06_tabular_data_gpu.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fZAgXt0Biex6"
      },
      "source": [
        "[![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/biodatlab/deep-learning-skooldio/blob/master/06_tabular_data_gpu.ipynb)\n",
        "\n",
        "\n",
        "## **Predicting second hand car price from Rodkaidee: Regression**\n",
        "\n",
        "This notebook contains the following steps for predicting second hand car price from [Rodkaidee website](https://rod.kaidee.com/)\n",
        "\n",
        "- Import libraries\n",
        "- Prepare data: Categorical and numerical features\n",
        "- Train linear regression model\n",
        "- Train one-layer neural network model for regression (equivalent to linear regression)\n",
        "- Train multi-layer neural network model for regression"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "zxcAGilIZI0C"
      },
      "outputs": [],
      "source": [
        "import os\n",
        "import os.path as op\n",
        "\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "from tqdm.auto import tqdm\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "\n",
        "from sklearn.linear_model import LinearRegression\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.preprocessing import OneHotEncoder, MinMaxScaler\n",
        "from sklearn.metrics import mean_squared_error\n",
        "import joblib"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1aoH2TxDZI0E"
      },
      "source": [
        "## **Data Preparation**\n",
        "\n",
        "- Download data from [`github.com/biodatlab/deep-learning-skooldio`](https://github.com/biodatlab/deep-learning-skooldio)\n",
        "- Clean and split data"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NUR3qwLXieyE",
        "outputId": "ad20996b-0a20-47da-fa58-7a17094194a1"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Cloning into 'deep-learning-skooldio'...\n",
            "remote: Enumerating objects: 1889, done.\u001b[K\n",
            "remote: Counting objects: 100% (183/183), done.\u001b[K\n",
            "remote: Compressing objects: 100% (94/94), done.\u001b[K\n",
            "remote: Total 1889 (delta 84), reused 166 (delta 84), pack-reused 1706\u001b[K\n",
            "Receiving objects: 100% (1889/1889), 14.42 MiB | 7.92 MiB/s, done.\n",
            "Resolving deltas: 100% (95/95), done.\n"
          ]
        }
      ],
      "source": [
        "!git clone https://github.com/biodatlab/deep-learning-skooldio"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "wCZuVhV7ZI0F",
        "outputId": "37553ed2-0c02-4077-f735-60ada1ac9b57"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "        Brand    Model  Engine    Segment  Year Mileage       Province  \\\n",
              "0       Mazda   Mazda2     1.3  B-Segment  2022  15,640  กรุงเทพมหานคร   \n",
              "1      Suzuki  Celerio     1.0    Eco Car  2021   9,545  กรุงเทพมหานคร   \n",
              "2  Mitsubishi   Mirage     1.2    Eco Car  2021  10,900        นนทบุรี   \n",
              "3      Nissan    March     1.2    Eco Car  2021  12,000  กรุงเทพมหานคร   \n",
              "4  Mitsubishi  Attrage     1.2    Eco Car  2021  25,706      กาญจนบุรี   \n",
              "\n",
              "     Price   Color  \n",
              "0  499,000    Gray  \n",
              "1  315,000   White  \n",
              "2  335,000    Gray  \n",
              "3  339,000  Silver  \n",
              "4  350,000    Gray  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-717538c3-8e5b-462a-afc3-eda06097d631\">\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>Brand</th>\n",
              "      <th>Model</th>\n",
              "      <th>Engine</th>\n",
              "      <th>Segment</th>\n",
              "      <th>Year</th>\n",
              "      <th>Mileage</th>\n",
              "      <th>Province</th>\n",
              "      <th>Price</th>\n",
              "      <th>Color</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Mazda</td>\n",
              "      <td>Mazda2</td>\n",
              "      <td>1.3</td>\n",
              "      <td>B-Segment</td>\n",
              "      <td>2022</td>\n",
              "      <td>15,640</td>\n",
              "      <td>กรุงเทพมหานคร</td>\n",
              "      <td>499,000</td>\n",
              "      <td>Gray</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Suzuki</td>\n",
              "      <td>Celerio</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Eco Car</td>\n",
              "      <td>2021</td>\n",
              "      <td>9,545</td>\n",
              "      <td>กรุงเทพมหานคร</td>\n",
              "      <td>315,000</td>\n",
              "      <td>White</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Mitsubishi</td>\n",
              "      <td>Mirage</td>\n",
              "      <td>1.2</td>\n",
              "      <td>Eco Car</td>\n",
              "      <td>2021</td>\n",
              "      <td>10,900</td>\n",
              "      <td>นนทบุรี</td>\n",
              "      <td>335,000</td>\n",
              "      <td>Gray</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Nissan</td>\n",
