3 /** \page TopicCustomizingEigen Customizing/Extending Eigen
5 Eigen can be extended in several ways, for instance, by defining global methods, \ref ExtendingMatrixBase "by adding custom methods to MatrixBase", adding support to \ref CustomScalarType "custom types" etc.
9 \section ExtendingMatrixBase Extending MatrixBase (and other classes)
11 In this section we will see how to add custom methods to MatrixBase. Since all expressions and matrix types inherit MatrixBase, adding a method to MatrixBase make it immediately available to all expressions ! A typical use case is, for instance, to make Eigen compatible with another API.
13 You certainly know that in C++ it is not possible to add methods to an existing class. So how that's possible ? Here the trick is to include in the declaration of MatrixBase a file defined by the preprocessor token \c EIGEN_MATRIXBASE_PLUGIN:
17 #ifdef EIGEN_MATRIXBASE_PLUGIN
18 #include EIGEN_MATRIXBASE_PLUGIN
22 Therefore to extend MatrixBase with your own methods you just have to create a file with your method declaration and define EIGEN_MATRIXBASE_PLUGIN before you include any Eigen's header file.
24 You can extend many of the other classes used in Eigen by defining similarly named preprocessor symbols. For instance, define \c EIGEN_ARRAYBASE_PLUGIN if you want to extend the ArrayBase class. A full list of classes that can be extended in this way and the corresponding preprocessor symbols can be found on our page \ref TopicPreprocessorDirectives.
26 Here is an example of an extension file for adding methods to MatrixBase: \n
29 inline Scalar at(uint i, uint j) const { return this->operator()(i,j); }
30 inline Scalar& at(uint i, uint j) { return this->operator()(i,j); }
31 inline Scalar at(uint i) const { return this->operator[](i); }
32 inline Scalar& at(uint i) { return this->operator[](i); }
34 inline RealScalar squaredLength() const { return squaredNorm(); }
35 inline RealScalar length() const { return norm(); }
36 inline RealScalar invLength(void) const { return fast_inv_sqrt(squaredNorm()); }
38 template<typename OtherDerived>
39 inline Scalar squaredDistanceTo(const MatrixBase<OtherDerived>& other) const
40 { return (derived() - other.derived()).squaredNorm(); }
42 template<typename OtherDerived>
43 inline RealScalar distanceTo(const MatrixBase<OtherDerived>& other) const
44 { return internal::sqrt(derived().squaredDistanceTo(other)); }
46 inline void scaleTo(RealScalar l) { RealScalar vl = norm(); if (vl>1e-9) derived() *= (l/vl); }
48 inline Transpose<Derived> transposed() {return this->transpose();}
49 inline const Transpose<Derived> transposed() const {return this->transpose();}
51 inline uint minComponentId(void) const { int i; this->minCoeff(&i); return i; }
52 inline uint maxComponentId(void) const { int i; this->maxCoeff(&i); return i; }
54 template<typename OtherDerived>
55 void makeFloor(const MatrixBase<OtherDerived>& other) { derived() = derived().cwiseMin(other.derived()); }
56 template<typename OtherDerived>
57 void makeCeil(const MatrixBase<OtherDerived>& other) { derived() = derived().cwiseMax(other.derived()); }
59 const CwiseUnaryOp<internal::scalar_add_op<Scalar>, Derived>
60 operator+(const Scalar& scalar) const
61 { return CwiseUnaryOp<internal::scalar_add_op<Scalar>, Derived>(derived(), internal::scalar_add_op<Scalar>(scalar)); }
63 friend const CwiseUnaryOp<internal::scalar_add_op<Scalar>, Derived>
64 operator+(const Scalar& scalar, const MatrixBase<Derived>& mat)
65 { return CwiseUnaryOp<internal::scalar_add_op<Scalar>, Derived>(mat.derived(), internal::scalar_add_op<Scalar>(scalar)); }
68 Then one can the following declaration in the config.h or whatever prerequisites header file of his project:
70 #define EIGEN_MATRIXBASE_PLUGIN "MatrixBaseAddons.h"
73 \section InheritingFromMatrix Inheriting from Matrix
75 Before inheriting from Matrix, be really, i mean REALLY sure that using
76 EIGEN_MATRIX_PLUGIN is not what you really want (see previous section).
77 If you just need to add few members to Matrix, this is the way to go.
79 An example of when you actually need to inherit Matrix, is when you have
80 several layers of heritage such as MyVerySpecificVector1,MyVerySpecificVector1 -> MyVector1 -> Matrix and.
