lgbm_stock_project.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "machine_shape": "hm",
      "gpuType": "L4"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tmtuzOuwoJiw",
        "outputId": "061505ad-c8e1-4edf-c133-3295c6f2ea44"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Requirement already satisfied: seaborn in /usr/local/lib/python3.10/dist-packages (0.13.2)\n",
            "Requirement already satisfied: numpy!=1.24.0,>=1.20 in /usr/local/lib/python3.10/dist-packages (from seaborn) (1.26.4)\n",
            "Requirement already satisfied: pandas>=1.2 in /usr/local/lib/python3.10/dist-packages (from seaborn) (2.2.2)\n",
            "Requirement already satisfied: matplotlib!=3.6.1,>=3.4 in /usr/local/lib/python3.10/dist-packages (from seaborn) (3.7.1)\n",
            "Requirement already satisfied: contourpy>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (1.3.0)\n",
            "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.10/dist-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (0.12.1)\n",
            "Requirement already satisfied: fonttools>=4.22.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (4.54.1)\n",
            "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (1.4.7)\n",
            "Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (24.1)\n",
            "Requirement already satisfied: pillow>=6.2.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (10.4.0)\n",
            "Requirement already satisfied: pyparsing>=2.3.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (3.1.4)\n",
            "Requirement already satisfied: python-dateutil>=2.7 in /usr/local/lib/python3.10/dist-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (2.8.2)\n",
            "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.10/dist-packages (from pandas>=1.2->seaborn) (2024.2)\n",
            "Requirement already satisfied: tzdata>=2022.7 in /usr/local/lib/python3.10/dist-packages (from pandas>=1.2->seaborn) (2024.2)\n",
            "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.10/dist-packages (from python-dateutil>=2.7->matplotlib!=3.6.1,>=3.4->seaborn) (1.16.0)\n",
            "Collecting pyalgotrade\n",
            "  Downloading PyAlgoTrade-0.20.tar.gz (84 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m84.6/84.6 kB\u001b[0m \u001b[31m3.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25h  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "Requirement already satisfied: matplotlib in /usr/local/lib/python3.10/dist-packages (from pyalgotrade) (3.7.1)\n",
            "Requirement already satisfied: numpy in /usr/local/lib/python3.10/dist-packages (from pyalgotrade) (1.26.4)\n",
            "Requirement already satisfied: python-dateutil in /usr/local/lib/python3.10/dist-packages (from pyalgotrade) (2.8.2)\n",
            "Requirement already satisfied: pytz in /usr/local/lib/python3.10/dist-packages (from pyalgotrade) (2024.2)\n",
            "Requirement already satisfied: requests in /usr/local/lib/python3.10/dist-packages (from pyalgotrade) (2.32.3)\n",
            "Collecting retrying (from pyalgotrade)\n",
            "  Downloading retrying-1.3.4-py3-none-any.whl.metadata (6.9 kB)\n",
            "Requirement already satisfied: scipy in /usr/local/lib/python3.10/dist-packages (from pyalgotrade) (1.13.1)\n",
            "Requirement already satisfied: six in /usr/local/lib/python3.10/dist-packages (from pyalgotrade) (1.16.0)\n",
            "Requirement already satisfied: tornado in /usr/local/lib/python3.10/dist-packages (from pyalgotrade) (6.3.3)\n",
            "Requirement already satisfied: tweepy in /usr/local/lib/python3.10/dist-packages (from pyalgotrade) (4.14.0)\n",
            "Collecting ws4py>=0.3.4 (from pyalgotrade)\n",
            "  Downloading ws4py-0.5.1.tar.gz (51 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m51.4/51.4 kB\u001b[0m \u001b[31m4.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25h  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "Requirement already satisfied: contourpy>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib->pyalgotrade) (1.3.0)\n",
            "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.10/dist-packages (from matplotlib->pyalgotrade) (0.12.1)\n",
            "Requirement already satisfied: fonttools>=4.22.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib->pyalgotrade) (4.54.1)\n",
            "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib->pyalgotrade) (1.4.7)\n",
            "Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib->pyalgotrade) (24.1)\n",
            "Requirement already satisfied: pillow>=6.2.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib->pyalgotrade) (10.4.0)\n",
            "Requirement already satisfied: pyparsing>=2.3.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib->pyalgotrade) (3.1.4)\n",
            "Requirement already satisfied: charset-normalizer<4,>=2 in /usr/local/lib/python3.10/dist-packages (from requests->pyalgotrade) (3.3.2)\n",
            "Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests->pyalgotrade) (3.10)\n",
            "Requirement already satisfied: urllib3<3,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests->pyalgotrade) (2.2.3)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests->pyalgotrade) (2024.8.30)\n",
            "Requirement already satisfied: oauthlib<4,>=3.2.0 in /usr/local/lib/python3.10/dist-packages (from tweepy->pyalgotrade) (3.2.2)\n",
            "Requirement already satisfied: requests-oauthlib<2,>=1.2.0 in /usr/local/lib/python3.10/dist-packages (from tweepy->pyalgotrade) (1.3.1)\n",
            "Downloading retrying-1.3.4-py3-none-any.whl (11 kB)\n",
            "Building wheels for collected packages: pyalgotrade, ws4py\n",
            "  Building wheel for pyalgotrade (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for pyalgotrade: filename=PyAlgoTrade-0.20-py3-none-any.whl size=159512 sha256=33e87bd86f76f4513363d3b46a275f4aed0dc2c99d0cafdd44c9608fc6c00cb8\n",
            "  Stored in directory: /root/.cache/pip/wheels/72/46/5c/ec695f4cbbdfbf40269fa8a4096c418670af83ed9aa40d669f\n",
            "  Building wheel for ws4py (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for ws4py: filename=ws4py-0.5.1-py3-none-any.whl size=45229 sha256=7c4d07ea4e2e147aaca70151477572e12687daaa47de66ee4e3c9cc125df40a4\n",
            "  Stored in directory: /root/.cache/pip/wheels/2e/7c/ad/d9c746276bf024d44296340869fcb169f1e5d80fb147351a57\n",
            "Successfully built pyalgotrade ws4py\n",
            "Installing collected packages: ws4py, retrying, pyalgotrade\n",
            "Successfully installed pyalgotrade-0.20 retrying-1.3.4 ws4py-0.5.1\n",
            "Collecting bayesian-optimization\n",
            "  Downloading bayesian_optimization-1.5.1-py3-none-any.whl.metadata (16 kB)\n",
            "Collecting colorama<0.5.0,>=0.4.6 (from bayesian-optimization)\n",
            "  Downloading colorama-0.4.6-py2.py3-none-any.whl.metadata (17 kB)\n",
            "Requirement already satisfied: numpy>=1.25 in /usr/local/lib/python3.10/dist-packages (from bayesian-optimization) (1.26.4)\n",
            "Requirement already satisfied: scikit-learn<2.0.0,>=1.0.0 in /usr/local/lib/python3.10/dist-packages (from bayesian-optimization) (1.5.2)\n",
            "Requirement already satisfied: scipy<2.0.0,>=1.0.0 in /usr/local/lib/python3.10/dist-packages (from bayesian-optimization) (1.13.1)\n",
            "Requirement already satisfied: joblib>=1.2.0 in /usr/local/lib/python3.10/dist-packages (from scikit-learn<2.0.0,>=1.0.0->bayesian-optimization) (1.4.2)\n",
            "Requirement already satisfied: threadpoolctl>=3.1.0 in /usr/local/lib/python3.10/dist-packages (from scikit-learn<2.0.0,>=1.0.0->bayesian-optimization) (3.5.0)\n",
            "Downloading bayesian_optimization-1.5.1-py3-none-any.whl (28 kB)\n",
            "Downloading colorama-0.4.6-py2.py3-none-any.whl (25 kB)\n",
            "Installing collected packages: colorama, bayesian-optimization\n",
            "Successfully installed bayesian-optimization-1.5.1 colorama-0.4.6\n"
          ]
        }
      ],
      "source": [
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.optim as optim\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "!pip install seaborn\n",
        "import seaborn as sns\n",
        "import os\n",
        "\n",
        "import random\n",
        "from torch.utils.data import Dataset, DataLoader\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "!pip install pyalgotrade\n",
        "from pyalgotrade import strategy\n",
        "from pyalgotrade.barfeed import yahoofeed\n",
        "from pyalgotrade.technical import ma\n",
        "from pyalgotrade.stratanalyzer import returns\n",
        "from pyalgotrade.bar import BasicBar, Frequency\n",
        "from pyalgotrade.barfeed import membf\n",
        "import datetime\n",
        "from pyalgotrade.stratanalyzer import returns, trades, sharpe\n",
        "import logging\n",
        "\n",
        "import sklearn\n",
        "from sklearn.metrics import roc_auc_score\n",
        "\n",
        "!pip install bayesian-optimization\n",
        "from bayes_opt import BayesianOptimization"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "LGBM import"
      ],
      "metadata": {
        "id": "uWuQRpDBoTPy"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install lightgbm\n",
