Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -1,4 +1,4 @@
|
|
| 1 |
-
# HUGING_FACE_V4.2(輕量AI版).py - 已整合 XGBoost 模型
|
| 2 |
|
| 3 |
# 系統套件
|
| 4 |
import os
|
|
@@ -18,28 +18,38 @@ import requests
|
|
| 18 |
import time # 引用 time 模組以處理時間戳
|
| 19 |
|
| 20 |
# ========================= 引用外部模組 START =========================
|
|
|
|
| 21 |
from Bert_predict import BertPredictor
|
|
|
|
|
|
|
| 22 |
from model_predictor import XGBoostModel
|
| 23 |
# ========================== 引用外部模組 END ==========================
|
| 24 |
|
| 25 |
# ========================= 全域設定 START =========================
|
|
|
|
| 26 |
USE_ADVANCED_MODEL = True
|
|
|
|
|
|
|
|
|
|
| 27 |
ANALYSIS_CACHE = {}
|
|
|
|
| 28 |
CACHE_DURATION_SECONDS = 8 * 60 * 60
|
| 29 |
# ========================== CACHE 設定 END ==========================
|
| 30 |
|
|
|
|
| 31 |
try:
|
| 32 |
print("正在初始化 XGBoost 預測模型...")
|
| 33 |
xgb_model = XGBoostModel(default_model='xgboost_model')
|
| 34 |
print("XGBoost 預測模型初始化成功。")
|
| 35 |
except Exception as e:
|
| 36 |
print(f"錯誤:XGBoost 預測模型初始化失敗 - {e}")
|
|
|
|
| 37 |
USE_ADVANCED_MODEL = False
|
| 38 |
xgb_model = None
|
| 39 |
print("警告:已自動切換回簡易統計模型模式。")
|
| 40 |
# ========================== 全域設定 END ==========================
|
| 41 |
|
| 42 |
-
# 台股代號對應表
|
| 43 |
TAIWAN_STOCKS = {
|
| 44 |
'元大台灣50': '0050.TW', '台積電': '2330.TW', '聯發科': '2454.TW',
|
| 45 |
'鴻海': '2317.TW', '台達電': '2308.TW', '廣達': '2382.TW', '富邦金': '2881.TW',
|
|
@@ -58,7 +68,7 @@ TAIWAN_STOCKS = {
|
|
| 58 |
'譜瑞-KY': '4966.TWO', '貿聯-KY': '3665.TW', '騰雲': '6870.TWO', '穩懋': '3105.TWO'
|
| 59 |
}
|
| 60 |
|
| 61 |
-
# 產業分類
|
| 62 |
INDUSTRY_MAPPING = {
|
| 63 |
'0050.TW': 'ETF', '2330.TW': '半導體', '2454.TW': '半導體', '2317.TW': '電子組件',
|
| 64 |
'2308.TW': '電子', '2382.TW': '電子', '2881.TW': '金融', '2891.TW': '金融',
|
|
@@ -78,6 +88,7 @@ INDUSTRY_MAPPING = {
|
|
| 78 |
}
|
| 79 |
|
| 80 |
def get_stock_data(symbol, period='1y'):
|
|
|
|
| 81 |
try:
|
| 82 |
stock = yf.Ticker(symbol)
|
| 83 |
data = stock.history(period=period)
|
|
@@ -92,6 +103,7 @@ def get_stock_data(symbol, period='1y'):
|
|
| 92 |
return pd.DataFrame()
|
| 93 |
|
| 94 |
def simple_statistical_predict(data, predict_days=5):
|
|
|
|
| 95 |
if len(data) < 60: return None
|
| 96 |
prices = data['Close'].values
|
| 97 |
ma_short = np.mean(prices[-5:])
|
|
@@ -106,42 +118,101 @@ def simple_statistical_predict(data, predict_days=5):
|
|
| 106 |
change_pct = ((predicted_price - prices[-1]) / prices[-1]) * 100
|
| 107 |
return {'predicted_price': predicted_price, 'change_pct': change_pct, 'confidence': max(0.6, 1 - volatility * 2)}
|
| 108 |
|
|
|
|
|
|
|
| 109 |
def advanced_xgboost_predict(data, predict_days):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 110 |
if xgb_model is None or data.empty:
|
|
|
|
| 111 |
return None
|
| 112 |
-
|
| 113 |
-
|
| 114 |
-
|
| 115 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 116 |
return None
|
| 117 |
-
|
| 118 |
try:
|
| 119 |
-
|
| 120 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 121 |
prediction_key = day_to_key_map.get(predict_days)
|
|
|
|
| 122 |
if prediction_key is None or prediction_key not in predictions:
|
| 123 |
-
print(f"警告: XGBoost 模型沒有提供 {predict_days}
|
| 124 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 125 |
predicted_price = predictions[prediction_key]
|
| 126 |
current_price = data['Close'].iloc[-1]
|
| 127 |
change_pct = ((predicted_price - current_price) / current_price) * 100
|
| 128 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 129 |
except Exception as e:
|
| 130 |
print(f"執行 XGBoost 預測時發生錯誤: {e}")
|
|
|
|
|
|
|
| 131 |
return None
|
| 132 |
|
| 133 |
def get_prediction(data, predict_days=5):
|
|
|
|
|
|
|
|
|
|
|
|
|
| 134 |
if USE_ADVANCED_MODEL:
|
| 135 |
print(f"模式: 進階XGBoost模型 | 預測天期: {predict_days}天")
|
|
|
|
| 136 |
prediction = advanced_xgboost_predict(data, predict_days)
|
|
|
|
| 137 |
if prediction is not None:
|
| 138 |
return prediction
|
| 139 |
else:
|
| 140 |
print("進階模型預測失敗或無對應天期,自動降級為簡易統計模型。")
|
|
|
|
|
|
|
| 141 |
print(f"模式: 簡易統計模型 | 預測天期: {predict_days}天")
|
| 142 |
return simple_statistical_predict(data, predict_days)
|
| 143 |
|
| 144 |
def calculate_technical_indicators(df):
|
|
|
|
| 145 |
if df.empty: return df
|
| 146 |
df['MA5'] = df['Close'].rolling(window=5).mean()
|
| 147 |
df['MA20'] = df['Close'].rolling(window=20).mean()
|
|
@@ -178,43 +249,42 @@ def calculate_technical_indicators(df):
|
|
| 178 |
df['ADX'] = df['DX'].ewm(com=13, adjust=False).mean()
|
| 179 |
return df
|
| 180 |
|
| 181 |
-
#
|
| 182 |
def calculate_volume_profile(df, num_bins=50):
|
| 183 |
-
if df.empty or 'High' not in df.columns or 'Low' not in df.columns or 'Volume' not in df.columns:
|
| 184 |
-
return None, None, None
|
| 185 |
-
|
| 186 |
-
# 建立一個包含所需欄位的臨時 DataFrame
|
| 187 |
-
df_temp = pd.DataFrame({
|
| 188 |
-
'High': df['High'],
|
| 189 |
-
'Low': df['Low'],
|
| 190 |
-
'Close': df['Close'],
|
| 191 |
-
'Volume': df['Volume']
|
| 192 |
-
})
|
| 193 |
-
|
| 194 |
-
# 一次性移除任何欄位包含 NaN 的整行資料
|
| 195 |
-
df_temp.dropna(inplace=True)
|
| 196 |
-
|
| 197 |
-
if df_temp.empty:
|
| 198 |
return None, None, None
|
| 199 |
-
|
| 200 |
-
# 從清理過的 DataFrame 中獲取資料
|
| 201 |
-
all_prices = np.concatenate([df_temp['High'].values, df_temp['Low'].values])
|
| 202 |
|
| 203 |
-
|
|
|
|
|
|
|
| 204 |
return None, None, None
|
| 205 |
-
|
|
|
|
| 206 |
min_price, max_price = all_prices.min(), all_prices.max()
|
| 207 |
|
| 208 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 209 |
return None, None, None
|
| 210 |
-
|
| 211 |
-
price_for_volume = (df_temp['High'] + df_temp['Low'] + df_temp['Close']) / 3
|
| 212 |
-
weights = df_temp['Volume'].values
|
| 213 |
|
| 214 |
-
|
| 215 |
-
|
| 216 |
-
|
| 217 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 218 |
|
| 219 |
def get_business_climate_data():
|
| 220 |
try:
|
|
