_kFwd + d_log_C + d_log_exp;

        return reverseRate * log_total_derivative; // Return the derivative of the reverse rate with respect to T9

    }

    bool GraphEngine::isUsingReverseReactions() const {
        return m_useReverseReactions;
    }

    void GraphEngine::setUseReverseReactions(const bool useReverse) {
        m_useReverseReactions = useReverse;
    }

    int GraphEngine::getSpeciesIndex(const fourdst::atomic::Species &species) const {
        return m_speciesToIndexMap.at(species); // Returns the index of the species in the stoichiometry matrix
    }

    std::vector<double> GraphEngine::mapNetInToMolarAbundanceVector(const NetIn &netIn) const {
        std::vector<double> Y(m_networkSpecies.size(), 0.0); // Initialize with zeros
        for (const auto& [symbol, entry] : netIn.composition) {
            Y[getSpeciesIndex(entry.isotope())] = netIn.composition.getMolarAbundance(symbol); // Map species to their molar abundance
        }
        return Y; // Return the vector of molar abundances
    }

    PrimingReport GraphEngine::primeEngine(const NetIn &netIn) {
        NetIn fullNetIn;
        fourdst::composition::Composition composition;

        std::vector<std::string> symbols;
        symbols.reserve(m_networkSpecies.size());
        for (const auto &symbol: m_networkSpecies) {
            symbols.emplace_back(symbol.name());
        }
        composition.registerSymbol(symbols);
        for (const auto& [symbol, entry] : netIn.composition) {
            if (m_networkSpeciesMap.contains(symbol)) {
                composition.setMassFraction(symbol, entry.mass_fraction());
            } else {
                composition.setMassFraction(symbol, 0.0);
            }
        }
        composition.finalize(true);
        fullNetIn.composition = composition;
        fullNetIn.temperature = netIn.temperature;
        fullNetIn.density = netIn.density;

        auto primingReport = primeNetwork(fullNetIn, *this);

        return primingReport;
    }

    BuildDepthType GraphEngine::getDepth() const {
        return m_depth;
    }

    void GraphEngine::rebuild(const fourdst::composition::Composition& comp, const BuildDepthType depth) {
        if (depth != m_depth) {
            m_depth = depth;
            m_reactions = build_reaclib_nuclear_network(comp, m_depth, false);
            syncInternalMaps(); // Resync internal maps after changing the depth
        } else {
            LOG_DEBUG(m_logger, "Rebuild requested with the same depth. No changes made to the network.");
        }
    }

    StepDerivatives<double> GraphEngine::calculateAllDerivativesUsingPrecomputation(
        const std::vector<double> &Y_in,
        const std::vector<double> &bare_rates,
        const std::vector<double> &bare_reverse_rates,
        const double T9,
        const double rho
    ) const {
        // --- Calculate screening factors ---
        const std::vector<double> screeningFactors = m_screeningModel->calculateScreeningFactors(
            m_reactions,
            m_networkSpecies,
            Y_in,
            T9,
            rho
        );

        // --- Optimized loop ---
        std::vector<double> molarReactionFlows;
        molarReactionFlows.reserve(m_precomputedReactions.size());

        for (const auto& precomp : m_precomputedReactions) {
            double forwardAbundanceProduct = 1.0;
            // bool below_threshold = false;
            for (size_t i = 0; i < precomp.unique_reactant_indices.size(); ++i) {
                const size_t reactantIndex = precomp.unique_reactant_indices[i];
                const int power = precomp.reactant_powers[i];
                // const double abundance = Y_in[reactantIndex];
                // if (abundance < MIN_ABUNDANCE_THRESHOLD) {
                //     below_threshold = true;
                //     break;
                // }

                forwardAbundanceProduct *= std::pow(Y_in[reactantIndex], power);
            }
            // if (below_threshold) {
            //     molarReactionFlows.push_back(0.0);
            //     continue; // Skip this reaction if any reactant is below the abundance threshold
            // }

