Mastering Scatterplots in ggplot2: The Complete Guide

Introduction

Scatterplots are the foundational tool for visualizing the relationship between two continuous variables. By observing the distribution of points, we can identify correlations, detect outliers, and uncover hidden clusters within our data.

1. Environment Setup

We use the tidyverse library for data manipulation and plotting.

Code

library(tidyverse)

2. Basic to Customized Scatterplots

The Core Geometry

A basic scatterplot maps two continuous variables. In this example, we examine the relationship between Sepal Length and Sepal Width from the iris dataset.

Code

iris %>% 
  ggplot(aes(x = Sepal.Length, y = Sepal.Width)) +
  geom_point() +
  labs(title = "Basic Scatterplot",
       x = "Sepal Length (cm)",
       y = "Sepal Width (cm)")

Aesthetic Customization

To make points more distinct, we can use specific shapes that support both color (outline) and fill (interior).

Code

iris %>% 
  ggplot(aes(x = Sepal.Length, y = Sepal.Width)) +
  geom_point(size = 3, 
             shape = 21, 
             color = "black", 
             fill = "salmon", 
             alpha = 0.7) + 
  labs(title = "Customized Scatterplot: Fill, Color, and Alpha")

3. Encoding Information (Color & Size)

Scatterplots can represent more than just two variables by mapping additional data to Color and Size.

Using color to represent a third categorical variable helps visualize how different groups cluster.

Code

iris %>% 
  ggplot(aes(x = Sepal.Length, y = Sepal.Width, color = Species)) +
  geom_point(size = 3) +
  scale_color_brewer(palette = "Set1") + 
  labs(title = "Clustering by Species")

By mapping size to a third numerical variable, you create a “Bubble Chart,” allowing you to see three dimensions of data at once.

Code

iris %>%
  ggplot(aes(x = Sepal.Length, y = Sepal.Width, size = Petal.Length)) +
  geom_point(alpha = 0.6) +
  labs(title = "Size Encodes Petal Length")

4. Statistical Trends & Faceting

Adding a Regression Line

To identify the mathematical trend, we add a linear regression line using geom_smooth().

Code

iris %>%
  ggplot(aes(x = Sepal.Length, y = Sepal.Width)) +
  geom_point(alpha = 0.5) +
  geom_smooth(method = "lm", color = "red", se = FALSE) + 
  labs(title = "Scatterplot with Linear Trend Line")

Faceting (Small Multiples)

When points overlap significantly (overplotting), splitting the graph into facets for each category is a systemic way to improve clarity.

Code

iris %>%
  ggplot(aes(x = Sepal.Length, y = Sepal.Width, color = Species)) +
  geom_point() +
  facet_wrap(~Species) + 
  theme(legend.position = "none") +
  labs(title = "Faceted by Species")

5. Advanced Publication-Ready Plot

Combining all systemic elements: manual color scales, centered titles, and statistical overlays.

Code

iris %>%
  ggplot(aes(x = Sepal.Length, y = Sepal.Width)) +
  geom_point(aes(color = Species, shape = Species), size = 3, alpha = 0.8) +
  geom_smooth(method = "lm", color = "black", linetype = "dashed", se = FALSE) +
  scale_color_manual(values = c("setosa" = "#A88EF2", 
                                "versicolor" = "#FBB222", 
                                "virginica" = "#4CAF50")) +
  labs(title = "Advanced Scatterplot Analysis",
       subtitle = "Sepal Dimensions with Linear Trends across Species",
       x = "Sepal Length (cm)",
       y = "Sepal Width (cm)",
       caption = "Dataset: Iris | Prepared by Abdullah Al Shamim") +
  theme(legend.position = "bottom",
        plot.title = element_text(face = "bold", size = 14))

Systemic Summary Checklist

Aesthetic Mapping	Usage	Variable Type
Color	Grouping or Intensity	Categorical or Continuous
Size	Weight or Magnitude	Continuous
Shape	Differentiation	Categorical
Alpha	Managing Overplotting	Numerical (0-1)