log_probability not implemented for zero_inflated poisson HMM

[hfealr1111]

Hi, I am trying to infer and analyze hidden states in neuron spikings with ZIP-HMM uninitialized and fit the model to data.
model = DenseHMM([ZeroInflated(Poisson()), ZeroInflated(Poisson()), ZeroInflated(Poisson())], max_iter=1000, verbose=True)
However, it shows that

/usr/local/lib/python3.10/dist-packages/pomegranate/distributions/_distribution.py in log_probability(self, X)
     62 
     63         def log_probability(self, X):
---> 64                 raise NotImplementedError
     65 
     66         def fit(self, X, sample_weight=None):

NotImplementedError: 

I understand that zero_inflated is a wrapper so it shouldn't have any dedicated log_probability function. So, I wish to confirm with you that ZeroInflated(Poisson()) could be used in hmm this way. If so, I wish you could kindly provide a solution to this. Thanks in advance!

[jmschrei]

Yes, this is an error on my side. I will look into a solution.

[hfealr1111]

Thanks, I really appreciate it!