FIFA_machine_learning.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "toc_visible": true,
      "authorship_tag": "ABX9TyMavFBpinpBrJAAILx0S1Zh",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/lnalice/FIFA-fee-forcasting/blob/master/FIFA_machine_learning.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 123,
      "metadata": {
        "id": "74oMnO-UCLZD"
      },
      "outputs": [],
      "source": [
        "from google.colab import files\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from scipy.stats import uniform, randint\n",
        "\n",
        "# setting package\n",
        "import warnings\n",
        "warnings.filterwarnings('always')\n",
        "warnings.filterwarnings('ignore')\n",
        "\n",
        "# standard data\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "\n",
        "# model\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import mean_squared_error\n",
        "from sklearn.metrics import mean_absolute_error\n",
        "from sklearn.model_selection import RandomizedSearchCV\n",
        "from sklearn.model_selection import cross_val_score, cross_validate\n",
        "\n",
        "# LGBM\n",
        "from lightgbm import LGBMRegressor\n",
        "\n",
        "from sklearn.metrics import confusion_matrix, accuracy_score, precision_score, recall_score, f1_score, roc_auc_score\n",
        "\n",
        "#\n",
        "# from sklearn.tree import DecisionTreeRegressor\n",
        "# from sklearn.model_selection import KFold\n",
        "# from sklearn.tree import plot_tree\n",
        "\n",
        "\n",
        "# from sklearn.ensemble import RandomForestRegressor, GradientBoostingRegressor"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "FIFA_files = files.upload();\n",
        "\n",
        "train_df = pd.read_csv('/content/FIFA_train.csv')\n",
        "test_df = pd.read_csv('/content/FIFA_test.csv')\n",
        "submission = pd.read_csv('/content/submission.csv')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 148
        },
        "id": "2WzXCjo7CTwr",
        "outputId": "16fbf899-3eda-4e45-dd40-a2d593e480ed"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-56853432-e39d-4fa1-9535-a673a24b9df7\" name=\"files[]\" multiple disabled\n",
              "        style=\"border:none\" />\n",
              "     <output id=\"result-56853432-e39d-4fa1-9535-a673a24b9df7\">\n",
              "      Upload widget is only available when the cell has been executed in the\n",
              "      current browser session. Please rerun this cell to enable.\n",
              "      </output>\n",
              "      <script>// Copyright 2017 Google LLC\n",
              "//\n",
              "// Licensed under the Apache License, Version 2.0 (the \"License\");\n",
              "// you may not use this file except in compliance with the License.\n",
              "// You may obtain a copy of the License at\n",
              "//\n",
              "//      http://www.apache.org/licenses/LICENSE-2.0\n",
              "//\n",
              "// Unless required by applicable law or agreed to in writing, software\n",
              "// distributed under the License is distributed on an \"AS IS\" BASIS,\n",
              "// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n",
              "// See the License for the specific language governing permissions and\n",
              "// limitations under the License.\n",
              "\n",
              "/**\n",
              " * @fileoverview Helpers for google.colab Python module.\n",
              " */\n",
              "(function(scope) {\n",
              "function span(text, styleAttributes = {}) {\n",
              "  const element = document.createElement('span');\n",
              "  element.textContent = text;\n",
              "  for (const key of Object.keys(styleAttributes)) {\n",
              "    element.style[key] = styleAttributes[key];\n",
              "  }\n",
              "  return element;\n",
              "}\n",
              "\n",
              "// Max number of bytes which will be uploaded at a time.\n",
              "const MAX_PAYLOAD_SIZE = 100 * 1024;\n",
              "\n",
              "function _uploadFiles(inputId, outputId) {\n",
              "  const steps = uploadFilesStep(inputId, outputId);\n",
              "  const outputElement = document.getElementById(outputId);\n",
              "  // Cache steps on the outputElement to make it available for the next call\n",
              "  // to uploadFilesContinue from Python.\n",
              "  outputElement.steps = steps;\n",
              "\n",
              "  return _uploadFilesContinue(outputId);\n",
              "}\n",
              "\n",
              "// This is roughly an async generator (not supported in the browser yet),\n",
              "// where there are multiple asynchronous steps and the Python side is going\n",
              "// to poll for completion of each step.\n",
              "// This uses a Promise to block the python side on completion of each step,\n",
              "// then passes the result of the previous step as the input to the next step.\n",
              "function _uploadFilesContinue(outputId) {\n",
              "  const outputElement = document.getElementById(outputId);\n",
              "  const steps = outputElement.steps;\n",
              "\n",
              "  const next = steps.next(outputElement.lastPromiseValue);\n",
              "  return Promise.resolve(next.value.promise).then((value) => {\n",
              "    // Cache the last promise value to make it available to the next\n",
              "    // step of the generator.\n",
              "    outputElement.lastPromiseValue = value;\n",
              "    return next.value.response;\n",
              "  });\n",
              "}\n",
              "\n",
              "/**\n",
              " * Generator function which is called between each async step of the upload\n",
              " * process.\n",
              " * @param {string} inputId Element ID of the input file picker element.\n",
              " * @param {string} outputId Element ID of the output display.\n",
              " * @return {!Iterable<!Object>} Iterable of next steps.\n",
              " */\n",
              "function* uploadFilesStep(inputId, outputId) {\n",
              "  const inputElement = document.getElementById(inputId);\n",
              "  inputElement.disabled = false;\n",
              "\n",
              "  const outputElement = document.getElementById(outputId);\n",
              "  outputElement.innerHTML = '';\n",
              "\n",
              "  const pickedPromise = new Promise((resolve) => {\n",
              "    inputElement.addEventListener('change', (e) => {\n",
              "      resolve(e.target.files);\n",
              "    });\n",
              "  });\n",
              "\n",
              "  const cancel = document.createElement('button');\n",
              "  inputElement.parentElement.appendChild(cancel);\n",
              "  cancel.textContent = 'Cancel upload';\n",
              "  const cancelPromise = new Promise((resolve) => {\n",
              "    cancel.onclick = () => {\n",
              "      resolve(null);\n",
              "    };\n",
              "  });\n",
              "\n",
              "  // Wait for the user to pick the files.\n",
              "  const files = yield {\n",
              "    promise: Promise.race([pickedPromise, cancelPromise]),\n",
              "    response: {\n",
              "      action: 'starting',\n",
              "    }\n",
              "  };\n",
              "\n",
              "  cancel.remove();\n",
              "\n",
              "  // Disable the input element since further picks are not allowed.\n",
              "  inputElement.disabled = true;\n",
              "\n",
              "  if (!files) {\n",
              "    return {\n",
              "      response: {\n",
              "        action: 'complete',\n",
              "      }\n",
              "    };\n",
              "  }\n",
              "\n",
              "  for (const file of files) {\n",
              "    const li = document.createElement('li');\n",
              "    li.append(span(file.name, {fontWeight: 'bold'}));\n",
              "    li.append(span(\n",
              "        `(${file.type || 'n/a'}) - ${file.size} bytes, ` +\n",
              "        `last modified: ${\n",