            const double bare_rate = bare_rates[precomp.reaction_index];
            const double screeningFactor = screeningFactors[precomp.reaction_index];
            const size_t numReactants = m_reactions[precomp.reaction_index].reactants().size();
            const size_t numProducts = m_reactions[precomp.reaction_index].products().size();

            const double forwardMolarReactionFlow =
                    screeningFactor *
                    bare_rate *
                    precomp.symmetry_factor *
                    forwardAbundanceProduct *
                    std::pow(rho, numReactants >  1 ? numReactants - 1 : 0.0);

            double reverseMolarReactionFlow = 0.0;
            if (precomp.reverse_symmetry_factor != 0.0 and m_useReverseReactions) {
                const double bare_reverse_rate = bare_reverse_rates[precomp.reaction_index];
                double reverseAbundanceProduct = 1.0;
                for (size_t i = 0; i < precomp.unique_product_indices.size(); ++i) {
                    reverseAbundanceProduct *= std::pow(Y_in[precomp.unique_product_indices[i]], precomp.product_powers[i]);
                }
                reverseMolarReactionFlow = screeningFactor *
                    bare_reverse_rate *
                    precomp.reverse_symmetry_factor *
                    reverseAbundanceProduct *
                    std::pow(rho, numProducts > 1 ? numProducts - 1 : 0.0);
            }

            molarReactionFlows.push_back(forwardMolarReactionFlow - reverseMolarReactionFlow);

        }

        // --- Assemble molar abundance derivatives ---
        StepDerivatives<double> result;
        result.dydt.assign(m_networkSpecies.size(), 0.0); // Initialize derivatives to zero
        for (size_t j = 0; j < m_precomputedReactions.size(); ++j) {
            const auto& precomp = m_precomputedReactions[j];
            const double R_j = molarReactionFlows[j];

            for (size_t i = 0; i < precomp.affected_species_indices.size(); ++i) {
                const size_t speciesIndex = precomp.affected_species_indices[i];
                const int stoichiometricCoefficient = precomp.stoichiometric_coefficients[i];

                // Update the derivative for this species
                result.dydt[speciesIndex] += static_cast<double>(stoichiometricCoefficient) * R_j;
            }
        }

        // --- Calculate the nuclear energy generation rate ---
        double massProductionRate = 0.0; // [mol][s^-1]
        for (size_t i = 0; i < m_networkSpecies.size(); ++i) {
            const auto& species = m_networkSpecies[i];
            massProductionRate += result.dydt[i] * species.mass() * m_constants.u;
        }
        result.nuclearEnergyGenerationRate = -massProductionRate * m_constants.Na * m_constants.c * m_constants.c; // [erg][s^-1][g^-1]
        return result;

    }

    // --- Generate Stoichiometry Matrix ---
    void GraphEngine::generateStoichiometryMatrix() {
        LOG_TRACE_L1(m_logger, "Generating stoichiometry matrix...");

        // Task 1: Set dimensions and initialize the matrix
        size_t numSpecies = m_networkSpecies.size();
        size_t numReactions = m_reactions.size();
        m_stoichiometryMatrix.resize(numSpecies, numReactions, false);

        LOG_TRACE_L1(m_logger, "Stoichiometry matrix initialized with dimensions: {} rows (species) x {} columns (reactions).",
                 numSpecies, numReactions);

        // Task 2: Populate the stoichiometry matrix
        // Iterate through all reactions, assign them a column index, and fill in their stoichiometric coefficients.
        size_t reactionColumnIndex = 0;
        for (const auto& reaction : m_reactions) {
            // Get the net stoichiometry for the current reaction
            std::unordered_map<fourdst::atomic::Species, int> netStoichiometry = reaction.stoichiometry();

