In fcorra/mod_clustsig: Modification of Significant Cluster Analysis

MODclustsig

This package is a redistribution of the original clustsig package: Significant Cluster Analysis written by Douglas Whitaker and Mary Christman.

I made the following changes:

1. I included the Hellinger Distance ("hellinger") as a method to calculate the distance matrix. The package original version allow you to pass any function to calculate the distance matrix to be used by hclust (ei using the method.distance argument). However, since we will need to work with the Hellinger Distance quite often I thought that it was a good idea to have the method included by default.
2. The Hellinger Distance matrix is calculated in C++ to speed up the process a little bit. This means that the package now imports form Rcpp evalCpp.
3. I made a class simprof to handle the result of the simprof function and implemented the simprof.plot as a method for this class (instead of just a function as before).

Everything else remains the same. If you intend to use this redistribution for anything else but passing the Hellinger Distance to hclust, I think it would be better to use Whitaker & Christman's original version. Anyways, you are free to use it if you want. I'll be available to answer any questions that you may have.

Thank you Douglas Whitaker and Mary Christman, you save my day!

Installation

Install the current development from github via:

remotes::install_github("fcorra/MODclustsig")

fcorra/mod_clustsig documentation built on Jan. 24, 2020, 1:26 a.m.