{ "cells": [ { "cell_type": "code", "id": "79a7758178bafdd3", "metadata": { "jupyter": { "source_hidden": true }, "ExecuteTime": { "end_time": "2025-04-09T14:48:04.783264Z", "start_time": "2025-04-09T14:48:04.779754Z" } }, "source": [ "# %load_ext autoreload\n", "# %autoreload 2\n", "\n", "import pandas as pd\n", "import warnings\n", "\n", "warnings.filterwarnings(\"ignore\")\n", "\n", "pd.set_option('display.max_columns', None)\n" ], "outputs": [], "execution_count": 25 }, { "cell_type": "code", "id": "a79cafb06a7e0e43", "metadata": { "scrolled": true, "ExecuteTime": { "end_time": "2025-04-09T14:48:50.995967Z", "start_time": "2025-04-09T14:48:04.834658Z" } }, "source": [ "from utils.utils import read_and_merge_h5_data\n", "\n", "print('daily data')\n", "df = read_and_merge_h5_data('../../data/daily_data.h5', key='daily_data',\n", " columns=['ts_code', 'trade_date', 'open', 'close', 'high', 'low', 'vol', 'pct_chg'],\n", " df=None)\n", "\n", "print('daily basic')\n", "df = read_and_merge_h5_data('../../data/daily_basic.h5', key='daily_basic',\n", " columns=['ts_code', 'trade_date', 'turnover_rate', 'pe_ttm', 'circ_mv', 'volume_ratio',\n", " 'is_st'], df=df, join='inner')\n", "df = df[df['trade_date'] >= '2021-01-01']\n", "\n", "print('stk limit')\n", "df = read_and_merge_h5_data('../../data/stk_limit.h5', key='stk_limit',\n", " columns=['ts_code', 'trade_date', 'pre_close', 'up_limit', 'down_limit'],\n", " df=df)\n", "print('money flow')\n", "df = read_and_merge_h5_data('../../data/money_flow.h5', key='money_flow',\n", " columns=['ts_code', 'trade_date', 'buy_sm_vol', 'sell_sm_vol', 'buy_lg_vol', 'sell_lg_vol',\n", " 'buy_elg_vol', 'sell_elg_vol', 'net_mf_vol'],\n", " df=df)\n", "print('cyq perf')\n", "df = read_and_merge_h5_data('../../data/cyq_perf.h5', key='cyq_perf',\n", " columns=['ts_code', 'trade_date', 'his_low', 'his_high', 'cost_5pct', 'cost_15pct',\n", " 'cost_50pct',\n", " 'cost_85pct', 'cost_95pct', 'weight_avg', 'winner_rate'],\n", " df=df)\n", "print(df.info())" ], "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "daily data\n", "daily basic\n", "inner merge on ['ts_code', 'trade_date']\n", "stk limit\n", "left merge on ['ts_code', 'trade_date']\n", "money flow\n", "left merge on ['ts_code', 'trade_date']\n", "cyq perf\n", "left merge on ['ts_code', 'trade_date']\n", "\n", "RangeIndex: 5113004 entries, 0 to 5113003\n", "Data columns (total 31 columns):\n", " # Column Dtype \n", "--- ------ ----- \n", " 0 ts_code object \n", " 1 trade_date datetime64[ns]\n", " 2 open float64 \n", " 3 close float64 \n", " 4 high float64 \n", " 5 low float64 \n", " 6 vol float64 \n", " 7 pct_chg float64 \n", " 8 turnover_rate float64 \n", " 9 pe_ttm float64 \n", " 10 circ_mv float64 \n", " 11 volume_ratio float64 \n", " 12 is_st bool \n", " 13 up_limit float64 \n", " 14 down_limit float64 \n", " 15 buy_sm_vol float64 \n", " 16 sell_sm_vol float64 \n", " 17 buy_lg_vol float64 \n", " 18 sell_lg_vol float64 \n", " 19 buy_elg_vol float64 \n", " 20 sell_elg_vol float64 \n", " 21 net_mf_vol float64 \n", " 22 his_low float64 \n", " 23 his_high float64 \n", " 24 cost_5pct float64 \n", " 25 cost_15pct float64 \n", " 26 cost_50pct float64 \n", " 27 cost_85pct float64 \n", " 28 cost_95pct float64 \n", " 29 weight_avg float64 \n", " 30 winner_rate float64 \n", "dtypes: bool(1), datetime64[ns](1), float64(28), object(1)\n", "memory usage: 1.1+ GB\n", "None\n" ] } ], "execution_count": 26 }, { "cell_type": "code", "id": "cac01788dac10678", "metadata": { "ExecuteTime": { "end_time": "2025-04-09T14:48:56.021834Z", "start_time": "2025-04-09T14:48:51.037076Z" } }, "source": [ "print('industry')\n", "industry_df = read_and_merge_h5_data('../../data/industry_data.h5', key='industry_data',\n", " columns=['ts_code', 'l2_code', 'in_date'],\n", " df=None, on=['ts_code'], join='left')\n", "\n", "\n", "def merge_with_industry_data(df, industry_df):\n", " # 确保日期字段是 datetime 类型\n", " df['trade_date'] = pd.to_datetime(df['trade_date'])\n", " industry_df['in_date'] = pd.to_datetime(industry_df['in_date'])\n", "\n", " # 对 industry_df 按 ts_code 和 in_date 排序\n", " industry_df_sorted = industry_df.sort_values(['in_date', 'ts_code'])\n", "\n", " # 对原始 df 按 ts_code 和 trade_date 排序\n", " df_sorted = df.sort_values(['trade_date', 'ts_code'])\n", "\n", " # 使用 merge_asof 进行向后合并\n", " merged = pd.merge_asof(\n", " df_sorted,\n", " industry_df_sorted,\n", " by='ts_code', # 按 ts_code 分组\n", " left_on='trade_date',\n", " right_on='in_date',\n", " direction='backward'\n", " )\n", "\n", " # 获取每个 ts_code 的最早 in_date 记录\n", " min_in_date_per_ts = (industry_df_sorted\n", " .groupby('ts_code')\n", " .first()\n", " .reset_index()[['ts_code', 'l2_code']])\n", "\n", " # 填充未匹配到的记录（trade_date 早于所有 in_date 的情况）\n", " merged['l2_code'] = merged['l2_code'].fillna(\n", " merged['ts_code'].map(min_in_date_per_ts.set_index('ts_code')['l2_code'])\n", " )\n", "\n", " # 保留需要的列并重置索引\n", " result = merged.reset_index(drop=True)\n", " return result\n", "\n", "\n", "# 使用示例\n", "df = merge_with_industry_data(df, industry_df)\n", "# print(mdf[mdf['ts_code'] == '600751.SH'][['ts_code', 'trade_date', 'l2_code']])" ], "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "industry\n" ] } ], "execution_count": 27 }, { "cell_type": "code", "id": "c4e9e1d31da6dba6", "metadata": { "ExecuteTime": { "end_time": "2025-04-09T14:48:56.082954Z", "start_time": "2025-04-09T14:48:56.026849Z" } }, "source": [ "def calculate_indicators(df):\n", " \"\"\"\n", " 计算四个指标：当日涨跌幅、5日移动平均、RSI、MACD。\n", " \"\"\"\n", " df = df.sort_values('trade_date')\n", " df['daily_return'] = (df['close'] - df['pre_close']) / df['pre_close'] * 100\n", " # df['5_day_ma'] = df['close'].rolling(window=5).mean()\n", " delta = df['close'].diff()\n", " gain = delta.where(delta > 0, 0)\n", " loss = -delta.where(delta < 0, 0)\n", " avg_gain = gain.rolling(window=14).mean()\n", " avg_loss = loss.rolling(window=14).mean()\n", " rs = avg_gain / avg_loss\n", " df['RSI'] = 100 - (100 / (1 + rs))\n", "\n", " # 计算MACD\n", " ema12 = df['close'].ewm(span=12, adjust=False).mean()\n", " ema26 = df['close'].ewm(span=26, adjust=False).mean()\n", " df['MACD'] = ema12 - ema26\n", " df['Signal_line'] = df['MACD'].ewm(span=9, adjust=False).mean()\n", " df['MACD_hist'] = df['MACD'] - df['Signal_line']\n", "\n", " # 4. 情绪因子1：市场上涨比例（Up Ratio）\n", " df['up_ratio'] = df['daily_return'].apply(lambda x: 1 if x > 0 else 0)\n", " df['up_ratio_20d'] = df['up_ratio'].rolling(window=20).mean() # 过去20天上涨比例\n", "\n", " # 5. 情绪因子2：成交量变化率（Volume Change Rate）\n", " df['volume_mean'] = df['vol'].rolling(window=20).mean() # 过去20天的平均成交量\n", " df['volume_change_rate'] = (df['vol'] - df['volume_mean']) / df['volume_mean'] * 100 # 成交量变化率\n", "\n", " # 6. 情绪因子3：波动率（Volatility）\n", " df['volatility'] = df['daily_return'].rolling(window=20).std() # 过去20天的日收益率标准差\n", "\n", " # 7. 情绪因子4：成交额变化率（Amount Change Rate）\n", " df['amount_mean'] = df['amount'].rolling(window=20).mean() # 过去20天的平均成交额\n", " df['amount_change_rate'] = (df['amount'] - df['amount_mean']) / df['amount_mean'] * 100 # 成交额变化率\n", "\n", " return df\n", "\n", "\n", "def generate_index_indicators(h5_filename):\n", " df = pd.read_hdf(h5_filename, key='index_data')\n", " df['trade_date'] = pd.to_datetime(df['trade_date'], format='%Y%m%d')\n", " df = df.sort_values('trade_date')\n", "\n", " # 计算每个ts_code的相关指标\n", " df_indicators = []\n", " for ts_code in df['ts_code'].unique():\n", " df_index = df[df['ts_code'] == ts_code].copy()\n", " df_index = calculate_indicators(df_index)\n", " df_indicators.append(df_index)\n", "\n", " # 合并所有指数的结果\n", " df_all_indicators = pd.concat(df_indicators, ignore_index=True)\n", "\n", " # 保留trade_date列，并将同一天的数据按ts_code合并成一行\n", " df_final = df_all_indicators.pivot_table(\n", " index='trade_date',\n", " columns='ts_code',\n", " values=['daily_return', 'RSI', 'MACD', 'Signal_line',\n", " 'MACD_hist', 'up_ratio_20d', 'volume_change_rate', 'volatility',\n", " 'amount_change_rate', 'amount_mean'],\n", " aggfunc='last'\n", " )\n", "\n", " df_final.columns = [f\"{col[1]}_{col[0]}\" for col in df_final.columns]\n", " df_final = df_final.reset_index()\n", "\n", " return df_final\n", "\n", "\n", "# 使用函数\n", "h5_filename = '../../data/index_data.h5'\n", "index_data = generate_index_indicators(h5_filename)\n", "index_data = index_data.dropna()\n" ], "outputs": [], "execution_count": 28 }, { "cell_type": "code", "id": "a735bc02ceb4d872", "metadata": { "jupyter": { "source_hidden": true }, "ExecuteTime": { "end_time": "2025-04-09T14:48:56.140563Z", "start_time": "2025-04-09T14:48:56.110420Z" } }, "source": [ "import numpy as np\n", "import talib\n", "\n", "\n", "def get_rolling_factor(df):\n", " old_columns = df.columns.tolist()[:]\n", "\n", " # 按股票和日期排序（如果尚未排序）\n", " df = df.sort_values(by=['ts_code', 'trade_date'])\n", "\n", " grouped = df.groupby('ts_code', group_keys=False)\n", "\n", " window = 20\n", " df['_is_positive'] = (df['pct_chg'] > 0).astype(int)\n", " df['_is_negative'] = (df['pct_chg'] < 0).astype(int)\n", " df['cat_is_positive'] = (df['pct_chg'] > 0).astype(int)\n", "\n", " # 分离正负收益率 (用于计算各自的均值和平方均值)\n", " # 注意：这里我们保留原始收益率用于计算，而不是 clip 到 0\n", " df['_pos_returns'] = df['pct_chg'].where(df['pct_chg'] > 0, 0) # 非正设为0，便于求和\n", " df['_neg_returns'] = df['pct_chg'].where(df['pct_chg'] < 0, 0) # 非负设为0，便于求和\n", "\n", " # 计算收益率的平方 (用于计算 E[X^2])\n", " df['_pos_returns_sq'] = np.square(df['_pos_returns'])\n", " df['_neg_returns_sq'] = np.square(df['_neg_returns']) # 平方后负数变正\n", "\n", " # 4. 计算滚动统计量 (使用内置函数，速度较快)\n", " # 计算正收益日的统计量\n", " rolling_pos_count = grouped['_is_positive'].rolling(window, min_periods=max(1, window // 2)).sum()\n", " rolling_pos_sum = grouped['_pos_returns'].rolling(window, min_periods=max(1, window // 2)).sum()\n", " rolling_pos_sum_sq = grouped['_pos_returns_sq'].rolling(window, min_periods=max(1, window // 2)).sum()\n", "\n", " # 计算负收益日的统计量\n", " rolling_neg_count = grouped['_is_negative'].rolling(window, min_periods=max(1, window // 2)).sum()\n", " rolling_neg_sum = grouped['_neg_returns'].rolling(window, min_periods=max(1, window // 2)).sum()\n", " rolling_neg_sum_sq = grouped['_neg_returns_sq'].rolling(window, min_periods=max(1, window // 2)).sum()\n", "\n", " # 5. 计算方差和标准差\n", " pos_mean_sq = rolling_pos_sum_sq / rolling_pos_count\n", " pos_mean = rolling_pos_sum / rolling_pos_count\n", " pos_var = pos_mean_sq - np.square(pos_mean)\n", " pos_var = pos_var.where(rolling_pos_count >= 2, np.nan).clip(lower=0)\n", " upside_vol = np.sqrt(pos_var)\n", "\n", " neg_mean_sq = rolling_neg_sum_sq / rolling_neg_count\n", " neg_mean = rolling_neg_sum / rolling_neg_count # 注意 neg_mean 是负数\n", " neg_var = neg_mean_sq - np.square(neg_mean)\n", " neg_var = neg_var.where(rolling_neg_count >= 2, np.nan).clip(lower=0)\n", " downside_vol = np.sqrt(neg_var)\n", "\n", " # rolling 操作后结果带有 MultiIndex，需要去除股票代码层级以便合并\n", " df['upside_vol'] = upside_vol.reset_index(level=0, drop=True)\n", " df['downside_vol'] = downside_vol.reset_index(level=0, drop=True)\n", "\n", " df['vol_ratio'] = df['upside_vol'] / df['downside_vol']\n", " df['vol_ratio'] = df['vol_ratio'].replace([np.inf, -np.inf], np.nan).fillna(0) # 或 fillna(np.nan)\n", "\n", " df['return_skew'] = grouped['pct_chg'].rolling(window=5).skew().reset_index(0, drop=True)\n", " df['return_kurtosis'] = grouped['pct_chg'].rolling(window=5).kurt().reset_index(0, drop=True)\n", "\n", " # 因子 1：短期成交量变化率\n", " df['volume_change_rate'] = (\n", " grouped['vol'].rolling(window=2).mean() /\n", " grouped['vol'].rolling(window=10).mean() - 1\n", " ).reset_index(level=0, drop=True) # 确保索引对齐\n", "\n", " # 因子 2：成交量突破信号\n", " max_volume = grouped['vol'].rolling(window=5).max().reset_index(level=0, drop=True) # 确保索引对齐\n", " df['cat_volume_breakout'] = (df['vol'] > max_volume)\n", "\n", " # 因子 3：换手率均线偏离度\n", " mean_turnover = grouped['turnover_rate'].rolling(window=3).mean().reset_index(level=0, drop=True)\n", " std_turnover = grouped['turnover_rate'].rolling(window=3).std().reset_index(level=0, drop=True)\n", " df['turnover_deviation'] = (df['turnover_rate'] - mean_turnover) / std_turnover\n", "\n", " # 因子 4：换手率激增信号\n", " df['cat_turnover_spike'] = (df['turnover_rate'] > mean_turnover + 2 * std_turnover)\n", "\n", " # 因子 5：量比均值\n", " df['avg_volume_ratio'] = grouped['volume_ratio'].rolling(window=3).mean().reset_index(level=0, drop=True)\n", "\n", " # 因子 6：量比突破信号\n", " max_volume_ratio = grouped['volume_ratio'].rolling(window=5).max().reset_index(level=0, drop=True)\n", " df['cat_volume_ratio_breakout'] = (df['volume_ratio'] > max_volume_ratio)\n", "\n", " df['vol_spike'] = grouped.apply(\n", " lambda x: pd.Series(x['vol'].rolling(20).mean(), index=x.index)\n", " )\n", " df['vol_std_5'] = grouped['vol'].pct_change().rolling(window=5).std()\n", "\n", " # 计算 ATR\n", " df['atr_14'] = grouped.apply(\n", " lambda x: pd.Series(talib.ATR(x['high'].values, x['low'].values, x['close'].values, timeperiod=14),\n", " index=x.index)\n", " )\n", " df['atr_6'] = grouped.apply(\n", " lambda x: pd.Series(talib.ATR(x['high'].values, x['low'].values, x['close'].values, timeperiod=6),\n", " index=x.index)\n", " )\n", "\n", " # 计算 OBV 及其均线\n", " df['obv'] = grouped.apply(\n", " lambda x: pd.Series(talib.OBV(x['close'].values, x['vol'].values), index=x.index)\n", " )\n", " print(df.columns)\n", " df['maobv_6'] = grouped.apply(\n", " lambda x: pd.Series(talib.SMA(x['obv'].values, timeperiod=6), index=x.index)\n", " )\n", "\n", " df['rsi_3'] = grouped.apply(\n", " lambda x: pd.Series(talib.RSI(x['close'].values, timeperiod=3), index=x.index)\n", " )\n", " # df['rsi_6'] = grouped.apply(\n", " # lambda x: pd.Series(talib.RSI(x['close'].values, timeperiod=6), index=x.index)\n", " # )\n", " # df['rsi_9'] = grouped.apply(\n", " # lambda x: pd.Series(talib.RSI(x['close'].values, timeperiod=9), index=x.index)\n", " # )\n", "\n", " # 计算 return_10 和 return_20\n", " df['return_5'] = grouped['close'].apply(lambda x: x / x.shift(5) - 1)\n", " # df['return_10'] = grouped['close'].apply(lambda x: x / x.shift(10) - 1)\n", " df['return_20'] = grouped['close'].apply(lambda x: x / x.shift(20) - 1)\n", "\n", " # df['avg_close_5'] = grouped['close'].apply(lambda x: x.rolling(window=5).mean() / x)\n", "\n", " # 计算标准差指标\n", " df['std_return_5'] = grouped['close'].apply(lambda x: x.pct_change().rolling(window=5).std())\n", " # df['std_return_15'] = grouped['close'].apply(lambda x: x.pct_change().rolling(window=15).std())\n", " # df['std_return_25'] = grouped['close'].apply(lambda x: x.pct_change().rolling(window=25).std())\n", " df['std_return_90'] = grouped['close'].apply(lambda x: x.pct_change().rolling(window=90).std())\n", " df['std_return_90_2'] = grouped['close'].apply(lambda x: x.shift(10).pct_change().rolling(window=90).std())\n", "\n", " # 计算 EMA 指标\n", " df['_ema_5'] = grouped['close'].apply(\n", " lambda x: pd.Series(talib.EMA(x.values, timeperiod=5), index=x.index)\n", " )\n", " df['_ema_13'] = grouped['close'].apply(\n", " lambda x: pd.Series(talib.EMA(x.values, timeperiod=13), index=x.index)\n", " )\n", " df['_ema_20'] = grouped['close'].apply(\n", " lambda x: pd.Series(talib.EMA(x.values, timeperiod=20), index=x.index)\n", " )\n", " df['_ema_60'] = grouped['close'].apply(\n", " lambda