I am trying to solve an eigenvalue problem and I am concerned with finding the smallest positive imaginary eigenvalue.
How to find it using EPSSolve?
See https://slepc.upv.es/documentation/slepc.pdf#page=32: -eps_smallest_imaginary. If you are only finding negative eigenvalues, I guess you’ll have to shift your problem a bit to recover the first positive eigenvalue.
by shift do you mean using “shift” in sparams ?
No, it will be -eps_target. Please read SLEPc documentation.
So, what’s the best algorithm today in Freefem++ for eigenvalue analysis ?
As I read here EPSSolve, and on the Freefem website Arpack is used. Indeed, it would be interesting to have the advices or a summary of the freefem experts.
Thanks in advance.
As I read here EPSSolve, and on the Freefem website Arpack is used
What does that even mean?
The best algorithm will of course depend on what you are trying to solve. In FreeFEM, SLEPc is the most flexible approach (as it can solve linear, nonlinear, polynomial eigenproblems, with a variety of algorithms, Krylov–Schur, LOBPCG, power method…).
Sorry, I mean for linear problems like A X = lambda B X.