        "import lightgbm as lgbm"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4Bs-ELnjoNl7",
        "outputId": "b470443a-87c9-4078-d474-4527a80ec22c"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Requirement already satisfied: lightgbm in /usr/local/lib/python3.10/dist-packages (4.5.0)\n",
            "Requirement already satisfied: numpy>=1.17.0 in /usr/local/lib/python3.10/dist-packages (from lightgbm) (1.26.4)\n",
            "Requirement already satisfied: scipy in /usr/local/lib/python3.10/dist-packages (from lightgbm) (1.13.1)\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/dask/dataframe/__init__.py:42: FutureWarning: \n",
            "Dask dataframe query planning is disabled because dask-expr is not installed.\n",
            "\n",
            "You can install it with `pip install dask[dataframe]` or `conda install dask`.\n",
            "This will raise in a future version.\n",
            "\n",
            "  warnings.warn(msg, FutureWarning)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rma8zn68oNn5",
        "outputId": "c4d9b5eb-bace-4fe2-c582-94090ce6cadb"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "data preprocessing"
      ],
      "metadata": {
        "id": "u1GwyByOoRgj"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# file path\n",
        "folder_path = '/content/drive/MyDrive/kospi_sim/'\n",
        "csv_files = [f for f in os.listdir(folder_path) if f.endswith('.csv')]\n",
        "\n",
        "# save each csv file to df list\n",
        "df = []\n",
        "for i , file_name in enumerate(csv_files):\n",
        "  file_path = os.path.join(folder_path, file_name)\n",
        "  p=pd.read_csv(file_path)\n",
        "  df.append(p)"
      ],
      "metadata": {
        "id": "ZvB2_PfYoNp-"
      },
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# rename columns , drop 'date' , 'Adj Close'\n",
        "stock_len = int(len(df))\n",
        "for i in range(stock_len):\n",
        "  if 'Adj Close' in df[i].columns:\n",
        "    df[i] = df[i].drop(['Adj Close'] , axis = 1)\n",
        "  if 'Date' in df[i].columns:\n",
        "    df[i] = df[i].drop(['Date'] , axis = 1)\n",
        "for i in range(len(df)):\n",
        "  df[i] = df[i].rename(columns={'close': 'Close', 'open': 'Open' , 'low': 'Low', 'high' : 'High' , 'volume' : 'Volume'})"
      ],
      "metadata": {
        "id": "V-8SylQMoNr5"
      },
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# if 0 exist, delete\n",
        "for i in range(len(df)):\n",
        "  s = df[i]\n",
        "  s = s.drop(s[s['Close'] == 0].index , axis=0)\n",
        "  s = s.drop(s[s['Open'] == 0].index , axis=0)\n",
        "  s = s.drop(s[s['Volume'] == 0].index , axis=0)\n",
        "  s = s.dropna(axis=0)\n",
        "  df[i] = s"
      ],
      "metadata": {
        "id": "I9ek7zHkoNuB"
      },
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# generating target col\n",
        "stock_len = int(len(df))\n",
        "for i in range(stock_len):\n",
        "  s = df[i]\n",
        "  mid = pd.DataFrame()\n",
        "\n",
        "  for j in range(1,11):\n",
        "    mid[f'close_{j}'] = s['Close'].shift(-j)\n",
        "  for j in range(1,11):\n",
        "    mid[f'open_{j}'] = s['Open'].shift(-j)\n",
        "\n",
        "  mid['target'] = ((s['Close'] - mid['close_1'])/mid['close_1']) + 1\n",
        "  mid = mid.dropna(axis=0)\n",
        "  mid[mid > 300000] = 300000\n",
        "  mid[mid < -300000] = -300000\n",
        "  df[i] = mid"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "collapsed": true,
        "id": "HXIAx5Yxotjk",
        "outputId": "98e7f8f8-2bf3-4a07-b543-f0d1367e56ba"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid > 300000] = 300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n",
            "<ipython-input-8-d8a0cc7e48a1>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  mid[mid < -300000] = -300000\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "s = df[0]\n",
        "for i in range(1,len(df)):\n",
        "  s = pd.concat([s,df[i]])\n",
        "df = s"
      ],
      "metadata": {
        "id": "-BAVH7s6ruZW"
      },
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#결측값 제거\n",
        "df = df.dropna(axis = 0)\n",
        "\n",
        "# data, target 분리\n",
        "df_data = df.iloc[:,:-1]\n",
        "df_target = df.iloc[:,-1]\n",
        "\n",
        "# target 클래스화\n",
        "df_where = df_target.where(df_target >= 1, -1)\n",
        "df_target = df_where.where(df_where < 1, 1)\n",
        "\n",
        "# scaling\n",
        "df_data['row_max'] = df_data.max(axis=1)\n",
        "df_data['row_min'] = df_data.min(axis = 1)\n",
        "\n",
        "df_data = df_data.apply(lambda row: (row - row['row_min']) / (row['row_max'] - row['row_min']), axis=1)\n",
        "df_data.drop(['row_max','row_min'] , axis= 1 , inplace=True)"
      ],
      "metadata": {
        "id": "xhhFEZlSotob"
      },
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "second data preprocessing"
      ],
      "metadata": {
        "id": "yb-81n6_ECVx"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "df_data['close_2_/_close_3'] = df_data['close_2'] / df_data['close_3']\n",
        "df_data['close_1_clse_2_close_3'] = df_data['close_1'] * df_data['close_2'] * df_data['close_3']"
      ],
      "metadata": {
        "id": "h24udzWqEE0P"
      },
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "dataframe -> tensor"
      ],
      "metadata": {
        "id": "diVEO9EQo9Bh"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def df_to_tensor(df):\n",
        "  return torch.tensor(df.values)"
      ],
      "metadata": {
        "id": "XGPr44noo8lG"
      },
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Heatmap"
      ],
      "metadata": {
        "id": "8j-2NrHBpB63"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def heatmap(df , df_final):\n",
        "  # 데이터프레임의 상관계수 계산\n",
        "\n",
        "  df['pred'] = df_final # df_final.shape : (n,1)\n",
        "  corr = df.corr()\n",
        "\n",
        "  # 상관계수 히트맵 그리기\n",
        "  plt.figure(figsize=(8, 6))\n",
        "  sns.heatmap(corr, annot=False, cmap='coolwarm', linewidths=0.5)\n",
        "  plt.title('Correlation Heatmap')\n",
        "  plt.show()"
      ],
      "metadata": {
        "id": "_le4Wsa3pBXE"
      },
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "further processing"
      ],
      "metadata": {
        "id": "8gx4dahEpEOZ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def data_preprocessing(df):\n",
        "  return df"
      ],
      "metadata": {
        "id": "es1XC4L7pE9r"
      },
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def feature_importance(lgb_model):\n",
        "  lgbm.plot_importance(lgb_model, max_num_features=10, importance_type='split')  # 'split' or 'gain' 옵션 가능\n",
        "  plt.title(\"Feature Importance using LightGBM\")\n",
        "  plt.show()"
      ],
      "metadata": {
        "id": "ZEINIxbfpGag"
      },
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "kospi , kosdaq_list"
      ],
      "metadata": {
        "id": "PjPSMz7npHjV"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "kosdaq = ['900110', '900270', '900260', '900290', '900300', '900310', '900340', '000250', '000440', '001000', '001540', '001810', '001840', '002230', '002290', '002680', '002800', '003100', '003310', '003380', '003800', '004590', '004650', '004780', '005160', '005290', '005670', '005710', '005860', '005990', '006050', '006140', '006620', '006730', '006910', '006920', '007330', '007370', '007390', '007530', '007680', '007720', '007770', '007820', '008290', '008370', '008470', '008830', '009300', '009520', '009620', '009730', '009780', '010170', '010240', '010280', '010470', '011040', '011080', '011320', '011370', '011560', '012340', '012620', '012700', '012790', '012860', '013030', '013120', '013310', '013720', '013810', '013990', '014100', '014190', '014200', '014470', '014570', '014620', '014940', '014970', '015710', '015750', '016100', '016250', '016600', '016670', '016790', '016920', '017000', '017250', '017480', '017510', '017650', '017890', '018000', '018120', '018290', '018310', '018620', '018680', '018700', '019010', '019210', '019540', '019550', '019570', '019590', '019660', '019770', '019990', '020180', '020400', '020710', '021040', '021045', '021080', '021320', '021650', '021880', '022220', '023160', '023410', '023440', '023460', '023600', '023760', '023770', '023790', '023900', '023910', '024060', '024120', '024740', '024800', '024810', '024830', '024840', '024850', '024880', '024910', '024940', '024950', '025320', '025440', '025550', '025770', '025870', '025880', '025900', '025950', '025980', '026040', '026150', '026910', '027040', '027050', '027360', '027580', '027710', '027830', '028080', '028300', '029480', '029960', '030350', '030520', '030530', '030960', '031310', '031330', '031510', '031860', '031980', '032080', '032190', '032280', '032300', '032500', '032540', '032580', '032620', '032680', '032685', '032750', '032790', '032800', '032820', '032850', '032860', '032940', '032960', '032980', '033050', '033100', '033130', '033160', '033170', '033200', '033230', '033290', '033310', '033320', '033340', '033500', '033540', '033560', '033640', '033790', '033830', '034810', '03