x: pd.Series(talib.EMA(x.values, timeperiod=60), index=x.index)\n", " )\n", "\n", " # 计算 act_factor1, act_factor2, act_factor3, act_factor4\n", " df['act_factor1'] = grouped['_ema_5'].apply(\n", " lambda x: np.arctan((x / x.shift(1) - 1) * 100) * 57.3 / 50\n", " )\n", " df['act_factor2'] = grouped['_ema_13'].apply(\n", " lambda x: np.arctan((x / x.shift(1) - 1) * 100) * 57.3 / 40\n", " )\n", " df['act_factor3'] = grouped['_ema_20'].apply(\n", " lambda x: np.arctan((x / x.shift(1) - 1) * 100) * 57.3 / 21\n", " )\n", " df['act_factor4'] = grouped['_ema_60'].apply(\n", " lambda x: np.arctan((x / x.shift(1) - 1) * 100) * 57.3 / 10\n", " )\n", "\n", " # 根据 trade_date 截面计算排名\n", " df['rank_act_factor1'] = df.groupby('trade_date', group_keys=False)['act_factor1'].rank(ascending=False, pct=True)\n", " df['rank_act_factor2'] = df.groupby('trade_date', group_keys=False)['act_factor2'].rank(ascending=False, pct=True)\n", " df['rank_act_factor3'] = df.groupby('trade_date', group_keys=False)['act_factor3'].rank(ascending=False, pct=True)\n", "\n", " df['log(circ_mv)'] = np.log(df['circ_mv'])\n", "\n", " window_high_volume = 5\n", " window_close_stddev = 20\n", " period_delta = 5\n", "\n", " # 计算每只股票的滚动协方差\n", " def calculate_rolling_cov(group):\n", " return group['high'].rolling(window_high_volume).cov(group['vol'])\n", "\n", " df['cov'] = grouped.apply(calculate_rolling_cov)\n", "\n", " # 计算每只股票的协方差差分\n", " def calculate_delta_cov(group):\n", " return group['cov'].diff(period_delta)\n", "\n", " df['delta_cov'] = grouped.apply(calculate_delta_cov)\n", "\n", " # 计算每只股票的滚动标准差\n", " def calculate_stddev_close(group):\n", " return group['close'].rolling(window_close_stddev).std()\n", "\n", " df['_stddev_close'] = grouped.apply(calculate_stddev_close)\n", " df['_rank_stddev'] = df.groupby('trade_date')['_stddev_close'].rank(pct=True)\n", " df['alpha_22_improved'] = -1 * df['delta_cov'] * df['_rank_stddev']\n", "\n", "\n", " df['alpha_003'] = np.where(df['high'] != df['low'],\n", " (df['close'] - df['open']) / (df['high'] - df['low']),\n", " 0)\n", "\n", " df['alpha_007'] = grouped.apply(lambda x: x['close'].rolling(5).corr(x['vol']))\n", " df['alpha_007'] = df.groupby('trade_date', group_keys=False)['alpha_007'].rank(ascending=True, pct=True)\n", "\n", " df['alpha_013'] = grouped['close'].transform(lambda x: x.rolling(5).sum() - x.rolling(20).sum())\n", " df['alpha_013'] = df.groupby('trade_date', group_keys=False)['alpha_013'].rank(ascending=True, pct=True)\n", "\n", " df['cat_up_limit'] = (df['close'] == df['up_limit']) # 是否涨停（1表示涨停，0表示未涨停）\n", " df['cat_down_limit'] = (df['close'] == df['down_limit']) # 是否跌停（1表示跌停，0表示未跌停）\n", " df['up_limit_count_10d'] = grouped['cat_up_limit'].rolling(window=10, min_periods=1).sum().reset_index(level=0,\n", " drop=True)\n", " df['down_limit_count_10d'] = grouped['cat_down_limit'].rolling(window=10, min_periods=1).sum().reset_index(level=0,\n", " drop=True)\n", "\n", " # 3. 最近连续涨跌停天数\n", " def calculate_consecutive_limits(series):\n", " \"\"\"\n", " 计算连续涨停/跌停天数。\n", " \"\"\"\n", " consecutive_up = series * (series.groupby((series != series.shift()).cumsum()).cumcount() + 1)\n", " consecutive_down = series * (series.groupby((series != series.shift()).cumsum()).cumcount() + 1)\n", " return consecutive_up, consecutive_down\n", "\n", " # 连续涨停天数\n", " df['consecutive_up_limit'] = grouped['cat_up_limit'].apply(\n", " lambda x: calculate_consecutive_limits(x)[0]\n", " )\n", "\n", " df['vol_break'] = np.where((df['close'] > df['cost_85pct']) & (df['volume_ratio'] > 2), 1, 0)\n", "\n", " df['weight_roc5'] = grouped['weight_avg'].apply(lambda x: x.pct_change(5))\n", "\n", " def rolling_corr(group):\n", " roc_close = group['close'].pct_change()\n", " roc_weight = group['weight_avg'].pct_change()\n", " return roc_close.rolling(10).corr(roc_weight)\n", "\n", " df['price_cost_divergence'] = grouped.apply(rolling_corr)\n", "\n", " df['smallcap_concentration'] = (1 / df['log(circ_mv)']) * (df['cost_85pct'] - df['cost_15pct'])\n", "\n", " # 16. 筹码稳定性指数 (20日波动率)\n", " df['weight_std20'] = grouped['weight_avg'].apply(lambda x: x.rolling(20).std())\n", " df['cost_stability'] = df['weight_std20'] / grouped['weight_avg'].transform(lambda x: x.rolling(20).mean())\n", "\n", " # 17. 成本区间突破标记\n", " df['high_cost_break_days'] = grouped.apply(lambda g: g['close'].gt(g['cost_95pct']).rolling(5).sum())\n", "\n", " # 20. 筹码-流动性风险\n", " df['liquidity_risk'] = (df['cost_95pct'] - df['cost_5pct']) * (\n", " 1 / grouped['vol'].transform(lambda x: x.rolling(10).mean()))\n", "\n", " # 7. 市值波动率因子 (使用 grouped)\n", " df['turnover_std'] = grouped['turnover_rate'].transform(lambda x: x.rolling(window=20).std())\n", " df['mv_volatility'] = grouped.apply(lambda x: x['turnover_std'] / x['log(circ_mv)'])\n", "\n", " # 8. 市值成长性因子\n", " df['volume_growth'] = grouped['vol'].pct_change(periods=20)\n", " df['mv_growth'] = df['volume_growth'] / df['log(circ_mv)']\n", "\n", " # AR 指标\n", " df[\"ar\"] = grouped.apply(lambda x: (x[\"high\"].div(x[\"open\"]).rolling(3).sum()) / (x[\"open\"].div(x[\"low\"]).rolling(3).sum()) * 100)\n", "\n", " # BR 指标\n", " df[\"pre_close\"] = grouped[\"close\"].shift(1)\n", " df[\"br_up\"] = (df[\"high\"] - df[\"pre_close\"]).clip(lower=0)\n", " df[\"br_down\"] = (df[\"pre_close\"] - df[\"low\"]).clip(lower=0)\n", " df[\"br\"] = grouped.apply(lambda x: (x[\"br_up\"].rolling(3).sum()) / (x[\"br_down\"].rolling(3).sum()) * 100)\n", "\n", " # ARBR\n", " df['arbr'] = df['ar'] - df['br']\n", " df.drop(columns=[\"pre_close\", \"br_up\", \"br_down\", 'ar', 'br'], inplace=True)\n", "\n", " df.drop(columns=['weight_std20'], inplace=True, errors='ignore')\n", " df.drop(\n", " columns=['_is_positive', '_is_negative', '_pos_returns', '_neg_returns', '_pos_returns_sq', '_neg_returns_sq'],\n", " inplace=True, errors='ignore')\n", " new_columns = [col for col in df.columns.tolist()[:] if col not in old_columns]\n", "\n", " return df, new_columns\n", "\n", "\n", "def get_simple_factor(df):\n", " old_columns = df.columns.tolist()[:]\n", " df = df.sort_values(by=['ts_code', 'trade_date'])\n", "\n", " alpha = 0.5\n", " df['momentum_factor'] = df['volume_change_rate'] + alpha * df['turnover_deviation']\n", " df['resonance_factor'] = df['volume_ratio'] * df['pct_chg']\n", " df['log_close'] = np.log(df['close'])\n", "\n", " df['cat_vol_spike'] = df['vol'] > 2 * df['vol_spike']\n", "\n", " df['up'] = (df['high'] - df[['close', 'open']].max(axis=1)) / df['close']\n", " df['down'] = (df[['close', 'open']].min(axis=1) - df['low']) / df['close']\n", "\n", " df['obv-maobv_6'] = df['obv'] - df['maobv_6']\n", "\n", " # 计算比值指标\n", " df['std_return_5 / std_return_90'] = df['std_return_5'] / df['std_return_90']\n", " # df['std_return_5 / std_return_25'] = df['std_return_5'] / df['std_return_25']\n", "\n", " # 计算标准差差值\n", " df['std_return_90 - std_return_90_2'] = df['std_return_90'] - df['std_return_90_2']\n", "\n", " # df['cat_af1'] = df['act_factor1'] > 0\n", " df['cat_af2'] = df['act_factor2'] > df['act_factor1']\n", " df['cat_af3'] = df['act_factor3'] > df['act_factor2']\n", " df['cat_af4'] = df['act_factor4'] > df['act_factor3']\n", "\n", " # 计算 act_factor5 和 act_factor6\n", " df['act_factor5'] = df['act_factor1'] + df['act_factor2'] + df['act_factor3'] + df['act_factor4']\n", " df['act_factor6'] = (df['act_factor1'] - df['act_factor2']) / np.sqrt(\n", " df['act_factor1'] ** 2 + df['act_factor2'] ** 2)\n", "\n", " df['active_buy_volume_large'] = df['buy_lg_vol'] / df['net_mf_vol']\n", " df['active_buy_volume_big'] = df['buy_elg_vol'] / df['net_mf_vol']\n", " df['active_buy_volume_small'] = df['buy_sm_vol'] / df['net_mf_vol']\n", "\n", " df['buy_lg_vol_minus_sell_lg_vol'] = (df['buy_lg_vol'] - df['sell_lg_vol']) / df['net_mf_vol']\n", " df['buy_elg_vol_minus_sell_elg_vol'] = (df['buy_elg_vol'] - df['sell_elg_vol']) / df['net_mf_vol']\n", "\n", " df['log(circ_mv)'] = np.log(df['circ_mv'])\n", "\n", " df['ctrl_strength'] = (df['cost_85pct'] - df['cost_15pct']) / (df['his_high'] - df['his_low'])\n", "\n", " df['low_cost_dev'] = (df['close'] - df['cost_5pct']) / (df['cost_50pct'] - df['cost_5pct'])\n", "\n", " df['asymmetry'] = (df['cost_95pct'] - df['cost_50pct']) / (df['cost_50pct'] - df['cost_5pct'])\n", "\n", " df['lock_factor'] = df['turnover_rate'] * (\n", " 1 - (df['cost_95pct'] - df['cost_5pct']) / (df['his_high'] - df['his_low']))\n", "\n", " df['cat_vol_break'] = (df['close'] > df['cost_85pct']) & (df['volume_ratio'] > 2)\n", "\n", " df['cost_atr_adj'] = (df['cost_95pct'] - df['cost_5pct']) / df['atr_14']\n", "\n", " # 12. 小盘股筹码集中度\n", " df['smallcap_concentration'] = (1 / df['log(circ_mv)']) * (df['cost_85pct'] - df['cost_15pct'])\n", "\n", " df['cat_golden_resonance'] = ((df['close'] > df['weight_avg']) &\n", " (df['volume_ratio'] > 1.5) &\n", " (df['winner_rate'] > 0.7))\n", "\n", " df['mv_turnover_ratio'] = df['turnover_rate'] / df['log(circ_mv)']\n", "\n", " df['mv_adjusted_volume'] = df['vol'] / df['log(circ_mv)']\n", "\n", " df['mv_weighted_turnover'] = df['turnover_rate'] * (1 / df['log(circ_mv)'])\n", "\n", " df['nonlinear_mv_volume'] = df['vol'] / df['log(circ_mv)']\n", "\n", " df['mv_volume_ratio'] = df['volume_ratio'] / df['log(circ_mv)']\n", "\n", " df['mv_momentum'] = df['turnover_rate'] * df['volume_ratio'] / df['log(circ_mv)']\n", "\n", " drop_columns = [col for col in df.columns if col.startswith('_')]\n", " df.drop(columns=drop_columns, inplace=True, errors='ignore')\n", "\n", " new_columns = [col for col in df.columns.tolist()[:] if col not in old_columns]\n", " return df, new_columns\n" ], "outputs": [], "execution_count": 29 }, { "cell_type": "code", "id": "53f86ddc0677a6d7", "metadata": { "scrolled": true, "ExecuteTime": { "end_time": "2025-04-09T14:49:01.535858Z", "start_time": "2025-04-09T14:48:56.166990Z" } }, "source": [ "from utils.factor import get_act_factor\n", "\n", "\n", "def read_industry_data(h5_filename):\n", " # 读取 H5 文件中所有的行业数据\n", " industry_data = pd.read_hdf(h5_filename, key='sw_daily', columns=[\n", " 'ts_code', 'trade_date', 'open', 'close', 'high', 'low', 'pe', 'pb', 'vol'\n", " ]) # 假设 H5 文件的键是 'industry_data'\n", " industry_data = industry_data.sort_values(by=['ts_code', 'trade_date'])\n", " industry_data = industry_data.reindex()\n", " industry_data['trade_date'] = pd.to_datetime(industry_data['trade_date'], format='%Y%m%d')\n", "\n", " grouped = industry_data.groupby('ts_code', group_keys=False)\n", " industry_data['obv'] = grouped.apply(\n", " lambda x: pd.Series(talib.OBV(x['close'].values, x['vol'].values), index=x.index)\n", " )\n", " industry_data['return_5'] = grouped['close'].apply(lambda x: x / x.shift(5) - 1)\n", " industry_data['return_20'] = grouped['close'].apply(lambda x: x / x.shift(20) - 1)\n", "\n", " industry_data = get_act_factor(industry_data, cat=False)\n", " industry_data = industry_data.sort_values(by=['trade_date', 'ts_code'])\n", "\n", " # # 计算每天每个 ts_code 的因子和当天所有 ts_code 的中位数的偏差\n", " # factor_columns = ['obv', 'return_5', 'return_20', 'act_factor1', 'act_factor2', 'act_factor3', 'act_factor4'] # 因子列\n", " #\n", " # for factor in factor_columns:\n", " # if factor in industry_data.columns:\n", " # # 计算每天每个 ts_code 的因子值与当天所有 ts_code 的中位数的偏差\n", " # industry_data[f'{factor}_deviation'] = industry_data.groupby('trade_date')[factor].transform(\n", " # lambda x: x - x.mean())\n", "\n", " industry_data['return_5_percentile'] = industry_data.groupby('trade_date')['return_5'].transform(\n", " lambda x: x.rank(pct=True))\n", " industry_data['return_20_percentile'] = industry_data.groupby('trade_date')['return_20'].transform(\n", " lambda x: x.rank(pct=True))\n", " industry_data = industry_data.drop(columns=['open', 'close', 'high', 'low', 'pe', 'pb', 'vol'])\n", "\n", " industry_data = industry_data.rename(\n", " columns={col: f'industry_{col}' for col in industry_data.columns if col not in ['ts_code', 'trade_date']})\n", "\n", " industry_data = industry_data.rename(columns={'ts_code': 'cat_l2_code'})\n", " return industry_data\n", "\n", "\n", "industry_df = read_industry_data('../../data/sw_daily.h5')\n" ], "outputs": [], "execution_count": 30 }, { "cell_type": "code", "id": "dbe2fd8021b9417f", "metadata": { "ExecuteTime": { "end_time": "2025-04-09T14:49:01.569403Z", "start_time": "2025-04-09T14:49:01.565741Z" } }, "source": [ "origin_columns = df.columns.tolist()\n", "origin_columns = [col for col in origin_columns if\n", " col not in ['turnover_rate', 'pe_ttm', 'volume_ratio', 'vol', 'pct_chg', 'l2_code', 'winner_rate']]\n", "origin_columns = [col for col in origin_columns if col not in index_data.columns]\n", "origin_columns = [col for col in origin_columns if 'cyq' not in col]\n", "print(origin_columns)" ], "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "['ts_code', 'open', 'close', 'high', 'low', 'circ_mv', 'is_st', 'up_limit', 'down_limit', 'buy_sm_vol', 'sell_sm_vol', 'buy_lg_vol', 'sell_lg_vol', 'buy_elg_vol', 'sell_elg_vol', 'net_mf_vol', 'his_low', 'his_high', 'cost_5pct', 'cost_15pct', 'cost_50pct', 'cost_85pct', 'cost_95pct', 'weight_avg', 'in_date']\n" ] } ], "execution_count": 31 }, { "cell_type": "code", "id": "85c3e3d0235ffffa", "metadata": { "ExecuteTime": { "end_time": "2025-04-09T14:50:18.282461Z", "start_time": "2025-04-09T14:49:01.600962Z" } }, "source": [ "def filter_data(df):\n", " # df = df.groupby('trade_date').apply(lambda x: x.nlargest(1000, 'act_factor1'))\n", " df = df[~df['is_st']]\n", " df = df[~df['ts_code'].str.endswith('BJ')]\n", " df = df[~df['ts_code'].str.startswith('30')]\n", " df = df[~df['ts_code'].str.startswith('68')]\n", " df = df[~df['ts_code'].str.startswith('8')]\n", " df = df[df['trade_date'] >= '20180101']\n", " if 'in_date' in df.columns:\n", " df = df.drop(columns=['in_date'])\n", " df = df.reset_index(drop=True)\n", " return df\n", "\n", "\n", "df = filter_data(df)\n", "# df = get_technical_factor(df)\n", "# df = get_act_factor(df)\n", "# df = get_money_flow_factor(df)\n", "# df = get_alpha_factor(df)\n", "# df = get_limit_factor(df)\n", "# df = get_cyp_perf_factor(df)\n", "# df = get_mv_factors(df)\n", "df, _ = get_rolling_factor(df)\n", "df, _ = get_simple_factor(df)\n", "# df = df.merge(industry_df, on=['l2_code', 'trade_date'], how='left')\n", "df = df.rename(columns={'l2_code': 'cat_l2_code'})\n", "# df = df.merge(index_data, on='trade_date', how='left')\n", "\n", "print(df.info())" ], "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Index(['ts_code', 'trade_date', 'open', 'close', 'high', 'low', 'vol',\n", " 'pct_chg', 'turnover_rate', 'pe_ttm', 'circ_mv', 'volume_ratio',\n", " 'is_st', 'up_limit', 'down_limit', 'buy_sm_vol', 'sell_sm_vol',\n", " 'buy_lg_vol', 'sell_lg_vol', 'buy_elg_vol', 'sell_elg_vol',\n", " 'net_mf_vol', 'his_low', 'his_high', 'cost_5pct', 'cost_15pct',\n", " 'cost_50pct', 'cost_85pct', 'cost_95pct', 'weight_avg', 'winner_rate',\n", " 'l2_code', '_is_positive', '_is_negative', 'cat_is_positive',\n", " '_pos_returns', '_neg_returns', '_pos_returns_sq', '_neg_returns_sq',\n", " 'upside_vol', 'downside_vol', 'vol_ratio', 'return_skew',\n", " 'return_kurtosis', 'volume_change_rate', 'cat_volume_breakout',\n", " 