Skip to main content

Crate rustmc_core

Crate rustmc_core 

Source

Re-exports§

pub use data::BindError;
pub use data::DataBinding;
pub use data::DataInputs;
pub use data::DataSchema;
pub use data::DataSlot;
pub use data::MatrixBinding;
pub use data::SlotKind;

Modules§

autodiff
bayesian_ar
Conjugate Bayesian autoregression and posterior-predictive forecasting.
bayesian_forecast
Seeded Bayesian inference and posterior prediction for a local-level model.
bayesian_regression
Joint Gaussian regression and structural-state Gibbs/FFBS inference.
bayesian_seasonal
Conjugate Bayesian structural seasonal local-level inference.
bayesian_trend
Bayesian inference and posterior prediction for a local-linear-trend model.
data
Dataset schemas and validated, re-bindable payloads.
decision
Decisions from predictive draws.
diagnostics
MCMC diagnostic computations: R-hat, ESS, MCSE, quantiles.
distributions
dynamic_glm
Joint latent-Gaussian dynamic regressions, with block elliptical slice inference. Gaussian prior scales and NB dispersion are fixed model specifications. Group coefficients equal an uncertain population coefficient plus a Gaussian deviation; group random walks can additionally share a common random walk.
elliptical_slice
Elliptical slice sampling for a standard-normal latent block.
forecast_batch
Independent batch execution with stable cell identity and bounded workers.
forecast_common
Pieces shared by the conjugate Gaussian forecasting samplers.
forecast_evaluation
Calibration and point accuracy of forecasts scored against realised outcomes.
graph
hierarchical
Joint partial pooling for ragged collections of Gaussian series.
hmc
hurdle
A Bernoulli hurdle with dynamic lognormal positive amounts.
mass_matrix
model
Declarative graph models shared by Python and standalone Rust applications.
nuts
No-U-Turn Sampler (NUTS) — Hoffman & Gelman (2014) with multinomial sampling (Betancourt 2017).
observation
Shared observation simulation and conditional means for graph models.
param_ref
Build-time validation of parameter references.
prior_sampling
Analytic prior draws in unconstrained coordinates, without arbitrary floors.
reconciliation
Forecast reconciliation: making the forecasts of a hierarchy add up.
runoff
Integer-event payment development with prefix censoring and an explicit tail.
sampler
seeding
Deterministic stream separation for seeded RNGs.
state_space
Linear Gaussian state-space filtering, smoothing, and forecasting.
structural
Composable structural models with exact Gaussian FFBS and conjugate variance updates. Student-t observations use Gamma precision mixtures with fixed df. Initial state priors are independent of innovation variances; x[-1] is included in every state draw so every transition contributes to the variance update.
target
Native custom log densities sampled by the same HMC/NUTS kernels as graphs.
time_design
Regression columns on an integer calendar: known-date interventions, piecewise-linear trend changepoints, Fourier terms, and named columns the caller supplies.