Archived

This topic is now archived and is closed to further replies.

Eigenvalues

This topic is 5113 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

That link doesn''t provide an explaination of how to find the eigen values/vectors, only one use for them...

In any case, to find the eigen values you have to solve the characteristic equation. So, if we have a matrix A, its characteristic equation is (using Y for lamda):

det(A - YI)

which will yield a polynomial (note: det(YI - A) also works). Find the roots and you have your eigen values. To find the eigen vectors you have to solve the system of equations given by:

(A - YI)v = 0

for each Y. This can be done using Gauss-Jordan elimination.

Share this post


Link to post
Share on other sites
step by step? there´s not many steps involved (I´ll take the notation of above post):

1) Av = yv (definition of eigenvalue y of eigenvector v)
2) Av = yv = y*I*v (multiplication by one)
3) => Av - yIv = 0 ( by applying -yIv to the equation)
4) => (A - yI)v = 0 (v in not zero by definition -you couldn´t define an eigenvalue then- so the matrix A-yI must not be regular (det != 0) for the equation to have a solution.)
5) => det(A-yI) = 0
6) Solve the polynomial (5) to obtain a set of eigenvalues {y}.
7) Plug the eigenvalues y into (4) and solve the resulting linear system of equations to get the corresponding eigenvectors. Be aware that while v=0 looks like a valid result you allready skipped that in step (4)->(5).

btw.: I should have read the link so I could have saved my time writing this because from the link it´s one click to what I just wrote.

Share this post


Link to post
Share on other sites