v.0.0.7 get information from SuperPlot

Author: quantixed

DESCRIPTION

@@ -1,6 +1,6 @@
 Package: SuperPlotR
 Title: Making SuperPlots in R
-Version: 0.0.6
+Version: 0.0.7
 Authors@R:
     person(given = "Stephen J",
            family = "Royle",
@@ -25,7 +25,8 @@ Imports:
     ggplot2,
     dplyr,
     ggforce,
-    cowplot
+    cowplot,
+    rlang
 URL: https://github.com/quantixed/SuperPlotR, https://quantixed.github.io/SuperPlotR/
 BugReports: https://github.com/quantixed/SuperPlotR/issues
 Suggests: 

NAMESPACE

@@ -6,11 +6,14 @@ export(get_fp_colour)
 export(get_sp_colours)
 export(get_sp_shapes)
 export(get_sp_stats)
+export(get_sp_summary)
+export(representative)
 export(superplot)
 import(cowplot)
 import(dplyr)
 import(ggforce)
 import(ggplot2)
+importFrom(rlang,":=")
 importFrom(stats,TukeyHSD)
 importFrom(stats,kruskal.test)
 importFrom(stats,median)

NEWS.md

@@ -1,5 +1,14 @@
 # SuperPlotR (development version)
 
+## SuperPlotR 0.0.7
+
+* Added ability to find representative datapoints with `representative()`
+* Added ability to retrieve information about superplot using the option
+  `info = TRUE`. User is prompted to do this if there are unequal numbers of
+  replicates in each condition
+* Added `get_sp_summary()` to retrieve the data frame that is overlaid
+  on the superplot
+
 ## SuperPlotR 0.0.6
 
 * Fixed factoring bug.

R/get_sp_info.R

@@ -0,0 +1,132 @@
+#' Information about a SuperPlot
+#'
+#' This function prints information about a SuperPlot, it runs internally from
+#' superplot()
+#'
+#' @param df data frame with at least three columns: meas, cond, repl
+#' @param meas character name of column with measurement (e.g. intensity)
+#' @param cond character name of column with condition (e.g. Control, WT)
+#' @param repl character name of column with replicate (e.g. unique experiment
+#' identifiers)
+#' @param pal argument passed to pal
+#' @param xlab argument passed to xlab
+#' @param ylab argument passed to ylab
+#' @param datadist argument passed to datadist
+#' @param size argument passed to size
+#' @param alpha argument passed to alpha
+#' @param bars argument passed to bars
+#' @param linking argument passed to linking
+#' @param rep_summary argument passed to rep_summary
+#' @param shapes argument passed to shapes
+#' @param fsize argument passed to fsize
+#' @param gg argument passed to gg
+#' @param stats argument passed to stats
+#' @param stats_test argument passed to stats_test
+#' @param ... additional arguments
+#'
+#' @import dplyr
+#' @importFrom stats sd median
+#' @importFrom rlang :=
+#'
+#' @returns none
+#' @keywords internal
+
+# this function will take all the arguments sent from superplot
+get_sp_info <- function(df,
+                      meas, cond, repl,
+                      pal, xlab, ylab, datadist, size, alpha, bars,
+                      linking, rep_summary, shapes, fsize, gg,
+                      stats, stats_test,
+                      ...) {
+  ncond <- nrepl <- NULL
+  rep_mean <- rep_median <- NULL
+
+  # args are already validated
+
+  # how many unique values in cond and repl?
+  ncond <- df %>%
+    pull(!!sym(cond)) %>%
+    unique() %>%
+    length()
+  nrepl <- df %>%
+    pull(!!sym(repl)) %>%
+    unique() %>%
+    length()
+
+  # calculate summary statistics
+  summary_df <- get_sp_summary(df = df,
+                             meas = meas, cond = cond, repl = repl)
+
+  # get colour values for the repl column
+  sp_colours <- get_sp_colours(nrepl, pal)
+  sp_shapes <- get_sp_shapes(nrepl, shapes)
+
+  # add repl from summary_df to sp_colour and sp_shapes
+  # repl is converted to factor so that the colours and shapes
+  # are assigned in the order of the factor levels
+  # but only if repl is not already a factor
+  if (!is.factor(df[[repl]])) {
+    summary_df <- summary_df %>%
+      mutate(!!sym(repl) := factor(!!sym(repl), levels = unique(df[[repl]])))
+  }
+  summary_df <- summary_df %>%
+    mutate(
+      sp_colour = factor(!!sym(repl), levels = unique(df[[repl]]),
+                         labels = sp_colours),
+      sp_shape = factor(!!sym(repl), levels = unique(df[[repl]]),
+                        labels = sp_shapes)
+    )
+
+  ## print information about the plot
+  message("SuperPlot information")
+  message("=====================")
+  message("Number of conditions: ", ncond)
+  message("Number of replicates: ", nrepl)
+  message("Number of data points: ", nrow(df))
+  message("Number of summary points: ", nrow(summary_df))
+  message("=====================")
+  message("Colour palette: ", pal)
+  message("Data distribution: ", datadist)
+  message("Summary statistic: ", rep_summary)
+  if( bars != "") {
+    message("Bars: ", bars)
+  } else {
+    message("No bars")
+  }
+  message("X-axis label: ", xlab)
+  message("Y-axis label: ", ylab)
+  if (linking == TRUE) {
+    message("Linking: ", linking)
+  }
+  if (shapes == TRUE) {
+    message("Shapes: ", shapes)
+  }
+  message("Point sizes: ", size[1], " (individual), ",
+          size[2], " (summary)")
+  message("Alpha for points: ", alpha[1], " (individual), ",
+          alpha[2], " (summary)")
+  message("Font size: ", fsize)
+  if (stats== TRUE) {
+    message("Statistics: ", stats_test)
+  } else {
+    message("No statistics")
+  }
+  message("=====================")
+  message("Colours for replicates: ", paste(sp_colours, collapse = ", "))
+  message("Shapes for replicates: ", paste(sp_shapes, collapse = ", "))
+
+  if (nrow(summary_df) != ncond * nrepl) {
+    s_summary_df <- summary_df %>%
+      group_by(!!sym(cond)) %>%
+      summarise(
+        n_replicates = n()
+      )
+    message("Unequal number of replicates per condition:")
+
+    print(s_summary_df)
+  }
+
+  message("=====================")
+  message("Summary statistics:")
+  print(summary_df)
+}