            // Iterate through the species and their coefficients in the stoichiometry map
            for (const auto& [species, coefficient] : netStoichiometry) {
                // Find the row index for this species
                auto it = m_speciesToIndexMap.find(species);
                if (it != m_speciesToIndexMap.end()) {
                    const size_t speciesRowIndex = it->second;
                    // Set the matrix element. Boost.uBLAS handles sparse insertion.
                    m_stoichiometryMatrix(speciesRowIndex, reactionColumnIndex) = coefficient;
                } else {
                    // This scenario should ideally not happen if m_networkSpeciesMap and m_speciesToIndexMap are correctly synced
                    LOG_ERROR(m_logger, "CRITICAL ERROR: Species '{}' from reaction '{}' stoichiometry not found in species to index map.",
                             species.name(), reaction.id());
                    m_logger -> flush_log();
                    throw std::runtime_error("Species not found in species to index map: " + std::string(species.name()));
                }
            }
            reactionColumnIndex++; // Move to the next column for the next reaction
        }

        LOG_TRACE_L1(m_logger, "Stoichiometry matrix population complete. Number of non-zero elements: {}.",
                 m_stoichiometryMatrix.nnz()); // Assuming nnz() exists for compressed_matrix
    }

    StepDerivatives<double> GraphEngine::calculateAllDerivatives(
        const std::vector<double> &Y_in,
        const double T9,
        const double rho
    ) const {
        return calculateAllDerivatives<double>(Y_in, T9, rho);
    }

    StepDerivatives<ADDouble> GraphEngine::calculateAllDerivatives(
        const std::vector<ADDouble> &Y_in,
        const ADDouble &T9,
        const ADDouble &rho
    ) const {
        return calculateAllDerivatives<ADDouble>(Y_in, T9, rho);
    }

    void GraphEngine::setScreeningModel(const screening::ScreeningType model) {
        m_screeningModel = screening::selectScreeningModel(model);
        m_screeningType = model;
    }

    screening::ScreeningType GraphEngine::getScreeningModel() const {
        return m_screeningType;
    }

    void GraphEngine::setPrecomputation(const bool precompute) {
        m_usePrecomputation = precompute;
    }

    bool GraphEngine::isPrecomputationEnabled() const {
        return m_usePrecomputation;
    }

    const partition::PartitionFunction & GraphEngine::getPartitionFunction() const {
        return *m_partitionFunction;
    }

    double GraphEngine::calculateMolarReactionFlow(
        const reaction::Reaction &reaction,
        const std::vector<double> &Y,
        const double T9,
        const double rho
    ) const {
        return calculateMolarReactionFlow<double>(reaction, Y, T9, rho);
    }

    void GraphEngine::generateJacobianMatrix(
        const std::vector<double> &Y_dynamic,
        const double T9,
        const double rho
    ) {

        LOG_TRACE_L1(m_logger, "Generating jacobian matrix for T9={}, rho={}..", T9, rho);
        const size_t numSpecies = m_networkSpecies.size();

        // 1. Pack the input variables into a vector for CppAD
        std::vector<double> adInput(numSpecies + 2, 0.0); // +2 for T9 and rho
        for (size_t i = 0; i < numSpecies; ++i) {
            adInput[i] = Y_dynamic[i];
        }
        adInput[numSpecies]     = T9;  // T9
        adInput[numSpecies + 1] = rho; // rho
        LOG_DEBUG(
            m_logger,
            "AD Input to jacobian {}",
            [&]() -> std::string {
                std::stringstream ss;
                ss << std::scientific << std::setprecision(5);
                for (size_t i = 0; i < adInput.size(); ++i) {
                    ss << adInput[i];
                    if (i < adInput.size() - 1) {
                        ss << ", ";
                    }
                }
                return ss.str();
            }());

        // 2. Calculate the full jacobian
        const std::vector<double> dotY = m_rhsADFun.Jacobian(adInput);

