ProStock/src/experiment/regression.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. 导入依赖"
   ]
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-08T03:32:40.563481Z",
     "start_time": "2026-03-08T03:32:39.994884Z"
    }
   },
   "source": [
    "import os\n",
    "from datetime import datetime\n",
    "from typing import List\n",
    "\n",
    "import polars as pl\n",
    "\n",
    "from src.factors import FactorEngine\n",
    "from src.training import (\n",
    "    DateSplitter,\n",
    "    LightGBMModel,\n",
    "    STFilter,\n",
    "    StandardScaler,\n",
    "    StockFilterConfig,\n",
    "    StockPoolManager,\n",
    "    Trainer,\n",
    "    Winsorizer,\n",
    "    NullFiller,\n",
    ")\n",
    "from src.training.config import TrainingConfig"
   ],
   "outputs": [],
   "execution_count": 1
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. 定义辅助函数"
   ]
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-08T03:32:40.574743Z",
     "start_time": "2026-03-08T03:32:40.571165Z"
    }
   },
   "cell_type": "code",
   "source": [
    "def create_factors_with_strings(engine: FactorEngine, factor_definitions: dict, label_factor: dict) -> List[str]:\n",
    "    print(\"=\" * 80)\n",
    "    print(\"使用字符串表达式定义因子\")\n",
    "    print(\"=\" * 80)\n",
    "\n",
    "    # 注册所有特征因子\n",
    "    print(\"\\n注册特征因子:\")\n",
    "    for name, expr in factor_definitions.items():\n",
    "        engine.add_factor(name, expr)\n",
    "        print(f\"  - {name}: {expr}\")\n",
    "\n",
    "    # 注册 label 因子\n",
    "    print(\"\\n注册 Label 因子:\")\n",
    "    for name, expr in label_factor.items():\n",
    "        engine.add_factor(name, expr)\n",
    "        print(f\"  - {name}: {expr}\")\n",
    "\n",
    "    # 从字典自动获取特征列\n",
    "    feature_cols = list(factor_definitions.keys())\n",
    "\n",
    "    print(f\"\\n特征因子数: {len(feature_cols)}\")\n",
    "    print(f\"Label: {list(label_factor.keys())[0]}\")\n",
    "    print(f\"已注册因子总数: {len(engine.list_registered())}\")\n",
    "\n",
    "    return feature_cols\n",
    "\n",
    "\n",
    "def prepare_data(\n",
    "    engine: FactorEngine,\n",
    "    feature_cols: List[str],\n",
    "    start_date: str,\n",
    "    end_date: str,\n",
    ") -> pl.DataFrame:\n",
    "    print(\"\\n\" + \"=\" * 80)\n",
    "    print(\"准备数据\")\n",
    "    print(\"=\" * 80)\n",
    "\n",
    "    # 计算因子（全市场数据）\n",
    "    print(f\"\\n计算因子: {start_date} - {end_date}\")\n",
    "    factor_names = feature_cols + [\"return_5\"]  # 包含 label\n",
    "\n",
    "    data = engine.compute(\n",
    "        factor_names=factor_names,\n",
    "        start_date=start_date,\n",
    "        end_date=end_date,\n",
    "    )\n",
    "\n",
    "    print(f\"数据形状: {data.shape}\")\n",
    "    print(f\"数据列: {data.columns}\")\n",
    "    print(f\"\\n前5行预览:\")\n",
    "    print(data.head())\n",
    "\n",
    "    return data"
   ],
   "outputs": [],
   "execution_count": 2
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. 配置参数\n",
    "\n",
    "### 3.1 因子定义"
   ]
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-08T03:32:40.585729Z",
     "start_time": "2026-03-08T03:32:40.579837Z"
    }
   },
   "cell_type": "code",
   "source": [
    "# 特征因子定义字典：新增因子只需在此处添加一行\n",
    "FACTOR_DEFINITIONS = {\n",
    "    # 1. 价格动量因子\n",
    "    \"ma5\": \"ts_mean(close, 5)\",\n",
    "    \"ma10\": \"ts_mean(close, 10)\",\n",
    "    \"ma20\": \"ts_mean(close, 20)\",\n",
    "    \"ma_ratio\": \"ts_mean(close, 5) / ts_mean(close, 20) - 1\",\n",
    "    # 2. 波动率因子\n",
    "    \"volatility_5\": \"ts_std(close, 5)\",\n",
    "    \"volatility_20\": \"ts_std(close, 20)\",\n",
    "    \"vol_ratio\": \"ts_std(close, 5) / (ts_std(close, 20) + 1e-8)\",\n",
    "    # 3. 收益率动量因子\n",
    "    \"return_10\": \"(close / ts_delay(close, 10)) - 1\",\n",
    "    \"return_20\": \"(close / ts_delay(close, 20)) - 1\",\n",
    "    # 4. 收益率变化因子\n",
    "    \"return_diff\": \"(close / ts_delay(close, 5)) - 1 - ((close / ts_delay(close, 10)) - 1)\",\n",
    "    # 5. 成交量因子\n",
    "    \"vol_ma5\": \"ts_mean(vol, 5)\",\n",
    "    \"vol_ma20\": \"ts_mean(vol, 20)\",\n",
    "    \"vol_ratio\": \"ts_mean(vol, 5) / (ts_mean(vol, 20) + 1e-8)\",\n",
    "    # 6. 市值因子（截面排名）\n",
    "    \"market_cap_rank\": \"cs_rank(total_mv)\",\n",
    "    # 7. 价格位置因子\n",
    "    \"high_low_ratio\": \"(close - ts_min(low, 20)) / (ts_max(high, 20) - ts_min(low, 20) + 1e-8)\",\n",
    "    \"n_income_rank\": \"cs_rank(n_income)\",  # 净利润截面排名\n",
    "    # 8. 财务数据因子（来自利润表 financial_income）\n",
    "    \"operate_profit_rank\": \"cs_rank(operate_profit)\",  # 营业利润截面排名\n",
    "    \"total_profit_rank\": \"cs_rank(total_profit)\",  # 利润总额截面排名\n",
    "    \"ebit_rank\": \"cs_rank(ebit)\",  # 息税前利润截面排名\n",
    "    \"ebitda_rank\": \"cs_rank(ebitda)\",  # 息税折旧摊销前利润截面排名\n",
    "    # 9. 财务数据因子（来自资产负债表 financial_balance）\n",
    "    \"total_liab_rank\": \"cs_rank(total_liab)\",  # 总负债截面排名\n",
    "    \"money_cap_rank\": \"cs_rank(money_cap)\",  # 货币资金截面排名\n",
    "    # 10. 财务数据因子（来自现金流量表 financial_cashflow）\n",
    "    \"n_cashflow_act_rank\": \"cs_rank(n_cashflow_act)\",  # 经营活动现金流净额截面排名\n",
    "    # 11. 财务估值因子\n",
    "    \"profit_to_market_cap\": \"n_income / (total_mv + 1e-8)\",  # 净利润率（净利润/市值）\n",
    "    \"cashflow_to_market_cap\": \"n_cashflow_act / (total_mv + 1e-8)\",  # 经营现金流/市值\n",
    "    \"operate_profit_to_market_cap\": \"operate_profit / (total_mv + 1e-8)\",  # 营业利润/市值\n",
    "}\n",
    "\n",
    "# Label 因子定义（不参与训练，用于计算目标）\n",
    "LABEL_FACTOR = {\n",
    "    \"return_5\": \"(ts_delay(close, -5) / close) - 1\",  # 未来5日收益率\n",
    "}"
   ],
   "outputs": [],
   "execution_count": 3
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.2 训练参数配置"
   ]
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-08T03:32:40.593766Z",
     "start_time": "2026-03-08T03:32:40.590001Z"
    }
   },
   "source": [
    "# 日期范围配置（正确的 train/val/test 三分法）\n",
    "# Train: 用于训练模型参数\n",
    "# Val: 用于验证/早停/调参（位于 train 之后，test 之前）\n",
    "# Test: 仅用于最终评估，完全独立于训练过程\n",
    "TRAIN_START = \"20200101\"\n",
    "TRAIN_END = \"20231231\"\n",
    "VAL_START = \"20240101\"\n",
    "VAL_END = \"20241231\"\n",
    "TEST_START = \"20250101\"\n",
    "TEST_END = \"20251231\"\n",
    "\n",
    "# 模型参数配置\n",
    "MODEL_PARAMS = {\n",
    "    \"objective\": \"regression\",\n",
    "    \"metric\": \"mae\",  # 改为 MAE，对异常值更稳健\n",
    "    # 树结构控制（防过拟合核心）\n",
    "    \"num_leaves\": 20,  # 从31降为20，降低模型复杂度\n",
    "    \"max_depth\": 4,  # 显式限制深度，防止过度拟合噪声\n",
    "    \"min_child_samples\": 50,  # 叶子最小样本数，防止学习极端样本\n",
    "    \"min_child_weight\": 0.001,\n",
    "    # 学习参数\n",
    "    \"learning_rate\": 0.01,  # 降低学习率，配合更多树\n",
    "    \"n_estimators\": 1000,  # 增加树数量，配合早停\n",
    "    # 采样策略（关键防过拟合）\n",
    "    \"subsample\": 0.8,  # 每棵树随机采样80%数据（行采样）\n",
    "    \"subsample_freq\": 5,  # 每5轮迭代进行一次 subsample\n",
    "    \"colsample_bytree\": 0.8,  # 每棵树随机选择80%特征（列采样）\n",
    "    # 正则化\n",
    "    \"reg_alpha\": 0.1,  # L1正则，增加稀疏性\n",
    "    \"reg_lambda\": 1.0,  # L2正则，平滑权重\n",
    "    # 数值稳定性\n",
    "    \"verbose\": -1,\n",
    "    \"random_state\": 42,\n",
    "}\n",
