Regarding Transition Probabilities about iohmm

[ghcsung]

You probably have a covariate (input) associated with the transition probabilities. For the first question I did:

# the number of zeros inside np.array() should be equal to the number of your covariates, I have two inputs.

empty = np.array([[1, 0, 0]]) # 1 for constant

SHMM0.model_transition[0].fit_intercept = False
SHMM0.model_transition[1].fit_intercept = False

print(np.exp(SHMM0.model_transition[0].predict_log_proba(empty)))
print(np.exp(SHMM0.model_transition[1].predict_log_proba(empty)))

This gets the transition probabilities for when your covariates are all 0 (except the constant).