481K', '034940', '034950', '035080', '035200', '035290', '035460', '035600', '035610', '035620', '035760', '035810', '035890', '035900', '036000', '036010', '036030', '036090', '036120', '036170', '036180', '036190', '036200', '036220', '036480', '036540', '036560', '036620', '036630', '036640', '036670', '036690', '036710', '036800', '036810', '036830', '036890', '036930', '037030', '037070', '037230', '037330', '037350', '037370', '037400', '037440', '037460', '037760', '037950', '038010', '038060', '038070', '038110', '038290', '038340', '038390', '038460', '038500', '038530', '038540', '038620', '038680', '038870', '038880', '038950', '039010', '039020', '039030', '039200', '039240', '039290', '039310', '039340', '039420', '039440', '039560', '039610', '039740', '039830', '039840', '039860', '039980', '040160', '040300', '040350', '040420', '040610', '040910', '041020', '041190', '041440', '041460', '041510', '041520', '041590', '041830', '041910', '041920', '041930', '041960', '042000', '042040', '042110', '042370', '042420', '042500', '042510', '042520', '042600', '042940', '043090', '043100', '043150', '043200', '043220', '043260', '043340', '043360', '043370', '043590', '043610', '043650', '043710', '043910', '044060', '044180', '044340', '044480', '044490', '044780', '044960', '045060', '045100', '045300', '045340', '045390', '045510', '045520', '045660', '045970', '046070', '046120', '046210', '046310', '046390', '046440', '046890', '046940', '046970', '047080', '047310', '047560', '047770', '047820', '047920', '048410', '048430', '048470', '048530', '048550', '048770', '048830', '048870', '048910', '049070', '049080', '049120', '049180', '049430', '049470', '049480', '049520', '049550', '049630', '049720', '049830', '049950', '049960', '050090', '050110', '050120', '050760', '050860', '050890', '050960', '051160', '051360', '051370', '051380']\n",
        "kospi = ['000020', '000040', '000050', '000070', '000075', '000080', '000087', '000100', '000105', '000120', '000140', '000145', '000150', '000155', '000157', '000180', '000210', '000215', '000220', '000225', '000227', '000230', '000240', '000270', '000300', '000320', '000325', '000370', '000390', '000400', '000430', '000480', '000490', '000500', '000520', '000540', '000545', '000590', '000640', '000650', '000660', '000670', '000680', '000700', '000720', '000725', '000760', '000810', '000815', '000850', '000860', '000880', '000885', '00088K', '000890', '000910', '000950', '000970', '000990', '001020', '001040', '001045', '00104K', '001060', '001065', '001067', '001070', '001080', '001120', '001130', '001140', '001200', '001210', '001230', '001250', '001260', '001270', '001275', '001290', '001340', '001360', '001380', '001390', '001420', '001430', '001440', '001450', '001460', '001465', '001470', '001500', '001510', '001515', '001520', '001525', '001527', '001530', '001550', '001560', '001570', '001620', '001630', '001680', '001685', '001720', '001740', '001750', '001755', '001770', '001780', '001790', '001795', '001800', '001820', '001940', '002020', '002025', '002030', '002070', '002100', '002140', '002150', '002170', '002200', '002210', '002220', '002240', '002310', '002320', '002350', '002355', '002360', '002380', '002390', '002410', '002420', '002450', '002460', '002600', '002620', '002630', '002690', '002700', '002710', '002720', '002760', '002780', '002785', '002787', '002790', '002795', '00279K', '002810', '002820', '002840', '002870', '002880', '002900', '002920', '002960', '002990', '002995', '003000', '003010', '003030', '003060', '003070', '003075', '003080', '003090', '003120', '003160', '003200', '003220', '003230', '003240', '003280', '003300', '003350', '003460', '003465', '003470', '003475', '003480', '003490', '003495', '003520', '003530', '003535', '003540', '003545', '003547', '003550', '003555', '003560', '003570', '003580', '003610', '003620', '003650', '003670', '003680', '003690', '003720', '003780', '003830', '003850', '003920', '003925', '003960', '004000', '004020', '004060', '004080', '004090', '004100', '004105', '004140', '004150', '004170', '004250', '004255', '004270', '004310', '004360', '004365', '004370', '004380', '004410', '004415', '004430', '004440', '004450', '004490', '004540', '004545', '004560', '004690', '004700', '004710', '004720', '004770', '004800', '004830', '004835', '004840', '004870', '004890', '004910', '004920', '004960', '004970', '004980', '004985', '004990', '00499K', '005010', '005030', '005070', '005090', '005110', '005180', '005250', '005257', '005300', '005305', '005320', '005360', '005380', '005385', '005387', '005389', '005390', '005420', '005430', '005440', '005490', '005500', '005610', '005680', '005690', '005720', '005725', '005740', '005745', '005750', '005800', '005810', '005820', '005830', '005850', '005870', '005880', '005930', '005935', '005940', '005945', '005950', '005960', '005965', '006040', '006060', '006090', '006110', '006120', '006125', '006200', '006220', '006260', '006280', '006340', '006345', '006360', '006370', '006380', '006390', '006400', '006405', '006490', '006570', '006650', '006660', '006740', '006800', '006805', '00680K', '006840', '006880', '006890', '006980', '007070', '007110', '007120', '007160', '007210', '007280', '007310', '007340', '007460', '007540', '007570', '007575', '007590', '007610', '007660', '007690', '007700', '007810', '007815', '00781K', '007860', '007980', '008040', '008060', '00806K', '008110', '008250', '008260', '008350', '008355', '008420', '008490', '008500', '008600', '008700', '008730', '008770', '008775', '008870', '008930', '008970', '009070', '009140', '009150', '009155', '009160', '009180', '009190', '009200', '009240', '009270', '009290', '009310', '009320', '009410', '009415', '009420', '009440', '009450', '009460', '009470', '009540', '009580', '009680', '009770', '009810', '009830', '009835', '009900', '009970']"
      ],
      "metadata": {
        "id": "S6YapGSspIGk"
      },
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "params"
      ],
      "metadata": {
        "id": "aURTb_RpGA15"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "bayesian_params = {\n",
        "    'num_leaves' : (100,200) , 'min_child_samples': (100,200),\n",
        "    'min_child_weight':(1,50),'sub_sample':(0.5,1.0),'colsample_bytree':(0.5,1.0),\n",
        "    'max_bin':(10,500), 'reg_lambda':(0.001,10),'reg_alpha':(0.001,1)\n",
        "}"
      ],
      "metadata": {
        "id": "jOU1XrZSFg6o"
      },
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "LGBM"
      ],
      "metadata": {
        "id": "VG0IhIXEpRbg"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "early stop , verbose"
      ],
      "metadata": {
        "id": "N0gURgbr9mqA"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "X_train, X_test, y_train, y_test = train_test_split(df_data, df_target, test_size=0.2, random_state=42)\n",
        "\n",
        "def lgb(num_leaves,min_child_samples,min_child_weight,sub_sample,colsample_bytree,\n",
        "        max_bin,reg_lambda,reg_alpha):\n",
        "\n",
        "  params = {\n",
        "    \"n_estimators\":1000,'learning_rate':0.01,'max_depth':-1,\n",
        "    'num_leaves':int(round(num_leaves)),'min_child_samples':int(round(min_child_samples)),\n",
        "    'min_child_weight':int(round(min_child_weight)),'sub_sample':max(min(sub_sample, 1), 0),\n",
        "    'colsample_bytree':max(min(colsample_bytree, 1), 0),'max_bin': max(int(round(max_bin)),10),\n",
        "    'reg_lambda':max(reg_lambda,0),'reg_alpha':max(reg_alpha, 0),\n",
        "    'random_state' : 42,\n",
        "}\n",
        "\n",
        "  # 하이퍼파라미터 변경\n",
        "  lgb_model = lgbm.LGBMClassifier(**params)\n",
        "  lgb_model.fit(X_train, y_train , eval_set = [(X_test, y_test)])\n",
        "\n",
        "  valid_proba = lgb_model.predict_proba(X_test)[:,1]\n",
        "  roc_auc = roc_auc_score(y_test , valid_proba)\n",
        "\n",
        "  return roc_auc"
      ],
      "metadata": {
        "id": "4YDv-AjzpRyH"
      },
      "execution_count": 20,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "bayesian LGBM experiment"
      ],
      "metadata": {
        "id": "ow8ucn7qHiFf"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "lgbm_bayesian = BayesianOptimization(f=lgb, pbounds=bayesian_params)\n",
        "lgbm_bayesian.maximize(init_points=5, n_iter=5)"
      ],
      "metadata": {
        "id": "nWMQpXjNHhZu",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "collapsed": true,
        "outputId": "d71a1bac-f3ae-439a-9da3-d97f439d80c6"
      },
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "|   iter    |  target   | colsam... |  max_bin  | min_ch... | min_ch... | num_le... | reg_alpha | reg_la... | sub_sa... |\n",
            "-------------------------------------------------------------------------------------------------------------------------\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] Number of positive: 94279, number of negative: 91542\n",
            "[LightGBM] [Info] Auto-choosing col-wise multi-threading, the overhead of testing was 0.008162 seconds.\n",
            "You can set `force_col_wise=true` to remove the overhead.\n",
            "[LightGBM] [Info] Total Bins 4290\n",
            "[LightGBM] [Info] Number of data points in the train set: 185821, number of used features: 22\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507365 -> initscore=0.029461\n",
