Inheritance diagram for vtb::linalg::Symmetric_matrix< T >:
Collaboration diagram for vtb::linalg::Symmetric_matrix< T >:
Public Member Functions | |
| Lower_triangular_matrix_ldlt< T > | cholesky_indefinite () const |
| Vector< T > | eigen_values (T eps, int nsweeps) const |
| Compute eigen values. More... | |
 Public Member Functions inherited from vtb::linalg::Matrix< T > | |
| Matrix (base_type rhs) | |
| bool | is_square () const |
| bool | is_symmetric () const |
| bool | is_symmetric (T eps) const |
| bool | is_positive_definite () const |
| bool | is_toeplitz () const |
| bool | is_toeplitz (T eps) const |
| Square_matrix< T > | inverse () const |
| Square_matrix< T > & | inverse_self () |
| Matrix< T > | transpose () const |
| void | transpose_self () |
Detailed Description
template<class T>
class vtb::linalg::Symmetric_matrix< T >
Definition at line 179 of file linear_algebra.hh.
Member Function Documentation
◆ eigen_values()
template<class T >
| auto vtb::linalg::Symmetric_matrix< T >::eigen_values | ( | T | eps, |
| int | nsweeps | ||
| ) | const |
Compute eigen values.
- Date
- 2019-02-08
- Uses Jacobi algorithm for symmetric real matrices [golub2000eigen], [wang2018eigen].
Find \(a_{ij}\) with the greatest absolute value that do not lie on the diagonal.
Find sine and cosine by solving \(t^2 + 2\omega t - 1 = 0\), where \(\omega = (a_jj - a_ii) / 2a_ij\).
Rotate matrix.
Definition at line 159 of file linear_algebra.cc.
