Customizing Combined Plots

Ronald (Ryy) G. Thomas

2026-05-11

library(zzlongplot)
library(ggplot2)
library(patchwork)

Overview

When lplot() is called with plot_type = "both", it returns a patchwork object that places the observed values and change-from-baseline plots side by side. This object can be modified after creation using two patchwork operators:

• & applies a modification to all panels
• + applies a modification to the last panel only

This vignette demonstrates common post-creation customizations.

Example data

set.seed(42)
n <- 20

trial <- data.frame(
  subject_id = rep(1:n, each = 4),
  visit = rep(0:3, times = n),
  arm = rep(c("Drug", "Placebo"), each = 4 * n / 2)
)
trial$score <- 50 +
  ifelse(trial$arm == "Drug", -3, -0.5) * trial$visit +
  rnorm(nrow(trial), sd = 4)

Base combined plot

p <- lplot(trial, score ~ visit | arm, baseline_value = 0,
           plot_type = "both")
#> Warning: The `size` argument of `element_line()` is deprecated as of ggplot2 3.4.0.
#> ℹ Please use the `linewidth` argument instead.
#> ℹ The deprecated feature was likely used in the zzlongplot package.
#>   Please report the issue at <https://github.com/rgt47/zzlongplot/issues>.
#> This warning is displayed once per session.
#> Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
#> generated.
#> Warning: The `size` argument of `element_rect()` is deprecated as of ggplot2 3.4.0.
#> ℹ Please use the `linewidth` argument instead.
#> ℹ The deprecated feature was likely used in the zzlongplot package.
#>   Please report the issue at <https://github.com/rgt47/zzlongplot/issues>.
#> This warning is displayed once per session.
#> Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
#> generated.
p

Modifying all panels with &

The & operator passes a ggplot2 element to every panel in the composition. This is useful for global theme changes, font adjustments, or color overrides.

Change the theme

p & theme_classic()

Adjust font size

p & theme(text = element_text(size = 14))

Override group colors

p & scale_color_manual(values = c("Drug" = "#0072B2",
                                  "Placebo" = "#D55E00"))
#> Scale for colour is already present.
#> Adding another scale for colour, which will replace the existing scale.
#> Scale for colour is already present.
#> Adding another scale for colour, which will replace the existing scale.

Remove gridlines

p & theme(panel.grid.minor = element_blank(),
          panel.grid.major.x = element_blank())

Modifying a single panel with +

The + operator targets only the last panel (the change plot in a "both" layout). This allows selective modifications when the two panels need different treatment.

Add a reference line to the change plot

p + geom_hline(yintercept = 0, linetype = "dashed",
               color = "grey50")

Adding overall annotations

The plot_annotation() function from patchwork adds titles, subtitles, captions, and tag labels that span the entire composition, sitting above or below the individual panel labels.

p + plot_annotation(
  title = "Efficacy Results: Drug vs Placebo",
  subtitle = "Study 001, ITT Population",
  caption = "Error bars represent standard error of the mean.",
  tag_levels = "A"
)

Panel tags (tag_levels = "A") label each panel as (A), (B), etc., which is a common requirement for journal submissions.

Styling the annotation

Annotation text inherits from the plot theme but can be overridden:

p + plot_annotation(
  title = "Efficacy Results",
  tag_levels = "A",
  theme = theme(
    plot.title = element_text(face = "bold", size = 16),
    plot.tag = element_text(face = "bold")
  )
)

Combining multiple modifications

The operators can be chained. Use & for global changes first, then + for composition-level annotations.

(p &
  theme_minimal() &
  theme(legend.position = "bottom",
        text = element_text(size = 12))) +
  plot_annotation(
    title = "Phase III Trial: Primary Endpoint",
    tag_levels = "A"
  )

Controlling layout

The layout itself can be adjusted after creation using plot_layout():

p + plot_layout(widths = c(2, 1))

This produces a wider observed panel and a narrower change panel, which can be useful when the change plot carries less visual complexity.

Summary

Operator	Scope	Example
&	All panels	p & theme_bw()
+	Last panel (or annotation)	p + geom_hline(...)
+ plot_annotation()	Entire composition	titles, tags, captions
+ plot_layout()	Entire composition	widths, heights, guides

Because lplot() returns a standard patchwork object, any technique documented in the patchwork package applies directly.