            "[LightGBM] [Info] Start training from score 0.029461\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "| \u001b[39m1        \u001b[39m | \u001b[39m0.5514   \u001b[39m | \u001b[39m0.7875   \u001b[39m | \u001b[39m195.0    \u001b[39m | \u001b[39m146.9    \u001b[39m | \u001b[39m13.04    \u001b[39m | \u001b[39m154.0    \u001b[39m | \u001b[39m0.6951   \u001b[39m | \u001b[39m7.922    \u001b[39m | \u001b[39m0.8449   \u001b[39m |\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] Number of positive: 94279, number of negative: 91542\n",
            "[LightGBM] [Info] Auto-choosing col-wise multi-threading, the overhead of testing was 0.008213 seconds.\n",
            "You can set `force_col_wise=true` to remove the overhead.\n",
            "[LightGBM] [Info] Total Bins 7942\n",
            "[LightGBM] [Info] Number of data points in the train set: 185821, number of used features: 22\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507365 -> initscore=0.029461\n",
            "[LightGBM] [Info] Start training from score 0.029461\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "| \u001b[35m2        \u001b[39m | \u001b[35m0.5527   \u001b[39m | \u001b[35m0.7831   \u001b[39m | \u001b[35m361.3    \u001b[39m | \u001b[35m184.2    \u001b[39m | \u001b[35m10.15    \u001b[39m | \u001b[35m137.6    \u001b[39m | \u001b[35m0.7241   \u001b[39m | \u001b[35m9.693    \u001b[39m | \u001b[35m0.549    \u001b[39m |\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] Number of positive: 94279, number of negative: 91542\n",
            "[LightGBM] [Info] Auto-choosing col-wise multi-threading, the overhead of testing was 0.007665 seconds.\n",
            "You can set `force_col_wise=true` to remove the overhead.\n",
            "[LightGBM] [Info] Total Bins 7348\n",
            "[LightGBM] [Info] Number of data points in the train set: 185821, number of used features: 22\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507365 -> initscore=0.029461\n",
            "[LightGBM] [Info] Start training from score 0.029461\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "| \u001b[39m3        \u001b[39m | \u001b[39m0.5519   \u001b[39m | \u001b[39m0.5847   \u001b[39m | \u001b[39m333.8    \u001b[39m | \u001b[39m144.1    \u001b[39m | \u001b[39m16.84    \u001b[39m | \u001b[39m195.0    \u001b[39m | \u001b[39m0.3936   \u001b[39m | \u001b[39m3.372    \u001b[39m | \u001b[39m0.5851   \u001b[39m |\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] Number of positive: 94279, number of negative: 91542\n",
            "[LightGBM] [Info] Auto-choosing col-wise multi-threading, the overhead of testing was 0.008538 seconds.\n",
            "You can set `force_col_wise=true` to remove the overhead.\n",
            "[LightGBM] [Info] Total Bins 3520\n",
            "[LightGBM] [Info] Number of data points in the train set: 185821, number of used features: 22\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507365 -> initscore=0.029461\n",
            "[LightGBM] [Info] Start training from score 0.029461\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "| \u001b[39m4        \u001b[39m | \u001b[39m0.5521   \u001b[39m | \u001b[39m0.8985   \u001b[39m | \u001b[39m160.3    \u001b[39m | \u001b[39m147.6    \u001b[39m | \u001b[39m20.54    \u001b[39m | \u001b[39m158.6    \u001b[39m | \u001b[39m0.9257   \u001b[39m | \u001b[39m7.987    \u001b[39m | \u001b[39m0.9547   \u001b[39m |\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] Number of positive: 94279, number of negative: 91542\n",
            "[LightGBM] [Info] Auto-choosing col-wise multi-threading, the overhead of testing was 0.011242 seconds.\n",
            "You can set `force_col_wise=true` to remove the overhead.\n",
            "[LightGBM] [Info] Total Bins 2882\n",
            "[LightGBM] [Info] Number of data points in the train set: 185821, number of used features: 22\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507365 -> initscore=0.029461\n",
            "[LightGBM] [Info] Start training from score 0.029461\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "| \u001b[39m5        \u001b[39m | \u001b[39m0.5497   \u001b[39m | \u001b[39m0.8866   \u001b[39m | \u001b[39m131.2    \u001b[39m | \u001b[39m183.6    \u001b[39m | \u001b[39m35.58    \u001b[39m | \u001b[39m105.3    \u001b[39m | \u001b[39m0.8053   \u001b[39m | \u001b[39m5.016    \u001b[39m | \u001b[39m0.9958   \u001b[39m |\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] Number of positive: 94279, number of negative: 91542\n",
            "[LightGBM] [Info] Auto-choosing col-wise multi-threading, the overhead of testing was 0.008096 seconds.\n",
            "You can set `force_col_wise=true` to remove the overhead.\n",
            "[LightGBM] [Info] Total Bins 8008\n",
            "[LightGBM] [Info] Number of data points in the train set: 185821, number of used features: 22\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507365 -> initscore=0.029461\n",
            "[LightGBM] [Info] Start training from score 0.029461\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "| \u001b[39m6        \u001b[39m | \u001b[39m0.5515   \u001b[39m | \u001b[39m0.7106   \u001b[39m | \u001b[39m363.7    \u001b[39m | \u001b[39m182.5    \u001b[39m | \u001b[39m5.525    \u001b[39m | \u001b[39m131.0    \u001b[39m | \u001b[39m0.6602   \u001b[39m | \u001b[39m7.056    \u001b[39m | \u001b[39m0.8205   \u001b[39m |\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] Number of positive: 94279, number of negative: 91542\n",
            "[LightGBM] [Info] Auto-choosing col-wise multi-threading, the overhead of testing was 0.008108 seconds.\n",
            "You can set `force_col_wise=true` to remove the overhead.\n",
            "[LightGBM] [Info] Total Bins 4884\n",
            "[LightGBM] [Info] Number of data points in the train set: 185821, number of used features: 22\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507365 -> initscore=0.029461\n",
            "[LightGBM] [Info] Start training from score 0.029461\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "| \u001b[39m7        \u001b[39m | \u001b[39m0.5522   \u001b[39m | \u001b[39m0.7324   \u001b[39m | \u001b[39m221.6    \u001b[39m | \u001b[39m131.8    \u001b[39m | \u001b[39m29.07    \u001b[39m | \u001b[39m164.8    \u001b[39m | \u001b[39m0.2087   \u001b[39m | \u001b[39m3.583    \u001b[39m | \u001b[39m0.9754   \u001b[39m |\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] Number of positive: 94279, number of negative: 91542\n",
            "[LightGBM] [Info] Auto-choosing col-wise multi-threading, the overhead of testing was 0.007642 seconds.\n",
            "You can set `force_col_wise=true` to remove the overhead.\n",
            "[LightGBM] [Info] Total Bins 10164\n",
            "[LightGBM] [Info] Number of data points in the train set: 185821, number of used features: 22\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507365 -> initscore=0.029461\n",
            "[LightGBM] [Info] Start training from score 0.029461\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "| \u001b[39m8        \u001b[39m | \u001b[39m0.5502   \u001b[39m | \u001b[39m0.5978   \u001b[39m | \u001b[39m461.5    \u001b[39m | \u001b[39m105.1    \u001b[39m | \u001b[39m36.78    \u001b[39m | \u001b[39m110.7    \u001b[39m | \u001b[39m0.7425   \u001b[39m | \u001b[39m6.177    \u001b[39m | \u001b[39m0.6833   \u001b[39m |\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] Number of positive: 94279, number of negative: 91542\n",
            "[LightGBM] [Info] Auto-choosing col-wise multi-threading, the overhead of testing was 0.008211 seconds.\n",
            "You can set `force_col_wise=true` to remove the overhead.\n",
            "[LightGBM] [Info] Total Bins 7898\n",
            "[LightGBM] [Info] Number of data points in the train set: 185821, number of used features: 22\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507365 -> initscore=0.029461\n",
            "[LightGBM] [Info] Start training from score 0.029461\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "| \u001b[39m9        \u001b[39m | \u001b[39m0.5502   \u001b[39m | \u001b[39m0.9543   \u001b[39m | \u001b[39m358.9    \u001b[39m | \u001b[39m186.2    \u001b[39m | \u001b[39m15.37    \u001b[39m | \u001b[39m143.3    \u001b[39m | \u001b[39m0.1987   \u001b[39m | \u001b[39m2.694    \u001b[39m | \u001b[39m0.555    \u001b[39m |\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] Number of positive: 94279, number of negative: 91542\n",
            "[LightGBM] [Info] Auto-choosing col-wise multi-threading, the overhead of testing was 0.008130 seconds.\n",
            "You can set `force_col_wise=true` to remove the overhead.\n",
            "[LightGBM] [Info] Total Bins 4972\n",
            "[LightGBM] [Info] Number of data points in the train set: 185821, number of used features: 22\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507365 -> initscore=0.029461\n",
            "[LightGBM] [Info] Start training from score 0.029461\n",
            "[LightGBM] [Warning] Unknown parameter: sub_sample\n",
            "| \u001b[39m10       \u001b[39m | \u001b[39m0.5504   \u001b[39m | \u001b[39m0.6667   \u001b[39m | \u001b[39m226.5    \u001b[39m | \u001b[39m130.9    \u001b[39m | \u001b[39m24.63    \u001b[39m | \u001b[39m160.6    \u001b[39m | \u001b[39m0.3789   \u001b[39m | \u001b[39m5.377    \u001b[39m | \u001b[39m0.8656   \u001b[39m |\n",
