This PR attempts to test how well we are able to estimate initial state probabilities given a transition probability matrix.
The changes at this point are my first pass attempts to:
Test the compute_steady_state method
I've added a few simple test cases to confirm that the steady state estimation is able to recover the steady state for a simple "4x4 identify matrix" example and simple 2x2 matrix with all probabilities equal to .5
Test the gamma_0_estimation method
With a lot of help from @xjing76 I was able to add a test comparing the initial state probabilities to the estimated state probabilities. Currently there's only one test case (two states, 50-50 probability).
This PR attempts to test how well we are able to estimate initial state probabilities given a transition probability matrix.
The changes at this point are my first pass attempts to:
compute_steady_statemethodI've added a few simple test cases to confirm that the steady state estimation is able to recover the steady state for a simple "4x4 identify matrix" example and simple 2x2 matrix with all probabilities equal to .5
gamma_0_estimationmethodWith a lot of help from @xjing76 I was able to add a test comparing the initial state probabilities to the estimated state probabilities. Currently there's only one test case (two states, 50-50 probability).