    "\n",
    "# 数据处理器配置\n",
    "PROCESSOR_CONFIGS = [\n",
    "    {\"name\": \"winsorizer\", \"params\": {\"lower\": 0.01, \"upper\": 0.99}},\n",
    "    {\"name\": \"cs_standard_scaler\", \"params\": {}},\n",
    "]\n",
    "\n",
    "# 股票池筛选配置\n",
    "STOCK_FILTER_CONFIG = {\n",
    "    \"exclude_cyb\": True,  # 排除创业板\n",
    "    \"exclude_kcb\": True,  # 排除科创板\n",
    "    \"exclude_bj\": True,  # 排除北交所\n",
    "    \"exclude_st\": True,  # 排除ST股票\n",
    "}\n",
    "\n",
    "# 输出配置（相对于本文件所在目录）\n",
    "OUTPUT_DIR = \"output\"\n",
    "SAVE_PREDICTIONS = True\n",
    "PERSIST_MODEL = False\n",
    "\n",
    "# Top N 配置：每日推荐股票数量\n",
    "TOP_N = 2  # 可调整为 10, 20 等"
   ],
   "outputs": [],
   "execution_count": 4
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. 训练流程\n",
    "\n",
    "### 4.1 初始化组件"
   ]
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-08T03:32:46.956583Z",
     "start_time": "2026-03-08T03:32:40.597894Z"
    }
   },
   "source": [
    "print(\"\\n\" + \"=\" * 80)\n",
    "print(\"LightGBM 回归模型训练\")\n",
    "print(\"=\" * 80)\n",
    "\n",
    "# 1. 创建 FactorEngine\n",
    "print(\"\\n[1] 创建 FactorEngine\")\n",
    "engine = FactorEngine()\n",
    "\n",
    "# 2. 使用字符串表达式定义因子\n",
    "print(\"\\n[2] 定义因子（字符串表达式）\")\n",
    "feature_cols = create_factors_with_strings(engine, FACTOR_DEFINITIONS, LABEL_FACTOR)\n",
    "target_col = \"return_5\"\n",
    "\n",
    "# 3. 准备数据（使用模块级别的日期配置）\n",
    "print(\"\\n[3] 准备数据\")\n",
    "\n",
    "data = prepare_data(\n",
    "    engine=engine,\n",
    "    feature_cols=feature_cols,\n",
    "    start_date=TRAIN_START,\n",
    "    end_date=TEST_END,\n",
    ")\n",
    "\n",
    "# 4. 打印配置信息\n",
    "print(f\"\\n[配置] 训练期: {TRAIN_START} - {TRAIN_END}\")\n",
    "print(f\"[配置] 验证期: {VAL_START} - {VAL_END}\")\n",
    "print(f\"[配置] 测试期: {TEST_START} - {TEST_END}\")\n",
    "print(f\"[配置] 特征数: {len(feature_cols)}\")\n",
    "print(f\"[配置] 目标变量: {target_col}\")\n",
    "\n",
    "# 5. 创建模型\n",
    "model = LightGBMModel(params=MODEL_PARAMS)\n",
    "\n",
    "# 6. 创建数据处理器\n",
    "processors = [\n",
    "    NullFiller(strategy=\"mean\"),\n",
    "    Winsorizer(**PROCESSOR_CONFIGS[0][\"params\"]),\n",
    "    StandardScaler(exclude_cols=[\"ts_code\", \"trade_date\", target_col]),\n",
    "]\n",
    "\n",
    "# 7. 创建数据划分器（正确的 train/val/test 三分法）\n",
    "# Train: 训练模型参数 | Val: 验证/早停 | Test: 最终评估\n",
    "splitter = DateSplitter(\n",
    "    train_start=TRAIN_START,\n",
    "    train_end=TRAIN_END,\n",
    "    val_start=VAL_START,\n",
    "    val_end=VAL_END,\n",
    "    test_start=TEST_START,\n",
    "    test_end=TEST_END,\n",
    ")\n",
    "\n",
    "# 8. 创建股票池管理器\n",
    "pool_manager = StockPoolManager(\n",
    "    filter_config=StockFilterConfig(**STOCK_FILTER_CONFIG),\n",
    "    selector_config=None,  # 暂时不启用市值选择\n",
    "    data_router=engine.router,\n",
    ")\n",
    "\n",
    "# 9. 创建 ST 股票过滤器\n",
    "st_filter = STFilter(\n",
    "    data_router=engine.router,\n",
    ")\n",
    "\n",
    "# 10. 创建训练器\n",
    "trainer = Trainer(\n",
    "    model=model,\n",
    "    pool_manager=pool_manager,\n",
    "    processors=processors,\n",
    "    filters=[st_filter],\n",
    "    splitter=splitter,\n",
    "    target_col=target_col,\n",
    "    feature_cols=feature_cols,\n",
    "    persist_model=PERSIST_MODEL,\n",
    ")"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "================================================================================\n",
      "LightGBM 回归模型训练\n",
      "================================================================================\n",
      "\n",
      "[1] 创建 FactorEngine\n",
      "\n",
      "[2] 定义因子（字符串表达式）\n",
      "================================================================================\n",
      "使用字符串表达式定义因子\n",
      "================================================================================\n",
      "\n",
      "注册特征因子:\n",
      "  - ma5: ts_mean(close, 5)\n",
      "  - ma10: ts_mean(close, 10)\n",
      "  - ma20: ts_mean(close, 20)\n",
      "  - ma_ratio: ts_mean(close, 5) / ts_mean(close, 20) - 1\n",
      "  - volatility_5: ts_std(close, 5)\n",
      "  - volatility_20: ts_std(close, 20)\n",
      "  - vol_ratio: ts_mean(vol, 5) / (ts_mean(vol, 20) + 1e-8)\n",
      "  - return_10: (close / ts_delay(close, 10)) - 1\n",
      "  - return_20: (close / ts_delay(close, 20)) - 1\n",
      "  - return_diff: (close / ts_delay(close, 5)) - 1 - ((close / ts_delay(close, 10)) - 1)\n",
      "  - vol_ma5: ts_mean(vol, 5)\n",
      "  - vol_ma20: ts_mean(vol, 20)\n",
      "  - market_cap_rank: cs_rank(total_mv)\n",
      "  - high_low_ratio: (close - ts_min(low, 20)) / (ts_max(high, 20) - ts_min(low, 20) + 1e-8)\n",
      "  - n_income_rank: cs_rank(n_income)\n",
      "  - operate_profit_rank: cs_rank(operate_profit)\n",
      "  - total_profit_rank: cs_rank(total_profit)\n",
      "  - ebit_rank: cs_rank(ebit)\n",
      "  - ebitda_rank: cs_rank(ebitda)\n",
      "  - total_liab_rank: cs_rank(total_liab)\n",
      "  - money_cap_rank: cs_rank(money_cap)\n",
      "  - n_cashflow_act_rank: cs_rank(n_cashflow_act)\n",
      "  - profit_to_market_cap: n_income / (total_mv + 1e-8)\n",
      "  - cashflow_to_market_cap: n_cashflow_act / (total_mv + 1e-8)\n",
      "  - operate_profit_to_market_cap: operate_profit / (total_mv + 1e-8)\n",
      "\n",
      "注册 Label 因子:\n",
      "  - return_5: (ts_delay(close, -5) / close) - 1\n",
      "\n",
      "特征因子数: 25\n",
      "Label: return_5\n",
      "已注册因子总数: 26\n",
      "\n",
      "[3] 准备数据\n",
      "\n",
      "================================================================================\n",
      "准备数据\n",
      "================================================================================\n",
      "\n",
      "计算因子: 20200101 - 20251231\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "D:\\PyProject\\ProStock\\src\\data\\financial_loader.py:123: UserWarning: Sortedness of columns cannot be checked when 'by' groups provided\n",
      "  merged = df_price.join_asof(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "数据形状: (7044952, 42)\n",
      "数据列: ['ts_code', 'trade_date', 'vol', 'close', 'low', 'high', 'total_mv', 'f_ann_date', 'n_income', 'ebitda', 'total_profit', 'ebit', 'operate_profit', 'money_cap', 'total_liab', 'n_cashflow_act', 'ma5', 'ma10', 'ma20', 'ma_ratio', 'volatility_5', 'volatility_20', 'vol_ratio', 'return_10', 'return_20', 'return_diff', 'vol_ma5', 'vol_ma20', 'market_cap_rank', 'high_low_ratio', 'n_income_rank', 'operate_profit_rank', 'total_profit_rank', 'ebit_rank', 'ebitda_rank', 'total_liab_rank', 'money_cap_rank', 'n_cashflow_act_rank', 'profit_to_market_cap', 'cashflow_to_market_cap', 'operate_profit_to_market_cap', 'return_5']\n",
      "\n",
      "前5行预览:\n",
      "shape: (5, 42)\n",
      "┌───────────┬────────────┬───────────┬─────────┬───┬───────────┬───────────┬───────────┬───────────┐\n",
      "│ ts_code   ┆ trade_date ┆ vol       ┆ close   ┆ … ┆ profit_to ┆ cashflow_ ┆ operate_p ┆ return_5  │\n",
      "│ ---       ┆ ---        ┆ ---       ┆ ---     ┆   ┆ _market_c ┆ to_market ┆ rofit_to_ ┆ ---       │\n",
