huice/check_buy1_price_decline.py

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
检查买1量大于99的挂单后是否出现低于当前价格的数据

详细分析买1量大单出现后的价格行为，验证是否真的全部上涨
"""

import pandas as pd
import numpy as np
import os

def check_buy1_price_decline():
    """检查买1量大单后的价格下跌情况"""

    # 加载数据
    data_file = 'data/au2512_20251013.parquet'
    if not os.path.exists(data_file):
        print(f"数据文件不存在: {data_file}")
        return

    print("正在加载数据...")
    df = pd.read_parquet(data_file)
    print(f"数据加载成功: {len(df):,} 条记录")

    # 确保按数列号排序
    if '数列号' in df.columns:
        df = df.sort_values('数列号').reset_index(drop=True)

    # 找到买1量列
    buy1_col = None
    for col in df.columns:
        if '买1量' in col:
            buy1_col = col
            break

    if buy1_col is None:
        print("未找到买1量列")
        return

    print(f"使用买1量列: {buy1_col}")

    # 筛选买1量大于99的数据
    threshold = 99
    large_buy1_mask = df[buy1_col] > threshold
    large_buy1_orders = df[large_buy1_mask].copy().reset_index(drop=True)

    print(f"\n买1量 > {threshold} 的记录数: {len(large_buy1_orders):,}")
    print(f"占总记录比例: {len(large_buy1_orders)/len(df)*100:.2f}%")

    if len(large_buy1_orders) == 0:
        print("未找到符合条件的记录")
        return

    # 详细分析每个大单后的价格行为
    print(f"\n" + "="*80)
    print("买1量大单后价格行为详细分析")
    print("="*80)

    decline_100_count = 0
    decline_200_count = 0
    detailed_results = []

    for idx, row in large_buy1_orders.iterrows():
        current_seq_num = row['数列号'] if '数列号' in row else idx
        current_price = row['成交价']
        buy1_volume = row[buy1_col]

        # 获取后续100笔数据
        future_mask_100 = df['数列号'] >= current_seq_num + 1
        future_data_100 = df[future_mask_100].head(100)

        # 获取后续200笔数据
        future_mask_200 = df['数列号'] >= current_seq_num + 1
        future_data_200 = df[future_mask_200].head(200)

        # 分析100笔内的价格行为
        if len(future_data_100) > 0:
            min_price_100 = future_data_100['成交价'].min()
            max_price_100 = future_data_100['成交价'].max()
            price_decline_100 = current_price - min_price_100  # 价格下跌幅度
            price_rise_100 = max_price_100 - current_price      # 价格上涨幅度

            # 检查是否出现低于当前价格的情况
            has_decline_100 = min_price_100 < current_price

            if has_decline_100:
                decline_100_count += 1
                decline_info_100 = f"是 (最低: {min_price_100:.2f}, 下跌: {price_decline_100:.2f})"
            else:
                decline_info_100 = f"否 (最低: {min_price_100:.2f})"

        # 分析200笔内的价格行为
        if len(future_data_200) > 0:
            min_price_200 = future_data_200['成交价'].min()
            max_price_200 = future_data_200['成交价'].max()
            price_decline_200 = current_price - min_price_200  # 价格下跌幅度
            price_rise_200 = max_price_200 - current_price      # 价格上涨幅度

            # 检查是否出现低于当前价格的情况
            has_decline_200 = min_price_200 < current_price

            if has_decline_200:
                decline_200_count += 1
                decline_info_200 = f"是 (最低: {min_price_200:.2f}, 下跌: {price_decline_200:.2f})"
            else:
                decline_info_200 = f"否 (最低: {min_price_200:.2f})"