@@ -246,19 +316,27 @@ def get_pmi_data():
|
|
| 246 |
return pd.DataFrame()
|
| 247 |
|
| 248 |
def generate_gemini_analysis(stock_name, stock_symbol, period, data):
|
|
|
|
|
|
|
|
|
|
| 249 |
api_key = os.getenv("GEMINI_API_KEY")
|
| 250 |
if not api_key:
|
| 251 |
return "無法讀取 GEMINI API 金鑰", "請在系統環境變數中設定您的金鑰"
|
|
|
|
| 252 |
try:
|
| 253 |
genai.configure(api_key=api_key)
|
| 254 |
model = genai.GenerativeModel('gemini-1.5-flash')
|
|
|
|
| 255 |
price_change = ((data['Close'].iloc[-1] - data['Close'].iloc[0]) / data['Close'].iloc[0]) * 100
|
| 256 |
rsi_current = data['RSI'].iloc[-1]
|
| 257 |
macd_current = data['MACD'].iloc[-1]
|
| 258 |
macd_signal_current = data['MACD_Signal'].iloc[-1]
|
| 259 |
industry = INDUSTRY_MAPPING.get(stock_symbol, '綜合')
|
|
|
|
| 260 |
prompt = f"""
|
| 261 |
-
|
|
|
|
|
|
|
| 262 |
**股票資訊:**
|
| 263 |
- **公司名稱:** {stock_name} ({stock_symbol})
|
| 264 |
- **分析期間:** 最近 {period}
|
|
@@ -266,12 +344,23 @@ def generate_gemini_analysis(stock_name, stock_symbol, period, data):
|
|
| 266 |
- **期間價格變動:** {price_change:+.2f}%
|
| 267 |
- **目前 RSI 指標:** {rsi_current:.2f}
|
| 268 |
- **目前 MACD 指標:** MACD線為 {macd_current:.3f}, 信號線為 {macd_signal_current:.3f}
|
|
|
|
| 269 |
**你的任務:**
|
| 270 |
-
1. **基本面分析 (約 150 字):**
|
| 271 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 272 |
**輸出格式:**
|
| 273 |
-
請嚴格按照以下格式回傳,使用"$$"作為兩個段落之間的分隔符:
|
|
|
|
| 274 |
"""
|
|
|
|
| 275 |
response = model.generate_content(prompt)
|
| 276 |
parts = response.text.split('$$')
|
| 277 |
if len(parts) == 2:
|
|
@@ -279,7 +368,9 @@ def generate_gemini_analysis(stock_name, stock_symbol, period, data):
|
|
| 279 |
market_outlook = parts[1].strip()
|
| 280 |
return dcc.Markdown(fundamental_analysis), dcc.Markdown(market_outlook)
|
| 281 |
else:
|
|
|
|
| 282 |
return dcc.Markdown("無法解析 Gemini 回應,請稍後再試。"), dcc.Markdown(response.text)
|
|
|
|
| 283 |
except Exception as e:
|
| 284 |
error_message = f"呼叫 Gemini API 時發生錯誤: {str(e)}"
|
| 285 |
print(error_message)
|
|
@@ -287,7 +378,6 @@ def generate_gemini_analysis(stock_name, stock_symbol, period, data):
|
|
| 287 |
|
| 288 |
# 建立 Dash 應用程式
|
| 289 |
app = dash.Dash(__name__, suppress_callback_exceptions=True)
|
| 290 |
-
server = app.server
|
| 291 |
|
| 292 |
try:
|
| 293 |
print("正在初始化新聞情緒��析模型...")
|
|
@@ -297,7 +387,7 @@ except Exception as e:
|
|
| 297 |
print(f"錯誤:新聞情緒分析模型初始化失敗 - {e}")
|
| 298 |
predictor = None
|
| 299 |
|
| 300 |
-
# 應用程式佈局
|
| 301 |
app.layout = html.Div([
|
| 302 |
html.H1("台股分析儀表板", style={'text-align': 'center', 'margin-bottom': '30px'}),
|
| 303 |
html.Div([
|
|
@@ -316,6 +406,7 @@ app.layout = html.Div([
|
|
| 316 |
]),
|
| 317 |
html.Div([dcc.Graph(id='taiex-prediction-chart')], style={'margin-top': '20px'})
|
| 318 |
], style={'background': 'linear-gradient(135deg, #667eea 0%, #764ba2 100%)','padding': '25px','border-radius': '15px','box-shadow': '0 8px 25px rgba(0,0,0,0.15)','color': 'white','margin-bottom': '40px'}),
|
|
|
|
| 319 |
html.Div([
|
| 320 |
html.H3("📰 市場情緒與新聞分析", style={'color': '#E74C3C', 'margin-bottom': '20px'}),
|
| 321 |
html.Div([
|
|
@@ -329,6 +420,7 @@ app.layout = html.Div([
|
|
| 329 |
], style={'width': '48%', 'display': 'inline-block', 'margin-left': '4%'})
|
| 330 |
])
|
| 331 |
], style={'margin-top': '30px','padding': '20px','background': 'white','border-radius': '10px','box-shadow': '0 2px 10px rgba(0,0,0,0.1)'}),
|
|
|
|
| 332 |
html.Div([
|
| 333 |
html.H3("景氣燈號與 PMI 分析"),
|
| 334 |
html.Div([
|
|
@@ -336,10 +428,11 @@ app.layout = html.Div([
|
|
| 336 |
html.Div([dcc.Graph(id='pmi-chart')], style={'width': '48%', 'display': 'inline-block', 'margin-left': '2%'})
|
| 337 |
])
|
| 338 |
], style={'margin-top': '30px'}),
|
|
|
|
| 339 |
html.Div([
|
| 340 |
html.Div([
|
| 341 |
html.Label("選擇股票:"),
|
| 342 |
-
dcc.Dropdown(id='stock-dropdown', options=[{'label': name, 'value': symbol} for name, symbol in TAIWAN_STOCKS.items()], value='
|
| 343 |
], style={'width': '30%', 'display': 'inline-block', 'vertical-align': 'top'}),
|
| 344 |
html.Div([
|
| 345 |
html.Label("時間範圍:"),
|
|
@@ -352,6 +445,7 @@ app.layout = html.Div([
|
|
| 352 |
dcc.Dropdown(id='chart-type', options=[{'label': '線圖', 'value': 'line'},{'label': '蠟燭圖', 'value': 'candlestick'}], value='candlestick', style={'margin-bottom': '10px'})
|
| 353 |
], style={'width': '30%', 'display': 'inline-block', 'margin-left': '5%', 'vertical-align': 'top'})
|
| 354 |
], style={'margin-bottom': '30px'}),
|
|
|
|
| 355 |
html.Div(id='stock-info-cards', style={'margin-bottom': '30px'}),
|
| 356 |
html.Div([html.Div([dcc.Graph(id='price-chart')], style={'width': '100%', 'display': 'inline-block', 'vertical-align': 'top'})]),
|
| 357 |
html.Div([
|
|
@@ -404,7 +498,6 @@ app.layout = html.Div([
|
|
| 404 |
], style={'margin-top': '30px','padding': '20px','background': 'white','border-radius': '10px','box-shadow': '0 2px 10px rgba(0,0,0,0.1)'}),
|
| 405 |
])
|
| 406 |
|
| 407 |
-
# 所有 Callback 函式 (省略)
|
| 408 |
@app.callback(
|
| 409 |
[dash.dependencies.Output('taiex-prediction-results', 'children'),
|
| 410 |
dash.dependencies.Output('taiex-prediction-chart', 'figure')],
|
|
@@ -413,18 +506,26 @@ app.layout = html.Div([
|
|
| 413 |
def update_taiex_prediction(predict_days):
|
| 414 |
data = get_stock_data('^TWII', '2y')
|
| 415 |
if data.empty: return html.Div("無法獲取台指期資料"), {}
|
|
|
|
|
|
|
| 416 |
final_prediction = get_prediction(data, predict_days)
|
| 417 |
-
|
|
|
|
| 418 |
current_price, last_date = data['Close'].iloc[-1], data.index[-1]
|
| 419 |
predicted_price, change_pct, confidence = final_prediction['predicted_price'], final_prediction['change_pct'], final_prediction['confidence']
|
| 420 |
-
|
|
|
|
| 421 |
intervals_to_predict = prediction_paths.get(predict_days, [predict_days])
|
| 422 |
prediction_dates, prediction_prices = [last_date], [current_price]
|
|
|
|
| 423 |
for days in intervals_to_predict:
