Trait net_ensembles::traits::Ensemble

pub trait Ensemble<S, Res> {
    fn undo_step(&mut self, step: S) -> Res;
    fn undo_step_quiet(&mut self, step: S);
    fn randomize(&mut self);
    fn mc_step(&mut self) -> S;

    fn mc_steps(&mut self, count: usize) -> Vec<S> { ... }
    fn simple_sample<F>(&mut self, times: usize, f: F)
    where
        F: FnMut(&Self),
    { ... }
    fn simple_sample_vec<F, G>(&mut self, times: usize, f: F) -> Vec<G>
    where
        F: FnMut(&Self) -> G,
    { ... }
    fn undo_steps(&mut self, steps: Vec<S>) -> Vec<Res> { ... }
    fn undo_steps_quiet(&mut self, steps: Vec<S>) { ... }
}

Sampling an Ensemble

includes:

• monte carlo steps
• simple sampling

Required methods

fn undo_step(&mut self, step: S) -> Res

• undo a monte carlo step, return result-state
• if you want to undo more than one step see undo_steps

fn undo_step_quiet(&mut self, step: S)

• undo a monte carlo step, panic on invalid result state
• for undoing multiple steps see undo_steps_quiet

fn randomize(&mut self)

Randomizes self according to model

• this is intended for creation of initial sample
• used in simple_sample and simple_sample_vec

fn mc_step(&mut self) -> S

Monte Carlo step

• use this to perform a Monte Carlo step
• for doing multiple mc steps at once, use mc_steps

Loading content...

Provided methods

fn mc_steps(&mut self, count: usize) -> Vec<S>

Monte Carlo steps

• use this to perform multiple Monte Carlo steps at once
• result Vec<S> can be used to undo the steps with self.undo_steps(result)

fn simple_sample<F>(&mut self, times: usize, f: F) where
    F: FnMut(&Self), 

do the following times times:

1. f(self)
2. self.randomize()

fn simple_sample_vec<F, G>(&mut self, times: usize, f: F) -> Vec<G> where
    F: FnMut(&Self) -> G, 

do the following times times:

1. res = f(self)
2. self.randomize()

res is collected into Vector

fn undo_steps(&mut self, steps: Vec<S>) -> Vec<Res>

Undo Monte Carlo steps

• Note: uses undo_step in correct order and returns result

Important:

• look at specific implementation of undo_step, every thing mentioned there applies to each step

fn undo_steps_quiet(&mut self, steps: Vec<S>)

Undo Monte Carlo steps

• Note: uses undo_step_quiet in correct order

Important:

• look at specific implementation of undo_step_quiet, every thing mentioned there applies to each step

Loading content...

Implementors

impl<T, R> Ensemble<ErStepC, ErStepC> for ErEnsembleC<T, R> where
    T: Node,
    R: Rng, 

fn randomize(&mut self)

Randomizes the edges according to Er probabilities

• this is used by ErEnsembleC::new to create the initial topology
• you can use this for sampling the ensemble
• runs in O(vertices * vertices)

fn mc_step(&mut self) -> ErStepC

Monte Carlo step

• use this to perform a Monte Carlo step
• result ErStepC can be used to undo the step with self.undo_step(result)

fn undo_step(&mut self, step: ErStepC) -> ErStepC

Undo a Monte Carlo step

• adds removed edge, or removes added edge, or does nothing
• if it returns an Err value, you probably used the function wrong

Important:

Restored graph is the same as before the random step except the order of nodes in the adjacency list might be shuffled!

fn undo_step_quiet(&mut self, step: ErStepC)

Undo a Monte Carlo step

• adds removed edge, or removes added edge, or does nothing
• if it returns an Err value, you probably used the function wrong

Important:

Restored graph is the same as before the random step except the order of nodes in the adjacency list might be shuffled!

impl<T, R> Ensemble<ErStepM, ErStepM> for ErEnsembleM<T, R> where
    T: Node,
    R: Rng, 

fn undo_step(&mut self, step: ErStepM) -> ErStepM

• undo a monte carlo step, return result-state
• if you want to undo more than one step see undo_steps

fn undo_step_quiet(&mut self, step: ErStepM)

• undo a monte carlo step, panic on invalid result state
• for undoing multiple steps see undo_steps_quiet

fn randomize(&mut self)

Randomizes self according to model

• this is intended for creation of initial sample
• used in simple_sample and simple_sample_vec

fn mc_step(&mut self) -> ErStepM

Monte Carlo step

• use this to perform a Monte Carlo step
• for doing multiple mc steps at once, use mc_steps

Loading content...