libtrilinos-stokhos-dev 12.10.1-3 / usr / include / trilinos / Stokhos_RecurrenceBasis.hpp

// @HEADER
// ***********************************************************************
//
//                           Stokhos Package
//                 Copyright (2009) Sandia Corporation
//
// Under terms of Contract DE-AC04-94AL85000, there is a non-exclusive
// license for use of this work by or on behalf of the U.S. Government.
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// 1. Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
//
// 2. Redistributions in binary form must reproduce the above copyright
// notice, this list of conditions and the following disclaimer in the
// documentation and/or other materials provided with the distribution.
//
// 3. Neither the name of the Corporation nor the names of the
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY SANDIA CORPORATION "AS IS" AND ANY
// EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
// IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
// PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL SANDIA CORPORATION OR THE
// CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
// EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
// PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
// PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
// LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
// NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
// SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
//
// Questions? Contact Eric T. Phipps (etphipp@sandia.gov).
//
// ***********************************************************************
// @HEADER

#ifndef STOKHOS_RECURRENCEBASIS_HPP
#define STOKHOS_RECURRENCEBASIS_HPP

#include "Stokhos_OneDOrthogPolyBasis.hpp"

namespace Stokhos {

  //! Enumerated type for determining Smolyak growth policies
  enum GrowthPolicy {
    SLOW_GROWTH,
    MODERATE_GROWTH
  };

  /*!
   * \brief Implementation of OneDOrthogPolyBasis based on the general
   * three-term recurrence relationship:
   * \f[
   *    \gamma_{k+1}\psi_{k+1}(x) =
   *       (\delta_k x - \alpha_k)\psi_k(x) - \beta_k\psi_{k-1}(x)
   * \f]
   * for \f$k=0,\dots,P\f$ where \f$\psi_{-1}(x) = 0\f$,
   * \f$\psi_{0}(x) = 1/\gamma_0\f$,
   * and \f$\beta_{0} = 1 = \int d\lambda\f$.
   */
  /*!Derived classes implement the recurrence
   * relationship by implementing computeRecurrenceCoefficients().  If
   * \c normalize = \c true in the constructor, then the recurrence relationship
   * becomes:
   * \f[
   * \sqrt{\frac{\gamma_{k+1}\beta_{k+1}}{\delta_{k+1}\delta_k}} \psi_{k+1}(x) =
   *       (x - \alpha_k/\delta_k)\psi_k(x) -
   *       \sqrt{\frac{\gamma_k\beta_k}{\delta_k\delta_{k-1}}} \psi_{k-1}(x)
   * \f]
   * for \f$k=0,\dots,P\f$ where \f$\psi_{-1}(x) = 0\f$,
   * \f$\psi_{0}(x) = 1/\sqrt{\beta_0}\f$,
   * Note that a three term recurrence can always be defined with
   * \f$\gamma_k = \delta_k = 1\f$ in which case the polynomials are monic.
   * However typical normalizations of some polynomial families (see
   * Stokhos::LegendreBasis) require the extra terms.  Also, the quadrature
   * rule (points and weights) is the same regardless if the polynomials are
   * normalized.  However the normalization can affect other algorithms.
   */
  template <typename ordinal_type, typename value_type>
  class RecurrenceBasis :
    public OneDOrthogPolyBasis<ordinal_type, value_type> {
  public:

    //! Destructor
    virtual ~RecurrenceBasis();

    //! \name Implementation of Stokhos::OneDOrthogPolyBasis methods
    //@{

    //! Return order of basis (largest monomial degree \f$P\f$).
    virtual ordinal_type order() const;

    //! Return total size of basis (given by order() + 1).
    virtual ordinal_type size() const;

    //! Return array storing norm-squared of each basis polynomial
    /*!
     * Entry \f$l\f$ of returned array is given by \f$\langle\psi_l^2\rangle\f$
     * for \f$l=0,\dots,P\f$ where \f$P\f$ is given by order().
     */
    virtual const Teuchos::Array<value_type>& norm_squared() const;

    //! Return norm squared of basis polynomial \c i.
    virtual const value_type& norm_squared(ordinal_type i) const;