        // 3. Pack jacobian vector into sparse matrix
        m_jacobianMatrix.clear();
        for (size_t i = 0; i < numSpecies; ++i) {
            for (size_t j = 0; j < numSpecies; ++j) {
                const double value = dotY[i * (numSpecies + 2) + j];
                if (std::abs(value) > MIN_JACOBIAN_THRESHOLD) {
                    m_jacobianMatrix(i, j) = value;
                }
            }
        }
        // LOG_DEBUG(
        //     m_logger,
        //     "Final Jacobian is:\n{}",
        //     [&]() -> std::string {
        //         std::stringstream ss;
        //         ss << std::scientific << std::setprecision(5);
        //         for (size_t i = 0; i < m_jacobianMatrix.size1(); ++i) {
        //             for (size_t j = 0; j < m_jacobianMatrix.size2(); ++j) {
        //                 ss << m_jacobianMatrix(i, j);
        //                 if (j < m_jacobianMatrix.size2() - 1) {
        //                     ss << ", ";
        //                 }
        //             }
        //             ss << "\n";
        //         }
        //         return ss.str();
        //     }());
        LOG_TRACE_L1(m_logger, "Jacobian matrix generated with dimensions: {} rows x {} columns.", m_jacobianMatrix.size1(), m_jacobianMatrix.size2());
    }

    void GraphEngine::generateJacobianMatrix(
        const std::vector<double> &Y_dynamic,
        const double T9,
        const double rho,
        const SparsityPattern &sparsityPattern
    ) {
        // --- Pack the input variables into a vector for CppAD ---
        const size_t numSpecies = m_networkSpecies.size();
        std::vector<double> x(numSpecies + 2, 0.0);
        for (size_t i = 0; i < numSpecies; ++i) {
           x[i] = Y_dynamic[i];
        }
        x[numSpecies] = T9;
        x[numSpecies + 1] = rho;

        // --- Convert into CppAD Sparsity pattern ---
        const size_t nnz = sparsityPattern.size(); // Number of non-zero entries in the sparsity pattern
        std::vector<size_t> row_indices(nnz);
        std::vector<size_t> col_indices(nnz);

        for (size_t k = 0; k < nnz; ++k) {
            row_indices[k] = sparsityPattern[k].first;
            col_indices[k] = sparsityPattern[k].second;
        }

        std::vector<double> values(nnz);
        const size_t num_rows_jac = numSpecies;
        const size_t num_cols_jac = numSpecies + 2; // +2 for T9 and rho

        CppAD::sparse_rc<std::vector<size_t>> CppAD_sparsity_pattern(num_rows_jac, num_cols_jac, nnz);
        for (size_t k = 0; k < nnz; ++k) {
            CppAD_sparsity_pattern.set(k, sparsityPattern[k].first, sparsityPattern[k].second);
        }

        CppAD::sparse_rcv<std::vector<size_t>, std::vector<double>> jac_subset(CppAD_sparsity_pattern);

        m_rhsADFun.sparse_jac_rev(
            x,
            jac_subset, // Sparse Jacobian output
            m_full_jacobian_sparsity_pattern,
            "cppad",
            m_jac_work // Work vector for CppAD
        );

        // --- Convert the sparse Jacobian back to the Boost uBLAS format ---
        m_jacobianMatrix.clear();
        for (size_t k = 0; k < nnz; ++k) {
            const size_t row = jac_subset.row()[k];
            const size_t col = jac_subset.col()[k];
            const double value = jac_subset.val()[k];

            if (std::abs(value) > MIN_JACOBIAN_THRESHOLD) {
                m_jacobianMatrix(row, col) = value; // Insert into the sparse matrix
            }
        }
    }

    double GraphEngine::getJacobianMatrixEntry(const int i, const int j) const {
        LOG_TRACE_L3(m_logger, "Getting jacobian matrix entry for {},{} = {}", i, j, m_jacobianMatrix(i, j));
        return m_jacobianMatrix(i, j);
    }

    std::unordered_map<fourdst::atomic::Species, int> GraphEngine::getNetReactionStoichiometry(
        const reaction::Reaction &reaction
    ) {
        return reaction.stoichiometry();
    }

    int GraphEngine::getStoichiometryMatrixEntry(
        const int speciesIndex,
        const int reactionIndex
    ) const {
        return m_stoichiometryMatrix(speciesIndex, reactionIndex);
    }

    void GraphEngine::exportToDot(const std::string &filename) const {
        LOG_TRACE_L1(m_logger, "Exporting network graph to DOT file: {}", filename);

        std::ofstream dotFile(filename);
        if (!dotFile.is_open()) {
            LOG_ERROR(m_logger, "Failed to open file for writing: {}", filename);
            m_logger->flush_log();
            throw std::runtime_error("Failed to open file for writing: " + filename);
        }

        dotFile << "digraph NuclearReactionNetwork {\n";
        dotFile << "    graph [rankdir=LR, splines=true, overlap=false, bgcolor=\"#f0f0f0\"];\n";
        dotFile << "    node [shape=circle, style=filled, fillcolor=\"#a7c7e7\", fontname=\"Helvetica\"];\n";
        dotFile << "    edge [fontname=\"Helvetica\", fontsize=10];\n\n";