      "│ str       ┆ str        ┆ f64       ┆ f64     ┆   ┆ ap        ┆ _cap      ┆ market_ca ┆ f64       │\n",
      "│           ┆            ┆           ┆         ┆   ┆ ---       ┆ ---       ┆ p         ┆           │\n",
      "│           ┆            ┆           ┆         ┆   ┆ f64       ┆ f64       ┆ ---       ┆           │\n",
      "│           ┆            ┆           ┆         ┆   ┆           ┆           ┆ f64       ┆           │\n",
      "╞═══════════╪════════════╪═══════════╪═════════╪═══╪═══════════╪═══════════╪═══════════╪═══════════╡\n",
      "│ 000001.SZ ┆ 20200102   ┆ 1.5302e6  ┆ 1841.69 ┆ … ┆ 721.52104 ┆ 2580.5045 ┆ 938.15146 ┆ -0.004746 │\n",
      "│           ┆            ┆           ┆         ┆   ┆ 1         ┆ 33        ┆ 4         ┆           │\n",
      "│ 000001.SZ ┆ 20200103   ┆ 1.1162e6  ┆ 1875.53 ┆ … ┆ 708.50174 ┆ 2533.9412 ┆ 921.22323 ┆ -0.02852  │\n",
      "│           ┆            ┆           ┆         ┆   ┆ 4         ┆ 96        ┆ 7         ┆           │\n",
      "│ 000001.SZ ┆ 20200106   ┆ 862083.5  ┆ 1863.52 ┆ … ┆ 713.06736 ┆ 2550.2701 ┆ 927.15964 ┆ -0.004685 │\n",
      "│           ┆            ┆           ┆         ┆   ┆ 8         ┆ 5         ┆ 9         ┆           │\n",
      "│ 000001.SZ ┆ 20200107   ┆ 728607.56 ┆ 1872.26 ┆ … ┆ 709.74110 ┆ 2538.3738 ┆ 922.83470 ┆ -0.022743 │\n",
      "│           ┆            ┆           ┆         ┆   ┆ 6         ┆ 46        ┆ 6         ┆           │\n",
      "│ 000001.SZ ┆ 20200108   ┆ 847824.12 ┆ 1818.76 ┆ … ┆ 730.61584 ┆ 2613.0319 ┆ 949.97690 ┆ -0.008401 │\n",
      "│           ┆            ┆           ┆         ┆   ┆ 4         ┆ 01        ┆ 3         ┆           │\n",
      "└───────────┴────────────┴───────────┴─────────┴───┴───────────┴───────────┴───────────┴───────────┘\n",
      "\n",
      "[配置] 训练期: 20200101 - 20231231\n",
      "[配置] 验证期: 20240101 - 20241231\n",
      "[配置] 测试期: 20250101 - 20251231\n",
      "[配置] 特征数: 25\n",
      "[配置] 目标变量: return_5\n"
     ]
    }
   ],
   "execution_count": 5
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.2 执行训练"
   ]
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-08T03:32:54.939339Z",
     "start_time": "2026-03-08T03:32:46.966714Z"
    }
   },
   "source": [
    "print(\"\\n\" + \"=\" * 80)\n",
    "print(\"开始训练\")\n",
    "print(\"=\" * 80)\n",
    "\n",
    "# 步骤 1: 股票池筛选\n",
    "print(\"\\n[步骤 1/6] 股票池筛选\")\n",
    "print(\"-\" * 60)\n",
    "if pool_manager:\n",
    "    print(\"  执行每日独立筛选股票池...\")\n",
    "    filtered_data = pool_manager.filter_and_select_daily(data)\n",
    "    print(f\"  筛选前数据规模: {data.shape}\")\n",
    "    print(f\"  筛选后数据规模: {filtered_data.shape}\")\n",
    "    print(f\"  筛选前股票数: {data['ts_code'].n_unique()}\")\n",
    "    print(f\"  筛选后股票数: {filtered_data['ts_code'].n_unique()}\")\n",
    "    print(f\"  删除记录数: {len(data) - len(filtered_data)}\")\n",
    "else:\n",
    "    filtered_data = data\n",
    "    print(\"  未配置股票池管理器，跳过筛选\")"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "================================================================================\n",
      "开始训练\n",
      "================================================================================\n",
      "\n",
      "[步骤 1/6] 股票池筛选\n",
      "------------------------------------------------------------\n",
      "  执行每日独立筛选股票池...\n",
      "  筛选前数据规模: (7044952, 42)\n",
      "  筛选后数据规模: (4532198, 42)\n",
      "  筛选前股票数: 5678\n",
      "  筛选后股票数: 3359\n",
      "  删除记录数: 2512754\n"
     ]
    }
   ],
   "execution_count": 6
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-08T03:32:56.279010Z",
     "start_time": "2026-03-08T03:32:54.952714Z"
    }
   },
   "source": [
    "# 步骤 2: 划分训练/验证/测试集（正确的三分法）\n",
    "print(\"\\n[步骤 2/6] 划分训练集、验证集和测试集\")\n",
    "print(\"-\" * 60)\n",
    "if splitter:\n",
    "    # 正确的三分法：train用于训练，val用于验证/早停，test仅用于最终评估\n",
    "    train_data, val_data, test_data = splitter.split(filtered_data)\n",
    "    print(f\"  训练集数据规模: {train_data.shape}\")\n",
    "    print(f\"  验证集数据规模: {val_data.shape}\")\n",
    "    print(f\"  测试集数据规模: {test_data.shape}\")\n",
    "    print(f\"  训练集股票数: {train_data['ts_code'].n_unique()}\")\n",
    "    print(f\"  验证集股票数: {val_data['ts_code'].n_unique()}\")\n",
    "    print(f\"  测试集股票数: {test_data['ts_code'].n_unique()}\")\n",
    "    print(\n",
    "        f\"  训练集日期范围: {train_data['trade_date'].min()} - {train_data['trade_date'].max()}\"\n",
    "    )\n",
    "    print(\n",
    "        f\"  验证集日期范围: {val_data['trade_date'].min()} - {val_data['trade_date'].max()}\"\n",
    "    )\n",
    "    print(\n",
    "        f\"  测试集日期范围: {test_data['trade_date'].min()} - {test_data['trade_date'].max()}\"\n",
    "    )\n",
    "\n",
    "    print(\"\\n  训练集前5行预览:\")\n",
    "    print(train_data.head())\n",
    "    print(\"\\n  验证集前5行预览:\")\n",
    "    print(val_data.head())\n",
    "    print(\"\\n  测试集前5行预览:\")\n",
    "    print(test_data.head())\n",
    "else:\n",
    "    train_data = filtered_data\n",
    "    test_data = filtered_data\n",
    "    print(\"  未配置划分器，全部作为训练集\")"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "[步骤 2/6] 划分训练集、验证集和测试集\n",
      "------------------------------------------------------------\n",
      "  训练集数据规模: (2991506, 42)\n",
      "  验证集数据规模: (769485, 42)\n",
      "  测试集数据规模: (771207, 42)\n",
      "  训练集股票数: 3297\n",
      "  验证集股票数: 3220\n",
      "  测试集股票数: 3215\n",
      "  训练集日期范围: 20200102 - 20231229\n",
      "  验证集日期范围: 20240102 - 20241231\n",
      "  测试集日期范围: 20250102 - 20251231\n",
      "\n",
      "  训练集前5行预览:\n",
      "shape: (5, 42)\n",
      "┌───────────┬────────────┬───────────┬─────────┬───┬───────────┬───────────┬───────────┬───────────┐\n",
      "│ ts_code   ┆ trade_date ┆ vol       ┆ close   ┆ … ┆ profit_to ┆ cashflow_ ┆ operate_p ┆ return_5  │\n",
      "│ ---       ┆ ---        ┆ ---       ┆ ---     ┆   ┆ _market_c ┆ to_market ┆ rofit_to_ ┆ ---       │\n",
      "│ str       ┆ str        ┆ f64       ┆ f64     ┆   ┆ ap        ┆ _cap      ┆ market_ca ┆ f64       │\n",
      "│           ┆            ┆           ┆         ┆   ┆ ---       ┆ ---       ┆ p         ┆           │\n",
      "│           ┆            ┆           ┆         ┆   ┆ f64       ┆ f64       ┆ ---       ┆           │\n",
      "│           ┆            ┆           ┆         ┆   ┆           ┆           ┆ f64       ┆           │\n",
      "╞═══════════╪════════════╪═══════════╪═════════╪═══╪═══════════╪═══════════╪═══════════╪═══════════╡\n",
      "│ 000001.SZ ┆ 20200102   ┆ 1.5302e6  ┆ 1841.69 ┆ … ┆ 721.52104 ┆ 2580.5045 ┆ 938.15146 ┆ -0.004746 │\n",
      "│           ┆            ┆           ┆         ┆   ┆ 1         ┆ 33        ┆ 4         ┆           │\n",
      "│ 000002.SZ ┆ 20200102   ┆ 1012130.4 ┆ 4832.29 ┆ … ┆ 776.91820 ┆ 47.131053 ┆ 1140.2493 ┆ -0.011057 │\n",
      "│           ┆            ┆           ┆         ┆   ┆ 1         ┆           ┆ 95        ┆           │\n",
      "│ 000004.SZ ┆ 20200102   ┆ 17853.2   ┆ 90.75   ┆ … ┆ -69.58089 ┆ -52.61755 ┆ -24.82135 ┆ -0.000441 │\n",
      "│           ┆            ┆           ┆         ┆   ┆ 5         ┆ 4         ┆ 9         ┆           │\n",