              "      <td>March</td>\n",
              "      <td>1.2</td>\n",
              "      <td>Eco Car</td>\n",
              "      <td>2021</td>\n",
              "      <td>12,000</td>\n",
              "      <td>กรุงเทพมหานคร</td>\n",
              "      <td>339,000</td>\n",
              "      <td>Silver</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Mitsubishi</td>\n",
              "      <td>Attrage</td>\n",
              "      <td>1.2</td>\n",
              "      <td>Eco Car</td>\n",
              "      <td>2021</td>\n",
              "      <td>25,706</td>\n",
              "      <td>กาญจนบุรี</td>\n",
              "      <td>350,000</td>\n",
              "      <td>Gray</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-717538c3-8e5b-462a-afc3-eda06097d631')\"\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-717538c3-8e5b-462a-afc3-eda06097d631 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-717538c3-8e5b-462a-afc3-eda06097d631');\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": 3
        }
      ],
      "source": [
        "df = pd.read_csv(op.join(\"deep-learning-skooldio\", \"rodkaidee-dataset\", \"Rodkaidee Data.csv\"))\n",
        "df.head(5)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gYLnuMfdieyG"
      },
      "source": [
        "## Clean and split data "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "id": "V-P_xIbMieyG"
      },
      "outputs": [],
      "source": [
        "import os.path as op\n",
        "\n",
        "df[\"Price\"] = df[\"Price\"].apply(lambda x: float(str(x).replace(\",\", \"\")))\n",
        "df[\"Mileage\"] = df[\"Mileage\"].apply(lambda x: float(str(x).replace(\",\", \"\")))\n",
        "df_train, df_val = train_test_split(df, test_size=0.2, random_state=42)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JFqqEkUUZI0G"
      },
      "source": [
        "## Apply One-hot Encoding to categorical values\n",
        "\n",
        "For example, if a column has three unique values 'A', 'B', and 'C', rows that contain 'C' will be transformed to [0, 0, 1]."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "u2025L16ieyI"
      },
      "outputs": [],
      "source": [
        "CAT_COLUMNS = [\"Brand\", \"Model\", \"Engine\", \"Segment\", \"Province\", \"Color\"]\n",
        "ohe = OneHotEncoder(handle_unknown='ignore', sparse=False)\n",
        "\n",
        "cat_train_array = ohe.fit_transform(df_train[CAT_COLUMNS])\n",
        "cat_val_array = ohe.transform(df_val[CAT_COLUMNS])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TWBWpadpZI0I"
      },
      "source": [
        "## Normalize continuous values\n",
        "\n",
        "We will use `MinMaxScaler` to normalize continuous value.\n",
        "For example, column X has values from 0 to 1000000, so we need to normalize them to 0 to 1.\n",
        "Alternatively `StandardScaler` is also an option."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2zc0NpePZI0I",
        "outputId": "bd483a82-d5c5-4acc-a7f9-87a42ac29d92"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "MinMaxScaler()"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ],
      "source": [
        "year_scaler = MinMaxScaler()\n",
        "mileage_scaler = MinMaxScaler()\n",
        "price_scaler = MinMaxScaler()\n",
        "\n",
        "year_scaler.fit(df_train[[\"Year\"]])\n",
        "mileage_scaler.fit(df_train[[\"Mileage\"]])\n",
        "price_scaler.fit(df_train[[\"Price\"]])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TezNN7EHieyJ"
      },
      "source": [
        "## Apply one-hot encoding and normalization to create train and teset dataset"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "id": "jCNEWoRBieyK"
      },
      "outputs": [],
      "source": [
        "X_train = np.hstack([\n",
        "    ohe.transform(df_train[CAT_COLUMNS]),\n",
        "    year_scaler.transform(df_train[[\"Year\"]]),\n",
        "    mileage_scaler.transform(df_train[[\"Mileage\"]]),\n",
        "])\n",
        "y_train = price_scaler.transform(df_train[[\"Price\"]])\n",
        "\n",
        "X_val = np.hstack([\n",
        "    ohe.transform(df_val[CAT_COLUMNS]),\n",
        "    year_scaler.transform(df_val[[\"Year\"]]),\n",
        "    mileage_scaler.transform(df_val[[\"Mileage\"]]),\n",
        "])\n",
        "y_val = price_scaler.transform(df_val[[\"Price\"]])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ANIl6W4iieyK",
        "outputId": "d7d77c99-425b-449e-9046-8e02c67f688e"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['trained_models/price_scaler.joblib']"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ],
      "source": [
        "# save one-hot encoder and scaler objects with joblib\n",
        "save_dir = \"trained_models\"\n",
        "os.makedirs(save_dir, exist_ok=True)\n",
        "joblib.dump(ohe, op.join(save_dir, \"one_hot_encoder.joblib\"))\n",
        "joblib.dump(year_scaler, op.join(save_dir, \"year_scaler.joblib\"))\n",
        "joblib.dump(mileage_scaler, op.join(save_dir, \"mileage_scaler.joblib\"))\n",
        "joblib.dump(price_scaler, op.join(save_dir, \"price_scaler.joblib\"))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3wvVjwnkZI0K"
