Controlling font size of check_model() axis labels and plot titles

[peter-bastian]

I am looking for a way to control (in my case, reduce) the font sizes used in the plots generated by check_model().

Is there some way to do this? I have tried passing arguments to the print() method after capturing the check_model() output in a variable, like this:
plots <- check_model(m)
print(plots, main="this is a test")

But that does not work. Seems like I could pass something via the ... in check_model() to be used by see, but that is unclear. Would another solution be to create my own custom see theme?

I have also seen reference to a "base" parameter that could work to scale the fonts.

[bwiernik]

The current release of see doesn't make it easy to access these plots for further tweaking.

If you run the function below, you can then do:

diagnostics <- check_model(mod)
diag_plots <- plot(diagnostics, return_list = TRUE)

to get a list of the individual plots. These are regular ggplot objects, so you can add more geoms, themes, etc. to them as you like.

plot.see_check_model <- function (x, style = theme_lucid, 
                                  colors = c("#3aaf85", "#1b6ca8", "#cd201f"), 
                                  return_list = FALSE,
                                  ...) 
{
  orig_x <- x
  p <- list()
  if (isTRUE(return_list)) {
    panel <- FALSE
  } else {
    panel <- attr(x, "panel")
  }
  check <- attr(x, "check")
  size_point <- attr(x, "dot_size")
  size_line <- attr(x, "line_size")
  size_text <- attr(x, "text_size")
  alpha_level <- attr(x, "alpha")
  dot_alpha_level <- attr(x, "dot_alpha")
  detrend <- attr(x, "detrend")
  if (missing(style) && !is.null(attr(x, "theme"))) {
    theme_style <- unlist(strsplit(attr(x, "theme"), "::", 
                                   fixed = TRUE))
    style <- get(theme_style[2], asNamespace(theme_style[1]))
  }
  if (missing(colors)) {
    colors <- attr(x, "colors")
  }
  if (is.null(colors)) {
    colors <- c("#3aaf85", "#1b6ca8", "#cd201f")
  }
  colors <- unname(colors)
  if (is.null(alpha_level)) {
    alpha_level <- 0.2
  }
  if (is.null(dot_alpha_level)) {
    dot_alpha_level <- 0.8
  }
  if (is.null(check)) 
    check <- "all"
  if ("NCV" %in% names(x) && any(c("ncv", "linearity", "all") %in% 
                                 check)) {
    p$NCV <- see:::.plot_diag_linearity(x$NCV, size_point, size_line, 
                                  alpha_level, theme_style = style, colors = colors, 
                                  dot_alpha_level = dot_alpha_level)
  }
  if ("HOMOGENEITY" %in% names(x) && any(c("homogeneity", 
                                           "all") %in% check)) {
    p$HOMOGENEITY <- see:::.plot_diag_homogeneity(x$HOMOGENEITY, 
                                            size_point, size_line, alpha_level, theme_style = style, 
                                            colors = colors, dot_alpha_level = dot_alpha_level)
  }
  if ("VIF" %in% names(x) && any(c("vif", "all") %in% check)) {
    p$VIF <- see:::.plot_diag_vif(x$VIF, theme_style = style, 
                            colors = colors)
  }
  if ("OUTLIERS" %in% names(x) && any(c("outliers", "all") %in% 
                                      check)) {
    p$OUTLIERS <- see:::.plot_diag_outliers_new(x$INFLUENTIAL, 
                                          size_text = size_text, size_line = size_line, theme_style = style, 
                                          colors = colors, dot_alpha_level = dot_alpha_level)
  }
  if ("QQ" %in% names(x) && any(c("qq", "all") %in% check)) {
    p$QQ <- see:::.plot_diag_qq(x$QQ, size_point, size_line, alpha_level = alpha_level, 
                          detrend = detrend, theme_style = style, colors = colors, 
                          dot_alpha_level = dot_alpha_level)
  }
  if ("NORM" %in% names(x) && any(c("normality", "all") %in% 
                                  check)) {
    p$NORM <- see:::.plot_diag_norm(x$NORM, size_line, alpha_level = alpha_level, 
                              theme_style = style, colors = colors)
  }
  if ("REQQ" %in% names(x) && any(c("reqq", "all") %in% check)) {
    ps <- see:::.plot_diag_reqq(x$REQQ, size_point, size_line, 
                          alpha_level = alpha_level, theme_style = style, 
                          colors = colors, dot_alpha_level = dot_alpha_level)
    for (i in 1:length(ps)) {
      p[[length(p) + 1]] <- ps[[i]]
    }
  }
  if (panel) {
    suppressWarnings(suppressMessages(do.call(plots, p)))
  }
  else {
    suppressWarnings(suppressMessages(p))
  }
}

[krassowski]

It would be great if we could pass a named list of lists with extra ggplot layers to be applied to specific plots. In my use case I experience overplotting on the collinearity plot and would like to add scale_x_discrete(guide = guide_axis(n.dodge = 2)) to this one plot.

So maybe something like:

check_model(
    model,
    extra_layers = list(
        "collinearity"=list(scale_x_discrete(guide = guide_axis(n.dodge = 2))),
        "normality"=list(theme_bw(), labs(subtitle="my text"))
    )
)

What do you think?

[bwiernik]

Now that see uses patchwork to combine plots, I think it makes more sense to leave this sort of tweaking to editing post-plot-creation:

library(performance)
library(see)
library(ggplot2)
m <- lm(mpg ~ factor(cyl) + disp + hp, data = mtcars)
pp <- check_model(m)

p <- plot(pp)
p

p[[3]] <- p[[3]] + labs(y = "Variance Inflation\nFactor (VIF)")
p[[6]] <- p[[6]] + xlim(c(-5, 6))
p

^{Created on 2021-07-26 by the reprex package (v2.0.0)}