              "            file.lastModifiedDate ? file.lastModifiedDate.toLocaleDateString() :\n",
              "                                    'n/a'} - `));\n",
              "    const percent = span('0% done');\n",
              "    li.appendChild(percent);\n",
              "\n",
              "    outputElement.appendChild(li);\n",
              "\n",
              "    const fileDataPromise = new Promise((resolve) => {\n",
              "      const reader = new FileReader();\n",
              "      reader.onload = (e) => {\n",
              "        resolve(e.target.result);\n",
              "      };\n",
              "      reader.readAsArrayBuffer(file);\n",
              "    });\n",
              "    // Wait for the data to be ready.\n",
              "    let fileData = yield {\n",
              "      promise: fileDataPromise,\n",
              "      response: {\n",
              "        action: 'continue',\n",
              "      }\n",
              "    };\n",
              "\n",
              "    // Use a chunked sending to avoid message size limits. See b/62115660.\n",
              "    let position = 0;\n",
              "    do {\n",
              "      const length = Math.min(fileData.byteLength - position, MAX_PAYLOAD_SIZE);\n",
              "      const chunk = new Uint8Array(fileData, position, length);\n",
              "      position += length;\n",
              "\n",
              "      const base64 = btoa(String.fromCharCode.apply(null, chunk));\n",
              "      yield {\n",
              "        response: {\n",
              "          action: 'append',\n",
              "          file: file.name,\n",
              "          data: base64,\n",
              "        },\n",
              "      };\n",
              "\n",
              "      let percentDone = fileData.byteLength === 0 ?\n",
              "          100 :\n",
              "          Math.round((position / fileData.byteLength) * 100);\n",
              "      percent.textContent = `${percentDone}% done`;\n",
              "\n",
              "    } while (position < fileData.byteLength);\n",
              "  }\n",
              "\n",
              "  // All done.\n",
              "  yield {\n",
              "    response: {\n",
              "      action: 'complete',\n",
              "    }\n",
              "  };\n",
              "}\n",
              "\n",
              "scope.google = scope.google || {};\n",
              "scope.google.colab = scope.google.colab || {};\n",
              "scope.google.colab._files = {\n",
              "  _uploadFiles,\n",
              "  _uploadFilesContinue,\n",
              "};\n",
              "})(self);\n",
              "</script> "
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Saving FIFA_test.csv to FIFA_test.csv\n",
            "Saving FIFA_train.csv to FIFA_train.csv\n",
            "Saving submission.csv to submission.csv\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Data Preprocessing"
      ],
      "metadata": {
        "id": "w4HNecfANfNY"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 기존 데이터 확인"
      ],
      "metadata": {
        "id": "A70lbzpGhwVy"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "print(\"\\n--------------------- uniqueIndex ---------------------\")\n",
        "id = train_df['id'].unique();\n",
        "name = train_df['name'].unique();\n",
        "age = train_df['age'].unique();\n",
        "continent = train_df['continent'].unique();\n",
        "contract_until = train_df['contract_until'].unique();\n",
        "position = train_df['position'].unique();\n",
        "prefer_foot = train_df['prefer_foot'].unique();\n",
        "reputation = train_df['reputation'].unique();\n",
        "stat_overall = train_df['stat_overall'].unique();\n",
        "stat_potential = train_df['stat_potential'].unique();\n",
        "stat_skill_moves = train_df['stat_skill_moves'].unique();\n",
        "value = train_df['value'].unique();\n",
        "\n",
        "uniqueIndex = {\n",
        "    \"id\" : id,\n",
        "    \"name\" : name,\n",
        "    \"age\" : age,\n",
        "  \"continent\" : continent,\n",
        "  \"contract_until\" :contract_until,\n",
        "  \"position\" : position,\n",
        "  \"prefer_foot\" : prefer_foot,\n",
        "  \"reputation\" : reputation,\n",
        "  \"stat_overall\" : stat_overall,\n",
        "  \"stat_potential\" : stat_potential,\n",
        "  \"stat_skill_moves\" : stat_skill_moves,\n",
        "  \"value\" : value\n",
        "}\n",
        "\n",
        "uniqueIndex"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oJHoUIFnHB-q",
        "outputId": "2a57bae7-87e7-4b37-906d-b50aadc46a1b"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "--------------------- uniqueIndex ---------------------\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'id': array([    0,     3,     7, ..., 16941, 16942, 16948]),\n",
              " 'name': array(['L. Messi', 'De Gea', 'L. Suárez', ..., 'N. Fuentes', 'J. Milli',\n",
              "        'N. Christoffersson'], dtype=object),\n",
              " 'age': array([31, 27, 32, 25, 26, 29, 33, 30, 40, 24, 28, 34, 23, 22, 35, 36, 21,\n",
              "        18, 19, 37, 20, 39, 17, 38, 16]),\n",
              " 'continent': array(['south america', 'europe', 'africa', 'asia', 'oceania'],\n",
              "       dtype=object),\n",
              " 'contract_until': array(['2021', '2020', '2019', '2023', '2022', '2024', 'Jun 30, 2019',\n",
              "        '2026', 'Dec 31, 2018', '2018', '2025', 'Jun 30, 2020',\n",
              "        'May 31, 2020', 'May 31, 2019', 'Jan 31, 2019', 'Jan 1, 2019',\n",
              "        'Jan 12, 2019'], dtype=object),\n",
              " 'position': array(['ST', 'GK', 'DF', 'MF'], dtype=object),\n",
              " 'prefer_foot': array(['left', 'right'], dtype=object),\n",
              " 'reputation': array([5., 4., 3., 1., 2.]),\n",
              " 'stat_overall': array([94, 91, 90, 89, 88, 87, 86, 85, 84, 83, 82, 81, 80, 79, 78, 77, 76,\n",
              "        75, 74, 73, 72, 71, 70, 69, 68, 67, 66, 65, 64, 63, 62, 61, 60, 59,\n",
              "        58, 57, 56, 55, 54, 53, 52, 51, 50, 49, 48, 47]),\n",
              " 'stat_potential': array([94, 93, 91, 90, 92, 89, 88, 87, 86, 85, 84, 83, 82, 81, 80, 79, 78,\n",
              "        77, 76, 75, 74, 73, 72, 71, 70, 69, 68, 67, 66, 65, 64, 63, 62, 61,\n",
              "        60, 59, 58, 57, 56, 55, 54, 53, 52, 50, 48]),\n",
              " 'stat_skill_moves': array([4., 1., 3., 2., 5.]),\n",
              " 'value': array([1.105e+08, 7.200e+07, 8.000e+07, 5.100e+07, 6.800e+07, 4.400e+07,\n",
              "        6.000e+07, 6.300e+07, 5.800e+07, 5.350e+07, 5.150e+07, 3.800e+07,\n",
              "        2.700e+07, 6.950e+07, 5.950e+07, 6.200e+07, 7.350e+07, 5.900e+07,\n",
              "        4.600e+07, 3.600e+07, 5.700e+07, 4.000e+06, 6.400e+07, 6.250e+07,\n",
              "        4.500e+07, 4.650e+07, 3.400e+07, 5.200e+07, 4.450e+07, 5.500e+07,\n",
              "        3.650e+07, 4.550e+07, 4.350e+07, 3.000e+07, 1.800e+07, 2.150e+07,\n",
              "        5.050e+07, 5.400e+07, 2.850e+07, 3.200e+07, 3.750e+07, 3.900e+07,\n",
              "        2.600e+07, 3.300e+07, 3.850e+07, 9.000e+06, 3.700e+07, 2.200e+07,\n",
              "        1.400e+07, 4.050e+07, 3.550e+07, 3.150e+07, 4.200e+07, 2.500e+07,\n",
              "        2.950e+07, 2.400e+07, 3.100e+07, 2.750e+07, 3.050e+07, 2.900e+07,\n",
              "        1.650e+07, 1.900e+07, 1.550e+07, 4.000e+07, 4.100e+07, 2.800e+07,\n",
              "        3.450e+07, 3.250e+07, 2.000e+07, 3.500e+07, 2.450e+07, 2.650e+07,\n",
              "        2.100e+07, 2.550e+07, 1.750e+07, 1.300e+07, 8.000e+06, 6.000e+06,\n",
              "        1.950e+07, 2.250e+07, 1.850e+07, 2.350e+07, 1.700e+07, 1.250e+07,\n",
              "        2.300e+07, 1.500e+07, 1.350e+07, 4.800e+06, 3.000e+06, 1.600e+07,\n",
              "        1.100e+07, 7.000e+06, 1.000e+07, 5.500e+06, 2.050e+07, 1.050e+07,\n",
              "        4.500e+06, 1.200e+07, 1.450e+07, 1.150e+07, 8.500e+06, 1.700e+06,\n",
              "        1.000e+06, 9.500e+06, 7.500e+06, 6.500e+06, 5.000e+06, 2.400e+06,\n",
              "        2.900e+06, 4.100e+06, 2.100e+06, 3.600e+06, 2.700e+06, 2.500e+06,\n",