      },
      "source": [
        "## **Train a Neural Network to predict car price**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "id": "cRVX8bSxZI0L"
      },
      "outputs": [],
      "source": [
        "from torch.utils.data import Dataset, DataLoader\n",
        "\n",
        "class CarPriceDataset(Dataset):\n",
        "    def __init__(self, X, y = None):\n",
        "        self.X = X\n",
        "        if y is not None:\n",
        "            self.y = y\n",
        "        else:\n",
        "            self.y = None\n",
        "\n",
        "    def __len__(self):\n",
        "        return len(self.X)\n",
        "\n",
        "    def __getitem__(self, idx):\n",
        "        if self.y is not None:\n",
        "            return self.X[idx], self.y[idx]\n",
        "        else:\n",
        "            return self.X[idx]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "id": "AvHBJxavieyO"
      },
      "outputs": [],
      "source": [
        "train_dataset = CarPriceDataset(X_train, y_train)\n",
        "val_dataset = CarPriceDataset(X_val, y_val)\n",
        "\n",
        "train_dataloader = DataLoader(train_dataset, batch_size=32, shuffle=True)\n",
        "val_dataloader = DataLoader(val_dataset, batch_size=32, shuffle=False)\n",
        "\n",
        "# x_sample, y_sample = next(iter(train_dataloader))\n",
        "# x_sample, y_sample"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "id": "zSvDE-GjZI0L"
      },
      "outputs": [],
      "source": [
        "class CarPriceModel(nn.Module):\n",
        "    def __init__(self, input_size, output_size):\n",
        "        super().__init__()\n",
        "        self.linear = nn.Linear(input_size, output_size)\n",
        "\n",
        "    def forward(self, x):\n",
        "        out = self.linear(x)\n",
        "        return out"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Check if GPU is available and move the model to `cuda` if GPU is available"
      ],
      "metadata": {
        "id": "FOxKAdDl_JQE"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "id": "Er6ql5JE-6FQ"
      },
      "outputs": [],
      "source": [
        "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YXdDQUJXZI0L",
        "outputId": "7890a727-7bb3-4d7f-a6e4-f3bf8a74c3fe"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "138\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "CarPriceModel(\n",
              "  (linear): Linear(in_features=138, out_features=1, bias=True)\n",
              ")"
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ],
      "source": [
        "n_features, n_output = X_train.shape[1], y_train.shape[1]\n",
        "print(n_features)\n",
        "model = CarPriceModel(input_size=n_features, output_size=n_output)\n",
        "model = model.to(device) # move model to specified device\n",
        "model"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "id": "nMMvmJtzieyP"
      },
      "outputs": [],
      "source": [
        "def train(\n",
        "    model,\n",
        "    train_dataloader,\n",
        "    test_dataloader,\n",
        "    optimizer,\n",
        "    loss_fn = nn.MSELoss(reduction='mean'),\n",
        "    num_epochs = 100,\n",
        "    update_frequency = 10,\n",
        "    device = \"cpu\",\n",
        "):\n",
        "    for epoch in tqdm(range(num_epochs)):\n",
        "        # Set the model to training mode.\n",
        "        model.train()\n",
        "        for x, y in train_dataloader:\n",
        "            # move X, y to device\n",
        "            x, y = x.to(device), y.to(device)\n",
        "            # Forward pass\n",
        "            y_pred = model(x.float())\n",
        "            # Calculate the loss.\n",
        "            loss = loss_fn(y_pred, y.float())\n",
        "            # Backward pass.\n",
        "            loss.backward()\n",
        "            # Update the model parameters.\n",
        "            optimizer.step()\n",
        "            # Reset the gradients.\n",
        "            optimizer.zero_grad()\n",
        "\n",
        "        # Set the model to evaluation mode and calculate the test loss at the end of each epoch.\n",
        "        model.eval()\n",
        "        with torch.no_grad():\n",
        "            for x, y in test_dataloader:\n",
        "                x = x.to(device)\n",
        "                y = y.to(device)\n",
        "                # Forward pass.\n",
        "                y_pred = model(x.float())\n",
        "                # Calculate the loss.\n",
        "                loss = loss_fn(y_pred, y.float())\n",
        "\n",
        "            # Print the loss if the epoch is a multiple of 10.\n",
        "            if epoch % update_frequency == 0:\n",
        "                print(f\"Epoch {epoch}: {loss.item()}\")\n",
        "\n",
        "\n",
        "def predict(model, test_dataloader):\n",
        "    model.eval()\n",
        "    y_pred_list = []\n",
        "    y_true_list = []\n",
        "    for x, y in test_dataloader:\n",
        "        y_pred = model(x.float())\n",
