cpu_vxm.hpp

Go to the documentation of this file.

/**********************************************************************************/
/* This file is part of spla project                                              */
/* https://github.com/SparseLinearAlgebra/spla                                    */
/**********************************************************************************/
/* MIT License                                                                    */
/*                                                                                */
/* Copyright (c) 2023 SparseLinearAlgebra                                         */
/*                                                                                */
/* Permission is hereby granted, free of charge, to any person obtaining a copy   */
/* of this software and associated documentation files (the "Software"), to deal  */
/* in the Software without restriction, including without limitation the rights   */
/* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      */
/* copies of the Software, and to permit persons to whom the Software is          */
/* furnished to do so, subject to the following conditions:                       */
/*                                                                                */
/* The above copyright notice and this permission notice shall be included in all */
/* copies or substantial portions of the Software.                                */
/*                                                                                */
/* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     */
/* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       */
/* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    */
/* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         */
/* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  */
/* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  */
/* SOFTWARE.                                                                      */
/**********************************************************************************/
 
#ifndef SPLA_CPU_VXM_HPP
#define SPLA_CPU_VXM_HPP
 
#include <schedule/schedule_tasks.hpp>
 
#include <core/dispatcher.hpp>
#include <core/registry.hpp>
#include <core/tmatrix.hpp>
#include <core/top.hpp>
#include <core/tscalar.hpp>
#include <core/ttype.hpp>
#include <core/tvector.hpp>
 
#include <robin_hood.hpp>
 
namespace spla {
 
    template<typename T>
    class Algo_vxm_masked_cpu final : public RegistryAlgo {
    public:
        ~Algo_vxm_masked_cpu() override = default;
 
        std::string get_name() override {
            return "vxm_masked";
        }
 
        std::string get_description() override {
            return "sequential masked vector-matrix product on cpu";
        }
 
        Status execute(const DispatchContext& ctx) override {
            TIME_PROFILE_SCOPE("cpu/vxm");
 
            auto t = ctx.task.template cast_safe<ScheduleTask_vxm_masked>();
 
            auto r           = t->r.template cast_safe<TVector<T>>();
            auto mask        = t->mask.template cast_safe<TVector<T>>();
            auto v           = t->v.template cast_safe<TVector<T>>();
            auto M           = t->M.template cast_safe<TMatrix<T>>();
            auto op_multiply = t->op_multiply.template cast_safe<TOpBinary<T, T, T>>();
            auto op_add      = t->op_add.template cast_safe<TOpBinary<T, T, T>>();
            auto op_select   = t->op_select.template cast_safe<TOpSelect<T>>();
            auto init        = t->init.template cast_safe<TScalar<T>>();
 
            const T sum_init = init->get_value();
 
            r->validate_wd(FormatVector::CpuCoo);
            mask->validate_rw(FormatVector::CpuDense);
            v->validate_rw(FormatVector::CpuCoo);
            M->validate_rw(FormatMatrix::CpuLil);
 
            CpuCooVec<T>*         p_sparse_r   = r->template get<CpuCooVec<T>>();
            const CpuDenseVec<T>* p_dense_mask = mask->template get<CpuDenseVec<T>>();
            const CpuCooVec<T>*   p_sparse_v   = v->template get<CpuCooVec<T>>();
            const CpuLil<T>*      p_lil_M      = M->template get<CpuLil<T>>();
 
            auto& func_multiply = op_multiply->function;
            auto& func_add      = op_add->function;
            auto& func_select   = op_select->function;
 
            const uint N = p_sparse_v->values;
 
            robin_hood::unordered_flat_map<uint, T> r_tmp;
 
            for (uint idx = 0; idx < N; ++idx) {
                const uint v_i = p_sparse_v->Ai[idx];
                const T    v_x = p_sparse_v->Ax[idx];
 
                const auto& row = p_lil_M->Ar[v_i];
 
                for (const auto& j_x : row) {
                    const uint j = j_x.first;
 
                    if (func_select(p_dense_mask->Ax[j])) {
                        auto r_x = r_tmp.find(j);
 
                        if (r_x != r_tmp.end())
                            r_x->second = func_add(r_x->second, func_multiply(v_x, j_x.second));
                        else
                            r_tmp[j] = func_multiply(v_x, j_x.second);
                    }
                }
            }
 
            std::vector<std::pair<uint, T>> r_entries;
            r_entries.reserve(r_tmp.size());
            for (const auto& e : r_tmp) {
                r_entries.emplace_back(e.first, e.second);
            }
            std::sort(r_entries.begin(), r_entries.end());
 
            p_sparse_r->values = uint(r_tmp.size());
            p_sparse_r->Ai.reserve(r_tmp.size());
            p_sparse_r->Ax.reserve(r_tmp.size());
            for (const auto& e : r_entries) {
                p_sparse_r->Ai.push_back(e.first);
                p_sparse_r->Ax.push_back(e.second);
            }
 
            return Status::Ok;
        }
    };
 
}// namespace spla
 
#endif//SPLA_CPU_VXM_HPP