        // 1. Define all species as nodes
        dotFile << "    // --- Species Nodes ---\n";
        for (const auto& species : m_networkSpecies) {
            dotFile << "    \"" << species.name() << "\" [label=\"" << species.name() << "\"];\n";
        }
        dotFile << "\n";

        // 2. Define all reactions as intermediate nodes and connect them
        dotFile << "    // --- Reaction Edges ---\n";
        for (const auto& reaction : m_reactions) {
            // Create a unique ID for the reaction node
            std::string reactionNodeId = "reaction_" + std::string(reaction.id());

            // Define the reaction node (small, black dot)
            dotFile << "    \"" << reactionNodeId << "\" [shape=point, fillcolor=black, width=0.1, height=0.1, label=\"\"];\n";

            // Draw edges from reactants to the reaction node
            for (const auto& reactant : reaction.reactants()) {
                dotFile << "    \"" << reactant.name() << "\" -> \"" << reactionNodeId << "\";\n";
            }

            // Draw edges from the reaction node to products
            for (const auto& product : reaction.products()) {
                dotFile << "    \"" << reactionNodeId << "\" -> \"" << product.name() << "\" [label=\"" << reaction.qValue() << " MeV\"];\n";
            }
            dotFile << "\n";
        }

        dotFile << "}\n";
        dotFile.close();
        LOG_TRACE_L1(m_logger, "Successfully exported network to {}", filename);
    }

    void GraphEngine::exportToCSV(const std::string &filename) const {
        LOG_TRACE_L1(m_logger, "Exporting network graph to CSV file: {}", filename);

        std::ofstream csvFile(filename, std::ios::out | std::ios::trunc);
        if (!csvFile.is_open()) {
            LOG_ERROR(m_logger, "Failed to open file for writing: {}", filename);
            m_logger->flush_log();
            throw std::runtime_error("Failed to open file for writing: " + filename);
        }
        csvFile << "Reaction;Reactants;Products;Q-value;sources;rates\n";
        for (const auto& reaction : m_reactions) {
            // Dynamic cast to REACLIBReaction to access specific properties
            csvFile << reaction.id() << ";";
            // Reactants
            int count = 0;
            for (const auto& reactant : reaction.reactants()) {
                csvFile << reactant.name();
                if (++count < reaction.reactants().size()) {
                    csvFile << ",";
                }
            }
            csvFile << ";";
            count = 0;
            for (const auto& product : reaction.products()) {
                csvFile << product.name();
                if (++count < reaction.products().size()) {
                    csvFile << ",";
                }
            }
            csvFile << ";" << reaction.qValue() << ";";
            // Reaction coefficients
            auto sources = reaction.sources();
            count = 0;
            for (const auto& source : sources) {
                csvFile << source;
                if (++count < sources.size()) {
                    csvFile << ",";
                }
            }
            csvFile << ";";
            // Reaction coefficients
            count = 0;
            for (const auto& rates : reaction) {
                csvFile << rates;
                if (++count < reaction.size()) {
                    csvFile << ",";
                }
            }
            csvFile << "\n";
        }
        csvFile.close();
        LOG_TRACE_L1(m_logger, "Successfully exported network graph to {}", filename);
    }

