R/get.vc.correlation.groups.R

In software-analytics/Rnalytica: An R package of JIRA defect datasets and tool suites for explainable software analytics

#' Get correlation groups according to VarClus based on the absolute Spearman correlation coefficients between metrics
#'
#' This function makes life simple by providing a VarClus.
#' @param dataset  a data frame for data
#' @param metrics  a vector of characters or a vector of characters for independent variables
#' @param similarity a character for similarity measures (e.g., Spearman rank correlation), default = spearman
#' @param varclus.threshold a numeric for correlation coefficient threshold value
#' @importFrom Hmisc varclus
#' @keywords VarClus

get.vc.correlation.groups <- function(dataset, metrics, similarity = 'spearman', varclus.threshold = 0.7){
  
  # Remove constant metrics and categorical metrics
  metrics <- remove.constant.categorical(dataset, metrics)
  
  f <- as.formula(paste("~", paste(metrics, collapse = " + ")))
  vc <-
    Hmisc::varclus(f,
                   similarity = similarity,
                   data = dataset[, metrics],
                   trans = "abs")
  
  var.clusters <-
    cutree(vc$hclust, h = (1 - varclus.threshold))
  melted.data <- melt(var.clusters)
  varclus.correlation.groups <- data.frame(metrics = row.names(melted.data), rank = var.clusters)
  row.names(varclus.correlation.groups) <- NULL
  
  return(varclus.correlation.groups)
}