Struct SMCConfig 

pub struct SMCConfig {
    pub resampling_method: ResamplingMethod,
    pub ess_threshold: f64,
    pub rejuvenation_steps: usize,
}

Configuration options for Sequential Monte Carlo.

This struct controls various aspects of the SMC algorithm, allowing fine-tuning of performance and accuracy trade-offs.

Fields

• resampling_method - Algorithm used for particle resampling
• ess_threshold - ESS threshold that triggers resampling (as fraction of N)
• rejuvenation_steps - Number of MCMC moves after resampling to increase diversity

Examples

use fugue::*;

// Conservative configuration (less resampling, more rejuvenation)
let conservative_config = SMCConfig {
    resampling_method: ResamplingMethod::Systematic,
    ess_threshold: 0.2,  // Resample when ESS < 20% of particles
    rejuvenation_steps: 5, // 5 MCMC moves after resampling
};

// Aggressive configuration (frequent resampling, no rejuvenation)
let aggressive_config = SMCConfig {
    resampling_method: ResamplingMethod::Systematic,
    ess_threshold: 0.8,  // Resample when ESS < 80% of particles
    rejuvenation_steps: 0, // No rejuvenation
};

Fields

resampling_method: ResamplingMethod

Algorithm used for particle resampling.

ess_threshold: f64

ESS threshold that triggers resampling (as fraction of particle count).

rejuvenation_steps: usize

Number of MCMC moves after resampling to increase diversity.

Trait Implementations

impl Default for SMCConfig

fn default() -> Self

Returns the “default value” for a type.