r‑statistics.co by Selva Prabhakaran

About

ggplot2 Customization Exercises: 10 Theme & Scale Practice Problems — Solved Step-by-Step

Themes control how a ggplot2 chart looks — fonts, backgrounds, grid lines, legend placement — and scales control how data maps to colours, axis limits, and labels. These 10 exercises build your customization skills from basic theme swaps to a full reusable house style, each with runnable starter code and a step-by-step solution.

What Do Themes and Scales Control?

Themes and scales are the two customization layers in ggplot2. A theme changes the non-data elements — everything around your data points, bars, or lines. A scale changes how data values translate into visual properties like position, colour, and size.

Here's how a single plot transforms when you add one theme and one scale function. Run the code to see the before-and-after.

library(ggplot2) library(scales) # Default ggplot2 styling p_base <- ggplot(mtcars, aes(x = wt, y = mpg, colour = factor(cyl))) + geom_point(size = 3) + labs( title = "Fuel Efficiency vs Weight", x = "Weight (1000 lbs)", y = "Miles per Gallon", colour = "Cylinders" ) p_base #> Scatter plot with grey background, default colours # Now: one theme + one scale = a polished chart p_styled <- p_base + theme_minimal(base_size = 13) + scale_colour_brewer(palette = "Set1") p_styled #> Same data, but clean background and vibrant categorical colours


  


  






The data didn't change — only the presentation did. theme_minimal() stripped the grey background and border, and scale_colour_brewer() replaced the default hue cycle with a ColorBrewer palette designed for readability.



Here's a quick reference for the functions you'll practice in these exercises.







Customization
Function
Controls






Complete theme swap
theme_minimal(), theme_classic(), etc.
All non-data elements at once




Individual element
theme(element = ...)
One specific visual property




Text styling
element_text(size, face, angle, colour)
Fonts, rotation, colour




Background/border
element_rect(fill, colour, linewidth)
Rectangles (panels, legend, strips)




Line styling
element_line(colour, linewidth, linetype)
Grid lines, axis lines




Remove element
element_blank()
Delete any theme element




Categorical colour
scale_colour_brewer(palette)
Discrete colour palette




Manual colour
scale_fill_manual(values)
Hand-picked colours




Axis formatting
scale_y_continuous(labels, limits, breaks)
Numeric axis display







Key Insight
Themes change the stage, scales change the translation. A theme is like redecorating the room your chart lives in. A scale is like changing the language your data speaks — a number becomes a dollar amount, a category becomes a specific colour.





Try it: Take p_base and apply theme_classic() with scale_colour_brewer(palette = "Set2"). How does the look change from the styled version above?



  

    
# Try it: swap theme and palette
ex_swap <- p_base +
  # your code here
  NULL

ex_swap
#> Expected: clean axis lines, no grid, Set2 pastel colours

    

      
      
    

    

  

  





Click to reveal solution



  

    
ex_swap <- p_base +
  theme_classic(base_size = 13) +
  scale_colour_brewer(palette = "Set2")
ex_swap
#> Scatter with axis lines only, no grid, pastel Set2 palette

    

      
      
    

    

  

  





Explanation: theme_classic() removes grid lines and the background box, leaving only axis lines — a common journal style. "Set2" uses softer pastel tones compared to the bolder "Set1".








How Do You Switch and Layer Themes? (Exercises 1–3)



These first three exercises focus on theme() and its element functions. You'll apply built-in themes, remove elements with element_blank(), and style text with element_text(). Each exercise builds on a familiar built-in dataset.



Exercise 1: Apply a Built-In Theme



Dataset: mpg



Task: Create a boxplot of hwy (y-axis) grouped by class (x-axis). Apply theme_light(). Add the title "Highway MPG by Vehicle Class" and label the axes "Vehicle Class" and "Highway MPG".





