Eigenvector calculation; to get dominant eigen vector
显示 更早的评论
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 个评论
Birdman
2017-10-27
Are you trying to verify
k*V=m*V*D
Yunus Emre Bülbül
2017-10-27
编辑：Yunus Emre Bülbül
2017-10-27
回答（1 个）
John D'Errico
2017-10-27
编辑：John D'Errico
2017-10-27
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.
类别
在 帮助中心 和 File Exchange 中查找有关 Eigenvalues & Eigenvectors 的更多信息
2017-10-27
2017-10-27
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!
Translated by 
