Matrices |
Computations of eigen values and vectors are implemented in the following files:
analyziz/eigen_floating.hpp uses the latter solver with gradual precision doubling,
analyziz/eigen_ball.hpp includes certification with ball coefficients.
Mmx] |
use "analyziz"; type_mode? := true; |
Mmx] |
rnd () == { x == uniform_deviate (0.0, 1.0); return ball (x, 0.0000001); }; |
Mmx] |
M == [ rnd () | i in 0..3 || j in 0..3 ] |
:
Mmx] |
significant_digits := 5; // print 5 digits only eigen_solve M |
:
Mmx] |
v == eigenvalues M |
:
Mmx] |
radius v[0] |
: