@saehrimnir/druidjs / NNDescent

Class: NNDescent<T>

Defined in: knn/NNDescent.js:38

NN-Descent

An efficient graph-based approximate nearest neighbor search algorithm. It works by iteratively improving a neighbor graph using the fact that "neighbors of neighbors are likely to be neighbors".

Template

See

Paper

Extends

• KNN

Type Parameters

Type Parameter	Description
T extends number[] | Float64Array

Constructors

Constructor

new NNDescent<T>(elements: T[], parameters?: Partial<ParametersNNDescent>): NNDescent<T>;

Defined in: knn/NNDescent.js:55

Parameters

Parameter	Type	Description
elements	T[]	Called V in paper.
parameters	Partial<ParametersNNDescent>	-

Returns

NNDescent<T>

See

http://www.cs.princeton.edu/cass/papers/www11.pdf

Overrides

KNN.constructor

Properties

Property	Type	Inherited from	Defined in
_elements	T[]	KNN._elements	knn/KNN.js:14
_N	number	-	knn/NNDescent.js:62
_nndescent_elements	{ flag: boolean; index: number; value: T; }[]	-	knn/NNDescent.js:66
_parameters	ParametersNNDescent	KNN._parameters	knn/KNN.js:16
_randomizer	Randomizer	-	knn/NNDescent.js:63
_sample_size	number	-	knn/NNDescent.js:64
_type	"array" | "typed"	KNN._type	knn/KNN.js:18

Methods

add()

add(elements: T[]): NNDescent<T>;

Defined in: knn/NNDescent.js:152

Parameters

Parameter	Type	Description
elements	T[]	-

Returns

NNDescent<T>

---

search()

search(x: T, k?: number): {
  distance: number;
  element: T;
  index: number;
}[];

Defined in: knn/NNDescent.js:236

Parameters

Parameter	Type	Default value	Description
x	T	undefined	-
k?	number	5	Default is 5

Returns

{ distance: number; element: T; index: number; }[]

Overrides

KNN.search

---

search_by_index()

search_by_index(i: number, k?: number): {
  distance: number;
  element: T;
  index: number;
}[];

Defined in: knn/NNDescent.js:325

Parameters

Parameter	Type	Default value	Description
i	number	undefined	-
k?	number	5	Default is 5

Returns

{ distance: number; element: T; index: number; }[]

Overrides

KNN.search_by_index