Tugas_5_convolutional_neural_network.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "convolutional_neural_network.ipynb",
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/MrezaFd/Tugas-Besar-ML-059-055-/blob/main/Tugas_5_convolutional_neural_network.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pg3l1ejwCF9P"
      },
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_yhI15_dBJI1",
        "outputId": "15147091-f43e-46fb-c8e8-ca8925e603a3",
        "colab": {
          "resources": {
            "http://localhost:8080/nbextensions/google.colab/files.js": {
              "data": "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",
              "ok": true,
              "headers": [
                [
                  "content-type",
                  "application/javascript"
                ]
              ],
              "status": 200,
              "status_text": ""
            }
          },
          "base_uri": "https://localhost:8080/",
          "height": 94
        }
      },
      "source": [
        "from google.colab import files #import lib\n",
        "files.upload()"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-1725fa53-5038-4296-84bd-79f568d56195\" name=\"files[]\" multiple disabled\n",
              "        style=\"border:none\" />\n",
              "     <output id=\"result-1725fa53-5038-4296-84bd-79f568d56195\">\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 src=\"/nbextensions/google.colab/files.js\"></script> "
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Saving kaggle.json to kaggle.json\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'kaggle.json': b'{\"username\":\"hexazor\",\"key\":\"663a14597732196faa9d3257dd4ff4df\"}'}"
            ]
          },
          "metadata": {},
          "execution_count": 2
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T-PttF09A1jP"
      },
      "source": [
        "!mkdir -p ~/.kaggle #membuat directory bernama kaggle \n",
        "!cp kaggle.json ~/.kaggle/ #copy file kaggle.json upload tadi ke directory baru\n",
        "!chmod 600 ~/.kaggle/kaggle.json #berikan izin untuk file kaggle.jsonv"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OvexdoKwBZ5V",
        "outputId": "4fb07eea-ae7e-4a0a-bb58-e310068d2b17",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "#Unduh dataset\n",
        "!kaggle datasets download -d fanconic/skin-cancer-malignant-vs-benign"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading skin-cancer-malignant-vs-benign.zip to /content\n",
            " 97% 315M/325M [00:02<00:00, 141MB/s]\n",
            "100% 325M/325M [00:02<00:00, 129MB/s]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4rtnNrFsEDIf",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "552908e1-8278-4589-b211-dade882033b7"
      },
      "source": [
        "from zipfile import ZipFile\n",
        "file_name = \"skin-cancer-malignant-vs-benign.zip\"\n",
        "\n",
        "with ZipFile(file_name,'r') as zip :\n",
        "  zip.extractall()\n",
        "  print('Berhasil')"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Berhasil\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "31Pr3YSaDHnn"
      },
      "source": [
        "!mkdir a"
      ],
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "u6Wd9eg8DOao"
      },
      "source": [
        "!mv -t a test train"
      ],
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "y9L_1kqnEEjc"
      },
      "source": [
        "# Definisikan path untuk data train dan data validation\n",
        "import os\n",
        "base_dir = '/content/a'\n",
        "\n",
        "\n",
        "train_dir = os.path.join(base_dir, 'train')\n",
        "validation_dir = os.path.join(base_dir, 'test')"
      ],
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9gc0BCSREF0M"
      },
      "source": [
        "import cv2\n",
        "import numpy as np\n",
        "\n",
        "# Gather data train\n",
        "train_data = []\n",
        "train_label = []\n",
        "for r, d, f in os.walk(train_dir):\n",
        "    for file in f:\n",
        "        if \".jpg\" in file:\n",
        "            imagePath = os.path.join(r, file)\n",
        "            image = cv2.imread(imagePath)\n",
        "            image = cv2.resize(image, (150,150))\n",
        "            train_data.append(image)\n",
        "            label = imagePath.split(os.path.sep)[-2]\n",
        "            train_label.append(label)\n",