    std::unordered_map<fourdst::atomic::Species, double> GraphEngine::getSpeciesTimescales(
        const std::vector<double> &Y,
        const double T9,
        const double rho
    ) const {
        auto [dydt, _] = calculateAllDerivatives<double>(Y, T9, rho);
        std::unordered_map<fourdst::atomic::Species, double> speciesTimescales;
        speciesTimescales.reserve(m_networkSpecies.size());
        for (size_t i = 0; i < m_networkSpecies.size(); ++i) {
            double timescale = std::numeric_limits<double>::infinity();
            const auto species = m_networkSpecies[i];
            if (std::abs(dydt[i]) > 0.0) {
                timescale = std::abs(Y[i] / dydt[i]);
            }
            speciesTimescales.emplace(species, timescale);
        }
        return speciesTimescales;
    }

    void GraphEngine::update(const NetIn &netIn) {
        // No-op for GraphEngine, as it does not support manually triggering updates.
    }

    void GraphEngine::recordADTape() {
        LOG_TRACE_L1(m_logger, "Recording AD tape for the RHS calculation...");

        // Task 1: Set dimensions and initialize the matrix
        const size_t numSpecies = m_networkSpecies.size();
        if (numSpecies == 0) {
            LOG_ERROR(m_logger, "Cannot record AD tape: No species in the network.");
            m_logger->flush_log();
            throw std::runtime_error("Cannot record AD tape: No species in the network.");
        }
        const size_t numADInputs = numSpecies + 2; // Note here that by not letting T9 and rho be independent variables, we are constraining the network to a constant temperature and density during each evaluation.

        // --- CppAD Tape Recording ---
        // 1. Declare independent variable (adY)
        //    We also initialize the dummy variable for tape recording (these tell CppAD what the derivative chain looks like).
        //    Their numeric values are irrelevant except for in so far as they avoid numerical instabilities.

        // Distribute total mass fraction uniformly between species in the dummy variable space
        const auto uniformMassFraction = static_cast<CppAD::AD<double>>(1.0 / static_cast<double>(numSpecies));
        std::vector<CppAD::AD<double>> adInput(numADInputs, uniformMassFraction);
        adInput[numSpecies]     = 1.0; // Dummy T9
        adInput[numSpecies + 1] = 1.0; // Dummy rho

        // 3. Declare independent variables (what CppAD will differentiate wrt.)
        //    This also beings the tape recording process.
        CppAD::Independent(adInput);

        std::vector<CppAD::AD<double>> adY(numSpecies);
        for(size_t i = 0; i < numSpecies; ++i) {
            adY[i] = adInput[i];
        }
        const CppAD::AD<double> adT9  = adInput[numSpecies];
        const CppAD::AD<double> adRho = adInput[numSpecies + 1];


        // 5. Call the actual templated function
        // We let T9 and rho be constant, so we pass them as fixed values.
        auto [dydt, nuclearEnergyGenerationRate] = calculateAllDerivatives<CppAD::AD<double>>(adY, adT9, adRho);

        m_rhsADFun.Dependent(adInput, dydt);

        LOG_TRACE_L1(m_logger, "AD tape recorded successfully for the RHS calculation. Number of independent variables: {}.",
                 adInput.size());
    }

    void GraphEngine::collectAtomicReverseRateAtomicBases() {
        m_atomicReverseRates.clear();
        m_atomicReverseRates.reserve(m_reactions.size());

        for (const auto& reaction: m_reactions) {
            if (reaction.qValue() != 0.0) {
                m_atomicReverseRates.push_back(std::make_unique<AtomicReverseRate>(reaction, *this));
            } else {
                m_atomicReverseRates.push_back(nullptr);
            }
        }
    }

    void GraphEngine::precomputeNetwork() {
        LOG_TRACE_L1(m_logger, "Pre-computing constant components of GraphNetwork state...");

        // --- Reverse map for fast species lookups ---
        std::unordered_map<fourdst::atomic::Species, size_t> speciesIndexMap;
        for (size_t i = 0; i < m_networkSpecies.size(); ++i) {
            speciesIndexMap[m_networkSpecies[i]] = i;
        }

        m_precomputedReactions.clear();
        m_precomputedReactions.reserve(m_reactions.size());

        for (size_t i = 0; i < m_reactions.size(); ++i) {
            const auto& reaction = m_reactions[i];
            PrecomputedReaction precomp;
            precomp.reaction_index = i;

