ggplot2 theme_classic() in R: L-Shape Axis Theme for Papers

theme_classic() applies a publication-ready theme to ggplot2 plots with a white background, no gridlines, and L-shape axis lines along the bottom and left edges only. It is the default choice for scientific journals and academic papers that prefer visible axis ticks over background grids.

theme_classic()                                          # default classic theme
theme_classic(base_size = 14)                            # larger text for journals
theme_classic(base_family = "serif")                     # serif font for papers
theme_classic(base_size = 11, base_line_size = 0.5)      # thicker axis lines
p + theme_classic()                                      # apply to existing plot
p + theme_classic() + theme(legend.position = "top")     # move legend after base
p + theme_classic() + theme(panel.grid.major.y = element_line()) # add y-grid back
theme_set(theme_classic())                               # set as session default

What theme_classic() does in ggplot2

theme_classic() strips ggplot2's panel grid and replaces the full panel border with two axis lines along the bottom and left edges. The result looks like a hand-drawn scientific figure: white panel, visible ticks, no visual noise.

It is one of seven complete themes in ggplot2 (theme_grey, theme_minimal, theme_bw, theme_classic, theme_light, theme_dark, theme_void). Add it with + theme_classic() after geoms and labels.

Internally, theme_classic() builds on theme_bw() by removing panel gridlines and the panel border, then drawing a single line for axis.line.x and axis.line.y only. Axis ticks stay visible, which is why most journals accept it as is.

Syntax and arguments

theme_classic() accepts four optional arguments that scale the entire theme proportionally. The defaults match the standard 6-by-4 inch ggplot2 figure.

The four arguments:

• base_size: numeric. Base font size in points (default 11). All text elements scale relative to this value.
• base_family: character. Base font family (default ""). Common values are "serif", "sans", "mono", or a registered custom font name.
• base_line_size: numeric. Base width for line elements (default base_size/22). Controls axis line and tick thickness.
• base_rect_size: numeric. Base width for rectangle borders (default base_size/22). Has no visible effect on most plots since the panel border is already removed.

Common theme_classic() patterns

These six patterns cover the everyday uses of theme_classic(). Each example uses a different base plot so you can see how the theme adapts across scatter, bar, and line geoms.

1. Apply theme_classic() to a scatter plot

The grey panel disappears, the panel border is gone, and only the bottom and left axes are drawn. Axis ticks remain visible so readers can read values without a grid.

2. Scale up base_size for a journal figure

base_size = 14 enlarges every text element together. Most two-column journal layouts require 9 to 12pt labels at the final printed size; bumping base_size to 14 and saving at 4 inches wide keeps the rendered text in that range.

3. Switch to a serif font for papers

"serif", "sans", and "mono" work without setup. For Times New Roman or Computer Modern (LaTeX defaults), register the font with extrafont::font_import() or showtext::font_add() before passing its family name.

4. Apply theme_classic() to a bar chart

Bar charts often work better with theme_classic() than with theme_minimal() because the absence of gridlines lets each bar stand out without competing horizontal lines.

5. Add a subtle y-grid back for readability

6. Combine with theme() for fine-tuning

theme() always follows theme_classic(). Earlier theme calls on the same plot are overwritten by later ones, so the order in the chain decides who wins.

theme_classic() vs other ggplot2 themes

Pick the right base theme before customizing anything. The seven built-in themes each cover a different reader expectation, and choosing the wrong one means rewriting overrides you did not need.

If you need a full panel border, switch to theme_bw(). If readers must scan many data points for value lookup, switch to theme_minimal() or restore the grid as shown in pattern 5.

Set theme_classic() as the session default

Call theme_set(theme_classic()) once and every later ggplot call inherits the theme automatically. This avoids repeating + theme_classic() across dozens of plots in a manuscript script.

Reset with theme_set(theme_grey()) to return to ggplot2's default. Keep the global call at the top of your analysis script so every figure in the paper or report inherits the same look.

Common pitfalls

Pitfall 1: Calling theme() before theme_classic(). A chain like p + theme(text = element_text(size = 14)) + theme_classic() discards your size change because theme_classic() comes last and reapplies its own defaults. Reverse the order so theme() comes after.

Pitfall 2: Expecting gridlines to remain. Unlike theme_minimal() and theme_bw(), theme_classic() removes BOTH major and minor gridlines. If the reader needs a grid, add it back selectively with theme(panel.grid.major.y = element_line()) as shown in pattern 5.

Pitfall 3: Facet strip backgrounds look bare. theme_classic() strips the strip background fill, so facet_wrap() labels float on white. Restore them with theme(strip.background = element_rect(fill = "grey90", color = NA)) if your audience expects boxed strip headers.

Pitfall 4: Custom font silently falls back. If base_family = "Times New Roman" still shows the default sans font, R cannot find the font on your system. Register it first with extrafont or showtext, or use the safe values "serif", "sans", or "mono".

Try it yourself

Related ggplot2 themes and helpers

After mastering theme_classic(), explore its closest companions:

• theme_bw(), theme_minimal(), theme_light(), theme_void(): alternative built-in themes
• theme(): element-by-element customization for titles, legends, and axis lines
• element_text(), element_line(), element_rect(), element_blank(): element constructors
• theme_set(): global default theme for every plot in the session
• ggthemes package: bundles like theme_economist, theme_few, theme_tufte
• cowplot::theme_cowplot(): publication-ready theme modeled on theme_classic()

See the ggplot2 complete themes documentation.

FAQ

What is theme_classic() in ggplot2?

theme_classic() is a built-in ggplot2 theme that gives plots a white background, no gridlines, and L-shape axis lines along the bottom and left edges. Axis ticks remain visible. It is one of seven complete themes in ggplot2 and the standard choice for scientific journals and figures that prefer ticks over background grids.

What is the difference between theme_classic() and theme_minimal()?

theme_classic() removes all gridlines and draws two axis lines with visible ticks, producing the scientific paper look. theme_minimal() keeps light grey gridlines and removes both the panel border and axis ticks for a cleaner report look. Use theme_classic() when readers read precise values from ticks, and theme_minimal() when readers scan trends across a grid.

How is theme_classic() different from theme_bw()?

Both use a white background, but theme_bw() keeps gridlines and a full rectangular panel border, while theme_classic() strips the grid and the border, leaving only the bottom and left axis lines. Pick theme_classic() for journal submissions and theme_bw() for engineering reports that depend on the grid for value reads.

How do I add gridlines back to theme_classic()?

Append theme(panel.grid.major = element_line(color = "grey90")) after theme_classic() to restore both major grids. For a single direction, target panel.grid.major.x or panel.grid.major.y. Use a light colour like "grey90" so the grid stays subtle and does not compete with the data.

How do I set theme_classic() as the default for every plot?

Call theme_set(theme_classic()) at the top of your script. Every later ggplot() call inherits the theme automatically. Reset with theme_set(theme_grey()) to restore the ggplot2 default. This is the standard way to enforce a consistent look across all figures in a paper or report.