Multi_Class(One_vs_all).ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Multi-Class(One vs. all).ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/saman-nia/MultiClass-Classification/blob/master/Multi_Class(One_vs_all).ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "metadata": {
        "id": "fnNyDnpnv3uK",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# Install TF 2.0 Alpha\n",
        "!pip install tf-nightly-gpu-2.0-preview"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "lBdNSN5rv9X2",
        "colab_type": "code",
        "outputId": "abdf7124-2e6b-4398-f46c-102c66daed41",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "import tensorflow as tf\n",
        "print('Check the Version of TF: ', tf.__version__)"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2.0.0-dev20190420\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "colab_type": "code",
        "id": "_Xg79b2X6FY4",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import warnings\n",
        "import random\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "#--------------------------------------------------------------\n",
        "from sklearn.feature_extraction.text import TfidfVectorizer\n",
        "from sklearn.feature_extraction.text import CountVectorizer\n",
        "from sklearn.preprocessing import MultiLabelBinarizer\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.multiclass import OneVsRestClassifier\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn import metrics\n",
        "from sklearn.preprocessing import OneHotEncoder\n",
        "from sklearn.preprocessing import LabelEncoder\n",
        "#--------------------------------------------------------------\n",
        "from IPython.display import clear_output"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "7HmeEC5kq4eD",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# Fetch Dataset from Github:\n",
        "url_data = 'https://raw.githubusercontent.com/saman-nia/Multi-label-Text-Classification-Features-with-Tensorflow/master/data/data_set.csv'\n",
        "data = pd.read_csv(url_data).dropna()\n",
        "# Dataset is now stored in a Pandas Dataframe"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "DDftFewn9n_x",
        "colab_type": "code",
        "outputId": "e8f6d56e-b49e-481f-c3f8-e83e18890340",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 289
        }
      },
      "cell_type": "code",
      "source": [
        "data.head()"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>name</th>\n",
              "      <th>description</th>\n",
              "      <th>price</th>\n",
              "      <th>shop</th>\n",
              "      <th>brand</th>\n",
              "      <th>labels</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>88136</td>\n",
              "      <td>Herren Schiesser Shirt 2er Pack ohne Arm unifa...</td>\n",
              "      <td>Shirt 2er Pack - American T-Shirts von Schiess...</td>\n",
              "      <td>22.000000</td>\n",
              "      <td>Herrenausstatter</td>\n",
              "      <td>Schiesser</td>\n",
              "      <td>Bekleidung,Shirts,T-Shirts,Wäsche,Unterhemden,...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>88247</td>\n",
              "      <td>Herren bruno banani Hipshorts Swim rot</td>\n",
              "      <td>Trendige Bademode bei Herrenausstatter: Swim B...</td>\n",
              "      <td>23.950001</td>\n",
              "      <td>Herrenausstatter</td>\n",
              "      <td>Bruno Banani</td>\n",
              "      <td>Bekleidung,Bademode,Badehosen,Bade Pants,Hosen...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>323683</td>\n",
              "      <td>Herren bruno banani Hipshorts Magical schwarz ...</td>\n",
              "      <td>Magical - die Serie aus Microfaser mit Elastha...</td>\n",
              "      <td>16.950001</td>\n",
              "      <td>Herrenausstatter</td>\n",
              "      <td>Bruno Banani</td>\n",
              "      <td>Bekleidung,Hosen,Kurze Hosen,Shorts,Wäsche,Unt...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>390726</td>\n",
              "      <td>MIA PlateauSandalette schwarz</td>\n",
              "      <td>Vagabond MIA PlateauSandalette schwarz: WOW - ...</td>\n",
              "      <td>74.950000</td>\n",
              "      <td>Zalando</td>\n",
              "      <td>Vagabond</td>\n",
              "      <td>Schuhe,Plateau Schuhe,Plateau Sandaletten,Sand...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>986780</td>\n",
              "      <td>Herren bruno banani Happy You V-Shirt schw. ge...</td>\n",
              "      <td>Happy U - V-Shirt aus elastischer Microfaser v...</td>\n",
              "      <td>29.950001</td>\n",
              "      <td>Herrenausstatter</td>\n",