        "        prediction = y_pred.detach().numpy()\n",
        "        y = y.numpy()\n",
        "        y_pred_list.extend(prediction)\n",
        "        y_true_list.extend(y)\n",
        "    y_pred_list = np.concatenate(y_pred_list)\n",
        "    y_true_list = np.concatenate(y_true_list)\n",
        "    return y_pred_list, y_true_list"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 257,
          "referenced_widgets": [
            "c1aa5cdf0de547b7aaba392b3d3e8e48",
            "f540f2025bb3430fa085fd4f41015bb8",
            "349418f485f54431bdae9a4445726f65",
            "34996dbe809a4a66bd907c6626d3aee7",
            "0220695ba0f84d34b6db8b091035f345",
            "21e00de5c35b4b44a65d9159a724bac1",
            "b9589ac96f3a4e86a0042f6388e32eed",
            "8de6521f697144d6994626c3c12cb17b",
            "fb910cfb2bf24d60a8b41aa47fc7b698",
            "ca5764747b644cf8acd47a922cd18aea",
            "0d46fdf40ce44398ba369b0b70616693"
          ]
        },
        "id": "KUPnOccyZI0M",
        "outputId": "eaf96cd9-be59-4271-fb88-5ba06c03e508"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "  0%|          | 0/120 [00:00<?, ?it/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "c1aa5cdf0de547b7aaba392b3d3e8e48"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 0: 0.007401853799819946\n",
            "Epoch 10: 0.0008336789906024933\n",
            "Epoch 20: 0.0008057209779508412\n",
            "Epoch 30: 0.001273913774639368\n",
            "Epoch 40: 0.0007082221563905478\n",
            "Epoch 50: 0.001047421363182366\n",
            "Epoch 60: 0.0007900574710220098\n",
            "Epoch 70: 0.0006823980947956443\n",
            "Epoch 80: 0.000877499463967979\n",
            "Epoch 90: 0.0007753869867883623\n",
            "Epoch 100: 0.0008727338863536716\n",
            "Epoch 110: 0.0009233366581611335\n"
          ]
        }
      ],
      "source": [
        "# Define the loss function and the optimizer.\n",
        "learning_rate = 1e-3\n",
        "loss_fn = nn.MSELoss(reduction=\"mean\")\n",
        "optimizer = torch.optim.Adam(model.parameters(), lr=learning_rate)\n",
        "\n",
        "# Train the model.\n",
        "num_epochs = 120\n",
        "update_frequency = 10\n",
        "\n",
        "train(\n",
        "    model = model,\n",
        "    train_dataloader = train_dataloader,\n",
        "    test_dataloader = val_dataloader,\n",
        "    optimizer = optimizer,\n",
        "    loss_fn = loss_fn,\n",
        "    num_epochs = num_epochs,\n",
        "    update_frequency = update_frequency,\n",
        "    device = device,\n",
        ")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lIYyFW3QZI0M"
      },
      "source": [
        "### Evaluate the model on the test set with Mean Squared Error"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "id": "pbGtQYsRZI0M"
      },
      "outputs": [],
      "source": [
        "# Get the predictions for the test set\n",
        "y_pred, y_true = predict(model.cpu(), val_dataloader)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BmcaecNgZI0M",
        "outputId": "f993edbf-615d-4e22-ded6-04b3ac0cdcba"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.0031842023004687962"
            ]
          },
          "metadata": {},
          "execution_count": 19
        }
      ],
      "source": [
        "mean_squared_error(y_true, y_pred)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "S6AvZKUvZI0N",
        "outputId": "ac85c8ef-8b24-493a-d354-725eacde9a87"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "# Plot a perfect prediction line.\n",
        "plt.plot([0, 1], [0, 1],  \"--\", color=\"black\",)\n",
        "# Plot the predictions vs the actual values.\n",
        "plt.scatter(y_pred, y_true, alpha=0.2)\n",
        "\n",
        "plt.xlabel(\"Predictions\")\n",
        "plt.ylabel(\"Actual\")\n",
        "plt.title(\"Predictions vs Actual\")\n",
        "plt.xlim(0, 1)\n",
        "plt.ylim(0, 1)\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "id": "5GDQ78C-ieyR"
      },
      "outputs": [],
      "source": [
        "torch.save(model.state_dict(), \"trained_models/carprice_model_mse_0003.pth\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "msNYb_DBieyR"
      },
      "source": [
        "## **Train a Neural Network with multiple layers (MLP)**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {
        "id": "pKqBHUT-ieyR"
      },
      "outputs": [],
      "source": [
        "class CarPriceTwoLayerModel(nn.Module):\n",
        "    def __init__(self, input_size, output_size, intermediate_dim = 10):\n",
        "        super().__init__()\n",
        "        self.linear1 = nn.Linear(input_size, intermediate_dim)\n",
        "        self.linear2 = nn.Linear(intermediate_dim, output_size)\n",
        "        self.relu = nn.ReLU()\n",
        "\n",
        "    def forward(self, x):\n",
        "        x = self.linear1(x)\n",
        "        x = self.relu(x)\n",
        "        x = self.linear2(x)\n",
        "        return x\n",
        "\n",
        "model_two_layers = CarPriceTwoLayerModel(input_size=n_features, output_size=n_output)\n",
