Update app.py

app.py

@@ -8,17 +8,6 @@ from plotly.subplots import make_subplots
 from google.oauth2.service_account import Credentials
 import gspread
 
-# Google Sheets credentials
-SCOPE = ['https://www.googleapis.com/auth/spreadsheets']
-SERVICE_ACCOUNT_FILE = "realtime-441511-f5708eabdf26.json"
-SPREADSHEET_URL = "https://docs.google.com/spreadsheets/d/1tIsXCbB8P6ZxdnZNnv7S7BBWbbT7lrSjW990zG-vQAA/edit?gid=0#gid=0"
-
-# Streamlit app setup
-st.title("Booking.com 台南飯店資料爬取與分析")
-st.sidebar.header("功能選擇")
-mode = st.sidebar.selectbox("選擇模式", ["資料爬取", "資料視覺化", "上傳至 Google Sheet"])
-
-@st.cache_data
 def scrape_booking_hotel():
     url = "https://www.booking.com/searchresults.zh-tw.html"
     headers = {
@@ -35,94 +24,126 @@ def scrape_booking_hotel():
         'dest_id': '-2637868',
         'dest_type': 'city'
     }
+
     try:
         response = requests.get(url, headers=headers, params=params)
         response.raise_for_status()
         soup = BeautifulSoup(response.text, 'html.parser')
+
         hotels_data = []
         hotel_cards = soup.find_all('div', {'data-testid': 'property-card'})
 
         for hotel in hotel_cards:
-            name = hotel.find('div', {'data-testid': 'title'}).text.strip() if hotel.find('div', {'data-testid': 'title'}) else "無資料"
-            price_elem = hotel.find('span', {'data-testid': 'price-and-discounted-price'})
-            price = price_elem.text.strip().replace('TWD', '').replace(',', '').strip() if price_elem else "無資料"
-            rating_elem = hotel.find('div', {'class': 'a3b8729ab1'})
-            rating = rating_elem.text.strip() if rating_elem else "無評分"
-            description_elem = hotel.find('div', {'data-testid': 'recommended-units'})
-            if description_elem:
-                room_type = description_elem.find('h4').text.strip() if description_elem.find('h4') else ""
-                bed_info = description_elem.find('div').text.strip() if description_elem.find('div') else ""
-                cancellation = "可免費取消" if description_elem.find('strong', text='可免費取消') else ""
-                payment = "無需訂金" if description_elem.find('strong', text='無需訂金') else ""
-                description = f"{room_type} | {bed_info} | {cancellation} | {payment}".strip(' |')
-            else:
-                description = "無說明"
-
-            hotels_data.append({'飯店名稱': name, '價格': price, '評分': rating, '說明': description})
-        return pd.DataFrame(hotels_data).drop_duplicates()
-    except requests.RequestException:
-        st.error("無法從網站獲取資料")
-        return pd.DataFrame()
+            try:
+                name = hotel.find('div', {'data-testid': 'title', 'class': 'f6431b446c'}).text.strip() or "無資料"
+                price = hotel.find('span', {'data-testid': 'price-and-discounted-price', 'class': 'f6431b446c'}).text.strip() or "無資料"
+                price = price.replace('TWD', '').replace(' ', '').replace(',', '').strip()
 
-def clean_rating(x):
-    if pd.isna(x) or x == '無評分':
-        return 0
-    return float(str(x).replace('分數', '').replace('分', ''))
+                rating_container = hotel.find('div', {'class': 'a3b8729ab1'})
+                rating = rating_container.find('div', {'class': 'ac4a7896c7'}).text.strip() if rating_container else "無評分"
 
-def upload_to_google_sheets(df):
-    creds = Credentials.from_service_account_file(SERVICE_ACCOUNT_FILE, scopes=SCOPE)
-    gs = gspread.authorize(creds)
-    sheet = gs.open_by_url(SPREADSHEET_URL)
-    worksheet = sheet.get_worksheet(0)
-    worksheet.update([df.columns.values.tolist()] + df.astype(str).values.tolist())
-    return "資料已成功上傳到 Google Sheet！"
+                description_elem = hotel.find('div', {'data-testid': 'recommended-units'})
+                if description_elem:
+                    room_type = description_elem.find('h4', {'class': 'abf093bdfe'}).text.strip() if description_elem.find('h4', {'class': 'abf093bdfe'}) else ""
+                    bed_info = description_elem.find('div', {'class': 'abf093bdfe'}).text.strip() if description_elem.find('div', {'class': 'abf093bdfe'}) else ""
+                    cancellation = "可免費取消" if description_elem.find('strong', text='可免費取消') else ""
+                    payment = "無需訂金" if description_elem.find('strong', text='無需訂金') else ""
+                    description = f"{room_type} | {bed_info} | {cancellation} | {payment}".strip(' |')
+                else:
+                    description = "無說明"
+
+                hotels_data.append({
+                    '飯店名稱': name,
+                    '價格': price,
+                    '評分': rating,
+                    '說明': description
+                })
+
+            except AttributeError as e:
+                print(f"解析飯店資訊時發生錯誤: {e}")
+                continue
+
+        df = pd.DataFrame(hotels_data)
+        df = df.drop_duplicates()
+        return df
+
+    except requests.RequestException as e:
+        print(f"請求發生錯誤: {e}")
+        return pd.DataFrame()
 
