Python Fast Track Coding Course for Students in Kaohsiung

1. Two Families of Categorical Plots

Seaborn splits categorical charts into aggregate summaries and distribution/spread views.

• Aggregate: sns.countplot (frequencies), sns.barplot (means + variation)
• Distribution: sns.boxplot, sns.violinplot, sns.stripplot, sns.swarmplot

2. Sample for Clarity

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

df = pd.read_csv("healthcare_patient_analytics_seaborn.csv")
sample_df = df.sample(n=500, random_state=42)

3. Count and Bar Plots

plt.figure(figsize=(8, 5))
sns.countplot(data=sample_df, x="Treatment Type")
plt.title("Patient Count by Treatment Type")
plt.show()

plt.figure(figsize=(9, 5))
sns.barplot(data=sample_df, x="Department", y="Treatment Cost")
plt.xticks(rotation=30)
plt.title("Average Treatment Cost by Department")
plt.show()

In the sample, Surgery was most frequent; Cardiology had the highest average cost.

4. Box and Violin Plots

plt.figure(figsize=(9, 5))
sns.boxplot(data=sample_df, x="Department", y="Recovery Score")
plt.xticks(rotation=30)
plt.title("Recovery Score Distribution Across Departments")
plt.show()

plt.figure(figsize=(9, 5))
sns.violinplot(data=sample_df, x="Department", y="Recovery Score")
plt.xticks(rotation=30)
plt.title("Recovery Score Kernel Density by Department")
plt.show()

Box: Q1–Q3 box, median line, whiskers, outlier points. Violin: same summary idea plus KDE silhouette width for density.

5. Strip Plot and Overlay

plt.figure(figsize=(10, 6))
sns.boxplot(data=sample_df, x="Department", y="Recovery Score")
sns.stripplot(
    data=sample_df,
    x="Department",
    y="Recovery Score",
    color="black",
    alpha=0.4,
    jitter=True
)
plt.xticks(rotation=30)
plt.title("Box Plot Overlaid with Strip Plot")
plt.show()

Jittered points show individual cases against quartile boundaries and whisker outliers.