|
|
|
|
| 424 |
interim_prediction = get_prediction(data, days)
|
| 425 |
if interim_prediction:
|
| 426 |
prediction_dates.append(last_date + timedelta(days=days))
|
| 427 |
prediction_prices.append(interim_prediction['predicted_price'])
|
|
|
|
|
|
|
| 428 |
color, arrow = ('red', '📈') if change_pct >= 0 else ('green', '📉')
|
| 429 |
result_card = html.Div([
|
| 430 |
html.H4(f"{predict_days}日後預測結果", style={'margin': '0 0 15px 0', 'color': 'white'}),
|
|
@@ -472,22 +573,82 @@ def update_stock_info(selected_stock):
|
|
| 472 |
dash.dependencies.Input('period-dropdown', 'value'),
|
| 473 |
dash.dependencies.Input('chart-type', 'value')]
|
| 474 |
)
|
| 475 |
-
|
|
|
|
| 476 |
data = get_stock_data(selected_stock, period)
|
| 477 |
-
if data.empty:
|
|
|
|
|
|
|
| 478 |
data = calculate_technical_indicators(data)
|
| 479 |
stock_name = [name for name, symbol in TAIWAN_STOCKS.items() if symbol == selected_stock][0]
|
| 480 |
-
|
|
|
|
|
|
|
|
|
|
| 481 |
if chart_type == 'candlestick':
|
| 482 |
-
fig.add_trace(go.Candlestick(
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 483 |
else:
|
| 484 |
-
fig.add_trace(
|
| 485 |
-
|
| 486 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 487 |
bin_edges, volume_per_bin, price_centers = calculate_volume_profile(data, num_bins=50)
|
| 488 |
-
|
| 489 |
-
|
| 490 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 491 |
return fig
|
| 492 |
|
| 493 |
@app.callback(
|
|
@@ -573,15 +734,30 @@ def update_industry_analysis(selected_stock):
|
|
| 573 |
data = get_stock_data(symbol, '1mo')
|
| 574 |
if not data.empty and len(data) > 1:
|
| 575 |
return_pct = ((data['Close'].iloc[-1] - data['Close'].iloc[0]) / data['Close'].iloc[0]) * 100
|
| 576 |
-
performance_data.append({
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 577 |
if not performance_data:
|
| 578 |
fig = go.Figure().add_annotation(text="無法計算產業資料", showarrow=False)
|
| 579 |
fig.update_layout(title="近一月市場波動最大標的", height=400)
|
| 580 |
return fig
|
| 581 |
df_performance = pd.DataFrame(performance_data)
|
| 582 |
df_top_movers = df_performance.sort_values(by='絕對波動', ascending=False).head(10)
|
| 583 |
-
fig = px.pie(
|
| 584 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 585 |
fig.update_layout(height=400, showlegend=False)
|
| 586 |
return fig
|
| 587 |
|
|
@@ -619,28 +795,41 @@ def update_business_climate_chart(selected_stock):
|
|
| 619 |
def update_analysis_text(selected_stock, period):
|
| 620 |
cache_key = f"{selected_stock}-{period}"
|
| 621 |
current_time = time.time()
|
|
|
|
| 622 |
if cache_key in ANALYSIS_CACHE:
|
| 623 |
cached_data = ANALYSIS_CACHE[cache_key]
|
| 624 |
if current_time - cached_data['timestamp'] < CACHE_DURATION_SECONDS:
|
| 625 |
print(f"從快取載入分析: {cache_key}")
|
| 626 |
return cached_data['technical'], cached_data['fundamental'], cached_data['outlook']
|
| 627 |
-
|
|
|
|
| 628 |
data = get_stock_data(selected_stock, period)
|
| 629 |
stock_name = [name for name, symbol in TAIWAN_STOCKS.items() if symbol == selected_stock][0]
|
| 630 |
if data.empty or len(data) < 20:
|
| 631 |
return "資料不足,無法分析", "資料不足,無法分析", "資料不足,無法分析"
|
|
|
|
| 632 |
data = calculate_technical_indicators(data)
|
|
|
|
| 633 |
price_change = ((data['Close'].iloc[-1] - data['Close'].iloc[0]) / data['Close'].iloc[0]) * 100
|
| 634 |
rsi_current = data['RSI'].iloc[-1] if not pd.isna(data['RSI'].iloc[-1]) else 50
|
| 635 |
macd_current = data['MACD'].iloc[-1] if not pd.isna(data['MACD'].iloc[-1]) else 0
|
| 636 |
macd_signal_current = data['MACD_Signal'].iloc[-1] if not pd.isna(data['MACD_Signal'].iloc[-1]) else 0
|
|
|
|
| 637 |
technical_text = html.Div([
|
| 638 |
html.P([html.Strong("價格趨勢:"), f"在最近 {period} 期間內,{stock_name} 股價呈現", html.Span(f"{'上漲' if price_change > 5 else '下跌' if price_change < -5 else '盤整'}", style={'color': 'red' if price_change > 5 else 'green' if price_change < -5 else 'orange', 'font-weight': 'bold'}), f"走勢,累計變動 {price_change:+.1f}%。"]),
|
| 639 |
html.P([html.Strong("RSI 指標:"), f"目前的 RSI 值為 {rsi_current:.1f},", html.Span("處於超買區(>70)" if rsi_current > 70 else "處於超賣區(<30)" if rsi_current < 30 else "在正常範圍內", style={'color': 'green' if rsi_current > 70 else 'red' if rsi_current < 30 else 'blue', 'font-weight': 'bold'}), "。"]),
|
| 640 |
html.P([html.Strong("MACD 指標:"), f"MACD 快線 ({macd_current:.3f}) 目前", html.Span("高於" if macd_current > macd_signal_current else "低於", style={'color': 'red' if macd_current > macd_signal_current else 'green', 'font-weight': 'bold'}), f" Signal 慢線 ({macd_signal_current:.3f}),", f"顯示市場動能偏向{'多頭' if macd_current > macd_signal_current else '空頭'}。"]),
|
| 641 |
])
|
|
|
|
| 642 |
fundamental_text, market_outlook_text = generate_gemini_analysis(stock_name, selected_stock, period, data)
|
| 643 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 644 |
return technical_text, fundamental_text, market_outlook_text
|
| 645 |
|
| 646 |
@app.callback(
|
|
@@ -661,20 +850,32 @@ def update_pmi_chart(selected_stock):
|
|
| 661 |
return fig
|
| 662 |
|
| 663 |
def summarize_news_with_gemini(news_list: list) -> str:
|
|
|
|
|
|
|
|
|
|
| 664 |
api_key = os.getenv("GEMINI_API_KEY")
|
| 665 |
if not api_key:
|
| 666 |
return "錯誤:找不到 GEMINI_API_KEY。請在 Hugging Face Secrets 中設定。"
|
|
|
|
| 667 |
try:
|
| 668 |
genai.configure(api_key=api_key)
|
| 669 |
model = genai.GenerativeModel('gemini-1.5-flash')
|
|
|
|
| 670 |
formatted_news = "\n".join([f"- {news}" for news in news_list])
|
|
|
|
| 671 |
prompt = f"""
|
| 672 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 673 |
英文新聞標題如下:
|
| 674 |
{formatted_news}
|
| 675 |
"""
|
|
|
|
| 676 |
response = model.generate_content(prompt)
|
| 677 |
return response.text
|
|
|
|
| 678 |
except Exception as e:
|
| 679 |
print(f"呼叫 Gemini API 時發生錯誤: {e}")
|
| 680 |
return f"無法生成新聞摘要,請稍後再試。錯誤訊息:{e}"
|
|
@@ -699,7 +900,7 @@ def update_comparison_analysis(selected_stocks, period):