        "model_two_layers = model_two_layers.to(device) # move to device"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 257,
          "referenced_widgets": [
            "b4ea85b390154d49bf249dfd3e4b5dd2",
            "a9a3f4cf6ffc4dc787e691ee0874f3a5",
            "b498fcb81c4b4408861bdabfb640aa2a",
            "927046f13ba0412c9a64cee2ffae9202",
            "a33a1e0e935f473bafb43066123aacfa",
            "9269308d64bd4b728ae250920e5ea133",
            "aef3f17690104c2b88b44e59e4191d81",
            "a15a698660a240a38d7250ff84d552f6",
            "8ee78b6c120d481cbe71ccb310d54512",
            "24d0ccf9eb504cc7919f06409cd783d8",
            "d500eaaf27c04a279c395db5405cdb7e"
          ]
        },
        "id": "X1S1mS7WieyR",
        "outputId": "0f406416-5c3d-4b42-a791-d48d1aa30566"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "  0%|          | 0/120 [00:00<?, ?it/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "b4ea85b390154d49bf249dfd3e4b5dd2"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 0: 0.005613044835627079\n",
            "Epoch 10: 0.0002780476934276521\n",
            "Epoch 20: 0.00025078855105675757\n",
            "Epoch 30: 0.00023882044479250908\n",
            "Epoch 40: 0.0004488053382374346\n",
            "Epoch 50: 0.0006681648665107787\n",
            "Epoch 60: 0.0007082157535478473\n",
            "Epoch 70: 0.0005855925264768302\n",
            "Epoch 80: 0.0006355448858812451\n",
            "Epoch 90: 0.0006763400742784142\n",
            "Epoch 100: 0.0006260552327148616\n",
            "Epoch 110: 0.0007233883952721953\n"
          ]
        }
      ],
      "source": [
        "# Define the loss function and the optimizer.\n",
        "learning_rate = 1e-3 # = 0.001\n",
        "loss_fn = nn.MSELoss(reduction=\"mean\")\n",
        "optimizer = torch.optim.Adam(model_two_layers.parameters(), lr=learning_rate)\n",
        "\n",
        "# Train the model.\n",
        "num_epochs = 120\n",
        "update_frequency = 10\n",
        "\n",
        "train(\n",
        "    model=model_two_layers,\n",
        "    train_dataloader=train_dataloader,\n",
        "    test_dataloader=val_dataloader,\n",
        "    optimizer=optimizer,\n",
        "    loss_fn=loss_fn,\n",
        "    num_epochs=num_epochs,\n",
        "    update_frequency=update_frequency,\n",
        "    device=device,\n",
        ")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GxA3ZJDvieyS",
        "outputId": "ec5f07f6-cd2a-4b1b-c4c4-808955d84ef8"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.0017051365135979713"
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ],
      "source": [
        "y_test, y_true_dl = predict(model_two_layers.cpu(), val_dataloader)\n",
        "mse_score = mean_squared_error(y_test, y_true_dl)\n",
        "mse_score"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "2Ys3PJFJieyS",
        "outputId": "6cd78143-19f5-4bca-f94f-0eec37bca6a0"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "# Plot a perfect prediction line.\n",
        "plt.plot([0, 1], [0, 1],  \"--\", color=\"black\",)\n",
        "# Plot the predictions vs the actual values.\n",
        "plt.scatter(y_true_dl, y_test, alpha=0.2)\n",
        "\n",
        "plt.xlabel(\"Predictions value\")\n",
        "plt.ylabel(\"Actual value\")\n",
        "plt.title(\"Predictions vs Actual\")\n",
        "plt.xlim(0, 1)\n",
        "plt.ylim(0, 1)\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 26,
      "metadata": {
        "id": "t2X8VCqDieyT"
      },
      "outputs": [],
      "source": [
        "torch.save(model_two_layers.state_dict(), f\"trained_models/carprice_two_layer_model_mse_00018.pth\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NgiMit2wZI0N"
      },
      "source": [
        "## **Use the models to predict on new data**\n",
        "\n",
        "- Create a model\n",
        "- Load one-hot-encoder and scaler\n",
        "- Transform dataframe\n",
        "- Predict"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 27,
      "metadata": {
        "id": "alui9M_cieyT"
      },
      "outputs": [],
      "source": [
        "# load data (similar to previous code)\n",
        "\n",
        "df = pd.read_csv(op.join(\"deep-learning-skooldio\", \"rodkaidee-dataset\", \"Rodkaidee Data.csv\"))\n",
        "df[\"Price\"] = df[\"Price\"].apply(lambda x: float(x.replace(\",\", \"\")))\n",
        "df[\"Mileage\"] = df[\"Mileage\"].apply(lambda x: float(x.replace(\",\", \"\")))\n",
        "\n",
        "_, df_test = train_test_split(df, test_size=0.2, random_state=42)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 28,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hn3rN6SBieyU",
        "outputId": "b8040b72-e9b1-44a9-d8b2-a0ef53ce2203"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<All keys matched successfully>"
            ]
          },
          "metadata": {},
          "execution_count": 28
        }
      ],
      "source": [
        "# create model and load trained weights\n",
        "class CarPriceTwoLayerModel(nn.Module):\n",
        "    def __init__(self, input_size, output_size, intermediate_dim = 10):\n",
