pyLGM

General-purpose latent Gaussian models for Python — the model class behind INLA. Fit fixed effects, random effects, temporal (RW/AR1) and spatial (CAR) structure under Gaussian, Poisson, or Bernoulli likelihoods, from a Pandas (or Spark) DataFrame, with exact-Gaussian and Laplace inference engines, empirical Bayes / MAP-II hyperparameter estimation, and INLA-style posterior integration.

Install

pip install pylgm
pip install "pylgm[spark]"   # optional: Spark data boundary

Requires Python ≥ 3.11.

30-second example

A Poisson model with a per-region random intercept, fit with the Laplace engine:

import pandas as pd
from pylgm import Fixed, IID, LGM, Poisson

# Counts of events per region over time, with a covariate x.
frame = pd.DataFrame({
    "region": ["north", "north", "north", "south", "south", "south"],
    "time":   [1, 2, 3, 1, 2, 3],
    "x":      [0.0, 0.5, 1.0, 0.0, 0.5, 1.0],
    "count":  [3, 5, 8, 2, 3, 5],
})

model = LGM(
    response="count",
    likelihood=Poisson(),                                 # canonical log link
    predictor=Fixed("1 + x")                              # fixed effects
        + IID("region", index="region", precision=2.0),  # random intercept per region
    panel=("region",),
    time="time",
)

result = model.fit(frame, engine="laplace")
print("fitted_mean:", result.fitted_mean.round(3).tolist())
# fitted_mean: [3.254, 4.986, 8.15, 2.201, 3.373, 5.513]

The same model can be declared in YAML and loaded with pylgm.config.load_model. See the general LGM example, which fits the same data through both the Python and YAML frontends, and the other 13 runnable scripts under examples/.

What's in the box

Documentation

Full docs: https://ardea00.github.io/pylgm/ — or browse the docs/ folder. Start with the index.

Scope and roadmap

pyLGM 0.3 is a bounded foundation release. What is and isn't in it, and where it's going, are documented in the roadmap and the internals / release policy page.

Development

See docs/development.md. License: MIT.