Add geometric and harmonic means

DESCRIPTION

@@ -1,7 +1,7 @@
 Package: jaspDescriptives
 Type: Package
 Title: Descriptives Module for JASP
-Version: 0.95.0
+Version: 0.95.1
 Date: 2022-09-12
 Author: JASP Team
 Website: jasp-stats.org

R/descriptives.R

@@ -380,7 +380,8 @@ DescriptivesInternal <- function(jaspResults, dataset, options) {
     stats$addFootnote(message = gettextf("Excluded %1$i rows from the analysis that correspond to the missing values of the split-by variable %2$s", numberMissingSplitBy, options$splitBy))
 
   stats$dependOn(c(
-    "splitBy", "variables", "quantilesForEqualGroupsNumber", "percentileValues", "mode", "median", "mean",
+    "splitBy", "variables", "quantilesForEqualGroupsNumber", "percentileValues", "mode", "median",
+    "meanArithmetic", "meanGeometric", "meanHarmonic",
     "seMean", "sd", "coefficientOfVariation", "variance", "skewness", "kurtosis", "shapiroWilkTest",
     "range", "iqr", "mad", "madRobust", "minimum", "maximum", "sum", "quartiles", "quantilesForEqualGroups",
     "percentiles", "descriptivesTableTransposed", "valid", "missing", "meanCi", "meanCiLevel", "meanCiMethod",
@@ -414,8 +415,8 @@ DescriptivesInternal <- function(jaspResults, dataset, options) {
     meanCiOvertitle <- NULL
     sdCiOvertitle <- NULL
     varianceCiOvertitle <- NULL
-    meanCiUbTitle <- gettextf("%s%% CI Mean Upper", formattedMeanCiPercent)
-    meanCiLbTitle <- gettextf("%s%% CI Mean Lower", formattedMeanCiPercent)
+    meanCiUbTitle <- gettextf("%s%% CI A. Mean Upper", formattedMeanCiPercent)
+    meanCiLbTitle <- gettextf("%s%% CI A. Mean Lower", formattedMeanCiPercent)
     sdCiUbTitle <- gettextf("%s%% CI Std. Dev. Upper", formattedSdCiPercent)
     sdCiLbTitle <- gettextf("%s%% CI Std. Dev. Lower", formattedSdCiPercent)
     varianceCiUbTitle <- gettextf("%s%% CI Variance Upper", formattedVarianceCiPercent)
@@ -426,10 +427,12 @@ DescriptivesInternal <- function(jaspResults, dataset, options) {
   if (options$missing)                        stats$addColumnInfo(name="Missing",                     title=gettext("Missing"),                 type="integer")
   if (options$mode)                           stats$addColumnInfo(name="Mode",                        title=gettext("Mode"),                    type="mixed")
   if (options$median)                         stats$addColumnInfo(name="Median",                      title=gettext("Median"),                  type="number")
-  if (options$mean)                           stats$addColumnInfo(name="Mean",                        title=gettext("Mean"), 				            type="number")
-  if (options$seMean)                         stats$addColumnInfo(name="Std. Error of Mean",          title=gettext("Std. Error of Mean"),      type="number")
+  if (options$meanArithmetic)                 stats$addColumnInfo(name="MeanArithmetic",              title=gettext("Mean (arithmetic)"),       type="number")
+  if (options$seMean)                         stats$addColumnInfo(name="Std. Error of Mean",          title=gettext("Std. Error of A. Mean"),   type="number")
   if (options$meanCi) {                       stats$addColumnInfo(name="MeanCILB",                    title=meanCiLbTitle,                      type="number", overtitle = meanCiOvertitle)
                                               stats$addColumnInfo(name="MeanCIUB",                    title=meanCiUbTitle,                      type="number", overtitle = meanCiOvertitle)}
+  if (options$meanGeometric)                  stats$addColumnInfo(name="MeanGeometric",               title=gettext("Mean (geometric)"),        type="number")
+  if (options$meanHarmonic)                   stats$addColumnInfo(name="MeanHarmonic",                title=gettext("Mean (harmonic)"),         type="number")
   if (options$sd)                             stats$addColumnInfo(name="Std. Deviation",              title=gettext("Std. Deviation"),          type="number")
   if (options$sdCi) {                         stats$addColumnInfo(name="SdCILB",                      title=sdCiLbTitle,                        type="number", overtitle = sdCiOvertitle)
                                               stats$addColumnInfo(name="SdCIUB",                      title=sdCiUbTitle,                        type="number", overtitle = sdCiOvertitle)}
@@ -581,9 +584,11 @@ DescriptivesInternal <- function(jaspResults, dataset, options) {
   resultsCol[["Missing"]]                 <- if (options$missing) rows - length(na.omitted)
 