      "│ 000005.SZ ┆ 20200102   ┆ 104134.12 ┆ 29.1    ┆ … ┆ 142.55925 ┆ 385.57490 ┆ 208.12520 ┆ 0.022337  │\n",
      "│           ┆            ┆           ┆         ┆   ┆ 6         ┆ 4         ┆ 2         ┆           │\n",
      "│ 000006.SZ ┆ 20200102   ┆ 124751.76 ┆ 191.3   ┆ … ┆ 633.27582 ┆ 650.95370 ┆ 819.10495 ┆ 0.012964  │\n",
      "│           ┆            ┆           ┆         ┆   ┆           ┆ 3         ┆ 5         ┆           │\n",
      "└───────────┴────────────┴───────────┴─────────┴───┴───────────┴───────────┴───────────┴───────────┘\n",
      "\n",
      "  验证集前5行预览:\n",
      "shape: (5, 42)\n",
      "┌───────────┬────────────┬───────────┬─────────┬───┬───────────┬───────────┬───────────┬───────────┐\n",
      "│ ts_code   ┆ trade_date ┆ vol       ┆ close   ┆ … ┆ profit_to ┆ cashflow_ ┆ operate_p ┆ return_5  │\n",
      "│ ---       ┆ ---        ┆ ---       ┆ ---     ┆   ┆ _market_c ┆ to_market ┆ rofit_to_ ┆ ---       │\n",
      "│ str       ┆ str        ┆ f64       ┆ f64     ┆   ┆ ap        ┆ _cap      ┆ market_ca ┆ f64       │\n",
      "│           ┆            ┆           ┆         ┆   ┆ ---       ┆ ---       ┆ p         ┆           │\n",
      "│           ┆            ┆           ┆         ┆   ┆ f64       ┆ f64       ┆ ---       ┆           │\n",
      "│           ┆            ┆           ┆         ┆   ┆           ┆           ┆ f64       ┆           │\n",
      "╞═══════════╪════════════╪═══════════╪═════════╪═══╪═══════════╪═══════════╪═══════════╪═══════════╡\n",
      "│ 000001.SZ ┆ 20240102   ┆ 1.1584e6  ┆ 1074.93 ┆ … ┆ 2217.6093 ┆ 6486.3743 ┆ 2744.2180 ┆ -0.003256 │\n",
      "│           ┆            ┆           ┆         ┆   ┆ 09        ┆ 45        ┆ 84        ┆           │\n",
      "│ 000002.SZ ┆ 20240102   ┆ 811106.29 ┆ 1844.3  ┆ … ┆ 1736.4093 ┆ 19.432701 ┆ 2329.7434 ┆ -0.026601 │\n",
      "│           ┆            ┆           ┆         ┆   ┆ 99        ┆           ┆ 1         ┆           │\n",
      "│ 000004.SZ ┆ 20240102   ┆ 28867.0   ┆ 65.59   ┆ … ┆ -168.7552 ┆ -184.4013 ┆ -192.7135 ┆ -0.014789 │\n",
      "│           ┆            ┆           ┆         ┆   ┆ 72        ┆ 85        ┆ 84        ┆           │\n",
      "│ 000005.SZ ┆ 20240102   ┆ 63028.0   ┆ 10.38   ┆ … ┆ -96.94997 ┆ -295.0388 ┆ -46.06373 ┆ -0.05395  │\n",
      "│           ┆            ┆           ┆         ┆   ┆ 7         ┆ 72        ┆ 6         ┆           │\n",
      "│ 000006.SZ ┆ 20240102   ┆ 261947.19 ┆ 177.64  ┆ … ┆ -6.971845 ┆ -51.5536  ┆ -5.32671  ┆ -0.013454 │\n",
      "└───────────┴────────────┴───────────┴─────────┴───┴───────────┴───────────┴───────────┴───────────┘\n",
      "\n",
      "  测试集前5行预览:\n",
      "shape: (5, 42)\n",
      "┌───────────┬────────────┬───────────┬─────────┬───┬───────────┬───────────┬───────────┬───────────┐\n",
      "│ ts_code   ┆ trade_date ┆ vol       ┆ close   ┆ … ┆ profit_to ┆ cashflow_ ┆ operate_p ┆ return_5  │\n",
      "│ ---       ┆ ---        ┆ ---       ┆ ---     ┆   ┆ _market_c ┆ to_market ┆ rofit_to_ ┆ ---       │\n",
      "│ str       ┆ str        ┆ f64       ┆ f64     ┆   ┆ ap        ┆ _cap      ┆ market_ca ┆ f64       │\n",
      "│           ┆            ┆           ┆         ┆   ┆ ---       ┆ ---       ┆ p         ┆           │\n",
      "│           ┆            ┆           ┆         ┆   ┆ f64       ┆ f64       ┆ ---       ┆           │\n",
      "│           ┆            ┆           ┆         ┆   ┆           ┆           ┆ f64       ┆           │\n",
      "╞═══════════╪════════════╪═══════════╪═════════╪═══╪═══════════╪═══════════╪═══════════╪═══════════╡\n",
      "│ 000001.SZ ┆ 20250102   ┆ 1.8196e6  ┆ 1460.57 ┆ … ┆ 1791.1304 ┆ 6183.5904 ┆ 2158.1117 ┆ -0.002622 │\n",
      "│           ┆            ┆           ┆         ┆   ┆ 08        ┆ 38        ┆ 45        ┆           │\n",
      "│ 000002.SZ ┆ 20250102   ┆ 1.1827e6  ┆ 1291.92 ┆ … ┆ -1933.116 ┆ -1110.658 ┆ -1729.069 ┆ -0.022509 │\n",
      "│           ┆            ┆           ┆         ┆   ┆ 105       ┆ 303       ┆ 737       ┆           │\n",
      "│ 000004.SZ ┆ 20250102   ┆ 119760.37 ┆ 57.63   ┆ … ┆ -199.1144 ┆ -126.8907 ┆ -197.3308 ┆ -0.064897 │\n",
      "│           ┆            ┆           ┆         ┆   ┆ 31        ┆ 63        ┆ 47        ┆           │\n",
      "│ 000006.SZ ┆ 20250102   ┆ 307195.1  ┆ 288.12  ┆ … ┆ -646.1294 ┆ 74.343232 ┆ -637.5489 ┆ -0.048278 │\n",
      "│           ┆            ┆           ┆         ┆   ┆ 33        ┆           ┆ 17        ┆           │\n",
      "│ 000007.SZ ┆ 20250102   ┆ 68219.01  ┆ 58.15   ┆ … ┆ 6.740918  ┆ 108.91759 ┆ 22.556002 ┆ 0.015649  │\n",
      "│           ┆            ┆           ┆         ┆   ┆           ┆ 8         ┆           ┆           │\n",
      "└───────────┴────────────┴───────────┴─────────┴───┴───────────┴───────────┴───────────┴───────────┘\n"
     ]
    }
   ],
   "execution_count": 7
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-08T03:32:57.302228Z",
     "start_time": "2026-03-08T03:32:56.290367Z"
    }
   },
   "source": [
    "# 步骤 3: 训练集数据处理\n",
    "print(\"\\n[步骤 3/6] 训练集数据处理\")\n",
    "print(\"-\" * 60)\n",
    "fitted_processors = []\n",
    "if processors:\n",
    "    for i, processor in enumerate(processors, 1):\n",
    "        print(\n",
    "            f\"  [{i}/{len(processors)}] 应用处理器: {processor.__class__.__name__}\"\n",
    "        )\n",
    "        train_data_before = len(train_data)\n",
    "        train_data = processor.fit_transform(train_data)\n",
    "        train_data_after = len(train_data)\n",
    "        fitted_processors.append(processor)\n",
    "        print(f\"      处理前记录数: {train_data_before}\")\n",
    "        print(f\"      处理后记录数: {train_data_after}\")\n",
    "        if train_data_before != train_data_after:\n",
    "            print(f\"      删除记录数: {train_data_before - train_data_after}\")\n",
    "\n",
    "print(\"\\n  训练集处理后前5行预览:\")\n",
    "print(train_data.head())\n",
    "print(f\"\\n  训练集特征统计:\")\n",
    "print(f\"    特征数: {len(feature_cols)}\")\n",
    "print(f\"    样本数: {len(train_data)}\")\n",
    "print(f\"    缺失值统计:\")\n",
    "for col in feature_cols[:5]:  # 只显示前5个特征的缺失值\n",
    "    null_count = train_data[col].null_count()\n",
    "    if null_count > 0:\n",
    "        print(\n",
    "            f\"      {col}: {null_count} ({null_count / len(train_data) * 100:.2f}%)\"\n",
    "        )"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "[步骤 3/6] 训练集数据处理\n",
      "------------------------------------------------------------\n",
      "  [1/3] 应用处理器: NullFiller\n",
      "      处理前记录数: 2991506\n",
      "      处理后记录数: 2991506\n",
      "  [2/3] 应用处理器: Winsorizer\n",
      "      处理前记录数: 2991506\n",
      "      处理后记录数: 2991506\n",
      "  [3/3] 应用处理器: StandardScaler\n",
      "      处理前记录数: 2991506\n",
      "      处理后记录数: 2991506\n",
      "\n",
      "  训练集处理后前5行预览:\n",
      "shape: (5, 42)\n",
      "┌───────────┬───────────┬───────────┬───────────┬───┬───────────┬───────────┬───────────┬──────────┐\n",
      "│ ts_code   ┆ trade_dat ┆ vol       ┆ close     ┆ … ┆ profit_to ┆ cashflow_ ┆ operate_p ┆ return_5 │\n",
      "│ ---       ┆ e         ┆ ---       ┆ ---       ┆   ┆ _market_c ┆ to_market ┆ rofit_to_ ┆ ---      │\n",