'turnover_deviation', 'cat_turnover_spike', 'avg_volume_ratio',\n", " 'cat_volume_ratio_breakout', 'vol_spike', 'vol_std_5', 'atr_14',\n", " 'atr_6', 'obv'],\n", " dtype='object')\n", "\n", "Index: 3142516 entries, 0 to 3142515\n", "Columns: 119 entries, ts_code to mv_momentum\n", "dtypes: bool(12), datetime64[ns](1), float64(101), int32(2), int64(1), object(2)\n", "memory usage: 2.5+ GB\n", "None\n" ] } ], "execution_count": 32 }, { "cell_type": "code", "id": "f4f16d63ad18d1bc", "metadata": { "jupyter": { "source_hidden": true }, "ExecuteTime": { "end_time": "2025-04-09T14:50:18.967476Z", "start_time": "2025-04-09T14:50:18.940918Z" } }, "source": [ "def create_deviation_within_dates(df, feature_columns):\n", " groupby_col = 'cat_l2_code' # 使用 trade_date 进行分组\n", " new_columns = {}\n", " ret_feature_columns = feature_columns[:]\n", "\n", " # 自动选择所有数值型特征\n", " num_features = [col for col in feature_columns if 'cat' not in col and 'index' not in col]\n", "\n", " # num_features = ['vol', 'pct_chg', 'turnover_rate', 'volume_ratio', 'cat_vol_spike', 'obv', 'maobv_6', 'return_5', 'return_10', 'return_20', 'std_return_5', 'std_return_15', 'std_return_90', 'std_return_90_2', 'act_factor1', 'act_factor2', 'act_factor3', 'act_factor4', 'act_factor5', 'act_factor6', 'rank_act_factor1', 'rank_act_factor2', 'rank_act_factor3', 'active_buy_volume_large', 'active_buy_volume_big', 'active_buy_volume_small', 'alpha_022', 'alpha_003', 'alpha_007', 'alpha_013']\n", " num_features = [col for col in num_features if 'cat' not in col and 'industry' not in col]\n", " num_features = [col for col in num_features if 'limit' not in col]\n", " num_features = [col for col in num_features if 'cyq' not in col]\n", "\n", " # 遍历所有数值型特征\n", " for feature in num_features:\n", " if feature == 'trade_date': # 不需要对 'trade_date' 计算偏差\n", " continue\n", "\n", " # grouped_mean = df.groupby(['trade_date'])[feature].transform('mean')\n", " # deviation_col_name = f'deviation_mean_{feature}'\n", " # new_columns[deviation_col_name] = df[feature] - grouped_mean\n", " # ret_feature_columns.append(deviation_col_name)\n", "\n", " grouped_mean = df.groupby(['trade_date', groupby_col])[feature].transform('mean')\n", " deviation_col_name = f'deviation_mean_{feature}'\n", " new_columns[deviation_col_name] = df[feature] - grouped_mean\n", " ret_feature_columns.append(deviation_col_name)\n", "\n", " # 将新计算的偏差特征与原始 DataFrame 合并\n", " df = pd.concat([df, pd.DataFrame(new_columns)], axis=1)\n", "\n", " # for feature in ['obv', 'return_20', 'act_factor1', 'act_factor2', 'act_factor3', 'act_factor4']:\n", " # df[f'deviation_industry_{feature}'] = df[feature] - df[f'industry_{feature}']\n", "\n", " return df, ret_feature_columns\n" ], "outputs": [], "execution_count": 33 }, { "cell_type": "code", "id": "40e6b68a91b30c79", "metadata": { "ExecuteTime": { "end_time": "2025-04-09T14:50:19.189566Z", "start_time": "2025-04-09T14:50:19.074786Z" } }, "source": [ "import pandas as pd\n", "\n", "from scipy.stats import ks_2samp, wasserstein_distance\n", "from sklearn.metrics import roc_auc_score\n", "from sklearn.model_selection import train_test_split\n", "from sklearn.preprocessing import StandardScaler\n", "\n", "\n", "def remove_shifted_features(train_data, feature_columns, ks_threshold=0.05, wasserstein_threshold=0.1, size=0.8,\n", " log=True):\n", " dropped_features = []\n", "\n", " all_dates = sorted(train_data['trade_date'].unique().tolist()) # 获取所有唯一的 trade_date\n", " split_date = all_dates[int(len(all_dates) * size)] # 划分点为倒数第 validation_days 天\n", " train_data_split = train_data[train_data['trade_date'] < split_date] # 训练集\n", " val_data_split = train_data[train_data['trade_date'] >= split_date] # 验证集\n", "\n", " # **统计数据漂移**\n", " numeric_columns = train_data_split.select_dtypes(include=['float64', 'int64']).columns\n", " numeric_columns = [col for col in numeric_columns if col in feature_columns]\n", " for feature in numeric_columns:\n", " ks_stat, p_value = ks_2samp(train_data_split[feature], val_data_split[feature])\n", " wasserstein_dist = wasserstein_distance(train_data_split[feature], val_data_split[feature])\n", "\n", " if p_value < ks_threshold or wasserstein_dist > wasserstein_threshold:\n", " dropped_features.append(feature)\n", " if log:\n", " print(f\"检测到 {len(dropped_features)} 个可能漂移的特征: {dropped_features}\")\n", "\n", " # **应用阈值进行最终筛选**\n", " filtered_features = [f for f in feature_columns if f not in dropped_features]\n", "\n", " return filtered_features, dropped_features\n", "\n", "\n", "def remove_outliers_label_percentile(label: pd.Series, lower_percentile: float = 0.01, upper_percentile: float = 0.99,\n", " log=True):\n", " if not (0 <= lower_percentile < upper_percentile <= 1):\n", " raise ValueError(\"Percentile values must satisfy 0 <= lower_percentile < upper_percentile <= 1.\")\n", "\n", " # Calculate lower and upper bounds based on percentiles\n", " lower_bound = label.quantile(lower_percentile)\n", " upper_bound = label.quantile(upper_percentile)\n", "\n", " # Filter out values outside the bounds\n", " filtered_label = label[(label >= lower_bound) & (label <= upper_bound)]\n", "\n", " # Print the number of removed outliers\n", " if log:\n", " print(f\"Removed {len(label) - len(filtered_label)} outliers.\")\n", " return filtered_label\n", "\n", "\n", "def calculate_risk_adjusted_target(df, days=5):\n", " df = df.sort_values(by=['ts_code', 'trade_date'])\n", "\n", " df['future_close'] = df.groupby('ts_code')['close'].shift(-days)\n", " df['future_open'] = df.groupby('ts_code')['open'].shift(-1)\n", " df['future_return'] = (df['future_close'] - df['future_open']) / df['future_open']\n", "\n", " df['future_volatility'] = df.groupby('ts_code')['future_return'].rolling(days, min_periods=1).std().reset_index(\n", " level=0, drop=True)\n", " sharpe_ratio = df['future_return'] * df['future_volatility']\n", " sharpe_ratio.replace([np.inf, -np.inf], np.nan, inplace=True)\n", "\n", " return sharpe_ratio\n", "\n", "\n", "def calculate_score(df, days=5, lambda_param=1.0):\n", " def calculate_max_drawdown(prices):\n", " peak = prices.iloc[0] # 初始化峰值\n", " max_drawdown = 0 # 初始化最大回撤\n", "\n", " for price in prices:\n", " if price > peak:\n", " peak = price # 更新峰值\n", " else:\n", " drawdown = (peak - price) / peak # 计算当前回撤\n", " max_drawdown = max(max_drawdown, drawdown) # 更新最大回撤\n", "\n", " return max_drawdown\n", "\n", " def compute_stock_score(stock_df):\n", " stock_df = stock_df.sort_values(by=['trade_date'])\n", " future_return = stock_df['future_return']\n", " # 使用已有的 pct_chg 字段计算波动率\n", " volatility = stock_df['pct_chg'].rolling(days).std().shift(-days)\n", " max_drawdown = stock_df['close'].rolling(days).apply(calculate_max_drawdown, raw=False).shift(-days)\n", " score = future_return - lambda_param * max_drawdown\n", " return score\n", "\n", " # # 确保 DataFrame 按照股票代码和交易日期排序\n", " # df = df.sort_values(by=['ts_code', 'trade_date'])\n", "\n", " # 对每个股票分别计算 score\n", " df['score'] = df.groupby('ts_code').apply(compute_stock_score).reset_index(level=0, drop=True)\n", "\n", " return df['score']\n", "\n", "\n", "def remove_highly_correlated_features(df, feature_columns, threshold=0.9):\n", " numeric_features = df[feature_columns].select_dtypes(include=[np.number]).columns.tolist()\n", " if not numeric_features:\n", " raise ValueError(\"No numeric features found in the provided data.\")\n", "\n", " corr_matrix = df[numeric_features].corr().abs()\n", " upper = corr_matrix.where(np.triu(np.ones(corr_matrix.shape), k=1).astype(bool))\n", " to_drop = [column for column in upper.columns if any(upper[column] > threshold)]\n", " remaining_features = [col for col in feature_columns if col not in to_drop\n", " or 'act' in col or 'af' in col]\n", " return remaining_features\n", "\n", "\n", "import pandas as pd\n", "from sklearn.preprocessing import StandardScaler\n", "\n", "\n", "def cross_sectional_standardization(df, features):\n", " df_sorted = df.sort_values(by='trade_date') # 按时间排序\n", " df_standardized = df_sorted.copy()\n", "\n", " for date in df_sorted['trade_date'].unique():\n", " # 获取当前时间点的数据\n", " current_data = df_standardized[df_standardized['trade_date'] == date]\n", "\n", " # 只对指定特征进行标准化\n", " scaler = StandardScaler()\n", " standardized_values = scaler.fit_transform(current_data[features])\n", "\n", " # 将标准化结果重新赋值回去\n", " df_standardized.loc[df_standardized['trade_date'] == date, features] = standardized_values\n", "\n", " return df_standardized\n", "\n", "\n", "import numpy as np\n", "import pandas as pd\n", "import statsmodels.api as sm\n", "\n", "from concurrent.futures import ProcessPoolExecutor\n", "\n", "\n", "def neutralize_manual(df, features, industry_col, mkt_cap_col):\n", " \"\"\" 手动实现简单回归以提升速度 \"\"\"\n", "\n", " for col in features:\n", " residuals = []\n", " for _, group in df.groupby(industry_col):\n", " if len(group) > 1:\n", " x = np.log(group[mkt_cap_col]) # 市值对数\n", " y = group[col] # 因子值\n", " beta = np.cov(y, x)[0, 1] / np.var(x) # 计算斜率\n", " alpha = np.mean(y) - beta * np.mean(x) # 计算截距\n", " resid = y - (alpha + beta * x) # 计算残差\n", " residuals.extend(resid)\n", " else:\n", " residuals.extend(group[col]) # 样本不足时保留原值\n", "\n", " df[col] = residuals\n", "\n", " return df\n", "\n", "\n", "import gc\n", "\n", "gc.collect()\n", "\n", "\n", "def mad_filter(df, features, n=3):\n", " for col in features:\n", " median = df[col].median()\n", " mad = np.median(np.abs(df[col] - median))\n", " upper = median + n * mad\n", " lower = median - n * mad\n", " df[col] = np.clip(df[col], lower, upper) # 截断极值\n", " return df\n", "\n", "\n", "def percentile_filter(df, features, lower_percentile=0.01, upper_percentile=0.99):\n", " for col in features:\n", " # 按日期分组计算上下百分位数\n", " lower_bound = df.groupby('trade_date')[col].transform(\n", " lambda x: x.quantile(lower_percentile)\n", " )\n", " upper_bound = df.groupby('trade_date')[col].transform(\n", " lambda x: x.quantile(upper_percentile)\n", " )\n", " # 截断超出范围的值\n", " df[col] = np.clip(df[col], lower_bound, upper_bound)\n", " return df\n", "\n", "\n", "from scipy.stats import iqr\n", "\n", "\n", "def iqr_filter(df, features):\n", " for col in features:\n", " df[col] = df.groupby('trade_date')[col].transform(\n", " lambda x: (x - x.median()) / iqr(x) if iqr(x) != 0 else x\n", " )\n", " return df\n", "\n", "\n", "def quantile_filter(df, features, lower_quantile=0.01, upper_quantile=0.99, window=60):\n", " df = df.copy()\n", " for col in features:\n", " # 计算 rolling 统计量，需要按日期进行 groupby\n", " rolling_lower = df.groupby('trade_date')[col].transform(\n", " lambda x: x.rolling(window=min(len(x), window)).quantile(lower_quantile))\n", " rolling_upper = df.groupby('trade_date')[col].transform(\n", " lambda x: x.rolling(window=min(len(x), window)).quantile(upper_quantile))\n", "\n", " # 对数据进行裁剪\n", " df[col] = np.clip(df[col], rolling_lower, rolling_upper)\n", "\n", " return df\n", "\n", "def time_series_quantile_filter(df, features, lower_quantile=0.01, upper_quantile=0.99, window=60):\n", " df = df.copy()\n", " # 确保按股票和时间排序\n", " df = df.sort_values(['ts_code', 'trade_date'])\n", " grouped = df.groupby('ts_code')\n", " for col in features:\n", " # 对每个股票的时间序列计算滚动分位数\n", " rolling_lower = grouped[col].rolling(window=window, min_periods=window // 2).quantile(lower_quantile)\n", " rolling_upper = grouped[col].rolling(window=window, min_periods=window // 2).quantile(upper_quantile)\n", " # rolling结果带有多重索引，需要对齐\n", " rolling_lower = rolling_lower.reset_index(level=0, drop=True)\n", " rolling_upper = rolling_upper.reset_index(level=0, drop=True)\n", " # 应用 clip\n", " df[col] = np.clip(df[col], rolling_lower, rolling_upper)\n", " return df\n", "\n", "def cross_sectional_quantile_filter(df, features, lower_quantile=0.01, upper_quantile=0.99):\n", " df = df.copy()\n", " grouped = df.groupby('trade_date')\n", " for col in features:\n", " # 计算每日截面的分位数边界\n", " lower_bound = grouped[col].transform(lambda x: x.quantile(lower_quantile))\n", " upper_bound = grouped[col].transform(lambda x: x.quantile(upper_quantile))\n", " # 应用 clip\n", " df[col] = np.clip(df[col], lower_bound, upper_bound)\n", " return df" ], "outputs": [], "execution_count": 34 }, { "cell_type": "code", "id": "1c46817a-b5dd-4bec-8bb4-e6e80bfd9d66", "metadata": { "ExecuteTime": { "end_time": "2025-04-09T14:50:19.205714Z", "start_time": "2025-04-09T14:50:19.197581Z" } }, "source": [ "# print(test_data.head()[['act_factor1', 'act_factor2', 'ts_code', 'trade_date']])" ], "outputs": [], "execution_count": 35 }, { "cell_type": "code", "id": "da2bb202843d9275", "metadata": { "jupyter": { "source_hidden": true }, "ExecuteTime": { "end_time": "2025-04-09T14:50:19.291833Z", "start_time": "2025-04-09T14:50:19.280710Z" } }, "source": [ "from sklearn.preprocessing import StandardScaler\n", "import lightgbm as lgb\n", "import matplotlib.pyplot as plt\n", "from sklearn.decomposition import PCA\n", "\n", "\n", "def train_light_model(train_data_df, params, feature_columns, callbacks, evals,\n", " print_feature_importance=True, num_boost_round=100,\n", " validation_days=180, use_pca=False, split_date=None): # 新增参数：validation_days\n", " # 确保数据按时间排序\n", " train_data_df = train_data_df.sort_values(by='trade_date')\n", "\n", " numeric_columns = train_data_df.select_dtypes(include=['float64', 'int64']).columns\n", " numeric_columns = [col for col in numeric_columns if col in feature_columns]\n", " # X_train.loc[:, numeric_columns] = scaler.fit_transform(X_train[numeric_columns])\n", " # X_val.loc[:, numeric_columns] = scaler.transform(X_val[numeric_columns])\n", " # train_data_df = cross_sectional_standardization(train_data_df, numeric_columns)\n", "\n", " # 去除标签为空的样本\n", " train_data_df = train_data_df.dropna(subset=['label'])\n", " # print('原始训练集大小: ', len(train_data_df))\n", "\n", " # 按时间顺序划分训练集和验证集\n", " if split_date is None:\n", " all_dates = train_data_df['trade_date'].unique() # 获取所有唯一的 trade_date\n", " if validation_days == 0:\n", " split_date = all_dates[-1]\n", " else:\n", " split_date = all_dates[-validation_days] # 划分点为倒数第 validation_days 天\n", " if validation_days == 0:\n", " train_data_split = train_data_df\n", " else:\n", " train_data_split = train_data_df[train_data_df['trade_date'] < split_date] # 训练集\n", " val_data_split = train_data_df[train_data_df['trade_date'] >= split_date] # 验证集\n", "\n", " # 打印划分结果\n", " print(f\"划分后的训练集大小: {len(train_data_split)}, 验证集大小: {len(val_data_split)}\")\n", "\n", " # 提取特征和标签\n", " X_train = train_data_split[feature_columns]\n", " y_train = train_data_split['label']\n", "\n", " # 标准化数值特征\n", " scaler = StandardScaler()\n", "\n", " # 计算每个 trade_date 内的样本数（LTR 需要 group 信息）\n", " train_groups = train_data_split.groupby('trade_date').size().tolist()\n", " val_groups = val_data_split.groupby('trade_date').size().tolist()\n", "\n", " # 处理类别特征\n", " categorical_feature = [col for col in feature_columns if 'cat' in col]\n", "\n", " # 计算权重（基于时间）\n", " # trade_date = train_data_split['trade_date'] # 交易日期\n", " # weights = (trade_date - trade_date.min()).dt.days / (trade_date.max() - trade_date.min()).days + 1\n", " # weights = train_data_split.groupby('trade_date')['std_return_5'].transform(\n", " # lambda