            "=========================================================================================================================\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(lgbm_bayesian.max)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qsnIunzu1nHp",
        "outputId": "dc578fd2-c9a4-4987-e5ae-9e5c2f709eab"
      },
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "{'target': 0.5526661311522713, 'params': {'colsample_bytree': 0.7831466414605907, 'max_bin': 361.29744148260824, 'min_child_samples': 184.23254188176085, 'min_child_weight': 10.151206630547922, 'num_leaves': 137.6089493616313, 'reg_alpha': 0.7241178770980159, 'reg_lambda': 9.693277696369794, 'sub_sample': 0.548974484268904}}\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "0.5526661311522713"
      ],
      "metadata": {
        "id": "BLljHbj38okj"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "tuned lgbm fit"
      ],
      "metadata": {
        "id": "xr7BSHvWH-8x"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "lgb_model = lgbm.LGBMClassifier(\n",
        "    n_estimators=500,\n",
        "    learning_rate=0.02,\n",
        "    max_depth=14,\n",
        "    num_leaves=57,\n",
        "    reg_alpha=0.5427965,\n",
        "    reg_lambda=5.489,\n",
        "    min_child_samples=109,\n",
        "    min_child_weight=15,\n",
        "    colsample_bytree=0.571133,\n",
        "    max_bin=333\n",
        ")\n",
        "\n",
        "lgb_model.fit(X_train,y_train,eval_set=[(X_test,y_test)])"
      ],
      "metadata": {
        "id": "-HW9MY-AH1W2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 271
        },
        "outputId": "0b01056c-ff28-48c7-ee54-3db54d7f61f0"
      },
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[LightGBM] [Info] Number of positive: 94279, number of negative: 91542\n",
            "[LightGBM] [Info] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001461 seconds.\n",
            "You can set `force_row_wise=true` to remove the overhead.\n",
            "And if memory is not enough, you can set `force_col_wise=true`.\n",
            "[LightGBM] [Info] Total Bins 7326\n",
            "[LightGBM] [Info] Number of data points in the train set: 185821, number of used features: 22\n",
            "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507365 -> initscore=0.029461\n",
            "[LightGBM] [Info] Start training from score 0.029461\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LGBMClassifier(colsample_bytree=0.571133, learning_rate=0.02, max_bin=333,\n",
              "               max_depth=14, min_child_samples=109, min_child_weight=15,\n",
              "               n_estimators=500, num_leaves=57, reg_alpha=0.5427965,\n",
              "               reg_lambda=5.489)"
            ],
            "text/html": [
              "<style>#sk-container-id-1 {\n",
              "  /* Definition of color scheme common for light and dark mode */\n",
              "  --sklearn-color-text: black;\n",
              "  --sklearn-color-line: gray;\n",
              "  /* Definition of color scheme for unfitted estimators */\n",
              "  --sklearn-color-unfitted-level-0: #fff5e6;\n",
              "  --sklearn-color-unfitted-level-1: #f6e4d2;\n",
              "  --sklearn-color-unfitted-level-2: #ffe0b3;\n",
              "  --sklearn-color-unfitted-level-3: chocolate;\n",
              "  /* Definition of color scheme for fitted estimators */\n",
              "  --sklearn-color-fitted-level-0: #f0f8ff;\n",
              "  --sklearn-color-fitted-level-1: #d4ebff;\n",
              "  --sklearn-color-fitted-level-2: #b3dbfd;\n",
              "  --sklearn-color-fitted-level-3: cornflowerblue;\n",
              "\n",
              "  /* Specific color for light theme */\n",
              "  --sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));\n",
              "  --sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, white)));\n",
              "  --sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));\n",
              "  --sklearn-color-icon: #696969;\n",
              "\n",
              "  @media (prefers-color-scheme: dark) {\n",
              "    /* Redefinition of color scheme for dark theme */\n",
              "    --sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));\n",
              "    --sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, #111)));\n",
              "    --sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));\n",
              "    --sklearn-color-icon: #878787;\n",
              "  }\n",
              "}\n",
              "\n",
              "#sk-container-id-1 {\n",
              "  color: var(--sklearn-color-text);\n",
              "}\n",
              "\n",
              "#sk-container-id-1 pre {\n",
              "  padding: 0;\n",
              "}\n",
              "\n",
              "#sk-container-id-1 input.sk-hidden--visually {\n",
              "  border: 0;\n",
              "  clip: rect(1px 1px 1px 1px);\n",
              "  clip: rect(1px, 1px, 1px, 1px);\n",
              "  height: 1px;\n",
              "  margin: -1px;\n",
              "  overflow: hidden;\n",
              "  padding: 0;\n",
              "  position: absolute;\n",
              "  width: 1px;\n",
              "}\n",
              "\n",
              "#sk-container-id-1 div.sk-dashed-wrapped {\n",
              "  border: 1px dashed var(--sklearn-color-line);\n",
              "  margin: 0 0.4em 0.5em 0.4em;\n",
              "  box-sizing: border-box;\n",
              "  padding-bottom: 0.4em;\n",
              "  background-color: var(--sklearn-color-background);\n",
              "}\n",
              "\n",
              "#sk-container-id-1 div.sk-container {\n",
              "  /* jupyter's `normalize.less` sets `[hidden] { display: none; }`\n",
              "     but bootstrap.min.css set `[hidden] { display: none !important; }`\n",
              "     so we also need the `!important` here to be able to override the\n",
              "     default hidden behavior on the sphinx rendered scikit-learn.org.\n",
              "     See: https://github.com/scikit-learn/scikit-learn/issues/21755 */\n",
              "  display: inline-block !important;\n",
              "  position: relative;\n",
              "}\n",
              "\n",
              "#sk-container-id-1 div.sk-text-repr-fallback {\n",
              "  display: none;\n",
              "}\n",
              "\n",
              "div.sk-parallel-item,\n",
              "div.sk-serial,\n",
              "div.sk-item {\n",
              "  /* draw centered vertical line to link estimators */\n",
              "  background-image: linear-gradient(var(--sklearn-color-text-on-default-background), var(--sklearn-color-text-on-default-background));\n",
              "  background-size: 2px 100%;\n",
              "  background-repeat: no-repeat;\n",
              "  background-position: center center;\n",
              "}\n",
              "\n",
              "/* Parallel-specific style estimator block */\n",
              "\n",
              "#sk-container-id-1 div.sk-parallel-item::after {\n",
              "  content: \"\";\n",
              "  width: 100%;\n",
              "  border-bottom: 2px solid var(--sklearn-color-text-on-default-background);\n",
              "  flex-grow: 1;\n",
              "}\n",
              "\n",
              "#sk-container-id-1 div.sk-parallel {\n",
              "  display: flex;\n",
              "  align-items: stretch;\n",
              "  justify-content: center;\n",
              "  background-color: var(--sklearn-color-background);\n",
              "  position: relative;\n",
              "}\n",
              "\n",
              "#sk-container-id-1 div.sk-parallel-item {\n",
              "  display: flex;\n",
              "  flex-direction: column;\n",
              "}\n",
              "\n",
              "#sk-container-id-1 div.sk-parallel-item:first-child::after {\n",
              "  align-self: flex-end;\n",
              "  width: 50%;\n",
              "}\n",
              "\n",
              "#sk-container-id-1 div.sk-parallel-item:last-child::after {\n",
              "  align-self: flex-start;\n",
              "  width: 50%;\n",
              "}\n",
              "\n",
              "#sk-container-id-1 div.sk-parallel-item:only-child::after {\n",
              "  width: 0;\n",
              "}\n",
              "\n",
              "/* Serial-specific style estimator block */\n",
              "\n",
              "#sk-container-id-1 div.sk-serial {\n",
              "  display: flex;\n",
              "  flex-direction: column;\n",
              "  align-items: center;\n",
              "  background-color: var(--sklearn-color-background);\n",
              "  padding-right: 1em;\n",
              "  padding-left: 1em;\n",
              "}\n",
              "\n",
              "\n",
              "/* Toggleable style: style used for estimator/Pipeline/ColumnTransformer box that is\n",
              "clickable and can be expanded/collapsed.\n",
              "- Pipeline and ColumnTransformer use this feature and define the default style\n",
              "- Estimators will overwrite some part of the style using the `sk-estimator` class\n",