|
|
| 699 |
normalized_prices = (data['Close'] / data['Close'].iloc[0]) * 100
|
| 700 |
fig.add_trace(go.Scatter(x=data.index, y=normalized_prices, mode='lines', name=stock_name, line=dict(width=2)))
|
| 701 |
total_return = ((data['Close'].iloc[-1] / data['Close'].iloc[0]) - 1) * 100
|
| 702 |
-
volatility = data['Close'].pct_change(
|
| 703 |
comparison_data.append({'name': stock_name, 'return': total_return, 'volatility': volatility, 'current_price': data['Close'].iloc[-1]})
|
| 704 |
fig.update_layout(title=f'股票績效比較 - {period}', xaxis_title='日期', yaxis_title='相對績效 (基期=100)', height=400, hovermode='x unified')
|
| 705 |
if comparison_data:
|
|
@@ -721,7 +922,9 @@ def update_sentiment_analysis(selected_stock):
|
|
| 721 |
error_fig = go.Figure().add_annotation(text="情緒指標模型載入失敗", showarrow=False)
|
| 722 |
error_fig.update_layout(height=200)
|
| 723 |
return dcc.Graph(figure=error_fig), html.P("新聞分析模型載入失敗,請檢查後台日誌。")
|
|
|
|
| 724 |
sentiment_score_raw = predictor.get_news_index()
|
|
|
|
| 725 |
if sentiment_score_raw is not None:
|
| 726 |
sentiment_score_normalized = (sentiment_score_raw + 1) * 50
|
| 727 |
sentiment_score_normalized = max(0, min(100, sentiment_score_normalized))
|
|
@@ -746,15 +949,21 @@ def update_sentiment_analysis(selected_stock):
|
|
| 746 |
error_fig = go.Figure().add_annotation(text="今日尚無情緒分數", showarrow=False)
|
| 747 |
error_fig.update_layout(height=200)
|
| 748 |
gauge_content = dcc.Graph(figure=error_fig)
|
|
|
|
| 749 |
top_news_list = predictor.get_news()
|
| 750 |
news_content = None
|
|
|
|
| 751 |
if top_news_list and isinstance(top_news_list, list):
|
| 752 |
summary_text = summarize_news_with_gemini(top_news_list)
|
| 753 |
-
news_content = dcc.Markdown(summary_text, style={
|
|
|
|
|
|
|
|
|
|
| 754 |
elif top_news_list == []:
|
| 755 |
news_content = html.P("昨日無重大相關新聞。", style={'text-align': 'center', 'padding-top': '50px'})
|
| 756 |
else:
|
| 757 |
news_content = html.P("讀取新聞時發生錯誤。", style={'text-align': 'center', 'padding-top': '50px'})
|
|
|
|
| 758 |
return gauge_content, news_content
|
| 759 |
|
| 760 |
# 主程式執行
|
|
|
|
| 1 |
+
# HUGING_FACE_V4.2(輕量AI版).py - 已整合 XGBoost 模型
|
| 2 |
|
| 3 |
# 系統套件
|
| 4 |
import os
|
|
|
|
| 18 |
import time # 引用 time 模組以處理時間戳
|
| 19 |
|
| 20 |
# ========================= 引用外部模組 START =========================
|
| 21 |
+
# 引用您組員的預測器程式
|
| 22 |
from Bert_predict import BertPredictor
|
| 23 |
+
|
| 24 |
+
# 【修改 1】: 匯入 XGBoostModel 類別
|
| 25 |
from model_predictor import XGBoostModel
|
| 26 |
# ========================== 引用外部模組 END ==========================
|
| 27 |
|
| 28 |
# ========================= 全域設定 START =========================
|
| 29 |
+
# 【修改 2】: 將開關設為 True 來啟用您的 XGBoost 模型
|
| 30 |
USE_ADVANCED_MODEL = True
|
| 31 |
+
|
| 32 |
+
# ========================= CACHE 設定 START =========================
|
| 33 |
+
# 分析結果的快取字典
|
| 34 |
ANALYSIS_CACHE = {}
|
| 35 |
+
# 快取有效時間(秒),例如:8 小時 = 8 * 60 * 60 = 28800 秒
|
| 36 |
CACHE_DURATION_SECONDS = 8 * 60 * 60
|
| 37 |
# ========================== CACHE 設定 END ==========================
|
| 38 |
|
| 39 |
+
# 【修改 3】: 在應用程式啟動時,預先載入 XGBoost 模型
|
| 40 |
try:
|
| 41 |
print("正在初始化 XGBoost 預測模型...")
|
| 42 |
xgb_model = XGBoostModel(default_model='xgboost_model')
|
| 43 |
print("XGBoost 預測模型初始化成功。")
|
| 44 |
except Exception as e:
|
| 45 |
print(f"錯誤:XGBoost 預測模型初始化失敗 - {e}")
|
| 46 |
+
# 如果模型載入失敗,則強制關閉進階模型開關,退回簡易模式
|
| 47 |
USE_ADVANCED_MODEL = False
|
| 48 |
xgb_model = None
|
| 49 |
print("警告:已自動切換回簡易統計模型模式。")
|
| 50 |
# ========================== 全域設定 END ==========================
|
| 51 |
|
| 52 |
+
# 台股代號對應表
|
| 53 |
TAIWAN_STOCKS = {
|
| 54 |
'元大台灣50': '0050.TW', '台積電': '2330.TW', '聯發科': '2454.TW',
|
| 55 |
'鴻海': '2317.TW', '台達電': '2308.TW', '廣達': '2382.TW', '富邦金': '2881.TW',
|
|
|
|
| 68 |
'譜瑞-KY': '4966.TWO', '貿聯-KY': '3665.TW', '騰雲': '6870.TWO', '穩懋': '3105.TWO'
|
| 69 |
}
|
| 70 |
|
| 71 |
+
# 產業分類
|
| 72 |
INDUSTRY_MAPPING = {
|
| 73 |
'0050.TW': 'ETF', '2330.TW': '半導體', '2454.TW': '半導體', '2317.TW': '電子組件',
|
| 74 |
'2308.TW': '電子', '2382.TW': '電子', '2881.TW': '金融', '2891.TW': '金融',
|
|
|
|
| 88 |
}
|
| 89 |
|
| 90 |
def get_stock_data(symbol, period='1y'):
|
| 91 |
+
"""獲取股票資料"""
|
| 92 |
try:
|
| 93 |
stock = yf.Ticker(symbol)
|
| 94 |
data = stock.history(period=period)
|
|
|
|
| 103 |
return pd.DataFrame()
|
| 104 |
|
| 105 |
def simple_statistical_predict(data, predict_days=5):
|
| 106 |
+
"""【備用模型】簡化的統計預測模型。"""
|
| 107 |
if len(data) < 60: return None
|
| 108 |
prices = data['Close'].values
|
| 109 |
ma_short = np.mean(prices[-5:])
|
|
|
|
| 118 |
change_pct = ((predicted_price - prices[-1]) / prices[-1]) * 100
|
| 119 |
return {'predicted_price': predicted_price, 'change_pct': change_pct, 'confidence': max(0.6, 1 - volatility * 2)}
|
| 120 |
|
| 121 |
+
# 【修改 4】: 建立一個新的函式來處理 XGBoost 模型的輸入和輸出
|
| 122 |
+
# 修正後的 advanced_xgboost_predict 函數
|
| 123 |
def advanced_xgboost_predict(data, predict_days):
|
| 124 |
+
"""
|
| 125 |
+
【進階模型橋接函式】
|
| 126 |
+
- 準備 XGBoost 模型所需的輸入 DataFrame。
|
| 127 |
+
- 呼叫模型進行預測。
|
| 128 |
+
- 將模型的輸出格式轉換為主程式所需的格式。
|
| 129 |
+
"""
|
| 130 |
if xgb_model is None or data.empty:
|
| 131 |
+
print("XGBoost 模型未載入或數據為空")
|
| 132 |
return None
|
| 133 |
+
|
| 134 |
+
# 1. 準備輸入資料
|
| 135 |
+
# 確保數據有足夠的歷史記錄
|
| 136 |
+
if len(data) < 20:
|
| 137 |
+
print("歷史數據不足,無法使用 XGBoost 模型")
|
| 138 |
+
return None
|
| 139 |
+
|
| 140 |
+
# 使用最新的資料點來進行未來預測
|
| 141 |
+
input_df = data.tail(1).copy()
|
| 142 |
+
|
| 143 |
+
# 檢查必要欄位是否存在
|
| 144 |
+
required_columns = ['Open', 'High', 'Low', 'Close', 'Volume']
|
| 145 |
+
missing_columns = [col for col in required_columns if col not in input_df.columns]
|
| 146 |
+
if missing_columns:
|
| 147 |
+
print(f"缺少必要欄位: {missing_columns}")
|
| 148 |
return None
|
| 149 |
+
|
| 150 |
try:
|
| 151 |
+
# 2. 呼叫模型預測
|
| 152 |
+
print(f"呼叫 XGBoost 模型進行 {predict_days} 天預測...")
