""" Day 104 - 数据可视化 03 - 实战案例：完整 EDA 报告（Titanic 数据探索性分析） 运行方式：python3 03-eda-report.py 输出文件：titanic_eda_report.png """ import matplotlib.pyplot as plt import seaborn as sns import pandas as pd import numpy as np # ============================================================ # 中文字体配置 # ============================================================ plt.rcParams['font.sans-serif'] = ['SimHei', 'DejaVu Sans'] plt.rcParams['axes.unicode_minus'] = False sns.set_theme(style="whitegrid") # ============================================================ # 1. 加载并预处理数据 # ============================================================ print("🚢 Loading Titanic dataset...") titanic = sns.load_dataset('titanic') # 用中位数填充年龄缺失值 median_age = titanic['age'].median() titanic['age'] = titanic['age'].fillna(median_age) # 创建年龄段 titanic['age_group'] = pd.cut(titanic['age'], bins=[0, 12, 18, 35, 55, 80], labels=['Child', 'Teen', 'Young', 'Middle', 'Senior']) # 创建家庭规模 titanic['family_size'] = titanic['sibsp'] + titanic['parch'] + 1 print(f"✅ 数据集形状: {titanic.shape}") print(f"✅ 缺失值处理: age 列用中位数 {median_age} 填充") # ============================================================ # 2. 生成 6 子图 EDA 报告 # ============================================================ print("\n📊 Generating EDA report...") fig = plt.figure(figsize=(18, 14)) fig.suptitle('🚢 Titanic Survival Analysis - EDA Report', fontsize=18, fontweight='bold', y=0.98) # --- 图1: 总体生存率 --- ax1 = fig.add_subplot(3, 3, 1) survival_counts = titanic['survived'].value_counts() colors = ['#e74c3c', '#2ecc71'] wedges, texts, autotexts = ax1.pie( survival_counts, labels=['Died', 'Survived'], autopct='%1.1f%%', colors=colors, startangle=90, explode=(0, 0.05), textprops={'fontsize': 11} ) ax1.set_title('Overall Survival Rate', fontsize=13, fontweight='bold') # --- 图2: 性别 vs 生存率 --- ax2 = fig.add_subplot(3, 3, 2) sex_survival = titanic.groupby('sex')['survived'].mean() * 100 bars = ax2.bar(sex_survival.index, sex_survival.values, color=['#3498db', '#e91e63'], edgecolor='black') for bar, val in zip(bars, sex_survival.values): ax2.text(bar.get_x() + bar.get_width()/2, bar.get_height() + 1, f'{val:.1f}%', ha='center', fontweight='bold', fontsize=12) ax2.set_title('Survival Rate by Gender', fontsize=13, fontweight='bold') ax2.set_ylabel('Survival Rate (%)') ax2.set_ylim(0, 100) # --- 图3: 船舱等级 vs 生存率 --- ax3 = fig.add_subplot(3, 3, 3) class_survival = titanic.groupby('class')['survived'].mean() * 100 colors_class = ['#f39c12', '#3498db', '#e74c3c'] bars = ax3.bar(class_survival.index, class_survival.values, color=colors_class, edgecolor='black') for bar, val in zip(bars, class_survival.values): ax3.text(bar.get_x() + bar.get_width()/2, bar.get_height() + 1, f'{val:.1f}%', ha='center', fontweight='bold', fontsize=12) ax3.set_title('Survival Rate by Class', fontsize=13, fontweight='bold') ax3.set_ylabel('Survival Rate (%)') ax3.set_ylim(0, 100) # --- 图4: 年龄分布（分组） --- ax4 = fig.add_subplot(3, 3, 4) age_survival = titanic.groupby(['age_group', 'survived']).size().unstack(fill_value=0) age_survival.plot(kind='bar', stacked=True, ax=ax4, color=colors, edgecolor='black') ax4.set_title('Survival by Age Group', fontsize=13, fontweight='bold') ax4.set_xlabel('Age Group') ax4.set_ylabel('Count') ax4.legend(['Died', 