              "*/\n",
              "\n",
              "/* Pipeline and ColumnTransformer style (default) */\n",
              "\n",
              "#sk-container-id-1 div.sk-toggleable {\n",
              "  /* Default theme specific background. It is overwritten whether we have a\n",
              "  specific estimator or a Pipeline/ColumnTransformer */\n",
              "  background-color: var(--sklearn-color-background);\n",
              "}\n",
              "\n",
              "/* Toggleable label */\n",
              "#sk-container-id-1 label.sk-toggleable__label {\n",
              "  cursor: pointer;\n",
              "  display: block;\n",
              "  width: 100%;\n",
              "  margin-bottom: 0;\n",
              "  padding: 0.5em;\n",
              "  box-sizing: border-box;\n",
              "  text-align: center;\n",
              "}\n",
              "\n",
              "#sk-container-id-1 label.sk-toggleable__label-arrow:before {\n",
              "  /* Arrow on the left of the label */\n",
              "  content: \"▸\";\n",
              "  float: left;\n",
              "  margin-right: 0.25em;\n",
              "  color: var(--sklearn-color-icon);\n",
              "}\n",
              "\n",
              "#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {\n",
              "  color: var(--sklearn-color-text);\n",
              "}\n",
              "\n",
              "/* Toggleable content - dropdown */\n",
              "\n",
              "#sk-container-id-1 div.sk-toggleable__content {\n",
              "  max-height: 0;\n",
              "  max-width: 0;\n",
              "  overflow: hidden;\n",
              "  text-align: left;\n",
              "  /* unfitted */\n",
              "  background-color: var(--sklearn-color-unfitted-level-0);\n",
              "}\n",
              "\n",
              "#sk-container-id-1 div.sk-toggleable__content.fitted {\n",
              "  /* fitted */\n",
              "  background-color: var(--sklearn-color-fitted-level-0);\n",
              "}\n",
              "\n",
              "#sk-container-id-1 div.sk-toggleable__content pre {\n",
              "  margin: 0.2em;\n",
              "  border-radius: 0.25em;\n",
              "  color: var(--sklearn-color-text);\n",
              "  /* unfitted */\n",
              "  background-color: var(--sklearn-color-unfitted-level-0);\n",
              "}\n",
              "\n",
              "#sk-container-id-1 div.sk-toggleable__content.fitted pre {\n",
              "  /* unfitted */\n",
              "  background-color: var(--sklearn-color-fitted-level-0);\n",
              "}\n",
              "\n",
              "#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {\n",
              "  /* Expand drop-down */\n",
              "  max-height: 200px;\n",
              "  max-width: 100%;\n",
              "  overflow: auto;\n",
              "}\n",
              "\n",
              "#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {\n",
              "  content: \"▾\";\n",
              "}\n",
              "\n",
              "/* Pipeline/ColumnTransformer-specific style */\n",
              "\n",
              "#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
              "  color: var(--sklearn-color-text);\n",
              "  background-color: var(--sklearn-color-unfitted-level-2);\n",
              "}\n",
              "\n",
              "#sk-container-id-1 div.sk-label.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
              "  background-color: var(--sklearn-color-fitted-level-2);\n",
              "}\n",
              "\n",
              "/* Estimator-specific style */\n",
              "\n",
              "/* Colorize estimator box */\n",
              "#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
              "  /* unfitted */\n",
              "  background-color: var(--sklearn-color-unfitted-level-2);\n",
              "}\n",
              "\n",
              "#sk-container-id-1 div.sk-estimator.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
              "  /* fitted */\n",
              "  background-color: var(--sklearn-color-fitted-level-2);\n",
              "}\n",
              "\n",
              "#sk-container-id-1 div.sk-label label.sk-toggleable__label,\n",
              "#sk-container-id-1 div.sk-label label {\n",
              "  /* The background is the default theme color */\n",
              "  color: var(--sklearn-color-text-on-default-background);\n",
              "}\n",
              "\n",
              "/* On hover, darken the color of the background */\n",
              "#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {\n",
              "  color: var(--sklearn-color-text);\n",
              "  background-color: var(--sklearn-color-unfitted-level-2);\n",
              "}\n",
              "\n",
              "/* Label box, darken color on hover, fitted */\n",
              "#sk-container-id-1 div.sk-label.fitted:hover label.sk-toggleable__label.fitted {\n",
              "  color: var(--sklearn-color-text);\n",
              "  background-color: var(--sklearn-color-fitted-level-2);\n",
              "}\n",
              "\n",
              "/* Estimator label */\n",
              "\n",
              "#sk-container-id-1 div.sk-label label {\n",
              "  font-family: monospace;\n",
              "  font-weight: bold;\n",
              "  display: inline-block;\n",
              "  line-height: 1.2em;\n",
              "}\n",
              "\n",
              "#sk-container-id-1 div.sk-label-container {\n",
              "  text-align: center;\n",
              "}\n",
              "\n",
              "/* Estimator-specific */\n",
              "#sk-container-id-1 div.sk-estimator {\n",
              "  font-family: monospace;\n",
              "  border: 1px dotted var(--sklearn-color-border-box);\n",
              "  border-radius: 0.25em;\n",
              "  box-sizing: border-box;\n",
              "  margin-bottom: 0.5em;\n",
              "  /* unfitted */\n",
              "  background-color: var(--sklearn-color-unfitted-level-0);\n",
              "}\n",
              "\n",
              "#sk-container-id-1 div.sk-estimator.fitted {\n",
              "  /* fitted */\n",
              "  background-color: var(--sklearn-color-fitted-level-0);\n",
              "}\n",
              "\n",
              "/* on hover */\n",
              "#sk-container-id-1 div.sk-estimator:hover {\n",
              "  /* unfitted */\n",
              "  background-color: var(--sklearn-color-unfitted-level-2);\n",
              "}\n",
              "\n",
              "#sk-container-id-1 div.sk-estimator.fitted:hover {\n",
              "  /* fitted */\n",
              "  background-color: var(--sklearn-color-fitted-level-2);\n",
              "}\n",
              "\n",
              "/* Specification for estimator info (e.g. \"i\" and \"?\") */\n",
              "\n",
              "/* Common style for \"i\" and \"?\" */\n",
              "\n",
              ".sk-estimator-doc-link,\n",
              "a:link.sk-estimator-doc-link,\n",
              "a:visited.sk-estimator-doc-link {\n",
              "  float: right;\n",
              "  font-size: smaller;\n",
              "  line-height: 1em;\n",
              "  font-family: monospace;\n",
              "  background-color: var(--sklearn-color-background);\n",
              "  border-radius: 1em;\n",
              "  height: 1em;\n",
              "  width: 1em;\n",
              "  text-decoration: none !important;\n",
              "  margin-left: 1ex;\n",
              "  /* unfitted */\n",
              "  border: var(--sklearn-color-unfitted-level-1) 1pt solid;\n",
              "  color: var(--sklearn-color-unfitted-level-1);\n",
              "}\n",
              "\n",
              ".sk-estimator-doc-link.fitted,\n",
              "a:link.sk-estimator-doc-link.fitted,\n",
              "a:visited.sk-estimator-doc-link.fitted {\n",
              "  /* fitted */\n",
              "  border: var(--sklearn-color-fitted-level-1) 1pt solid;\n",
              "  color: var(--sklearn-color-fitted-level-1);\n",
              "}\n",
              "\n",
              "/* On hover */\n",
              "div.sk-estimator:hover .sk-estimator-doc-link:hover,\n",
              ".sk-estimator-doc-link:hover,\n",
              "div.sk-label-container:hover .sk-estimator-doc-link:hover,\n",
              ".sk-estimator-doc-link:hover {\n",
              "  /* unfitted */\n",
              "  background-color: var(--sklearn-color-unfitted-level-3);\n",
              "  color: var(--sklearn-color-background);\n",
              "  text-decoration: none;\n",
              "}\n",
              "\n",
              "div.sk-estimator.fitted:hover .sk-estimator-doc-link.fitted:hover,\n",
              ".sk-estimator-doc-link.fitted:hover,\n",
              "div.sk-label-container:hover .sk-estimator-doc-link.fitted:hover,\n",
              ".sk-estimator-doc-link.fitted:hover {\n",
              "  /* fitted */\n",
              "  background-color: var(--sklearn-color-fitted-level-3);\n",
              "  color: var(--sklearn-color-background);\n",
              "  text-decoration: none;\n",
              "}\n",
              "\n",
              "/* Span, style for the box shown on hovering the info icon */\n",
