Eigenvector calculation; to get dominant eigen vector
6 vues (au cours des 30 derniers jours)
Afficher commentaires plus anciens
I have a question regarding to eigenvector. The below simple script results in wrong eigenvectors.
mscalar=100; %kips/g
%sqrt(k/m) is calculated as 55.187
%kscalar
kscalar=mscalar*55.187^2;
k=kscalar*[2 -1 0 0 0;-1 2 -1 0 0;0 -1 2 -1 0;0 0 -1 2 -1;0 0 0 -1 1];
m=mscalar*[1 0 0 0 0;0 1 0 0 0;0 0 1 0 0;0 0 0 1 0;0 0 0 0 1];
[V,D]=eig(k,m);
V
sqrt(D)
2 commentaires
Réponses (1)
John D'Errico
le 27 Oct 2017
Modifié(e) : John D'Errico
le 27 Oct 2017
The immediate answer is that eigenvectors are defined only to within a constant multiplier. Multiply an eigenvector by ANY constant, and it still satisfies the classic relationship
A*v = lambda*v
eig returns eigenvectors normalized to have unit norm, which is pretty standard. But in fact, there can always be an arbitrary factor of -1 in there. Just flip the signs on some of your eigenvectors, and nothing changes. The result is still completely valid. So sometimes you just get the wrong sign. People are always confused by that.
Note: If some of the eigenvalues had multiplicity greater than 1, there are other reasons why you can get differing eigenvectors. But that does not happen here.
0 commentaires
Voir également
Catégories
En savoir plus sur Linear Algebra dans Help Center et File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!