Learn R Programming

⚠️There's a newer version (0.0-4) of this package.Take me there.

clusternor (version 0.0-2)

A Parallel Clustering Non-Uniform Memory Access ('NUMA') Optimized Package

Description

The clustering 'NUMA' Optimized Routines package or 'clusternor' is a highly optimized package for performing clustering in parallel with accelerations specifically targeting multi-core Non-Uniform Memory Access ('NUMA') hardware architectures. Disa Mhembere, Da Zheng, Carey E. Priebe, Joshua T. Vogelstein, Randal Burns (2019) .

Copy Link

Version

Install

install.packages('clusternor')

Monthly Downloads

21

Version

0.0-2

License

Apache License 2.0

Issues

Pull Requests

Stars

Forks

Repository

https://github.com/neurodata/knorR

Maintainer

Disa Mhembere

Last Published

March 25th, 2019

Functions in clusternor (0.0-2)

Perform the k-means++ clustering algorithm on a data matrix.

Perform parallel hierarchical clustering on a data matrix.

Perform Fuzzy C-means clustering on a data matrix. A soft variant of the kmeans algorithm where each data point are assigned a contribution weight to each cluster

Perform spherical k-means clustering on a data matrix. Similar to the k-means algorithm differing only in that data features are min-max normalized the dissimilarity metric is Cosine distance.

Perform a parallel hierarchical clustering using the x-means algorithm

Perform k-medoids clustering on a data matrix. After initialization the k-medoids algorithm partitions data by testing which data member of a cluster Ci may make a better candidate as medoid (centroid) by reducing the sum of distance (usually taxi), then running a reclustering step with updated medoids.

Perform k-means clustering on a data matrix.

MiniBatchKmeans

A randomized dataset sub-sample algorithm that approximates the k-means algorithm. See: https://www.eecs.tufts.edu/~dsculley/papers/fastkmeans.pdf

A small example of centroids of dim: (8,5) used as for micro-benchmarks of the clusternor package. The data are randomly generated.

A small dataset of dim: (50,5) used as for micro-benchmarks of the clusternor package. The data are randomly generated hence a clear number of clusters will be hard to find.