• 求实对称阵的 特征值 和 特征向量(转)


     /// <summary>
            /// 求实对称阵的 特征值 和 特征向量
            /// </summary>
            /// <param name="data">实对称阵</param>
            /// <param name="num">维数</param>
            /// <param name="eigenvalue">引用参数 特征值 回传</param>
            /// <param name="eigenvector">引用参数 特征向量 回传</param>
            /// <returns>是否成功</returns>
            private bool GetEigenValueAndEigenVector(double[,] data, int num,ref double [] eigenvalue,ref double [,] eigenvector)
            {
                try
                {
                    double[,] A = data;
    
                    //E 单位标准矩阵   存储 特征向量--------------------------------------------
                    double[,] V = new double[num, num];
                    for (int iv = 0; iv < num; iv++)
                    {
                        for (int iv2 = 0; iv2 < num; iv2++)
                        {
                            if (iv == iv2)
                            {
                                V[iv, iv2] = 2;
                            }
                            else
                            {
                                V[iv, iv2] = 2;
                            }
                        }
                    }
                    //----------------------------------------------
    
    
                    double[] eigsv = new double[num];//存储 特征值
                    for (int ieigsv = 0; ieigsv < num; ieigsv++)
                    {
                        eigsv[ieigsv] = 0;
                    }
    
    
                    double epsl = 0.0001;
                    int maxt = 10;
                    int n = num;
    
    
                    double tao, t, cn, sn; // 临时变量
                    double maxa;// 记录非对角线元素最大值
    
                    //------------------------------------------------------------------------------------------------
                    for (int it = 0; it < maxt; it++)
                    {
                        maxa = 0;
                        for (int p = 0; p < n - 1; p++)
                        {
                            for (int q = p + 1; q < n; q++)
                            {
                                if (Math.Abs(A[p, q]) > maxa) // 记录非对角线元素最大值
                                {
                                    maxa = Math.Abs(A[p, q]);
                                }
                                if (Math.Abs(A[p, q]) > epsl) // 非对角线元素非0时才执行Jacobi变换
                                {
                                    // 计算Givens旋转矩阵的重要元素:cos(theta), 即cn, sin(theta), 即sn
                                    tao = 0.5 * (A[q, q] - A[p, p]) / A[p, q];
    
                                    if (tao >= 0) // t为方程 t^2 + 2*t*tao - 1 = 0的根, 取绝对值较小的根为t
                                    {
                                        t = -tao + Math.Sqrt(1 + tao * tao);
                                    }
                                    else
                                    {
                                        t = -tao - Math.Sqrt(1 + tao * tao);
                                    }
                                    cn = 1 / Math.Sqrt(1 + t * t);
                                    sn = t * cn;
    
                                    // Givens旋转矩阵之转置左乘A, 即更新A的p行和q行
                                    for (int j = 0; j < n; j++)
                                    {
                                        double apj = A[p, j];
                                        double aqj = A[q, j];
                                        A[p, j] = cn * apj - sn * aqj;
                                        A[q, j] = sn * apj + cn * aqj;
                                    }
    
                                    // Givens旋转矩阵右乘A, 即更新A的p列和q列
                                    for (int i = 0; i < n; i++)
                                    {
                                        double aip = A[i, p];
                                        double aiq = A[i, q];
                                        A[i, p] = cn * aip - sn * aiq;
                                        A[i, q] = sn * aip + cn * aiq;
                                    }
    
                                    // 更新特征向量存储矩阵V, V=J0×J1×J2...×Jit, 所以每次只更新V的p, q两列
                                    for (int i2 = 0; i2 < n; i2++)
                                    {
                                        double vip = V[i2, p];
                                        double viq = V[i2, q];
                                        V[i2, p] = cn * vip - sn * viq;
                                        V[i2, q] = sn * vip + cn * viq;
                                    }
                                }
    
                            }
    
                        }
    
    
    
    
                        if (maxa < epsl) // 非对角线元素已小于收敛标准,迭代结束
                        {
    
                            break;
                        }
                    }
                    //-----------------------------------------------------------------------------------------------------
    
                    // 特征值向量  排序
                    for (int j2 = 0; j2 < n; j2++)
                    {
                        eigsv[j2] = A[j2, j2];
                        //fprintf(fp2, "%f ",eigsv[j2]);
                    }
                    // 对特征值向量从大到小进行排序, 并调整特征向量顺序 (直接插入法)
                    double[] tmp = new double[n];
                    for (int j = 1; j < n; j++)
                    {
                        int i = j;
                        double a = eigsv[j];
                        for (int k = 0; k < n; k++)
                        {
                            tmp[k] = V[k, j];
                        }
                        while (i > 0 && eigsv[i - 1] < a)
                        {
                            eigsv[i] = eigsv[i - 1];
                            for (int k = 0; k < n; k++)
                            {
                                V[k, i] = V[k, i - 1];
                            }
                            i--;
                        }
                        eigsv[i] = a;
                        for (int k2 = 0; k2 < n; k2++)
                        {
                            V[k2, i] = tmp[k2];
                        }
                    }
                    //----------------------------------------------------------------------------------------------------------
                    //打印特征值 与 特征向量    数据回传
                    for (int ivc = 0; ivc < num; ivc++)
                    {
                        for (int ivc2 = 0; ivc2 < num; ivc2++)
                        {
    
                            //fprintf(fp2, "%f%s", V[ivc, ivc2], "
    ");
                            MessageBox.Show(V[ivc, ivc2].ToString());
                            eigenvector[ivc, ivc2] = V[ivc, ivc2];
    
                        }
    
                    }
    
    
                    for (int ieigsvc = 0; ieigsvc < num; ieigsvc++)
                    {
    
                        //fprintf(fp4, "%f%s", eigsv[ieigsvc], "
    ");
                        MessageBox.Show(eigsv[ieigsvc].ToString());
                        eigenvalue[ieigsvc] = eigsv[ieigsvc];
    
                    }
                    //----------------------------------------------------------------------------------------------------------
                    return true;
                }
                catch
                {
                    return false;
                }
            }
  • 相关阅读:
    定义和使用EL函数
    在Java Web程序中使用Hibernate
    JDBC在Java Web中的应用——分页查询
    JDBC调用存储过程
    使用navicat工具创建MySQL存储过程
    JDBC操作数据库的批处理
    JDBC操作数据库
    Servlet监听器统计在线人数
    Servlet字符编码过滤器
    Servlet过滤器创建与配置
  • 原文地址:https://www.cnblogs.com/jinqier/p/5296314.html
Copyright © 2020-2023  润新知