precluster: Pre-cluster points with k-means

In kmknn: k-Means for k-Nearest Neighbors

Description

Perform k-means clustering in preparation for a nearest-neighbors search.

Usage

precluster(X, ...)

Arguments

X	A numeric matrix where rows correspond to data points and columns correspond to variables (i.e., dimensions).
...	Further arguments to pass to kmeans.

Details

This function is automatically called by findKNN and related functions. However, it can be called directly by the user to save time if multiple queries are to be performed to the same X.

Value

A list containing:

• data, a numeric matrix with points in the columns and dimensions in the rows, i.e., transposed relative to the input. Points have also been reordered to improve data locality during the nearest-neighbor search. Specifically, points in the same cluster are contiguous and ordered by increasing distance from the cluster center.

• clusters, itself a list containing:

  • centers, a numeric matrix of cluster center coordinates where each column corresponds to a cluster.

  • info, another list of length equal to the number of clusters. Each entry corresponds to a column of centers (let's say cluster j) and is a list of length 2. The first element is the zero-index of the first cell in the output X that is assigned to j. The second element is the distance of each point in the cluster from the cluster center.

• order, an integer vector specifying the new ordering of points in X.

Author(s)

Aaron Lun

See Also

kmeans, findKNN, queryKNN, findNeighbors

Examples

Y <- matrix(rnorm(100000), ncol=20)
out <- precluster(Y)
str(out)

kmknn documentation built on Nov. 1, 2018, 4:21 a.m.