81 MyVerySpecificVector3,MyVerySpecificVector4 -> MyVector2 -> Matrix.
83 In order for your object to work within the %Eigen framework, you need to
84 define a few members in your inherited class.
86 Here is a minimalistic example:\n
88 class MyVectorType : public Eigen::VectorXd
91 MyVectorType(void):Eigen::VectorXd() {}
93 typedef Eigen::VectorXd Base;
95 // This constructor allows you to construct MyVectorType from Eigen expressions
96 template<typename OtherDerived>
97 MyVectorType(const Eigen::MatrixBase<OtherDerived>& other)
98 : Eigen::Vector3d(other)
101 // This method allows you to assign Eigen expressions to MyVectorType
102 template<typename OtherDerived>
103 MyVectorType & operator= (const Eigen::MatrixBase <OtherDerived>& other)
105 this->Base::operator=(other);
111 This is the kind of error you can get if you don't provide those methods
113 error: no match for ‘operator=’ in ‘delta =
114 (((Eigen::MatrixBase<Eigen::Matrix<std::complex<float>, 10000, 1, 2, 10000,
115 1> >*)(& delta)) + 8u)->Eigen::MatrixBase<Derived>::cwise [with Derived =
116 Eigen::Matrix<std::complex<float>, 10000, 1, 2, 10000,
117 1>]().Eigen::Cwise<ExpressionType>::operator* [with OtherDerived =
118 Eigen::Matrix<std::complex<float>, 10000, 1, 2, 10000, 1>, ExpressionType =
119 Eigen::Matrix<std::complex<float>, 10000, 1, 2, 10000, 1>](((const
120 Eigen::MatrixBase<Eigen::Matrix<std::complex<float>, 10000, 1, 2, 10000, 1>
121 >&)(((const Eigen::MatrixBase<Eigen::Matrix<std::complex<float>, 10000, 1,
122 >2, 10000, 1> >*)((const spectral1d*)where)) + 8u)))’
125 \anchor user_defined_scalars \section CustomScalarType Using custom scalar types
127 By default, Eigen currently supports standard floating-point types (\c float, \c double, \c std::complex<float>, \c std::complex<double>, \c long \c double), as well as all native integer types (e.g., \c int, \c unsigned \c int, \c short, etc.), and \c bool.
128 On x86-64 systems, \c long \c double permits to locally enforces the use of x87 registers with extended accuracy (in comparison to SSE).
130 In order to add support for a custom type \c T you need:
131 -# make sure the common operator (+,-,*,/,etc.) are supported by the type \c T
132 -# add a specialization of struct Eigen::NumTraits<T> (see \ref NumTraits)
133 -# define the math functions that makes sense for your type. This includes standard ones like sqrt, pow, sin, tan, conj, real, imag, etc, as well as abs2 which is Eigen specific.
134 (see the file Eigen/src/Core/MathFunctions.h)
136 The math function should be defined in the same namespace than \c T, or in the \c std namespace though that second approach is not recommended.
138 Here is a concrete example adding support for the Adolc's \c adouble type. <a href="https://projects.coin-or.org/ADOL-C">Adolc</a> is an automatic differentiation library. The type \c adouble is basically a real value tracking the values of any number of partial derivatives.
141 #ifndef ADOLCSUPPORT_H
142 #define ADOLCSUPPORT_H
144 #define ADOLC_TAPELESS
145 #include <adolc/adouble.h>
146 #include <Eigen/Core>
150 template<> struct NumTraits<adtl::adouble>
151 : NumTraits<double> // permits to get the epsilon, dummy_precision, lowest, highest functions
153 typedef adtl::adouble Real;
154 typedef adtl::adouble NonInteger;
155 typedef adtl::adouble Nested;
161 RequireInitialization = 1,
172 inline const adouble& conj(const adouble& x) { return x; }
173 inline const adouble& real(const adouble& x) { return x; }
174 inline adouble imag(const adouble&) { return 0.; }
175 inline adouble abs(const adouble& x) { return fabs(x); }
176 inline adouble abs2(const adouble& x) { return x*x; }
180 #endif // ADOLCSUPPORT_H
184 \sa \ref TopicPreprocessorDirectives