              "        3.200e+06, 4.700e+06, 3.400e+06, 4.900e+06, 4.300e+06, 2.300e+06,\n",
              "        3.100e+06, 3.900e+06, 2.200e+06, 4.400e+06, 9.000e+05, 3.700e+06,\n",
              "        1.600e+06, 3.800e+06, 1.900e+06, 4.500e+05, 2.000e+06, 4.200e+06,\n",
              "        7.750e+05, 6.500e+05, 7.500e+05, 2.800e+06, 1.300e+06, 1.200e+06,\n",
              "        4.600e+06, 3.500e+06, 3.300e+06, 2.600e+06, 2.700e+05, 5.500e+05,\n",
              "        9.500e+05, 1.100e+06, 1.800e+06, 9.750e+05, 1.400e+06, 7.250e+05,\n",
              "        4.250e+05, 2.100e+05, 8.750e+05, 1.500e+06, 3.250e+05, 8.000e+05,\n",
              "        8.500e+05, 1.600e+05, 6.750e+05, 1.200e+05, 6.250e+05, 2.400e+05,\n",
              "        8.250e+05, 9.250e+05, 5.000e+05, 6.000e+05, 5.750e+05, 2.500e+05,\n",
              "        7.000e+05, 3.500e+05, 3.000e+05, 3.750e+05, 7.000e+04, 4.750e+05,\n",
              "        5.250e+05, 1.400e+05, 2.300e+05, 4.000e+05, 6.000e+04, 5.000e+04,\n",
              "        1.000e+05, 1.800e+05, 9.000e+04, 2.900e+05, 2.000e+05, 1.500e+05,\n",
              "        4.000e+04, 2.600e+05, 2.200e+05, 2.800e+05, 1.300e+05, 1.700e+05,\n",
              "        3.000e+04, 1.100e+05, 1.900e+05, 2.000e+04, 8.000e+04, 1.000e+04])}"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 데이터 전처리 시작\n",
        "\n",
        "### train_df → new_train_df → tidy_train_df\n",
        "- new_train_df : 데이터 형식 변환\n",
        "- tidy_train_df : 왜도 높은 칼럼 로그변환, 더미화, id/name 열 삭제"
      ],
      "metadata": {
        "id": "ndHlpKk1NmqH"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "\n",
        "## (1) new_train_df\n",
        "### value 수정 (형식변환, 의미변환)\n",
        "- contract_until\n",
        "  - 수정 전 : string 타입, 형식 다름\n",
        "  - 수정 후 : int 타입, 형식 맞추기(ex 2019)\n",
        "  - 완   료 : 계약 기간(ex 2024-2018= 6)\n",
        "\n",
        "- value\n",
        "  - 수정 전 : 값 그대로\n",
        "  - 수정 후 : log(value)\n",
        "\n",
        "<!-- get_dummies : 범주형 칼럼 더미화해서 표준화함 -->\n",
        "<!-- np.logp1 : 왜도가 높은 컬럼 로그 변환 -->\n"
      ],
      "metadata": {
        "id": "TBtj8kb4eg0g"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "'''\n",
        "contract_until\n",
        "'''\n",
        "oldDate = []\n",
        "newDate = []\n",
        "for e in train_df[\"contract_until\"] :\n",
        "    oldDate.append(e)\n",
        "    newDate.append(int(e[-4:])-2018)\n",
        "\n",
        "new_train_df = train_df.replace(oldDate, newDate)\n",
        "uniqueIndex[\"contract_until\"] = new_train_df[\"contract_until\"].unique()\n",
        "\n",
        "'''\n",
        "value\n",
        "'''\n",
        "oldValue = []\n",
        "newValue= []\n",
        "for e in train_df[\"value\"] :\n",
        "    oldValue.append(e)\n",
        "    newValue.append(np.log1p(e))\n",
        "\n",
        "new_train_df = new_train_df.replace(oldValue, newValue)\n",
        "uniqueIndex[\"value\"] = new_train_df[\"value\"].unique()\n",
        "\n",
        "print(\"\\n--------------------- new_train_df ---------------------\")\n",
        "print(new_train_df)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HboG7eA9E4yT",
        "outputId": "b09e6c8b-2797-496b-a6fe-8a3679aa707e"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "--------------------- new_train_df ---------------------\n",
            "           id                name  age      continent  contract_until  \\\n",
            "0         0.0            L. Messi   31  south america               3   \n",
            "1         3.0              De Gea   27         europe               2   \n",
            "2         7.0           L. Suárez   31  south america               3   \n",
            "3         8.0        Sergio Ramos   32         europe               2   \n",
            "4         9.0            J. Oblak   25         europe               3   \n",
            "...       ...                 ...  ...            ...             ...   \n",
            "8927  16925.0          S. Adewusi   18         africa               1   \n",
            "8928  16936.0           C. Ehlich   19         europe               2   \n",
            "8929  16941.0          N. Fuentes   18  south america               3   \n",
            "8930  16942.0            J. Milli   18         europe               3   \n",
            "8931  16948.0  N. Christoffersson   19         europe               2   \n",
            "\n",
            "     position prefer_foot  reputation  stat_overall  stat_potential  \\\n",
            "0          ST        left         5.0            94              94   \n",
            "1          GK       right         4.0            91              93   \n",
            "2          ST       right         5.0            91              91   \n",
            "3          DF       right         4.0            91              91   \n",
            "4          GK       right         3.0            90              93   \n",
            "...       ...         ...         ...           ...             ...   \n",
            "8927       MF       right         1.0            48              63   \n",
            "8928       DF       right         1.0            47              59   \n",
            "8929       DF       right         1.0            47              64   \n",
            "8930       GK       right         1.0            47              65   \n",
            "8931       ST       right         1.0            47              63   \n",
            "\n",
            "      stat_skill_moves      value  \n",
            "0                  4.0  18.520526  \n",
            "1                  1.0  18.092177  \n",
            "2                  3.0  18.197537  \n",
            "3                  3.0  17.747336  \n",
            "4                  1.0  18.035018  \n",
            "...                ...        ...  \n",
            "8927               3.0  11.002117  \n",
            "8928               2.0  10.596660  \n",
            "8929               2.0  10.819798  \n",
            "8930               1.0  10.819798  \n",
            "8931               2.0  11.002117  \n",
            "\n",
            "[8932 rows x 12 columns]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## (2) tidy_train_df\n",
        "\n",
        "### 왜도가 높은 컬럼 로그 변환\n",
        "- name : 열 삭제\n",
        "- reputation : 로그 변환\n",
        "  - (1) 0이나 0에 매우 가까운 값이 없다.\n",
        "  - (2) 음수가 없다\n",
        "  - 즉, 로그 변환 가능하며, 왜도가 크기 때문에 로그 변환을 함\n",
        "\n",
        "### 범주형 칼럼 더미화\n",
        "- continent, position, prefer_foot 더미화\n",
        "- id : 열 삭제"
      ],
      "metadata": {
        "id": "R2YNq_zNDaE6"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "'''\n",
        "이름과 더미화할 데이터 드롭\n",
        "'''\n",
        "print(\"\\n--------------------- tidy_train_df (drop) ---------------------\")\n",
        "tidy_train_df = new_train_df.drop(columns =['name', 'continent', 'position', 'prefer_foot'])\n",
        "print(tidy_train_df.head())\n",
        "\n",
        "'''\n",
        "왜도확인\n",
        "'''\n",
        "print(\"\\n--------------------- skew ---------------------\")\n",
        "print(tidy_train_df.skew())\n",
        "tidy_train_df.hist(bins=20, figsize=(10,10))\n",
        "\n",
        "\n",
        "'''\n",
        "reputation - 로그 변환\n",
        "(1) 0이나 0에 매우 가까운 값이 없다.\n",
        "(2) 음수가 없다\n",
        "즉, 로그 변환 가능하다\n",
        "'''\n",
        "print(\"\\n--------------------- tidy_train_df (log) ---------------------\")\n",
        "tidy_train_df['reputation'] = np.log(tidy_train_df['reputation'])\n",
        "print(tidy_train_df.head())\n",
        "\n",
        "\n",
        "\n",
        "'''\n",
        "더미화 - continent, position, prefer_foot\n",
        "'''\n",
        "print(\"\\n--------------------- tidy_train_df (dummies) ---------------------\")\n",
        "id = tidy_train_df['id']\n",
        "dummies_continent = pd.get_dummies(new_train_df['continent'], prefix='cnt', dummy_na=False)\n",
        "dummies_position = pd.get_dummies(new_train_df['position'], prefix='pos', dummy_na=False)\n",
        "dummies_prefer_foot = pd.get_dummies(new_train_df['prefer_foot'], prefix='foot', dummy_na=False)\n",
        "dummies_df = pd.concat([id, dummies_continent,dummies_position, dummies_prefer_foot], axis=1)\n",
        "dummies_df.head()\n",
        "\n",
        "'''\n",
        "merge\n",