        # 记录详细结果
        detailed_results.append({
            '序号': idx + 1,
            '数列号': current_seq_num,
            '当前价格': current_price,
            '买1量': buy1_volume,
            '100笔样本数': len(future_data_100),
            '100笔最低价': min_price_100 if len(future_data_100) > 0 else None,
            '100笔最高价': max_price_100 if len(future_data_100) > 0 else None,
            '100笔是否下跌': has_decline_100 if len(future_data_100) > 0 else None,
            '100笔下跌幅度': price_decline_100 if len(future_data_100) > 0 else None,
            '100笔上涨幅度': price_rise_100 if len(future_data_100) > 0 else None,
            '200笔样本数': len(future_data_200),
            '200笔最低价': min_price_200 if len(future_data_200) > 0 else None,
            '200笔最高价': max_price_200 if len(future_data_200) > 0 else None,
            '200笔是否下跌': has_decline_200 if len(future_data_200) > 0 else None,
            '200笔下跌幅度': price_decline_200 if len(future_data_200) > 0 else None,
            '200笔上涨幅度': price_rise_200 if len(future_data_200) > 0 else None,
        })

    # 保存详细结果
    results_df = pd.DataFrame(detailed_results)
    results_df.to_csv('large_orders/buy1_detailed_price_analysis.csv', index=False, encoding='utf-8-sig')

    # 打印汇总统计
    print(f"\n【汇总统计】")
    print(f"总样本数: {len(large_buy1_orders)}")
    print(f"100笔内出现价格下跌的样本数: {decline_100_count} ({decline_100_count/len(large_buy1_orders)*100:.1f}%)")
    print(f"200笔内出现价格下跌的样本数: {decline_200_count} ({decline_200_count/len(large_buy1_orders)*100:.1f}%)")

    # 打印详细分析
    print(f"\n【详细分析 - 前15个样本】")
    print(f"{'序号':>4} {'数列号':>8} {'当前价格':>10} {'买1量':>8} {'100笔下跌?':>12} {'200笔下跌?':>12}")
    print("-" * 80)

    for i, result in enumerate(detailed_results[:15]):
        decline_100_status = "是" if result['100笔是否下跌'] else "否"
        decline_200_status = "是" if result['200笔是否下跌'] else "否"

        print(f"{result['序号']:>4} {result['数列号']:>8.0f} {result['当前价格']:>10.2f} "
              f"{result['买1量']:>8.0f} {decline_100_status:>12} {decline_200_status:>12}")

    # 分析下跌样本的详细信息
    decline_samples_100 = [r for r in detailed_results if r['100笔是否下跌']]
    decline_samples_200 = [r for r in detailed_results if r['200笔是否下跌']]

    if decline_samples_100:
        print(f"\n【100笔内出现价格下跌的样本详情】")
        for result in decline_samples_100:
            print(f"序号{result['序号']}: 当前价格={result['当前价格']:.2f}, "
                  f"最低价格={result['100笔最低价']:.2f}, "
                  f"下跌幅度={result['100笔下跌幅度']:.2f}元, "
                  f"买1量={result['买1量']:.0f}手")

    if decline_samples_200:
        print(f"\n【200笔内出现价格下跌的样本详情】")
        for result in decline_samples_200:
            print(f"序号{result['序号']}: 当前价格={result['当前价格']:.2f}, "
                  f"最低价格={result['200笔最低价']:.2f}, "
                  f"下跌幅度={result['200笔下跌幅度']:.2f}元, "
                  f"买1量={result['买1量']:.0f}手")

    # 价格波动统计
    print(f"\n【价格波动统计】")
    if len(detailed_results) > 0:
        all_declines_100 = [r['100笔下跌幅度'] for r in detailed_results if r['100笔下跌幅度'] is not None]
        all_rises_100 = [r['100笔上涨幅度'] for r in detailed_results if r['100笔上涨幅度'] is not None]

        if all_declines_100:
            avg_decline_100 = np.mean(all_declines_100)
            max_decline_100 = np.max(all_declines_100)
            print(f"100笔内平均下跌幅度: {avg_decline_100:.3f}元")
            print(f"100笔内最大下跌幅度: {max_decline_100:.3f}元")

        if all_rises_100:
            avg_rise_100 = np.mean(all_rises_100)
            max_rise_100 = np.max(all_rises_100)
            print(f"100笔内平均上涨幅度: {avg_rise_100:.3f}元")
            print(f"100笔内最大上涨幅度: {max_rise_100:.3f}元")

    print(f"\n分析完成！详细结果已保存到: large_orders/buy1_detailed_price_analysis.csv")

    return decline_100_count, decline_200_count

if __name__ == "__main__":
    check_buy1_price_decline()