-# Visualization functions
 def create_price_rating_scatter(df):
-    fig = px.scatter(df, x='價格', y='評分', text='飯店名稱', size='價格', color='評分',
-                     title='台南飯店價格與評分關係圖', labels={'價格': '房價 (TWD)', '評分': '評分 (0-10)'})
-    fig.update_layout(height=600, title_x=0.5)
+    fig = px.scatter(
+        df,
+        x='價格',
+        y='評分',
+        text='飯店名稱',
+        size='價格',
+        color='評分',
+        title='台南飯店價格與評分關係圖',
+        labels={'價格': '房價 (TWD)', '評分': '評分 (0-10)'}
+    )
+    fig.update_traces(textposition='top center', marker=dict(sizeref=2.*max(df['價格'])/(40.**2)))
+    fig.update_layout(height=600, showlegend=True, title_x=0.5, title_font_size=20)
     return fig
 
 def create_price_distribution(df):
     fig = go.Figure()
     fig.add_trace(go.Histogram(x=df['價格'], name='價格分布', nbinsx=10, marker_color='rgb(55, 83, 109)'))
     fig.add_trace(go.Box(x=df['價格'], name='價格箱型圖', marker_color='rgb(26, 118, 255)'))
-    fig.update_layout(title_text='台南飯店價格分布', title_x=0.5, height=500)
+    fig.update_layout(title_text='台南飯店價格分布', title_x=0.5, title_font_size=20, xaxis_title='價格 (TWD)', yaxis_title='數量', height=500, bargap=0.2, showlegend=True)
     return fig
 
-# Streamlit app implementation
-if mode == "資料爬取":
-    st.header("爬取台南飯店資料")
-    if st.button("開始爬取"):
-        df = scrape_booking_hotel()
-        if not df.empty:
-            st.dataframe(df)
-            df.to_csv('booking_hotels_tainan.csv', index=False, encoding='utf-8-sig')
-            st.success("資料爬取成功，已儲存至 booking_hotels_tainan.csv")
-        else:
-            st.error("未能成功爬取資料")
-
-elif mode == "資料視覺化":
-    st.header("分析與視覺化")
-    try:
-        df = pd.read_csv('booking_hotels_tainan.csv', encoding='utf-8-sig')
-        df['價格'] = pd.to_numeric(df['價格'], errors='coerce')
-        df['評分'] = df['評分'].apply(clean_rating)
-        df = df.dropna(subset=['價格'])
-
-        scatter_fig = create_price_rating_scatter(df)
-        st.plotly_chart(scatter_fig)
-        dist_fig = create_price_distribution(df)
-        st.plotly_chart(dist_fig)
-    except Exception as e:
-        st.error(f"讀取或分析資料時發生錯誤：{e}")
-
-elif mode == "上傳至 Google Sheet":
-    st.header("上傳資料至 Google Sheet")
-    try:
-        df = pd.read_csv('booking_hotels_tainan.csv', encoding='utf-8-sig')
-        result = upload_to_google_sheets(df)
-        st.success(result)
-    except Exception as e:
-        st.error(f"上傳資料時發生錯誤：{e}")
+def create_rating_box_by_price_range(df):
+    fig = px.box(df, x='價格區間', y='評分', title='不同價格區間的評分分布', labels={'價格區間': '價格類型', '評分': '評分 (0-10)'}, color='價格區間')
+    fig.update_layout(title_x=0.5, title_font_size=20, height=500, showlegend=False)
+    return fig
+
+def create_hotel_comparison(df):
+    fig = make_subplots(specs=[[{"secondary_y": True}]])
+    df_sorted = df.sort_values('評分', ascending=True)
+    fig.add_trace(go.Bar(x=df_sorted['飯店名稱'], y=df_sorted['評分'], name="評分", marker_color='rgb(55, 83, 109)'))
+    fig.add_trace(go.Scatter(x=df_sorted['飯店名稱'], y=df_sorted['價格'], name="價格", marker_color='rgb(26, 118, 255)'), secondary_y=True)
+    fig.update_layout(title_text='台南飯店評分與價格比較', title_x=0.5, title_font_size=20, height=700, showlegend=True, xaxis_tickangle=45)
+    fig.update_yaxes(title_text="評分", secondary_y=False)
+    fig.update_yaxes(title_text="價格 (TWD)", secondary_y=True)
+    return fig
+
+def update_google_sheet(df):
+    scope = ['https://www.googleapis.com/auth/spreadsheets']
+    creds = Credentials.from_service_account_file("realtime-441511-f5708eabdf26.json", scopes=scope)
+    gs = gspread.authorize(creds)
+    sheet = gs.open_by_url('https://docs.google.com/spreadsheets/d/1tIsXCbB8P6ZxdnZNnv7S7BBWbbT7lrSjW990zG-vQAA/edit?gid=0#gid=0')
+    worksheet = sheet.get_worksheet(0)
+    worksheet.update([df.columns.values.tolist()] + df.astype(str).values.tolist())
+    st.success("Data updated to Google Sheet successfully!")
+
+def main():
+    st.set_page_config(page_title="Booking.com Hotel Analysis")
+    st.title("Booking.com Hotel Analysis")
+
+    df = scrape_booking_hotel()
+
+    st.subheader("Hotel Data")
+    st.dataframe(df)
+
+    st.subheader("Price vs Rating Scatter Plot")
+    scatter_fig = create_price_rating_scatter(df)
+    st.plotly_chart(scatter_fig)
+
+    st.subheader("Price Distribution")
+    dist_fig = create_price_distribution(df)
+    st.plotly_chart(dist_fig)
+
+    st.subheader("Rating by Price Range")
+    box_fig = create_rating_box_by_price_range(df)
+    st.plotly_chart(box_fig)
+
+    st.subheader("Hotel Comparison")
+    comparison_fig = create_hotel_comparison(df)
+    st.plotly_chart(comparison_fig)
+
+    if st.button("Update Google Sheet"):
+        update_google_sheet(df)
+
+if __name__ == "__main__":
+    main()