pybocd 1.1.0

No project description provided

pybocd

---

pybocd implements Bayesian Online Changepoint Detection (BOCD) with two built-in observation models:

• BOCD-NIG: conjugate Normal–Inverse–Gamma model for unknown mean and variance.
• BOCD-GMM: particle-based Gaussian mixture model for multimodal data and outliers.

Installation

pip install pybocd

If you install from source for development, ensure numpy, scipy, and pandas are available.

Quickstart — BOCD-NIG

from pybocd import BOCDNIG

# prior: m0, kappa0, alpha0, beta0; l: expected run length; threshold: pruning
model = BOCDNIG(m_0=0.0, kappa_0=1.0, alpha_0=1.0, beta_0=1.0, l=200.0, threshold=1e-4)

for x in stream_of_floats:
	model.add_data(x)
	# model.run_length_dist and model.run_length show current posterior

Quickstart — BOCD-GMM

from pybocd import BOCDGMM

# The GMM-based BOCD requires several hyperparameters — see source for details.
model = BOCDGMM(alpha_0=2.0, beta_0=2.0, m_0=0.0, kappa_0=1.0,
				alpha_p_0=2.0, beta_p_0=2.0, mu_p_0=0.0, sigma_p_sq_0=1.0,
				jitter_mu=0.01, jitter_sigma_sq=0.01, jitter_tau_sq=0.01, jitter_pi=0.01,
				l=200.0, m=20, n=200, init_particle_n=50)

for x in stream_of_floats:
	model.add_data(x)

Documentation & Links

• Documentation: https://zeyu-xie.github.io/pybocd/
• PyPI: https://pypi.org/project/pybocd/
• Examples repository: https://github.com/Zeyu-Xie/pybocd-Examples

Data Attributes

During development, data sourced from Westermo is used, which is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) License. We gratefully acknowledge Westermo for making this real-world telemetry data available for research and development purposes.

License

pybocd is distributed under the terms of the GPL-3.0-or-later license.