pub fn reduce_angle<T: Float>(angle: T) -> T
Stable angle reduction for trigonometric functions
Reduces angle to [-π, π] range while preserving precision for large angles.