I do not really see the motivation for this question but if you really want to do this you should start with the measurement of the complex S-parameters (S11, S12, S12 , S22) over the frequency range of interest .Since there is a nearly infinite number of possibilities you need some assumption and a priori knowledge (e.g. is it coax or waveguide isolator. Which type of isolator (e. Faraday rotation of disk etc. You may consider doing a 3 D Xray tomography to look inside without opening it. The easiest is of course to take it apart (if permitted) and measure the ferrite.There are not so many options for ferrites used in isolators (or just ask the supplier if he want to tell you those details)