        "        super().__init__()\n",
        "        self.linear1 = nn.Linear(input_size, intermediate_dim)\n",
        "        self.linear2 = nn.Linear(intermediate_dim, output_size)\n",
        "        self.relu = nn.ReLU()\n",
        "\n",
        "    def forward(self, x):\n",
        "        x = self.linear1(x)\n",
        "        x = self.relu(x)\n",
        "        x = self.linear2(x)\n",
        "        return x\n",
        "\n",
        "model = CarPriceTwoLayerModel(138, 1)\n",
        "model.load_state_dict(torch.load(\"trained_models/carprice_two_layer_model_mse_00018.pth\"))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 29,
      "metadata": {
        "id": "G8mX3agUieyU"
      },
      "outputs": [],
      "source": [
        "# load one-hot encoder \n",
        "one_hot_encoder_path = \"trained_models/one_hot_encoder.joblib\"\n",
        "year_scaler_path = \"trained_models/year_scaler.joblib\"\n",
        "mileage_scaler_path = \"trained_models/mileage_scaler.joblib\"\n",
        "price_scaler_path = \"trained_models/price_scaler.joblib\"\n",
        "\n",
        "ohe = joblib.load(one_hot_encoder_path)\n",
        "year_scaler = joblib.load(year_scaler_path)\n",
        "mileage_scaler = joblib.load(mileage_scaler_path)\n",
        "price_scaler = joblib.load(price_scaler_path)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 30,
      "metadata": {
        "id": "03oN8M-YieyU"
      },
      "outputs": [],
      "source": [
        "# prepare test dataset\n",
        "CAT_COLUMNS = [\"Brand\", \"Model\", \"Engine\", \"Segment\", \"Province\", \"Color\"]\n",
        "X_test = np.hstack([\n",
        "    ohe.transform(df_val[CAT_COLUMNS]),\n",
        "    year_scaler.transform(df_val[[\"Year\"]]),\n",
        "    mileage_scaler.transform(df_val[[\"Mileage\"]]),\n",
        "])\n",
        "test_dataset = CarPriceDataset(X_test)\n",
        "test_dataloader = DataLoader(test_dataset, batch_size=32, shuffle=False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 31,
      "metadata": {
        "id": "7lxFxVRQieyU"
      },
      "outputs": [],
      "source": [
        "def predict(model, test_dataloader):\n",
        "    model.eval()\n",
        "    y_pred_list = []\n",
        "    for x in test_dataloader:\n",
        "        y_pred = model(x.float())\n",
        "        prediction = y_pred.detach().numpy()\n",
        "        y_pred_list.extend(prediction)\n",
        "    y_pred_list = np.concatenate(y_pred_list)\n",
        "    return y_pred_list"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 32,
      "metadata": {
        "id": "sAD1TXndieyV"
      },
      "outputs": [],
      "source": [
        "y_pred = predict(model, test_dataloader)\n",
        "y_pred_price = price_scaler.inverse_transform(y_pred.reshape(-1, 1))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 33,
      "metadata": {
        "id": "_n-SMmE4ieyV"
      },
      "outputs": [],
      "source": [
        "df_test[\"Pred_Price\"] = y_pred_price"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 34,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 363
        },
        "id": "zHcoSFbwieyV",
        "outputId": "16ea1f93-8b95-4d76-cc03-8667567131cd"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       Brand   Model  Engine    Segment  Year   Mileage       Province  \\\n",
              "2759  Toyota    Vios     1.5  B-Segment  2015   62639.0  กรุงเทพมหานคร   \n",
              "2846  Toyota   Altis     1.6  C-Segment  2015   65615.0          สงขลา   \n",
              "6682   Honda   Civic     2.0  C-Segment  2007  203000.0      สิงห์บุรี   \n",
              "3262  Toyota   Yaris     1.2    Eco Car  2014  124007.0        ราชบุรี   \n",
              "486    Mazda  Mazda2     1.3  B-Segment  2019   51957.0  กรุงเทพมหานคร   \n",
              "3616  Nissan  Sylphy     1.6  C-Segment  2013  120000.0        นนทบุรี   \n",
              "3590  Nissan  Almera     1.2    Eco Car  2013  130000.0      เชียงใหม่   \n",
              "3246  Toyota    Vios     1.5  B-Segment  2014   80000.0      เชียงใหม่   \n",
              "6503  Nissan   Tiida     1.6  C-Segment  2007  140000.0        นนทบุรี   \n",
              "7209   Honda    City     1.5  B-Segment  2003  166508.0  นครศรีธรรมราช   \n",
              "\n",
              "         Price   Color     Pred_Price  \n",
              "2759  338000.0    Gray  346732.000000  \n",
              "2846  385000.0  Silver  420232.718750  \n",
              "6682  295000.0    Gray  307101.750000  \n",
              "3262  339000.0   White  339311.625000  \n",
              "486   439000.0    Blue  415036.812500  \n",
              "3616  219000.0   Black  304913.125000  \n",
              "3590  199000.0   Black  236784.375000  \n",
              "3246  329000.0   Black  345957.250000  \n",
              "6503  128000.0   Black  150061.468750  \n",
              "7209  129000.0    Gold  117058.953125  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-8c04fea9-c5ba-46f6-97ad-7bb390e3a19d\">\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>Brand</th>\n",
              "      <th>Model</th>\n",
