-
Notifications
You must be signed in to change notification settings - Fork 10
Eigenbounds
william-dawson edited this page Feb 21, 2018
·
2 revisions
In the Eigenbounds module, there are methods for quickly estimating the largest and smallest eigenvalue of a matrix. These methods are particularly important when one tries to scale the spectrum of a matrix to fit inside a certain interval.
Two methods are implemented in this module. First, is a method based on Gershgorin's circle theorem: https://en.wikipedia.org/wiki/Gershgorin_circle_theorem
The second is based on the power method: https://en.wikipedia.org/wiki/Power_iteration
The power method gives a tighter bounds, but is currently only implemented for computing the largest magnitude eigenvalue of a matrix.
No citation is needed for these methods.