   if (columnType == "scale") {
-    resultsCol[["Median"]]                  <- .descriptivesDescriptivesTable_subFunction_OptionChecker(options$median,            na.omitted, median)
-    resultsCol[["Mean"]]                    <- .descriptivesDescriptivesTable_subFunction_OptionChecker(options$mean,              na.omitted, mean)
-    resultsCol[["Std. Error of Mean"]]      <- .descriptivesDescriptivesTable_subFunction_OptionChecker(options$seMean, na.omitted, function(param) { sd(param)/sqrt(length(param))} )
+    resultsCol[["Median"]]                  <- .descriptivesDescriptivesTable_subFunction_OptionChecker(options$median,             na.omitted, median)
+    resultsCol[["MeanArithmetic"]]          <- .descriptivesDescriptivesTable_subFunction_OptionChecker(options$meanArithmetic,     na.omitted, mean)
+    resultsCol[["Std. Error of Mean"]]      <- .descriptivesDescriptivesTable_subFunction_OptionChecker(options$seMean,             na.omitted, function(param) { sd(param)/sqrt(length(param))} )
+    resultsCol[["MeanGeometric"]]           <- .descriptivesDescriptivesTable_subFunction_OptionChecker(options$meanGeometric,      na.omitted, .geometricMean )
+    resultsCol[["MeanHarmonic"]]            <- .descriptivesDescriptivesTable_subFunction_OptionChecker(options$meanHarmonic,       na.omitted, .harmonicMean )
     resultsCol[["Std. Deviation"]]          <- .descriptivesDescriptivesTable_subFunction_OptionChecker(options$sd, na.omitted, sd)
     resultsCol[["Coefficient of Variation"]]<- .descriptivesDescriptivesTable_subFunction_OptionChecker(options$coefficientOfVariation,      na.omitted, function(param) { sd(param) / mean(param)})
     resultsCol[["MAD"]]                     <- .descriptivesDescriptivesTable_subFunction_OptionChecker(options$mad,               na.omitted, function(param) { mad(param, constant = 1) } )
@@ -2109,7 +2114,7 @@ DescriptivesInternal <- function(jaspResults, dataset, options) {
   container[[plotName]] <- jaspPlot
 }
 
-.descriptivesHeatmapAggregateData <- function(variable, groups, fun = c("identity", "mean", "median", "mode", "length")) {
+.descriptivesHeatmapAggregateData <- function(variable, groups, fun = c("identity", "meanArithmetic", "median", "mode", "length")) {
   fun <- match.arg(fun)
   mode <- function(x) {
     levels <- levels(x)
@@ -2523,3 +2528,19 @@ DescriptivesInternal <- function(jaspResults, dataset, options) {
 
   container[[plotName]] <- densPlot
 }
+
+.geometricMean <- function(x) {
+  if (any(x <= 0)) return(NA)
+
+  return(
+    exp(mean(log(x)))
+  )
+}
+
+.harmonicMean <- function(x) {
+  if (any(x <= 0)) return(NA)
+
+  return(
+    1 / mean(1/x)
+  )
+}

