awesomePMC

awesomePMC (distribution awesomepmc on PyPI, import package pmcprg) implements pairwise Markov chains (PMC) — and the hidden Markov chains they contain — with the dependence between consecutive observations modelled by bivariate copulas: simulation, supervised classification, unsupervised estimation (ICE, SEM, GICE) with automatic copula and margin selection, missing-data imputation and forecasting, and a PyQt6 GUI.

This site is the API reference: generated from the package's docstrings and published for every release, so it documents the version on PyPI (see Installation to build it locally). It does not repeat the project's narrative documentation — for the feature tour, the model TOML format, the CLI, the GUI, worked examples and the scientific references, see the README and, for how to contribute, CONTRIBUTING.md.

Reference implementation of

S. Derrode, W. Pieczynski. Unsupervised data classification using pairwise Markov chains with automatic copulas selection. Computational Statistics & Data Analysis 63 (2013), 81-98. doi:10.1016/j.csda.2013.01.027

S. Derrode, W. Pieczynski. Unsupervised classification using hidden Markov chain with unknown noise copulas and margins. Signal Processing 128 (2016), 8-17. doi:10.1016/j.sigpro.2016.03.008

Layout of this site

• Installation — installing the package and its optional extras, and building this site locally.
• API reference — one page per sub-package: pmcprg.copulas (the copula registry and the shared CopulaVirt/BivariateLaw API, fitting and diagnostics helpers), pmcprg.pmc (the PMC/HMC model, inference, missing-data handling and the ICE/SEM estimators), pmcprg.missing (missingness patterns and evaluation metrics) and pmcprg.diagnostics (goodness-of-fit tests).
• State labelling — how the package numbers and identifies the hidden states of a fitted model, and what a user should expect across estimation runs.
• Changelog — the project's CHANGELOG.md.