R/get_sp_summary.R

@@ -0,0 +1,60 @@
+#' Get summary statistics for SuperPlot data
+#'
+#' This function calculates summary statistics (mean and median) for each
+#' combination of condition and replicate in a SuperPlot dataset.
+#'
+#' It is useful to get teh summary data and computing further stats or making
+#' plots. The function only needs 4 parameters, anything else is ignored. This
+#' means you can simply exchange get_sp_summary() for your superplot() call and
+#' get the correct data frame to work with.
+#'
+#' @param df A data frame containing the experimental data
+#' @param meas Character string specifying the name of the column containing
+#'   the measurements/values to summarize (e.g., "intensity", "speed")
+#' @param cond Character string specifying the name of the column containing
+#'   the experimental conditions (e.g., "Treatment", "Genotype")
+#' @param repl Character string specifying the name of the column containing
+#'   the replicate identifiers (e.g., "Replicate", "Experiment")
+#' @param ... Additional arguments (ignored)
+#'
+#' @return A data frame with columns for condition, replicate, rep_mean
+#'   (mean of measurements), and rep_median (median of measurements)
+#'
+#' @import dplyr
+#' @importFrom stats median
+#'
+#' @export
+#'
+#' @examples
+#' # Using the built-in dataset
+#' get_sp_summary(lord_jcb, "Speed", "Treatment", "Replicate")
+#'
+get_sp_summary <- function(df, meas, cond, repl, ...) {
+  # Validate inputs
+  if (!is.data.frame(df)) {
+    stop("df must be a data frame")
+  }
+
+  if (!meas %in% names(df)) {
+    stop("meas column '", meas, "' not found in data frame")
+  }
+
+  if (!cond %in% names(df)) {
+    stop("cond column '", cond, "' not found in data frame")
+  }
+
+  if (!repl %in% names(df)) {
+    stop("repl column '", repl, "' not found in data frame")
+  }
+
+  # Calculate summary statistics grouped by condition and replicate
+  sdf <- df %>%
+    group_by(!!sym(cond), !!sym(repl)) %>%
+    summarise(
+      rep_mean = mean(!!sym(meas), na.rm = TRUE),
+      rep_median = median(!!sym(meas), na.rm = TRUE),
+      .groups = "drop"
+    )
+
+  return(sdf)
+}