              "      <td>Bruno Banani</td>\n",
              "      <td>Bekleidung,Shirts,T-Shirts,V-Shirts,Wäsche,Unt...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       id                                               name  \\\n",
              "0   88136  Herren Schiesser Shirt 2er Pack ohne Arm unifa...   \n",
              "1   88247             Herren bruno banani Hipshorts Swim rot   \n",
              "2  323683  Herren bruno banani Hipshorts Magical schwarz ...   \n",
              "3  390726                      MIA PlateauSandalette schwarz   \n",
              "4  986780  Herren bruno banani Happy You V-Shirt schw. ge...   \n",
              "\n",
              "                                         description      price  \\\n",
              "0  Shirt 2er Pack - American T-Shirts von Schiess...  22.000000   \n",
              "1  Trendige Bademode bei Herrenausstatter: Swim B...  23.950001   \n",
              "2  Magical - die Serie aus Microfaser mit Elastha...  16.950001   \n",
              "3  Vagabond MIA PlateauSandalette schwarz: WOW - ...  74.950000   \n",
              "4  Happy U - V-Shirt aus elastischer Microfaser v...  29.950001   \n",
              "\n",
              "               shop         brand  \\\n",
              "0  Herrenausstatter     Schiesser   \n",
              "1  Herrenausstatter  Bruno Banani   \n",
              "2  Herrenausstatter  Bruno Banani   \n",
              "3           Zalando      Vagabond   \n",
              "4  Herrenausstatter  Bruno Banani   \n",
              "\n",
              "                                              labels  \n",
              "0  Bekleidung,Shirts,T-Shirts,Wäsche,Unterhemden,...  \n",
              "1  Bekleidung,Bademode,Badehosen,Bade Pants,Hosen...  \n",
              "2  Bekleidung,Hosen,Kurze Hosen,Shorts,Wäsche,Unt...  \n",
              "3  Schuhe,Plateau Schuhe,Plateau Sandaletten,Sand...  \n",
              "4  Bekleidung,Shirts,T-Shirts,V-Shirts,Wäsche,Unt...  "
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        }
      ]
    },
    {
      "metadata": {
        "id": "JGpv5QSOlfDh",
        "colab_type": "code",
        "outputId": "d0a5413f-b37e-4984-de42-93b2fd931f59",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "data.shape"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(25879, 7)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        }
      ]
    },
    {
      "metadata": {
        "id": "s6Fk4Mpeq4RE",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# Convert category feature to string for extract text\n",
        "data['name'] = data.name.astype('str')\n",
        "data['description'] = data.description.astype('str')\n",
        "data['brand'] = data.brand.astype('str')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "colab_type": "code",
        "id": "lHwJ6-ADUMWK",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# Get raw text from products\n",
        "txt = data[['name', 'description', 'brand']].apply(lambda x: ' '.join(x),\n",
        "                                                   axis=1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "uKKGEDviCNyk",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# Get all level tags for each product\n",
        "y = data['labels'].apply(lambda x: set(x.split(','))) # split the tags in each row"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "colab_type": "code",
        "id": "2xgFl2S9UMSB",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# Split the data between training and test set\n",
        "txt_train, txt_test, y_train, y_test = train_test_split(txt, y, test_size=0.3, \n",
        "                                                        random_state=112358)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "colab_type": "code",
        "id": "fnFLVjkZUMKc",
        "outputId": "70830c76-4f20-4fb0-f417-4d00af9b9861",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "cell_type": "code",
      "source": [
        "# Create a word vectorizer, and train it on the text of the training set\n",
        "# Get the vectorized training set text as output\n",
        "vectorizer = CountVectorizer()\n",
        "X_train = vectorizer.fit_transform(txt_train)\n",
        "print(\"\\nNumber of text features: %s\" % X_train.shape[1])"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Number of text features: 41470\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "3tqRbztVfbXQ",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# Vectorize the test set text\n",
        "X_test = vectorizer.transform(txt_test)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "colab_type": "code",
        "id": "bgN3ysGlUycI",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# One-hot encoder for the product tags\n",
