Function block_bootstrap_resample 

pub fn block_bootstrap_resample<R: Rng>(
    data: &[f64],
    block_size: usize,
    rng: &mut R,
) -> Vec<f64>

Perform block bootstrap resampling on timing measurements.

Resamples measurements from a single class (fixed or random) while preserving autocorrelation structure. Blocks of consecutive measurements are sampled with replacement.

Arguments

• data - Slice of timing measurements for one class
• block_size - Size of contiguous blocks to resample
• rng - Random number generator

Returns

A new vector of resampled measurements with the same length as input.

Algorithm

1. Divide data into overlapping blocks of size block_size
2. Sample blocks with replacement until we have n samples
3. Truncate to exactly n samples