Python 房价空间分析完结篇：区域对比、时间序列与结论

REGIONS = {
    "京津冀": ["北京", "天津", "石家庄", "唐山", "秦皇岛"],
    "长三角": ["上海", "南京", "无锡", "杭州", "宁波", "合肥", "温州", "金华", "徐州", "扬州", "蚌埠", "安庆"],
    "珠三角": ["广州", "深圳", "惠州", "韶关", "湛江"],
    "海西+海南": ["福州", "厦门", "泉州", "海口", "三亚"],
    "中部": ["武汉", "长沙", "郑州", "南昌", "太原", "洛阳", "平顶山", "宜昌", "襄阳", "岳阳", "常德", "九江", "赣州", "济宁", "烟台", "济南", "青岛"],
    "西部": ["重庆", "成都", "西安", "贵阳", "昆明", "兰州", "西宁", "银川", "乌鲁木齐", "南宁", "桂林", "北海", "泸州", "南充", "遵义", "大理", "呼和浩特", "包头"],
    "东北": ["沈阳", "大连", "长春", "哈尔滨", "丹东", "锦州", "吉林", "牡丹江"],
}

def assign_region(city):
    for region, cities in REGIONS.items():
        if city in cities:
            return region
    return "其他"

gdf['region'] = gdf['CITY'].map(assign_region)

import matplotlib.pyplot as plt

fig, axes = plt.subplots(1, 2, figsize=(18, 7))

# 左图：区域平均累计涨幅
reg_cum = gdf.groupby('region')['cum_return'].agg(['mean', 'std']).sort_values('mean', ascending=True)
colors_cum = ['#4e79a7' if r != '海西+海南' else '#e15759' for r in reg_cum.index]

axes[0].barh(reg_cum.index, reg_cum['mean'], xerr=reg_cum['std'],
             color=colors_cum, alpha=0.85, edgecolor='white', capsize=4)
for i, (name, row) in enumerate(reg_cum.iterrows()):
    axes[0].text(row['mean'] + 1, i, f\"{row['mean']:.1f}%\", va='center', fontsize=9, fontweight='bold')
axes[0].set_xlabel('平均累计涨幅（%）')
axes[0].set_title('区域平均累计涨幅', fontweight='bold')
axes[0].grid(True, axis='x', alpha=0.2)

# 右图：区域平均波动率
reg_vol = gdf.groupby('region')['volatility'].agg(['mean', 'std']).sort_values('mean', ascending=True)
reg_vol['mean_pct'] = reg_vol['mean'] * 100
reg_vol['std_pct'] = reg_vol['std'] * 100

axes[1].barh(reg_vol.index, reg_vol['mean_pct'], xerr=reg_vol['std_pct'],
             color=colors_cum, alpha=0.85, edgecolor='white', capsize=4)
for i, (name, row) in enumerate(reg_vol.iterrows()):
    axes[1].text(row['mean_pct'] + 0.5, i, f\"{row['mean_pct']:.1f}%\", va='center', fontsize=9, fontweight='bold')
axes[1].set_xlabel('平均波动率（%）')
axes[1].set_title('区域平均波动率', fontweight='bold')
axes[1].grid(True, axis='x', alpha=0.2)

plt.suptitle('区域对比：涨幅 vs 波动率', fontsize=15, fontweight='bold', y=1.02)
plt.tight_layout()
plt.savefig('regional_comparison.png', dpi=150, bbox_inches='tight')

cities = ['北京', '上海', '深圳', '广州', '成都', '武汉', '温州', '三亚', '大连', '贵阳']

fig, ax = plt.subplots(figsize=(14, 8))
for city in cities:
    c = df_mom[df_mom['CITY'] == city].sort_values('DATE')
    cum_idx = (c['index_value'] / 100).cumprod() * 100
    ax.plot(range(len(cum_idx)), cum_idx, label=city, linewidth=1.5, alpha=0.8)

ax.axhline(100, color='gray', linestyle='--', linewidth=0.8, alpha=0.5)

# 标注关键政策节点
events = {24: '2008 金融危机', 96: '2014 限购放松', 120: '2016 去库存', 180: '2021 调控收紧'}
for month, label in events.items():
    ax.axvline(month, color='red', linestyle=':', linewidth=0.8, alpha=0.4)
    ax.text(month + 1, ax.get_ylim()[1] * 0.95, label, fontsize=8, color='red', alpha=0.7)

ax.set_xlabel('月份（2006.01 起）')
ax.set_ylabel('累计指数（基期=100）')
ax.set_title('10 城房价指数累计走势（2006-2026）', fontweight='bold')
ax.legend(loc='upper left', fontsize=9)
ax.grid(True, alpha=0.2)
plt.savefig('time_series.png', dpi=150, bbox_inches='tight')