'Survived'], loc='upper left') ax4.tick_params(axis='x', rotation=0) # --- 图5: 票价分布 vs 生存 --- ax5 = fig.add_subplot(3, 3, 5) sns.histplot(data=titanic, x='fare', hue='survived', kde=True, palette={0: '#e74c3c', 1: '#2ecc71'}, ax=ax5, bins=30, alpha=0.6) ax5.set_title('Fare Distribution by Survival', fontsize=13, fontweight='bold') ax5.set_xlabel('Fare ($)') ax5.set_xlim(0, 300) # --- 图6: 相关性热力图 --- ax6 = fig.add_subplot(3, 3, 6) features = ['survived', 'pclass', 'age', 'sibsp', 'parch', 'fare', 'family_size'] corr_matrix = titanic[features].corr() mask = np.triu(np.ones_like(corr_matrix, dtype=bool)) sns.heatmap(corr_matrix, mask=mask, annot=True, cmap='RdYlBu_r', center=0, fmt='.2f', linewidths=0.5, ax=ax6, cbar_kws={'shrink': 0.8}) ax6.set_title('Feature Correlation', fontsize=13, fontweight='bold') # --- 图7: 登船港口 --- ax7 = fig.add_subplot(3, 3, 7) embarked_survival = titanic.groupby('embarked')['survived'].mean() * 100 bars = ax7.bar(embarked_survival.index, embarked_survival.values, color=['#9b59b6', '#1abc9c', '#e67e22'], edgecolor='black') for bar, val in zip(bars, embarked_survival.values): ax7.text(bar.get_x() + bar.get_width()/2, bar.get_height() + 1, f'{val:.1f}%', ha='center', fontweight='bold') ax7.set_title('Survival by Embarkation', fontsize=13, fontweight='bold') ax7.set_ylabel('Survival Rate (%)') # --- 图8: 家庭规模 --- ax8 = fig.add_subplot(3, 3, 8) family_survival = titanic.groupby('family_size')['survived'].mean() * 100 ax8.plot(family_survival.index, family_survival.values, marker='o', linewidth=2, markersize=8, color='#2c3e50') ax8.fill_between(family_survival.index, family_survival.values, alpha=0.2, color='#3498db') ax8.set_title('Survival by Family Size', fontsize=13, fontweight='bold') ax8.set_xlabel('Family Size') ax8.set_ylabel('Survival Rate (%)') ax8.set_xticks(range(1, titanic['family_size'].max() + 1)) ax8.grid(True, alpha=0.3) # --- 图9: 关键发现文本 --- ax9 = fig.add_subplot(3, 3, 9) ax9.axis('off') findings = """ 🔑 KEY FINDINGS 1. Gender: Women survived at 74.2% vs Men at 18.9% 2. Class: 1st class had 63% survival vs 3rd class 24% 3. Age: Children (0-12) had higher survival rates 4. Fare: Higher fare = higher survival probability 5. Family: Size 2-4 had best survival, solo had worst """ ax9.text(0.05, 0.95, findings, transform=ax9.transAxes, fontsize=11, verticalalignment='top', fontfamily='monospace', bbox=dict(boxstyle='round,pad=0.8', facecolor='lightyellow', edgecolor='orange', alpha=0.9)) plt.tight_layout(rect=[0, 0, 1, 0.96]) plt.savefig('titanic_eda_report.png', dpi=150, bbox_inches='tight') print("✅ titanic_eda_report.png saved") # ============================================================ # 3. 输出分析摘要 # ============================================================ print("\n" + "=" * 50) print("📋 ANALYSIS SUMMARY") print("=" * 50) print(f"Total passengers: {len(titanic)}") print(f"Overall survival rate: {titanic['survived'].mean()*100:.1f}%") print(f"\nBy Gender:") print(titanic.groupby('sex')['survived'].mean().map(lambda x: f"{x*100:.1f}%")) print(f"\nBy Class:") print(titanic.groupby('class')['survived'].mean().map(lambda x: f"{x*100:.1f}%")) print(f"\nBy Age Group:") print(titanic.groupby('age_group')['survived'].mean().map(lambda x: f"{x*100:.1f}%")) print("\n🎉 EDA Report Complete!")