R/descriptivesWrapper.R

@@ -41,8 +41,8 @@
 #' @param customHistogramPosition, Options for separate bins of the histogram
 #' \itemize{
 #'   \item \code{"identity"}: Identity: Bars are layered on top of each other, with transparency often used to distinguish overlapping data.
-#'   \item \code{"stack"}: Stack: Bars are stacked vertically, combining counts across categories within each bin.
 #'   \item \code{"dodge"}: Dodge: Bars are placed side-by-side, allowing for easy comparison of different categories within each bin.
+#'   \item \code{"stack"}: Stack: Bars are stacked vertically, combining counts across categories within each bin.
 #' }
 #' @param densityPlot, visualizes data distributions, adapting to the type of variable selected. For scale variables, it generates histograms or density plots to show continuous data distributions. For categorical variables, it creates bar plots displaying counts or proportions of each category.
 #'    Defaults to \code{FALSE}.
@@ -75,10 +75,14 @@
 #'    Defaults to \code{FALSE}.
 #' @param maximum, Maximum value of the data points.
 #'    Defaults to \code{TRUE}.
-#' @param mean, Arithmetic mean of the data points
+#' @param meanArithmetic, Arithmetic mean of the data points.
 #'    Defaults to \code{TRUE}.
 #' @param meanCiLevel, width of the confidence interval.
 #' @param meanCiMethod, How should the confidence interval be computed? By default, we use a `T model`, which yields results identical to a one-sample t-test. Alternative options are a normal model ($\bar{x} \pm z_{95} \times SE$), or `Bootstrap`.
+#' @param meanGeometric, Geometric mean of the data points; defined only for strictly positive variables.
+#'    Defaults to \code{FALSE}.
+#' @param meanHarmonic, Harmonic mean of the data points; defined only for strictly positive variables.
+#'    Defaults to \code{FALSE}.
 #' @param median, Median of the data points.
 #'    Defaults to \code{FALSE}.
 #' @param minimum, Minimum value of the data points.
@@ -100,7 +104,7 @@
 #' @param sdCi, a confidence interval for the standard deviation based on bootstrap samples.
 #'    Defaults to \code{FALSE}.
 #' @param sdCiMethod, How should the confidence interval be computed? By default, we use an analytical approach (chi-square). The alternative option is `Bootstrap`
-#' @param seMean, Standard error of the mean.
+#' @param seMean, Standard error of the arithmetic mean.
 #'    Defaults to \code{FALSE}.
 #' @param shapiroWilkTest, Shapiro-Wilk test
 #'    Defaults to \code{FALSE}.
@@ -121,126 +125,128 @@
 #'    Defaults to \code{FALSE}.
 #' @param varianceCiMethod, How should the confidence interval be computed? By default, we use a analytical approach (chi-square). The alternative option is `Bootstrap`
 Descriptives <- function(
-          data = NULL,
-          version = "0.95",
-          formula = NULL,
-          associationMatrixUse = "everything",
-          boxPlot = FALSE,
-          boxPlotBoxPlot = TRUE,
-          boxPlotColourPalette = FALSE,
-          boxPlotJitter = FALSE,
-          boxPlotOutlierLabel = FALSE,
-          boxPlotViolin = FALSE,
-          ciBootstrapSamples = 1000,
-          coefficientOfVariation = FALSE,
-          colorPalette = "colorblind",
-          correlation = FALSE,
-          correlationPlots = FALSE,
-          covariance = FALSE,
-          customHistogramPosition = "stack",
-          densityPlot = FALSE,
-          densityPlotCategoricalType = "count",
-          densityPlotSeparate = list(types = list(), value = ""),
-          densityPlotTransparency = 20,
-          densityPlotType = "density",
-          descriptivesTableTransposed = FALSE,
-          distributionAndCorrelationPlotDensity = FALSE,
-          distributionAndCorrelationPlotHistogramBinWidthType = "sturges",
-          distributionAndCorrelationPlotHistogramManualNumberOfBins = 30,
-          distributionAndCorrelationPlotRugMarks = FALSE,
-          distributionPlots = FALSE,
-          dotPlot = FALSE,
-          frequencyTables = FALSE,
-          frequencyTablesMaximumDistinctValues = 10,
-          heatmapDisplayValue = FALSE,
-          heatmapDisplayValueRelativeTextSize = 1,
-          heatmapHorizontalAxis = list(types = list(), value = ""),
-          heatmapLegend = FALSE,
-          heatmapPlot = FALSE,
-          heatmapStatisticContinuous = "mean",
-          heatmapStatisticDiscrete = "mode",
-          heatmapTileWidthHeightRatio = 1,
-          heatmapVerticalAxis = list(types = list(), value = ""),
-          intervalPlot = FALSE,
-          iqr = FALSE,
-          kurtosis = FALSE,
-          likertPlot = FALSE,
-          likertPlotAdjustableFontSize = "normal",
-          likertPlotAssumeVariablesSameLevel = FALSE,
-          mad = FALSE,
-          madRobust = FALSE,
-          maximum = TRUE,
-          mean = TRUE,