  


    
# Exercise 1: boxplot with theme_light
p1 <- ggplot(mpg, aes(x = class, y = hwy)) +
  # your code here
  NULL

p1
#> Expected: boxplot with 7 classes, light theme, titled axes


    


      

      

    


    


  


  







Click to reveal solution



  

    
p1 <- ggplot(mpg, aes(x = class, y = hwy)) +
  geom_boxplot(fill = "#7FB3D8") +
  theme_light() +
  labs(
    title = "Highway MPG by Vehicle Class",
    x = "Vehicle Class",
    y = "Highway MPG"
  )
p1
#> Boxplot: compact and subcompact have highest medians (~28-29),
#> pickups and SUVs lowest (~17). Light grey grid, white background.

    

      
      
    

    

  

  





Explanation: theme_light() gives a clean white background with subtle grey grid lines — less stark than theme_bw(), more structured than theme_minimal(). The fill argument colours the boxes, making group boundaries easier to see.






Exercise 2: Remove Grid Lines and Style Line Elements



Dataset: mtcars



Task: Create a scatter plot of wt (x) vs mpg (y). Start with theme_bw(), then add theme() to:



    


  1. Remove minor grid lines entirely
    2. 


  2. Change major grid lines to dashed, light grey ("grey85")
    3. 


  3. Make the panel border thicker (linewidth 1.5)
    4. 







  


    
# Exercise 2: customize grid lines
p2 <- ggplot(mtcars, aes(x = wt, y = mpg)) +
  geom_point(size = 3, colour = "#4B6FA5") +
  theme_bw() +
  theme(
    # your code here
  ) +
  labs(title = "Weight vs Fuel Efficiency", x = "Weight (1000 lbs)", y = "MPG")

p2
#> Expected: scatter with no minor grid, dashed major grid, thick border


    


      

      

    


    


  


  







Click to reveal solution



  

    
p2 <- ggplot(mtcars, aes(x = wt, y = mpg)) +
  geom_point(size = 3, colour = "#4B6FA5") +
  theme_bw() +
  theme(
    panel.grid.minor = element_blank(),
    panel.grid.major = element_line(colour = "grey85", linetype = "dashed"),
    panel.border = element_rect(colour = "black", linewidth = 1.5)
  ) +
  labs(title = "Weight vs Fuel Efficiency", x = "Weight (1000 lbs)", y = "MPG")
p2
#> Scatter plot: no minor grid, dashed light grey major grid,
#> thick black border. Heavier cars cluster at lower MPG.

    

      
      
    

    

  

  





Explanation: element_blank() removes an element completely — no line, no space. element_line() controls line properties: linetype = "dashed" draws dashes instead of solid lines, and colour = "grey85" makes them subtle. element_rect() styles rectangles like the panel border.






Exercise 3: Customize Axis and Title Text



Dataset: diamonds (random sample of 500 rows)



Task: Create a bar chart counting diamonds by cut. Then customize the text:



    


  1. Rotate x-axis labels 45 degrees (with hjust = 1 for alignment)
    2. 


  2. Set the plot title to 16pt bold
    3. 


  3. Make axis titles 12pt and bold
    4. 


  4. Apply theme_minimal() as the base
    5. 







  


    
# Exercise 3: text customization
set.seed(42)
d_small <- diamonds[sample(nrow(diamonds), 500), ]

p3 <- ggplot(d_small, aes(x = cut)) +
  geom_bar(fill = "#A5C882") +
  theme_minimal() +
  theme(
    # your code here
  ) +
  labs(title = "Diamond Quality Distribution", x = "Cut Quality", y = "Count")

p3
#> Expected: bar chart with angled labels, bold title at 16pt, bold axis titles


    


      

      

    


    


  


  







Click to reveal solution



  

    
set.seed(42)
d_small <- diamonds[sample(nrow(diamonds), 500), ]

p3 <- ggplot(d_small, aes(x = cut)) +
  geom_bar(fill = "#A5C882") +
  theme_minimal() +
  theme(
    axis.text.x = element_text(angle = 45, hjust = 1),
    plot.title = element_text(size = 16, face = "bold"),
    axis.title = element_text(size = 12, face = "bold")
  ) +
  labs(title = "Diamond Quality Distribution", x = "Cut Quality", y = "Count")
p3
#> Bar chart: Ideal has the most diamonds, Fair the fewest.
#> X-axis labels angled at 45°, large bold title, bold axis titles.