              "      <th>Engine</th>\n",
              "      <th>Segment</th>\n",
              "      <th>Year</th>\n",
              "      <th>Mileage</th>\n",
              "      <th>Province</th>\n",
              "      <th>Price</th>\n",
              "      <th>Color</th>\n",
              "      <th>Pred_Price</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>2759</th>\n",
              "      <td>Toyota</td>\n",
              "      <td>Vios</td>\n",
              "      <td>1.5</td>\n",
              "      <td>B-Segment</td>\n",
              "      <td>2015</td>\n",
              "      <td>62639.0</td>\n",
              "      <td>กรุงเทพมหานคร</td>\n",
              "      <td>338000.0</td>\n",
              "      <td>Gray</td>\n",
              "      <td>346732.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2846</th>\n",
              "      <td>Toyota</td>\n",
              "      <td>Altis</td>\n",
              "      <td>1.6</td>\n",
              "      <td>C-Segment</td>\n",
              "      <td>2015</td>\n",
              "      <td>65615.0</td>\n",
              "      <td>สงขลา</td>\n",
              "      <td>385000.0</td>\n",
              "      <td>Silver</td>\n",
              "      <td>420232.718750</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6682</th>\n",
              "      <td>Honda</td>\n",
              "      <td>Civic</td>\n",
              "      <td>2.0</td>\n",
              "      <td>C-Segment</td>\n",
              "      <td>2007</td>\n",
              "      <td>203000.0</td>\n",
              "      <td>สิงห์บุรี</td>\n",
              "      <td>295000.0</td>\n",
              "      <td>Gray</td>\n",
              "      <td>307101.750000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3262</th>\n",
              "      <td>Toyota</td>\n",
              "      <td>Yaris</td>\n",
              "      <td>1.2</td>\n",
              "      <td>Eco Car</td>\n",
              "      <td>2014</td>\n",
              "      <td>124007.0</td>\n",
              "      <td>ราชบุรี</td>\n",
              "      <td>339000.0</td>\n",
              "      <td>White</td>\n",
              "      <td>339311.625000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>486</th>\n",
              "      <td>Mazda</td>\n",
              "      <td>Mazda2</td>\n",
              "      <td>1.3</td>\n",
              "      <td>B-Segment</td>\n",
              "      <td>2019</td>\n",
              "      <td>51957.0</td>\n",
              "      <td>กรุงเทพมหานคร</td>\n",
              "      <td>439000.0</td>\n",
              "      <td>Blue</td>\n",
              "      <td>415036.812500</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3616</th>\n",
              "      <td>Nissan</td>\n",
              "      <td>Sylphy</td>\n",
              "      <td>1.6</td>\n",
              "      <td>C-Segment</td>\n",
              "      <td>2013</td>\n",
              "      <td>120000.0</td>\n",
              "      <td>นนทบุรี</td>\n",
              "      <td>219000.0</td>\n",
              "      <td>Black</td>\n",
              "      <td>304913.125000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3590</th>\n",
              "      <td>Nissan</td>\n",
              "      <td>Almera</td>\n",
              "      <td>1.2</td>\n",
              "      <td>Eco Car</td>\n",
              "      <td>2013</td>\n",
              "      <td>130000.0</td>\n",
              "      <td>เชียงใหม่</td>\n",
              "      <td>199000.0</td>\n",
              "      <td>Black</td>\n",
              "      <td>236784.375000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3246</th>\n",
              "      <td>Toyota</td>\n",
              "      <td>Vios</td>\n",
              "      <td>1.5</td>\n",
              "      <td>B-Segment</td>\n",
              "      <td>2014</td>\n",
              "      <td>80000.0</td>\n",
              "      <td>เชียงใหม่</td>\n",
              "      <td>329000.0</td>\n",
              "      <td>Black</td>\n",
              "      <td>345957.250000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6503</th>\n",
              "      <td>Nissan</td>\n",
              "      <td>Tiida</td>\n",
              "      <td>1.6</td>\n",
              "      <td>C-Segment</td>\n",
              "      <td>2007</td>\n",
              "      <td>140000.0</td>\n",
              "      <td>นนทบุรี</td>\n",
              "      <td>128000.0</td>\n",
              "      <td>Black</td>\n",
              "      <td>150061.468750</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7209</th>\n",
              "      <td>Honda</td>\n",
              "      <td>City</td>\n",
              "      <td>1.5</td>\n",
              "      <td>B-Segment</td>\n",
              "      <td>2003</td>\n",
              "      <td>166508.0</td>\n",
              "      <td>นครศรีธรรมราช</td>\n",
              "      <td>129000.0</td>\n",
              "      <td>Gold</td>\n",