-          meanCi = FALSE,
-          meanCiLevel = 0.95,
-          meanCiMethod = "oneSampleTTest",
-          median = FALSE,
-          minimum = TRUE,
-          missing = TRUE,
-          mode = FALSE,
-          paretoPlot = FALSE,
-          paretoPlotRule = FALSE,
-          paretoPlotRuleCi = 0.95,
-          percentileValues = list(),
-          percentiles = FALSE,
-          pieChart = FALSE,
-          plotHeight = 320,
-          plotWidth = 480,
-          qqPlot = FALSE,
-          quantilesForEqualGroups = FALSE,
-          quantilesForEqualGroupsNumber = 4,
-          quartiles = FALSE,
-          range = FALSE,
-          scatterPlot = FALSE,
-          scatterPlotGraphTypeAbove = "density",
-          scatterPlotGraphTypeRight = "density",
-          scatterPlotLegend = TRUE,
-          scatterPlotRegressionLine = TRUE,
-          scatterPlotRegressionLineCi = TRUE,
-          scatterPlotRegressionLineCiLevel = 0.95,
-          scatterPlotRegressionLineType = "linear",
-          sd = TRUE,
-          sdCi = FALSE,
-          sdCiLevel = 0.95,
-          sdCiMethod = "chiSquaredModel",
-          seMean = FALSE,
-          shapiroWilkTest = FALSE,
-          skewness = FALSE,
-          splitBy = list(types = list(), value = ""),
-          statisticsValuesAreGroupMidpoints = FALSE,
-          stemAndLeaf = FALSE,
-          stemAndLeafScale = 1,
-          sum = FALSE,
-          valid = TRUE,
-          variables = list(types = list(), value = list()),
-          variance = FALSE,
-          varianceCi = FALSE,
-          varianceCiLevel = 0.95,
-          varianceCiMethod = "chiSquaredModel") {
+    data = NULL,
+    version = "0.95.1",
+    formula = NULL,
+    associationMatrixUse = "everything",
+    boxPlot = FALSE,
+    boxPlotBoxPlot = TRUE,
+    boxPlotColourPalette = FALSE,
+    boxPlotJitter = FALSE,
+    boxPlotOutlierLabel = FALSE,
+    boxPlotViolin = FALSE,
+    ciBootstrapSamples = 1000,
+    coefficientOfVariation = FALSE,
+    colorPalette = "colorblind",
+    correlation = FALSE,
+    correlationPlots = FALSE,
+    covariance = FALSE,
+    customHistogramPosition = "stack",
+    densityPlot = FALSE,
+    densityPlotCategoricalType = "count",
+    densityPlotSeparate = list(types = list(), value = ""),
+    densityPlotTransparency = 20,
+    densityPlotType = "density",
+    descriptivesTableTransposed = FALSE,
+    distributionAndCorrelationPlotDensity = FALSE,
+    distributionAndCorrelationPlotHistogramBinWidthType = "sturges",
+    distributionAndCorrelationPlotHistogramManualNumberOfBins = 30,
+    distributionAndCorrelationPlotRugMarks = FALSE,
+    distributionPlots = FALSE,
+    dotPlot = FALSE,
+    frequencyTables = FALSE,
+    frequencyTablesMaximumDistinctValues = 10,
+    heatmapDisplayValue = FALSE,
+    heatmapDisplayValueRelativeTextSize = 1,
+    heatmapHorizontalAxis = list(types = list(), value = ""),
+    heatmapLegend = FALSE,
+    heatmapPlot = FALSE,
+    heatmapStatisticContinuous = "mean",
+    heatmapStatisticDiscrete = "mode",
+    heatmapTileWidthHeightRatio = 1,
+    heatmapVerticalAxis = list(types = list(), value = ""),
+    intervalPlot = FALSE,
+    iqr = FALSE,
+    kurtosis = FALSE,
+    likertPlot = FALSE,
+    likertPlotAdjustableFontSize = "normal",
+    likertPlotAssumeVariablesSameLevel = FALSE,
+    mad = FALSE,
+    madRobust = FALSE,
+    maximum = TRUE,
+    meanArithmetic = TRUE,
+    meanCi = FALSE,
+    meanCiLevel = 0.95,
+    meanCiMethod = "oneSampleTTest",
+    meanGeometric = FALSE,
+    meanHarmonic = FALSE,
+    median = FALSE,
+    minimum = TRUE,
+    missing = TRUE,
+    mode = FALSE,
+    paretoPlot = FALSE,
+    paretoPlotRule = FALSE,
+    paretoPlotRuleCi = 0.95,
+    percentileValues = list(),
+    percentiles = FALSE,
+    pieChart = FALSE,
+    plotHeight = 320,
+    plotWidth = 480,
+    qqPlot = FALSE,
+    quantilesForEqualGroups = FALSE,
+    quantilesForEqualGroupsNumber = 4,
+    quartiles = FALSE,
+    range = FALSE,
+    scatterPlot = FALSE,
+    scatterPlotGraphTypeAbove = "density",
+    scatterPlotGraphTypeRight = "density",
+    scatterPlotLegend = TRUE,
+    scatterPlotRegressionLine = TRUE,
+    scatterPlotRegressionLineCi = TRUE,
+    scatterPlotRegressionLineCiLevel = 0.95,
+    scatterPlotRegressionLineType = "linear",
+    sd = TRUE,
+    sdCi = FALSE,
+    sdCiLevel = 0.95,
+    sdCiMethod = "chiSquaredModel",
+    seMean = FALSE,
+    shapiroWilkTest = FALSE,
+    skewness = FALSE,
+    splitBy = list(types = list(), value = ""),
+    statisticsValuesAreGroupMidpoints = FALSE,
+    stemAndLeaf = FALSE,
+    stemAndLeafScale = 1,
+    sum = FALSE,
+    valid = TRUE,
+    variables = list(types = list(), value = list()),
+    variance = FALSE,
+    varianceCi = FALSE,
+    varianceCiLevel = 0.95,
+    varianceCiMethod = "chiSquaredModel") {
 
