module KDTree.ts #

type KDTreeNode #

export type KDTreeNode = {
  depth: number; 
  dataPoint: Point; 
  screenPoint: NumericPoint; 
  originalIndex: number; 
};

An entry in the KDTree, should probably be made more generic to support indexing more information, as these nodes are returned on queries as well

• depth number
• dataPoint Point
• screenPoint NumericPoint
• originalIndex number

type KDTree #

export type KDTree = {
  node: KDTreeNode;
  left: KDTree | null;
  right: KDTree | null;
};

Binary space parition of KDTreeNodes which are used to index dataSet points for mouse over detection

• node KDTreeNode
• left KDTree
• right KDTree

type KDTreeSearchResult #

export type KDTreeSearchResult = {
  node: KDTreeNode;
  screenDistanceSq: number;
};

the result type for search, returns the closest Node that matches as well as the squared screen distance to it

• node KDTreeNode
• screenDistanceSq number

function buildIndex #

(nodes: KDTreeNode[], depth?: number, xSorter?: NodeSorter,
ySorter?: NodeSorter) => KDTree

Parameters #

• nodes KDTreeNode
• depth number
• xSorter NodeSorter
• ySorter NodeSorter

Returns KDTree #

function dumpIndex #

(tree: KDTree, depth?: number) => void

Parameters #

• tree KDTree
• depth number

Returns void #

function nearestNeighbor #

(queryPoint: NumericPoint, index: KDTree, maxDistance?: number,
depth?: number, bestMatch?: KDTreeSearchResult, ignoreX?:
boolean, ignoreY?: boolean) => KDTreeSearchResult

Parameters #

• queryPoint NumericPoint
• index KDTree
• maxDistance number
• depth number
• bestMatch KDTreeSearchResult
• ignoreX boolean
• ignoreY boolean

Returns KDTreeSearchResult #