/compaRe

R package with two functions for similarity measurement and clustering. It uses dynamic programming to measure spatial similarity between two datasets. It can be used to generate a similarity matrix as input to the clustering function. The clustering function reports maximal cliques as clusters and dense areas as communities.

Primary LanguageRMIT LicenseMIT

COMPARE

COMPARE is an R package with two functions for similarity measurement and clustering. It employs a dynamic programming algorithm to measure spatial similarity between two given datasets. One can then use this function to generate a similarity/affinity matrix which can then be input to the clustering function. The clustering function reports maximal cliques as clusters and dense areas as communities. COMPARE

  1. is data-size independent
  2. can effectively circumvent moderate background noise (like batch effect)
  3. does not need dimension reduction
  4. does not need subsampling

How it works

For more information about its algorithms check out COMPARE-suite's similarity and clustering modules. COMPARE-suite is a pipeline for high-throughput and high-content multiparameter screening built on COMPARE.

Installation

Refer to Q&A.

Usage

Read the accompanying help page of the package for more details about usage.

FAQ

Refer to Q&A.

Report issues

Report any bug/issue on the Issue page.

How to cite compaRe

Refer to Q&A.