x: x / x.mean()\n", " # )\n", " ud = sorted(train_data_split[\"trade_date\"].unique().tolist())\n", " date_weights = {date: weight * weight for date, weight in zip(ud, np.linspace(1, 10, len(ud)))}\n", " params['weight'] = train_data_split[\"trade_date\"].map(date_weights).tolist()\n", "\n", " train_dataset = lgb.Dataset(\n", " X_train, label=y_train, group=train_groups,\n", " categorical_feature=categorical_feature\n", " )\n", "\n", " if validation_days > 0:\n", " X_val = val_data_split[feature_columns]\n", " y_val = val_data_split['label']\n", " val_groups = val_data_split.groupby('trade_date').size().tolist()\n", " val_dataset = lgb.Dataset(\n", " X_val, label=y_val, group=val_groups,\n", " categorical_feature=categorical_feature\n", " )\n", " # 训练模型\n", " model = lgb.train(\n", " params, train_dataset, num_boost_round=num_boost_round,\n", " valid_sets=[train_dataset, val_dataset], valid_names=['train', 'valid'],\n", " callbacks=callbacks\n", " )\n", " else:\n", " model = lgb.train(\n", " params, train_dataset, num_boost_round=num_boost_round, callbacks=callbacks\n", " )\n", "\n", " # 打印特征重要性（如果需要）\n", " if print_feature_importance:\n", " lgb.plot_metric(evals)\n", " lgb.plot_importance(model, importance_type='split', max_num_features=20)\n", " plt.show()\n", "\n", " return model, scaler, None" ], "outputs": [], "execution_count": 36 }, { "cell_type": "code", "id": "ff19e3f1e051a489", "metadata": { "ExecuteTime": { "end_time": "2025-04-09T14:51:32.646948Z", "start_time": "2025-04-09T14:50:19.359656Z" } }, "source": [ "\n", "days = 2\n", "df = df.sort_values(by=['ts_code', 'trade_date'])\n", "# df['future_return'] = df.groupby('ts_code', group_keys=False)['close'].apply(lambda x: x.shift(-days) / x - 1)\n", "df['future_return'] = (df.groupby('ts_code')['close'].shift(-days) - df.groupby('ts_code')['open'].shift(-1)) / \\\n", " df.groupby('ts_code')['open'].shift(-1)\n", "df['future_volatility'] = (\n", " df.groupby('ts_code')['pct_chg']\n", " .transform(lambda x: x.rolling(days).std().shift(-days))\n", ")\n", "df['future_score'] = calculate_score(df, days=2, lambda_param=0.3)\n", "df['label'] = df.groupby('trade_date', group_keys=False)['future_score'].transform(\n", " lambda x: pd.qcut(x, q=20, labels=False, duplicates='drop')\n", ")\n", "# df['label'] = df.groupby('trade_date', group_keys=False)['future_score'].transform(\n", "# lambda x: pd.qcut(x.rank(method='first'), q=20, labels=False, duplicates='raise')\n", "# )\n", "# df['future_score'] = (\n", "# 0.7 * df['future_return']\n", "# * 0.3 * df['future_volatility']\n", "# )" ], "outputs": [], "execution_count": 37 }, { "cell_type": "code", "id": "27dba27b2e108316", "metadata": { "ExecuteTime": { "end_time": "2025-04-09T14:51:45.505400Z", "start_time": "2025-04-09T14:51:32.722125Z" } }, "source": [ "def select_pre_zt_stocks_dynamic(stock_df):\n", " def select_stocks(group):\n", " max_stocks = 150\n", " initial_data = group.nlargest(100, 'return_20')\n", " unique_labels = initial_data['label'].nunique()\n", "\n", " if unique_labels >= 20 or unique_labels == 0: # 包含标签种类为0的情况\n", " return initial_data\n", "\n", " for i in range(110, max_stocks + 1, 10):\n", " data = group.nlargest(i, 'return_20')\n", " unique_labels = data['label'].nunique()\n", " if unique_labels >= 20:\n", " return data\n", "\n", " return group.nlargest(max_stocks, 'return_20') # 如果循环结束仍未找到足够标签，则返回最大数量的股票\n", "\n", " stock_df = stock_df.groupby('trade_date', group_keys=False).apply(select_stocks)\n", " return stock_df\n", "\n", "\n", "pdf = select_pre_zt_stocks_dynamic(df[(df['trade_date'] >= '2022-01-01') & (df['trade_date'] <= '2029-04-07')])\n", "print(pdf['trade_date'].max())\n", "\n", "# pdf['label'] = pdf.groupby('trade_date', group_keys=False)['future_score'].transform(\n", "# lambda x: pd.qcut(x, q=20, labels=False, duplicates='drop')\n", "# )" ], "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "2025-04-07 00:00:00\n" ] } ], "execution_count": 38 }, { "cell_type": "code", "id": "ca96fb81e17c4a90", "metadata": { "ExecuteTime": { "end_time": "2025-04-09T14:51:46.074673Z", "start_time": "2025-04-09T14:51:45.512822Z" } }, "source": [ "pdf = pdf.merge(industry_df, on=['cat_l2_code', 'trade_date'], how='left')\n", "pdf = pdf.replace([np.inf, -np.inf], np.nan)\n", "\n", "feature_columns = [col for col in pdf.columns if col in pdf.columns]\n", "feature_columns = [col for col in feature_columns if col not in ['trade_date',\n", " 'ts_code',\n", " 'label']]\n", "feature_columns = [col for col in feature_columns if 'future' not in col]\n", "feature_columns = [col for col in feature_columns if 'label' not in col]\n", "feature_columns = [col for col in feature_columns if 'score' not in col]\n", "feature_columns = [col for col in feature_columns if 'gen' not in col]\n", "feature_columns = [col for col in feature_columns if 'pe_ttm' not in col]\n", "feature_columns = [col for col in feature_columns if 'volatility' not in col]\n", "feature_columns = [col for col in feature_columns if 'cat_l2_code' not in col]\n", "feature_columns = [col for col in feature_columns if col not in origin_columns]\n", "feature_columns = [col for col in feature_columns if not col.startswith('_')]\n", "# feature_columns = [col for col in feature_columns if col not in ['ts_code', 'trade_date', 'vol_std_5', 'cov', 'delta_cov', 'alpha_22_improved', 'alpha_007', 'consecutive_up_limit', 'mv_volatility', 'volume_growth', 'mv_growth', 'arbr']]\n", "\n", "print(feature_columns)\n", "numeric_columns = pdf.select_dtypes(include=['float64', 'int64']).columns\n", "numeric_columns = [col for col in numeric_columns if col in feature_columns]\n", "\n", "\n", "# filter_index = pdf['future_volatility'].between(pdf['future_volatility'].quantile(0.01),\n", "# pdf['future_volatility'].quantile(0.99)) | filter_index" ], "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "['vol', 'pct_chg', 'turnover_rate', 'volume_ratio', 'winner_rate', 'cat_is_positive', 'upside_vol', 'downside_vol', 'vol_ratio', 'return_skew', 'return_kurtosis', 'volume_change_rate', 'cat_volume_breakout', 'turnover_deviation', 'cat_turnover_spike', 'avg_volume_ratio', 'cat_volume_ratio_breakout', 'vol_spike', 'vol_std_5', 'atr_14', 'atr_6', 'obv', 'maobv_6', 'rsi_3', 'return_5', 'return_20', 'std_return_5', 'std_return_90', 'std_return_90_2', 'act_factor1', 'act_factor2', 'act_factor3', 'act_factor4', 'rank_act_factor1', 'rank_act_factor2', 'rank_act_factor3', 'log(circ_mv)', 'cov', 'delta_cov', 'alpha_22_improved', 'alpha_003', 'alpha_007', 'alpha_013', 'cat_up_limit', 'cat_down_limit', 'up_limit_count_10d', 'down_limit_count_10d', 'consecutive_up_limit', 'vol_break', 'weight_roc5', 'smallcap_concentration', 'cost_stability', 'high_cost_break_days', 'liquidity_risk', 'turnover_std', 'volume_growth', 'mv_growth', 'arbr', 'momentum_factor', 'resonance_factor', 'log_close', 'cat_vol_spike', 'up', 'down', 'obv-maobv_6', 'std_return_5 / std_return_90', 'std_return_90 - std_return_90_2', 'cat_af2', 'cat_af3', 'cat_af4', 'act_factor5', 'act_factor6', 'active_buy_volume_large', 'active_buy_volume_big', 'active_buy_volume_small', 'buy_lg_vol_minus_sell_lg_vol', 'buy_elg_vol_minus_sell_elg_vol', 'ctrl_strength', 'low_cost_dev', 'asymmetry', 'lock_factor', 'cat_vol_break', 'cost_atr_adj', 'cat_golden_resonance', 'mv_turnover_ratio', 'mv_adjusted_volume', 'mv_weighted_turnover', 'nonlinear_mv_volume', 'mv_volume_ratio', 'mv_momentum', 'industry_obv', 'industry_return_5', 'industry_return_20', 'industry__ema_5', 'industry__ema_13', 'industry__ema_20', 'industry__ema_60', 'industry_act_factor1', 'industry_act_factor2', 'industry_act_factor3', 'industry_act_factor4', 'industry_act_factor5', 'industry_act_factor6', 'industry_rank_act_factor1', 'industry_rank_act_factor2', 'industry_rank_act_factor3', 'industry_return_5_percentile', 'industry_return_20_percentile']\n" ] } ], "execution_count": 39 }, { "cell_type": "code", "id": "81d4570663ae21d7", "metadata": { "ExecuteTime": { "end_time": "2025-04-09T14:52:38.849524Z", "start_time": "2025-04-09T14:51:46.113849Z" } }, "source": [ "\n", "# pdf = time_series_quantile_filter(pdf, numeric_columns)\n", "pdf = cross_sectional_quantile_filter(pdf, numeric_columns)\n", "pdf = cross_sectional_standardization(pdf, numeric_columns)\n", "\n", "pdf = pdf.sort_values(by=['ts_code', 'trade_date'])\n", "\n", "filter_index = pdf['future_return'].between(pdf['future_return'].quantile(0.01), pdf['future_return'].quantile(0.99))\n", "\n", "feature_columns = remove_highly_correlated_features(pdf, feature_columns)\n", "\n", "# print(pdf[pdf['trade_date'] == '2025-03-26'][['ts_code', 'trade_date', 'pct_chg']])\n", "#\n", "# pdf = select_pre_zt_stocks_dynamic(df[(df['trade_date'] >= '2022-03-26') & (df['trade_date'] <= '2029-03-26')])\n", "# pdf = pdf.sort_values(['ts_code', 'trade_date'])\n", "# pdf = pdf.replace([np.inf, -np.inf], np.nan)\n", "#\n", "# # pdf = time_series_quantile_filter(pdf, numeric_columns)\n", "#\n", "# pdf = cross_sectional_standardization(pdf, numeric_columns)\n", "#\n", "# print(pdf[pdf['trade_date'] == '2025-03-26'][['ts_code', 'trade_date', 'pct_chg']])" ], "outputs": [], "execution_count": 40 }, { "cell_type": "code", "id": "92428d543f4727ad", "metadata": { "ExecuteTime": { "end_time": "2025-04-09T14:52:38.972693Z", "start_time": "2025-04-09T14:52:38.887044Z" } }, "source": [ "# print('train data size: ', len(train_data))\n", "\n", "label_gain = list(range(len(df['label'].unique())))\n", "label_gain = [gain * gain for gain in label_gain]\n", "light_params = {\n", " 'label_gain': label_gain,\n", " 'objective': 'lambdarank',\n", " 'metric': 'ndcg',\n", " 'learning_rate': 0.03,\n", " 'num_leaves': 32,\n", " # 'min_data_in_leaf': 128,\n", " 'max_depth': 8,\n", " 'max_bin': 32,\n", " 'feature_fraction': 0.7,\n", " # 'bagging_fraction': 0.7,\n", " 'bagging_freq': 5,\n", " 'lambda_l1': 0.1,\n", " 'lambda_l2': 0.1,\n", " 'boosting': 'gbdt',\n", " 'verbosity': -1,\n", " 'extra_trees': True,\n", " 'max_position': 5,\n", " 'ndcg_at': 1,\n", " 'quant_train_renew_leaf': True,\n", " 'lambdarank_truncation_level': 3,\n", " # 'lambdarank_position_bias_regularization': 1,\n", " 'seed': 7\n", "}\n", "evals = {}\n", "\n", "gc.collect()" ], "outputs": [ { "data": { "text/plain": [ "0" ] }, "execution_count": 41, "metadata": {}, "output_type": "execute_result" } ], "execution_count": 41 }, { "cell_type": "code", "id": "8f134d435f71e9e2", "metadata": { "ExecuteTime": { "end_time": "2025-04-09T14:52:39.087454Z", "start_time": "2025-04-09T14:52:39.016514Z" } }, "source": [ "gc.collect()\n", "\n", "\n", "def rolling_train_predict(df, train_days, test_days, feature_columns_origin, days=5, use_pca=False, validation_days=60,\n", " filter_index=None, params=None):\n", " # 1. 按照交易日期排序\n", " unique_dates = df[df['trade_date'] >= '2020-01-01']['trade_date'].unique().tolist()\n", " unique_dates = sorted(unique_dates)\n", " n = len(unique_dates)\n", "\n", " # 2. 计算需要跳过的天数，使后续窗口对齐\n", " extra_days = (n - train_days) % test_days\n", " start_index = extra_days # 从此索引开始滚动\n", "\n", " predictions_list = []\n", "\n", " for start in range(start_index, n - train_days - test_days + 1, test_days):\n", "\n", " train_dates = unique_dates[start: start + train_days]\n", " test_dates = unique_dates[start + train_days: start + train_days + test_days]\n", "\n", " # 根据日期筛选数据\n", " # train_data = df[df['trade_date'].isin(train_dates)]\n", " train_data = df[filter_index & df['trade_date'].isin(train_dates)]\n", " test_data = df[df['trade_date'].isin(test_dates)]\n", "\n", " train_data = train_data.sort_values('trade_date')\n", " test_data = test_data.sort_values('trade_date')\n", "\n", " feature_columns, _ = remove_shifted_features(train_data, feature_columns_origin, size=0.8, log=False)\n", "\n", " train_data = train_data.dropna(subset=feature_columns)\n", " train_data = train_data.dropna(subset=['label'])\n", " train_data = train_data.reset_index(drop=True)\n", "\n", " # print(test_data.tail())\n", " test_data = test_data.dropna(subset=feature_columns)\n", " # test_data = test_data.dropna(subset=['label'])\n", " test_data = test_data.reset_index(drop=True)\n", "\n", " # print(len(train_data))\n", " # print(f\"最小日期: {train_data['trade_date'].min().strftime('%Y-%m-%d')}\")\n", " print(f\"train_data最大日期: {train_data['trade_date'].max().strftime('%Y-%m-%d')}\")\n", " # # print(len(test_data))\n", " # print(f\"最小日期: {test_data['trade_date'].min().strftime('%Y-%m-%d')}\")\n", " print(f\"test_data最大日期: {test_data['trade_date'].max().strftime('%Y-%m-%d')}\")\n", "\n", " cat_columns = [col for col in df.columns if col.startswith('cat')]\n", " for col in cat_columns:\n", " train_data[col] = train_data[col].astype('category')\n", " test_data[col] = test_data[col].astype('category')\n", "\n", " label_gain = list(range(len(train_data['label'].unique())))\n", " label_gain = [(gain + 1) * (gain + 1) for gain in label_gain]\n", " params['label_gain'] = label_gain\n", "\n", " # ud = train_data[\"trade_date\"].unique()\n", " # date_weights = {date: weight for date, weight in zip(ud, np.linspace(1, 2, len(unique_dates)))}\n", " # params['weight'] = train_data[\"trade_date\"].map(date_weights).tolist()\n", "\n", " # print(f'feature_columns: {feature_columns}')\n", " # feature_contri = [2 if feat.startswith('act_factor') else 1 for feat in feature_columns]\n", " # params['feature_contri'] = feature_contri\n", " model, _, _ = train_light_model(train_data.dropna(subset=['label']),\n", " params, feature_columns,\n", " [lgb.log_evaluation(period=100),\n", " lgb.callback.record_evaluation(evals),\n", " # lgb.early_stopping(100, first_metric_only=True)\n", " ], evals,\n", " num_boost_round=100, validation_days=validation_days,\n", " print_feature_importance=False, use_pca=False)\n", "\n", " score_df = test_data.copy()\n", " score_df['score'] = model.predict(score_df[feature_columns])\n", " score_df = score_df.loc[score_df.groupby('trade_date')['score'].idxmax()]\n", " score_df = score_df[['trade_date', 'score', 'ts_code']]\n", " predictions_list.append(score_df)\n", "\n", " final_predictions = pd.concat(predictions_list, ignore_index=True)\n", " return final_predictions\n", "\n", "\n", "from concurrent.futures import ThreadPoolExecutor\n", "\n", "\n", "def worker(train_data, test_data, train_days, test_days, feature_columns_origin, unique_dates, start, filter_index,\n", " validation_days, params=None):\n", " train_data = train_data.sort_values('trade_date')\n", " test_data = test_data.sort_values('trade_date')\n", "\n", " # feature_columns, _ = remove_shifted_features(train_data, feature_columns_origin, size=0.8, log=False)\n", "\n", " train_data = train_data.dropna(subset=feature_columns + ['label'])\n", " train_data = train_data.reset_index(drop=True)\n", "\n", " test_data = test_data.dropna(subset=feature_columns)\n", " test_data = test_data.reset_index(drop=True)\n", "\n", " cat_columns = [col for col in df.columns if col.startswith('cat')]\n", " for col in cat_columns:\n", " if col in train_data.columns:\n", " train_data[col] = train_data[col].astype('category')\n", " if col in test_data.columns:\n", " test_data[col] = test_data[col].astype('category')\n", "\n", " label_gain = list(range(len(train_data['label'].unique())))\n", " label_gain = [(gain + 1) * (gain + 1) for gain in label_gain]\n", " current_light_params = params.copy()\n", " current_light_params['label_gain'] = label_gain\n", "\n", " model, _, _ = train_light_model(train_data.dropna(subset=['label']),\n", " current_light_params, feature_columns,\n", " [lgb.log_evaluation(period=100),\n", " lgb.callback.record_evaluation(evals),\n", " # lgb.early_stopping(100, first_metric_only=True)\n", " ], evals,\n", " num_boost_round=100, validation_days=validation_days,\n", " print_feature_importance=False, use_pca=False)\n", "\n", " score_df = test_data.copy()\n", " score_df['score'] = model.predict(score_df[feature_columns])\n", " if not score_df.empty:\n", " score_df = score_df.loc[score_df.groupby('trade_date')['score'].idxmax()]\n", " score_df = score_df[['trade_date', 'score', 'ts_code']]\n", " return score_df\n", " return None\n", "\n", "\n", "def rolling_train_predict_mt(df, train_days, test_days, feature_columns_origin, days=5, use_pca=False,\n", " validation_days=60, filter_index=None, num_threads=4):\n", " unique_dates = df['trade_date'].unique().tolist()\n", " unique_dates = sorted(unique_dates)\n", " n = len(unique_dates)\n", " extra_days = (n - train_days) % test_days\n", " start_index = extra_days\n", "\n", " predictions_list = []\n", " with ThreadPoolExecutor(max_workers=num_threads) as executor:\n", " futures = []\n", " for start in range(start_index, n - train_days - test_days + 1, test_days):\n", " train_dates = unique_dates[start: start + train_days]\n", " test_dates = unique_dates[start + train_days: start + train_days + test_days]\n", "\n", " train_data = df[filter_index & df['trade_date'].isin(train_dates)].copy()\n", " test_data = df[df['trade_date'].isin(test_dates)].copy()\n", " future = executor.submit(worker, train_data, test_data, train_days, test_days, feature_columns_origin,\n", " unique_dates, start, filter_index, validation_days)\n", " futures.append(future)\n", "\n", " for future in futures:\n", " result = future.result()\n", " if result is not None:\n", " predictions_list.append(result)\n", "\n", " final_predictions = pd.concat(predictions_list, ignore_index=True)\n", " return final_predictions\n" ], "outputs": [], "execution_count": 42 }, { "cell_type": "code", "id": "63235069-dc59-48fb-961a-e80373e41a61", "metadata": { "editable": true, "scrolled": true, "slideshow": { "slide_type": "" }, "tags": [], "ExecuteTime": { "end_time": "2025-04-09T14:53:52.944545Z", "start_time": "2025-04-09T14:52:39.125461Z" } }, "source": [ "\n", "gc.collect()\n", "\n", "print('finish')\n", "# qdf = qdf[qdf['trade_date'] >= '2022-01-01']\n", "\n", "final_predictions = rolling_train_predict(\n", " pdf[(pdf['trade_date'] >= '2022-12-01') & (pdf['trade_date'] <= '2029-03-26')], 5, 1, feature_columns,\n", " days=days, validation_days=0, filter_index=filter_index, params=light_params)\n", "final_predictions.to_csv('predictions_test.tsv', index=False)\n" ], "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "finish\n", "train_data最大日期: 2022-12-07\n", "test_data最大日期: 2022-12-08\n", "划分后的训练集大小: 487, 验证集大小: 90\n", "train_data最大日期: 2022-12-08\n", "test_data最大日期: 2022-12-09\n", "划分后的训练集大小: 497, 验证集大小: 104\n", "train_data最大日期: 2022-12-09\n", "test_data最大日期: 2022-12-12\n", "划分后的训练集大小: 481, 验证集大小: 94\n", "train_data最大日期: 2022-12-12\n", "test_data最大日期: 2022-12-13\n", "划分后的训练集大小: 492, 验证集大小: 102\n", "train_data最大日期: 2022-12-13\n", "test_data最大日期: 2022-12-14\n", "划分后的训练集大小: 531, 验证集大小: 141\n", "train_data最大日期: 2022-12-14\n", "test_data最大日期: 2022-12-15\n", "划分后的训练集大小: 536, 验证集大小: 95\n", "train_data最大日期: 2022-12-15\n", "test_data最大日期: 2022-12-16\n", "划分后的训练集大小: 568, 验证集大小: 136\n", "train_data最大日期: 2022-12-16\n", "test_data最大日期: 2022-12-19\n", "划分后的训练集大小: 566, 验证集大小: 92\n", "train_data最大日期: 2022-12-19\n", "test_data最大日期: 2022-12-20\n", "划分后的训练集大小: 600, 验证集大小: 136\n", "train_data最大日期: 2022-12-20\n", "test_data最大日期: 2022-12-21\n", "划分后的训练集大小: 595, 验证集大小: 136\n", "train_data最大日期: 2022-12-21\n", "test_data最大日期: 2022-12-22\n", "划分后的训练集大小: 593, 验证集大小: 93\n", "train_data最大日期: 2022-12-22\n", "test_data最大日期: 2022-12-23\n", "划分后的训练集大小: 577, 验证集大小: 120\n", "train_data最大日期: 2022-12-23\n", "test_data最大日期: 2022-12-26\n", "划分后的训练集大小: 578, 验证集大小: 93\n", "train_data最大日期: 2022-12-26\n", "test_data最大日期: 2022-12-27\n", "划分后的训练集大小: 537, 验证集大小: 95\n", "train_data最大日期: 2022-12-27\n", "test_data最大日期: 2022-12-28\n", "划分后的训练集大小: 492, 验证集大小: 91\n", "train_data最大日期: 2022-12-28\n", "test_data最大日期: 2022-12-29\n", "划分后的训练集大小: 491, 验证集大小: 92\n", "train_data最大日期: 2022-12-29\n", "test_data最大日期: 2022-12-30\n", "划分后的训练集大小: 509, 验证集大小: 138\n", "train_data最大日期: 2022-12-30\n", "test_data最大日期: 2023-01-03\n", "划分后的训练集大小: 548, 验证集大小: 132\n", "train_data最大日期: 2023-01-03\n", "test_data最大日期: 2023-01-04\n", "划分后的训练集大小: 596, 验证集大小: 143\n", "train_data最大日期: 2023-01-04\n", "test_data最大日期: 2023-01-05\n", "划分后的训练集大小: 641, 验证集大小: 136\n", "train_data最大日期: 2023-01-05\n", "test_data最大日期: 2023-01-06\n", "划分后的训练集大小: 692, 验证集大小: 143\n", "train_data最大日期: 2023-01-06\n", "test_data最大日期: 2023-01-09\n", "划分后的训练集大小: 650, 验证集大小: 96\n", "train_data最大日期: 2023-01-09\n", "test_data最大日期: 2023-01-10\n", "划分后的训练集大小: 613, 验证集大小: 95\n", "train_data最大日期: 2023-01-10\n", "test_data最大日期: 2023-01-11\n", "划分后的训练集大小: 574, 验证集大小: 104\n", "train_data最大日期: 2023-01-11\n", "test_data最大日期: 2023-01-12\n", "划分后的训练集大小: 530, 验证集大小: 92\n", "train_data最大日期: 2023-01-12\n", "test_data最大日期: 2023-01-13\n", "划分后的训练集大小: 529, 验证集大小: 142\n", "train_data最大日期: 2023-01-13\n", "test_data最大日期: 2023-01-16\n", "划分后的训练集大小: 532, 验证集大小: 99\n", "train_data最大日期: 2023-01-16\n", "test_data最大日期: 2023-01-17\n", "划分后的训练集大小: 548, 验证集大小: 111\n", "train_data最大日期: 2023-01-17\n", "test_data最大日期: 2023-01-18\n", "划分后的训练集大小: 572, 验证集大小: 128\n", "train_data最大日期: 2023-01-18\n", "test_data最大日期: 2023-01-19\n", "划分后的训练集大小: 594, 验证集大小: 114\n", "train_data最大日期: 2023-01-19\n", "test_data最大日期: 2023-01-20\n", "划分后的训练集大小: 587, 验证集大小: 135\n", "train_data最大日期: 2023-01-20\n", "test_data最大日期: 2023-01-30\n", "划分后的训练集大小: 584, 验证集大小: 96\n", "train_data最大日期: 2023-01-30\n", "test_data最大日期: 2023-01-31\n", "划分后的训练集大小: 598, 验证集大小: 125\n", "train_data最大日期: 2023-01-31\n", "test_data最大日期: 2023-02-01\n", "划分后的训练集大小: 561, 验证集大小: 91\n", "train_data最大日期: 2023-02-01\n", "test_data最大日期: 2023-02-02\n", "划分后的训练集大小: 552, 验证集大小: 105\n", "train_data最大日期: 2023-02-02\n", "test_data最大日期: 2023-02-03\n", "划分后的训练集大小: 517, 验证集大小: 100\n", "train_data最大日期: 2023-02-03\n", "test_data最大日期: 2023-02-06\n", "划分后的训练集大小: 543, 验证集大小: 122\n", "train_data最大日期: 2023-02-06\n", "test_data最大日期: 2023-02-07\n", "划分后的训练集大小: 535, 验证集大小: 117\n", "train_data最大日期: 2023-02-07\n", "test_data最大日期: 2023-02-08\n", "划分后的训练集大小: 589, 验证集大小: 145\n", "train_data最大日期: 2023-02-08\n", "test_data最大日期: 2023-02-09\n", "划分后的训练集大小: 629, 验证集大小: 145\n", "train_data最大日期: 2023-02-09\n", "test_data最大日期: 2023-02-10\n", "划分后的训练集大小: 675, 验证集大小: 146\n", "train_data最大日期: 2023-02-10\n", "test_data最大日期: 2023-02-13\n", "划分后的训练集大小: 679, 验证集大小: 126\n", "train_data最大日期: 2023-02-13\n", "test_data最大日期: 2023-02-14\n", "划分后的训练集大小: 697, 验证集大小: 135\n", "train_data最大日期: 2023-02-14\n", "test_data最大日期: 2023-02-15\n", "划分后的训练集大小: 698, 验证集大小: 146\n", "train_data最大日期: 2023-02-15\n", "test_data最大日期: 2023-02-16\n", "划分后的训练集大小: 698, 验证集大小: 145\n", "train_data最大日期: 2023-02-16\n", "test_data最大日期: 2023-02-17\n", "划分后的训练集大小: 696, 验证集大小: 144\n", "train_data最大日期: 2023-02-17\n", "test_data最大日期: 2023-02-20\n", "划分后的训练集大小: 665, 验证集大小: 95\n", "train_data最大日期: 2023-02-20\n", "test_data最大日期: 2023-02-21\n", "划分后的训练集大小: 665, 验证集大小: 135\n", "train_data最大日期: 2023-02-21\n", "test_data最大日期: 2023-02-22\n", "划分后的训练集大小: 615, 验证集大小: 96\n", "train_data最大日期: 2023-02-22\n", "test_data最大日期: 2023-02-23\n", "划分后的训练集大小: 604, 验证集大小: 134\n", "train_data最大日期: 2023-02-23\n", "test_data最大日期: 2023-02-24\n", "划分后的训练集大小: 559, 验证集大小: 99\n", "train_data最大日期: 2023-02-24\n", "test_data最大日期: 2023-02-27\n", "划分后的训练集大小: 611, 验证集大小: 147\n", "train_data最大日期: 2023-02-27\n", "test_data最大日期: 2023-02-28\n", "划分后的训练集大小: 623, 验证集大小: 147\n", "train_data最大日期: 2023-02-28\n", "test_data最大日期: 2023-03-01\n", "划分后的训练集大小: 624, 验证集大小: 97\n", "train_data最大日期: 2023-03-01\n", "test_data最大日期: 2023-03-02\n", "划分后的训练集大小: 584, 验证集大小: 94\n", "train_data最大日期: 2023-03-02\n", "test_data最大日期: 2023-03-03\n", "划分后的训练集大小: 579, 验证集大小: 94\n", "train_data最大日期: 2023-03-03\n", "test_data最大日期: 2023-03-06\n", "划分后的训练集大小: 577, 验证集大小: 145\n", "train_data最大日期: 2023-03-06\n", "test_data最大日期: 2023-03-07\n", "划分后的训练集大小: 529, 验证集大小: 99\n", "train_data最大日期: 2023-03-07\n", "test_data最大日期: 2023-03-08\n", "划分后的训练集大小: 548, 验证集大小: 116\n", "train_data最大日期: 2023-03-08\n", "test_data最大日期: 2023-03-09\n", "划分后的训练集大小: 570, 验证集大小: 116\n", "train_data最大日期: 2023-03-09\n", "test_data最大日期: 2023-03-10\n", "划分后的训练集大小: 623, 验证集大小: 147\n", "train_data最大日期: 2023-03-10\n", "test_data最大日期: 2023-03-13\n", "划分后的训练集大小: 623, 验证集大小: 145\n", "train_data最大日期: 2023-03-13\n", "test_data最大日期: 2023-03-14\n", "划分后的训练集大小: 632, 验证集大小: 108\n", "train_data最大日期: 2023-03-14\n", "test_data最大日期: 2023-03-15\n", "划分后的训练集大小: 661, 验证集大小: 145\n", "train_data最大日期: 2023-03-15\n", "test_data最大日期: 2023-03-16\n", "划分后的训练集大小: 673, 验证集大小: 128\n", "train_data最大日期: 2023-03-16\n", "test_data最大日期: 2023-03-17\n", "划分后的训练集大小: 672, 验证集大小: 146\n", "train_data最大日期: 2023-03-17\n", "test_data最大日期: 2023-03-20\n", "划分后的训练集大小: 634, 验证集大小: 107\n", "train_data最大日期: 2023-03-20\n", "test_data最大日期: 2023-03-21\n", "划分后的训练集大小: 653, 验证集大小: 127\n", "train_data最大日期: 2023-03-21\n", "test_data最大日期: 2023-03-22\n", "划分后的训练集大小: 604, 验证集大小: 96\n", "train_data最大日期: 2023-03-22\n", "test_data最大日期: 2023-03-23\n", "划分后的训练集大小: 623, 验证集大小: 147\n", "train_data最大日期: 2023-03-23\n", "test_data最大日期: 2023-03-24\n", "划分后的训练集大小: 619, 验证集大小: 142\n", "train_data最大日期: 2023-03-24\n", "test_data最大日期: 2023-03-27\n", "划分后的训练集大小: 637, 验证集大小: 125\n", "train_data最大日期: 2023-03-27\n", "test_data最大日期: 2023-03-28\n", "划分后的训练集大小: 607, 验证集大小: 97\n", "train_data最大日期: 2023-03-28\n", "test_data最大日期: 2023-03-29\n", "划分后的训练集大小: 607, 验证集大小: 96\n", "train_data最大日期: 2023-03-29\n", "test_data最大日期: 2023-03-30\n", "划分后的训练集大小: 586, 验证集大小: 126\n", "train_data最大日期: 2023-03-30\n", "test_data最大日期: 2023-03-31\n", "划分后的训练集大小: 540, 验证集大小: 96\n", "train_data最大日期: 2023-03-31\n", "test_data最大日期: 2023-04-03\n", "划分后的训练集大小: 559, 验证集大小: 144\n", "train_data最大日期: 2023-04-03\n", "test_data最大日期: 2023-04-04\n", "划分后的训练集大小: 605, 验证集大小: 143\n", "train_data最大日期: 2023-04-04\n", "test_data最大日期: 2023-04-06\n", "划分后的训练集大小: 621, 验证集大小: 112\n", "train_data最大日期: 2023-04-06\n", "test_data最大日期: 2023-04-07\n", "划分后的训练集大小: 636, 验证集大小: 141\n", "train_data最大日期: 2023-04-07\n", "test_data最大日期: 2023-04-10\n", "划分后的训练集大小: 686, 验证集大小: 146\n", "train_data最大日期: 2023-04-10\n", "test_data最大日期: 2023-04-11\n", "划分后的训练集大小: 684, 验证集大小: 142\n", "train_data最大日期: 2023-04-11\n", "test_data最大日期: 2023-04-12\n", "划分后的训练集大小: 654, 验证集大小: 113\n", "train_data最大日期: 2023-04-12\n", "test_data最大日期: 2023-04-13\n", "划分后的训练集大小: 676, 验证集大小: 134\n", "train_data最大日期: 2023-04-13\n", "test_data最大日期: 2023-04-14\n", "划分后的训练集大小: 677, 验证集大小: 142\n", "train_data最大日期: 2023-04-14\n", "test_data最大日期: 2023-04-17\n", "划分后的训练集大小: 664, 验证集大小: 133\n", "train_data最大日期: 2023-04-17\n", "test_data最大日期: 2023-04-18\n", "划分后的训练集大小: 619, 验证集大小: 97\n", "train_data最大日期: 2023-04-18\n", "test_data最大日期: 2023-04-19\n", "划分后的训练集大小: 643, 验证集大小: 137\n", "train_data最大日期: 2023-04-19\n", "test_data最大日期: 2023-04-20\n", "划分后的训练集大小: 605, 验证集大小: 96\n", "train_data最大日期: 2023-04-20\n", "test_data最大日期: 2023-04-21\n", "划分后的训练集大小: 608, 验证集大小: 145\n", "train_data最大日期: 2023-04-21\n", "test_data最大日期: 2023-04-24\n", "划分后的训练集大小: 608, 验证集大小: 133\n", "train_data最大日期: 2023-04-24\n", "test_data最大日期: 2023-04-25\n", "划分后的训练集大小: 655, 验证集大小: 144\n", "train_data最大日期: 2023-04-25\n", "test_data最大日期: 2023-04-26\n", "划分后的训练集大小: 661, 验证集大小: 143\n", "train_data最大日期: 2023-04-26\n", "test_data最大日期: 2023-04-27\n", "划分后的训练集大小: 662, 验证集大小: 97\n", "train_data最大日期: 2023-04-27\n", "test_data最大日期: 2023-04-28\n", "划分后的训练集大小: 639, 验证集大小: 122\n", "train_data最大日期: 2023-04-28\n", "test_data最大日期: 2023-05-04\n", "划分后的训练集大小: 650, 验证集大小: 144\n", "train_data最大日期: 2023-05-04\n", "test_data最大日期: 2023-05-05\n", "划分后的训练集大小: 621, 验证集大小: 115\n", "train_data最大日期: 2023-05-05\n", "test_data最大日期: 2023-05-08\n", "划分后的训练集大小: 618, 验证集大小: 140\n", "train_data最大日期: 2023-05-08\n", "test_data最大日期: 2023-05-09\n", "划分后的训练集大小: 666, 验证集大小: 145\n", "train_data最大日期: 2023-05-09\n", "test_data最大日期: 2023-05-10\n", "划分后的训练集大小: 638, 验证集大小: 94\n", "train_data最大日期: 2023-05-10\n", "test_data最大日期: 2023-05-11\n", "划分后的训练集大小: 618, 验证集大小: 124\n", "train_data最大日期: 2023-05-11\n", "test_data最大日期: 2023-05-12\n", "划分后的训练集大小: 647, 验证集大小: 144\n", "train_data最大日期: 2023-05-12\n", "test_data最大日期: 2023-05-15\n", "划分后的训练集大小: 650, 验证集大小: 143\n", "train_data最大日期: 2023-05-15\n", "test_data最大日期: 2023-05-16\n", "划分后的训练集大小: 608, 验证集大小: 103\n", "train_data最大日期: 2023-05-16\n", "test_data最大日期: 2023-05-17\n", "划分后的训练集大小: 611, 验证集大小: 97\n", "train_data最大日期: 2023-05-17\n", "test_data最大日期: 2023-05-18\n", "划分后的训练集大小: 594, 验证集大小: 107\n", "train_data最大日期: 2023-05-18\n", "test_data最大日期: 2023-05-19\n", "划分后的训练集大小: 565, 验证集大小: 115\n", "train_data最大日期: 2023-05-19\n", "test_data最大日期: 2023-05-22\n", "划分后的训练集大小: 527, 验证集大小: 105\n", "train_data最大日期: 2023-05-22\n", "test_data最大日期: 2023-05-23\n", "划分后的训练集大小: 518, 验证集大小: 94\n", "train_data最大日期: 2023-05-23\n", "test_data最大日期: 2023-05-24\n", "划分后的训练集大小: 542, 验证集大小: 121\n", "train_data最大日期: 2023-05-24\n", "test_data最大日期: 2023-05-25\n", "划分后的训练集大小: 530, 验证集大小: 95\n", "train_data最大日期: 2023-05-25\n", "test_data最大日期: 2023-05-26\n", "划分后的训练集大小: 506, 验证集大小: 91\n", "train_data最大日期: 2023-05-26\n", "test_data最大日期: 2023-05-29\n", "划分后的训练集大小: 497, 验证集大小: 96\n", "train_data最大日期: 2023-05-29\n", "test_data最大日期: 2023-05-30\n", "划分后的训练集大小: 512, 验证集大小: 109\n", "train_data最大日期: 2023-05-30\n", "test_data最大日期: 2023-05-31\n", "划分后的训练集大小: 530, 验证集大小: 139\n", "train_data最大日期: 2023-05-31\n", "test_data最大日期: 2023-06-01\n", "划分后的训练集大小: 574, 验证集大小: 139\n", "train_data最大日期: 2023-06-01\n", "test_data最大日期: 2023-06-02\n", "划分后的训练集大小: 595, 验证集大小: 112\n", "train_data最大日期: 2023-06-02\n", "test_data最大日期: 2023-06-05\n", "划分后的训练集大小: 643, 验证集大小: 144\n", "train_data最大日期: 2023-06-05\n", "test_data最大日期: 2023-06-06\n", "划分后的训练集大小: 645, 验证集大小: 111\n", "train_data最大日期: 2023-06-06\n", "test_data最大日期: 2023-06-07\n", "划分后的训练集大小: 596, 验证集大小: 90\n", "train_data最大日期: 2023-06-07\n", "test_data最大日期: 2023-06-08\n", "划分后的训练集大小: 600, 验证集大小: 143\n", "train_data最大日期: 2023-06-08\n", "test_data最大日期: 2023-06-09\n", "划分后的训练集大小: 582, 验证集大小: 94\n", "train_data最大日期: 2023-06-09\n", "test_data最大日期: 2023-06-12\n", "划分后的训练集大小: 536, 验证集大小: 98\n", "train_data最大日期: 2023-06-12\n", "test_data最大日期: 2023-06-13\n", "划分后的训练集大小: 561, 验证集大小: 136\n", "train_data最大日期: 2023-06-13\n", "test_data最大日期: 2023-06-14\n", "划分后的训练集大小: 611, 验证集大小: 140\n", "train_data最大日期: 2023-06-14\n", "test_data最大日期: 2023-06-15\n", "划分后的训练集大小: 613, 验证集大小: 145\n", "train_data最大日期: 2023-06-15\n", "test_data最大日期: 2023-06-16\n", "划分后的训练集大小: 665, 验证集大小: 146\n", "train_data最大日期: 2023-06-16\n", "test_data最大日期: 