|
| 153 |
+
predictions = xgb_model.predict('xgboost_model', input_df)
|
| 154 |
+
|
| 155 |
+
# 3. 根據 predict_days 解析輸出
|
| 156 |
+
# 建立預測天數到模型輸出鍵的映射
|
| 157 |
+
day_to_key_map = {
|
| 158 |
+
1: 'Close_t0_pred', # 假��� t0 代表 1 天後
|
| 159 |
+
5: 'Close_t5_pred',
|
| 160 |
+
10: 'Close_t10_pred',
|
| 161 |
+
20: 'Close_t20_pred'
|
| 162 |
+
}
|
| 163 |
+
|
| 164 |
+
# 找到對應的預測鍵
|
| 165 |
prediction_key = day_to_key_map.get(predict_days)
|
| 166 |
+
|
| 167 |
if prediction_key is None or prediction_key not in predictions:
|
| 168 |
+
print(f"警告: XGBoost 模型沒有提供 {predict_days} 天的預測結果。可用鍵值: {list(predictions.keys())}")
|
| 169 |
+
# 如果沒有對應的預測期間,嘗試使用最接近的
|
| 170 |
+
available_days = [1, 5, 10, 20]
|
| 171 |
+
closest_day = min(available_days, key=lambda x: abs(x - predict_days))
|
| 172 |
+
prediction_key = day_to_key_map[closest_day]
|
| 173 |
+
print(f"使用最接近的預測期間: {closest_day} 天")
|
| 174 |
+
|
| 175 |
predicted_price = predictions[prediction_key]
|
| 176 |
current_price = data['Close'].iloc[-1]
|
| 177 |
change_pct = ((predicted_price - current_price) / current_price) * 100
|
| 178 |
+
|
| 179 |
+
# 4. 包裝成主程式所需的格式
|
| 180 |
+
result = {
|
| 181 |
+
'predicted_price': float(predicted_price),
|
| 182 |
+
'change_pct': float(change_pct),
|
| 183 |
+
'confidence': 0.85 # XGBoost 模型通常有較高的信心度
|
| 184 |
+
}
|
| 185 |
+
|
| 186 |
+
print(f"XGBoost 預測成功: 當前價格={current_price:.2f}, 預測價格={predicted_price:.2f}, 變化={change_pct:.2f}%")
|
| 187 |
+
return result
|
| 188 |
+
|
| 189 |
except Exception as e:
|
| 190 |
print(f"執行 XGBoost 預測時發生錯誤: {e}")
|
| 191 |
+
import traceback
|
| 192 |
+
traceback.print_exc()
|
| 193 |
return None
|
| 194 |
|
| 195 |
def get_prediction(data, predict_days=5):
|
| 196 |
+
"""
|
| 197 |
+
【【模型預測控制器】】
|
| 198 |
+
根據 USE_ADVANCED_MODEL 的設定,呼叫對應的預測模型。
|
| 199 |
+
"""
|
| 200 |
if USE_ADVANCED_MODEL:
|
| 201 |
print(f"模式: 進階XGBoost模型 | 預測天期: {predict_days}天")
|
| 202 |
+
# 【修改 5】: 呼叫新的 XGBoost 橋接函式
|
| 203 |
prediction = advanced_xgboost_predict(data, predict_days)
|
| 204 |
+
# 如果進階模型預測失敗,則自動降級使用簡易模型
|
| 205 |
if prediction is not None:
|
| 206 |
return prediction
|
| 207 |
else:
|
| 208 |
print("進階模型預測失敗或無對應天期,自動降級為簡易統計模型。")
|
| 209 |
+
|
| 210 |
+
# 預設或降級時執行簡易模型
|
| 211 |
print(f"模式: 簡易統計模型 | 預測天期: {predict_days}天")
|
| 212 |
return simple_statistical_predict(data, predict_days)
|
| 213 |
|
| 214 |
def calculate_technical_indicators(df):
|
| 215 |
+
"""計算技術指標"""
|
| 216 |
if df.empty: return df
|
| 217 |
df['MA5'] = df['Close'].rolling(window=5).mean()
|
| 218 |
df['MA20'] = df['Close'].rolling(window=20).mean()
|
|
|
|
| 249 |
df['ADX'] = df['DX'].ewm(com=13, adjust=False).mean()
|
| 250 |
return df
|
| 251 |
|
| 252 |
+
# 修正後的 calculate_volume_profile 函數
|
| 253 |
def calculate_volume_profile(df, num_bins=50):
|
| 254 |
+
if df.empty or 'High' not in df.columns or 'Low' not in df.columns or 'Volume' not in df.columns:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 255 |
return None, None, None
|
|
|
|
|
|
|
|
|
|
| 256 |
|
| 257 |
+
# 確保沒有 NaN 值
|
| 258 |
+
df_clean = df.dropna(subset=['High', 'Low', 'Close', 'Volume'])
|
| 259 |
+
if df_clean.empty:
|
| 260 |
return None, None, None
|
| 261 |
+
|
| 262 |
+
all_prices = np.concatenate([df_clean['High'].values, df_clean['Low'].values])
|
| 263 |
min_price, max_price = all_prices.min(), all_prices.max()
|
| 264 |
|
| 265 |
+
# 使用典型價格 (High + Low + Close) / 3 作為價格指標
|
| 266 |
+
price_for_volume = (df_clean['High'] + df_clean['Low'] + df_clean['Close']) / 3
|
| 267 |
+
|
| 268 |
+
# 移除 NaN 值並確保對應的權重也被移除
|
| 269 |
+
price_indicator = price_for_volume.dropna()
|
| 270 |
+
corresponding_volume = df_clean['Volume'].loc[price_indicator.index]
|
| 271 |
+
|
| 272 |
+
# 再次檢查是否有空數據
|
| 273 |
+
if len(price_indicator) == 0 or len(corresponding_volume) == 0:
|
| 274 |
return None, None, None
|
|
|
|
|
|
|
|
|
|
| 275 |
|
| 276 |
+
try:
|
| 277 |
+
hist, bin_edges = np.histogram(
|
| 278 |
+
price_indicator.values,
|
| 279 |
+
bins=num_bins,
|
| 280 |
+
range=(min_price, max_price),
|
| 281 |
+
weights=corresponding_volume.values
|
| 282 |
+
)
|
| 283 |
+
price_centers = (bin_edges[:-1] + bin_edges[1:]) / 2
|
| 284 |
+
return bin_edges, hist, price_centers
|
| 285 |
+
except Exception as e:
|
| 286 |
+
print(f"Volume profile 計算錯誤: {e}")
|
| 287 |
+
return None, None, None
|
| 288 |
|
| 289 |
def get_business_climate_data():
|
| 290 |
try:
|
|
|
|
| 316 |
return pd.DataFrame()
|
| 317 |
|
| 318 |
def generate_gemini_analysis(stock_name, stock_symbol, period, data):
|
| 319 |
+
"""
|
| 320 |
+
使用 Gemini API 生成基本面和市場展望分析。
|
| 321 |
+
"""
|
| 322 |
api_key = os.getenv("GEMINI_API_KEY")
|
| 323 |
if not api_key:
|
| 324 |
return "無法讀取 GEMINI API 金鑰", "請在系統環境變數中設定您的金鑰"
|
| 325 |
+
|
| 326 |
try:
|
| 327 |
genai.configure(api_key=api_key)
|
| 328 |
model = genai.GenerativeModel('gemini-1.5-flash')
|
| 329 |
+
|
| 330 |
price_change = ((data['Close'].iloc[-1] - data['Close'].iloc[0]) / data['Close'].iloc[0]) * 100
|
| 331 |
rsi_current = data['RSI'].iloc[-1]
|
| 332 |
macd_current = data['MACD'].iloc[-1]
|
| 333 |
macd_signal_current = data['MACD_Signal'].iloc[-1]
|
| 334 |
industry = INDUSTRY_MAPPING.get(stock_symbol, '綜合')
|
| 335 |
+
|
| 336 |
prompt = f"""
|
| 337 |
+
請扮演一位專業、資深的台灣股市金融分析師。
|