        "\n",
        "train_data = np.array(train_data)\n",
        "train_label = np.array(train_label)"
      ],
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FuoKFhhVEG_O"
      },
      "source": [
        "# Gather data validation\n",
        "val_data = []\n",
        "val_label = []\n",
        "for r, d, f in os.walk(validation_dir):\n",
        "    for file in f:\n",
        "        if \".jpg\" in file:\n",
        "            imagePath = os.path.join(r, file)\n",
        "            image = cv2.imread(imagePath)\n",
        "            image = cv2.resize(image, (150,150))\n",
        "            val_data.append(image)\n",
        "            label = imagePath.split(os.path.sep)[-2]\n",
        "            val_label.append(label)\n",
        "\n",
        "val_data = np.array(val_data)\n",
        "val_label = np.array(val_label)"
      ],
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kiXbwrN3ELqv"
      },
      "source": [
        "<h3>Data Preprocessing</h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "m5VhnvtrEI1I",
        "outputId": "5fc65261-2a96-4fbc-96b1-55e9cc2dcfd6"
      },
      "source": [
        "# Tampilkan shape dari data train dan data validation\n",
        "print(\"Train Data = \", train_data.shape)\n",
        "print(\"Train Label = \", train_label.shape)\n",
        "print(\"Validation Data = \", val_data.shape)\n",
        "print(\"Validation Label = \", val_label.shape)"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Data =  (2637, 150, 150, 3)\n",
            "Train Label =  (2637,)\n",
            "Validation Data =  (660, 150, 150, 3)\n",
            "Validation Label =  (660,)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "q1LVp6dCENne",
        "outputId": "0ce8c7bc-25df-4ece-d0ab-5cf40be84d2a"
      },
      "source": [
        "# Normalisasi dataset\n",
        "print(\"Data sebelum di-normalisasi \", train_data[0][0][0])\n",
        "\n",
        "x_train = train_data.astype('float32') / 255.0\n",
        "x_val = val_data.astype('float32') / 255.0\n",
        "print(\"Data setelah di-normalisasi \", x_train[0][0][0])"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Data sebelum di-normalisasi  [ 88 101 145]\n",
            "Data setelah di-normalisasi  [0.34509805 0.39607844 0.5686275 ]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tc7FivC6EPhm",
        "outputId": "5d9c78dc-4520-44b7-d253-cfd839623625"
      },
      "source": [
        "# Transformasi label encoder\n",
        "from sklearn.preprocessing import LabelEncoder\n",
        "\n",
        "print(\"Label sebelum di-encoder \", train_label[995:2005])\n",
        "\n",
        "lb = LabelEncoder()\n",
        "y_train = lb.fit_transform(train_label)\n",
        "y_val = lb.fit_transform(val_label)\n",
        "\n",
        "print(\"Label setelah di-encoder \", y_train[995:2005])"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Label sebelum di-encoder  ['malignant' 'malignant' 'malignant' ... 'benign' 'benign' 'benign']\n",
            "Label setelah di-encoder  [1 1 1 ... 0 0 0]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4jVWbOhOEzQJ"
      },
      "source": [
        "<h3>Definisikan Model</h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "usLiFiwAEvVb"
      },
      "source": [
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import InputLayer, Dense, Conv2D, MaxPool2D, Flatten"
      ],
      "execution_count": 21,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rDwzxacpE1ei"
      },
      "source": [
        "# Feature Extraction Layer\n",
        "model = Sequential()\n",
        "\n",
        "model.add(InputLayer(input_shape=[150,150,3]))\n",
        "model.add(Conv2D(filters=16, kernel_size=3, strides=1, padding='same', activation='relu'))\n",
        "model.add(MaxPool2D(pool_size=2, padding='same'))\n",
        "model.add(Conv2D(filters=32, kernel_size=3, strides=1, padding='same', activation='relu'))\n",
        "model.add(MaxPool2D(pool_size=2, padding='same'))\n",
        "model.add(Conv2D(filters=64, kernel_size=3, strides=1, padding='same', activation='relu'))\n",
        "model.add(MaxPool2D(pool_size=2, padding='same'))\n",
        "model.add(Flatten())"
      ],
      "execution_count": 22,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Sud1AUF0E2k8"
      },
      "source": [
        "# Fully Connected Layer\n",
        "model.add(Dense(128, activation='relu'))\n",