        "'''\n",
        "tidy_train_df = pd.merge(tidy_train_df,dummies_df,  how='left')\n",
        "tidy_train_df.columns\n",
        "\n",
        "'''\n",
        "순서 바꾸기 & drop id\n",
        "'''\n",
        "before_value = tidy_train_df.columns[1:7].to_list()\n",
        "after_value = dummies_df.columns[1:].to_list()\n",
        "new_col = before_value + after_value + ['value']\n",
        "tidy_train_df=tidy_train_df[new_col]\n",
        "tidy_train_df.head()"
      ],
      "metadata": {
        "id": "VRHrpH87EGnC",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "4b86c38f-5d25-47d4-ff7e-db93715629c6"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "--------------------- tidy_train_df (drop) ---------------------\n",
            "    id  age  contract_until  reputation  stat_overall  stat_potential  \\\n",
            "0  0.0   31               3         5.0            94              94   \n",
            "1  3.0   27               2         4.0            91              93   \n",
            "2  7.0   31               3         5.0            91              91   \n",
            "3  8.0   32               2         4.0            91              91   \n",
            "4  9.0   25               3         3.0            90              93   \n",
            "\n",
            "   stat_skill_moves      value  \n",
            "0               4.0  18.520526  \n",
            "1               1.0  18.092177  \n",
            "2               3.0  18.197537  \n",
            "3               3.0  17.747336  \n",
            "4               1.0  18.035018  \n",
            "\n",
            "--------------------- skew ---------------------\n",
            "id                  0.131275\n",
            "age                 0.356824\n",
            "contract_until      0.493837\n",
            "reputation          3.774495\n",
            "stat_overall        0.026723\n",
            "stat_potential      0.285905\n",
            "stat_skill_moves    0.086536\n",
            "value               0.308798\n",
            "dtype: float64\n",
            "\n",
            "--------------------- tidy_train_df (log) ---------------------\n",
            "    id  age  contract_until  reputation  stat_overall  stat_potential  \\\n",
            "0  0.0   31               3    1.609438            94              94   \n",
            "1  3.0   27               2    1.386294            91              93   \n",
            "2  7.0   31               3    1.609438            91              91   \n",
            "3  8.0   32               2    1.386294            91              91   \n",
            "4  9.0   25               3    1.098612            90              93   \n",
            "\n",
            "   stat_skill_moves      value  \n",
            "0               4.0  18.520526  \n",
            "1               1.0  18.092177  \n",
            "2               3.0  18.197537  \n",
            "3               3.0  17.747336  \n",
            "4               1.0  18.035018  \n",
            "\n",
            "--------------------- tidy_train_df (dummies) ---------------------\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   age  contract_until  reputation  stat_overall  stat_potential  \\\n",
              "0   31               3    1.609438            94              94   \n",
              "1   27               2    1.386294            91              93   \n",
              "2   31               3    1.609438            91              91   \n",
              "3   32               2    1.386294            91              91   \n",
              "4   25               3    1.098612            90              93   \n",
              "\n",
              "   stat_skill_moves  cnt_africa  cnt_asia  cnt_europe  cnt_oceania  \\\n",
              "0               4.0           0         0           0            0   \n",
              "1               1.0           0         0           1            0   \n",
              "2               3.0           0         0           0            0   \n",
              "3               3.0           0         0           1            0   \n",
              "4               1.0           0         0           1            0   \n",
              "\n",
              "   cnt_south america  pos_DF  pos_GK  pos_MF  pos_ST  foot_left  foot_right  \\\n",
              "0                  1       0       0       0       1          1           0   \n",
              "1                  0       0       1       0       0          0           1   \n",
              "2                  1       0       0       0       1          0           1   \n",
              "3                  0       1       0       0       0          0           1   \n",
              "4                  0       0       1       0       0          0           1   \n",
              "\n",
              "       value  \n",
              "0  18.520526  \n",
              "1  18.092177  \n",
              "2  18.197537  \n",
              "3  17.747336  \n",
              "4  18.035018  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-5f42b1cc-3692-4bbd-a4c6-b764f152b34e\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>age</th>\n",
              "      <th>contract_until</th>\n",
              "      <th>reputation</th>\n",
              "      <th>stat_overall</th>\n",
              "      <th>stat_potential</th>\n",
              "      <th>stat_skill_moves</th>\n",
              "      <th>cnt_africa</th>\n",
              "      <th>cnt_asia</th>\n",
              "      <th>cnt_europe</th>\n",
              "      <th>cnt_oceania</th>\n",
              "      <th>cnt_south america</th>\n",
              "      <th>pos_DF</th>\n",
              "      <th>pos_GK</th>\n",
              "      <th>pos_MF</th>\n",
              "      <th>pos_ST</th>\n",
              "      <th>foot_left</th>\n",
              "      <th>foot_right</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>31</td>\n",
              "      <td>3</td>\n",
              "      <td>1.609438</td>\n",
              "      <td>94</td>\n",
              "      <td>94</td>\n",
              "      <td>4.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>18.520526</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>27</td>\n",
              "      <td>2</td>\n",
              "      <td>1.386294</td>\n",
              "      <td>91</td>\n",
              "      <td>93</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>18.092177</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>31</td>\n",
              "      <td>3</td>\n",
              "      <td>1.609438</td>\n",
              "      <td>91</td>\n",
              "      <td>91</td>\n",
              "      <td>3.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>18.197537</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>32</td>\n",
              "      <td>2</td>\n",
              "      <td>1.386294</td>\n",
              "      <td>91</td>\n",
              "      <td>91</td>\n",
              "      <td>3.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>17.747336</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>25</td>\n",
              "      <td>3</td>\n",
              "      <td>1.098612</td>\n",
              "      <td>90</td>\n",
              "      <td>93</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>18.035018</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-5f42b1cc-3692-4bbd-a4c6-b764f152b34e')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-5f42b1cc-3692-4bbd-a4c6-b764f152b34e button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-5f42b1cc-3692-4bbd-a4c6-b764f152b34e');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 7
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x720 with 9 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tidy_train_df.info()"
      ],
      "metadata": {
        "id": "maLh_cp4DHhh",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "43470ba9-02a2-489d-9b9c-7b4b93f4b320"
      },
      "execution_count": 62,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "Int64Index: 8932 entries, 0 to 8931\n",
            "Data columns (total 18 columns):\n",
            " #   Column             Non-Null Count  Dtype  \n",
            "---  ------             --------------  -----  \n",
            " 0   age                8932 non-null   int64  \n",
