I need to build a hidden Markov model in R, which is divided into 1 and 2, two states. When the hidden state is 1, the observed variables obey the exponential distribution, and when the hidden state is 2, the observed variables obey the Gaussian distribution, which means that observed variables obey continuous distribution. enter image description here
I need to get the parameters of the exponential distribution and the Gaussian distribution, and the most likely sequence of hidden states.
My first question is: Which package can meet my need?
My idea is to use depmisS4 packgage, with below code:
mod <- depmix(response = cases ~ 1, data = influenza, nstates = 2, family = list(gaussian(), exp())
trstart = runif(4))
Is my code correct? How do I get a sequence of possible hidden states through the code?