laplace-language

A Laplace library of time series building blocks for Stan — the classical ARIMA family, seasonal variants, exponential smoothing, GARCH volatility, smooth-transition regime switching, and count autoregression. Every model ships a mean recursion, a log density, a simulator, and a forecaster.

A Laplace library of Gaussian process building blocks — twenty covariance kernels, marginal and latent GP models, posterior prediction, and helpers for hierarchical (per-group) GPs. Import it into any .laplace model and call it with namespaced calls (gaussian_process::function_name(...)).

A Laplace library of latent time series components for Stan — local level, local linear trend, stochastic and trigonometric seasonality, stochastic volatility, and latent-Gaussian counts.

A Laplace library of parametric nonlinear functions for pharmacokinetics, dose-response modeling, and growth/decay processes — ready to import into any .laplace model with namespaced calls (kinetics::function_name(...)).

Laplace is a source-to-source preprocessor for Stan: it compiles .laplace files down to plain .stan files, adding a package manager and namespaced imports (pkg::func()) on top of a language that has neither natively.

A Laplace library of spline building blocks for Stan — B-spline, M-spline, I-spline, natural cubic and tensor-product bases, difference and derivative penalties, non-centred P-spline priors, and helpers for hierarchical (per-group) splines. Import it into any .laplace model and call it with namespaced calls (splines::function_name(...)).

A Laplace library of parametric survival models — Gompertz, log-logistic, Gompertz–Makeham, PWE — with right-censored likelihoods, pointwise log-likelihoods for LOO, survival curves, and event-time simulators.

A thin R wrapper around the Laplace compiler and cmdstanr, so you can go from a .laplace source file to a running cmdstanr model without touching the terminal.