            " 1   contract_until     8932 non-null   int64  \n",
            " 2   reputation         8932 non-null   float64\n",
            " 3   stat_overall       8932 non-null   int64  \n",
            " 4   stat_potential     8932 non-null   int64  \n",
            " 5   stat_skill_moves   8932 non-null   float64\n",
            " 6   cnt_africa         8932 non-null   uint8  \n",
            " 7   cnt_asia           8932 non-null   uint8  \n",
            " 8   cnt_europe         8932 non-null   uint8  \n",
            " 9   cnt_oceania        8932 non-null   uint8  \n",
            " 10  cnt_south america  8932 non-null   uint8  \n",
            " 11  pos_DF             8932 non-null   uint8  \n",
            " 12  pos_GK             8932 non-null   uint8  \n",
            " 13  pos_MF             8932 non-null   uint8  \n",
            " 14  pos_ST             8932 non-null   uint8  \n",
            " 15  foot_left          8932 non-null   uint8  \n",
            " 16  foot_right         8932 non-null   uint8  \n",
            " 17  value              8932 non-null   float64\n",
            "dtypes: float64(3), int64(4), uint8(11)\n",
            "memory usage: 654.2 KB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Machine Learning\n"
      ],
      "metadata": {
        "id": "hHZze1hKOZj6"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "##  Train data\n",
        "### train_input\n",
        "- id, age를 제외한 모든 열\n",
        "- StandardScaler()\n",
        "\n",
        "### train_taget\n",
        "- value 열 "
      ],
      "metadata": {
        "id": "GxKdfoFLg4IP"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "print(\"\\n--------------------- train_input ---------------------\")\n",
        "train_input= tidy_train_df.drop(columns =['value'])\n",
        "ss = StandardScaler()\n",
        "train_input = ss.fit_transform(train_input)\n",
        "print(train_input)\n",
        "\n",
        "print(\"\\n--------------------- train_target ---------------------\")\n",
        "train_target = new_train_df.iloc[:,-1]\n",
        "print(train_target)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tadrxFzAICe2",
        "outputId": "4a7b2056-e299-4640-eae5-30186a234536"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "--------------------- train_input ---------------------\n",
            "[[ 1.2493085   0.63595162  6.08585496 ...  2.05881269  1.80651169\n",
            "  -1.80651169]\n",
            " [ 0.38635718 -0.15342023  5.19671935 ... -0.48571684 -0.55355302\n",
            "   0.55355302]\n",
            " [ 1.2493085   0.63595162  6.08585496 ...  2.05881269 -0.55355302\n",
            "   0.55355302]\n",
            " ...\n",
            " [-1.5552833   0.63595162 -0.32709558 ... -0.48571684 -0.55355302\n",
            "   0.55355302]\n",
            " [-1.5552833   0.63595162 -0.32709558 ... -0.48571684 -0.55355302\n",
            "   0.55355302]\n",
            " [-1.33954546 -0.15342023 -0.32709558 ...  2.05881269 -0.55355302\n",
            "   0.55355302]]\n",
            "\n",
            "--------------------- train_target ---------------------\n",
            "0       18.520526\n",
            "1       18.092177\n",
            "2       18.197537\n",
            "3       17.747336\n",
            "4       18.035018\n",
            "          ...    \n",
            "8927    11.002117\n",
            "8928    10.596660\n",
            "8929    10.819798\n",
            "8930    10.819798\n",
            "8931    11.002117\n",
            "Name: value, Length: 8932, dtype: float64\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Test data"
      ],
      "metadata": {
        "id": "zitRjnoMITq3"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "test_df.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 357
        },
        "id": "p4TVGkgTJlU2",
        "outputId": "ce050e32-9445-4524-f370-28e9f437393e"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   id               name  age      continent contract_until position  \\\n",
              "0   1  Cristiano Ronaldo   33         europe           2022       ST   \n",
              "1   2          Neymar Jr   26  south america           2022       ST   \n",
              "2   4       K. De Bruyne   27         europe           2023       MF   \n",
              "3   5          E. Hazard   27         europe           2020       ST   \n",
              "4   6          L. Modrić   32         europe           2020       MF   \n",
              "\n",
              "  prefer_foot  reputation  stat_overall  stat_potential  stat_skill_moves  \n",
              "0       right         5.0            94              94               5.0  \n",
              "1       right         5.0            92              93               5.0  \n",
              "2       right         4.0            91              92               4.0  \n",
              "3       right         4.0            91              91               4.0  \n",
              "4       right         4.0            91              91               4.0  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-57cbe9be-0297-4bbe-a6be-4e90ea0df1dd\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>name</th>\n",
              "      <th>age</th>\n",
              "      <th>continent</th>\n",
              "      <th>contract_until</th>\n",
              "      <th>position</th>\n",
              "      <th>prefer_foot</th>\n",
              "      <th>reputation</th>\n",
              "      <th>stat_overall</th>\n",
              "      <th>stat_potential</th>\n",
              "      <th>stat_skill_moves</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>Cristiano Ronaldo</td>\n",
              "      <td>33</td>\n",
              "      <td>europe</td>\n",
              "      <td>2022</td>\n",
              "      <td>ST</td>\n",
              "      <td>right</td>\n",
              "      <td>5.0</td>\n",
              "      <td>94</td>\n",
              "      <td>94</td>\n",
              "      <td>5.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2</td>\n",
              "      <td>Neymar Jr</td>\n",
              "      <td>26</td>\n",
              "      <td>south america</td>\n",
              "      <td>2022</td>\n",
              "      <td>ST</td>\n",
              "      <td>right</td>\n",
              "      <td>5.0</td>\n",
              "      <td>92</td>\n",
              "      <td>93</td>\n",
              "      <td>5.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>4</td>\n",
              "      <td>K. De Bruyne</td>\n",
              "      <td>27</td>\n",
              "      <td>europe</td>\n",
              "      <td>2023</td>\n",
              "      <td>MF</td>\n",
              "      <td>right</td>\n",
              "      <td>4.0</td>\n",
              "      <td>91</td>\n",
              "      <td>92</td>\n",
              "      <td>4.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>5</td>\n",
              "      <td>E. Hazard</td>\n",
              "      <td>27</td>\n",
              "      <td>europe</td>\n",
              "      <td>2020</td>\n",
              "      <td>ST</td>\n",
              "      <td>right</td>\n",
              "      <td>4.0</td>\n",
              "      <td>91</td>\n",
              "      <td>91</td>\n",
              "      <td>4.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>6</td>\n",
              "      <td>L. Modrić</td>\n",
              "      <td>32</td>\n",
              "      <td>europe</td>\n",
              "      <td>2020</td>\n",
              "      <td>MF</td>\n",
              "      <td>right</td>\n",
              "      <td>4.0</td>\n",
              "      <td>91</td>\n",
              "      <td>91</td>\n",
              "      <td>4.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-57cbe9be-0297-4bbe-a6be-4e90ea0df1dd')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-57cbe9be-0297-4bbe-a6be-4e90ea0df1dd button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-57cbe9be-0297-4bbe-a6be-4e90ea0df1dd');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "'''\n",
        "contract_until\n",
        "'''\n",
        "oldDate = []\n",
        "newDate = []\n",
        "for e in test_df[\"contract_until\"] :\n",
        "    oldDate.append(e)\n",
        "    newDate.append(int(e[-4:])-2018)\n",
        "\n",
