Ecosystem

Ecosystem

Everything below is installable today. Each library has its own repository and its own README with full function reference and examples, so the entries here stay deliberately short.

Libraries

Package	Covers	Tag	Author	Repository
gaussian_process	Twenty covariance kernels, marginal and latent GPs, posterior prediction, per-group GPs	0.1.0	@mlatinov	laplace-gaussian_processes-
splines	B-spline, M-spline, I-spline, natural cubic and tensor-product bases, penalties, P-spline priors	0.1.0	@mlatinov	laplace-splines
survival	Gompertz, log-logistic, Gompertz–Makeham and piecewise-exponential survival, with censoring, LOO and simulators	0.1.0	@mlatinov	laplace-survival
ts	Observation-driven time series: ARIMA and seasonal variants, exponential smoothing, GARCH, regime switching, count autoregression	0.1.0	@mlatinov	laplace-ts
lts	Latent time series components: local level and trend, stochastic and trigonometric seasonality, stochastic volatility, latent counts	0.1.0	@mlatinov	laplace-lts
transformations	Centring, scaling, standardizing, clipping, ranking and other data transformations	0.1.0	@mlatinov	laplace-transform
kinetics	Dose-response, pharmacokinetics, and growth/decay curves	0.1.1	@mlatinov	laplace-kinetics

All of these keep the package in a laplace/ subdirectory, so they install with the same shape of command:

laplace add <package> \
  --git https://github.com/mlatinov/<repository> --tag <tag> --subdir laplace

For example:

laplace add gaussian_process \
  --git https://github.com/mlatinov/laplace-gaussian_processes- --tag 0.1.0 --subdir laplace

Then import it in your model and call it as gaussian_process::rbf_cov(...). See Getting Started for the full walkthrough.

ts and lts are companions and are meant to be used together: ts computes its recursion from the data, lts samples a hidden path.

Tooling

Tool	What it does	Repository
laplace	The compiler and package manager itself	laplace
cmdlaplacer	R package: build .laplace files and run them through cmdstanr without leaving R	cmdlaplacer
laplace-lsp	Language server for .laplace files	laplace_tools
vscode-laplace	Syntax highlighting and editor support for VS Code and Positron	laplace_tools

Getting your library listed

This page is open to libraries by anyone, not just to the ones above. Laplace exists so that people can share methods they have already written and be credited for them, and this page is where that credit lives.

If you have written a library that meets the requirements in Writing a Library, open an issue on the laplace repository with:

• the repository link and the tag to install,
• the package name as declared in its manifest,
• one sentence describing what it covers,
• how you would like to be credited. The listing keeps your name and links to your repository. The library stays yours: Laplace resolves it straight from your git repository, and nothing is vendored or copied here.