    

      
      
    

    

  

  





Explanation: axis.text.x controls only x-axis tick labels. angle = 45 rotates them, and hjust = 1 right-aligns so they don't overlap the ticks. plot.title and axis.title each accept element_text() — face = "bold" makes text bold, size sets the point size.






Tip
Always set hjust = 1 when angling x-axis labels. Without it, the rotated text anchors at the centre and floats away from the tick marks. With hjust = 1, the right edge of each label sits directly above its tick.



How Do You Control Colours with Scales? (Exercises 4–6)



Scales control the mapping between data values and visual properties. These exercises focus on colour and axis scales — the two you'll use most often when customizing real charts.



Exercise 4: Apply a ColorBrewer Palette



Dataset: iris



Task: Create a scatter plot of Sepal.Length (x) vs Petal.Length (y), coloured by Species. Apply scale_colour_brewer(palette = "Dark2"). Set the legend title to "Iris Species" and apply theme_minimal().





  


    
# Exercise 4: ColorBrewer palette
p4 <- ggplot(iris, aes(x = Sepal.Length, y = Petal.Length, colour = Species)) +
  geom_point(size = 2.5) +
  # your code here
  labs(x = "Sepal Length (cm)", y = "Petal Length (cm)")

p4
#> Expected: scatter with 3 species in Dark2 colours, clean theme


    


      

      

    


    


  


  







Click to reveal solution



  

    
p4 <- ggplot(iris, aes(x = Sepal.Length, y = Petal.Length, colour = Species)) +
  geom_point(size = 2.5) +
  scale_colour_brewer(palette = "Dark2") +
  theme_minimal() +
  labs(
    x = "Sepal Length (cm)",
    y = "Petal Length (cm)",
    colour = "Iris Species"
  )
p4
#> Scatter: setosa clusters bottom-left (short petals), virginica top-right.
#> Dark2 palette: teal, orange, purple — high contrast, colourblind-safe.

    

      
      
    

    

  

  





Explanation: scale_colour_brewer() picks a palette from the ColorBrewer project, designed by cartographers for visual clarity. "Dark2" is a qualitative palette — distinct, saturated colours for categorical data. The colour argument inside labs() renames the legend title.






Exercise 5: Assign Custom Colours with scale_fill_manual



Dataset: mtcars



Task: Compute mean mpg by cyl using aggregate(). Create a bar chart (geom_col) of mean MPG per cylinder group with fill = factor(cyl). Assign these exact colours:



    


  • 4 cylinders → "#2C7BB6" (blue)
    • 


  • 6 cylinders → "#ABD9E9" (light blue)
    • 


  • 8 cylinders → "#D7191C" (red)
    • 





Apply theme_minimal() and remove the legend (it's redundant with the x-axis).





  


    
# Exercise 5: custom fill colours
cyl_means <- aggregate(mpg ~ cyl, data = mtcars, FUN = mean)

p5 <- ggplot(cyl_means, aes(x = factor(cyl), y = mpg, fill = factor(cyl))) +
  geom_col() +
  # your code here
  labs(title = "Mean Fuel Efficiency by Cylinder Count",
       x = "Cylinders", y = "Mean MPG")

p5
#> Expected: 3 bars in blue, light blue, red — no legend


    


      

      

    


    


  


  







Click to reveal solution



  

    
cyl_means <- aggregate(mpg ~ cyl, data = mtcars, FUN = mean)

p5 <- ggplot(cyl_means, aes(x = factor(cyl), y = mpg, fill = factor(cyl))) +
  geom_col() +
  scale_fill_manual(values = c("4" = "#2C7BB6", "6" = "#ABD9E9", "8" = "#D7191C")) +
  theme_minimal() +
  theme(legend.position = "none") +
  labs(
    title = "Mean Fuel Efficiency by Cylinder Count",
    x = "Cylinders",
    y = "Mean MPG"
  )
p5
#> 4-cyl: ~26.7 mpg (blue), 6-cyl: ~19.7 (light blue), 8-cyl: ~15.1 (red).
#> No legend — colour + x-axis labels tell the story.

