1c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath// This file is part of Eigen, a lightweight C++ template library
2c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath// for linear algebra.
3c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath//
4c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath// Copyright (C) 2009-2010 Gael Guennebaud <gael.guennebaud@inria.fr>
5c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath//
6c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath// This Source Code Form is subject to the terms of the Mozilla
7c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath// Public License v. 2.0. If a copy of the MPL was not distributed
8c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath// with this file, You can obtain one at http://mozilla.org/MPL/2.0/.
9c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
10c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath#ifndef EIGEN_SELFCWISEBINARYOP_H
11c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath#define EIGEN_SELFCWISEBINARYOP_H
12c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
13c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamathnamespace Eigen {
14c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
15c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath/** \class SelfCwiseBinaryOp
16c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  * \ingroup Core_Module
17c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  *
18c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  * \internal
19c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  *
20c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  * \brief Internal helper class for optimizing operators like +=, -=
21c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  *
22c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  * This is a pseudo expression class re-implementing the copyCoeff/copyPacket
23c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  * method to directly performs a +=/-= operations in an optimal way. In particular,
24c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  * this allows to make sure that the input/output data are loaded only once using
25c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  * aligned packet loads.
26c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  *
27c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  * \sa class SwapWrapper for a similar trick.
28c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  */
29c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
30c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamathnamespace internal {
31c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamathtemplate<typename BinaryOp, typename Lhs, typename Rhs>
32c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamathstruct traits<SelfCwiseBinaryOp<BinaryOp,Lhs,Rhs> >
33c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  : traits<CwiseBinaryOp<BinaryOp,Lhs,Rhs> >
34c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath{
35c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  enum {
36c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    // Note that it is still a good idea to preserve the DirectAccessBit
37c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    // so that assign can correctly align the data.
38c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    Flags = traits<CwiseBinaryOp<BinaryOp,Lhs,Rhs> >::Flags | (Lhs::Flags&DirectAccessBit) | (Lhs::Flags&LvalueBit),
39c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    OuterStrideAtCompileTime = Lhs::OuterStrideAtCompileTime,
40c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    InnerStrideAtCompileTime = Lhs::InnerStrideAtCompileTime
41c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  };
42c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath};
43c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath}
44c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
45c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamathtemplate<typename BinaryOp, typename Lhs, typename Rhs> class SelfCwiseBinaryOp
46c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  : public internal::dense_xpr_base< SelfCwiseBinaryOp<BinaryOp, Lhs, Rhs> >::type
47c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath{
48c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  public:
49c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
50c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    typedef typename internal::dense_xpr_base<SelfCwiseBinaryOp>::type Base;
51c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    EIGEN_DENSE_PUBLIC_INTERFACE(SelfCwiseBinaryOp)
52c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
53c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    typedef typename internal::packet_traits<Scalar>::type Packet;
54c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
55c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    inline SelfCwiseBinaryOp(Lhs& xpr, const BinaryOp& func = BinaryOp()) : m_matrix(xpr), m_functor(func) {}
56c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
57c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    inline Index rows() const { return m_matrix.rows(); }
58c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    inline Index cols() const { return m_matrix.cols(); }
59c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    inline Index outerStride() const { return m_matrix.outerStride(); }
60c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    inline Index innerStride() const { return m_matrix.innerStride(); }
61c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    inline const Scalar* data() const { return m_matrix.data(); }
62c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
63c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    // note that this function is needed by assign to correctly align loads/stores
64c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    // TODO make Assign use .data()
65c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    inline Scalar& coeffRef(Index row, Index col)
66c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    {
67c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      EIGEN_STATIC_ASSERT_LVALUE(Lhs)
68c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      return m_matrix.const_cast_derived().coeffRef(row, col);
69c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    }
70c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    inline const Scalar& coeffRef(Index row, Index col) const
71c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    {
72c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      return m_matrix.coeffRef(row, col);
73c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    }
74c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
75c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    // note that this function is needed by assign to correctly align loads/stores
76c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    // TODO make Assign use .data()
77c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    inline Scalar& coeffRef(Index index)
78c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    {
79c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      EIGEN_STATIC_ASSERT_LVALUE(Lhs)
80c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      return m_matrix.const_cast_derived().coeffRef(index);
81c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    }
82c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    inline const Scalar& coeffRef(Index index) const
83c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    {
84c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      return m_matrix.const_cast_derived().coeffRef(index);
85c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    }
86c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
87c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    template<typename OtherDerived>
88c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    void copyCoeff(Index row, Index col, const DenseBase<OtherDerived>& other)
89c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    {
90c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      OtherDerived& _other = other.const_cast_derived();
91c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      eigen_internal_assert(row >= 0 && row < rows()
92c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath                         && col >= 0 && col < cols());
93c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      Scalar& tmp = m_matrix.coeffRef(row,col);
94c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      tmp = m_functor(tmp, _other.coeff(row,col));
95c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    }
96c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
97c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    template<typename OtherDerived>
98c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    void copyCoeff(Index index, const DenseBase<OtherDerived>& other)
99c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    {
