orx-shapes/org.openrndr.extra.shapes.eigenmodes/BoundaryEigenMode

BoundaryEigenMode

data class BoundaryEigenMode(val eigenvalue: Double, val frequencyRatio: Double, val contour: RectifiedContour, val ts: List<Double>, val values: DoubleArray)(source)

One eigenmode of a RectifiedContour's outline, as produced by RectifiedContour.extractBoundaryEigenModes.

Unlike EigenMode, this only considers the boundary curve itself, not the shape's interior -- it's the eigenbasis of the contour sampled as a closed graph, a generalization of the sine/cosine basis used by classical Fourier shape descriptors to non-uniformly detailed curves.

Constructors

BoundaryEigenMode

constructor(eigenvalue: Double, frequencyRatio: Double, contour: RectifiedContour, ts: List<Double>, values: DoubleArray)

Properties

contour

val contour: RectifiedContour

The rectified contour that was sampled. Rectification is what makes equal steps in ts equal steps in arc length, which is what the underlying graph Laplacian assumes of its nodes.

eigenvalue

val eigenvalue: Double

The Laplacian eigenvalue of this mode; larger means higher spatial frequency.

frequencyRatio

val frequencyRatio: Double

This mode's frequency relative to the fundamental mode. Mode 0 is always the trivial, patternless zero-eigenvalue constant mode (frequencyRatio == 0.0); the fundamental is mode 1.

val n: Int

Number of points/nodes around the curve.

val ts: List<Double>

The (rectified) contour parameters sampled around the contour, in order; ts[i] is where values[i] was taken, and contour.position(ts[i]) recovers the world-space point.

values

val values: DoubleArray

The mode shape, one value per parameter in ts.

Functions

equals

open operator override fun equals(other: Any?): Boolean

hashCode

open override fun hashCode(): Int

position

fun position(index: Int): Vector2

The world-space position that sample point index (0 until n) was taken at.

value

fun value(t: Double): Double

Samples this mode at point-index coordinate t, which wraps around modulo n (so both -0.5 and n - 0.5 land halfway between the last and first point), using linear interpolation between the nearest sample points.