      "│ str       ┆ ---       ┆ f64       ┆ f64       ┆   ┆ ap        ┆ _cap      ┆ market_ca ┆ f64      │\n",
      "│           ┆ str       ┆           ┆           ┆   ┆ ---       ┆ ---       ┆ p         ┆          │\n",
      "│           ┆           ┆           ┆           ┆   ┆ f64       ┆ f64       ┆ ---       ┆          │\n",
      "│           ┆           ┆           ┆           ┆   ┆           ┆           ┆ f64       ┆          │\n",
      "╞═══════════╪═══════════╪═══════════╪═══════════╪═══╪═══════════╪═══════════╪═══════════╪══════════╡\n",
      "│ 000001.SZ ┆ 20200102  ┆ 4.749919  ┆ 7.139221  ┆ … ┆ 1.228197  ┆ 2.569741  ┆ 1.426297  ┆ -0.00474 │\n",
      "│           ┆           ┆           ┆           ┆   ┆           ┆           ┆           ┆ 6        │\n",
      "│ 000002.SZ ┆ 20200102  ┆ 2.92576   ┆ 7.139221  ┆ … ┆ 1.353457  ┆ -0.145372 ┆ 1.83338   ┆ -0.01105 │\n",
      "│           ┆           ┆           ┆           ┆   ┆           ┆           ┆           ┆ 7        │\n",
      "│ 000004.SZ ┆ 20200102  ┆ -0.574944 ┆ 0.095339  ┆ … ┆ -0.560579 ┆ -0.252276 ┆ -0.513404 ┆ -0.00044 │\n",
      "│           ┆           ┆           ┆           ┆   ┆           ┆           ┆           ┆ 1        │\n",
      "│ 000005.SZ ┆ 20200102  ┆ -0.271162 ┆ -0.301277 ┆ … ┆ -0.080905 ┆ 0.217351  ┆ -0.044183 ┆ 0.022337 │\n",
      "│ 000006.SZ ┆ 20200102  ┆ -0.19857  ┆ 0.742213  ┆ … ┆ 1.028664  ┆ 0.501768  ┆ 1.186504  ┆ 0.012964 │\n",
      "└───────────┴───────────┴───────────┴───────────┴───┴───────────┴───────────┴───────────┴──────────┘\n",
      "\n",
      "  训练集特征统计:\n",
      "    特征数: 25\n",
      "    样本数: 2991506\n",
      "    缺失值统计:\n",
      "      ma5: 11541 (0.39%)\n",
      "      ma10: 25950 (0.87%)\n",
      "      ma20: 54850 (1.83%)\n",
      "      ma_ratio: 54850 (1.83%)\n",
      "      volatility_5: 11541 (0.39%)\n"
     ]
    }
   ],
   "execution_count": 8
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-08T03:33:14.426213Z",
     "start_time": "2026-03-08T03:32:57.306610Z"
    }
   },
   "source": [
    "# 步骤 4: 训练模型\n",
    "print(\"\\n[步骤 4/6] 训练模型\")\n",
    "print(\"-\" * 60)\n",
    "print(f\"  模型类型: LightGBM\")\n",
    "print(f\"  训练样本数: {len(train_data)}\")\n",
    "print(f\"  特征数: {len(feature_cols)}\")\n",
    "print(f\"  目标变量: {target_col}\")\n",
    "\n",
    "X_train = train_data.select(feature_cols)\n",
    "y_train = train_data.select(target_col).to_series()\n",
    "\n",
    "print(f\"\\n  目标变量统计:\")\n",
    "print(f\"    均值: {y_train.mean():.6f}\")\n",
    "print(f\"    标准差: {y_train.std():.6f}\")\n",
    "print(f\"    最小值: {y_train.min():.6f}\")\n",
    "print(f\"    最大值: {y_train.max():.6f}\")\n",
    "print(f\"    缺失值: {y_train.null_count()}\")\n",
    "\n",
    "print(\"\\n  开始训练...\")\n",
    "model.fit(X_train, y_train)\n",
    "print(\"  训练完成！\")"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "[步骤 4/6] 训练模型\n",
      "------------------------------------------------------------\n",
      "  模型类型: LightGBM\n",
      "  训练样本数: 2991506\n",
      "  特征数: 25\n",
      "  目标变量: return_5\n",
      "\n",
      "  目标变量统计:\n",
      "    均值: 0.001610\n",
      "    标准差: 0.059623\n",
      "    最小值: -0.155098\n",
      "    最大值: 0.212842\n",
      "    缺失值: 0\n",
      "\n",
      "  开始训练...\n",
      "  训练完成！\n"
     ]
    }
   ],
   "execution_count": 9
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-08T03:33:14.547157Z",
     "start_time": "2026-03-08T03:33:14.431622Z"
    }
   },
   "source": [
    "# 步骤 5: 测试集数据处理\n",
    "print(\"\\n[步骤 5/6] 测试集数据处理\")\n",
    "print(\"-\" * 60)\n",
    "if processors and test_data is not train_data:\n",
    "    for i, processor in enumerate(fitted_processors, 1):\n",
    "        print(\n",
    "            f\"  [{i}/{len(fitted_processors)}] 应用处理器: {processor.__class__.__name__}\"\n",
    "        )\n",
    "        test_data_before = len(test_data)\n",
    "        test_data = processor.transform(test_data)\n",
    "        test_data_after = len(test_data)\n",
    "        print(f\"      处理前记录数: {test_data_before}\")\n",
    "        print(f\"      处理后记录数: {test_data_after}\")\n",
    "else:\n",
    "    print(\"  跳过测试集处理\")"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "[步骤 5/6] 测试集数据处理\n",
      "------------------------------------------------------------\n",
      "  [1/3] 应用处理器: NullFiller\n",
      "      处理前记录数: 771207\n",
      "      处理后记录数: 771207\n",
      "  [2/3] 应用处理器: Winsorizer\n",
      "      处理前记录数: 771207\n",
      "      处理后记录数: 771207\n",
      "  [3/3] 应用处理器: StandardScaler\n",
      "      处理前记录数: 771207\n",
      "      处理后记录数: 771207\n"
     ]
    }
   ],
   "execution_count": 10
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-08T03:33:15.644580Z",
     "start_time": "2026-03-08T03:33:14.556580Z"
    }
   },
   "source": [
    "# 步骤 6: 生成预测\n",
    "print(\"\\n[步骤 6/6] 生成预测\")\n",
    "print(\"-\" * 60)\n",
    "X_test = test_data.select(feature_cols)\n",
    "print(f\"  测试样本数: {len(X_test)}\")\n",
    "print(\"  预测中...\")\n",
    "predictions = model.predict(X_test)\n",
    "print(f\"  预测完成！\")\n",
    "\n",
    "print(f\"\\n  预测结果统计:\")\n",
    "print(f\"    均值: {predictions.mean():.6f}\")\n",
    "print(f\"    标准差: {predictions.std():.6f}\")\n",
    "print(f\"    最小值: {predictions.min():.6f}\")\n",
    "print(f\"    最大值: {predictions.max():.6f}\")\n",
    "\n",
    "# 保存结果到 trainer\n",
    "trainer.results = test_data.with_columns([pl.Series(\"prediction\", predictions)])"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "[步骤 6/6] 生成预测\n",
      "------------------------------------------------------------\n",
      "  测试样本数: 771207\n",
      "  预测中...\n",
      "  预测完成！\n",
      "\n",
      "  预测结果统计:\n",
      "    均值: -0.002611\n",
      "    标准差: 0.006952\n",
      "    最小值: -0.104909\n",
      "    最大值: 0.093437\n"
     ]
    }
   ],
   "execution_count": 11
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.3 训练指标曲线"
   ]
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-08T03:33:20.917840Z",
     "start_time": "2026-03-08T03:33:15.648621Z"
    }
   },
   "source": [
    "print(\"\\n\" + \"=\" * 80)\n",
    "print(\"训练指标曲线\")\n",
    "print(\"=\" * 80)\n",
    "\n",
    "# 重新训练以收集指标（因为之前的训练没有保存评估结果）\n",
    "print(\"\\n重新训练模型以收集训练指标...\")\n",
    "\n",
    "import lightgbm as lgb\n",
    "\n",
    "# 准备数据（使用 val 做验证，test 不参与训练过程）\n",
    "X_train_np = X_train.to_numpy()\n",
    "y_train_np = y_train.to_numpy()\n",
    "X_val_np = val_data.select(feature_cols).to_numpy()\n",
    "y_val_np = val_data.select(target_col).to_series().to_numpy()\n",
    "\n",
    "# 创建数据集\n",
    "train_dataset = lgb.Dataset(X_train_np, label=y_train_np)\n",
    "val_dataset = lgb.Dataset(X_val_np, label=y_val_np, reference=train_dataset)\n",
    "\n",
    "# 用于存储评估结果\n",
    "evals_result = {}\n",
    "\n",
    "# 使用与原模型相同的参数重新训练\n",
    "# 正确的三分法：train用于训练，val用于验证，test不参与训练过程\n",
    "# 添加早停：如果验证指标连续100轮没有改善则停止训练\n",
    "booster_with_eval = lgb.train(\n",
    "    MODEL_PARAMS,\n",
    "    train_dataset,\n",
    "    num_boost_round=MODEL_PARAMS.get(\"n_estimators\", 100),\n",
    "    valid_sets=[train_dataset, val_dataset],\n",
    "    valid_names=[\"train\", \"val\"],\n",
