My eig function worked once and then broke

Hello everyone,

I am currently trying to solve a vibrant system, and I need its eigenvalues and eigenvectors. I so used the eig() function, as following:

[EigVec, EigVal] = eig(K\M, 'vector') % K the stiffness matrix, M the mass matrix

I tried it once and got the exact right results. I modified some values in the mass matrix and now this line isn't working anymore. I get the following error:

Error using sym/eig

Too many input arguments.

Error in Homework2 (line 347)

[EigVect, EigVal] = eig(K\M, 'vector');

I also tried:

[EigVec, EigVal] = eig(K\M)

And now I get absurd values.

Any ideas of what happened?

[Christine Tobler]

It sounds like originally your matrices didn't contain symbolic variables, so the eig command accepted the 'vector' option (eig for symbolic variables doesn't accept any option strings). Then you changed M so that it's now a symbolic variable - and those have quite different behavior, since it's not numerical calculation anymore but symbolic.

Can't really tell more since I don't know what you mean by absurd values, or what you expect in terms of symbolic calculation (for example, is M a specific number just in stored symbolically, or does it contain symbolic unknowns)?

We really need reproduction steps of what K and M are, and what you're hoping to compute, to help any more.

Answer 1

[Bruno Luong]

Don't use symbolic variable to build your matrices.

[Benjamin KOENIG]

I banned all symbolic variables from my code, but it doesn't look like it changed anything. I still have the error and absurd values.

[Bruno Luong]

"Error using sym/eig"

implies you have symbolic variable.

If

eig(K\M)

returns "absurd" result perhaps your matrices input is not correct ("I modified some values in the mass matrix" but what?) . But you don't tell us what ther are and why it is absurd. The only absurd thing to me is that you expect useful help from so little information.