Atomic Correlation Pairs Plot — CorPairsPlotAtomic • plotthis

Core implementation for drawing a correlation pairs (scatterplot matrix) grid. This is the workhorse behind the exported CorPairsPlot — it takes a single data frame (no split_by support) and returns a patchwork object.

The grid arranges all pairwise scatter plots of selected columns. The upper or lower triangle displays correlation tiles (fill) while the opposite triangle displays scatter plots with regression lines. The diagonal can show density plots, violin plots, histograms, box plots, or a simple diagonal line.

The function supports four layout orientations (layout), three correlation methods (cor_method), configurable diagonal plots using other plotthis functions, and custom correlation tile formatting.

Usage

CorPairsPlotAtomic(
  data,
  columns = NULL,
  group_by = NULL,
  group_by_sep = "_",
  group_name = NULL,
  diag_type = NULL,
  diag_args = list(),
  layout = c(".\\", "\\.", "/.", "./"),
  cor_method = c("pearson", "spearman", "kendall"),
  cor_palette = "RdBu",
  cor_palcolor = NULL,
  cor_size = 3,
  cor_format = "corr: {round(corr, 2)}",
  cor_fg = "black",
  cor_bg = "white",
  cor_bg_r = 0.1,
  theme = "theme_this",
  theme_args = list(),
  palette = ifelse(is.null(group_by), "Spectral", "Paired"),
  palcolor = NULL,
  palreverse = FALSE,
  title = NULL,
  subtitle = NULL,
  facet_by = NULL,
  legend.position = "right",
  legend.direction = "vertical",
  seed = 8525,
  ...
)

Arguments

data: A data frame.
columns: A character vector of column names to include in the pairs plot. When NULL (default), all columns except group_by are used. At least two columns are required.
group_by: A character vector of column names to colour the scatter points by. Each unique combination becomes a separate group. Required for diag_type = "violin" and diag_type = "box". Multiple columns are concatenated with group_by_sep.
group_by_sep: A character string to separate concatenated group_by columns. Default "_".
group_name: A character string used as the colour legend title in the scatter plots. When NULL, the group_by column name is used.
diag_type: A character string specifying the plot type for diagonal cells. One of "density", "violin", "histogram", "box", or "none" (diagonal line). Default: "density" (no group_by) or "violin" (with group_by).
diag_args: A named list of additional arguments passed to the diagonal plot function (DensityPlot, ViolinPlot, Histogram, or BoxPlot). Default: list().
layout: A character string specifying the layout orientation. One of the following codes (dot = scatter, backslash/slash = diagonal): .\, \\., /., ./. Default: .\.
cor_method: A character string specifying the correlation method for the fill tiles. One of "pearson", "spearman", "kendall". Default: "pearson".
cor_palette: A character string specifying the colour palette for the correlation fill tiles. Default: "RdBu".
cor_palcolor: A character vector of custom colours used to create the correlation tile palette. When NULL, the palette's default colours are used.
cor_size: A numeric value specifying the font size of the correlation text in the fill tiles. Default: 3.
cor_format: A character string specifying a glue template for formatting the correlation text. The template is evaluated by glue::glue() with access to corr (the correlation value), x, and y (the column names). Default: "corr: \{round(corr, 2)\}".
cor_fg: A character string specifying the colour of the correlation text. Default: "black".
cor_bg: A character string specifying the background colour of the correlation text boxes. Default: "white".
cor_bg_r: A numeric value specifying the corner radius of the correlation text background boxes. Default: 0.1.
theme: A character string or a theme class (i.e. ggplot2::theme_classic) specifying the theme to use. Default is "theme_this".
theme_args: A list of arguments to pass to the theme function.
palette: A character string specifying the palette to use. A named list or vector can be used to specify the palettes for different split_by values.
palcolor: A character string specifying the color to use in the palette. A named list can be used to specify the colors for different split_by values. If some values are missing, the values from the palette will be used (palcolor will be NULL for those values).
palreverse: A logical value indicating whether to reverse the palette. Default is FALSE.
title: A character string specifying the title of the plot. A function can be used to generate the title based on the default title. This is useful when split_by is used and the title needs to be dynamic.
subtitle: A character string specifying the subtitle of the plot.
facet_by: A character string specifying the column name of the data frame to facet the plot. Otherwise, the data will be split by split_by and generate multiple plots and combine them into one using patchwork::wrap_plots
legend.position: A character string specifying the position of the legend. if waiver(), for single groups, the legend will be "none", otherwise "right".
legend.direction: A character string specifying the direction of the legend.
seed: The random seed to use. Default is 8525.
...: Additional arguments passed to the underlying CorPlot for the scatter plot cells.

Value

A patchwork object with height and width attributes (in inches) attached.

Details

theme and theme_args are supported and passed to each individual cell plot.

Architecture

ggplot dispatch — selects gglogger::ggplot or ggplot2::ggplot based on getOption("plotthis.gglogger.enabled").
facet_by guard — raises an error if facet_by is provided (not supported in pairs plots; use split_by instead).
Column resolution — when columns is NULL, all columns (except group_by) are used. Otherwise, columns are validated via check_columns(). At least two columns are required.
Default diag_type — when diag_type is NULL, it defaults to "density" (no group_by) or "violin" (with group_by).
Layout determination — get_plot_info() is called for each cell (i, j) to determine:
- Type: layout (diagonal line), "cor" (scatter plot with regression), "fill" (correlation tile).
- Axis/label positions: which sides of the cell should show x-axis, y-axis, x-label, and y-label, based on the cell's location in the matrix and the chosen layout.
Cell rendering — get_plot() draws each cell based on its info$type:
- Diagonal, diag_type = "none" — draws an X or backslash diagonal line via geom_line().
- Diagonal, "density" — delegates to DensityPlot().
- Diagonal, "violin" — delegates to ViolinPlot(); requires group_by.
- Diagonal, "histogram" — delegates to Histogram().
- Diagonal, "box" — delegates to BoxPlot(); requires group_by.
- Off-diagonal, "cor" — delegates to CorPlot() for the scatter plot with regression line.
- Off-diagonal, "fill" — draws a geom_tile() filled by the correlation value with scale_fill_gradientn() and displays the formatted correlation via ggrepel::geom_text_repel().
Axis/label positioning — axes, axis titles, and labels are positioned at the margins of the full grid based on info$xaxis, info$yaxis, info$xlab, info$ylab using scale_x_*/scale_y_* with appropriate position arguments.
Layout assembly — all cell plots are arranged into a matrix via patchwork::wrap_plots() with ncol = length(columns). Guides are collected, and the title/subtitle are added via plot_annotation(). The result is wrapped in patchwork::wrap_elements() so the title displays correctly.
Dimension calculation — calculate_plot_dimensions() computes height and width attributes from base_height = sqrt(length(columns)) * 4, a fixed aspect ratio of 1, and legend metrics.