    "    callbacks=[\n",
    "        lgb.record_evaluation(evals_result),\n",
    "        lgb.early_stopping(stopping_rounds=100, verbose=True),\n",
    "    ],\n",
    ")\n",
    "\n",
    "print(\"训练完成，指标已收集\")\n",
    "\n",
    "# 获取指标名称\n",
    "metric_name = list(evals_result[\"train\"].keys())[0]\n",
    "print(f\"\\n评估指标: {metric_name}\")\n",
    "\n",
    "# 提取训练和验证指标\n",
    "train_metric = evals_result[\"train\"][metric_name]\n",
    "val_metric = evals_result[\"val\"][metric_name]\n",
    "\n",
    "# 显示早停信息\n",
    "actual_rounds = len(train_metric)\n",
    "expected_rounds = MODEL_PARAMS.get(\"n_estimators\", 100)\n",
    "print(f\"\\n[早停信息]\")\n",
    "print(f\"  配置的最大轮数: {expected_rounds}\")\n",
    "print(f\"  实际训练轮数: {actual_rounds}\")\n",
    "if actual_rounds < expected_rounds:\n",
    "    print(f\"  早停状态: 已触发（连续100轮验证指标未改善）\")\n",
    "else:\n",
    "    print(f\"  早停状态: 未触发（达到最大轮数）\")\n",
    "\n",
    "print(f\"\\n最终指标:\")\n",
    "print(f\"  训练 {metric_name}: {train_metric[-1]:.6f}\")\n",
    "print(f\"  验证 {metric_name}: {val_metric[-1]:.6f}\")"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "================================================================================\n",
      "训练指标曲线\n",
      "================================================================================\n",
      "\n",
      "重新训练模型以收集训练指标...\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "Early stopping, best iteration is:\n",
      "[175]\ttrain's l1: 0.0422897\tval's l1: 0.0535436\n",
      "训练完成，指标已收集\n",
      "\n",
      "评估指标: l1\n",
      "\n",
      "[早停信息]\n",
      "  配置的最大轮数: 1000\n",
      "  实际训练轮数: 275\n",
      "  早停状态: 已触发（连续100轮验证指标未改善）\n",
      "\n",
      "最终指标:\n",
      "  训练 l1: 0.042114\n",
      "  验证 l1: 0.053595\n"
     ]
    }
   ],
   "execution_count": 12
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-08T03:33:21.167778Z",
     "start_time": "2026-03-08T03:33:20.923061Z"
    }
   },
   "source": [
    "# 绘制训练指标曲线\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(12, 6))\n",
    "\n",
    "# 绘制训练集和验证集的指标曲线（注意：val用于验证，test不参与训练）\n",
    "iterations = range(1, len(train_metric) + 1)\n",
    "ax.plot(iterations, train_metric, label=f\"Train {metric_name}\", linewidth=2, color=\"blue\")\n",
    "ax.plot(iterations, val_metric, label=f\"Validation {metric_name}\", linewidth=2, color=\"red\")\n",
    "\n",
    "ax.set_xlabel(\"Iteration\", fontsize=12)\n",
    "ax.set_ylabel(metric_name.upper(), fontsize=12)\n",
    "ax.set_title(f\"Training and Validation {metric_name.upper()} Curve\", fontsize=14, fontweight=\"bold\")\n",
    "ax.legend(fontsize=10)\n",
    "ax.grid(True, alpha=0.3)\n",
    "\n",
    "# 标记最佳验证指标点（用于早停决策）\n",
    "best_iter = val_metric.index(min(val_metric))\n",
    "best_metric = min(val_metric)\n",
    "ax.axvline(x=best_iter + 1, color=\"green\", linestyle=\"--\", alpha=0.7, label=f\"Best Iteration ({best_iter + 1})\")\n",
    "ax.scatter([best_iter + 1], [best_metric], color=\"green\", s=100, zorder=5)\n",
    "ax.annotate(\n",
    "    f\"Best: {best_metric:.6f}\\nIter: {best_iter + 1}\",\n",
    "    xy=(best_iter + 1, best_metric),\n",
    "    xytext=(best_iter + 1 + len(iterations) * 0.1, best_metric),\n",
    "    fontsize=9,\n",
    "    arrowprops=dict(arrowstyle=\"->\", color=\"green\", alpha=0.7),\n",
    ")\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n",
    "print(f\"\\n[指标分析]\")\n",
    "print(f\"  最佳验证 {metric_name}: {best_metric:.6f}\")\n",
    "print(f\"  最佳迭代轮数: {best_iter + 1}\")\n",
    "print(f\"  早停建议: 如果验证指标连续10轮不下降，建议在第 {best_iter + 1} 轮停止训练\")\n",
    "print(f\"\\n[重要提醒] 验证集仅用于早停/调参，测试集完全独立于训练过程！\")"
   ],
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 1200x600 with 1 Axes>"
      ],
      "image/png": "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
     },
     "metadata": {},
     "output_type": "display_data",
     "jetTransient": {
      "display_id": null
     }
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "[指标分析]\n",
      "  最佳验证 l1: 0.053544\n",
      "  最佳迭代轮数: 175\n",
      "  早停建议: 如果验证指标连续10轮不下降，建议在第 175 轮停止训练\n",
      "\n",
      "[重要提醒] 验证集仅用于早停/调参，测试集完全独立于训练过程！\n"
     ]
    }
   ],
   "execution_count": 13
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.4 查看结果"
   ]
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-08T03:33:21.213471Z",
     "start_time": "2026-03-08T03:33:21.183426Z"
    }
   },
   "source": [
    "print(\"\\n\" + \"=\" * 80)\n",
    "print(\"训练结果\")\n",
    "print(\"=\" * 80)\n",
    "\n",
    "results = trainer.results\n",
    "\n",
    "print(f\"\\n结果数据形状: {results.shape}\")\n",
    "print(f\"结果列: {results.columns}\")\n",
    "print(f\"\\n结果前10行预览:\")\n",
    "print(results.head(10))\n",
    "print(f\"\\n结果后5行预览:\")\n",
    "print(results.tail())\n",
    "\n",
    "print(f\"\\n每日预测样本数统计:\")\n",
    "daily_counts = results.group_by(\"trade_date\").agg(pl.len()).sort(\"trade_date\")\n",
    "print(f\"  最小: {daily_counts['len'].min()}\")\n",
    "print(f\"  最大: {daily_counts['len'].max()}\")\n",
    "print(f\"  平均: {daily_counts['len'].mean():.2f}\")\n",
    "\n",
    "# 展示某一天的前10个预测结果\n",
    "sample_date = results[\"trade_date\"][0]\n",
    "sample_data = results.filter(results[\"trade_date\"] == sample_date).head(10)\n",
    "print(f\"\\n示例日期 {sample_date} 的前10条预测:\")\n",
    "print(sample_data.select([\"ts_code\", \"trade_date\", target_col, \"prediction\"]))"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "================================================================================\n",
      "训练结果\n",
      "================================================================================\n",
      "\n",
      "结果数据形状: (771207, 43)\n",
      "结果列: ['ts_code', 'trade_date', 'vol', 'close', 'low', 'high', 'total_mv', 'f_ann_date', 'n_income', 'ebitda', 'total_profit', 'ebit', 'operate_profit', 'money_cap', 'total_liab', 'n_cashflow_act', 'ma5', 'ma10', 'ma20', 'ma_ratio', 'volatility_5', 'volatility_20', 'vol_ratio', 'return_10', 'return_20', 'return_diff', 'vol_ma5', 'vol_ma20', 'market_cap_rank', 'high_low_ratio', 'n_income_rank', 'operate_profit_rank', 'total_profit_rank', 'ebit_rank', 'ebitda_rank', 'total_liab_rank', 'money_cap_rank', 'n_cashflow_act_rank', 'profit_to_market_cap', 'cashflow_to_market_cap', 'operate_profit_to_market_cap', 'return_5', 'prediction']\n",
      "\n",
      "结果前10行预览:\n",
      "shape: (10, 43)\n",
      "┌───────────┬───────────┬───────────┬───────────┬───┬───────────┬───────────┬───────────┬──────────┐\n",
      "│ ts_code   ┆ trade_dat ┆ vol       ┆ close     ┆ … ┆ cashflow_ ┆ operate_p ┆ return_5  ┆ predicti │\n",
      "│ ---       ┆ e         ┆ ---       ┆ ---       ┆   ┆ to_market ┆ rofit_to_ ┆ ---       ┆ on       │\n",
      "│ str       ┆ ---       ┆ f64       ┆ f64       ┆   ┆ _cap      ┆ market_ca ┆ f64       ┆ ---      │\n",
      "│           ┆ str       ┆           ┆           ┆   ┆ ---       ┆ p         ┆           ┆ f64      │\n",
      "│           ┆           ┆           ┆           ┆   ┆ f64       ┆ ---       ┆           ┆          │\n",
      "│           ┆           ┆           ┆           ┆   ┆           ┆ f64       ┆           ┆          │\n",
