RayleighRitzEigenSolver.h

Go to the documentation of this file.

/******************************************************************************
 * Copyright (c) 2021.                                                        *
 * The Regents of the University of Michigan and DFT-EFE developers.          *
 *                                                                            *
 * This file is part of the DFT-EFE code.                                     *
 *                                                                            *
 * DFT-EFE is free software: you can redistribute it and/or modify            *
 *   it under the terms of the Lesser GNU General Public License as           *
 *   published by the Free Software Foundation, either version 3 of           *
 *   the License, or (at your option) any later version.                      *
 *                                                                            *
 * DFT-EFE is distributed in the hope that it will be useful, but             *
 *   WITHOUT ANY WARRANTY; without even the implied warranty                  *
 *   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.                     *
 *   See the Lesser GNU General Public License for more details.              *
 *                                                                            *
 * You should have received a copy of the GNU Lesser General Public           *
 *   License at the top level of DFT-EFE distribution.  If not, see           *
 *   <https://www.gnu.org/licenses/>.                                         *
 ******************************************************************************/
 
/*
 * @author Avirup Sircar
 */
 
#ifndef dftefeRayleighRitzEigenSolver_h
#define dftefeRayleighRitzEigenSolver_h
 
#include <utils/MemorySpaceType.h>
#include <linearAlgebra/BlasLapackTypedef.h>
#include <linearAlgebra/LinAlgOpContext.h>
#include <linearAlgebra/Vector.h>
#include <linearAlgebra/MultiVector.h>
#include <linearAlgebra/LinearAlgebraTypes.h>
#include <linearAlgebra/OperatorContext.h>
#include <memory>
#include <linearAlgebra/ElpaScalapackManager.h>
 
namespace dftefe
{
  namespace linearAlgebra
  {
    template <typename ValueTypeOperator,
              typename ValueTypeOperand,
              utils::MemorySpace memorySpace>
    class RayleighRitzEigenSolver
    {
    public:
      using ValueType =
        blasLapack::scalar_type<ValueTypeOperator, ValueTypeOperand>;
      using RealType = blasLapack::real_type<ValueType>;
      using OpContext =
        OperatorContext<ValueTypeOperator, ValueTypeOperand, memorySpace>;
 
    public:
      RayleighRitzEigenSolver(
        const size_type             eigenVectorBatchSize,
        const ElpaScalapackManager &elpaScala,
        std::shared_ptr<const utils::mpi::MPIPatternP2P<memorySpace>>
                                                      mpiPatternP2P,
        std::shared_ptr<LinAlgOpContext<memorySpace>> linAlgOpContext,
        const bool                                    useScalpack = true);
 
      ~RayleighRitzEigenSolver() = default;
 
      // In this case we can solve the KohnSham GHEP if the X's are B
      // orthogonalized. X_MO^T H X_MO Q = \lambda X_MO^T M X_MO Q , or, H X =
      // \lambda M X if X = X_MO Q
      EigenSolverError
      solve(const OpContext &                           A,
            MultiVector<ValueTypeOperand, memorySpace> &X,
            std::vector<RealType> &                     eigenValues,
            MultiVector<ValueType, memorySpace> &       eigenVectors,
            bool computeEigenVectors = false);
 
      // In this case we solve the Kohn Sham GHEP if X's are othogonalized.
      // X_O^T H X_O Q = \lambda X_O^T M X_O Q , or,
      // H X = \lambda M X if X = X_O Q
      EigenSolverError
      solve(const OpContext &                           A,
            const OpContext &                           B,
            MultiVector<ValueTypeOperand, memorySpace> &X,
            std::vector<RealType> &                     eigenValues,
            MultiVector<ValueType, memorySpace> &       eigenVectors,
            bool computeEigenVectors = false);
 
      // /**  In this case we solve the Kohn Sham GHEP for any general X
      // * Performs cholesky factorization for orthogonalization internally.
      // * SConj = X^T M X
      // * SConj=LConj*L^{T}
      // * Lconj^{-1} compute
      // * compute HSConjProj= Lconj^{-1}*HConjProj*(Lconj^{-1})^C  (C denotes
      // *     conjugate transpose LAPACK notation)
      // * compute standard eigendecomposition HSConjProj: {QConjPrime,D}
      // * HSConjProj=QConjPrime*D*QConjPrime^{C} QConj={Lc^{-1}}^{C}*QConjPrime
      // *     rotate the basis in the subspace
      // * X^{T}={QConjPrime}^{C}*LConj^{-1}*X^{T}, stored in the column major
      // *     format In the above we use Q^{T}={QConjPrime}^{C}*LConj^{-1}
      // * In other words, X_O^T H X_O Q = \lambda X_O^T M X_O Q , or,
      // * H X = \lambda M X if X = X_O Q where X_O = X((LInv)^C) (C is conj
      // trans)
      // **/
      // EigenSolverError
      // solveGEPNoOrtho(const OpContext &                 A,
      //       const OpContext &                           B,
      //       MultiVector<ValueTypeOperand, memorySpace> &X,
      //       std::vector<RealType> &                     eigenValues,
      //       MultiVector<ValueType, memorySpace> &       eigenVectors,
      //       bool computeEigenVectors = false);
 
    private:
      void
      computeXTransOpX(MultiVector<ValueTypeOperand, memorySpace> &  X,
                       utils::MemoryStorage<ValueType, memorySpace> &S,
                       const OpContext &                             Op,
                       const bool &useBatched = true);
 
      void
      computeXTransOpX(MultiVector<ValueTypeOperand, memorySpace> &X,
                       const std::shared_ptr<const ProcessGrid> &  processGrid,
                       ScaLAPACKMatrix<ValueType> &overlapMatPar,
                       const OpContext &           Op);
 
      std::shared_ptr<MultiVector<ValueType, memorySpace>> d_XinBatchSmall,
        d_XinBatch, d_XoutBatchSmall, d_XoutBatch;
 
      size_type       d_batchSizeSmall;
      const size_type d_eigenVecBatchSize;
 
      const ElpaScalapackManager *d_elpaScala;
      const bool                  d_useELPA;
      const bool                  d_useScalapack;
 
    }; // end of class RayleighRitzEigenSolver
  }    // end of namespace linearAlgebra
} // end of namespace dftefe
#include <linearAlgebra/RayleighRitzEigenSolver.t.cpp>
#endif // dftefeRayleighRitzEigenSolver_h