Module window 

Rolling window aggregations (sum, mean, min, max). Sliding-window functions for CJC.

Provides window_sum, window_mean, window_min, window_max over arrays of numeric values. These are commonly used in time-series analysis, data science pipelines, and signal processing.

Determinism

• window_sum and window_mean use Kahan summation for numerically stable, deterministic results.
• All window functions produce the same output for the same input on every invocation.

Semantics

All window functions take (data: &[f64], window_size: usize) and return a Vec<f64> of length data.len() - window_size + 1. That is, they produce one output per valid (full) window position.

If window_size is 0 or greater than data.len(), the result is empty.

Functions

window_max

Sliding-window maximum.

window_mean

Sliding-window mean with Kahan summation.

window_min

Sliding-window minimum.

window_sum

Sliding-window sum with Kahan summation.