              ".sk-estimator-doc-link span {\n",
              "  display: none;\n",
              "  z-index: 9999;\n",
              "  position: relative;\n",
              "  font-weight: normal;\n",
              "  right: .2ex;\n",
              "  padding: .5ex;\n",
              "  margin: .5ex;\n",
              "  width: min-content;\n",
              "  min-width: 20ex;\n",
              "  max-width: 50ex;\n",
              "  color: var(--sklearn-color-text);\n",
              "  box-shadow: 2pt 2pt 4pt #999;\n",
              "  /* unfitted */\n",
              "  background: var(--sklearn-color-unfitted-level-0);\n",
              "  border: .5pt solid var(--sklearn-color-unfitted-level-3);\n",
              "}\n",
              "\n",
              ".sk-estimator-doc-link.fitted span {\n",
              "  /* fitted */\n",
              "  background: var(--sklearn-color-fitted-level-0);\n",
              "  border: var(--sklearn-color-fitted-level-3);\n",
              "}\n",
              "\n",
              ".sk-estimator-doc-link:hover span {\n",
              "  display: block;\n",
              "}\n",
              "\n",
              "/* \"?\"-specific style due to the `<a>` HTML tag */\n",
              "\n",
              "#sk-container-id-1 a.estimator_doc_link {\n",
              "  float: right;\n",
              "  font-size: 1rem;\n",
              "  line-height: 1em;\n",
              "  font-family: monospace;\n",
              "  background-color: var(--sklearn-color-background);\n",
              "  border-radius: 1rem;\n",
              "  height: 1rem;\n",
              "  width: 1rem;\n",
              "  text-decoration: none;\n",
              "  /* unfitted */\n",
              "  color: var(--sklearn-color-unfitted-level-1);\n",
              "  border: var(--sklearn-color-unfitted-level-1) 1pt solid;\n",
              "}\n",
              "\n",
              "#sk-container-id-1 a.estimator_doc_link.fitted {\n",
              "  /* fitted */\n",
              "  border: var(--sklearn-color-fitted-level-1) 1pt solid;\n",
              "  color: var(--sklearn-color-fitted-level-1);\n",
              "}\n",
              "\n",
              "/* On hover */\n",
              "#sk-container-id-1 a.estimator_doc_link:hover {\n",
              "  /* unfitted */\n",
              "  background-color: var(--sklearn-color-unfitted-level-3);\n",
              "  color: var(--sklearn-color-background);\n",
              "  text-decoration: none;\n",
              "}\n",
              "\n",
              "#sk-container-id-1 a.estimator_doc_link.fitted:hover {\n",
              "  /* fitted */\n",
              "  background-color: var(--sklearn-color-fitted-level-3);\n",
              "}\n",
              "</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>LGBMClassifier(colsample_bytree=0.571133, learning_rate=0.02, max_bin=333,\n",
              "               max_depth=14, min_child_samples=109, min_child_weight=15,\n",
              "               n_estimators=500, num_leaves=57, reg_alpha=0.5427965,\n",
              "               reg_lambda=5.489)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator fitted sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" checked><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label fitted sk-toggleable__label-arrow fitted\">&nbsp;LGBMClassifier<span class=\"sk-estimator-doc-link fitted\">i<span>Fitted</span></span></label><div class=\"sk-toggleable__content fitted\"><pre>LGBMClassifier(colsample_bytree=0.571133, learning_rate=0.02, max_bin=333,\n",
              "               max_depth=14, min_child_samples=109, min_child_weight=15,\n",
              "               n_estimators=500, num_leaves=57, reg_alpha=0.5427965,\n",
              "               reg_lambda=5.489)</pre></div> </div></div></div></div>"
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "feature importance visualization"
      ],
      "metadata": {
        "id": "l56MC3zHpT-i"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "feature_importance(lgb_model)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "7LqUFnNopVIV",
        "outputId": "fb1d1e8c-3baf-4dcf-8984-b50932ac0f61"
      },
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "heat map visualization"
      ],
      "metadata": {
        "id": "qU7iRUJBpV2p"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "pred = lgb_model.predict(X_test)\n",
        "heatmap(X_test , pred)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 690
        },
        "id": "QILUshLGpW3H",
        "outputId": "7f1886ec-b17c-4368-ff07-79687f52709a"
      },
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x600 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "inference"
      ],
      "metadata": {
        "id": "jvBaklp6-wNf"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "answer_prob = sum((pred - y_test)**2)/len(pred)\n",
        "answer_prob"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bNG1ToOY-vVm",
        "outputId": "fee42798-df84-485c-d206-8845430e6828"
      },
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1.8643017048389874"
            ]
          },
          "metadata": {},
          "execution_count": 27
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Back testing"
      ],
      "metadata": {
        "id": "NTcUVD1cpilj"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "class DeepLearningStrategy(strategy.BacktestingStrategy):\n",
        "    def __init__(self, feed, instrument, model):\n",
        "        super(DeepLearningStrategy, self).__init__(feed)\n",
        "        self.__instrument = instrument\n",
        "        self.__position = None\n",
        "        self.__model = model\n",
        "        self.__prices = []  # 종가 데이터를 저장할 리스트\n",
        "        self.__opens=[]\n",
        "        self.__volumes = []  # 거래량 데이터를 저장할 리스트\n",
        "        self.__buy_date = None  # 매수 날짜 저장 변수\n",
        "        self.__target_price = None  # 목표가 저장 변수\n",
        "        self.feed=feed\n",
        "\n",
        "    def onBars(self, bars):\n",
        "        bar = bars[self.__instrument]\n",
        "        self.__prices.append(bar.getClose())   # 종가 데이터를 저장\n",
        "        self.__opens.append(bar.getOpen())\n",
        "        current_date = bar.getDateTime()\n",
        "\n",
        "        # 데이터가 있는지 확인\n",
        "        date_times = self.feed[self.__instrument].getDateTimes()\n",
        "        if len(date_times) > 0:\n",
        "            date_len = len(date_times)-1\n",
        "            self.__last_date = date_times[date_len]  # 마지막 날짜 저장\n",
        "        else:\n",
        "            self.__last_date = None  # 데이터가 없을 경우 None으로 설정\n",
        "\n",
        "        # 최소 50일치 데이터가 있어야 예측 가능\n",
        "        if len(self.__prices) < 13:\n",
        "            return\n",
        "\n",
        "        # 마지막 20일의 거래량 데이터 가져오기\n",
        "        price = []\n",
        "        for i in range(-1, -11, -1):  # 마지막 20일간의 데이터\n",
        "            price_t = self.__prices[i]\n",
        "            price.append(price_t)\n",
        "        open = []\n",
        "        for i in range(-1, -11, -1):  # 마지막 20일간의 데이터\n",
        "            open_t = self.__opens[i]\n",
        "            open.append(open_t)\n",
        "\n",
        "        price=torch.Tensor(price).unsqueeze(1)\n",
        "        price=price.reshape(1,-1)\n",
        "        open=torch.Tensor(open).unsqueeze(1)\n",
        "        open=open.reshape(1,-1)\n",
        "\n",
        "        input_0 = (price[:,1] / price[:,2]).unsqueeze(1)\n",
        "        input_1 = (price[:,0] * price[:,1] * price[:,2]).unsqueeze(1)\n",
        "\n",
        "        x = torch.cat([price,open,input_0,input_1],axis=1)\n",
        "\n",
        "        x = (x - x.min())/(x.max() - x.min())\n",
        "\n",
        "        # 딥러닝 모델로 다음 날 가격 예측\n",
        "        #self.__model.eval()\n",
        "        prediction = self.__model.predict(x)\n",
        "        #predicted_classes = int(100*prediction[0])\n",
        "        #predicted_classes = torch.argmax(prediction, dim=1)\n",
        "\n",
        "        # 간단한 매수/매도 로직\n",
        "        if self.__position is None:\n",
        "            if prediction >= 1:  # 예측된 가격이 현재 가격보다 크면 매수\n",
        "                available_cash = self.getBroker().getCash()  # 현재 자산(현금)\n",
        "                price_per_share = bar.getClose()  # 현재 주가\n",
        "                current_date = bar.getDateTime()\n",
        "\n",
        "                # 구매 가능한 최대 주식 수 계산 (수수료 및 추가 비용을 15%로 가정)\n",
        "                max_shares = int(available_cash / (price_per_share * 1.01))\n",
        "\n",
        "                if max_shares > 0:\n",
        "                    self.__position = self.enterLong(self.__instrument, max_shares)\n",
        "                    self.__buy_date = bar.getDateTime()  # 매수 날짜를 저장\n",
        "                    self.info(f\"매수 메시지: {bar.getDateTime()}에 {bar.getPrice()}에 매수함\")\n",
        "                    self.__target_price = bar.getClose() * 1.15  # 목표가를 설정 (4% 상승)\n",
        "                    self.__current_price = bar.getClose()\n",
        "        elif self.__position is not None and not self.__position.exitActive():\n",
        "          high_price_today = bar.getHigh()  # 오늘의 최고가\n",