        "new_test_df = test_df.replace(oldDate, newDate)\n",
        "uniqueIndex[\"contract_until\"] = new_test_df[\"contract_until\"].unique()\n",
        "\n",
        "'''\n",
        "이름과 더미화할 데이터 드롭\n",
        "'''\n",
        "print(\"\\n--------------------- tidy_test_df (drop) ---------------------\")\n",
        "tidy_test_df = new_test_df.drop(columns =['name', \n",
        "                    'continent', 'position', 'prefer_foot'])\n",
        "print(tidy_test_df.head())\n",
        "\n",
        "'''\n",
        "왜도확인\n",
        "'''\n",
        "print(\"\\n--------------------- skew ---------------------\")\n",
        "print(tidy_test_df.skew())\n",
        "tidy_test_df.hist(bins=20, figsize=(10,10))\n",
        "\n",
        "\n",
        "'''\n",
        "reputation - 로그 변환\n",
        "(1) 0이나 0에 매우 가까운 값이 없다.\n",
        "(2) 음수가 없다\n",
        "즉, 로그 변환 가능하다\n",
        "'''\n",
        "print(\"\\n--------------------- tidy_test_df (log) ---------------------\")\n",
        "tidy_test_df['reputation'] = np.log(tidy_test_df['reputation'])\n",
        "print(tidy_test_df.head())\n",
        "\n",
        "\n",
        "\n",
        "'''\n",
        "더미화 - continent, position, prefer_foot\n",
        "'''\n",
        "print(\"\\n--------------------- tidy_test_df (dummies) ---------------------\")\n",
        "id = tidy_test_df['id']\n",
        "dummies_continent = pd.get_dummies(new_test_df['continent'], prefix='cnt', dummy_na=False)\n",
        "dummies_position = pd.get_dummies(new_test_df['position'], prefix='pos', dummy_na=False)\n",
        "dummies_prefer_foot = pd.get_dummies(new_test_df['prefer_foot'], prefix='foot', dummy_na=False)\n",
        "dummies_df = pd.concat([id, dummies_continent,dummies_position, dummies_prefer_foot], axis=1)\n",
        "print(dummies_df.head())\n",
        "\n",
        "'''\n",
        "merge\n",
        "'''\n",
        "tidy_test_df = pd.merge(tidy_test_df,dummies_df,  how='left')\n",
        "tidy_test_df.columns\n",
        "\n",
        "'''\n",
        "순서 바꾸기 & drop id\n",
        "'''\n",
        "before_value = tidy_test_df.columns[1:7].to_list()\n",
        "after_value = dummies_df.columns[1:].to_list()\n",
        "new_col = before_value + after_value\n",
        "tidy_test_df=tidy_test_df[new_col]\n",
        "tidy_test_df.head()\n",
        "\n",
        "print(\"\\n--------------------- tidy_test_df (completed) ---------------------\")\n",
        "print(tidy_test_df.head())\n",
        "print(tidy_test_df.info())\n",
        "\n",
        "\n",
        "'''\n",
        "test_input\n",
        "'''\n",
        "print(\"\\n--------------------- test_input ---------------------\")\n",
        "ss = StandardScaler()\n",
        "test_input = ss.fit_transform(tidy_test_df)\n",
        "print(test_input)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "HQtwhakkIYa5",
        "outputId": "cd1a8883-516c-49bc-af23-6f51c818b5a3"
      },
      "execution_count": 83,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "--------------------- tidy_test_df (drop) ---------------------\n",
            "   id  age  contract_until  reputation  stat_overall  stat_potential  \\\n",
            "0   1   33               4         5.0            94              94   \n",
            "1   2   26               4         5.0            92              93   \n",
            "2   4   27               5         4.0            91              92   \n",
            "3   5   27               2         4.0            91              91   \n",
            "4   6   32               2         4.0            91              91   \n",
            "\n",
            "   stat_skill_moves  \n",
            "0               5.0  \n",
            "1               5.0  \n",
            "2               4.0  \n",
            "3               4.0  \n",
            "4               4.0  \n",
            "\n",
            "--------------------- skew ---------------------\n",
            "id                  0.118753\n",
            "age                 0.364907\n",
            "contract_until      0.476073\n",
            "reputation          3.644128\n",
            "stat_overall        0.134543\n",
            "stat_potential      0.324914\n",
            "stat_skill_moves    0.134095\n",
            "dtype: float64\n",
            "\n",
            "--------------------- tidy_test_df (log) ---------------------\n",
            "   id  age  contract_until  reputation  stat_overall  stat_potential  \\\n",
            "0   1   33               4    1.609438            94              94   \n",
            "1   2   26               4    1.609438            92              93   \n",
            "2   4   27               5    1.386294            91              92   \n",
            "3   5   27               2    1.386294            91              91   \n",
            "4   6   32               2    1.386294            91              91   \n",
            "\n",
            "   stat_skill_moves  \n",
            "0               5.0  \n",
            "1               5.0  \n",
            "2               4.0  \n",
            "3               4.0  \n",
            "4               4.0  \n",
            "\n",
            "--------------------- tidy_test_df (dummies) ---------------------\n",
            "   id  cnt_africa  cnt_asia  cnt_europe  cnt_oceania  cnt_south america  \\\n",
            "0   1           0         0           1            0                  0   \n",
            "1   2           0         0           0            0                  1   \n",
            "2   4           0         0           1            0                  0   \n",
            "3   5           0         0           1            0                  0   \n",
            "4   6           0         0           1            0                  0   \n",
            "\n",
            "   pos_DF  pos_GK  pos_MF  pos_ST  foot_left  foot_right  \n",
            "0       0       0       0       1          0           1  \n",
            "1       0       0       0       1          0           1  \n",
            "2       0       0       1       0          0           1  \n",
            "3       0       0       0       1          0           1  \n",
            "4       0       0       1       0          0           1  \n",
            "\n",
            "--------------------- tidy_test_df (completed) ---------------------\n",
            "   age  contract_until  reputation  stat_overall  stat_potential  \\\n",
            "0   33               4    1.609438            94              94   \n",
            "1   26               4    1.609438            92              93   \n",
            "2   27               5    1.386294            91              92   \n",
            "3   27               2    1.386294            91              91   \n",
            "4   32               2    1.386294            91              91   \n",
            "\n",
            "   stat_skill_moves  cnt_africa  cnt_asia  cnt_europe  cnt_oceania  \\\n",
            "0               5.0           0         0           1            0   \n",
            "1               5.0           0         0           0            0   \n",
            "2               4.0           0         0           1            0   \n",
            "3               4.0           0         0           1            0   \n",
            "4               4.0           0         0           1            0   \n",
            "\n",
            "   cnt_south america  pos_DF  pos_GK  pos_MF  pos_ST  foot_left  foot_right  \n",
            "0                  0       0       0       0       1          0           1  \n",
            "1                  1       0       0       0       1          0           1  \n",
            "2                  0       0       0       1       0          0           1  \n",
            "3                  0       0       0       0       1          0           1  \n",
            "4                  0       0       0       1       0          0           1  \n",
            "<class 'pandas.core.frame.DataFrame'>\n",
            "Int64Index: 3828 entries, 0 to 3827\n",
            "Data columns (total 17 columns):\n",
            " #   Column             Non-Null Count  Dtype  \n",
            "---  ------             --------------  -----  \n",
            " 0   age                3828 non-null   int64  \n",