| 338 |
+
我將提供一檔台股的即時技術指標數據,請你基於這些數據,結合你對這家公司、其所在產業以及當前市場趨勢的理解,為我生成一段專業的「基本面分析」和一段「市場展望與投資建議」。
|
| 339 |
+
|
| 340 |
**股票資訊:**
|
| 341 |
- **公司名稱:** {stock_name} ({stock_symbol})
|
| 342 |
- **分析期間:** 最近 {period}
|
|
|
|
| 344 |
- **期間價格變動:** {price_change:+.2f}%
|
| 345 |
- **目前 RSI 指標:** {rsi_current:.2f}
|
| 346 |
- **目前 MACD 指標:** MACD線為 {macd_current:.3f}, 信號線為 {macd_signal_current:.3f}
|
| 347 |
+
|
| 348 |
**你的任務:**
|
| 349 |
+
1. **基本面分析 (約 150 字):**
|
| 350 |
+
- 評論這家公司的產業地位、近期營運亮點或挑戰。
|
| 351 |
+
- 提及任何可能影響其基本面的關鍵因素 (例如:財報、法說會、政策、供應鏈變化等)。
|
| 352 |
+
- 請用專業、客觀的語氣撰寫。
|
| 353 |
+
|
| 354 |
+
2. **市場展望與投資建議 (約 150 字):**
|
| 355 |
+
- 基於上述所有資訊,提供對該股票的短期和中期市場展望。
|
| 356 |
+
- 提出具體的投資建議,例如:適合何種類型的投資人、潛在的風險點。
|
| 357 |
+
- 請直接提供分析內容,不要包含任何問候語。
|
| 358 |
+
|
| 359 |
**輸出格式:**
|
| 360 |
+
請嚴格按照以下格式回傳,使用"$$"作為兩個段落之間的分隔符:
|
| 361 |
+
[基本面分析內容]$$[市場展望與投資建議內容]
|
| 362 |
"""
|
| 363 |
+
|
| 364 |
response = model.generate_content(prompt)
|
| 365 |
parts = response.text.split('$$')
|
| 366 |
if len(parts) == 2:
|
|
|
|
| 368 |
market_outlook = parts[1].strip()
|
| 369 |
return dcc.Markdown(fundamental_analysis), dcc.Markdown(market_outlook)
|
| 370 |
else:
|
| 371 |
+
# Fallback for unexpected response format
|
| 372 |
return dcc.Markdown("無法解析 Gemini 回應,請稍後再試。"), dcc.Markdown(response.text)
|
| 373 |
+
|
| 374 |
except Exception as e:
|
| 375 |
error_message = f"呼叫 Gemini API 時發生錯誤: {str(e)}"
|
| 376 |
print(error_message)
|
|
|
|
| 378 |
|
| 379 |
# 建立 Dash 應用程式
|
| 380 |
app = dash.Dash(__name__, suppress_callback_exceptions=True)
|
|
|
|
| 381 |
|
| 382 |
try:
|
| 383 |
print("正在初始化新聞情緒��析模型...")
|
|
|
|
| 387 |
print(f"錯誤:新聞情緒分析模型初始化失敗 - {e}")
|
| 388 |
predictor = None
|
| 389 |
|
| 390 |
+
# 應用程式佈局
|
| 391 |
app.layout = html.Div([
|
| 392 |
html.H1("台股分析儀表板", style={'text-align': 'center', 'margin-bottom': '30px'}),
|
| 393 |
html.Div([
|
|
|
|
| 406 |
]),
|
| 407 |
html.Div([dcc.Graph(id='taiex-prediction-chart')], style={'margin-top': '20px'})
|
| 408 |
], style={'background': 'linear-gradient(135deg, #667eea 0%, #764ba2 100%)','padding': '25px','border-radius': '15px','box-shadow': '0 8px 25px rgba(0,0,0,0.15)','color': 'white','margin-bottom': '40px'}),
|
| 409 |
+
|
| 410 |
html.Div([
|
| 411 |
html.H3("📰 市場情緒與新聞分析", style={'color': '#E74C3C', 'margin-bottom': '20px'}),
|
| 412 |
html.Div([
|
|
|
|
| 420 |
], style={'width': '48%', 'display': 'inline-block', 'margin-left': '4%'})
|
| 421 |
])
|
| 422 |
], style={'margin-top': '30px','padding': '20px','background': 'white','border-radius': '10px','box-shadow': '0 2px 10px rgba(0,0,0,0.1)'}),
|
| 423 |
+
|
| 424 |
html.Div([
|
| 425 |
html.H3("景氣燈號與 PMI 分析"),
|
| 426 |
html.Div([
|
|
|
|
| 428 |
html.Div([dcc.Graph(id='pmi-chart')], style={'width': '48%', 'display': 'inline-block', 'margin-left': '2%'})
|
| 429 |
])
|
| 430 |
], style={'margin-top': '30px'}),
|
| 431 |
+
|
| 432 |
html.Div([
|
| 433 |
html.Div([
|
| 434 |
html.Label("選擇股票:"),
|
| 435 |
+
dcc.Dropdown(id='stock-dropdown', options=[{'label': name, 'value': symbol} for name, symbol in TAIWAN_STOCKS.items()], value='0050.TW', style={'margin-bottom': '10px'})
|
| 436 |
], style={'width': '30%', 'display': 'inline-block', 'vertical-align': 'top'}),
|
| 437 |
html.Div([
|
| 438 |
html.Label("時間範圍:"),
|
|
|
|
| 445 |
dcc.Dropdown(id='chart-type', options=[{'label': '線圖', 'value': 'line'},{'label': '蠟燭圖', 'value': 'candlestick'}], value='candlestick', style={'margin-bottom': '10px'})
|
| 446 |
], style={'width': '30%', 'display': 'inline-block', 'margin-left': '5%', 'vertical-align': 'top'})
|
| 447 |
], style={'margin-bottom': '30px'}),
|
| 448 |
+
|
| 449 |
html.Div(id='stock-info-cards', style={'margin-bottom': '30px'}),
|
| 450 |
html.Div([html.Div([dcc.Graph(id='price-chart')], style={'width': '100%', 'display': 'inline-block', 'vertical-align': 'top'})]),
|
| 451 |
html.Div([
|
|
|
|
| 498 |
], style={'margin-top': '30px','padding': '20px','background': 'white','border-radius': '10px','box-shadow': '0 2px 10px rgba(0,0,0,0.1)'}),
|
| 499 |
])
|
| 500 |
|
|
|
|
| 501 |
@app.callback(
|
| 502 |
[dash.dependencies.Output('taiex-prediction-results', 'children'),
|
| 503 |
dash.dependencies.Output('taiex-prediction-chart', 'figure')],
|
|
|
|
| 506 |
def update_taiex_prediction(predict_days):
|
| 507 |
data = get_stock_data('^TWII', '2y')
|
| 508 |
if data.empty: return html.Div("無法獲取台指期資料"), {}
|
| 509 |
+
|
| 510 |
+
# === 呼叫 get_prediction 控制器,它會自動選擇模型 ===
|
| 511 |
final_prediction = get_prediction(data, predict_days)
|
| 512 |
+
|
| 513 |
+
if final_prediction is None: return html.Div("資料不足,無法進行預測"), {}
|
| 514 |
current_price, last_date = data['Close'].iloc[-1], data.index[-1]
|
| 515 |
predicted_price, change_pct, confidence = final_prediction['predicted_price'], final_prediction['change_pct'], final_prediction['confidence']
|
| 516 |
+
|
| 517 |
+
prediction_paths = {1: [1], 5: [1, 5], 10: [1, 5, 10], 20: [1, 10, 20], 60: [1, 10, 20, 60]}
|
| 518 |
intervals_to_predict = prediction_paths.get(predict_days, [predict_days])
|
| 519 |
prediction_dates, prediction_prices = [last_date], [current_price]
|
| 520 |
+
|
| 521 |
for days in intervals_to_predict:
|
| 522 |
+
# === 迴圈內也使用統一的預測控制器 ===
|
| 523 |
interim_prediction = get_prediction(data, days)
