Struct WeightedSampler 

pub struct WeightedSampler { /* private fields */ }

Pre-computed weighted sampler for O(log n) repeated sampling.

Builds a cumulative distribution table from weights, then uses binary search for each sample.

Algorithm

CDF-based weighted sampling with binary search.

Complexity

• Construction: O(n)
• Sampling: O(log n)

Examples

use u_numflow::random::{create_rng, WeightedSampler};
let weights = vec![1.0, 2.0, 3.0, 4.0];
let sampler = WeightedSampler::new(&weights).unwrap();
let mut rng = create_rng(42);
let idx = sampler.sample(&mut rng);
assert!(idx < 4);

Implementations

impl WeightedSampler

pub fn new(weights: &[f64]) -> Option<Self>

Creates a new weighted sampler from the given weights.

Returns

• None if weights is empty or all weights are zero/negative.

pub fn sample<R: Rng>(&self, rng: &mut R) -> usize

Samples a random index according to the weights.

Complexity

O(log n) via binary search.

pub fn len(&self) -> usize

Returns the number of categories.

pub fn is_empty(&self) -> bool

Returns true if there are no categories.

pub fn total_weight(&self) -> f64

Returns the total weight.