rm center; add scale

NAMESPACE

@@ -1,7 +1,7 @@
 # Generated by roxygen2: do not edit by hand
 
-export(center)
 export(preprocess)
+export(scale)
 import(dplyr)
 import(ggplot2)
 import(tidyr)

R/scale.R

@@ -0,0 +1,102 @@
+#'
+#' Centering and/or scaling
+#'
+#' @description
+#' Apply centering and/or scaling to the data.
+#'
+#' @param dataSet A data frame containing the data signals.
+#'
+#' @param names A vector of strings (default = c("gender", "treatment", "replicate"))
+#' specifying the names of the attribute columns.
+#'
+#' @param center A boolean (default = TRUE) specifying whether to apply centering.
+#'
+#' @param scale A string (default = "auto") specifying which method of scaling to apply:
+#' \enumerate{
+#' \item "none" No scaling will be applied.
+#' \item "auto": Auto scaling.
+#' \item "level": Level scaling.
+#' \item "pareto": Pareto scaling.
+#' \item "range": Range scaling.
+#' \item "vast": Vast scaling.
+#' }
+#' @param mu A numeric vector specifying the centering reference. The default centering
+#' reference is set to the mean value.
+#'
+#' @param sigma A numeric vector specifying the scaling reference based on the corresponding
+#' definition. The default scaling reference is set to the standard deviation when
+#' \code{center = TRUE} and the adjusted root mean square when \code{center = FALSE}.
+#'
+#' @details
+#' The function executes the following:
+#' \enumerate{
+#' \item Plots the mean-variance relationship.
+#' \item Centers and/or scales the data.
+#' \item Plots the mean-variance relationship again for comparison.
+#' }
+#'
+#' @returns The centered and/or scaled data.
+#'
+#' @autoglobal
+#'
+#' @export
+
+scale <- function(dataSet,
+                  names = c("gender", "treatment", "replicate"),
+                  center = TRUE, scale = "auto",
+                  mu = NULL, sigma = NULL) {
+
+  ## organize the data for centering
+  dataPoints <- dataSet %>%
+    select(-any_of(names))
+
+  ## calculate and plot a mean-variance plot
+  plotPre <- meanVarPlot(dataPoints, title = "Pre-Pretreatment")
+  print(plotPre)
+
+  if (is.null(mu)) {
+    mu <- colMeans(dataPoints, na.rm = TRUE)
+  }
+
+  ## centering
+  if (center) {
+    dataPoints <- sweep(dataPoints, 2L, mu, check.margin = FALSE)
+  }
+
+  if (is.null(sigma)) {
+    sigma <- apply(dataPoints, 2, function(col) {
+      col <- col[!is.na(col)]
+      sqrt(sum(col^2) / max(1L, length(col)-1L))
+    })
+  }
+
+  ## scaling
+  if (scale == "none") {
+    result <- dataPoints
+  } else if (scale == "auto") {
+    result <- sweep(dataPoints, 2L, sigma, `/`, check.margin = FALSE)
+  } else if (scale == "level") {
+    result <- sweep(dataPoints, 2L, mu, `/`, check.margin = FALSE)
+  } else if (scale == "pareto") {
+    result <- sweep(dataPoints, 2L, sqrt(sigma), `/`, check.margin = FALSE)
+  } else if (scale == "range") {
+    range_diff <- apply(dataPoints, 2, function(col) {
+      res <- max(col, na.rm = TRUE) - min(col, na.rm = TRUE)
+      res <- ifelse(res == 0, 1, res)
+      return(res)
+    })
+    result <- sweep(dataPoints, 2L, range_diff, `/`, check.margin = FALSE)
+  } else if (scale == "vast") {
+    result <- sweep(dataPoints, 2L, mu/sigma^2, `*`, check.margin = FALSE)
+  }
+
+  ## calculate and plot a mean-variance plot
+  plotPost <- meanVarPlot(result,  title = "Post-Pretreatment")
+  print(plotPost)
+
+  ## recombine the labels and centered and/or scaled data into a single data frame
+  result <- cbind(dataSet[,names], result)
+
+  ## return the centered and/or scaled data
+  return(result)
+}

docs/pkgdown.yml

@@ -2,7 +2,7 @@ pandoc: '3.2'
 pkgdown: 2.0.9
 pkgdown_sha: ~
 articles: {}
-last_built: 2024-06-02T06:31Z
+last_built: 2024-06-03T07:13Z
 urls:
   reference: https://uconn-scs.github.io/metastat/reference
   article: https://uconn-scs.github.io/metastat/articles