        "model.add(Dense(1, activation='sigmoid'))"
      ],
      "execution_count": 23,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XW-KE9ePE466",
        "outputId": "a13caa65-b5b7-40aa-fe3d-2615835f3aa8"
      },
      "source": [
        "# Print model summary\n",
        "print(model.summary())"
      ],
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv2d (Conv2D)             (None, 150, 150, 16)      448       \n",
            "                                                                 \n",
            " max_pooling2d (MaxPooling2D  (None, 75, 75, 16)       0         \n",
            " )                                                               \n",
            "                                                                 \n",
            " conv2d_1 (Conv2D)           (None, 75, 75, 32)        4640      \n",
            "                                                                 \n",
            " max_pooling2d_1 (MaxPooling  (None, 38, 38, 32)       0         \n",
            " 2D)                                                             \n",
            "                                                                 \n",
            " conv2d_2 (Conv2D)           (None, 38, 38, 64)        18496     \n",
            "                                                                 \n",
            " max_pooling2d_2 (MaxPooling  (None, 19, 19, 64)       0         \n",
            " 2D)                                                             \n",
            "                                                                 \n",
            " flatten (Flatten)           (None, 23104)             0         \n",
            "                                                                 \n",
            " dense (Dense)               (None, 128)               2957440   \n",
            "                                                                 \n",
            " dense_1 (Dense)             (None, 1)                 129       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 2,981,153\n",
            "Trainable params: 2,981,153\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "None\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Tais64D7E5Zo",
        "outputId": "b8368bdf-e475-4ac5-e529-40f650f13a63"
      },
      "source": [
        "# Compile model\n",
        "from tensorflow.keras.optimizers import Adam\n",
        "\n",
        "model.compile(optimizer=Adam(lr=0.001), \n",
        "              loss='binary_crossentropy', \n",
        "              metrics=['acc'])"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/keras/optimizer_v2/adam.py:105: UserWarning: The `lr` argument is deprecated, use `learning_rate` instead.\n",
            "  super(Adam, self).__init__(name, **kwargs)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NTNBH19iE8kt"
      },
      "source": [
        "<h3>Training Model</h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "p4Vr8dyvE6pM",
        "outputId": "0122186a-d789-4914-fe82-733e502261c3"
      },
      "source": [
        "H = model.fit(x_train, y_train, batch_size=20, epochs=25, validation_data=(x_val, y_val))"
      ],
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/25\n",
            "132/132 [==============================] - 4s 29ms/step - loss: 0.1746 - acc: 0.9272 - val_loss: 0.3752 - val_acc: 0.8394\n",
            "Epoch 2/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 0.1888 - acc: 0.9196 - val_loss: 0.3938 - val_acc: 0.8470\n",
            "Epoch 3/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 0.1611 - acc: 0.9348 - val_loss: 0.5139 - val_acc: 0.8318\n",
            "Epoch 4/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 0.1541 - acc: 0.9405 - val_loss: 0.4447 - val_acc: 0.8333\n",
            "Epoch 5/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 0.1283 - acc: 0.9503 - val_loss: 0.4601 - val_acc: 0.8500\n",
            "Epoch 6/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 0.1065 - acc: 0.9606 - val_loss: 0.5811 - val_acc: 0.8273\n",
            "Epoch 7/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 0.1040 - acc: 0.9598 - val_loss: 0.6381 - val_acc: 0.8227\n",
            "Epoch 8/25\n",
            "132/132 [==============================] - 3s 23ms/step - loss: 0.0890 - acc: 0.9655 - val_loss: 0.5621 - val_acc: 0.8424\n",