100c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      OtherDerived& _other = other.const_cast_derived();
101c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      eigen_internal_assert(index >= 0 && index < m_matrix.size());
102c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      Scalar& tmp = m_matrix.coeffRef(index);
103c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      tmp = m_functor(tmp, _other.coeff(index));
104c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    }
105c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
106c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    template<typename OtherDerived, int StoreMode, int LoadMode>
107c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    void copyPacket(Index row, Index col, const DenseBase<OtherDerived>& other)
108c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    {
109c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      OtherDerived& _other = other.const_cast_derived();
110c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      eigen_internal_assert(row >= 0 && row < rows()
111c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath                        && col >= 0 && col < cols());
112c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      m_matrix.template writePacket<StoreMode>(row, col,
113c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath        m_functor.packetOp(m_matrix.template packet<StoreMode>(row, col),_other.template packet<LoadMode>(row, col)) );
114c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    }
115c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
116c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    template<typename OtherDerived, int StoreMode, int LoadMode>
117c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    void copyPacket(Index index, const DenseBase<OtherDerived>& other)
118c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    {
119c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      OtherDerived& _other = other.const_cast_derived();
120c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      eigen_internal_assert(index >= 0 && index < m_matrix.size());
121c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      m_matrix.template writePacket<StoreMode>(index,
122c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath        m_functor.packetOp(m_matrix.template packet<StoreMode>(index),_other.template packet<LoadMode>(index)) );
123c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    }
124c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
125c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    // reimplement lazyAssign to handle complex *= real
126c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    // see CwiseBinaryOp ctor for details
127c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    template<typename RhsDerived>
128c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    EIGEN_STRONG_INLINE SelfCwiseBinaryOp& lazyAssign(const DenseBase<RhsDerived>& rhs)
129c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    {
130c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      EIGEN_STATIC_ASSERT_SAME_MATRIX_SIZE(Lhs,RhsDerived)
131c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      EIGEN_CHECK_BINARY_COMPATIBILIY(BinaryOp,typename Lhs::Scalar,typename RhsDerived::Scalar);
132c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
133c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    #ifdef EIGEN_DEBUG_ASSIGN
134c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      internal::assign_traits<SelfCwiseBinaryOp, RhsDerived>::debug();
135c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    #endif
136c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      eigen_assert(rows() == rhs.rows() && cols() == rhs.cols());
137c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      internal::assign_impl<SelfCwiseBinaryOp, RhsDerived>::run(*this,rhs.derived());
138c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    #ifndef EIGEN_NO_DEBUG
139c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      this->checkTransposeAliasing(rhs.derived());
140c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    #endif
141c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      return *this;
142c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    }
143c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
144c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    // overloaded to honor evaluation of special matrices
145c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    // maybe another solution would be to not use SelfCwiseBinaryOp
146c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    // at first...
147c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    SelfCwiseBinaryOp& operator=(const Rhs& _rhs)
148c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    {
149c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      typename internal::nested<Rhs>::type rhs(_rhs);
150c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      return Base::operator=(rhs);
151c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    }
152c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
153c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    Lhs& expression() const
154c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    {
155c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      return m_matrix;
156c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    }
157c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
158c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    const BinaryOp& functor() const
159c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    {
160c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath      return m_functor;
161c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    }
162c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
163c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  protected:
164c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    Lhs& m_matrix;
165c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    const BinaryOp& m_functor;
166c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
167c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  private:
168c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath    SelfCwiseBinaryOp& operator=(const SelfCwiseBinaryOp&);
169c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath};
170c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
171c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamathtemplate<typename Derived>
172c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamathinline Derived& DenseBase<Derived>::operator*=(const Scalar& other)
173c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath{
174c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  typedef typename Derived::PlainObject PlainObject;
175c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  SelfCwiseBinaryOp<internal::scalar_product_op<Scalar>, Derived, typename PlainObject::ConstantReturnType> tmp(derived());
176c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  tmp = PlainObject::Constant(rows(),cols(),other);
177c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  return derived();
178c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath}
179c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
180c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamathtemplate<typename Derived>
181c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamathinline Derived& DenseBase<Derived>::operator/=(const Scalar& other)
182c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath{
183c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  typedef typename internal::conditional<NumTraits<Scalar>::IsInteger,
184c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath                                        internal::scalar_quotient_op<Scalar>,
185c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath                                        internal::scalar_product_op<Scalar> >::type BinOp;
186c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  typedef typename Derived::PlainObject PlainObject;
187c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  SelfCwiseBinaryOp<BinOp, Derived, typename PlainObject::ConstantReturnType> tmp(derived());
1887faaa9f3f0df9d23790277834d426c3d992ac3baCarlos Hernandez  Scalar actual_other;
1897faaa9f3f0df9d23790277834d426c3d992ac3baCarlos Hernandez  if(NumTraits<Scalar>::IsInteger)  actual_other = other;
1907faaa9f3f0df9d23790277834d426c3d992ac3baCarlos Hernandez  else                              actual_other = Scalar(1)/other;
1917faaa9f3f0df9d23790277834d426c3d992ac3baCarlos Hernandez  tmp = PlainObject::Constant(rows(),cols(), actual_other);
192c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath  return derived();
193c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath}
194c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
195c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath} // end namespace Eigen
196c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath
197c981c48f5bc9aefeffc0bcb0cc3934c2fae179ddNarayan Kamath#endif // EIGEN_SELFCWISEBINARYOP_H
198