Given #37, you should fill in the continuous_time_methods\matlab\tests\time_varying_optimal_stopping_diffusion_test.m tests with important tests to make sure you think the method is correct, and delivers sensible results. I threw in a few starting points, but they are not really verified.
[ ] Check that if we have constant $\mu,\sigma^2,S,u$ but multiple time periods, that the value functions and stopping point are the same.
[ ] Have the S function increasing over time, and make sure that both (a) the $x$ at which stopping occurs is increasing, but also that the point at which stopping occurs is strictly greater than it is for a constant S at that point (i.e., agents would wait around for the S to grow and stop less)
[ ] Do the same with a decreasing S
[ ] Do the same with a $u(t,x)$ that is time varying. Ensure it moves large enough that we get more than trivial changes in the soluion.
[ ] Try to have a time-varying $\mu(t,x)$ and make sure that the solution is sensible from an agents perspective.
Given #37, you should fill in the
continuous_time_methods\matlab\tests\time_varying_optimal_stopping_diffusion_test.mtests with important tests to make sure you think the method is correct, and delivers sensible results. I threw in a few starting points, but they are not really verified.