Plot scaled score variables as radar charts

This function plots the scaled score variables as radar charts using the superb package. It can plot radars for any combination of the Group, Cluster, or Subcluster grouping variables and selections of original or reduced study variables, typically obtained through get_longer() and filter_study_variables() (see examples).

The function allows for customisation of the radar charts with various options. Note that the default sizes are pretty small, this was designed to render best as small dimension PDF vector figures for journals.

Usage

plot_score_radars(
  df,
  groups,
  txt_big = 7,
  txt_smol = 5,
  dot_size = 0.8,
  lw_line = 0.2,
  lw_error = 0.2,
  y_off = 40,
  r_off = 0,
  l_off = 0,
  v_off = 0,
  key = 3,
  ...
)

Arguments

df

A data frame containing the variables to be plotted in long format, with the variable names in a Variable column and the associated values in a value column. This is for example the output of get_longer(study_data).

groups

An grouping variable among Group, Cluster, or Subcluster. Can be quoted or unquoted. The function will plot a radars for each level of this variable.

txt_big

Size of the text for the x-axis labels. Default is 7.

txt_smol

Size of the text for the y-axis labels. Default is 5.

dot_size

Size of the dots in the radar charts. Default is 0.8.

lw_line, lw_error

Line width for the main plot lines and error bars, respectively. Default is 0.2 for both.

y_off

Offset for the y-axis text to centre it. Default is 40.

r_off, l_off, v_off

Right, left and vertical offset for the plot margins. Default is 0 for all three.

key

Size of the legend key in mm. Default is 3.

...

Additional arguments passed to the superb function.

Value

A ggplot object containing the radar chart.

Examples

df_merged_long <- merge_clusters(
  df_raw     = study_data,
  df_red     = scale_reduce_vars(study_data),
  clustering = cluster_selected_vars(study_data)
  ) |>
  scale_vars() |>
  get_longer()

if (require("superb", quietly = TRUE)) {
# Groups on the original variables
p1 <-
  df_merged_long |>
  filter_study_variables("original") |>
  plot_score_radars(Group, r_off = 6, l_off = 6)

# Clusters on the reduced + validation variables
p2 <-
  df_merged_long |>
  filter_study_variables("reduced") |>
  plot_score_radars(Cluster, r_off = 6, l_off = 6)

# Subclusters on the reduced variables
p3 <-
  df_merged_long |>
  filter_study_variables("reduced_strict") |>
  plot_score_radars(Subcluster, r_off = 6, l_off = 6)

# Clusters on all the variables
p4 <- df_merged_long |> plot_score_radars(Cluster, r_off = 6, l_off = 6)

if (require("patchwork", quietly = TRUE)) {
 print(p1 + p2)
 print(p3 + p4)
} else print(p1)
}