Data Visualisation

Tell stories with your data

Why Visualise?

• Understand patterns quickly
• Communicate findings effectively
• Identify outliers and anomalies
• Explore relationships

Scatter Plots

Show relationships between two continuous variables

Step 1: Create Basic Canvas

Start with the data and aesthetic mappings:

library(ggplot2)

# Create the basic ggplot object with axes
ggplot(mtcars, aes(x = wt, y = mpg))

Step 2: Add Points

Add points to visualise individual observations:

# Add points to show the relationship
ggplot(mtcars, aes(x = wt, y = mpg)) +
  geom_point()

Step 3: Customise Points

Adjust size, transparency, and colour:

# Make points larger and semi-transparent
ggplot(mtcars, aes(x = wt, y = mpg)) +
  geom_point(size = 3, alpha = 0.6, colour = "steelblue")

Step 4: Add Trend Line

Add a smoothing layer to show the pattern:

# Add linear regression line
ggplot(mtcars, aes(x = wt, y = mpg)) +
  geom_point(size = 3, alpha = 0.6, colour = "steelblue") +
  geom_smooth(method = "lm", se = TRUE, colour = "red")

Step 5: Add Labels and Theme

Make it publication-ready with titles and formatting:

# Build the complete scatter plot
scatter_plot <- ggplot(mtcars, aes(x = wt, y = mpg)) +
  geom_point(size = 3, alpha = 0.6, colour = "steelblue") +
  geom_smooth(method = "lm", se = TRUE, colour = "red") +
  labs(
    title = "Fuel Efficiency vs Weight",
    subtitle = "mtcars dataset",
    x = "Weight (1000 lbs)",
    y = "Miles per Gallon",
    caption = "Data: Henderson & Velleman (1981)"
  ) +
  theme_minimal() +
  theme(plot.title = element_text(size = 14, face = "bold"))

Final Scatter Plot

View the complete publication-ready plot:

Bar Plots

Show counts or comparisons across categories

Step 1: Basic Bar Plot

Create bars for each category:

# Create basic bar plot
ggplot(mtcars, aes(x = factor(cyl))) +
  geom_bar()

Step 2: Add Colour

Customise the appearance with colours:

# Add custom fill colour
ggplot(mtcars, aes(x = factor(cyl))) +
  geom_bar(fill = "steelblue", colour = "black", alpha = 0.7)

Step 3: Add Labels

Add titles and axis labels:

# Build the complete bar plot
bar_plot <- ggplot(mtcars, aes(x = factor(cyl))) +
  geom_bar(fill = "steelblue", colour = "black", alpha = 0.7) +
  labs(
    title = "Number of Cars by Cylinders",
    x = "Cylinders",
    y = "Count"
  ) +
  theme_minimal()

Final Bar Plot

View the complete publication-ready plot:

Box Plots

Show distributions and identify outliers

Step 1: Basic Box Plot

Create a box plot to show distribution:

# Create basic box plot
ggplot(mtcars, aes(x = factor(cyl), y = mpg)) +
  geom_boxplot()

Step 2: Customise Box Plot

Add colours and improve aesthetics:

# Customised box plot with colour
ggplot(mtcars, aes(x = factor(cyl), y = mpg, fill = factor(cyl))) +
  geom_boxplot(alpha = 0.7) +
  geom_jitter(width = 0.1, alpha = 0.3)  # Add individual points

Step 3: Add Labels

Make it clear and publication-ready:

# Complete box plot with labels
my_boxplot <- ggplot(mtcars, aes(x = factor(cyl), y = mpg, fill = factor(cyl))) +
  geom_boxplot(alpha = 0.7) +
  geom_jitter(width = 0.1, alpha = 0.3) +
  labs(
    title = "MPG Distribution by Number of Cylinders",
    x = "Cylinders",
    y = "Miles per Gallon",
    fill = "Cylinders"
  ) +
  theme_minimal() +
  theme(legend.position = "bottom")

Final Box Plot

View the complete publication-ready plot:

Using Plots

Plots can also be assigned to objects for further modification or saving:

# Save plot to object
final_plot <- my_boxplot + theme_classic()

final_plot

Exporting Plots

You can save your plots using ggsave():

# Save the final plot as a PNG file
ggsave("final_plot.png", plot = final_plot,  width = 15,  height = 10, units = "cm", dpi = 300)

Key Resources

• R for Data Science - Chapter 3: Data Visualisation
• ggplot2 Documentation
• R Graphics Cookbook
• Visualization for Social Data Science

Key Takeaways

✅ Start with a basic canvas and build incrementally

✅ Choose the right geom for your data type

✅ Customise colours, size, and transparency strategically

✅ Add informative titles and axis labels

✅ Use themes to improve presentation

✅ Layer multiple geoms for richer visualisations

Python Content (Optional)

If you’re interested in Python, here are equivalent concepts

Scatter Plot in Python

Step-by-step construction of a scatter plot:

import matplotlib.pyplot as plt
import seaborn as sns
import pandas as pd

# Read data
df = sns.load_dataset('mpg')

# Step 1: Create figure and axis
fig, ax = plt.subplots(figsize=(10, 6))

# Step 2: Create scatter plot
ax.scatter(df['weight'], df['mpg'], s=100, alpha=0.6, color='steelblue')

# Step 3: Add labels
ax.set_xlabel('Weight')
ax.set_ylabel('Miles per Gallon')
ax.set_title('Fuel Efficiency vs Weight')
plt.grid(alpha=0.3)
plt.show()

Bar Plot in Python

Create a bar plot showing counts by category:

import matplotlib.pyplot as plt
import seaborn as sns

df = sns.load_dataset('mpg')

# Create figure
fig, ax = plt.subplots(figsize=(10, 6))

# Create bar plot
df['cylinders'].value_counts().sort_index().plot(
    kind='bar', ax=ax, color='steelblue', alpha=0.7
)

# Add labels
ax.set_title('Number of Cars by Cylinders')
ax.set_xlabel('Cylinders')
ax.set_ylabel('Count')
plt.show()

Box Plot in Python

Show distribution and outliers with a box plot:

import matplotlib.pyplot as plt
import seaborn as sns

df = sns.load_dataset('mpg')

# Create figure
fig, ax = plt.subplots(figsize=(10, 6))

# Create box plot with individual points
sns.boxplot(data=df, x='cylinders', y='mpg', ax=ax, palette='Set2')
sns.stripplot(data=df, x='cylinders', y='mpg', ax=ax, 
              color='black', alpha=0.3, size=3)

# Add labels
ax.set_title('MPG Distribution by Number of Cylinders')
ax.set_xlabel('Cylinders')
ax.set_ylabel('Miles per Gallon')
plt.show()