    

      
      
    

    

  

  





Explanation: scale_fill_manual() accepts a named vector — each name matches a level of the fill variable, each value is a colour. Using named values ensures the right colour always maps to the right group, even if the data order changes. legend.position = "none" hides the legend when it's redundant.






Exercise 6: Format Axis Labels and Set Limits



Dataset: diamonds (random sample of 1000 rows)



Task: Create a scatter plot of carat (x) vs price (y). Then:



    


  1. Format the y-axis as dollars with commas using scale_y_continuous(labels = scales::dollar)
    2. 


  2. Set the x-axis range from 0 to 3 using scale_x_continuous(limits = ...)
    3. 


  3. Apply theme_light()
    4. 







  


    
# Exercise 6: axis formatting
set.seed(99)
d_sample <- diamonds[sample(nrow(diamonds), 1000), ]

p6 <- ggplot(d_sample, aes(x = carat, y = price)) +
  geom_point(alpha = 0.4, colour = "#4B6FA5") +
  # your code here
  labs(title = "Diamond Price vs Carat", x = "Carat Weight", y = "Price")

p6
#> Expected: scatter with $ y-axis labels, x-axis 0-3, light theme


    


      

      

    


    


  


  







Click to reveal solution



  

    
set.seed(99)
d_sample <- diamonds[sample(nrow(diamonds), 1000), ]

p6 <- ggplot(d_sample, aes(x = carat, y = price)) +
  geom_point(alpha = 0.4, colour = "#4B6FA5") +
  scale_y_continuous(labels = scales::dollar) +
  scale_x_continuous(limits = c(0, 3)) +
  theme_light() +
  labs(title = "Diamond Price vs Carat", x = "Carat Weight", y = "Price")
p6
#> Scatter: exponential-looking relationship — price rises steeply above 1.5 carat.
#> Y-axis shows $0, $5,000, $10,000, $15,000. X-axis runs 0 to 3.

    

      
      
    

    

  

  





Explanation: scales::dollar is a label formatter from the scales package — it adds a $ prefix and comma separators. limits = c(0, 3) restricts the x-axis range and drops points outside that window. Any scales:: formatter (percent, comma, scientific) works the same way inside labels.






Warning
Setting axis limits with scale_x_continuous removes data outside the range. Points beyond the limits are silently dropped, which affects computed statistics like trend lines. If you want to zoom without removing data, use coord_cartesian(xlim = c(0, 3)) instead.



How Do You Customize Legends, Backgrounds, and Strips? (Exercises 7–8)



Legends, panel backgrounds, and facet strips are all theme elements controlled by element_rect() and element_text(). These two exercises practise the elements most people struggle with — positioning and styling the decorations around the data.



Exercise 7: Move and Style the Legend



Dataset: mpg



Task: Create a scatter plot of displ (x) vs hwy (y), coloured by drv (drive type). Then:



    


  1. Move the legend to the bottom of the plot
    2. 


  2. Give the legend a light grey background ("grey95") with a thin grey border
    3. 


  3. Set legend text to 11pt
    4. 


  4. Apply theme_minimal()
    5. 







  


    
# Exercise 7: legend customization
p7 <- ggplot(mpg, aes(x = displ, y = hwy, colour = drv)) +
  geom_point(size = 2.5) +
  theme_minimal() +
  theme(
    # your code here
  ) +
  labs(title = "Fuel Efficiency by Drive Type",
       x = "Engine Displacement (L)", y = "Highway MPG", colour = "Drive")

p7
#> Expected: legend at bottom, light grey box around it, 11pt text


    


      

      

    


    


  


  







Click to reveal solution



  

    
p7 <- ggplot(mpg, aes(x = displ, y = hwy, colour = drv)) +
  geom_point(size = 2.5) +
  theme_minimal() +
  theme(
    legend.position = "bottom",
    legend.background = element_rect(fill = "grey95", colour = "grey80", linewidth = 0.3),
    legend.text = element_text(size = 11)
  ) +
  labs(
    title = "Fuel Efficiency by Drive Type",
    x = "Engine Displacement (L)",
    y = "Highway MPG",
    colour = "Drive"
  )
p7
#> Scatter: front-wheel drive (f) clusters at lower displacement/higher mpg.
#> Legend sits below the plot in a subtle grey box.