2023-06-19\n", "划分后的训练集大小: 705, 验证集大小: 138\n", "train_data最大日期: 2023-06-19\n", "test_data最大日期: 2023-06-20\n", "划分后的训练集大小: 712, 验证集大小: 143\n", "train_data最大日期: 2023-06-20\n", "test_data最大日期: 2023-06-21\n", "划分后的训练集大小: 716, 验证集大小: 144\n", "train_data最大日期: 2023-06-21\n", "test_data最大日期: 2023-06-26\n", "划分后的训练集大小: 718, 验证集大小: 147\n", "train_data最大日期: 2023-06-26\n", "test_data最大日期: 2023-06-27\n", "划分后的训练集大小: 719, 验证集大小: 147\n", "train_data最大日期: 2023-06-27\n", "test_data最大日期: 2023-06-28\n", "划分后的训练集大小: 686, 验证集大小: 105\n", "train_data最大日期: 2023-06-28\n", "test_data最大日期: 2023-06-29\n", "划分后的训练集大小: 639, 验证集大小: 96\n", "train_data最大日期: 2023-06-29\n", "test_data最大日期: 2023-06-30\n", "划分后的训练集大小: 638, 验证集大小: 143\n", "train_data最大日期: 2023-06-30\n", "test_data最大日期: 2023-07-03\n", "划分后的训练集大小: 617, 验证集大小: 126\n", "train_data最大日期: 2023-07-03\n", "test_data最大日期: 2023-07-04\n", "划分后的训练集大小: 615, 验证集大小: 145\n", "train_data最大日期: 2023-07-04\n", "test_data最大日期: 2023-07-05\n", "划分后的训练集大小: 648, 验证集大小: 138\n", "train_data最大日期: 2023-07-05\n", "test_data最大日期: 2023-07-06\n", "划分后的训练集大小: 695, 验证集大小: 143\n", "train_data最大日期: 2023-07-06\n", "test_data最大日期: 2023-07-07\n", "划分后的训练集大小: 699, 验证集大小: 147\n", "train_data最大日期: 2023-07-07\n", "test_data最大日期: 2023-07-10\n", "划分后的训练集大小: 723, 验证集大小: 150\n", "train_data最大日期: 2023-07-10\n", "test_data最大日期: 2023-07-11\n", "划分后的训练集大小: 685, 验证集大小: 107\n", "train_data最大日期: 2023-07-11\n", "test_data最大日期: 2023-07-12\n", "划分后的训练集大小: 689, 验证集大小: 142\n", "train_data最大日期: 2023-07-12\n", "test_data最大日期: 2023-07-13\n", "划分后的训练集大小: 641, 验证集大小: 95\n", "train_data最大日期: 2023-07-13\n", "test_data最大日期: 2023-07-14\n", "划分后的训练集大小: 620, 验证集大小: 126\n", "train_data最大日期: 2023-07-14\n", "test_data最大日期: 2023-07-17\n", "划分后的训练集大小: 585, 验证集大小: 115\n", "train_data最大日期: 2023-07-17\n", "test_data最大日期: 2023-07-18\n", "划分后的训练集大小: 624, 验证集大小: 146\n", "train_data最大日期: 2023-07-18\n", "test_data最大日期: 2023-07-19\n", "划分后的训练集大小: 625, 验证集大小: 143\n", "train_data最大日期: 2023-07-19\n", "test_data最大日期: 2023-07-20\n", "划分后的训练集大小: 654, 验证集大小: 124\n", "train_data最大日期: 2023-07-20\n", "test_data最大日期: 2023-07-21\n", "划分后的训练集大小: 632, 验证集大小: 104\n", "train_data最大日期: 2023-07-21\n", "test_data最大日期: 2023-07-24\n", "划分后的训练集大小: 609, 验证集大小: 92\n", "train_data最大日期: 2023-07-24\n", "test_data最大日期: 2023-07-25\n", "划分后的训练集大小: 567, 验证集大小: 104\n", "train_data最大日期: 2023-07-25\n", "test_data最大日期: 2023-07-26\n", "划分后的训练集大小: 556, 验证集大小: 132\n", "train_data最大日期: 2023-07-26\n", "test_data最大日期: 2023-07-27\n", "划分后的训练集大小: 524, 验证集大小: 92\n", "train_data最大日期: 2023-07-27\n", "test_data最大日期: 2023-07-28\n", "划分后的训练集大小: 530, 验证集大小: 110\n", "train_data最大日期: 2023-07-28\n", "test_data最大日期: 2023-07-31\n", "划分后的训练集大小: 536, 验证集大小: 98\n", "train_data最大日期: 2023-07-31\n", "test_data最大日期: 2023-08-01\n", "划分后的训练集大小: 568, 验证集大小: 136\n", "train_data最大日期: 2023-08-01\n", "test_data最大日期: 2023-08-02\n", "划分后的训练集大小: 580, 验证集大小: 144\n", "train_data最大日期: 2023-08-02\n", "test_data最大日期: 2023-08-03\n", "划分后的训练集大小: 583, 验证集大小: 95\n", "train_data最大日期: 2023-08-03\n", "test_data最大日期: 2023-08-04\n", "划分后的训练集大小: 596, 验证集大小: 123\n", "train_data最大日期: 2023-08-04\n", "test_data最大日期: 2023-08-07\n", "划分后的训练集大小: 615, 验证集大小: 117\n", "train_data最大日期: 2023-08-07\n", "test_data最大日期: 2023-08-08\n", "划分后的训练集大小: 584, 验证集大小: 105\n", "train_data最大日期: 2023-08-08\n", "test_data最大日期: 2023-08-09\n", "划分后的训练集大小: 564, 验证集大小: 124\n", "train_data最大日期: 2023-08-09\n", "test_data最大日期: 2023-08-10\n", "划分后的训练集大小: 586, 验证集大小: 117\n", "train_data最大日期: 2023-08-10\n", "test_data最大日期: 2023-08-11\n", "划分后的训练集大小: 609, 验证集大小: 146\n", "train_data最大日期: 2023-08-11\n", "test_data最大日期: 2023-08-14\n", "划分后的训练集大小: 586, 验证集大小: 94\n", "train_data最大日期: 2023-08-14\n", "test_data最大日期: 2023-08-15\n", "划分后的训练集大小: 584, 验证集大小: 103\n", "train_data最大日期: 2023-08-15\n", "test_data最大日期: 2023-08-16\n", "划分后的训练集大小: 602, 验证集大小: 142\n", "train_data最大日期: 2023-08-16\n", "test_data最大日期: 2023-08-17\n", "划分后的训练集大小: 627, 验证集大小: 142\n", "train_data最大日期: 2023-08-17\n", "test_data最大日期: 2023-08-18\n", "划分后的训练集大小: 604, 验证集大小: 123\n", "train_data最大日期: 2023-08-18\n", "test_data最大日期: 2023-08-21\n", "划分后的训练集大小: 603, 验证集大小: 93\n", "train_data最大日期: 2023-08-21\n", "test_data最大日期: 2023-08-22\n", "划分后的训练集大小: 630, 验证集大小: 130\n", "train_data最大日期: 2023-08-22\n", "test_data最大日期: 2023-08-23\n", "划分后的训练集大小: 579, 验证集大小: 91\n", "train_data最大日期: 2023-08-23\n", "test_data最大日期: 2023-08-24\n", "划分后的训练集大小: 573, 验证集大小: 136\n", "train_data最大日期: 2023-08-24\n", "test_data最大日期: 2023-08-25\n", "划分后的训练集大小: 552, 验证集大小: 102\n", "train_data最大日期: 2023-08-25\n", "test_data最大日期: 2023-08-28\n", "划分后的训练集大小: 566, 验证集大小: 107\n", "train_data最大日期: 2023-08-28\n", "test_data最大日期: 2023-08-29\n", "划分后的训练集大小: 581, 验证集大小: 145\n", "train_data最大日期: 2023-08-29\n", "test_data最大日期: 2023-08-30\n", "划分后的训练集大小: 583, 验证集大小: 93\n", "train_data最大日期: 2023-08-30\n", "test_data最大日期: 2023-08-31\n", "划分后的训练集大小: 560, 验证集大小: 113\n", "train_data最大日期: 2023-08-31\n", "test_data最大日期: 2023-09-01\n", "划分后的训练集大小: 552, 验证集大小: 94\n", "train_data最大日期: 2023-09-01\n", "test_data最大日期: 2023-09-04\n", "划分后的训练集大小: 584, 验证集大小: 139\n", "train_data最大日期: 2023-09-04\n", "test_data最大日期: 2023-09-05\n", "划分后的训练集大小: 544, 验证集大小: 105\n", "train_data最大日期: 2023-09-05\n", "test_data最大日期: 2023-09-06\n", "划分后的训练集大小: 577, 验证集大小: 126\n", "train_data最大日期: 2023-09-06\n", "test_data最大日期: 2023-09-07\n", "划分后的训练集大小: 608, 验证集大小: 144\n", "train_data最大日期: 2023-09-07\n", "test_data最大日期: 2023-09-08\n", "划分后的训练集大小: 618, 验证集大小: 104\n", "train_data最大日期: 2023-09-08\n", "test_data最大日期: 2023-09-11\n", "划分后的训练集大小: 603, 验证集大小: 124\n", "train_data最大日期: 2023-09-11\n", "test_data最大日期: 2023-09-12\n", "划分后的训练集大小: 612, 验证集大小: 114\n", "train_data最大日期: 2023-09-12\n", "test_data最大日期: 2023-09-13\n", "划分后的训练集大小: 582, 验证集大小: 96\n", "train_data最大日期: 2023-09-13\n", "test_data最大日期: 2023-09-14\n", "划分后的训练集大小: 561, 验证集大小: 123\n", "train_data最大日期: 2023-09-14\n", "test_data最大日期: 2023-09-15\n", "划分后的训练集大小: 555, 验证集大小: 98\n", "train_data最大日期: 2023-09-15\n", "test_data最大日期: 2023-09-18\n", "划分后的训练集大小: 567, 验证集大小: 136\n", "train_data最大日期: 2023-09-18\n", "test_data最大日期: 2023-09-19\n", "划分后的训练集大小: 570, 验证集大小: 117\n", "train_data最大日期: 2023-09-19\n", "test_data最大日期: 2023-09-20\n", "划分后的训练集大小: 572, 验证集大小: 98\n", "train_data最大日期: 2023-09-20\n", "test_data最大日期: 2023-09-21\n", "划分后的训练集大小: 578, 验证集大小: 129\n", "train_data最大日期: 2023-09-21\n", "test_data最大日期: 2023-09-22\n", "划分后的训练集大小: 604, 验证集大小: 124\n", "train_data最大日期: 2023-09-22\n", "test_data最大日期: 2023-09-25\n", "划分后的训练集大小: 575, 验证集大小: 107\n", "train_data最大日期: 2023-09-25\n", "test_data最大日期: 2023-09-26\n", "划分后的训练集大小: 601, 验证集大小: 143\n", "train_data最大日期: 2023-09-26\n", "test_data最大日期: 2023-09-27\n", "划分后的训练集大小: 646, 验证集大小: 143\n", "train_data最大日期: 2023-09-27\n", "test_data最大日期: 2023-09-28\n", "划分后的训练集大小: 639, 验证集大小: 122\n", "train_data最大日期: 2023-09-28\n", "test_data最大日期: 2023-10-09\n", "划分后的训练集大小: 612, 验证集大小: 97\n", "train_data最大日期: 2023-10-09\n", "test_data最大日期: 2023-10-10\n", "划分后的训练集大小: 604, 验证集大小: 99\n", "train_data最大日期: 2023-10-10\n", "test_data最大日期: 2023-10-11\n", "划分后的训练集大小: 556, 验证集大小: 95\n", "train_data最大日期: 2023-10-11\n", "test_data最大日期: 2023-10-12\n", "划分后的训练集大小: 535, 验证集大小: 122\n", "train_data最大日期: 2023-10-12\n", "test_data最大日期: 2023-10-13\n", "划分后的训练集大小: 509, 验证集大小: 96\n", "train_data最大日期: 2023-10-13\n", "test_data最大日期: 2023-10-16\n", "划分后的训练集大小: 545, 验证集大小: 133\n", "train_data最大日期: 2023-10-16\n", "test_data最大日期: 2023-10-17\n", "划分后的训练集大小: 585, 验证集大小: 139\n", "train_data最大日期: 2023-10-17\n", "test_data最大日期: 2023-10-18\n", "划分后的训练集大小: 631, 验证集大小: 141\n", "train_data最大日期: 2023-10-18\n", "test_data最大日期: 2023-10-19\n", "划分后的训练集大小: 648, 验证集大小: 139\n", "train_data最大日期: 2023-10-19\n", "test_data最大日期: 2023-10-20\n", "划分后的训练集大小: 652, 验证集大小: 100\n", "train_data最大日期: 2023-10-20\n", "test_data最大日期: 2023-10-23\n", "划分后的训练集大小: 623, 验证集大小: 104\n", "train_data最大日期: 2023-10-23\n", "test_data最大日期: 2023-10-24\n", "划分后的训练集大小: 629, 验证集大小: 145\n", "train_data最大日期: 2023-10-24\n", "test_data最大日期: 2023-10-25\n", "划分后的训练集大小: 627, 验证集大小: 139\n", "train_data最大日期: 2023-10-25\n", "test_data最大日期: 2023-10-26\n", "划分后的训练集大小: 597, 验证集大小: 109\n", "train_data最大日期: 2023-10-26\n", "test_data最大日期: 2023-10-27\n", "划分后的训练集大小: 595, 验证集大小: 98\n", "train_data最大日期: 2023-10-27\n", "test_data最大日期: 2023-10-30\n", "划分后的训练集大小: 632, 验证集大小: 141\n", "train_data最大日期: 2023-10-30\n", "test_data最大日期: 2023-10-31\n", "划分后的训练集大小: 634, 验证集大小: 147\n", "train_data最大日期: 2023-10-31\n", "test_data最大日期: 2023-11-01\n", "划分后的训练集大小: 640, 验证集大小: 145\n", "train_data最大日期: 2023-11-01\n", "test_data最大日期: 2023-11-02\n", "划分后的训练集大小: 657, 验证集大小: 126\n", "train_data最大日期: 2023-11-02\n", "test_data最大日期: 2023-11-03\n", "划分后的训练集大小: 673, 验证集大小: 114\n", "train_data最大日期: 2023-11-03\n", "test_data最大日期: 2023-11-06\n", "划分后的训练集大小: 677, 验证集大小: 145\n", "train_data最大日期: 2023-11-06\n", "test_data最大日期: 2023-11-07\n", "划分后的训练集大小: 654, 验证集大小: 124\n", "train_data最大日期: 2023-11-07\n", "test_data最大日期: 2023-11-08\n", "划分后的训练集大小: 649, 验证集大小: 140\n", "train_data最大日期: 2023-11-08\n", "test_data最大日期: 2023-11-09\n", "划分后的训练集大小: 661, 验证集大小: 138\n", "train_data最大日期: 2023-11-09\n", "test_data最大日期: 2023-11-10\n", "划分后的训练集大小: 669, 验证集大小: 122\n", "train_data最大日期: 2023-11-10\n", "test_data最大日期: 2023-11-13\n", "划分后的训练集大小: 651, 验证集大小: 127\n", "train_data最大日期: 2023-11-13\n", "test_data最大日期: 2023-11-14\n", "划分后的训练集大小: 672, 验证集大小: 145\n", "train_data最大日期: 2023-11-14\n", "test_data最大日期: 2023-11-15\n", "划分后的训练集大小: 627, 验证集大小: 95\n", "train_data最大日期: 2023-11-15\n", "test_data最大日期: 2023-11-16\n", "划分后的训练集大小: 608, 验证集大小: 119\n", "train_data最大日期: 2023-11-16\n", "test_data最大日期: 2023-11-17\n", "划分后的训练集大小: 588, 验证集大小: 102\n", "train_data最大日期: 2023-11-17\n", "test_data最大日期: 2023-11-20\n", "划分后的训练集大小: 606, 验证集大小: 145\n", "train_data最大日期: 2023-11-20\n", "test_data最大日期: 2023-11-21\n", "划分后的训练集大小: 606, 验证集大小: 145\n", "train_data最大日期: 2023-11-21\n", "test_data最大日期: 2023-11-22\n", "划分后的训练集大小: 653, 验证集大小: 142\n", "train_data最大日期: 2023-11-22\n", "test_data最大日期: 2023-11-23\n", "划分后的训练集大小: 643, 验证集大小: 109\n", "train_data最大日期: 2023-11-23\n", "test_data最大日期: 2023-11-24\n", "划分后的训练集大小: 687, 验证集大小: 146\n", "train_data最大日期: 2023-11-24\n", "test_data最大日期: 2023-11-27\n", "划分后的训练集大小: 690, 验证集大小: 148\n", "train_data最大日期: 2023-11-27\n", "test_data最大日期: 2023-11-28\n", "划分后的训练集大小: 689, 验证集大小: 144\n", "train_data最大日期: 2023-11-28\n", "test_data最大日期: 2023-11-29\n", "划分后的训练集大小: 674, 验证集大小: 127\n", "train_data最大日期: 2023-11-29\n", "test_data最大日期: 2023-11-30\n", "划分后的训练集大小: 710, 验证集大小: 145\n", "train_data最大日期: 2023-11-30\n", "test_data最大日期: 2023-12-01\n", "划分后的训练集大小: 709, 验证集大小: 145\n", "train_data最大日期: 2023-12-01\n", "test_data最大日期: 2023-12-04\n", "划分后的训练集大小: 703, 验证集大小: 142\n", "train_data最大日期: 2023-12-04\n", "test_data最大日期: 2023-12-05\n", "划分后的训练集大小: 686, 验证集大小: 127\n", "train_data最大日期: 2023-12-05\n", "test_data最大日期: 2023-12-06\n", "划分后的训练集大小: 701, 验证集大小: 142\n", "train_data最大日期: 2023-12-06\n", "test_data最大日期: 2023-12-07\n", "划分后的训练集大小: 694, 验证集大小: 138\n", "train_data最大日期: 2023-12-07\n", "test_data最大日期: 2023-12-08\n", "划分后的训练集大小: 643, 验证集大小: 94\n", "train_data最大日期: 2023-12-08\n", "test_data最大日期: 2023-12-11\n", "划分后的训练集大小: 647, 验证集大小: 146\n", "train_data最大日期: 2023-12-11\n", "test_data最大日期: 2023-12-12\n", "划分后的训练集大小: 635, 验证集大小: 115\n", "train_data最大日期: 2023-12-12\n", "test_data最大日期: 2023-12-13\n", "划分后的训练集大小: 598, 验证集大小: 105\n", "train_data最大日期: 2023-12-13\n", "test_data最大日期: 2023-12-14\n", "划分后的训练集大小: 604, 验证集大小: 144\n", "train_data最大日期: 2023-12-14\n", "test_data最大日期: 2023-12-15\n", "划分后的训练集大小: 639, 验证集大小: 129\n", "train_data最大日期: 2023-12-15\n", "test_data最大日期: 2023-12-18\n", "划分后的训练集大小: 634, 验证集大小: 141\n", "train_data最大日期: 2023-12-18\n", "test_data最大日期: 2023-12-19\n", "划分后的训练集大小: 624, 验证集大小: 105\n", "train_data最大日期: 2023-12-19\n", "test_data最大日期: 2023-12-20\n", "划分后的训练集大小: 666, 验证集大小: 147\n", "train_data最大日期: 2023-12-20\n", "test_data最大日期: 2023-12-21\n", "划分后的训练集大小: 668, 验证集大小: 146\n", "train_data最大日期: 2023-12-21\n", "test_data最大日期: 2023-12-22\n", "划分后的训练集大小: 680, 验证集大小: 141\n", "train_data最大日期: 2023-12-22\n", "test_data最大日期: 2023-12-25\n", "划分后的训练集大小: 674, 验证集大小: 135\n", "train_data最大日期: 2023-12-25\n", "test_data最大日期: 2023-12-26\n", "划分后的训练集大小: 714, 验证集大小: 145\n", "train_data最大日期: 2023-12-26\n", "test_data最大日期: 2023-12-27\n", "划分后的训练集大小: 711, 验证集大小: 144\n", "train_data最大日期: 2023-12-27\n", "test_data最大日期: 2023-12-28\n", "划分后的训练集大小: 660, 验证集大小: 95\n", "train_data最大日期: 2023-12-28\n", "test_data最大日期: 2023-12-29\n", "划分后的训练集大小: 662, 验证集大小: 143\n", "train_data最大日期: 2023-12-29\n", "test_data最大日期: 2024-01-02\n", "划分后的训练集大小: 669, 验证集大小: 142\n", "train_data最大日期: 2024-01-02\n", "test_data最大日期: 2024-01-03\n", "划分后的训练集大小: 666, 验证集大小: 142\n", "train_data最大日期: 2024-01-03\n", "test_data最大日期: 2024-01-04\n", "划分后的训练集大小: 662, 验证集大小: 140\n", "train_data最大日期: 2024-01-04\n", "test_data最大日期: 2024-01-05\n", "划分后的训练集大小: 712, 验证集大小: 145\n", "train_data最大日期: 2024-01-05\n", "test_data最大日期: 2024-01-08\n", "划分后的训练集大小: 711, 验证集大小: 142\n", "train_data最大日期: 2024-01-08\n", "test_data最大日期: 2024-01-09\n", "划分后的训练集大小: 671, 验证集大小: 102\n", "train_data最大日期: 2024-01-09\n", "test_data最大日期: 2024-01-10\n", "划分后的训练集大小: 671, 验证集大小: 142\n", "train_data最大日期: 2024-01-10\n", "test_data最大日期: 2024-01-11\n", "划分后的训练集大小: 673, 验证集大小: 142\n", "train_data最大日期: 2024-01-11\n", "test_data最大日期: 2024-01-12\n", "划分后的训练集大小: 622, 验证集大小: 94\n", "train_data最大日期: 2024-01-12\n", "test_data最大日期: 2024-01-15\n", "划分后的训练集大小: 624, 验证集大小: 144\n", "train_data最大日期: 2024-01-15\n", "test_data最大日期: 2024-01-16\n", "划分后的训练集大小: 625, 验证集大小: 103\n", "train_data最大日期: 2024-01-16\n", "test_data最大日期: 2024-01-17\n", "划分后的训练集大小: 606, 验证集大小: 123\n", "train_data最大日期: 2024-01-17\n", "test_data最大日期: 2024-01-18\n", "划分后的训练集大小: 608, 验证集大小: 144\n", "train_data最大日期: 2024-01-18\n", "test_data最大日期: 2024-01-19\n", "划分后的训练集大小: 654, 验证集大小: 140\n", "train_data最大日期: 2024-01-19\n", "test_data最大日期: 2024-01-22\n", "划分后的训练集大小: 606, 验证集大小: 96\n", "train_data最大日期: 2024-01-22\n", "test_data最大日期: 2024-01-23\n", "划分后的训练集大小: 629, 验证集大小: 126\n", "train_data最大日期: 2024-01-23\n", "test_data最大日期: 2024-01-24\n", "划分后的训练集大小: 648, 验证集大小: 142\n", "train_data最大日期: 2024-01-24\n", "test_data最大日期: 2024-01-25\n", "划分后的训练集大小: 645, 验证集大小: 141\n", "train_data最大日期: 2024-01-25\n", "test_data最大日期: 2024-01-26\n", "划分后的训练集大小: 645, 验证集大小: 140\n", "train_data最大日期: 2024-01-26\n", "test_data最大日期: 2024-01-29\n", "划分后的训练集大小: 677, 验证集大小: 128\n", "train_data最大日期: 2024-01-29\n", "test_data最大日期: 2024-01-30\n", "划分后的训练集大小: 685, 验证集大小: 134\n", "train_data最大日期: 2024-01-30\n", "test_data最大日期: 2024-01-31\n", "划分后的训练集大小: 637, 验证集大小: 94\n", "train_data最大日期: 2024-01-31\n", "test_data最大日期: 2024-02-01\n", "划分后的训练集大小: 640, 验证集大小: 144\n", "train_data最大日期: 2024-02-01\n", "test_data最大日期: 2024-02-02\n", "划分后的训练集大小: 643, 验证集大小: 143\n", "train_data最大日期: 2024-02-02\n", "test_data最大日期: 2024-02-05\n", "划分后的训练集大小: 660, 验证集大小: 145\n", "train_data最大日期: 2024-02-05\n", "test_data最大日期: 2024-02-06\n", "划分后的训练集大小: 671, 验证集大小: 145\n", "train_data最大日期: 2024-02-06\n", "test_data最大日期: 2024-02-07\n", "划分后的训练集大小: 723, 验证集大小: 146\n", "train_data最大日期: 2024-02-07\n", "test_data最大日期: 2024-02-08\n", "划分后的训练集大小: 726, 验证集大小: 147\n", "train_data最大日期: 