Eigen用法_eigen 矩阵使用pow-优快云博客

这篇博客详细介绍了Eigen库的使用，包括矩阵和向量的初始化、赋值、算术运算、属性计算以及相似度度量。示例中展示了如何创建不同尺寸的矩阵和向量，进行加减乘除、转置、共轭、迹、乘积等操作，并计算Pearson相关系数和余弦相似度。此外，还涉及了向量的范数和统计属性的计算。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

Eigen网址：

https://eigen.tuxfamily.org/dox/classEigen_1_1MatrixBase.html

可以查阅相关类的介绍以及API

使用：

51     //initialize
 52     std::map<int, Eigen::VectorXf> mp;
 53     mp[1].resize(400);
 54 
 55     Eigen::VectorXf x1;
 56     cout << "size:" << x1.size() << endl;
 57 
 58     Eigen::Vector2d u(-2, 3), v(4, 5);
 59 
 60     Eigen::MatrixXd m(2, 3);//2行3列
 61     m << 1, 11, 12,         //初始化
 62       2, 22, 23;
 63 
 64     //assign
 65     m(0, 0) = 2;
 66     Eigen::VectorXf x;
 67     x.resize(3);
 68     for (int i = 0; i < 2; ++i)
 69         x(i, 0) = i;
 70     cout << "size:" << x.size() << endl;
 71     cout << "data:" << *(x.data()) << " " << *(x.data()+1)  << endl;
 72     x.resize(0);
 73     cout << "size:" << x.size() << endl;    

76     //arithmetic
 77     cout << "+\n" << u + v << endl;;
 78     cout << "-\n" << u - v << endl;;
 79     cout << "*\n" << u * 2 << endl;
 80     cout << "/\n" << u / 2 << endl;
 81 
 82     cout << "T\n" << u.transpose() << endl;
 83     cout << "-\n" << u.conjugate() << endl;
 84     cout << "*\n" << u.adjoint() << endl;
 85 
 86     //multiplication
 87     cout << "transpose\n" << u.transpose() * v << endl;
 88     cout << "dot\n" << u.dot(v) << endl;
 89 
 90     //math
 91     cout << "sum\n" << u.sum() << endl;
 92     cout << "mean\n" << u.mean() << endl;
 93     cout << "trace\n" << u.trace() << endl;//对角线和
 94     cout << "prod\n" << u.prod() << endl;//各元素间乘积
  95 
 96     //Similarity
 97     cout << "lpNorm---------" << endl;
 98     cout << (u - v).lpNorm<1>() << endl;
 99     cout << (u - v).lpNorm<2>() << endl;
100     cout << (u - v).lpNorm<Eigen::Infinity>() << endl;
101 
102     cout << "dot---------" << endl;
103     cout << u.dot(v) << endl;

105     //pearson
106     float mean1 = u.mean();
107     float mean2 = v.mean();
108     auto r1 = (u.array() - mean1);
109     auto r2 = (v.array() - mean2);
110     float t_dot = r1.matrix().dot(r2.matrix());
111 
112     Eigen::Vector2d t_v1 (r1.pow(2));
113     Eigen::Vector2d t_v2 (r2.pow(2));
114 
115     cout << "pearson---------" << endl;
116     float t_dot2 = std::sqrt(t_v1.sum() * t_v2.sum());
119     t_dot2 = std::sqrt(t_dot2);
120     cout << "t_dot2 " << t_dot2 << endl;
121     float res = t_dot / t_dot2;
122     cout << res << endl;


124     //cos
125     float c_v = u.dot(v);
126     Eigen::Vector2d t_c1 (r1.pow(2));
127     Eigen::Vector2d t_c2 (r2.pow(2));
128     cout << "cos---------" << endl;
129     cout << c_v / std::sqrt(t_c1.sum() * t_c2.sum()) << endl;

类关系：

Matrix类，继承于PlainObjectBase；

PlainObjectBase最终会使用MatrixBase/ArrayBase

MatrixBase继承于DenseBase

因此看接口的话，可以看DenseBase,MatrixBase,Matrix类