            "Epoch 9/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 0.0648 - acc: 0.9795 - val_loss: 0.5541 - val_acc: 0.8333\n",
            "Epoch 10/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 0.1123 - acc: 0.9602 - val_loss: 0.5206 - val_acc: 0.8424\n",
            "Epoch 11/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 0.0406 - acc: 0.9867 - val_loss: 0.5749 - val_acc: 0.8424\n",
            "Epoch 12/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 0.0303 - acc: 0.9898 - val_loss: 0.6938 - val_acc: 0.8394\n",
            "Epoch 13/25\n",
            "132/132 [==============================] - 3s 23ms/step - loss: 0.0377 - acc: 0.9863 - val_loss: 0.7018 - val_acc: 0.8576\n",
            "Epoch 14/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 0.0263 - acc: 0.9939 - val_loss: 0.7100 - val_acc: 0.8591\n",
            "Epoch 15/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 0.0275 - acc: 0.9924 - val_loss: 0.7683 - val_acc: 0.8364\n",
            "Epoch 16/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 0.0412 - acc: 0.9818 - val_loss: 0.9597 - val_acc: 0.8121\n",
            "Epoch 17/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 0.0414 - acc: 0.9829 - val_loss: 0.8799 - val_acc: 0.8333\n",
            "Epoch 18/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 0.0120 - acc: 0.9962 - val_loss: 0.8512 - val_acc: 0.8455\n",
            "Epoch 19/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 0.0020 - acc: 1.0000 - val_loss: 0.9562 - val_acc: 0.8424\n",
            "Epoch 20/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 0.0016 - acc: 1.0000 - val_loss: 0.9763 - val_acc: 0.8470\n",
            "Epoch 21/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 6.4493e-04 - acc: 1.0000 - val_loss: 1.0351 - val_acc: 0.8455\n",
            "Epoch 22/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 3.8696e-04 - acc: 1.0000 - val_loss: 1.0559 - val_acc: 0.8470\n",
            "Epoch 23/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 3.0699e-04 - acc: 1.0000 - val_loss: 1.0737 - val_acc: 0.8455\n",
            "Epoch 24/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 2.5763e-04 - acc: 1.0000 - val_loss: 1.0973 - val_acc: 0.8455\n",
            "Epoch 25/25\n",
            "132/132 [==============================] - 3s 24ms/step - loss: 2.1375e-04 - acc: 1.0000 - val_loss: 1.1079 - val_acc: 0.8455\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HuxjW35RFCZ6"
      },
      "source": [
        "<h3>Evaluasi Model</h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 298
        },
        "id": "1AM3ms3zE_AU",
        "outputId": "2ef96e5e-933e-4749-e40e-d009af713376"
      },
      "source": [
        "%matplotlib inline\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "plt.style.use(\"ggplot\")\n",
        "plt.figure()\n",
        "plt.plot(np.arange(0, 25), H.history[\"loss\"], label=\"train_loss\")\n",
        "plt.plot(np.arange(0, 25), H.history[\"val_loss\"], label=\"val_loss\")\n",
        "plt.title(\"Loss Plot\")\n",
        "plt.xlabel(\"Epoch #\")\n",
        "plt.ylabel(\"Loss\")\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 298
        },
        "id": "q1SJWudcFEDL",
        "outputId": "2d3bc1ea-ddc0-4d4e-c8fb-f43e4166c03e"
      },
      "source": [
        "plt.style.use(\"ggplot\")\n",
        "plt.figure()\n",
        "plt.plot(np.arange(0, 25), H.history[\"acc\"], label=\"train_acc\")\n",
        "plt.plot(np.arange(0, 25), H.history[\"val_acc\"], label=\"val_acc\")\n",
        "plt.title(\"Accuracy Plot\")\n",
        "plt.xlabel(\"Epoch #\")\n",
        "plt.ylabel(\"Acc\")\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "H_8mB3igFG-x",
        "outputId": "f7b8e0c4-6344-4d4b-cdfc-8a88dd4f2723"
      },
      "source": [
        "from sklearn.metrics import classification_report\n",
        "\n",
        "pred = model.predict(x_val)\n",
        "labels = (pred > 0.5).astype(np.int)\n",
        "\n",
        "print(classification_report(y_val, labels))"
      ],
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.87      0.84      0.86       360\n",
            "           1       0.82      0.85      0.83       300\n",
            "\n",
            "    accuracy                           0.85       660\n",
            "   macro avg       0.84      0.85      0.84       660\n",
            "weighted avg       0.85      0.85      0.85       660\n",
            "\n"
          ]
        }
      ]
    }
  ]
}