            // --- Precompute forward reaction information ---
            // Count occurrences for each reactant to determine powers and symmetry
            std::unordered_map<size_t, int> reactantCounts;
            for (const auto& reactant: reaction.reactants()) {
                size_t reactantIndex = speciesIndexMap.at(reactant);
                reactantCounts[reactantIndex]++;
            }

            double symmetryDenominator = 1.0;
            for (const auto& [index, count] : reactantCounts) {
                precomp.unique_reactant_indices.push_back(index);
                precomp.reactant_powers.push_back(count);

                symmetryDenominator *= std::tgamma(count + 1);
            }

            precomp.symmetry_factor = 1.0/symmetryDenominator;

            // --- Precompute reverse reaction information ---
            if (reaction.qValue() != 0.0) {
                std::unordered_map<size_t, int> productCounts;
                for (const auto& product : reaction.products()) {
                    productCounts[speciesIndexMap.at(product)]++;
                }
                double reverseSymmetryDenominator = 1.0;
                for (const auto& [index, count] : productCounts) {
                    precomp.unique_product_indices.push_back(index);
                    precomp.product_powers.push_back(count);
                    reverseSymmetryDenominator *= std::tgamma(count + 1);
                }

                precomp.reverse_symmetry_factor = 1.0/reverseSymmetryDenominator;
            } else {
                precomp.unique_product_indices.clear();
                precomp.product_powers.clear();
                precomp.reverse_symmetry_factor = 0.0; // No reverse reaction for Q = 0 reactions
            }

            // --- Precompute stoichiometry information ---
            const auto stoichiometryMap = reaction.stoichiometry();
            precomp.affected_species_indices.reserve(stoichiometryMap.size());
            precomp.stoichiometric_coefficients.reserve(stoichiometryMap.size());

            for (const auto& [species, coeff] : stoichiometryMap) {
                precomp.affected_species_indices.push_back(speciesIndexMap.at(species));
                precomp.stoichiometric_coefficients.push_back(coeff);
            }

            m_precomputedReactions.push_back(std::move(precomp));
        }
    }

    bool GraphEngine::AtomicReverseRate::forward(
        const size_t p,
        const size_t q,
        const CppAD::vector<bool> &vx,
        CppAD::vector<bool> &vy,
        const CppAD::vector<double> &tx,
        CppAD::vector<double> &ty
    ) {

        if ( p != 0) { return false; }
        const double T9 = tx[0];

        const double reverseRate = m_engine.calculateReverseRate(m_reaction, T9);
        // std::cout << m_reaction.peName() << " reverseRate: " << reverseRate << " at T9: " << T9 << "\n";
        ty[0] = reverseRate; // Store the reverse rate in the output vector

        if (vx.size() > 0) {
            vy[0] = vx[0];
        }
        return true;
    }

    bool GraphEngine::AtomicReverseRate::reverse(
        size_t q,
        const CppAD::vector<double> &tx,
        const CppAD::vector<double> &ty,
        CppAD::vector<double> &px,
        const CppAD::vector<double> &py
    ) {
        const double T9 = tx[0];
        const double reverseRate = ty[0];

        const double derivative = m_engine.calculateReverseRateTwoBodyDerivative(m_reaction, T9, reverseRate);
        // std::cout << m_reaction.peName() << " reverseRate Derivative: " << derivative << "\n";

        px[0] = py[0] * derivative; // Return the derivative of the reverse rate with respect to T9

        return true;
    }

    bool GraphEngine::AtomicReverseRate::for_sparse_jac(
        size_t q,
        const CppAD::vector<std::set<size_t>> &r,
        CppAD::vector<std::set<size_t>> &s
    ) {
        s[0] = r[0];
        return true;
    }

    bool GraphEngine::AtomicReverseRate::rev_sparse_jac(
        size_t q,
        const CppAD::vector<std::set<size_t>> &rt,
        CppAD::vector<std::set<size_t>> &st
    ) {
        st[0] = rt[0];
        return true;
    }
}