1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
|
namespace Eigen {
/** \page QuickStartGuide
<h1>Quick start guide</h1>
<h2>Simple example with fixed-size matrices and vectors</h2>
By fixed-size, we mean that the number of rows and columns are known at compile-time. In this case, Eigen avoids dynamic memory allocation and unroll loops. This is useful for very small sizes (typically up to 4x4).
<table><tr><td>
\include Tutorial_simple_example_fixed_size.cpp
</td>
<td>
output:
\include Tutorial_simple_example_fixed_size.out
</td></tr></table>
<h2>Simple example with dynamic-size matrices and vectors</h2>
Dynamic-size means that the number of rows and columns are not known at compile-time. In this case, they are stored as runtime variables and the arrays are dynamically allocated.
<table><tr><td>
\include Tutorial_simple_example_dynamic_size.cpp
</td>
<td>
output:
\include Tutorial_simple_example_dynamic_size.out
</td></tr></table>
<h2>Matrix and vector types</h2>
In Eigen, all kinds of dense matrices and vectors are represented by the template class Matrix. In most cases you can simply use one of the several convenient typedefs (\ref matrixtypedefs).
The template class Matrix takes a number of template parameters, but for now it is enough to understand the 3 first ones (and the others can then be left unspecified):
\code Matrix<Scalar, RowsAtCompileTime, ColsAtCompileTime> \endcode
\li \c Scalar is the scalar type, i.e. the type of the coefficients. That is, if you want a vector of floats, choose \c float here.
\li \c RowsAtCompileTime and \c ColsAtCompileTime are the number of rows and columns of the matrix as known at compile-time.
For example, \c Vector3d is a typedef for \code Matrix<double, 3, 1> \endcode.
What if the matrix has dynamic-size i.e. the number of rows or cols isn't known at compile-time? Then use the special value Eigen::Dynamic. For example, \c VectorXd is a typedef for \code Matrix<double, Dynamic, 1> \endcode.
<h2>Matrix and vector creation and initialization</h2>
For instance \code Matrix3f m = Matrix3f::Identity(); \endcode creates a 3x3 fixed size matrix of float
which is initialized to the identity matrix.
Similarly \code MatrixXcd m = MatrixXcd::Zero(rows,cols); \endcode creates a rows x cols matrix
of double precision complex which is initialized to zero. Here rows and cols do not have to be
known at compile-time. In "MatrixXcd", "X" stands for dynamic, "c" for complex, and "d" for double.
You can also initialize a matrix with all coefficients equal to one:
\code MatrixXi m = MatrixXi::Ones(rows,cols); \endcode
or to any constant value:
\code
MatrixXi m = MatrixXi::Constant(rows,cols,66);
Matrix4d m = Matrix4d::Constant(6.6);
\endcode
All these 4 matrix creation functions also exist with the "set" prefix:
\code
Matrix3f m3; MatrixXi mx; VectorXcf vec;
m3.setZero(); mx.setZero(rows,cols); vec.setZero(size);
m3.setIdentity(); mx.setIdentity(rows,cols); vec.setIdentity(size);
m3.setOnes(); mx.setOnes(rows,cols); vec.setOnes(size);
m3.setConstant(6.6); mx.setConstant(rows,cols,6.6); vec.setConstant(size,complex<float>(6,3));
\endcode
Finally, all the coefficients of a matrix can set using the comma initializer syntax:
<table><tr><td>
\include Tutorial_commainit_01.cpp
</td>
<td>
output:
\verbinclude Tutorial_commainit_01.out
</td></tr></table>
Eigen's comma initializer also allows to set the matrix per block making it much more powerful:
<table><tr><td>
\include Tutorial_commainit_02.cpp
</td>
<td>
output with rows=cols=5:
\verbinclude Tutorial_commainit_02.out
</td></tr></table>
<h2>Basic Linear Algebra</h2>
As long as you use mathematically well defined operators, you can basically write your matrix
and vector expressions using standard arithmetic operators:
\code
mat1 = mat1*1.5 + mat2 * mat3/4;
\endcode
\b dot \b product (inner product):
\code
scalar = vec1.dot(vec2);
\endcode
\b outer \b product:
\code
mat = vec1 * vec2.transpose();
\endcode
\b cross \b product: The cross product is defined in the Geometry module, you therefore have to include it first:
\code
#include <Eigen/Geometry>
vec3 = vec1.cross(vec2);
\endcode
By default, Eigen's only allows mathematically well defined operators.
However, thanks to the .cwise() operator prefix, Eigen's matrices also provide
a very powerful numerical container supporting most common coefficient wise operators:
* Coefficient wise product: \code mat3 = mat1.cwise() * mat2; \endcode
* Coefficient wise division: \code mat3 = mat1.cwise() / mat2; \endcode
* Coefficient wise reciprocal: \code mat3 = mat1.cwise().inverse(); \endcode
* Add a scalar to a matrix: \code mat3 = mat1.cwise() + scalar; \endcode
* Coefficient wise comparison: \code mat3 = mat1.cwise() < mat2; \endcode
* Finally, \c .cwise() offers many common numerical functions including abs, pow, exp, sin, cos, tan, e.g.:
\code mat3 = mat1.cwise().sin(); \endcode
<h2>Reductions</h2>
\code
scalar = mat.sum(); scalar = mat.norm(); scalar = mat.minCoeff();
vec = mat.colwise().sum(); vec = mat.colwise().norm(); vec = mat.colwise().minCoeff();
vec = mat.rowwise().sum(); vec = mat.rowwise().norm(); vec = mat.rowwise().minCoeff();
\endcode
Other natively supported reduction operations include maxCoeff(), norm2(), all() and any().
<h2>Sub matrices</h2>
<h2>Geometry features</h2>
<h2>Notes on performances</h2>
<h2>Advanced Linear Algebra</h2>
<h3>Solving linear problems</h3>
<h3>LU</h3>
<h3>Cholesky</h3>
<h3>QR</h3>
<h3>Eigen value problems</h3>
*/
}
|
