import pandas as pd
import matplotlib.pyplot as plt
plt.rcParams['font.sans-serif'] = ['SimHei']
plt.rcParams['axes.unicode_minus'] = False

# 先装读取xlsx必需的库（只运行一次即可）
import subprocess
import sys
subprocess.check_call([sys.executable, "-m", "pip", "install", "openpyxl"])

# 读取文件：强制用 openpyxl 引擎，文件名完全按你左边显示的写
age_abs = pd.read_excel('healthcare-dataset-age_abs (1).xlsx', engine='openpyxl')
stroke  = pd.read_excel('healthcare-dataset-stroke.xlsx', engine='openpyxl')

# 合并
data = pd.merge(age_abs, stroke, on='编号')
data = data[data['中风'] == '是'].reset_index(drop=True)

# 相关系数
corr_coef = data['平均血糖'].corr(data['体重指数'])
print(f"血糖-BMI相关系数：{corr_coef:.3f}")

# 散点图
plt.scatter(data['平均血糖'], data['体重指数'], s=8)
plt.xlabel("平均血糖")
plt.ylabel("体重指数")
plt.title("平均血糖与体重指数关系")
plt.show()