      "╞═══════════╪═══════════╪═══════════╪═══════════╪═══╪═══════════╪═══════════╪═══════════╪══════════╡\n",
      "│ 000001.SZ ┆ 20250102  ┆ 5.587779  ┆ 7.139221  ┆ … ┆ 4.533499  ┆ 3.606645  ┆ -0.002622 ┆ -0.00794 │\n",
      "│           ┆           ┆           ┆           ┆   ┆           ┆           ┆           ┆ 7        │\n",
      "│ 000002.SZ ┆ 20250102  ┆ 3.526191  ┆ 7.139221  ┆ … ┆ -1.386218 ┆ -3.946245 ┆ -0.022509 ┆ -0.00197 │\n",
      "│           ┆           ┆           ┆           ┆   ┆           ┆           ┆           ┆ 2        │\n",
      "│ 000004.SZ ┆ 20250102  ┆ -0.216144 ┆ -0.117733 ┆ … ┆ -0.331877 ┆ -0.860887 ┆ -0.064897 ┆ 0.006379 │\n",
      "│ 000006.SZ ┆ 20250102  ┆ 0.443786  ┆ 1.365091  ┆ … ┆ -0.116207 ┆ -1.747611 ┆ -0.048278 ┆ 0.002845 │\n",
      "│ 000007.SZ ┆ 20250102  ┆ -0.397614 ┆ -0.114388 ┆ … ┆ -0.079153 ┆ -0.417972 ┆ 0.015649  ┆ -0.00445 │\n",
      "│           ┆           ┆           ┆           ┆   ┆           ┆           ┆           ┆ 4        │\n",
      "│ 000008.SZ ┆ 20250102  ┆ 3.219998  ┆ -0.079069 ┆ … ┆ -0.385751 ┆ -1.051162 ┆ -0.066939 ┆ -0.00372 │\n",
      "│           ┆           ┆           ┆           ┆   ┆           ┆           ┆           ┆ 9        │\n",
      "│ 000009.SZ ┆ 20250102  ┆ 0.204049  ┆ 0.025538  ┆ … ┆ 0.245489  ┆ 0.579808  ┆ -0.036045 ┆ 0.003052 │\n",
      "│ 000010.SZ ┆ 20250102  ┆ 0.584268  ┆ -0.300634 ┆ … ┆ -0.781057 ┆ -1.022311 ┆ 0.092123  ┆ -0.01003 │\n",
      "│           ┆           ┆           ┆           ┆   ┆           ┆           ┆           ┆ 7        │\n",
      "│ 000011.SZ ┆ 20250102  ┆ -0.488782 ┆ -0.243891 ┆ … ┆ -1.733032 ┆ -0.464004 ┆ -0.022094 ┆ 0.000949 │\n",
      "│ 000012.SZ ┆ 20250102  ┆ -0.060654 ┆ 0.501156  ┆ … ┆ 0.722727  ┆ 0.567525  ┆ -0.029188 ┆ 0.001858 │\n",
      "└───────────┴───────────┴───────────┴───────────┴───┴───────────┴───────────┴───────────┴──────────┘\n",
      "\n",
      "结果后5行预览:\n",
      "shape: (5, 43)\n",
      "┌───────────┬───────────┬───────────┬───────────┬───┬───────────┬───────────┬──────────┬───────────┐\n",
      "│ ts_code   ┆ trade_dat ┆ vol       ┆ close     ┆ … ┆ cashflow_ ┆ operate_p ┆ return_5 ┆ predictio │\n",
      "│ ---       ┆ e         ┆ ---       ┆ ---       ┆   ┆ to_market ┆ rofit_to_ ┆ ---      ┆ n         │\n",
      "│ str       ┆ ---       ┆ f64       ┆ f64       ┆   ┆ _cap      ┆ market_ca ┆ f64      ┆ ---       │\n",
      "│           ┆ str       ┆           ┆           ┆   ┆ ---       ┆ p         ┆          ┆ f64       │\n",
      "│           ┆           ┆           ┆           ┆   ┆ f64       ┆ ---       ┆          ┆           │\n",
      "│           ┆           ┆           ┆           ┆   ┆           ┆ f64       ┆          ┆           │\n",
      "╞═══════════╪═══════════╪═══════════╪═══════════╪═══╪═══════════╪═══════════╪══════════╪═══════════╡\n",
      "│ 605588.SH ┆ 20251231  ┆ -0.601605 ┆ -0.161094 ┆ … ┆ 0.196523  ┆ -0.667804 ┆ null     ┆ 0.000246  │\n",
      "│ 605589.SH ┆ 20251231  ┆ -0.152478 ┆ -0.293364 ┆ … ┆ -0.343887 ┆ 0.27582   ┆ null     ┆ -0.001414 │\n",
      "│ 605598.SH ┆ 20251231  ┆ -0.248806 ┆ 0.149122  ┆ … ┆ -0.191229 ┆ -0.349198 ┆ null     ┆ -0.011185 │\n",
      "│ 605599.SH ┆ 20251231  ┆ -0.482522 ┆ -0.364967 ┆ … ┆ 1.283712  ┆ 0.931037  ┆ null     ┆ 0.00215   │\n",
      "│ 689009.SH ┆ 20251231  ┆ -0.481066 ┆ -0.119084 ┆ … ┆ 1.09954   ┆ 0.730255  ┆ null     ┆ -0.001826 │\n",
      "└───────────┴───────────┴───────────┴───────────┴───┴───────────┴───────────┴──────────┴───────────┘\n",
      "\n",
      "每日预测样本数统计:\n",
      "  最小: 3147\n",
      "  最大: 3186\n",
      "  平均: 3173.69\n",
      "\n",
      "示例日期 20250102 的前10条预测:\n",
      "shape: (10, 4)\n",
      "┌───────────┬────────────┬───────────┬────────────┐\n",
      "│ ts_code   ┆ trade_date ┆ return_5  ┆ prediction │\n",
      "│ ---       ┆ ---        ┆ ---       ┆ ---        │\n",
      "│ str       ┆ str        ┆ f64       ┆ f64        │\n",
      "╞═══════════╪════════════╪═══════════╪════════════╡\n",
      "│ 000001.SZ ┆ 20250102   ┆ -0.002622 ┆ -0.007947  │\n",
      "│ 000002.SZ ┆ 20250102   ┆ -0.022509 ┆ -0.001972  │\n",
      "│ 000004.SZ ┆ 20250102   ┆ -0.064897 ┆ 0.006379   │\n",
      "│ 000006.SZ ┆ 20250102   ┆ -0.048278 ┆ 0.002845   │\n",
      "│ 000007.SZ ┆ 20250102   ┆ 0.015649  ┆ -0.004454  │\n",
      "│ 000008.SZ ┆ 20250102   ┆ -0.066939 ┆ -0.003729  │\n",
      "│ 000009.SZ ┆ 20250102   ┆ -0.036045 ┆ 0.003052   │\n",
      "│ 000010.SZ ┆ 20250102   ┆ 0.092123  ┆ -0.010037  │\n",
      "│ 000011.SZ ┆ 20250102   ┆ -0.022094 ┆ 0.000949   │\n",
      "│ 000012.SZ ┆ 20250102   ┆ -0.029188 ┆ 0.001858   │\n",
      "└───────────┴────────────┴───────────┴────────────┘\n"
     ]
    }
   ],
   "execution_count": 14
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.4 保存结果"
   ]
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-08T03:35:07.142015Z",
     "start_time": "2026-03-08T03:35:06.791043Z"
    }
   },
   "cell_type": "code",
   "source": [
    "print(\"\\n\" + \"=\" * 80)\n",
    "print(\"保存预测结果\")\n",
    "print(\"=\" * 80)\n",
    "\n",
    "# 确保输出目录存在\n",
    "os.makedirs(OUTPUT_DIR, exist_ok=True)\n",
    "\n",
    "# 生成时间戳\n",
    "start_dt = datetime.strptime(TEST_START, \"%Y%m%d\")\n",
    "end_dt = datetime.strptime(TEST_END, \"%Y%m%d\")\n",
    "date_str = f\"{start_dt.strftime('%Y%m%d')}_{end_dt.strftime('%Y%m%d')}\"\n",
    "\n",
    "# 保存每日 Top N\n",
    "print(f\"\\n[1/1] 保存每日 Top {TOP_N} 股票...\")\n",
    "topn_output_path = os.path.join(OUTPUT_DIR, f\"regression_output.csv\")\n",
    "\n",
    "# 按日期分组，取每日 top N\n",
    "topn_by_date = []\n",
    "unique_dates = results[\"trade_date\"].unique().sort()\n",
    "for date in unique_dates:\n",
    "    day_data = results.filter(results[\"trade_date\"] == date)\n",
    "    # 按 prediction 降序排序，取前 N\n",
    "    topn = day_data.sort(\"prediction\", descending=True).head(TOP_N)\n",
    "    topn_by_date.append(topn)\n",
    "\n",
    "# 合并所有日期的 top N\n",
    "topn_results = pl.concat(topn_by_date)\n",
    "\n",
    "# 格式化日期并调整列顺序：日期、分数、股票\n",
    "topn_to_save = topn_results.select(\n",
    "    [\n",
    "        pl.col(\"trade_date\").str.slice(0, 4)\n",
    "        + \"-\"\n",
    "        + pl.col(\"trade_date\").str.slice(4, 2)\n",
    "        + \"-\"\n",
    "        + pl.col(\"trade_date\").str.slice(6, 2).alias(\"date\"),\n",
    "        pl.col(\"prediction\").alias(\"score\"),\n",
    "        pl.col(\"ts_code\"),\n",
    "    ]\n",
    ")\n",
    "topn_to_save.write_csv(topn_output_path, include_header=True)\n",
    "print(f\"  保存路径: {topn_output_path}\")\n",
    "print(f\"  保存行数: {len(topn_to_save)}（{len(unique_dates)}个交易日 × 每日top{TOP_N}）\")\n",
    "print(f\"\\n  预览（前15行）:\")\n",
    "print(topn_to_save.head(15))"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "================================================================================\n",
      "保存预测结果\n",
      "================================================================================\n",
      "\n",
      "[1/1] 保存每日 Top 2 股票...\n",
      "  保存路径: output\\regression_output.csv\n",
      "  保存行数: 486（243个交易日 × 每日top2）\n",
      "\n",
      "  预览（前15行）:\n",
      "shape: (15, 3)\n",