    //! Compute triple product tensor
    /*!
     * The \f$(i,j,k)\f$ entry of the tensor \f$C_{ijk}\f$ is given by
     * \f$C_{ijk} = \langle\Psi_i\Psi_j\Psi_k\rangle\f$ where \f$\Psi_l\f$
     * represents basis polynomial \f$l\f$ and \f$i,j=0,\dots,P\f$ where
     * \f$P\f$ is size()-1 and \f$k=0,\dots,p\f$ where \f$p\f$
     * is the supplied \c order.
     *
     * This method is implemented by computing \f$C_{ijk}\f$ using Gaussian
     * quadrature.
     */
    virtual Teuchos::RCP< Stokhos::Dense3Tensor<ordinal_type, value_type> >
    computeTripleProductTensor() const;

    //! Compute triple product tensor
    /*!
     * The \f$(i,j,k)\f$ entry of the tensor \f$C_{ijk}\f$ is given by
     * \f$C_{ijk} = \langle\Psi_i\Psi_j\Psi_k\rangle\f$ where \f$\Psi_l\f$
     * represents basis polynomial \f$l\f$ and \f$i,j=0,\dots,P\f$ where
     * \f$P\f$ is size()-1 and \f$k=0,\dots,p\f$ where \f$p\f$
     * is the supplied \c order.
     *
     * This method is implemented by computing \f$C_{ijk}\f$ using Gaussian
     * quadrature.
     */
    virtual
    Teuchos::RCP< Stokhos::Sparse3Tensor<ordinal_type, value_type> >
    computeSparseTripleProductTensor(ordinal_type order) const;

    //! Compute derivative double product tensor
    /*!
     * The \f$(i,j)\f$ entry of the tensor \f$B_{ij}\f$ is given by
     * \f$B_{ij} = \langle\psi_i'\psi_j\rangle\f$ where \f$\psi_l\f$
     * represents basis polynomial \f$l\f$ and \f$i,j=0,\dots,P\f$ where
     * \f$P\f$ is the order of the basis.
     *
     * This method is implemented by computing \f$B_{ij}\f$ using Gaussian
     * quadrature.
     */
    virtual Teuchos::RCP< Teuchos::SerialDenseMatrix<ordinal_type, value_type> > computeDerivDoubleProductTensor() const;

    //! Evaluate each basis polynomial at given point \c point
    /*!
     * Size of returned array is given by size(), and coefficients are
     * ordered from order 0 up to order order().
     */
    virtual void evaluateBases(const value_type& point,
                               Teuchos::Array<value_type>& basis_pts) const;

    /*!
     * \brief Evaluate basis polynomial given by order \c order at given
     * point \c point.
     */
    virtual value_type evaluate(const value_type& point,
                                ordinal_type order) const;

    //! Print basis to stream \c os
    virtual void print(std::ostream& os) const;

    //! Return string name of basis
    virtual const std::string& getName() const;

    /*!
     * \brief Compute quadrature points, weights, and values of
     * basis polynomials at given set of points \c points.
     */
    /*!
     * \c quad_order specifies the order to which the quadrature should be
     * accurate, not the number of quadrature points.  The number of points
     * is given by (\c quad_order + 1) / 2.   Note however the passed arrays
     * do NOT need to be sized correctly on input as they will be resized
     * appropriately.
     *
     * The quadrature points and weights are computed from the three-term
     * recurrence by solving a tri-diagional symmetric eigenvalue problem
     * (see Gene H. Golub and John H. Welsch, "Calculation of Gauss Quadrature
     * Rules", Mathematics of Computation, Vol. 23, No. 106 (Apr., 1969),
     * pp. 221-230).
     */
    virtual void
    getQuadPoints(ordinal_type quad_order,
                  Teuchos::Array<value_type>& points,
                  Teuchos::Array<value_type>& weights,
                  Teuchos::Array< Teuchos::Array<value_type> >& values) const;

    /*!
     * Return polynomial degree of exactness for a given number of quadrature
     * points.
     */
    virtual ordinal_type quadDegreeOfExactness(ordinal_type n) const;

    //! Evaluate coefficient growth rule for Smolyak-type bases
    virtual ordinal_type coefficientGrowth(ordinal_type n) const;

    //! Evaluate point growth rule for Smolyak-type bases
    virtual ordinal_type pointGrowth(ordinal_type n) const;

    //! Function pointer needed for level_to_order mappings
    typedef typename OneDOrthogPolyBasis<ordinal_type,value_type>::LevelToOrderFnPtr LevelToOrderFnPtr;

