chainsaddiction 0.1

Discrete time, finit state space, stationary Hidden Markov Model.

# ChainsAddiction

ChainsAddiction is a tool for simple training discrete-time Hidden Markov Models. It is written in C and features a numpy-based Python extension module.

## Installation Clone this repository, change to its root directory and issue

pip install .

## Working with the C API

## Working with the Python interpreter Calling Chains_addiction from Python is simple as pie. You just need to import it:

import chains_addiction as ca ca.hmm_poisson_fit_em(x, m, init_means, init_tpm, int_sd, max_iter=1000, tol=1e-5)

## Notes - Currently only Poisson-distributed HMM are implemented. - ChainsAddiction does not support Python 2. Specifically, it requires Python >= 3.5 and numpy >= 1.16.