Q&A 8 How do you compare group distributions using a boxplot?

8.1 Explanation

A boxplot is a standard way to visualize the distribution of a numerical variable across categories. It summarizes key statistics:

• Median (central line)
• Interquartile range (IQR) (box edges)
• Minimum and maximum (whiskers)
• Outliers (points outside whiskers)

Boxplots are ideal for detecting:

• Differences in central tendency
• Variation in spread
• Presence of outliers
• Asymmetry or skewness in the distribution

Adding color and overlaying raw data (e.g., strip plots) improves interpretability.

---

8.2 Python Code

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

# Load dataset
iris = pd.read_csv("data/iris.csv")

# Set style
sns.set(style="whitegrid")

# Basic boxplot
plt.figure(figsize=(8, 6))
sns.boxplot(data=iris, x="species", y="sepal_length", palette="viridis")
plt.title("Boxplot: Sepal Length by Species", fontsize=14)
plt.xlabel("Species")
plt.ylabel("Sepal Length")
plt.tight_layout()
plt.show()

# Boxplot with overlaid strip plot (raw points)
plt.figure(figsize=(8, 6))
sns.boxplot(data=iris, x="species", y="sepal_length", palette="viridis", width=0.6)
sns.stripplot(data=iris, x="species", y="sepal_length", color="black", alpha=0.5, jitter=True)
plt.title("Boxplot with Raw Points: Sepal Length by Species", fontsize=14)
plt.tight_layout()
plt.show()

/var/folders/m1/0dxpqygn2ds41kxkjgwtftr00000gn/T/ipykernel_75563/1537151029.py:13: FutureWarning: 

Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.

  sns.boxplot(data=iris, x="species", y="sepal_length", palette="viridis")

/var/folders/m1/0dxpqygn2ds41kxkjgwtftr00000gn/T/ipykernel_75563/1537151029.py:22: FutureWarning: 

Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.

  sns.boxplot(data=iris, x="species", y="sepal_length", palette="viridis", width=0.6)

8.3 R Code

library(readr)
library(ggplot2)

# Load dataset
iris <- read_csv("data/iris.csv")

# Basic boxplot
ggplot(iris, aes(x = species, y = sepal_length)) +
  geom_boxplot(fill = "lightblue", color = "black") +
  theme_minimal() +
  labs(title = "Boxplot: Sepal Length by Species", x = "Species", y = "Sepal Length")

# Boxplot with overlaid jittered points
ggplot(iris, aes(x = species, y = sepal_length)) +
  geom_boxplot(fill = "lightgreen", outlier.shape = NA) +
  geom_jitter(color = "black", width = 0.2, alpha = 0.5) +
  theme_minimal() +
  labs(title = "Boxplot with Raw Points: Sepal Length by Species", x = "Species", y = "Sepal Length")

---

✅ Boxplots offer a compact summary of distribution and spread for each category. When enhanced with color and raw points, they reveal both statistical structure and individual variation clearly.