              "      <td>117058.953125</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-8c04fea9-c5ba-46f6-97ad-7bb390e3a19d')\"\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-8c04fea9-c5ba-46f6-97ad-7bb390e3a19d 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-8c04fea9-c5ba-46f6-97ad-7bb390e3a19d');\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": 34
        }
      ],
      "source": [
        "df_test.head(10)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 35,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "kuSPPF0PieyV",
        "outputId": "3b541a9c-9dc1-46a8-cb91-e5088fd05d15"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "plt.plot([0, 1_000_000], [0, 1_000_000], color=\"black\", alpha=0.5, linestyle=\"--\")\n",
        "\n",
        "plt.scatter(df_test[\"Pred_Price\"].values, df_test[\"Price\"], alpha=0.2)\n",
        "plt.xlabel(\"Predictions (Baht)\")\n",
        "plt.ylabel(\"Actual (Baht)\")\n",
        "plt.title(\"Predictions vs Actual\")\n",
        "\n",
        "plt.xlim(0, 1_000_000)\n",
        "plt.ylim(0, 1_000_000)\n",
        "plt.show()"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "provenance": []
    },
    "gpuClass": "standard",
    "kernelspec": {
      "display_name": "Python 3.7.5 ('ml-agents')",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.7.5"
    },
    "vscode": {
      "interpreter": {
        "hash": "94499b8ebe41d82c56ca741edd2212fa3535c1025402838f7d8b62fb6e0f95b1"
      }
    },
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "c1aa5cdf0de547b7aaba392b3d3e8e48": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_f540f2025bb3430fa085fd4f41015bb8",
              "IPY_MODEL_349418f485f54431bdae9a4445726f65",
              "IPY_MODEL_34996dbe809a4a66bd907c6626d3aee7"
            ],
            "layout": "IPY_MODEL_0220695ba0f84d34b6db8b091035f345"
          }
        },
        "f540f2025bb3430fa085fd4f41015bb8": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_21e00de5c35b4b44a65d9159a724bac1",
            "placeholder": "​",
            "style": "IPY_MODEL_b9589ac96f3a4e86a0042f6388e32eed",
            "value": "100%"
          }
        },
        "349418f485f54431bdae9a4445726f65": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_8de6521f697144d6994626c3c12cb17b",
            "max": 120,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_fb910cfb2bf24d60a8b41aa47fc7b698",
            "value": 120
          }
        },
        "34996dbe809a4a66bd907c6626d3aee7": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_ca5764747b644cf8acd47a922cd18aea",
            "placeholder": "​",
            "style": "IPY_MODEL_0d46fdf40ce44398ba369b0b70616693",
            "value": " 120/120 [00:34&lt;00:00,  3.44it/s]"
          }
        },
        "0220695ba0f84d34b6db8b091035f345": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "21e00de5c35b4b44a65d9159a724bac1": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "b9589ac96f3a4e86a0042f6388e32eed": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "8de6521f697144d6994626c3c12cb17b": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "fb910cfb2bf24d60a8b41aa47fc7b698": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "ca5764747b644cf8acd47a922cd18aea": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "0d46fdf40ce44398ba369b0b70616693": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "b4ea85b390154d49bf249dfd3e4b5dd2": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_a9a3f4cf6ffc4dc787e691ee0874f3a5",
              "IPY_MODEL_b498fcb81c4b4408861bdabfb640aa2a",
              "IPY_MODEL_927046f13ba0412c9a64cee2ffae9202"
            ],
            "layout": "IPY_MODEL_a33a1e0e935f473bafb43066123aacfa"
          }
        },
        "a9a3f4cf6ffc4dc787e691ee0874f3a5": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_9269308d64bd4b728ae250920e5ea133",
            "placeholder": "​",
            "style": "IPY_MODEL_aef3f17690104c2b88b44e59e4191d81",
            "value": "100%"
          }
        },
        "b498fcb81c4b4408861bdabfb640aa2a": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_a15a698660a240a38d7250ff84d552f6",
            "max": 120,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_8ee78b6c120d481cbe71ccb310d54512",
            "value": 120
          }
        },
        "927046f13ba0412c9a64cee2ffae9202": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_24d0ccf9eb504cc7919f06409cd783d8",
            "placeholder": "​",
            "style": "IPY_MODEL_d500eaaf27c04a279c395db5405cdb7e",
            "value": " 120/120 [00:38&lt;00:00,  3.38it/s]"
          }
        },
        "a33a1e0e935f473bafb43066123aacfa": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "9269308d64bd4b728ae250920e5ea133": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "aef3f17690104c2b88b44e59e4191d81": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "a15a698660a240a38d7250ff84d552f6": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "8ee78b6c120d481cbe71ccb310d54512": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "24d0ccf9eb504cc7919f06409cd783d8": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "d500eaaf27c04a279c395db5405cdb7e": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        }
      }
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}