app.py

import streamlit as st
import requests
from bs4 import BeautifulSoup
import pandas as pd
import plotly.express as px
import plotly.graph_objects as go

# 定義爬取數據的函數
def fetch_data(hospital_url, table_id, hospital_name):
    response = requests.get(hospital_url)
    soup = BeautifulSoup(response.text, 'html.parser')
    table = soup.find('table', {'id': table_id})
    if not table:
        return pd.DataFrame()  # 若表格未找到，返回空的DataFrame
    
    rows = table.find_all('tr')
    data = []
    
    if hospital_name == "成大醫院":
        # 成大醫院的表格結構不同
        columns = [th.text.strip() for th in rows[0].find_all("th")]
        # 將 "病床類別" 改名為 "病床種類"
        columns = ['病床種類' if col == '病床類別' else col for col in columns]
        data_rows = rows[1:]
        for row in data_rows:
            row_data = [td.text.strip() for td in row.find_all("td")]
            if all(row_data):  # 確保數據不為空
                data.append(row_data)
    else:
        # 台南醫院和奇美醫院的表格結構類似
        columns = ['病床種類', '病床數', '住院人數', '空床數', '佔床率']
        for row in rows[1:]:  # 跳過標題行
            cols = row.find_all('td')
            if len(cols) == 5:
                row_data = [col.get_text(strip=True) for col in cols]
                if all(row_data):  # 確保數據不為空
                    data.append(row_data)
    
    df = pd.DataFrame(data, columns=columns)
    
    # 確保所有醫院的數據都有 '病床數' 列
    if '病床數' not in df.columns and '開放床數' in df.columns:
        df = df.rename(columns={'開放床數': '病床數'})
    
    return df

# Streamlit UI
st.title("醫院床位分配表爬取工具")

# 下拉式選單選擇醫院
hospital_options = {
    "台南醫院": {
        "url": "https://www.tmh.org.tw/tmh2016/ImpBD.aspx?Kind=2",
        "table_id": "ctl00_ContentPlaceHolder1_GV_Bed"
    },
    "奇美醫院": {
        "url": "https://www.chimei.org.tw/%E4%BD%94%E5%BA%8A%E7%8E%87%E6%9F%A5%E8%A9%A2/%E4%BD%94%E5%BA%8A%E7%8E%87%E6%9F%A5%E8%A9%A2.aspx?ihospital=10&ffloor=",
        "table_id": "DG1"
    },
    "成大醫院": {
        "url": "https://web.hosp.ncku.edu.tw/nckm/Bedstatus/BedStatus.aspx",
        "table_id": "GV_EmgInsure"
    }
}

selected_hospitals = st.multiselect("選擇醫院", list(hospital_options.keys()))

# 當用戶按下按鈕時，開始爬取數據
if st.button("爬取資料"):
    st.write("正在爬取資料...")
    progress_bar = st.progress(0)
    all_data = pd.DataFrame()

    for i, hospital_name in enumerate(selected_hospitals):
        hospital_data = hospital_options[hospital_name]
        df = fetch_data(hospital_data["url"], hospital_data["table_id"], hospital_name)
        if df.empty:
            st.warning(f"{hospital_name} 的數據爬取結果為空，請檢查是否存在問題。")
        else:
            df['醫院'] = hospital_name
            all_data = pd.concat([all_data, df], ignore_index=True)
        
        # 更新進度條
        progress_bar.progress((i + 1) / len(selected_hospitals))

    if not all_data.empty:
        st.write("爬取完成，合併的數據如下：")
        st.dataframe(all_data)

        # 下載數據為 CSV
        csv = all_data.to_csv(index=False).encode('utf-8-sig')
        st.download_button(
            label="下載數據為 CSV",
            data=csv,
            file_name='合併的床位分配資料.csv',
            mime='text/csv'
        )

        # 自動繪製圖表
        st.write("正在繪製圖表...")
        
        # 準備數據
        bed_column = '病床數'
        all_data[bed_column] = pd.to_numeric(all_data[bed_column], errors='coerce')
        bed_counts = all_data.groupby(['醫院', '病床種類'])[bed_column].sum().reset_index()
        
        # 繪製圓餅圖
        fig_pie = px.pie(bed_counts, values=bed_column, names='病床種類', title='各類型病床分佈',
                         hover_data=['醫院'], labels={bed_column:'病床數'})
        st.plotly_chart(fig_pie)

        # 繪製柱狀圖
        fig_bar = px.bar(bed_counts, x='醫院', y=bed_column, color='病床種類', title='醫院病床分佈',
                         labels={bed_column:'病床數'}, barmode='group')
        st.plotly_chart(fig_bar)

    else:
        st.error("沒有成功爬取任何數據。")