    //! Get sparse grid level_to_order mapping function
    /*!
     * Predefined functions are:
     *  webbur::level_to_order_linear_wn Symmetric Gaussian linear growth
     *  webbur::level_to_order_linear_nn Asymmetric Gaussian linear growth
     *  webbur::level_to_order_exp_cc    Clenshaw-Curtis exponential growth
     *  webbur::level_to_order_exp_gp    Gauss-Patterson exponential growth
     *  webbur::level_to_order_exp_hgk   Genz-Keister exponential growth
     *  webbur::level_to_order_exp_f2    Fejer-2 exponential growth
     */
    virtual LevelToOrderFnPtr getSparseGridGrowthRule() const {
      return sparse_grid_growth_rule; }

    //! Set sparse grid rule
    virtual void setSparseGridGrowthRule(LevelToOrderFnPtr ptr) {
      sparse_grid_growth_rule = ptr; }

    //@}

    //! Return recurrence coefficients defined by above formula
    virtual void getRecurrenceCoefficients(Teuchos::Array<value_type>& alpha,
                                           Teuchos::Array<value_type>& beta,
                                           Teuchos::Array<value_type>& delta,
                                           Teuchos::Array<value_type>& gamma) const;

    //! Evaluate basis polynomials and their derivatives at given point \c point
    virtual void evaluateBasesAndDerivatives(const value_type& point,
                                             Teuchos::Array<value_type>& vals,
                                             Teuchos::Array<value_type>& derivs) const;

    //! Set tolerance for zero in quad point generation
    virtual void setQuadZeroTol(value_type tol) {
      quad_zero_tol = tol; }

  protected:

    //! Constructor to be called by derived classes
    /*!
     * \c name is the name for the basis that will be displayed when
     * printing the basis, \c p is the order of the basis, \c normalize
     * indicates whether the basis polynomials should have unit-norm, and
     * \c quad_zero_tol is used to replace any quadrature point within this
     * tolerance with zero (which can help with duplicate removal in sparse
     * grid calculations).
     */
    RecurrenceBasis(const std::string& name, ordinal_type p, bool normalize,
                    GrowthPolicy growth = SLOW_GROWTH);

    //! Copy constructor with specified order
    RecurrenceBasis(ordinal_type p, const RecurrenceBasis& basis);

    //! Compute recurrence coefficients
    /*!
     * Derived classes should implement this method to compute their
     * recurrence coefficients.  \c n is the number of coefficients to compute.
     * Return value indicates whether coefficients correspond to normalized
     * (i.e., orthonormal) polynomials.
     *
     * Note:  Owing to the description above, \c gamma should be an array of
     * length n+1.
     */
    virtual bool
    computeRecurrenceCoefficients(ordinal_type n,
                                  Teuchos::Array<value_type>& alpha,
                                  Teuchos::Array<value_type>& beta,
                                  Teuchos::Array<value_type>& delta,
                                  Teuchos::Array<value_type>& gamma) const = 0;

    //! Setup basis after computing recurrence coefficients
    /*!
     * Derived classes should call this method after computing their recurrence
     * coefficients in their constructor to finish setting up the basis.
     */
    virtual void setup();

    //! Normalize coefficients
    void normalizeRecurrenceCoefficients(
      Teuchos::Array<value_type>& alpha,
      Teuchos::Array<value_type>& beta,
      Teuchos::Array<value_type>& delta,
      Teuchos::Array<value_type>& gamma) const;

  private:

    // Prohibit copying
    RecurrenceBasis(const RecurrenceBasis&);

    // Prohibit Assignment
    RecurrenceBasis& operator=(const RecurrenceBasis& b);

  protected:

    //! Name of basis
    std::string name;

    //! Order of basis
    ordinal_type p;

    //! Normalize basis
    bool normalize;

    //! Smolyak growth policy
    GrowthPolicy growth;

    //! Tolerance for quadrature points near zero
    value_type quad_zero_tol;

    //! Sparse grid growth rule (as determined by Pecos)
    LevelToOrderFnPtr sparse_grid_growth_rule;

    //! Recurrence \f$\alpha\f$ coefficients
    Teuchos::Array<value_type> alpha;

    //! Recurrence \f$\beta\f$ coefficients
    Teuchos::Array<value_type> beta;

    //! Recurrence \f$\delta\f$ coefficients
    Teuchos::Array<value_type> delta;

    //! Recurrence \f$\gamma\f$ coefficients
    Teuchos::Array<value_type> gamma;

    //! Norms
    Teuchos::Array<value_type> norms;

  }; // class RecurrenceBasis

} // Namespace Stokhos

// Include template definitions
#include "Stokhos_RecurrenceBasisImp.hpp"

#endif