{
  "nbformat": 4,
  "nbformat_minor": 5,
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "# Pandas 合成数据实验\n",
        "不是实际行情。下载配套CSV及pandas_lab.py到同目录，先按README安装依赖。运行下面单元可重现完整清洗、特征与周OHLC聚合；脚本内含数值断言。\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {},
      "execution_count": null,
      "outputs": [],
      "source": [
        "from pandas_lab import run\n",
        "result = run()\n",
        "assert result['valid'] == 240\n",
        "assert result['duplicates'] == 1\n",
        "assert result['rejected'] == 3\n",
        "result\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "## 独立练习\n",
        "把SMA3改成SMA5，说明前几行为何缺失；不要用未来数据回填。核对周开高低收的聚合规则，不用列平均代替。记录版本、输出与未运行项。"
      ]
    }
  ]
}