Struct SmoothingReport 

pub struct SmoothingReport {
    pub axes: Vec<usize>,
    pub segments: usize,
    pub interpolated_segments: usize,
    pub refinements: usize,
    pub fitting_rows: usize,
    pub checked_intervals: usize,
    pub max_errors: Vec<f64>,
}

Diagnostics recorded while constructing a smoothed waypoint path.

Obtain this report with Path::smoothing_report. axes, segments, and interpolated_segments describe the final representation. refinements records completed passes, while fitting_rows and checked_intervals accumulate work from the initial construction and subsequent local refinement. These values help explain tolerance cost and budget usage; they are not optimality or run-time guarantees.

In exact arithmetic, the Bernstein convex-hull property supplies a sufficient whole-interval bound. max_errors records the corresponding values computed in physical input units with ordinary floating-point arithmetic rather than outward-rounded interval arithmetic; they are numerical audit results, not formal certificates or Cartesian/FK errors.

Fields

axes: Vec<usize>

Selected input-row indices, in tolerance/report order.

segments: usize

Number of final polynomial spans shared by the selected axes.

This count is not guaranteed to be minimal or monotone as tolerance changes because initial knot placement and refinement are adaptive.

interpolated_segments: usize

Number of spans in the separate unselected-axis interpolant.

This is zero when every axis is selected and is not included in SmoothingReport::segments or the configured segment budget.

refinements: usize

Number of completed local knot-refinement passes.

fitting_rows: usize

Number of discrete or quadrature fitting rows assembled.

Rows assembled again during local refits are counted again; this is a work statistic, not the number of distinct input points. Quadrature rows used only for the third-derivative penalty are not included.

checked_intervals: usize

Number of reference-polyline intervals visited by numerical audits.

Revisited intervals and the final physical-unit full audit are included.

max_errors: Vec<f64>

Final whole-domain numerical absolute-error bound for each selected axis.

Values use physical input units and follow SmoothingReport::axes order. They are Bernstein-derived numerical bounds, not sampled maxima or outward-rounded certificates.

Trait Implementations

impl Clone for SmoothingReport

fn clone(&self) -> SmoothingReport

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for SmoothingReport

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.