2024-02-08\n", "test_data最大日期: 2024-02-19\n", "划分后的训练集大小: 730, 验证集大小: 147\n", "train_data最大日期: 2024-02-19\n", "test_data最大日期: 2024-02-20\n", "划分后的训练集大小: 682, 验证集大小: 97\n", "train_data最大日期: 2024-02-20\n", "test_data最大日期: 2024-02-21\n", "划分后的训练集大小: 635, 验证集大小: 98\n", "train_data最大日期: 2024-02-21\n", "test_data最大日期: 2024-02-22\n", "划分后的训练集大小: 635, 验证集大小: 146\n", "train_data最大日期: 2024-02-22\n", "test_data最大日期: 2024-02-23\n", "划分后的训练集大小: 633, 验证集大小: 145\n", "train_data最大日期: 2024-02-23\n", "test_data最大日期: 2024-02-26\n", "划分后的训练集大小: 630, 验证集大小: 144\n", "train_data最大日期: 2024-02-26\n", "test_data最大日期: 2024-02-27\n", "划分后的训练集大小: 626, 验证集大小: 93\n", "train_data最大日期: 2024-02-27\n", "test_data最大日期: 2024-02-28\n", "划分后的训练集大小: 671, 验证集大小: 143\n", "train_data最大日期: 2024-02-28\n", "test_data最大日期: 2024-02-29\n", "划分后的训练集大小: 619, 验证集大小: 94\n", "train_data最大日期: 2024-02-29\n", "test_data最大日期: 2024-03-01\n", "划分后的训练集大小: 582, 验证集大小: 108\n", "train_data最大日期: 2024-03-01\n", "test_data最大日期: 2024-03-04\n", "划分后的训练集大小: 533, 验证集大小: 95\n", "train_data最大日期: 2024-03-04\n", "test_data最大日期: 2024-03-05\n", "划分后的训练集大小: 533, 验证集大小: 93\n", "train_data最大日期: 2024-03-05\n", "test_data最大日期: 2024-03-06\n", "划分后的训练集大小: 484, 验证集大小: 94\n", "train_data最大日期: 2024-03-06\n", "test_data最大日期: 2024-03-07\n", "划分后的训练集大小: 485, 验证集大小: 95\n", "train_data最大日期: 2024-03-07\n", "test_data最大日期: 2024-03-08\n", "划分后的训练集大小: 521, 验证集大小: 144\n", "train_data最大日期: 2024-03-08\n", "test_data最大日期: 2024-03-11\n", "划分后的训练集大小: 521, 验证集大小: 95\n", "train_data最大日期: 2024-03-11\n", "test_data最大日期: 2024-03-12\n", "划分后的训练集大小: 525, 验证集大小: 97\n", "train_data最大日期: 2024-03-12\n", "test_data最大日期: 2024-03-13\n", "划分后的训练集大小: 575, 验证集大小: 144\n", "train_data最大日期: 2024-03-13\n", "test_data最大日期: 2024-03-14\n", "划分后的训练集大小: 603, 验证集大小: 123\n", "train_data最大日期: 2024-03-14\n", "test_data最大日期: 2024-03-15\n", "划分后的训练集大小: 553, 验证集大小: 94\n", "train_data最大日期: 2024-03-15\n", "test_data最大日期: 2024-03-18\n", "划分后的训练集大小: 552, 验证集大小: 94\n", "train_data最大日期: 2024-03-18\n", "test_data最大日期: 2024-03-19\n", "划分后的训练集大小: 593, 验证集大小: 138\n", "train_data最大日期: 2024-03-19\n", "test_data最大日期: 2024-03-20\n", "划分后的训练集大小: 537, 验证集大小: 88\n", "train_data最大日期: 2024-03-20\n", "test_data最大日期: 2024-03-21\n", "划分后的训练集大小: 554, 验证集大小: 140\n", "train_data最大日期: 2024-03-21\n", "test_data最大日期: 2024-03-22\n", "划分后的训练集大小: 592, 验证集大小: 132\n", "train_data最大日期: 2024-03-22\n", "test_data最大日期: 2024-03-25\n", "划分后的训练集大小: 640, 验证集大小: 142\n", "train_data最大日期: 2024-03-25\n", "test_data最大日期: 2024-03-26\n", "划分后的训练集大小: 637, 验证集大小: 135\n", "train_data最大日期: 2024-03-26\n", "test_data最大日期: 2024-03-27\n", "划分后的训练集大小: 671, 验证集大小: 122\n", "train_data最大日期: 2024-03-27\n", "test_data最大日期: 2024-03-28\n", "划分后的训练集大小: 645, 验证集大小: 114\n", "train_data最大日期: 2024-03-28\n", "test_data最大日期: 2024-03-29\n", "划分后的训练集大小: 657, 验证集大小: 144\n", "train_data最大日期: 2024-03-29\n", "test_data最大日期: 2024-04-01\n", "划分后的训练集大小: 621, 验证集大小: 106\n", "train_data最大日期: 2024-04-01\n", "test_data最大日期: 2024-04-02\n", "划分后的训练集大小: 626, 验证集大小: 140\n", "train_data最大日期: 2024-04-02\n", "test_data最大日期: 2024-04-03\n", "划分后的训练集大小: 619, 验证集大小: 115\n", "train_data最大日期: 2024-04-03\n", "test_data最大日期: 2024-04-08\n", "划分后的训练集大小: 649, 验证集大小: 144\n", "train_data最大日期: 2024-04-08\n", "test_data最大日期: 2024-04-09\n", "划分后的训练集大小: 609, 验证集大小: 104\n", "train_data最大日期: 2024-04-09\n", "test_data最大日期: 2024-04-10\n", "划分后的训练集大小: 598, 验证集大小: 95\n", "train_data最大日期: 2024-04-10\n", "test_data最大日期: 2024-04-11\n", "划分后的训练集大小: 601, 验证集大小: 143\n", "train_data最大日期: 2024-04-11\n", "test_data最大日期: 2024-04-12\n", "划分后的训练集大小: 582, 验证集大小: 96\n", "train_data最大日期: 2024-04-12\n", "test_data最大日期: 2024-04-15\n", "划分后的训练集大小: 528, 验证集大小: 90\n", "train_data最大日期: 2024-04-15\n", "test_data最大日期: 2024-04-16\n", "划分后的训练集大小: 552, 验证集大小: 128\n", "train_data最大日期: 2024-04-16\n", "test_data最大日期: 2024-04-17\n", "划分后的训练集大小: 552, 验证集大小: 95\n", "train_data最大日期: 2024-04-17\n", "test_data最大日期: 2024-04-18\n", "划分后的训练集大小: 554, 验证集大小: 145\n", "train_data最大日期: 2024-04-18\n", "test_data最大日期: 2024-04-19\n", "划分后的训练集大小: 553, 验证集大小: 95\n", "train_data最大日期: 2024-04-19\n", "test_data最大日期: 2024-04-22\n", "划分后的训练集大小: 609, 验证集大小: 146\n", "train_data最大日期: 2024-04-22\n", "test_data最大日期: 2024-04-23\n", "划分后的训练集大小: 629, 验证集大小: 148\n", "train_data最大日期: 2024-04-23\n", "test_data最大日期: 2024-04-24\n", "划分后的训练集大小: 631, 验证集大小: 97\n", "train_data最大日期: 2024-04-24\n", "test_data最大日期: 2024-04-25\n", "划分后的训练集大小: 582, 验证集大小: 96\n", "train_data最大日期: 2024-04-25\n", "test_data最大日期: 2024-04-26\n", "划分后的训练集大小: 593, 验证集大小: 106\n", "train_data最大日期: 2024-04-26\n", "test_data最大日期: 2024-04-29\n", "划分后的训练集大小: 593, 验证集大小: 146\n", "train_data最大日期: 2024-04-29\n", "test_data最大日期: 2024-04-30\n", "划分后的训练集大小: 580, 验证集大小: 135\n", "train_data最大日期: 2024-04-30\n", "test_data最大日期: 2024-05-06\n", "划分后的训练集大小: 581, 验证集大小: 98\n", "train_data最大日期: 2024-05-06\n", "test_data最大日期: 2024-05-07\n", "划分后的训练集大小: 598, 验证集大小: 113\n", "train_data最大日期: 2024-05-07\n", "test_data最大日期: 2024-05-08\n", "划分后的训练集大小: 587, 验证集大小: 95\n", "train_data最大日期: 2024-05-08\n", "test_data最大日期: 2024-05-09\n", "划分后的训练集大小: 536, 验证集大小: 95\n", "train_data最大日期: 2024-05-09\n", "test_data最大日期: 2024-05-10\n", "划分后的训练集大小: 494, 验证集大小: 93\n", "train_data最大日期: 2024-05-10\n", "test_data最大日期: 2024-05-13\n", "划分后的训练集大小: 540, 验证集大小: 144\n", "train_data最大日期: 2024-05-13\n", "test_data最大日期: 2024-05-14\n", "划分后的训练集大小: 573, 验证集大小: 146\n", "train_data最大日期: 2024-05-14\n", "test_data最大日期: 2024-05-15\n", "划分后的训练集大小: 623, 验证集大小: 145\n", "train_data最大日期: 2024-05-15\n", "test_data最大日期: 2024-05-16\n", "划分后的训练集大小: 673, 验证集大小: 145\n", "train_data最大日期: 2024-05-16\n", "test_data最大日期: 2024-05-17\n", "划分后的训练集大小: 679, 验证集大小: 99\n", "train_data最大日期: 2024-05-17\n", "test_data最大日期: 2024-05-20\n", "划分后的训练集大小: 634, 验证集大小: 99\n", "train_data最大日期: 2024-05-20\n", "test_data最大日期: 2024-05-21\n", "划分后的训练集大小: 633, 验证集大小: 145\n", "train_data最大日期: 2024-05-21\n", "test_data最大日期: 2024-05-22\n", "划分后的训练集大小: 603, 验证集大小: 115\n", "train_data最大日期: 2024-05-22\n", "test_data最大日期: 2024-05-23\n", "划分后的训练集大小: 555, 验证集大小: 97\n", "train_data最大日期: 2024-05-23\n", "test_data最大日期: 2024-05-24\n", "划分后的训练集大小: 555, 验证集大小: 99\n", "train_data最大日期: 2024-05-24\n", "test_data最大日期: 2024-05-27\n", "划分后的训练集大小: 595, 验证集大小: 139\n", "train_data最大日期: 2024-05-27\n", "test_data最大日期: 2024-05-28\n", "划分后的训练集大小: 550, 验证集大小: 100\n", "train_data最大日期: 2024-05-28\n", "test_data最大日期: 2024-05-29\n", "划分后的训练集大小: 581, 验证集大小: 146\n", "train_data最大日期: 2024-05-29\n", "test_data最大日期: 2024-05-30\n", "划分后的训练集大小: 609, 验证集大小: 125\n", "train_data最大日期: 2024-05-30\n", "test_data最大日期: 2024-05-31\n", "划分后的训练集大小: 610, 验证集大小: 100\n", "train_data最大日期: 2024-05-31\n", "test_data最大日期: 2024-06-03\n", "划分后的训练集大小: 571, 验证集大小: 100\n", "train_data最大日期: 2024-06-03\n", "test_data最大日期: 2024-06-04\n", "划分后的训练集大小: 590, 验证集大小: 119\n", "train_data最大日期: 2024-06-04\n", "test_data最大日期: 2024-06-05\n", "划分后的训练集大小: 543, 验证集大小: 99\n", "train_data最大日期: 2024-06-05\n", "test_data最大日期: 2024-06-06\n", "划分后的训练集大小: 536, 验证集大小: 118\n", "train_data最大日期: 2024-06-06\n", "test_data最大日期: 2024-06-07\n", "划分后的训练集大小: 575, 验证集大小: 139\n", "train_data最大日期: 2024-06-07\n", "test_data最大日期: 2024-06-11\n", "划分后的训练集大小: 574, 验证集大小: 99\n", "train_data最大日期: 2024-06-11\n", "test_data最大日期: 2024-06-12\n", "划分后的训练集大小: 554, 验证集大小: 99\n", "train_data最大日期: 2024-06-12\n", "test_data最大日期: 2024-06-13\n", "划分后的训练集大小: 601, 验证集大小: 146\n", "train_data最大日期: 2024-06-13\n", "test_data最大日期: 2024-06-14\n", "划分后的训练集大小: 629, 验证集大小: 146\n", "train_data最大日期: 2024-06-14\n", "test_data最大日期: 2024-06-17\n", "划分后的训练集大小: 617, 验证集大小: 127\n", "train_data最大日期: 2024-06-17\n", "test_data最大日期: 2024-06-18\n", "划分后的训练集大小: 614, 验证集大小: 96\n", "train_data最大日期: 2024-06-18\n", "test_data最大日期: 2024-06-19\n", "划分后的训练集大小: 653, 验证集大小: 138\n", "train_data最大日期: 2024-06-19\n", "test_data最大日期: 2024-06-20\n", "划分后的训练集大小: 604, 验证集大小: 97\n", "train_data最大日期: 2024-06-20\n", "test_data最大日期: 2024-06-21\n", "划分后的训练集大小: 555, 验证集大小: 97\n", "train_data最大日期: 2024-06-21\n", "test_data最大日期: 2024-06-24\n", "划分后的训练集大小: 553, 验证集大小: 125\n", "train_data最大日期: 2024-06-24\n", "test_data最大日期: 2024-06-25\n", "划分后的训练集大小: 604, 验证集大小: 147\n", "train_data最大日期: 2024-06-25\n", "test_data最大日期: 2024-06-26\n", "划分后的训练集大小: 563, 验证集大小: 97\n", "train_data最大日期: 2024-06-26\n", "test_data最大日期: 2024-06-27\n", "划分后的训练集大小: 563, 验证集大小: 97\n", "train_data最大日期: 2024-06-27\n", "test_data最大日期: 2024-06-28\n", "划分后的训练集大小: 583, 验证集大小: 117\n", "train_data最大日期: 2024-06-28\n", "test_data最大日期: 2024-07-01\n", "划分后的训练集大小: 555, 验证集大小: 97\n", "train_data最大日期: 2024-07-01\n", "test_data最大日期: 2024-07-02\n", "划分后的训练集大小: 505, 验证集大小: 97\n", "train_data最大日期: 2024-07-02\n", "test_data最大日期: 2024-07-03\n", "划分后的训练集大小: 553, 验证集大小: 145\n", "train_data最大日期: 2024-07-03\n", "test_data最大日期: 2024-07-04\n", "划分后的训练集大小: 561, 验证集大小: 105\n", "train_data最大日期: 2024-07-04\n", "test_data最大日期: 2024-07-05\n", "划分后的训练集大小: 589, 验证集大小: 145\n", "train_data最大日期: 2024-07-05\n", "test_data最大日期: 2024-07-08\n", "划分后的训练集大小: 639, 验证集大小: 147\n", "train_data最大日期: 2024-07-08\n", "test_data最大日期: 2024-07-09\n", "划分后的训练集大小: 639, 验证集大小: 97\n", "train_data最大日期: 2024-07-09\n", "test_data最大日期: 2024-07-10\n", "划分后的训练集大小: 591, 验证集大小: 97\n", "train_data最大日期: 2024-07-10\n", "test_data最大日期: 2024-07-11\n", "划分后的训练集大小: 632, 验证集大小: 146\n", "train_data最大日期: 2024-07-11\n", "test_data最大日期: 2024-07-12\n", "划分后的训练集大小: 581, 验证集大小: 94\n", "train_data最大日期: 2024-07-12\n", "test_data最大日期: 2024-07-15\n", "划分后的训练集大小: 555, 验证集大小: 121\n", "train_data最大日期: 2024-07-15\n", "test_data最大日期: 2024-07-16\n", "划分后的训练集大小: 552, 验证集大小: 94\n", "train_data最大日期: 2024-07-16\n", "test_data最大日期: 2024-07-17\n", "划分后的训练集大小: 592, 验证集大小: 137\n", "train_data最大日期: 2024-07-17\n", "test_data最大日期: 2024-07-18\n", "划分后的训练集大小: 543, 验证集大小: 97\n", "train_data最大日期: 2024-07-18\n", "test_data最大日期: 2024-07-19\n", "划分后的训练集大小: 546, 验证集大小: 97\n", "train_data最大日期: 2024-07-19\n", "test_data最大日期: 2024-07-22\n", "划分后的训练集大小: 561, 验证集大小: 136\n", "train_data最大日期: 2024-07-22\n", "test_data最大日期: 2024-07-23\n", "划分后的训练集大小: 564, 验证集大小: 97\n", "train_data最大日期: 2024-07-23\n", "test_data最大日期: 2024-07-24\n", "划分后的训练集大小: 543, 验证集大小: 116\n", "train_data最大日期: 2024-07-24\n", "test_data最大日期: 2024-07-25\n", "划分后的训练集大小: 592, 验证集大小: 146\n", "train_data最大日期: 2024-07-25\n", "test_data最大日期: 2024-07-26\n", "划分后的训练集大小: 593, 验证集大小: 98\n", "train_data最大日期: 2024-07-26\n", "test_data最大日期: 2024-07-29\n", "划分后的训练集大小: 583, 验证集大小: 126\n", "train_data最大日期: 2024-07-29\n", "test_data最大日期: 2024-07-30\n", "划分后的训练集大小: 632, 验证集大小: 146\n", "train_data最大日期: 2024-07-30\n", "test_data最大日期: 2024-07-31\n", "划分后的训练集大小: 650, 验证集大小: 134\n", "train_data最大日期: 2024-07-31\n", "test_data最大日期: 2024-08-01\n", "划分后的训练集大小: 650, 验证集大小: 146\n", "train_data最大日期: 2024-08-01\n", "test_data最大日期: 2024-08-02\n", "划分后的训练集大小: 694, 验证集大小: 142\n", "train_data最大日期: 2024-08-02\n", "test_data最大日期: 2024-08-05\n", "划分后的训练集大小: 681, 验证集大小: 113\n", "train_data最大日期: 2024-08-05\n", "test_data最大日期: 2024-08-06\n", "划分后的训练集大小: 658, 验证集大小: 123\n", "train_data最大日期: 2024-08-06\n", "test_data最大日期: 2024-08-07\n", "划分后的训练集大小: 671, 验证集大小: 147\n", "train_data最大日期: 2024-08-07\n", "test_data最大日期: 2024-08-08\n", "划分后的训练集大小: 669, 验证集大小: 144\n", "train_data最大日期: 2024-08-08\n", "test_data最大日期: 2024-08-09\n", "划分后的训练集大小: 671, 验证集大小: 144\n", "train_data最大日期: 2024-08-09\n", "test_data最大日期: 2024-08-12\n", "划分后的训练集大小: 676, 验证集大小: 118\n", "train_data最大日期: 2024-08-12\n", "test_data最大日期: 2024-08-13\n", "划分后的训练集大小: 648, 验证集大小: 95\n", "train_data最大日期: 2024-08-13\n", "test_data最大日期: 2024-08-14\n", "划分后的训练集大小: 618, 验证集大小: 117\n", "train_data最大日期: 2024-08-14\n", "test_data最大日期: 2024-08-15\n", "划分后的训练集大小: 574, 验证集大小: 100\n", "train_data最大日期: 2024-08-15\n", "test_data最大日期: 2024-08-16\n", "划分后的训练集大小: 568, 验证集大小: 138\n", "train_data最大日期: 2024-08-16\n", "test_data最大日期: 2024-08-19\n", "划分后的训练集大小: 596, 验证集大小: 146\n", "train_data最大日期: 2024-08-19\n", "test_data最大日期: 2024-08-20\n", "划分后的训练集大小: 650, 验证集大小: 149\n", "train_data最大日期: 2024-08-20\n", "test_data最大日期: 2024-08-21\n", "划分后的训练集大小: 681, 验证集大小: 148\n", "train_data最大日期: 2024-08-21\n", "test_data最大日期: 2024-08-22\n", "划分后的训练集大小: 698, 验证集大小: 117\n", "train_data最大日期: 2024-08-22\n", "test_data最大日期: 2024-08-23\n", "划分后的训练集大小: 708, 验证集大小: 148\n", "train_data最大日期: 2024-08-23\n", "test_data最大日期: 2024-08-26\n", "划分后的训练集大小: 708, 验证集大小: 146\n", "train_data最大日期: 2024-08-26\n", "test_data最大日期: 2024-08-27\n", "划分后的训练集大小: 704, 验证集大小: 145\n", "train_data最大日期: 2024-08-27\n", "test_data最大日期: 2024-08-28\n", "划分后的训练集大小: 654, 验证集大小: 98\n", "train_data最大日期: 2024-08-28\n", "test_data最大日期: 2024-08-29\n", "划分后的训练集大小: 642, 验证集大小: 105\n", "train_data最大日期: 2024-08-29\n", "test_data最大日期: 2024-08-30\n", "划分后的训练集大小: 591, 验证集大小: 97\n", "train_data最大日期: 2024-08-30\n", "test_data最大日期: 2024-09-02\n", "划分后的训练集大小: 551, 验证集大小: 106\n", "train_data最大日期: 2024-09-02\n", "test_data最大日期: 2024-09-03\n", "划分后的训练集大小: 511, 验证集大小: 105\n", "train_data最大日期: 2024-09-03\n", "test_data最大日期: 2024-09-04\n", "划分后的训练集大小: 511, 验证集大小: 98\n", "train_data最大日期: 2024-09-04\n", "test_data最大日期: 2024-09-05\n", "划分后的训练集大小: 500, 验证集大小: 94\n", "train_data最大日期: 2024-09-05\n", "test_data最大日期: 2024-09-06\n", "划分后的训练集大小: 498, 验证集大小: 95\n", "train_data最大日期: 2024-09-06\n", "test_data最大日期: 2024-09-09\n", "划分后的训练集大小: 534, 验证集大小: 142\n", "train_data最大日期: 2024-09-09\n", "test_data最大日期: 2024-09-10\n", "划分后的训练集大小: 550, 验证集大小: 121\n", "train_data最大日期: 2024-09-10\n", "test_data最大日期: 2024-09-11\n", "划分后的训练集大小: 549, 验证集大小: 97\n", "train_data最大日期: 2024-09-11\n", "test_data最大日期: 2024-09-12\n", "划分后的训练集大小: 579, 验证集大小: 124\n", "train_data最大日期: 2024-09-12\n", "test_data最大日期: 2024-09-13\n", "划分后的训练集大小: 588, 验证集大小: 104\n", "train_data最大日期: 2024-09-13\n", "test_data最大日期: 2024-09-18\n", "划分后的训练集大小: 544, 验证集大小: 98\n", "train_data最大日期: 2024-09-18\n", "test_data最大日期: 2024-09-19\n", "划分后的训练集大小: 526, 验证集大小: 103\n", "train_data最大日期: 2024-09-19\n", "test_data最大日期: 2024-09-20\n", "划分后的训练集大小: 564, 验证集大小: 135\n", "train_data最大日期: 2024-09-20\n", "test_data最大日期: 2024-09-23\n", "划分后的训练集大小: 584, 验证集大小: 144\n", "train_data最大日期: 2024-09-23\n", "test_data最大日期: 2024-09-24\n", "划分后的训练集大小: 569, 验证集大小: 89\n", "train_data最大日期: 2024-09-24\n", "test_data最大日期: 2024-09-25\n", "划分后的训练集大小: 602, 验证集大小: 131\n", "train_data最大日期: 2024-09-25\n", "test_data最大日期: 2024-09-26\n", "划分后的训练集大小: 592, 验证集大小: 93\n", "train_data最大日期: 2024-09-26\n", "test_data最大日期: 2024-09-27\n", "划分后的训练集大小: 566, 验证集大小: 109\n", "train_data最大日期: 2024-09-27\n", "test_data最大日期: 2024-09-30\n", "划分后的训练集大小: 514, 验证集大小: 92\n", "train_data最大日期: 2024-09-30\n", "test_data最大日期: 2024-10-08\n", "划分后的训练集大小: 524, 验证集大小: 99\n", "train_data最大日期: 2024-10-08\n", "test_data最大日期: 2024-10-09\n", "划分后的训练集大小: 473, 验证集大小: 80\n", "train_data最大日期: 2024-10-09\n", "test_data最大日期: 2024-10-10\n", "划分后的训练集大小: 502, 验证集大小: 122\n", "train_data最大日期: 2024-10-10\n", "test_data最大日期: 2024-10-11\n", "划分后的训练集大小: 535, 验证集大小: 142\n", "train_data最大日期: 2024-10-11\n", "test_data最大日期: 2024-10-14\n", "划分后的训练集大小: 588, 验证集大小: 145\n", "train_data最大日期: 2024-10-14\n", "test_data最大日期: 2024-10-15\n", "划分后的训练集大小: 596, 验证集大小: 107\n", "train_data最大日期: 2024-10-15\n", "test_data最大日期: 2024-10-16\n", "划分后的训练集大小: 664, 验证集大小: 148\n", "train_data最大日期: 2024-10-16\n", "test_data最大日期: 