-   defaultArgCalls <- formals(jaspDescriptives::Descriptives)
-   defaultArgs <- lapply(defaultArgCalls, eval)
-   options <- as.list(match.call())[-1L]
-   options <- lapply(options, eval)
-   defaults <- setdiff(names(defaultArgs), names(options))
-   options[defaults] <- defaultArgs[defaults]
-   options[["data"]] <- NULL
-   options[["version"]] <- NULL
+  defaultArgCalls <- formals(jaspDescriptives::Descriptives)
+  defaultArgs <- lapply(defaultArgCalls, eval)
+  options <- as.list(match.call())[-1L]
+  options <- lapply(options, eval)
+  defaults <- setdiff(names(defaultArgs), names(options))
+  options[defaults] <- defaultArgs[defaults]
+  options[["data"]] <- NULL
+  options[["version"]] <- NULL
 
 
-   if (!jaspBase::jaspResultsCalledFromJasp() && !is.null(data)) {
-      jaspBase::storeDataSet(data)
-   }
+  if (!jaspBase::jaspResultsCalledFromJasp() && !is.null(data)) {
+    jaspBase::storeDataSet(data)
+  }
 
-   if (!is.null(formula)) {
-      if (!inherits(formula, "formula")) {
-         formula <- as.formula(formula)
-      }
-      options$formula <- jaspBase::jaspFormula(formula, data)
-   }
-   optionsWithFormula <- c("associationMatrixUse", "colorPalette", "densityPlotSeparate", "distributionAndCorrelationPlotHistogramBinWidthType", "heatmapHorizontalAxis", "heatmapVerticalAxis", "likertPlotAdjustableFontSize", "meanCiMethod", "sdCiMethod", "splitBy", "variables", "varianceCiMethod")
-   for (name in optionsWithFormula) {
-      if ((name %in% optionsWithFormula) && inherits(options[[name]], "formula")) options[[name]] = jaspBase::jaspFormula(options[[name]], data)   }
+  if (!is.null(formula)) {
+    if (!inherits(formula, "formula")) {
+      formula <- as.formula(formula)
+    }
+    options$formula <- jaspBase::jaspFormula(formula, data)
+  }
+  optionsWithFormula <- c("associationMatrixUse", "colorPalette", "densityPlotSeparate", "distributionAndCorrelationPlotHistogramBinWidthType", "heatmapHorizontalAxis", "heatmapVerticalAxis", "likertPlotAdjustableFontSize", "meanCiMethod", "sdCiMethod", "splitBy", "variables", "varianceCiMethod")
+  for (name in optionsWithFormula) {
+    if ((name %in% optionsWithFormula) && inherits(options[[name]], "formula")) options[[name]] = jaspBase::jaspFormula(options[[name]], data)   }
 
-   return(jaspBase::runWrappedAnalysis("jaspDescriptives", "Descriptives", "Descriptives.qml", options, version, TRUE))
-}
+  return(jaspBase::runWrappedAnalysis("jaspDescriptives", "Descriptives", "Descriptives.qml", options, version, TRUE))
+}

inst/Description.qml

@@ -6,7 +6,7 @@ Description
 	name		: "jaspDescriptives"
 	title		: qsTr("Descriptives")
 	description	: qsTr("Explore the data with tables and plots")
-	version			: "0.95.0"
+	version			: "0.95.1"
 	author		: "JASP Team"
 	maintainer	: "JASP Team <jasp-stats.org>"
 	website		: "jasp-stats.org"

inst/Upgrades.qml

@@ -72,4 +72,14 @@ Upgrades
 			from: "heatmapDisplayLegend";		to: "heatmapLegend"								
 		}
 	}
+
+	Upgrade
+	{
+		functionName:	"Descriptives"
+		fromVersion:	"0.95.0"
+		toVersion:	"0.95.1"
+
+		// this resulted from adding geometric and harmonic means
+		ChangeRename { from: "mean";	to: "meanArithmetic" }
+	}
 }