NNDescentFn

Full name Microsoft.VisualBasic.DataMining.UMAP.NNDescentFn Assembly Microsoft.VisualBasic.DataMining.UMAP Members 1

01 Syntax

Microsoft.VisualBasic.DataMining.UMAP.NNDescentFn

02 Methods

Name	Overloads	Summary
NNDescent	1	run the nearest neighbour descent for build the approximated knn graph

03 Members

method NNDescent #

NNDescent(Double()(), Int32()(), Int32, Int32, Int32, Double, Double, Boolean, ParallelConfig)

run the nearest neighbour descent for build the approximated knn graph

Parameters

Name	Type	Description
`data`	`Double()()`	the input data matrix
`leafArray`	`Int32()()`	the leaf nodes of the random projection forest, which is used for seed the initial state of the neighbour graph.
`nNeighbors`	`Int32`	the number of the nearest neighbours
`nIters`	`Int32`	the max number of the descent iterations
`maxCandidates`	`Int32`	the max number of the candidate neighbours of each vertex
`delta`	`Double`	the early stop threshold of the descent iteration
`rho`	`Double`	the sample rate of the descent procedure
`rpTreeInit`	`Boolean`	init the neighbour graph via the random projection forest?
`parallelism`	ParallelConfig	the parallelism configuration