    

      
      
    

    

  

  





Explanation: legend.position = "bottom" moves the legend below the plot area. legend.background takes element_rect() — fill sets the background colour, colour the border, and linewidth the border thickness. legend.text controls the category labels inside the legend.






Exercise 8: Style Facet Strip Panels



Dataset: mpg



Task: Create a scatter plot of displ (x) vs hwy (y) faceted by drv. Customize:



    


  1. Set strip background to navy ("#2C3E50")
    2. 


  2. Set strip text to white, bold, 11pt
    3. 


  3. Set panel background to white
    4. 


  4. Keep theme_minimal() as the base
    5. 







  


    
# Exercise 8: facet strip styling
p8 <- ggplot(mpg, aes(x = displ, y = hwy)) +
  geom_point(colour = "#4B6FA5", alpha = 0.6) +
  facet_wrap(~drv) +
  theme_minimal() +
  theme(
    # your code here
  ) +
  labs(title = "MPG by Drive Type", x = "Displacement (L)", y = "Highway MPG")

p8
#> Expected: 3 facet panels with navy strip headers, white bold text


    


      

      

    


    


  


  







Click to reveal solution



  

    
p8 <- ggplot(mpg, aes(x = displ, y = hwy)) +
  geom_point(colour = "#4B6FA5", alpha = 0.6) +
  facet_wrap(~drv) +
  theme_minimal() +
  theme(
    strip.background = element_rect(fill = "#2C3E50", colour = "#2C3E50"),
    strip.text = element_text(colour = "white", face = "bold", size = 11),
    panel.background = element_rect(fill = "white", colour = NA)
  ) +
  labs(title = "MPG by Drive Type", x = "Displacement (L)", y = "Highway MPG")
p8
#> Three panels (4, f, r) with dark navy headers and white bold text.
#> Front-wheel cars show highest highway mpg, rear-wheel the lowest.

    

      
      
    

    

  

  





Explanation: strip.background controls the coloured bar above each facet panel. strip.text styles the text inside that bar. Setting both fill and colour of the strip to the same navy prevents a contrasting border from showing. panel.background fills each panel's plotting area.






Tip
Set the strip border colour to match the fill. By default, element_rect() draws a thin black border. If your strip fill is dark, that border is invisible. But on lighter fills, a mismatched border looks messy. Setting colour = fill_colour gives a clean edge.



Can You Build a Reusable Custom Theme? (Exercises 9–10)



Real-world reporting means applying the same visual style to dozens of charts. Writing a custom theme function saves you from copy-pasting theme() calls across every plot. These final exercises combine everything above into a reusable workflow.



Exercise 9: Build a theme_report() Function



Task: Write a function called theme_report() that returns a ggplot2 theme with these settings:



    


  • Base: theme_minimal(base_size = 13)
    • 


  • Plot title: 16pt, bold
    • 


  • Plot subtitle: 12pt, italic, grey ("grey40")
    • 


  • Axis titles: 12pt, bold
    • 


  • Plot caption: 9pt, grey ("grey50")
    • 


  • Major grid: light grey ("grey90"), dashed
    • 


  • Minor grid: removed
    • 


  • Legend position: bottom
    • 





Test it by applying theme_report() to a scatter of mtcars wt vs mpg.





  


    
# Exercise 9: build a reusable theme
theme_report <- function() {
  # your code here
}

# Test it
p9 <- ggplot(mtcars, aes(x = wt, y = mpg)) +
  geom_point(size = 3, colour = "#4B6FA5") +
  theme_report() +
  labs(
    title = "Weight vs Fuel Efficiency",
    subtitle = "Data from the 1974 Motor Trend magazine",
    caption = "Source: mtcars dataset",
    x = "Weight (1000 lbs)", y = "Miles per Gallon"
  )
p9
#> Expected: clean minimal plot with bold title, italic grey subtitle, dashed grid


    


      

      

    


    


  


  







