RampantTrackGeneration 0.0.1.2.2

Rampant on the Tracks's Track generation logic, leveraging Voronout and optimized for a web service.

RampantTrackGeneration is..

.. the Track generation logic for Rampant on the Tracks.

The logic is invoked by calling

@staticmethod
def generateTrack(
    diagramWidth: int,
    diagramHeight: int,
    numDiagramRegions: int,
    diagramEdgePercentageToProcess: float, 
    newConnectionAngleMinQuantile: float, 
    lonelyConnectionMinLengthQuantile: float, 
    connectionLengthVertexPadding: float, 
    connectionLengthNodeBuffer: float,
    destinationDistanceUpperQuantile: float
) -> Track:

in TrackGenerator.

Track

@dataclass(frozen=True)
class Track:
    nodes: dict[uuid4, Point]
    stops: dict[uuid4, tuple[Point]]
    edges: dict[uuid4, EdgeVertexInfo]
    startNode: uuid4
    destinationNode: uuid4

describes a set of edges, each a connection between two Points. nodes are the Points - stops are intervals on each edge, a gameplay mechanic that'll be elaborated later.

generateTrack derives the Track from a randomly generated Voronoi diagram.

It preserves the organic appeal of the diagram's shape - connectivity of varying lengths between unevenly spaced points - and goes on to enhance that by removing a subset of the smallest edges, replacing them with longer " reconnections ". The reconnections' intersections with other edges are calculated. Intersections that could be removed without newly isolating either of the involved Points are removed - this, and further trimming, increase the variance of the resulting Track's shape and its appeal in both visual and gameplay terms.

It uses Voronout to generate the diagram and networkX to model the diagram's transformation into a Track.

startNode and destinationNode reflect the gameplay - where the Walker the player is responsible for starts from, and where it must end up for the player to win.

RampantTrackGeneration works by..

.. doing the following:

• generating #numDiagramRegions Voronoi diagram sites (0 <= x <= diagramWidth, 0 <= y <= diagramHeight)
• generating the Voronoi diagram
• using newConnectionAngleMinQuantile to calculate initialDiagramMinAcceptableAngle, the minimum angle any new reconnection should make with any of the edges at either of its vertices
  • rejecting any reconnection that does not satisfy that constraint minimizes the probability of getting Track edges that make awkwardly small angles with other edges
• reconnecting diagramEdgePercentageToProcess * 100% of the smallest edges in the diagram
• removing all intersection edges created by reconnection that can be safely removed
• removing all " lonely " edges (ones where one vertex is only connected to that edge) with length <= lonelyConnectionMinLengthQuantile * 100% of all edge lengths
• adjusting all edges with length < minEdgeAdjustLength, either..
  • .. combining them with another collinear edge such that the combination would make the longest edge possible
  • .. deleting the edge, reconnecting all edges involving the vertex with the fewest neighbors to the other vertex
• placing Stops on the remaining edges
  • to avoid the awkwardness of placing on " too small edges ", we only place on edges whose length is greater than (connectionLengthVertexPadding + connectionLengthNodeBuffer) * 100% of edges
  • to space Stops organically on an edge, we place them at least connectionLengthVertexPadding * 100% of the edge length away from either of points - and make the distance between each Stop at least connectionLengthNodeBuffer * 100% of the edge length
• calculating Track.startNode and Track.destinationNode
  • Point.distance(<startNode>, <destinationNode>) must be >= destinationDistanceUpperQuantile * 100% of the distances non-startNodes have to startNode

The resulting enhancement can be seen in the below illustration of Voronoi diagram -> Track: