wm-simulation 9.2.1

//! wm-simulation — Monte Carlo simulation, counterfactual estimation, forecasting,
//! and prediction calibration.
//!
//! **N21**: Ports v2's `wm-evolution` MC suite capabilities for v4:
//!
//! - **Monte Carlo simulation** — Bayesian MC, Quasi-MC, sensitivity analysis
//! - **Counterfactual estimation** — synthetic control projection for causal impact
//! - **Forecasting** — time series forecasting with confidence intervals
//! - **Prediction calibration** — Brier scorecard with the Murphy decomposition
//! - **Bayesian optimization** — GP surrogates + Expected Improvement search
//! - **Information-theoretic measures** — entropy, mutual information
//!
//! This enables the SelfModel to forecast outcomes, the Dream cycle to
//! simulate counterfactuals, and the Homeostatic loop to simulate action
//! consequences before executing them.

#![forbid(unsafe_code)]

pub mod bayesian;
pub mod calibration;
pub mod claims;
pub mod counterfactual;
pub mod forecasting;
pub mod monte_carlo;
pub mod pce;
pub mod rare_event;
pub mod sde;
pub mod sensitivity;

pub use bayesian::{
    BayesianOptimizer, Expr, GaussianProcess, OptimizationStep, expected_improvement, norm_cdf,
    norm_pdf,
};
pub use calibration::{BrierScorecard, CalibrationBin, CalibrationPrediction, CalibrationStore};
pub use claims::{CALIBRATION_PRIOR_SAMPLES, Claim, ClaimStatus, ClaimsLedger, ValidationEvent};
pub use counterfactual::{CounterfactualEstimator, CounterfactualResult};
pub use forecasting::{ForecastMethod, ForecastResult, Forecaster};
pub use monte_carlo::{Distribution, McConfig, McResult, MonteCarloSimulator};
pub use pce::{Pce, SuperforecasterResult, latin_hypercube, superforecaster};
pub use rare_event::{ImportanceResult, SubsetResult, importance_sampling, subset_simulation};
pub use sde::{DriftType, MlMcResult, SdeConfig, SdeResult, Solver, solve, solve_mlmc};
pub use sensitivity::{SensitivityAnalyzer, SensitivityIndex, SensitivityResult};