Click to reveal solution



  

    
theme_report <- function() {
  theme_minimal(base_size = 13) +
    theme(
      plot.title = element_text(size = 16, face = "bold"),
      plot.subtitle = element_text(size = 12, face = "italic", colour = "grey40"),
      axis.title = element_text(size = 12, face = "bold"),
      plot.caption = element_text(size = 9, colour = "grey50"),
      panel.grid.major = element_line(colour = "grey90", linetype = "dashed"),
      panel.grid.minor = element_blank(),
      legend.position = "bottom"
    )
}

p9 <- ggplot(mtcars, aes(x = wt, y = mpg)) +
  geom_point(size = 3, colour = "#4B6FA5") +
  theme_report() +
  labs(
    title = "Weight vs Fuel Efficiency",
    subtitle = "Data from the 1974 Motor Trend magazine",
    caption = "Source: mtcars dataset",
    x = "Weight (1000 lbs)", y = "Miles per Gallon"
  )
p9
#> Bold title, italic grey subtitle, dashed light grid, no minor grid,
#> bold axis titles, grey caption at bottom-right.

    

      
      
    

    

  

  





Explanation: A custom theme function wraps theme_minimal() and layers your theme() tweaks on top. Because it's a function, you call theme_report() on any plot — no copy-pasting. The + operator inside the function merges the base theme with your overrides.






Key Insight
A custom theme function is the single most time-saving ggplot2 trick for professionals. Once you define theme_report(), every chart in your project gets consistent styling with one line of code. Update the function once, and every plot updates automatically.



Exercise 10: Publication-Ready Plot — Theme + Scale Combined



Dataset: airquality (built-in, May through September temperatures)



Task: Build a complete, publication-ready line chart:



    


  1. Compute mean Temp per Month using aggregate()
    2. 


  2. Create a line chart with points: geom_line() + geom_point(), colour by factor(Month)
    3. 


  3. Apply theme_report() from Exercise 9
    4. 


  4. Use scale_colour_brewer(palette = "YlOrRd") (yellow-orange-red for temperature)
    5. 


  5. Label the x-axis with month names using scale_x_continuous(breaks = 5:9, labels = month.abb[5:9])
    6. 


  6. Add title "Average Temperature by Month", subtitle "New York, May–September 1973", and caption "Source: airquality dataset"
    7. 


  7. Remove the legend — colour is redundant with the x-axis labels
    8. 







  


    
# Exercise 10: publication-ready plot
aq_summary <- aggregate(Temp ~ Month, data = airquality, FUN = mean)

p10 <- ggplot(aq_summary, aes(x = Month, y = Temp, colour = factor(Month))) +
  # your code here
  NULL

p10
#> Expected: line chart, month names on x-axis, warm colour palette, no legend


    


      

      

    


    


  


  







Click to reveal solution



  

    
aq_summary <- aggregate(Temp ~ Month, data = airquality, FUN = mean)

p10 <- ggplot(aq_summary, aes(x = Month, y = Temp, colour = factor(Month))) +
  geom_line(linewidth = 1.2, colour = "grey60") +
  geom_point(size = 4) +
  scale_colour_brewer(palette = "YlOrRd") +
  scale_x_continuous(breaks = 5:9, labels = month.abb[5:9]) +
  theme_report() +
  theme(legend.position = "none") +
  labs(
    title = "Average Temperature by Month",
    subtitle = "New York, May\u2013September 1973",
    caption = "Source: airquality dataset",
    x = "Month",
    y = "Temperature (\u00B0F)"
  )
p10
#> Line chart rising from May (~65°F) to July (~84°F), dipping slightly in Sep.
#> Points coloured yellow (May) through red (Sep). Clean, print-ready layout.

    

      
      
    

    

  

  





Explanation: The line connects months in grey (a neutral connector), while the points carry the colour encoding. scale_colour_brewer("YlOrRd") is a sequential palette — light yellow for cooler months, deep red for the hottest. month.abb[5:9] converts numeric months to abbreviations. Removing the legend avoids redundancy since the x-axis already shows month names.






Complete Example



Here's a full pipeline combining themes, scales, and facets into one polished chart. This uses concepts from every exercise above.





