fmcs/lppl

Memory-efficient open-universe probabilistic programming in modern C++. Deep dive here. Program synthesis here.

• Record-based (i.e., making trace-based approach explicit instead of implicit)
• Sample-based, with posterior queries by default computed online rather than after sampling is complete – defaulting to constant memory complexity
• Explicit PRNG state (reproducibility)
• Algebraic sum types everywhere instead of boxing
• Header-only

This should look familiar.

template<size_t N>
std::tuple<double, double> 
linear_regression(record_t<DTypes<Normal, Gamma>>& r, std::shared_ptr<data_1d<N>> data) {
    auto intercept = sample(r, "intercept", Normal(), rng);
    auto slope = sample(r, "slope", Normal(), rng);
    auto scale = sample(r, "scale", Gamma(), rng);
    for (size_t ix = 0; ix != N; ix++) {
        observe(
            r,
            "obs/" + std::to_string(ix),
            Normal(data->x[ix] * slope + intercept, scale),
            data->y[ix]
        );
    }
    return std::make_tuple(intercept, slope);
}

Details

• Documentation
• Tests
• Source (will attempt a download of a zip file to your machine) or on Gitlab

The lppl webpage and its subpages are licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.