data_store.hpp

// SPDX-FileCopyrightText: 2006-2024, Knut Reinert & Freie Universität Berlin
// SPDX-FileCopyrightText: 2016-2024, Knut Reinert & MPI für molekulare Genetik
// SPDX-License-Identifier: BSD-3-Clause
 
#pragma once
 
#include <cassert>   // for assert
#include <cinttypes> // for uint64_t
#include <cstddef>   // for size_t
#include <numeric>   // for iota
#include <string>    // for string
#include <vector>    // for vector
 
#include <hibf/layout/layout.hpp>      // for layout
#include <hibf/misc/timer.hpp>         // for concurrent_timer
#include <hibf/sketch/hyperloglog.hpp> // for hyperloglog
 
namespace seqan::hibf::layout
{
 
struct data_store
{
    struct previous_level
    {
        std::vector<size_t> bin_indices{};
        std::string num_of_bins;
 
        bool empty() const
        {
            assert(bin_indices.empty() == num_of_bins.empty());
            return bin_indices.empty();
        }
    };
 
    double false_positive_rate{};
 
    layout * hibf_layout; // Will be modified by {simple,hierarchical}_binning.
 
    std::vector<size_t> const * kmer_counts{}; // Pointed to data should not be modified.
 
    std::vector<sketch::hyperloglog> const * sketches{}; // Pointed to data should not be modified.
 
    std::vector<size_t> positions = [this]()
    {
        std::vector<size_t> ps;
        ps.resize(this->kmer_counts->size());
        std::iota(ps.begin(), ps.end(), 0);
        return ps;
    }(); // GCOVR_EXCL_LINE
 
    std::vector<double> fpr_correction{};
 
    previous_level previous{};
 
    std::vector<uint64_t> union_estimates{};
 
    bool user_bins_arranged{false};
 
    concurrent_timer union_estimation_timer{};
 
    concurrent_timer rearrangement_timer{};
};
 
} // namespace seqan::hibf::layout