lindermanlab / ssm

Bayesian learning and inference for state space models
MIT License
540 stars 196 forks source link

Unspecified numpy version in ssm raising type error #140

Open mshallow opened 2 years ago

mshallow commented 2 years ago

I'm testing out running the GLM-HMM jupyter notebook and running the line below raises a type error. code ran: new_glmhmm.permute(find_permutation(true_latents[0], new_glmhmm.most_likely_states(true_choices[0], input=inpts[0])))

error: TypeError: expected dtype object, got 'numpy.dtype[float64]'

The error points to both util.py and hmm.py, and it's unclear what version of numpy was used in these so I am unable to resolve the issue by changing my version.