pub fn rolling_min_deque(data: &[f64], window: usize) -> Vec<f64>Expand description
Deque-based rolling minimum - O(n) amortized complexity
Uses a monotonic deque to achieve O(1) amortized per-element cost. This is 10-100x faster than the naive O(n*w) approach for large windows.
ยงPerformance
- Expected speedup: 10-100x over O(n*w) scalar (depends on window size)
- Complexity: O(n) amortized