2024-10-17\n", "划分后的训练集大小: 683, 验证集大小: 141\n", "train_data最大日期: 2024-10-17\n", "test_data最大日期: 2024-10-18\n", "划分后的训练集大小: 658, 验证集大小: 117\n", "train_data最大日期: 2024-10-18\n", "test_data最大日期: 2024-10-21\n", "划分后的训练集大小: 655, 验证集大小: 142\n", "train_data最大日期: 2024-10-21\n", "test_data最大日期: 2024-10-22\n", "划分后的训练集大小: 686, 验证集大小: 138\n", "train_data最大日期: 2024-10-22\n", "test_data最大日期: 2024-10-23\n", "划分后的训练集大小: 652, 验证集大小: 114\n", "train_data最大日期: 2024-10-23\n", "test_data最大日期: 2024-10-24\n", "划分后的训练集大小: 596, 验证集大小: 85\n", "train_data最大日期: 2024-10-24\n", "test_data最大日期: 2024-10-25\n", "划分后的训练集大小: 603, 验证集大小: 124\n", "train_data最大日期: 2024-10-25\n", "test_data最大日期: 2024-10-28\n", "划分后的训练集大小: 576, 验证集大小: 115\n", "train_data最大日期: 2024-10-28\n", "test_data最大日期: 2024-10-29\n", "划分后的训练集大小: 552, 验证集大小: 114\n", "train_data最大日期: 2024-10-29\n", "test_data最大日期: 2024-10-30\n", "划分后的训练集大小: 541, 验证集大小: 103\n", "train_data最大日期: 2024-10-30\n", "test_data最大日期: 2024-10-31\n", "划分后的训练集大小: 582, 验证集大小: 126\n", "train_data最大日期: 2024-10-31\n", "test_data最大日期: 2024-11-01\n", "划分后的训练集大小: 565, 验证集大小: 107\n", "train_data最大日期: 2024-11-01\n", "test_data最大日期: 2024-11-04\n", "划分后的训练集大小: 583, 验证集大小: 133\n", "train_data最大日期: 2024-11-04\n", "test_data最大日期: 2024-11-05\n", "划分后的训练集大小: 612, 验证集大小: 143\n", "train_data最大日期: 2024-11-05\n", "test_data最大日期: 2024-11-06\n", "划分后的训练集大小: 647, 验证集大小: 138\n", "train_data最大日期: 2024-11-06\n", "test_data最大日期: 2024-11-07\n", "划分后的训练集大小: 621, 验证集大小: 100\n", "train_data最大日期: 2024-11-07\n", "test_data最大日期: 2024-11-08\n", "划分后的训练集大小: 601, 验证集大小: 87\n", "train_data最大日期: 2024-11-08\n", "test_data最大日期: 2024-11-11\n", "划分后的训练集大小: 606, 验证集大小: 138\n", "train_data最大日期: 2024-11-11\n", "test_data最大日期: 2024-11-12\n", "划分后的训练集大小: 598, 验证集大小: 135\n", "train_data最大日期: 2024-11-12\n", "test_data最大日期: 2024-11-13\n", "划分后的训练集大小: 598, 验证集大小: 138\n", "train_data最大日期: 2024-11-13\n", "test_data最大日期: 2024-11-14\n", "划分后的训练集大小: 638, 验证集大小: 140\n", "train_data最大日期: 2024-11-14\n", "test_data最大日期: 2024-11-15\n", "划分后的训练集大小: 691, 验证集大小: 140\n", "train_data最大日期: 2024-11-15\n", "test_data最大日期: 2024-11-18\n", "划分后的训练集大小: 695, 验证集大小: 142\n", "train_data最大日期: 2024-11-18\n", "test_data最大日期: 2024-11-19\n", "划分后的训练集大小: 659, 验证集大小: 99\n", "train_data最大日期: 2024-11-19\n", "test_data最大日期: 2024-11-20\n", "划分后的训练集大小: 642, 验证集大小: 121\n", "train_data最大日期: 2024-11-20\n", "test_data最大日期: 2024-11-21\n", "划分后的训练集大小: 642, 验证集大小: 140\n", "train_data最大日期: 2024-11-21\n", "test_data最大日期: 2024-11-22\n", "划分后的训练集大小: 634, 验证集大小: 132\n", "train_data最大日期: 2024-11-22\n", "test_data最大日期: 2024-11-25\n", "划分后的训练集大小: 610, 验证集大小: 118\n", "train_data最大日期: 2024-11-25\n", "test_data最大日期: 2024-11-26\n", "划分后的训练集大小: 642, 验证集大小: 131\n", "train_data最大日期: 2024-11-26\n", "test_data最大日期: 2024-11-27\n", "划分后的训练集大小: 656, 验证集大小: 135\n", "train_data最大日期: 2024-11-27\n", "test_data最大日期: 2024-11-28\n", "划分后的训练集大小: 654, 验证集大小: 138\n", "train_data最大日期: 2024-11-28\n", "test_data最大日期: 2024-11-29\n", "划分后的训练集大小: 656, 验证集大小: 134\n", "train_data最大日期: 2024-11-29\n", "test_data最大日期: 2024-12-02\n", "划分后的训练集大小: 675, 验证集大小: 137\n", "train_data最大日期: 2024-12-02\n", "test_data最大日期: 2024-12-03\n", "划分后的训练集大小: 678, 验证集大小: 134\n", "train_data最大日期: 2024-12-03\n", "test_data最大日期: 2024-12-04\n", "划分后的训练集大小: 672, 验证集大小: 129\n", "train_data最大日期: 2024-12-04\n", "test_data最大日期: 2024-12-05\n", "划分后的训练集大小: 670, 验证集大小: 136\n", "train_data最大日期: 2024-12-05\n", "test_data最大日期: 2024-12-06\n", "划分后的训练集大小: 646, 验证集大小: 110\n", "train_data最大日期: 2024-12-06\n", "test_data最大日期: 2024-12-09\n", "划分后的训练集大小: 626, 验证集大小: 117\n", "train_data最大日期: 2024-12-09\n", "test_data最大日期: 2024-12-10\n", "划分后的训练集大小: 625, 验证集大小: 133\n", "train_data最大日期: 2024-12-10\n", "test_data最大日期: 2024-12-11\n", "划分后的训练集大小: 611, 验证集大小: 115\n", "train_data最大日期: 2024-12-11\n", "test_data最大日期: 2024-12-12\n", "划分后的训练集大小: 618, 验证集大小: 143\n", "train_data最大日期: 2024-12-12\n", "test_data最大日期: 2024-12-13\n", "划分后的训练集大小: 639, 验证集大小: 131\n", "train_data最大日期: 2024-12-13\n", "test_data最大日期: 2024-12-16\n", "划分后的训练集大小: 624, 验证集大小: 102\n", "train_data最大日期: 2024-12-16\n", "test_data最大日期: 2024-12-17\n", "划分后的训练集大小: 612, 验证集大小: 121\n", "train_data最大日期: 2024-12-17\n", "test_data最大日期: 2024-12-18\n", "划分后的训练集大小: 609, 验证集大小: 112\n", "train_data最大日期: 2024-12-18\n", "test_data最大日期: 2024-12-19\n", "划分后的训练集大小: 603, 验证集大小: 137\n", "train_data最大日期: 2024-12-19\n", "test_data最大日期: 2024-12-20\n", "划分后的训练集大小: 617, 验证集大小: 145\n", "train_data最大日期: 2024-12-20\n", "test_data最大日期: 2024-12-23\n", "划分后的训练集大小: 619, 验证集大小: 104\n", "train_data最大日期: 2024-12-23\n", "test_data最大日期: 2024-12-24\n", "划分后的训练集大小: 611, 验证集大小: 113\n", "train_data最大日期: 2024-12-24\n", "test_data最大日期: 2024-12-25\n", "划分后的训练集大小: 640, 验证集大小: 141\n", "train_data最大日期: 2024-12-25\n", "test_data最大日期: 2024-12-26\n", "划分后的训练集大小: 647, 验证集大小: 144\n", "train_data最大日期: 2024-12-26\n", "test_data最大日期: 2024-12-27\n", "划分后的训练集大小: 635, 验证集大小: 133\n", "train_data最大日期: 2024-12-27\n", "test_data最大日期: 2024-12-30\n", "划分后的训练集大小: 657, 验证集大小: 126\n", "train_data最大日期: 2024-12-30\n", "test_data最大日期: 2024-12-31\n", "划分后的训练集大小: 645, 验证集大小: 101\n", "train_data最大日期: 2024-12-31\n", "test_data最大日期: 2025-01-02\n", "划分后的训练集大小: 598, 验证集大小: 94\n", "train_data最大日期: 2025-01-02\n", "test_data最大日期: 2025-01-03\n", "划分后的训练集大小: 550, 验证集大小: 96\n", "train_data最大日期: 2025-01-03\n", "test_data最大日期: 2025-01-06\n", "划分后的训练集大小: 541, 验证集大小: 124\n", "train_data最大日期: 2025-01-06\n", "test_data最大日期: 2025-01-07\n", "划分后的训练集大小: 510, 验证集大小: 95\n", "train_data最大日期: 2025-01-07\n", "test_data最大日期: 2025-01-08\n", "划分后的训练集大小: 554, 验证集大小: 145\n", "train_data最大日期: 2025-01-08\n", "test_data最大日期: 2025-01-09\n", "划分后的训练集大小: 573, 验证集大小: 113\n", "train_data最大日期: 2025-01-09\n", "test_data最大日期: 2025-01-10\n", "划分后的训练集大小: 612, 验证集大小: 135\n", "train_data最大日期: 2025-01-10\n", "test_data最大日期: 2025-01-13\n", "划分后的训练集大小: 625, 验证集大小: 137\n", "train_data最大日期: 2025-01-13\n", "test_data最大日期: 2025-01-14\n", "划分后的训练集大小: 629, 验证集大小: 99\n", "train_data最大日期: 2025-01-14\n", "test_data最大日期: 2025-01-15\n", "划分后的训练集大小: 629, 验证集大小: 145\n", "train_data最大日期: 2025-01-15\n", "test_data最大日期: 2025-01-16\n", "划分后的训练集大小: 663, 验证集大小: 147\n", "train_data最大日期: 2025-01-16\n", "test_data最大日期: 2025-01-17\n", "划分后的训练集大小: 619, 验证集大小: 91\n", "train_data最大日期: 2025-01-17\n", "test_data最大日期: 2025-01-20\n", "划分后的训练集大小: 612, 验证集大小: 130\n", "train_data最大日期: 2025-01-20\n", "test_data最大日期: 2025-01-21\n", "划分后的训练集大小: 661, 验证集大小: 148\n", "train_data最大日期: 2025-01-21\n", "test_data最大日期: 2025-01-22\n", "划分后的训练集大小: 660, 验证集大小: 144\n", "train_data最大日期: 2025-01-22\n", "test_data最大日期: 2025-01-23\n", "划分后的训练集大小: 653, 验证集大小: 140\n", "train_data最大日期: 2025-01-23\n", "test_data最大日期: 2025-01-24\n", "划分后的训练集大小: 708, 验证集大小: 146\n", "train_data最大日期: 2025-01-24\n", "test_data最大日期: 2025-01-27\n", "划分后的训练集大小: 722, 验证集大小: 144\n", "train_data最大日期: 2025-01-27\n", "test_data最大日期: 2025-02-05\n", "划分后的训练集大小: 668, 验证集大小: 94\n", "train_data最大日期: 2025-02-05\n", "test_data最大日期: 2025-02-06\n", "划分后的训练集大小: 656, 验证集大小: 132\n", "train_data最大日期: 2025-02-06\n", "test_data最大日期: 2025-02-07\n", "划分后的训练集大小: 609, 验证集大小: 93\n", "train_data最大日期: 2025-02-07\n", "test_data最大日期: 2025-02-10\n", "划分后的训练集大小: 582, 验证集大小: 119\n", "train_data最大日期: 2025-02-10\n", "test_data最大日期: 2025-02-11\n", "划分后的训练集大小: 577, 验证集大小: 139\n", "train_data最大日期: 2025-02-11\n", "test_data最大日期: 2025-02-12\n", "划分后的训练集大小: 608, 验证集大小: 125\n", "train_data最大日期: 2025-02-12\n", "test_data最大日期: 2025-02-13\n", "划分后的训练集大小: 617, 验证集大小: 141\n", "train_data最大日期: 2025-02-13\n", "test_data最大日期: 2025-02-14\n", "划分后的训练集大小: 629, 验证集大小: 105\n", "train_data最大日期: 2025-02-14\n", "test_data最大日期: 2025-02-17\n", "划分后的训练集大小: 651, 验证集大小: 141\n", "train_data最大日期: 2025-02-17\n", "test_data最大日期: 2025-02-18\n", "划分后的训练集大小: 656, 验证集大小: 144\n", "train_data最大日期: 2025-02-18\n", "test_data最大日期: 2025-02-19\n", "划分后的训练集大小: 645, 验证集大小: 114\n", "train_data最大日期: 2025-02-19\n", "test_data最大日期: 2025-02-20\n", "划分后的训练集大小: 598, 验证集大小: 94\n", "train_data最大日期: 2025-02-20\n", "test_data最大日期: 2025-02-21\n", "划分后的训练集大小: 636, 验证集大小: 143\n", "train_data最大日期: 2025-02-21\n", "test_data最大日期: 2025-02-24\n", "划分后的训练集大小: 639, 验证集大小: 144\n", "train_data最大日期: 2025-02-24\n", "test_data最大日期: 2025-02-25\n", "划分后的训练集大小: 634, 验证集大小: 139\n", "train_data最大日期: 2025-02-25\n", "test_data最大日期: 2025-02-26\n", "划分后的训练集大小: 663, 验证集大小: 143\n", "train_data最大日期: 2025-02-26\n", "test_data最大日期: 2025-02-27\n", "划分后的训练集大小: 701, 验证集大小: 132\n", "train_data最大日期: 2025-02-27\n", "test_data最大日期: 2025-02-28\n", "划分后的训练集大小: 694, 验证集大小: 136\n", "train_data最大日期: 2025-02-28\n", "test_data最大日期: 2025-03-03\n", "划分后的训练集大小: 670, 验证集大小: 120\n", "train_data最大日期: 2025-03-03\n", "test_data最大日期: 2025-03-04\n", "划分后的训练集大小: 671, 验证集大小: 140\n", "train_data最大日期: 2025-03-04\n", "test_data最大日期: 2025-03-05\n", "划分后的训练集大小: 673, 验证集大小: 145\n", "train_data最大日期: 2025-03-05\n", "test_data最大日期: 2025-03-06\n", "划分后的训练集大小: 650, 验证集大小: 109\n", "train_data最大日期: 2025-03-06\n", "test_data最大日期: 2025-03-07\n", "划分后的训练集大小: 659, 验证集大小: 145\n", "train_data最大日期: 2025-03-07\n", "test_data最大日期: 2025-03-10\n", "划分后的训练集大小: 682, 验证集大小: 143\n", "train_data最大日期: 2025-03-10\n", "test_data最大日期: 2025-03-11\n", "划分后的训练集大小: 638, 验证集大小: 96\n", "train_data最大日期: 2025-03-11\n", "test_data最大日期: 2025-03-12\n", "划分后的训练集大小: 640, 验证集大小: 147\n", "train_data最大日期: 2025-03-12\n", "test_data最大日期: 2025-03-13\n", "划分后的训练集大小: 674, 验证集大小: 143\n", "train_data最大日期: 2025-03-13\n", "test_data最大日期: 2025-03-14\n", "划分后的训练集大小: 678, 验证集大小: 149\n", "train_data最大日期: 2025-03-14\n", "test_data最大日期: 2025-03-17\n", "划分后的训练集大小: 642, 验证集大小: 107\n", "train_data最大日期: 2025-03-17\n", "test_data最大日期: 2025-03-18\n", "划分后的训练集大小: 691, 验证集大小: 145\n", "train_data最大日期: 2025-03-18\n", "test_data最大日期: 2025-03-19\n", "划分后的训练集大小: 633, 验证集大小: 89\n", "train_data最大日期: 2025-03-19\n", "test_data最大日期: 2025-03-20\n", "划分后的训练集大小: 610, 验证集大小: 120\n", "train_data最大日期: 2025-03-20\n", "test_data最大日期: 2025-03-21\n", "划分后的训练集大小: 601, 验证集大小: 140\n", "train_data最大日期: 2025-03-21\n", "test_data最大日期: 2025-03-24\n", "划分后的训练集大小: 629, 验证集大小: 135\n", "train_data最大日期: 2025-03-24\n", "test_data最大日期: 2025-03-25\n", "划分后的训练集大小: 596, 验证集大小: 112\n", "train_data最大日期: 2025-03-25\n", "test_data最大日期: 2025-03-26\n", "划分后的训练集大小: 651, 验证集大小: 144\n", "train_data最大日期: 2025-03-26\n", "test_data最大日期: 2025-03-27\n", "划分后的训练集大小: 671, 验证集大小: 140\n", "train_data最大日期: 2025-03-27\n", "test_data最大日期: 2025-03-28\n", "划分后的训练集大小: 623, 验证集大小: 92\n", "train_data最大日期: 2025-03-28\n", "test_data最大日期: 2025-03-31\n", "划分后的训练集大小: 633, 验证集大小: 145\n", "train_data最大日期: 2025-03-31\n", "test_data最大日期: 2025-04-01\n", "划分后的训练集大小: 666, 验证集大小: 145\n", "train_data最大日期: 2025-04-01\n", "test_data最大日期: 2025-04-02\n", "划分后的训练集大小: 637, 验证集大小: 115\n", "train_data最大日期: 2025-04-02\n", "test_data最大日期: 2025-04-03\n", "划分后的训练集大小: 620, 验证集大小: 123\n", "train_data最大日期: 2025-04-02\n", "test_data最大日期: 2025-04-07\n", "划分后的训练集大小: 528, 验证集大小: 123\n" ] } ], "execution_count": 43 }, { "cell_type": "code", "id": "e01fe33b-e30d-4bc6-bf40-de91e61862b4", "metadata": { "ExecuteTime": { "end_time": "2025-04-09T14:53:53.045759Z", "start_time": "2025-04-09T14:53:53.040829Z" } }, "source": [ "print(final_predictions[['ts_code', 'trade_date', 'score']].tail())" ], "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " ts_code trade_date score\n", "557 002851.SZ 2025-03-31 1.547811\n", "558 002847.SZ 2025-04-01 0.596230\n", "559 601595.SH 2025-04-02 -0.014094\n", "560 603353.SH 2025-04-03 1.512002\n", "561 600600.SH 2025-04-07 0.640985\n" ] } ], "execution_count": 44 }, { "cell_type": "code", "id": "0dc75517-c857-4f1d-8815-e807400a6d33", "metadata": { "ExecuteTime": { "end_time": "2025-04-09T14:53:53.055527Z", "start_time": "2025-04-09T14:53:53.049774Z" } }, "source": [ "# pdf1 = select_pre_zt_stocks_dynamic(df[(df['trade_date'] >= '2022-03-26') & (df['trade_date'] <= '2029-03-26')])\n", "# pdf1 = pdf1.replace([np.inf, -np.inf], np.nan)\n", "#\n", "#\n", "# pdf1 = time_series_quantile_filter(pdf1, numeric_columns)\n", "#\n", "# # pdf1 = cross_sectional_standardization(pdf1, numeric_columns)\n", "# # pdf1 = pdf1[pdf1['trade_date'] <= '2025-03-26']\n", "# # pdf1 = pdf1.sort_values(by=['ts_code', 'trade_date'])\n", "# filter_index1 = pdf1['future_return'].between(pdf1['future_return'].quantile(0.01), pdf1['future_return'].quantile(0.99))\n", "#\n", "# print('-----------------------------------------')\n", "# pdf2 = select_pre_zt_stocks_dynamic(df[(df['trade_date'] >= '2022-03-26') & (df['trade_date'] <= '2025-03-26')])\n", "# pdf2 = pdf2.replace([np.inf, -np.inf], np.nan)\n", "#\n", "#\n", "# pdf2 = time_series_quantile_filter(pdf2, numeric_columns)\n", "#\n", "# # pdf2 = cross_sectional_standardization(pdf2, numeric_columns)\n", "#\n", "# # pdf2 = pdf2[pdf2['trade_date'] <= '2025-03-26']\n", "# # pdf2 = pdf2.sort_values(by=['ts_code', 'trade_date'])\n", "# filter_index2 = pdf2['future_return'].between(pdf2['future_return'].quantile(0.01), pdf2['future_return'].quantile(0.99))\n" ], "outputs": [], "execution_count": 45 }, { "cell_type": "code", "id": "8299a6f461097f14", "metadata": { "ExecuteTime": { "end_time": "2025-04-09T14:53:53.135966Z", "start_time": "2025-04-09T14:53:53.131344Z" } }, "source": [ "# are_equal = pdf1[filter_index1].equals(pdf2[filter_index2])\n", "# print(are_equal) # 输出 True 或 False\n", "#\n", "# are_equal = pdf1.equals(pdf2)\n", "# print(are_equal) # 输出 True 或 False\n", "#\n", "# are_equal = filter_index1.equals(filter_index2)\n", "# print(are_equal) # 输出 True 或 False" ], "outputs": [], "execution_count": 46 }, { "cell_type": "code", "id": "3f5079aa2c937c22", "metadata": { "ExecuteTime": { "end_time": "2025-04-09T14:53:53.148195Z", "start_time": "2025-04-09T14:53:53.138983Z" } }, "source": [ "# final_predictions1 = rolling_train_predict(\n", "# pdf1[(pdf1['trade_date'] >= '2024-12-01') & (pdf1['trade_date'] <= '2029-03-26')], 5, 1, feature_columns,\n", "# days=days, validation_days=0, filter_index=filter_index1, params=light_params)\n", "# final_predictions.to_csv('test1.tsv', index=False)\n", "#\n", "# final_predictions2 = rolling_train_predict(\n", "# pdf2[(pdf2['trade_date'] >= '2024-12-01') & (pdf2['trade_date'] <= '2029-03-26')], 5, 1, feature_columns,\n", "# days=days, validation_days=0, filter_index=filter_index2, params=light_params)\n", "# final_predictions2.to_csv('test2.tsv', index=False)" ], "outputs": [], "execution_count": 47 }, { "cell_type": "code", "id": "199b12e7e20e4e6a", "metadata": { "ExecuteTime": { "end_time": "2025-04-09T14:53:53.186822Z", "start_time": "2025-04-09T14:53:53.184469Z" } }, "source": [ "# print(final_predictions1['trade_date'].max())\n", "# print(final_predictions2['trade_date'].max())\n", "#\n", "# are_equal = final_predictions1[(final_predictions1['trade_date'] >= '2022-12-01') & (final_predictions1['trade_date'] <= '2025-03-26')].equals(final_predictions2[(final_predictions2['trade_date'] >= '2022-12-01') & (final_predictions2['trade_date'] <= '2025-03-26')])\n", "# print(are_equal) # 输出 True 或 False" ], "outputs": [], "execution_count": 48 } ], "metadata": { "kernelspec": { "display_name": "Python 3 (ipykernel)", "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": "3.11.11" } }, "nbformat": 4, "nbformat_minor": 5 }