  


    
# Complete example: faceted bar chart with full customization
dia_summary <- aggregate(price ~ cut + color, data = diamonds, FUN = mean)

p_complete <- ggplot(dia_summary, aes(x = color, y = price, fill = cut)) +
  geom_col(position = "dodge") +
  scale_fill_brewer(palette = "Blues") +
  scale_y_continuous(labels = scales::dollar) +
  facet_wrap(~cut, nrow = 1) +
  theme_report() +
  theme(
    strip.background = element_rect(fill = "#2C3E50", colour = "#2C3E50"),
    strip.text = element_text(colour = "white", face = "bold", size = 10),
    axis.text.x = element_text(angle = 45, hjust = 1, size = 9),
    legend.position = "none"
  ) +
  labs(
    title = "Average Diamond Price by Colour and Cut",
    subtitle = "Higher colour grades (D\u2013F) don\u2019t always mean higher prices",
    caption = "Source: ggplot2 diamonds dataset",
    x = "Colour Grade",
    y = "Mean Price"
  )
p_complete
#> Five facet panels (Fair through Ideal), each showing 7 colour-grade bars.
#> Fair-cut diamonds in colour J are the priciest (~$4,800) — because they're
#> larger stones. Ideal-cut diamonds are cheaper on average — more small stones.
#> Navy strip headers, dollar y-axis, rotated x-labels, no legend (redundant with facets).


    


      

      

    


    


  


  






This chart reveals a counterintuitive pattern: lower-quality cuts can have higher average prices because cut quality and carat weight are negatively correlated. The customization layers — theme, scales, facets, strip styling — work together to make this insight immediately visible.



Summary







Exercise
Concept
Key Functions






1
Apply built-in theme
theme_light(), labs()




2
Remove and restyle grid lines
element_blank(), element_line(linetype)




3
Customize text elements
element_text(angle, size, face, hjust)




4
ColorBrewer categorical palette
scale_colour_brewer(palette)




5
Custom colours with named vector
scale_fill_manual(values)




6
Axis formatting and limits
scale_y_continuous(labels), scale_x_continuous(limits)




7
Legend position and background
legend.position, legend.background




8
Facet strip styling
strip.background, strip.text




9
Reusable custom theme function
theme_minimal() + theme(...) in a function




10
Publication-ready combined plot
Theme + scale + labs + legend removal







The customization pattern in every exercise is the same: start with a complete theme, then layer theme() tweaks and scale_*() functions on top. Once that pattern clicks, you can style any ggplot2 chart.



References



    


  1. Wickham, H. — ggplot2: Elegant Graphics for Data Analysis, 3rd Edition. Chapter 17: Themes. Link
    2. 


  2. ggplot2 documentation — theme() reference. Link
    3. 


  3. ggplot2 documentation — Scales reference. Link
    4. 


  4. ggplot2 documentation — Complete themes. Link
    5. 


  5. ColorBrewer 2.0 — Colour palettes for cartography and data visualization. Link
    6. 


  6. Scherer, C. — A ggplot2 Tutorial for Beautiful Plotting in R (2019). Link
    7. 





Continue Learning



    


  1. ggplot2 Themes — Deep dive into the theme() system, all 8 built-in themes, the element inheritance tree, and building custom house styles from scratch.
    2. 


  2. ggplot2 Scales — Full reference for every scale function: position, colour, fill, size, alpha, plus breaks, labels, limits, and transformations.
    3. 


  3. Publication-Ready Figures — Export polished charts for papers and presentations with the right DPI, dimensions, and file format.
    4. 




Related Tutorials
ggplot2 Scales: Control Every Axis, Colour, and Size — The Full ReferenceVisualization · 10 min readggplot2's Grammar of Graphics: The Mental Model That Makes Everything ClickVisualization · 10 min readggplot2 Themes: From theme_classic to Your Own Custom House StyleVisualization · 10 min readggplot2 Log Scale in R: When & How to Transform Axes (with Examples)Visualization · 9 min read



        


        
      


      

        

        
© 2016- Selva Prabhakaran.
          This work is licensed under the Creative Commons License.