|
| 524 |
if interim_prediction:
|
| 525 |
prediction_dates.append(last_date + timedelta(days=days))
|
| 526 |
prediction_prices.append(interim_prediction['predicted_price'])
|
| 527 |
+
|
| 528 |
+
# (後續繪圖邏輯不變)
|
| 529 |
color, arrow = ('red', '📈') if change_pct >= 0 else ('green', '📉')
|
| 530 |
result_card = html.Div([
|
| 531 |
html.H4(f"{predict_days}日後預測結果", style={'margin': '0 0 15px 0', 'color': 'white'}),
|
|
|
|
| 573 |
dash.dependencies.Input('period-dropdown', 'value'),
|
| 574 |
dash.dependencies.Input('chart-type', 'value')]
|
| 575 |
)
|
| 576 |
+
# 修正後的 update_price_chart callback 函數的相關部分
|
| 577 |
+
def update_price_chart_fixed(selected_stock, period, chart_type):
|
| 578 |
data = get_stock_data(selected_stock, period)
|
| 579 |
+
if data.empty:
|
| 580 |
+
return {}
|
| 581 |
+
|
| 582 |
data = calculate_technical_indicators(data)
|
| 583 |
stock_name = [name for name, symbol in TAIWAN_STOCKS.items() if symbol == selected_stock][0]
|
| 584 |
+
|
| 585 |
+
fig = make_subplots(rows=1, cols=2, shared_yaxes=True,
|
| 586 |
+
column_widths=[0.8, 0.2], horizontal_spacing=0.01)
|
| 587 |
+
|
| 588 |
if chart_type == 'candlestick':
|
| 589 |
+
fig.add_trace(go.Candlestick(
|
| 590 |
+
x=data.index,
|
| 591 |
+
open=data['Open'],
|
| 592 |
+
high=data['High'],
|
| 593 |
+
low=data['Low'],
|
| 594 |
+
close=data['Close'],
|
| 595 |
+
name=stock_name,
|
| 596 |
+
increasing_line_color='red',
|
| 597 |
+
decreasing_line_color='green'
|
| 598 |
+
), row=1, col=1)
|
| 599 |
else:
|
| 600 |
+
fig.add_trace(go.Scatter(
|
| 601 |
+
x=data.index,
|
| 602 |
+
y=data['Close'],
|
| 603 |
+
mode='lines',
|
| 604 |
+
name=stock_name
|
| 605 |
+
), row=1, col=1)
|
| 606 |
+
|
| 607 |
+
# 添加移動平均線
|
| 608 |
+
fig.add_trace(go.Scatter(
|
| 609 |
+
x=data.index,
|
| 610 |
+
y=data['MA5'],
|
| 611 |
+
mode='lines',
|
| 612 |
+
name='MA5',
|
| 613 |
+
line=dict(color='orange')
|
| 614 |
+
), row=1, col=1)
|
| 615 |
+
|
| 616 |
+
fig.add_trace(go.Scatter(
|
| 617 |
+
x=data.index,
|
| 618 |
+
y=data['MA20'],
|
| 619 |
+
mode='lines',
|
| 620 |
+
name='MA20',
|
| 621 |
+
line=dict(color='blue')
|
| 622 |
+
), row=1, col=1)
|
| 623 |
+
|
| 624 |
+
# 修正後的 Volume Profile 計算
|
| 625 |
bin_edges, volume_per_bin, price_centers = calculate_volume_profile(data, num_bins=50)
|
| 626 |
+
|
| 627 |
+
if volume_per_bin is not None and price_centers is not None:
|
| 628 |
+
fig.add_trace(go.Bar(
|
| 629 |
+
orientation='h',
|
| 630 |
+
y=price_centers,
|
| 631 |
+
x=volume_per_bin,
|
| 632 |
+
name='Volume Profile',
|
| 633 |
+
text=[f'{vol/1000:.0f}k' for vol in volume_per_bin],
|
| 634 |
+
textposition='auto',
|
| 635 |
+
marker=dict(
|
| 636 |
+
color='rgba(173, 216, 230, 0.6)',
|
| 637 |
+
line=dict(color='rgba(30, 144, 255, 0.8)', width=1)
|
| 638 |
+
)
|
| 639 |
+
), row=1, col=2)
|
| 640 |
+
|
| 641 |
+
fig.update_layout(
|
| 642 |
+
title_text=f'{stock_name} 股價走勢與成交量分佈',
|
| 643 |
+
height=500,
|
| 644 |
+
showlegend=True,
|
| 645 |
+
xaxis1=dict(title='日期', type='date', rangeslider_visible=False),
|
| 646 |
+
yaxis1=dict(title='價格 (TWD)'),
|
| 647 |
+
xaxis2=dict(title='成交量', showticklabels=True),
|
| 648 |
+
yaxis2=dict(showticklabels=False),
|
| 649 |
+
bargap=0.05
|
| 650 |
+
)
|
| 651 |
+
|
| 652 |
return fig
|
| 653 |
|
| 654 |
@app.callback(
|
|
|
|
| 734 |
data = get_stock_data(symbol, '1mo')
|
| 735 |
if not data.empty and len(data) > 1:
|
| 736 |
return_pct = ((data['Close'].iloc[-1] - data['Close'].iloc[0]) / data['Close'].iloc[0]) * 100
|
| 737 |
+
performance_data.append({
|
| 738 |
+
'股票': name,
|
| 739 |
+
'代碼': symbol,
|
| 740 |
+
'月報酬率(%)': return_pct,
|
| 741 |
+
'絕對波動': abs(return_pct)
|
| 742 |
+
})
|
| 743 |
if not performance_data:
|
| 744 |
fig = go.Figure().add_annotation(text="無法計算產業資料", showarrow=False)
|
| 745 |
fig.update_layout(title="近一月市場波動最大標的", height=400)
|
| 746 |
return fig
|
| 747 |
df_performance = pd.DataFrame(performance_data)
|
| 748 |
df_top_movers = df_performance.sort_values(by='絕對波動', ascending=False).head(10)
|
| 749 |
+
fig = px.pie(
|
| 750 |
+
df_top_movers,
|
| 751 |
+
values='絕對波動',
|
| 752 |
+
names='股票',
|
| 753 |
+
title='近一月市場波動最大 Top 10 標的',
|
| 754 |
+
hover_data={'月報酬率(%)': ':.2f'}
|
| 755 |
+
)
|
| 756 |
+
fig.update_traces(
|
| 757 |
+
textposition='inside',
|
| 758 |
+
textinfo='percent+label',
|
| 759 |
+
hovertemplate="<b>%{label}</b><br>月報酬率: %{customdata[0]:.2f}%<extra></extra>"
|
| 760 |
+
)
|
| 761 |
fig.update_layout(height=400, showlegend=False)
|
| 762 |
return fig
|
| 763 |
|
|
|
|
| 795 |
def update_analysis_text(selected_stock, period):
|
| 796 |
cache_key = f"{selected_stock}-{period}"
|
| 797 |
current_time = time.time()
|
| 798 |
+
|
| 799 |
if cache_key in ANALYSIS_CACHE:
|
| 800 |
cached_data = ANALYSIS_CACHE[cache_key]
|
| 801 |
if current_time - cached_data['timestamp'] < CACHE_DURATION_SECONDS:
|
| 802 |
print(f"從快取載入分析: {cache_key}")
|
| 803 |
return cached_data['technical'], cached_data['fundamental'], cached_data['outlook']
|
| 804 |
+
|
| 805 |
+
print(f"重新生成分析: {cache_key}")
|
| 806 |
data = get_stock_data(selected_stock, period)
|
| 807 |
stock_name = [name for name, symbol in TAIWAN_STOCKS.items() if symbol == selected_stock][0]
|
| 808 |
if data.empty or len(data) < 20:
|
| 809 |
return "資料不足,無法分析", "資料不足,無法分析", "資料不足,無法分析"
|
| 810 |
+
|
| 811 |
data = calculate_technical_indicators(data)
|
| 812 |
+
|
| 813 |
price_change = ((data['Close'].iloc[-1] - data['Close'].iloc[0]) / data['Close'].iloc[0]) * 100
|
| 814 |
rsi_current = data['RSI'].iloc[-1] if not pd.isna(data['RSI'].iloc[-1]) else 50
|
| 815 |
macd_current = data['MACD'].iloc[-1] if not pd.isna(data['MACD'].iloc[-1]) else 0
|
| 816 |
macd_signal_current = data['MACD_Signal'].iloc[-1] if not pd.isna(data['MACD_Signal'].iloc[-1]) else 0
|
| 817 |
+
|
| 818 |
technical_text = html.Div([
|
| 819 |
html.P([html.Strong("價格趨勢:"), f"在最近 {period} 期間內,{stock_name} 股價呈現", html.Span(f"{'上漲' if price_change > 5 else '下跌' if price_change < -5 else '盤整'}", style={'color': 'red' if price_change > 5 else 'green' if price_change < -5 else 'orange', 'font-weight': 'bold'}), f"走勢,累計變動 {price_change:+.1f}%。"]),
|
| 820 |
html.P([html.Strong("RSI 指標:"), f"目前的 RSI 值為 {rsi_current:.1f},", html.Span("處於超買區(>70)" if rsi_current > 70 else "處於超賣區(<30)" if rsi_current < 30 else "在正常範圍內", style={'color': 'green' if rsi_current > 70 else 'red' if rsi_current < 30 else 'blue', 'font-weight': 'bold'}), "。"]),
|
| 821 |
html.P([html.Strong("MACD 指標:"), f"MACD 快線 ({macd_current:.3f}) 目前", html.Span("高於" if macd_current > macd_signal_current else "低於", style={'color': 'red' if macd_current > macd_signal_current else 'green', 'font-weight': 'bold'}), f" Signal 慢線 ({macd_signal_current:.3f}),", f"顯示市場動能偏向{'多頭' if macd_current > macd_signal_current else '空頭'}。"]),
|
| 822 |
])
|
| 823 |
+
|
| 824 |
fundamental_text, market_outlook_text = generate_gemini_analysis(stock_name, selected_stock, period, data)
|
| 825 |
+
|
| 826 |
+
ANALYSIS_CACHE[cache_key] = {
|
| 827 |
+
'technical': technical_text,
|
| 828 |
+
'fundamental': fundamental_text,
|
| 829 |
+
'outlook': market_outlook_text,
|
| 830 |
+
'timestamp': current_time
|
| 831 |
+
}
|
| 832 |
+
|
| 833 |
return technical_text, fundamental_text, market_outlook_text
|
| 834 |
|
| 835 |
@app.callback(
|
|
|
|
| 850 |
return fig
|
| 851 |
|
| 852 |
def summarize_news_with_gemini(news_list: list) -> str:
|
| 853 |
+
"""
|
| 854 |
+
使用 Gemini API 將英文新聞標題列表摘要成一段繁體中文。
|
| 855 |
+
"""
|
| 856 |
api_key = os.getenv("GEMINI_API_KEY")
|
| 857 |
if not api_key:
|
| 858 |
return "錯誤:找不到 GEMINI_API_KEY。請在 Hugging Face Secrets 中設定。"
|
| 859 |
+
|
| 860 |
try:
|
| 861 |
genai.configure(api_key=api_key)
|
| 862 |
model = genai.GenerativeModel('gemini-1.5-flash')
|
| 863 |
+
|
| 864 |
formatted_news = "\n".join([f"- {news}" for news in news_list])
|
| 865 |
+
|
| 866 |
prompt = f"""
|
| 867 |
+
請扮演一位專業的金融市場分析師。
|
| 868 |
+
以下是幾則最新的英文財經新聞標題,請將它們整合成一段簡潔、流暢、約 200 字的繁體中文市場動態摘要,與利多哪些產業,利空哪些產業。
|
| 869 |
+
提供3段重點,
|
| 870 |
+
請專注於可能影響市場情緒和股價的關鍵資訊,並直接提供摘要內容,不要包含任何額外的問候語或說明。
|
| 871 |
+
|
| 872 |
英文新聞標題如下:
|
| 873 |
{formatted_news}
|
| 874 |
"""
|
| 875 |
+
|
| 876 |
response = model.generate_content(prompt)
|
| 877 |
return response.text
|
| 878 |
+
|
| 879 |
except Exception as e:
|
| 880 |
print(f"呼叫 Gemini API 時發生錯誤: {e}")
|
| 881 |
return f"無法生成新聞摘要,請稍後再試。錯誤訊息:{e}"
|
|
|
|
| 900 |
normalized_prices = (data['Close'] / data['Close'].iloc[0]) * 100
|
| 901 |
fig.add_trace(go.Scatter(x=data.index, y=normalized_prices, mode='lines', name=stock_name, line=dict(width=2)))
|
| 902 |
total_return = ((data['Close'].iloc[-1] / data['Close'].iloc[0]) - 1) * 100
|
| 903 |
+
volatility = data['Close'].pct_change().std() * np.sqrt(252) * 100
|
| 904 |
comparison_data.append({'name': stock_name, 'return': total_return, 'volatility': volatility, 'current_price': data['Close'].iloc[-1]})
|
| 905 |
fig.update_layout(title=f'股票績效比較 - {period}', xaxis_title='日期', yaxis_title='相對績效 (基期=100)', height=400, hovermode='x unified')
|
| 906 |
if comparison_data:
|
|
|
|
| 922 |
error_fig = go.Figure().add_annotation(text="情緒指標模型載入失敗", showarrow=False)
|
| 923 |
error_fig.update_layout(height=200)
|
| 924 |
return dcc.Graph(figure=error_fig), html.P("新聞分析模型載入失敗,請檢查後台日誌。")
|
| 925 |
+
|
| 926 |
sentiment_score_raw = predictor.get_news_index()
|
| 927 |
+
|
| 928 |
if sentiment_score_raw is not None:
|
| 929 |
sentiment_score_normalized = (sentiment_score_raw + 1) * 50
|
| 930 |
sentiment_score_normalized = max(0, min(100, sentiment_score_normalized))
|
|
|
|
| 949 |
error_fig = go.Figure().add_annotation(text="今日尚無情緒分數", showarrow=False)
|
| 950 |
error_fig.update_layout(height=200)
|
| 951 |
gauge_content = dcc.Graph(figure=error_fig)
|
| 952 |
+
|
| 953 |
top_news_list = predictor.get_news()
|
| 954 |
news_content = None
|
| 955 |
+
|
| 956 |
if top_news_list and isinstance(top_news_list, list):
|
| 957 |
summary_text = summarize_news_with_gemini(top_news_list)
|
| 958 |
+
news_content = dcc.Markdown(summary_text, style={
|
| 959 |
+
'margin': '8px 0', 'padding-left': '5px',
|
| 960 |
+
'font-size': '15px', 'line-height': '1.7'
|
| 961 |
+
})
|
| 962 |
elif top_news_list == []:
|
| 963 |
news_content = html.P("昨日無重大相關新聞。", style={'text-align': 'center', 'padding-top': '50px'})
|
| 964 |
else:
|
| 965 |
news_content = html.P("讀取新聞時發生錯誤。", style={'text-align': 'center', 'padding-top': '50px'})
|
| 966 |
+
|
| 967 |
return gauge_content, news_content
|
| 968 |
|
| 969 |
# 主程式執行
|