        "          low_price_today = bar.getLow()\n",
        "          current_date = bar.getDateTime()\n",
        "\n",
        "          if (current_date - self.__buy_date).days > 30 and high_price_today >= self.__current_price * 1.5:\n",
        "            self.info(f\"5일 경과 후 강제 매도: {current_date}에 {bar.getPrice()}에 매도함\")\n",
        "            self.__position.exitMarket()\n",
        "            self.__position = None  # 매도 후 포지션 초기화\n",
        "          elif (current_date - self.__buy_date).days > 400:\n",
        "            self.info(f\"5일 경과 후 강제 매도: {current_date}에 {bar.getPrice()}에 매도함\")\n",
        "            self.__position.exitMarket()\n",
        "            self.__position = None  # 매도 후 포지션 초기화"
      ],
      "metadata": {
        "id": "0JwQzQSWpi7Z"
      },
      "execution_count": 62,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def backtesting(df):\n",
        "  df['Date'] = pd.to_datetime(df['Date'], errors='coerce')\n",
        "\n",
        "  df.replace([np.inf], 300000, inplace=True)\n",
        "  df.replace([-np.inf], -300000, inplace=True)\n",
        "\n",
        "  filtered_df = df[\n",
        "    (df['Low'] <= df['Open']) &   # Low가 Open보다 크면 안 됨\n",
        "    (df['Low'] <= df['Close']) &  # Low가 Close보다 크면 안 됨\n",
        "    (df['Low'] <= df['High']) &   # Low가 High보다 크면 안 됨\n",
        "    (df['High'] >= df['Open']) &  # High는 Open보다 작으면 안 됨\n",
        "    (df['High'] >= df['Close']) & # High는 Close보다 작으면 안 됨\n",
        "    (df['Volume'] > 0)           # 거래량은 음수일 수 없음\n",
        "  ]\n",
        "\n",
        "  #  변환된 DataFrame을 다시 CSV 파일로 저장\n",
        "  filtered_df.to_csv('AAPL_fixed.csv', index=False)\n",
        "\n",
        "  # Yahoo 피드 준비\n",
        "  feed = yahoofeed.Feed()\n",
        "  feed.addBarsFromCSV(\"aapl\", \"AAPL_fixed.csv\")\n",
        "\n",
        "  # 전략 실행 (딥러닝 모델을 통합)\n",
        "  start = DeepLearningStrategy(feed, \"aapl\", lgb_model)\n",
        "\n",
        "  # 수익률 분석기 추가\n",
        "  returnsAnalyzer = returns.Returns()\n",
        "  start.attachAnalyzer(returnsAnalyzer)\n",
        "\n",
        "  # 트레이드 분석기 추가 (거래 내역 및 수익 확인)\n",
        "  tradesAnalyzer = trades.Trades()\n",
        "  start.attachAnalyzer(tradesAnalyzer)\n",
        "\n",
        "  # 백테스트 실행\n",
        "  initial_cash = start.getBroker().getCash()\n",
        "  logging.getLogger(\"broker.backtesting\").setLevel(logging.CRITICAL)\n",
        "  start.run()\n",
        "  profit = ((tradesAnalyzer.getProfits().sum() + tradesAnalyzer.getLosses().sum())/ initial_cash) * 100\n",
        "\n",
        "  return profit,tradesAnalyzer.getCount()"
      ],
      "metadata": {
        "id": "tt-dVUpJplPA"
      },
      "execution_count": 63,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "sum_profit =0\n",
        "sum_trade_num = 0\n",
        "cnt = 0\n",
        "zero_num = 0\n",
        "\n",
        "for stock in kosdaq:\n",
        "  cnt += 1\n",
        "  df = pd.read_csv('/content/drive/MyDrive/kosdaq_sim/' + stock + '.csv')\n",
        "  df[df.columns[0]] = pd.to_datetime(df.iloc[:, 0].astype(str), format='%Y%m%d')\n",
        "  df.rename(columns={df.columns[0]: 'Date'}, inplace=True)\n",
        "  df['Adj Close']=df['close']\n",
        "  df['Low']=df['low']\n",
        "  df['High']=df['high']\n",
        "  df['Open']=df['open']\n",
        "  df['Close']=df['close']\n",
        "  df['Volume']=df['volume']\n",
        "  df=df.drop(['low','high','open','close','volume'],axis=1)\n",
        "  a1 , a2 = backtesting(df)\n",
        "  if a2 == 0:\n",
        "    zero_num += 1\n",
        "  if a2 != 0:\n",
        "    sum_profit += a1-(a2/2)\n",
        "    sum_trade_num += a2\n",
        "    print(stock, a1-(a2/2), a2)\n",
        "print(sum_profit/(cnt-zero_num), sum_trade_num/(cnt-zero_num) , zero_num)"
      ],
      "metadata": {
        "id": "sePAfLjnpmTe",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 356
        },
        "outputId": "ec92c41a-939f-444e-86e3-45d5841416ec"
      },
      "execution_count": 64,
      "outputs": [
        {
          "output_type": "error",
          "ename": "Exception",
          "evalue": "Feed was empty",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mException\u001b[0m                                 Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-64-06dc73e6624f>\u001b[0m in \u001b[0;36m<cell line: 6>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     16\u001b[0m   \u001b[0mdf\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'Volume'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdf\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'volume'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     17\u001b[0m   \u001b[0mdf\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdrop\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'low'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m'high'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m'open'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m'close'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m'volume'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 18\u001b[0;31m   \u001b[0ma1\u001b[0m \u001b[0;34m,\u001b[0m \u001b[0ma2\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mbacktesting\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdf\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     19\u001b[0m   \u001b[0;32mif\u001b[0m \u001b[0ma2\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     20\u001b[0m     \u001b[0mzero_num\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m<ipython-input-63-8f2131412615>\u001b[0m in \u001b[0;36mbacktesting\u001b[0;34m(df)\u001b[0m\n\u001b[1;32m     35\u001b[0m   \u001b[0minitial_cash\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mstart\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mgetBroker\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mgetCash\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     36\u001b[0m   \u001b[0mlogging\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mgetLogger\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"broker.backtesting\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msetLevel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlogging\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mCRITICAL\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 37\u001b[0;31m   \u001b[0mstart\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrun\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     38\u001b[0m   \u001b[0mprofit\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtradesAnalyzer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mgetProfits\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0mtradesAnalyzer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mgetLosses\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m/\u001b[0m \u001b[0minitial_cash\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m*\u001b[0m \u001b[0;36m100\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     39\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/pyalgotrade/strategy/__init__.py\u001b[0m in \u001b[0;36mrun\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    517\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0monFinish\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__barFeed\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mgetCurrentBars\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    518\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 519\u001b[0;31m             \u001b[0;32mraise\u001b[0m \u001b[0mException\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"Feed was empty\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    520\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    521\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mstop\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mException\u001b[0m: Feed was empty"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "online learning"
      ],
      "metadata": {
        "id": "-TkJMk2NtWWb"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# 기존 모델 불러오기\n",
        "model = torch.load(\"trained_model.pth\")  # 훈련된 모델 로드"
      ],
      "metadata": {
        "id": "h9e1J_1bteSY"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model = ...\n",
        "\n",
        "# new dataset\n"
      ],
      "metadata": {
        "id": "NBevrFyjtZLQ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# 새로운 데이터에 대해 모델 재학습\n",
        "model.train()\n",
        "for epoch in range(3):  # 간단히 3 epoch 동안만 학습\n",
        "    for inputs, labels in new_loader:\n",
        "        optimizer.zero_grad()  # 기존 Gradient 초기화\n",
        "        outputs = model(inputs)  # 모델 예측\n",
        "        loss = criterion(outputs, labels)  # 손실 계산\n",
        "        loss.backward()  # Backpropagation 수행\n",
        "        optimizer.step()  # 모델 파라미터 업데이트\n",
        "        print(f\"Epoch {epoch+1}, Loss: {loss.item()}\")"
      ],
      "metadata": {
        "id": "Zrcl_E01tX0t"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# 학습된 모델 저장\n",
        "torch.save(model, \"online_trained_model.pth\")"
      ],
      "metadata": {
        "id": "JeOQTHRMtct5"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}