            " 1   contract_until     3828 non-null   int64  \n",
            " 2   reputation         3828 non-null   float64\n",
            " 3   stat_overall       3828 non-null   int64  \n",
            " 4   stat_potential     3828 non-null   int64  \n",
            " 5   stat_skill_moves   3828 non-null   float64\n",
            " 6   cnt_africa         3828 non-null   uint8  \n",
            " 7   cnt_asia           3828 non-null   uint8  \n",
            " 8   cnt_europe         3828 non-null   uint8  \n",
            " 9   cnt_oceania        3828 non-null   uint8  \n",
            " 10  cnt_south america  3828 non-null   uint8  \n",
            " 11  pos_DF             3828 non-null   uint8  \n",
            " 12  pos_GK             3828 non-null   uint8  \n",
            " 13  pos_MF             3828 non-null   uint8  \n",
            " 14  pos_ST             3828 non-null   uint8  \n",
            " 15  foot_left          3828 non-null   uint8  \n",
            " 16  foot_right         3828 non-null   uint8  \n",
            "dtypes: float64(2), int64(4), uint8(11)\n",
            "memory usage: 250.5 KB\n",
            "None\n",
            "\n",
            "--------------------- test_input ---------------------\n",
            "[[ 1.66439249  1.35670294  5.78318809 ...  2.08300555 -0.55964949\n",
            "   0.55964949]\n",
            " [ 0.15173949  1.35670294  5.78318809 ...  2.08300555 -0.55964949\n",
            "   0.55964949]\n",
            " [ 0.36783277  2.12530719  4.93405963 ... -0.48007553 -0.55964949\n",
            "   0.55964949]\n",
            " ...\n",
            " [-1.5770068   0.5880987  -0.34120823 ... -0.48007553 -0.55964949\n",
            "   0.55964949]\n",
            " [-1.5770068  -0.18050554 -0.34120823 ...  2.08300555 -0.55964949\n",
            "   0.55964949]\n",
            " [-1.36091352  0.5880987  -0.34120823 ... -0.48007553 -0.55964949\n",
            "   0.55964949]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x720 with 9 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## LGBM\n",
        "- Random Search\n",
        "  - 연속적인 매개변수의 최적값 찾을 때 유용\n",
        "  - 과적합 막기 위해 최적의 파라미터 값 찾기"
      ],
      "metadata": {
        "id": "q8Tc5iOhfw9Y"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "'''\n",
        "LGBM\n",
        "'''\n",
        "lgbm = LGBMRegressor()\n",
        "\n",
        "'''\n",
        "Random Search\n",
        "'''\n",
        "eval_set = [(train_input, train_target)]\n",
        "eval_metric = 'rmse'\n",
        "\n",
        "params = {\n",
        "    'num_leaves': randint(31,50),\n",
        "    'min_child_samples': randint(10,30),\n",
        "\t 'bagging_fraction' : uniform(0.5,0.1),\n",
        "\t 'feature_fraction' : uniform(0.5,0.1),\n",
        "\t 'learning_rate' : uniform(0.01, 0.1),\n",
        "\t 'early_stopping_round' : randint(5,10),\n",
        "}\n",
        "\n",
        "rs = RandomizedSearchCV(\n",
        "    estimator=lgbm, param_distributions=params, n_iter=100, cv=5, n_jobs=-1, \n",
        "    scoring='neg_mean_squared_error', random_state=42, verbose=1)\n",
        "rs.fit(train_input, train_target, eval_set = eval_set, eval_metric = eval_metric)\n",
        "\n",
        "print(\"best_params_ : \", rs.best_params_)\n",
        "print(\"train_score  : \", rs.score(train_input, train_target))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "n5oPc64AfvXe",
        "outputId": "e7bb82f3-e9d9-4c89-c0b9-4df6dd5045df"
      },
      "execution_count": 136,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Fitting 5 folds for each of 100 candidates, totalling 500 fits\n",
            "[LightGBM] [Warning] feature_fraction is set=0.561500722669917, colsample_bytree=1.0 will be ignored. Current value: feature_fraction=0.561500722669917\n",
            "[LightGBM] [Warning] bagging_fraction is set=0.5070416130849544, subsample=1.0 will be ignored. Current value: bagging_fraction=0.5070416130849544\n",
            "[1]\ttraining's rmse: 1.26308\ttraining's l2: 1.59537\n",
            "[2]\ttraining's rmse: 1.14109\ttraining's l2: 1.30208\n",
            "[3]\ttraining's rmse: 1.02652\ttraining's l2: 1.05375\n",
            "[4]\ttraining's rmse: 0.919737\ttraining's l2: 0.845915\n",
            "[5]\ttraining's rmse: 0.847596\ttraining's l2: 0.718419\n",
            "[6]\ttraining's rmse: 0.765173\ttraining's l2: 0.585489\n",
            "[7]\ttraining's rmse: 0.691446\ttraining's l2: 0.478097\n",
            "[8]\ttraining's rmse: 0.626384\ttraining's l2: 0.392357\n",
            "[9]\ttraining's rmse: 0.571766\ttraining's l2: 0.326916\n",
            "[10]\ttraining's rmse: 0.521344\ttraining's l2: 0.2718\n",
            "[11]\ttraining's rmse: 0.468771\ttraining's l2: 0.219746\n",
            "[12]\ttraining's rmse: 0.440427\ttraining's l2: 0.193976\n",
            "[13]\ttraining's rmse: 0.397182\ttraining's l2: 0.157753\n",
            "[14]\ttraining's rmse: 0.366329\ttraining's l2: 0.134197\n",
            "[15]\ttraining's rmse: 0.343418\ttraining's l2: 0.117936\n",
            "[16]\ttraining's rmse: 0.317043\ttraining's l2: 0.100516\n",
            "[17]\ttraining's rmse: 0.287188\ttraining's l2: 0.0824772\n",
            "[18]\ttraining's rmse: 0.27993\ttraining's l2: 0.0783608\n",
            "[19]\ttraining's rmse: 0.266067\ttraining's l2: 0.0707918\n",
            "[20]\ttraining's rmse: 0.248359\ttraining's l2: 0.0616821\n",
            "[21]\ttraining's rmse: 0.226324\ttraining's l2: 0.0512225\n",
            "[22]\ttraining's rmse: 0.207148\ttraining's l2: 0.0429102\n",
            "[23]\ttraining's rmse: 0.197334\ttraining's l2: 0.0389407\n",
            "[24]\ttraining's rmse: 0.189091\ttraining's l2: 0.0357555\n",
            "[25]\ttraining's rmse: 0.174161\ttraining's l2: 0.030332\n",
            "[26]\ttraining's rmse: 0.160931\ttraining's l2: 0.0258987\n",
            "[27]\ttraining's rmse: 0.153117\ttraining's l2: 0.023445\n",
            "[28]\ttraining's rmse: 0.14282\ttraining's l2: 0.0203977\n",
            "[29]\ttraining's rmse: 0.133981\ttraining's l2: 0.0179509\n",
            "[30]\ttraining's rmse: 0.12594\ttraining's l2: 0.0158608\n",
            "[31]\ttraining's rmse: 0.119664\ttraining's l2: 0.0143195\n",
            "[32]\ttraining's rmse: 0.115555\ttraining's l2: 0.0133529\n",
            "[33]\ttraining's rmse: 0.110063\ttraining's l2: 0.0121138\n",
            "[34]\ttraining's rmse: 0.106825\ttraining's l2: 0.0114116\n",
            "[35]\ttraining's rmse: 0.106074\ttraining's l2: 0.0112516\n",
            "[36]\ttraining's rmse: 0.104304\ttraining's l2: 0.0108794\n",
            "[37]\ttraining's rmse: 0.101941\ttraining's l2: 0.0103919\n",
            "[38]\ttraining's rmse: 0.0984323\ttraining's l2: 0.00968892\n",
            "[39]\ttraining's rmse: 0.0980321\ttraining's l2: 0.00961028\n",
            "[40]\ttraining's rmse: 0.0960414\ttraining's l2: 0.00922396\n",
            "[41]\ttraining's rmse: 0.0932\ttraining's l2: 0.00868624\n",
            "[42]\ttraining's rmse: 0.090392\ttraining's l2: 0.00817071\n",
            "[43]\ttraining's rmse: 0.0894333\ttraining's l2: 0.00799832\n",
            "[44]\ttraining's rmse: 0.0870004\ttraining's l2: 0.00756907\n",
            "[45]\ttraining's rmse: 0.0860154\ttraining's l2: 0.00739865\n",
            "[46]\ttraining's rmse: 0.0839529\ttraining's l2: 0.0070481\n",
            "[47]\ttraining's rmse: 0.0831729\ttraining's l2: 0.00691774\n",
            "[48]\ttraining's rmse: 0.0813271\ttraining's l2: 0.0066141\n",
            "[49]\ttraining's rmse: 0.0800372\ttraining's l2: 0.00640596\n",
            "[50]\ttraining's rmse: 0.0798453\ttraining's l2: 0.00637527\n",
            "[51]\ttraining's rmse: 0.078474\ttraining's l2: 0.00615817\n",
            "[52]\ttraining's rmse: 0.0775575\ttraining's l2: 0.00601517\n",
            "[53]\ttraining's rmse: 0.076842\ttraining's l2: 0.0059047\n",
            "[54]\ttraining's rmse: 0.0758208\ttraining's l2: 0.00574879\n",
            "[55]\ttraining's rmse: 0.0753659\ttraining's l2: 0.00568001\n",
            "[56]\ttraining's rmse: 0.075204\ttraining's l2: 0.00565564\n",
            "[57]\ttraining's rmse: 0.0746311\ttraining's l2: 0.00556979\n",
            "[58]\ttraining's rmse: 0.073716\ttraining's l2: 0.00543405\n",
            "[59]\ttraining's rmse: 0.0734339\ttraining's l2: 0.00539254\n",
            "[60]\ttraining's rmse: 0.0726614\ttraining's l2: 0.00527968\n",
