hyperloglog.hpp

Go to the documentation of this file.

// SPDX-FileCopyrightText: 2006-2025, Knut Reinert & Freie Universität Berlin
// SPDX-FileCopyrightText: 2016-2025, Knut Reinert & MPI für molekulare Genetik
// SPDX-FileCopyrightText: 2013 Hideaki Ohno <hide.o.j55{at}gmail.com>
// SPDX-License-Identifier: BSD-3-Clause AND MIT
 
#pragma once
 
#include <array>   // for array
#include <cstddef> // for size_t
#include <cstdint> // for uint64_t, uint8_t, uint32_t
#include <iosfwd>  // for istream, ostream
#include <vector>  // for vector
 
#include <cereal/access.hpp> // for access
#include <cereal/cereal.hpp> // for make_nvp, CEREAL_NVP
 
#include <hibf/contrib/aligned_allocator.hpp> // for aligned_allocator
#include <hibf/platform.hpp>
 
namespace seqan::hibf::sketch
{
 
class hyperloglog
{
public:
    hyperloglog(uint8_t const num_bits = 5u);
    hyperloglog(hyperloglog const &) = default;             
    hyperloglog & operator=(hyperloglog const &) = default; 
    hyperloglog(hyperloglog &&) = default;                  
    hyperloglog & operator=(hyperloglog &&) = default;      
    ~hyperloglog() = default;                               
 
 
    void add(uint64_t const value);
 
    double estimate() const;
 
    void merge(hyperloglog const & other);
 
    double merge_and_estimate(hyperloglog const & other);
 
    void reset();
 
    uint64_t data_size() const
    {
        return size;
    }
    uint64_t data_size() const {…}
 
    void store(std::ostream & os) const;
 
    void load(std::istream & is);
 
private:
    static constexpr std::array<float, 61> expectation_values = []() constexpr
    {
        std::array<float, 61> result{};
        for (size_t i = 0; i < 61; ++i)
            result[i] = 1.0f / (1ULL << i);
        return result;
    }();
 
    uint8_t bits{};
    uint64_t size{};
    uint64_t rank_mask{};
    double normalization_factor{};
    std::vector<uint8_t, seqan::hibf::contrib::aligned_allocator<uint8_t, 32u>> data{};
 
    friend class cereal::access;
 
    template <typename archive_t>
    void serialize(archive_t & archive)
    {
        uint32_t version{1};
        archive(CEREAL_NVP(version));
 
        archive(CEREAL_NVP(bits));
        archive(CEREAL_NVP(size));
        archive(CEREAL_NVP(rank_mask));
        archive(CEREAL_NVP(normalization_factor));
        archive(CEREAL_NVP(data));
    }
};
class hyperloglog {…};
 
} // namespace seqan::hibf::sketch