def marginal_likelihood(self, log=True, normalize=True):
lp = logmvbeta(self.initial_counts) + logmvbeta(self.transition_counts)
if normalize:
lp -= (logmvbeta(self.initial_alpha) + logmvbeta(self.transition_alpha))
return lp if log else np.exp(lp)
def marginal_likelihood(self, log=True, normalize=True):
lf1 = logmvbeta(self.counts)
if normalize:
lf1 -= logmvbeta(self.alpha)
lf2 = sum(component.marginal_likelihood(log=True, normalize=normalize) for component in self.components)
return lf1 + lf2 if log else np.exp(lf1 + lf2)