      "┌────────────┬──────────┬───────────┐\n",
      "│ trade_date ┆ score    ┆ ts_code   │\n",
      "│ ---        ┆ ---      ┆ ---       │\n",
      "│ str        ┆ f64      ┆ str       │\n",
      "╞════════════╪══════════╪═══════════╡\n",
      "│ 2025-01-02 ┆ 0.078766 ┆ 603007.SH │\n",
      "│ 2025-01-02 ┆ 0.065013 ┆ 603559.SH │\n",
      "│ 2025-01-03 ┆ 0.087979 ┆ 603007.SH │\n",
      "│ 2025-01-03 ┆ 0.067351 ┆ 603559.SH │\n",
      "│ 2025-01-06 ┆ 0.087962 ┆ 603007.SH │\n",
      "│ …          ┆ …        ┆ …         │\n",
      "│ 2025-01-09 ┆ 0.036131 ┆ 603848.SH │\n",
      "│ 2025-01-09 ┆ 0.027876 ┆ 603977.SH │\n",
      "│ 2025-01-10 ┆ 0.027436 ┆ 002952.SZ │\n",
      "│ 2025-01-10 ┆ 0.024641 ┆ 603848.SH │\n",
      "│ 2025-01-13 ┆ 0.024852 ┆ 603848.SH │\n",
      "└────────────┴──────────┴───────────┘\n"
     ]
    }
   ],
   "execution_count": 23
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.5 特征重要性"
   ]
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-08T03:35:07.151989Z",
     "start_time": "2026-03-08T03:35:07.148773Z"
    }
   },
   "source": [
    "importance = model.feature_importance()\n",
    "if importance is not None:\n",
    "    print(\"\\n特征重要性:\")\n",
    "    print(importance.sort_values(ascending=False))\n",
    "\n",
    "print(\"\\n\" + \"=\" * 80)\n",
    "print(\"训练完成！\")\n",
    "print(\"=\" * 80)"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "特征重要性:\n",
      "ebitda_rank                     14616.823972\n",
      "return_10                        2170.157647\n",
      "return_20                        1597.996920\n",
      "ma_ratio                         1346.640091\n",
      "vol_ratio                        1307.361400\n",
      "high_low_ratio                    918.782296\n",
      "vol_ma20                          828.398926\n",
      "ma20                              635.353193\n",
      "volatility_5                      631.979581\n",
      "return_diff                       617.399685\n",
      "vol_ma5                           491.980673\n",
      "market_cap_rank                   415.972949\n",
      "volatility_20                     276.516502\n",
      "profit_to_market_cap              256.683738\n",
      "ma5                               236.319446\n",
      "ma10                              173.155782\n",
      "ebit_rank                         142.423690\n",
      "operate_profit_to_market_cap      127.127625\n",
      "total_liab_rank                    97.979562\n",
      "n_income_rank                      77.705741\n",
      "operate_profit_rank                75.762452\n",
      "cashflow_to_market_cap             68.135267\n",
      "money_cap_rank                     58.562898\n",
      "n_cashflow_act_rank                47.001317\n",
      "total_profit_rank                  36.941945\n",
      "dtype: float64\n",
      "\n",
      "================================================================================\n",
      "训练完成！\n",
      "================================================================================\n"
     ]
    }
   ],
   "execution_count": 24
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. 可视化分析\n",
    "\n",
    "使用训练好的模型直接绘图。\n",
    "- **特征重要性图**：辅助特征选择\n",
    "- **决策树图**：理解决策逻辑"
   ]
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-08T03:35:07.163848Z",
     "start_time": "2026-03-08T03:35:07.158216Z"
    }
   },
   "source": [
    "# 导入可视化库\n",
    "import matplotlib.pyplot as plt\n",
    "import lightgbm as lgb\n",
    "import pandas as pd\n",
    "\n",
    "# 从封装的model中取出底层Booster\n",
    "booster = model.model\n",
    "print(f\"模型类型: {type(booster)}\")\n",
    "print(f\"特征数量: {len(feature_cols)}\")"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "模型类型: <class 'lightgbm.basic.Booster'>\n",
      "特征数量: 25\n"
     ]
    }
   ],
   "execution_count": 25
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5.1 绘制特征重要性（辅助特征选择）\n",
    "\n",
    "**解读**：\n",
    "- 重要性高的特征对模型贡献大\n",
    "- 重要性为0的特征可以考虑删除\n",
    "- 可以帮助理解哪些因子最有效"
   ]
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-08T03:35:07.289782Z",
     "start_time": "2026-03-08T03:35:07.170460Z"
    }
   },
   "cell_type": "code",
   "source": [
    "print(\"绘制特征重要性...\")\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(10, 8))\n",
    "lgb.plot_importance(\n",
    "    booster, \n",
    "    max_num_features=20,\n",
    "    importance_type='gain',\n",
    "    title='Feature Importance (Gain)',\n",
    "    ax=ax\n",
    ")\n",
    "ax.set_xlabel('Importance (Gain)')\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n",
    "# 打印重要性排名\n",
    "importance_gain = pd.Series(\n",
    "    booster.feature_importance(importance_type='gain'),\n",
    "    index=feature_cols\n",
    ").sort_values(ascending=False)\n",
    "\n",
    "print(\"\\n[特征重要性排名 - Gain]\")\n",
    "print(importance_gain)\n",
    "\n",
    "# 识别低重要性特征\n",
    "zero_importance = importance_gain[importance_gain == 0].index.tolist()\n",
    "if zero_importance:\n",
    "    print(f\"\\n[低重要性特征] 以下{len(zero_importance)}个特征重要性为0，可考虑删除:\")\n",
    "    for feat in zero_importance:\n",
    "        print(f\"  - {feat}\")\n",
    "else:\n",
    "    print(\"\\n所有特征都有一定重要性\")"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "绘制特征重要性...\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 1000x800 with 1 Axes>"
      ],
      "image/png": "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
     },
     "metadata": {},
     "output_type": "display_data",
     "jetTransient": {
      "display_id": null
     }
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "[特征重要性排名 - Gain]\n",
      "ebitda_rank                     14616.823972\n",
      "return_10                        2170.157647\n",
      "return_20                        1597.996920\n",
      "ma_ratio                         1346.640091\n",
      "vol_ratio                        1307.361400\n",
      "high_low_ratio                    918.782296\n",
      "vol_ma20                          828.398926\n",
      "ma20                              635.353193\n",
      "volatility_5                      631.979581\n",
      "return_diff                       617.399685\n",
      "vol_ma5                           491.980673\n",
      "market_cap_rank                   415.972949\n",
      "volatility_20                     276.516502\n",
      "profit_to_market_cap              256.683738\n",
      "ma5                               236.319446\n",
      "ma10                              173.155782\n",
      "ebit_rank                         142.423690\n",
      "operate_profit_to_market_cap      127.127625\n",
      "total_liab_rank                    97.979562\n",
      "n_income_rank                      77.705741\n",
      "operate_profit_rank                75.762452\n",
      "cashflow_to_market_cap             68.135267\n",
      "money_cap_rank                     58.562898\n",
      "n_cashflow_act_rank                47.001317\n",
      "total_profit_rank                  36.941945\n",
      "dtype: float64\n",
      "\n",
      "所有特征都有一定重要性\n"
     ]
    }
   ],
   "execution_count": 26
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}