ChainsAddiction

ChainsAddiction is an easy to use tool for time series analysis using discrete-time Hidden Markov Models. It is written in C as a numpy-based Python extension module.

Installation

Install from PyPi

We currently provide wheels for macOS and Windows AMD 64, which you can install from PyPI via:

python3 -m pip install chainsaddiction

Linux users have to build from source until we get that manylinux thing running.

Install from source

Before attemting to build ChainsAddiction from source, make sure you have

• Python >= 3.9
• pip, setuptools
• C compiler

installed and ready to go.

Then, clone the source code by typing the following command in your terminal app. Replace path/to/ca with the path to where ChainsAddiction should be cloned:

git clone https://github.com/teagum/chainsaddiction path/to/ca

Second, change to the root directory of your freshly cloned code repository:

cd path/to/ca

Third, instruct Python to build and install ChainsAddiction:

python3 -m pip install .

---

Notes

Currently only Poisson-distributed HMM are implemented.