# Help with Eigenvectors

Printable View

• Nov 12th 2011, 06:08 AM
hfullwood
Help with Eigenvectors
Hello,

Was wondering if anyone could help, i am trying to work out eigenvectors after working out the 3 eigenvalues for the below matrix:
[5 3 2
1 4 6
9 7 3]
My eigenvectors are correct however for 2 of the eigenvalues, my eigenvectors have opposite signs.
I have checked these in MathCAD so i know that the eigenvectors are correct however for some reason some of the signs are opposite.
I have attached my workings for the eigenvalue 1.65, the eigenvectors shown have opposite signs, can anyone spot why.

Thanks so much for your help.

Attachment 22720Attachment 22719
• Nov 12th 2011, 07:33 AM
Opalg
Re: Help with Eigenvectors
Quote:

Originally Posted by hfullwood
Hello,

Was wondering if anyone could help, i am trying to work out eigenvectors after working out the 3 eigenvalues for the below matrix:
[5 3 2
1 4 6
9 7 3]
My eigenvectors are correct however for 2 of the eigenvalues, my eigenvectors have opposite signs.
I have checked these in MathCAD so i know that the eigenvectors are correct however for some reason some of the signs are opposite.
I have attached my workings for the eigenvalue 1.65, the eigenvectors shown have opposite signs, can anyone spot why.

Thanks so much for your help.

Attachment 22720Attachment 22719

If x is an eigenvector then so is –x. Both answers are equally correct.
• Nov 12th 2011, 07:56 AM
hfullwood
Re: Help with Eigenvectors
Thanks so much for your reply.
Do you know if ive worked out the eigenvectors for all three eigenvalues and only two of sets have opposite signs, if ive followed the same method, would the other not come out with opposite signs too?
Sorry if this is an obvious question, why does the sign of the eigenvector not matter?

Thanks again...Hayley
• Nov 12th 2011, 01:22 PM
Deveno
Re: Help with Eigenvectors
suppose v is an eigenvector for the matrix A.

then Av = λv, for some eigenvalue λ.

but A(cv) = c(Av) = c(λv) = λ(cv), for any scalar c ≠ 0, so cv is likewise an eigenvector for A.

in other words, Au = λu for all u in span({v}).

in particular, all of the above is true when c = -1, as -1 is a non-zero scalar (in any field where char(F) ≠ 2).