            "[61]\ttraining's rmse: 0.0720952\ttraining's l2: 0.00519772\n",
            "[62]\ttraining's rmse: 0.0717859\ttraining's l2: 0.00515322\n",
            "[63]\ttraining's rmse: 0.0713818\ttraining's l2: 0.00509536\n",
            "[64]\ttraining's rmse: 0.0708136\ttraining's l2: 0.00501456\n",
            "[65]\ttraining's rmse: 0.0703658\ttraining's l2: 0.00495135\n",
            "[66]\ttraining's rmse: 0.0699969\ttraining's l2: 0.00489956\n",
            "[67]\ttraining's rmse: 0.0698722\ttraining's l2: 0.00488212\n",
            "[68]\ttraining's rmse: 0.0695215\ttraining's l2: 0.00483324\n",
            "[69]\ttraining's rmse: 0.0692492\ttraining's l2: 0.00479546\n",
            "[70]\ttraining's rmse: 0.0688808\ttraining's l2: 0.00474456\n",
            "[71]\ttraining's rmse: 0.0686383\ttraining's l2: 0.00471122\n",
            "[72]\ttraining's rmse: 0.0683421\ttraining's l2: 0.00467064\n",
            "[73]\ttraining's rmse: 0.068267\ttraining's l2: 0.00466038\n",
            "[74]\ttraining's rmse: 0.0678197\ttraining's l2: 0.00459951\n",
            "[75]\ttraining's rmse: 0.0676368\ttraining's l2: 0.00457473\n",
            "[76]\ttraining's rmse: 0.0674289\ttraining's l2: 0.00454666\n",
            "[77]\ttraining's rmse: 0.0672601\ttraining's l2: 0.00452393\n",
            "[78]\ttraining's rmse: 0.0670857\ttraining's l2: 0.00450049\n",
            "[79]\ttraining's rmse: 0.0667115\ttraining's l2: 0.00445042\n",
            "[80]\ttraining's rmse: 0.0662915\ttraining's l2: 0.00439456\n",
            "[81]\ttraining's rmse: 0.0660374\ttraining's l2: 0.00436094\n",
            "[82]\ttraining's rmse: 0.0657529\ttraining's l2: 0.00432344\n",
            "[83]\ttraining's rmse: 0.0656166\ttraining's l2: 0.00430553\n",
            "[84]\ttraining's rmse: 0.0652962\ttraining's l2: 0.00426359\n",
            "[85]\ttraining's rmse: 0.0652366\ttraining's l2: 0.00425581\n",
            "[86]\ttraining's rmse: 0.0650306\ttraining's l2: 0.00422898\n",
            "[87]\ttraining's rmse: 0.0646337\ttraining's l2: 0.00417752\n",
            "[88]\ttraining's rmse: 0.0644224\ttraining's l2: 0.00415025\n",
            "[89]\ttraining's rmse: 0.0642167\ttraining's l2: 0.00412379\n",
            "[90]\ttraining's rmse: 0.0639837\ttraining's l2: 0.00409391\n",
            "[91]\ttraining's rmse: 0.0637735\ttraining's l2: 0.00406706\n",
            "[92]\ttraining's rmse: 0.0636119\ttraining's l2: 0.00404647\n",
            "[93]\ttraining's rmse: 0.0632332\ttraining's l2: 0.00399843\n",
            "[94]\ttraining's rmse: 0.0631272\ttraining's l2: 0.00398504\n",
            "[95]\ttraining's rmse: 0.0628791\ttraining's l2: 0.00395378\n",
            "[96]\ttraining's rmse: 0.0626977\ttraining's l2: 0.003931\n",
            "[97]\ttraining's rmse: 0.0624794\ttraining's l2: 0.00390368\n",
            "[98]\ttraining's rmse: 0.0622958\ttraining's l2: 0.00388076\n",
            "[99]\ttraining's rmse: 0.0620138\ttraining's l2: 0.00384571\n",
            "[100]\ttraining's rmse: 0.0618615\ttraining's l2: 0.00382684\n",
            "best_params_ :  {'bagging_fraction': 0.5070416130849544, 'early_stopping_round': 6, 'feature_fraction': 0.561500722669917, 'learning_rate': 0.10900538501042632, 'min_child_samples': 11, 'num_leaves': 40}\n",
            "train_score  :  -0.0038268450437770546\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_pred = rs.predict(train_input)\n",
        "train_rmse = mean_squared_error(train_target, train_pred, squared=False)\n",
        "\n",
        "print(\"train RMSE : \", train_rmse)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Z2MLFGcuUJa0",
        "outputId": "b6047c11-6ed4-4e14-a71e-345ccac55e59"
      },
      "execution_count": 137,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "train RMSE :  0.0618614988807825\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Model evaluation"
      ],
      "metadata": {
        "id": "_lXJdUiQd2Z5"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "print(\"\\n--------------------- lgbm with random search ---------------------\")\n",
        "\n",
        "'''\n",
        "lgbm_best : set best_params_\n",
        "'''\n",
        "# categorical_col_name = ['cnt_africa', 'cnt_asia', 'cnt_europe', 'cnt_oceania', 'cnt_south america', 'pos_DF', 'pos_GK', 'pos_MF', 'pos_ST', 'foot_left', 'foot_right']\n",
        "categorical_col_num = [6,7,8,9,10,11,12,13,14,15,16]\n",
        "bp_rs = rs.best_params_\n",
        "\n",
        "lgbm_best = LGBMRegressor(num_leaves= bp_rs['num_leaves'], min_child_samples= bp_rs['min_child_samples'],\n",
        "                     bagging_fraction=bp_rs['bagging_fraction'], feature_fraction = bp_rs['feature_fraction'], \n",
        "                     early_stopping_round= bp_rs['early_stopping_round'], learning_rate= bp_rs['learning_rate'],\n",
        "                     categorical_feature= categorical_col_num)\n",
        "\n",
        "'''\n",
        "validate\n",
        "\n",
        "'''\n",
        "input_train, input_val, target_train, target_val = train_test_split(train_input, train_target, test_size=0.2, random_state=42)\n",
        "\n",
        "eval_set = [(input_val, target_val)]\n",
        "\n",
        "lgbm_best.fit(input_train, target_train, eval_set=eval_set, early_stopping_rounds=5, verbose=False)\n",
        "\n",
        "scores = lgbm_best.score(input_val, target_val)\n",
        "pred = lgbm_best.predict(input_val)\n",
        "\n",
        "print('SCORE                         : ', scores)\n",
        "print('MSE(mean squeared error)      : ', mean_squared_error(target_val, pred))\n",
        "print('RMSE(root mean squared error) : ', np.sqrt(mean_squared_error(target_val, pred)))\n",
        "print('MAE(mean absolute error)      : ', mean_absolute_error(target_val, pred))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RB2oq3IUGDFS",
        "outputId": "51a28867-c12f-4595-c005-c7e673de9535"
      },
      "execution_count": 138,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "--------------------- lgbm with random search ---------------------\n",
            "[LightGBM] [Warning] feature_fraction is set=0.561500722669917, colsample_bytree=1.0 will be ignored. Current value: feature_fraction=0.561500722669917\n",
            "[LightGBM] [Warning] bagging_fraction is set=0.5070416130849544, subsample=1.0 will be ignored. Current value: bagging_fraction=0.5070416130849544\n",
            "SCORE                         :  0.9971728361572244\n",
            "MSE(mean squeared error)      :  0.00605101542884672\n",
            "RMSE(root mean squared error) :  0.07778827308050179\n",
            "MAE(mean absolute error)      :  0.05230536377707307\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Submission\n",
        "- predict value with test_input\n",
        "- convert to submission.csv"
      ],
      "metadata": {
        "id": "FC9Fbh1egNbn"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "'''\n",
        "predict\n",
        "'''\n",
        "predict_tmp_data = lgbm_best.predict(test_input)\n",
        "\n",
        "'''\n",
        "replace value \n",
        "'''\n",
        "predict_data= []\n",
        "for e in predict_tmp_data :\n",
        "    predict_data.append(np.expm1(e))\n",
        "\n",
        "'''\n",
        "array to datagram\n",
        "'''\n",
        "submission = pd.DataFrame()\n",
        "submission['id'] = new_test_df['id']\n",
        "submission['value'] = predict_data\n",
        "submission.to_csv('submission.csv', index = False)\n",
        "\n",
        "print(submission)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oJuTqgjQdIn9",
        "outputId": "bda29544-80ae-4d04-822d-6a5c00cbccbd"
      },
      "execution_count": 139,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "         id         value\n",
            "0         1  3.527839e+07\n",
            "1         2  7.633133e+07\n",
            "2         4  7.443450e+07\n",
            "3         5  6.837572e+07\n",
            "4         6  5.360389e+07\n",
            "...     ...           ...\n",
            "3823  16924  5.546033e+04\n",
            "3824  16929  4.842492e+04\n",
            "3825  16932  5.603228e+04\n",
            "3826  16937  5.027882e+04\n",
            "3827  16943  4.879204e+04\n",
            "\n",
            "[3828 rows x 2 columns]\n"
          ]
        }
      ]
    }
  ]
}