        "label_binarizer = MultiLabelBinarizer().fit(y)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "colab_type": "code",
        "id": "OoNnWwLFUyje",
        "outputId": "189c4691-9857-48fa-a4c9-f8496e386a15",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "# convert the training and test set tags\n",
        "y_train_bin = label_binarizer.transform(y_train)\n",
        "y_test_bin = label_binarizer.transform(y_test)\n",
        "print(\"Number of tags: %s\" % y_train_bin.shape[1])"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Number of tags: 612\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "aaeF-NCw5dPm",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "**Build Machine Learning Model by LogisticRegression**\n"
      ]
    },
    {
      "metadata": {
        "id": "WSIHpu3ZcWZV",
        "colab_type": "code",
        "outputId": "1ffec054-5e43-4efa-d7db-a5b1b0b14cb4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "cell_type": "code",
      "source": [
        "# Create a one-versus-rest classifier for out multi-output problem\n",
        "print(\"Machine Learning model is in the training process... \")\n",
        "model = OneVsRestClassifier(LogisticRegression(), n_jobs=-1).fit(X_train.tocoo()\n",
        "                                                                 , y_train_bin)\n",
        "# Measure the accuracy of your model\n",
        "accuracyMLmodel = metrics.accuracy_score(y_test_bin, model.predict(X_test))\n",
        "clear_output()\n",
        "print(\"\\nThe accuracy of Machine Learning model is %.1f%%\" % (100 * accuracyMLmodel))"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "The accuracy of Machine Learning model is 19.5%\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "mGgjAJvrMFM5",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# Find the geometric progression value to set the hidden layer.\n",
        "output_layer = y_train_bin.shape[1]\n",
        "x = [2**i for i in range(2, 20)]\n",
        "for j in range(len(x)):\n",
        "  if x[j] > output_layer:\n",
        "    Hidden_layer = (x[j])\n",
        "    break"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "PfnZIFEtDwL8",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "input_layer = X_train.shape[1]\n",
        "hidden_layer = Hidden_layer"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "Zte4zrgDfl2C",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "f3f30206-8d26-4986-9469-b2debdc909b6"
      },
      "cell_type": "code",
      "source": [
        "X_train.shape"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(18115, 41470)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 19
        }
      ]
    },
    {
      "metadata": {
        "id": "5IpxE9eQNDcg",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        },
        "outputId": "14d3643c-3e75-4c31-ddbe-2bf6ee6314d8"
      },
      "cell_type": "code",
      "source": [
        "print('Input Layer Nodes are: ------>', input_layer)\n",
        "print('Hidden Layer Nodes are: ----->', hidden_layer)\n",
        "print('Output Layer Nodes are: ----->', output_layer)"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Input Layer Nodes are: ------> 41470\n",
            "Hidden Layer Nodes are: -----> 1024\n",
            "Output Layer Nodes are: -----> 612\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "H-Nrbis3pLa3",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "n_epoch = 5\n",
        "batch_size = 64\n",
        "csv_logger = tf.keras.callbacks.CSVLogger('training.log')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "tlIkiQS4599X",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "**Build Deep Learning Model by Sigmoid Function and Binary Crossentropy Lossing Function**\n"
      ]
    },
    {
      "metadata": {
        "id": "9O2wASsQa-vr",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "modelTF = tf.keras.models.Sequential()\n",
        "modelTF.add(tf.keras.layers.Dense(hidden_layer, input_dim = input_layer, activation='tanh')) # hidden layer\n",
        "modelTF.add(tf.keras.layers.Dense(output_layer, activation='sigmoid')) # output layer\n",
        "# I use (sigmoid + binary_crossentropy) to encounter text topic multi-label\n",
        "modelTF.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "WazbpfDyu9LF",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 221
        },
        "outputId": "3255e587-1885-4d73-ff78-d8befa8e57d8"
      },
      "cell_type": "code",
      "source": [
        "modelTF.fit(X_train.toarray(), y_train_bin, batch_size = batch_size, callbacks=[csv_logger],\n",
        "            epochs = n_epoch, validation_data=(X_test.toarray(), y_test_bin))"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 18115 samples, validate on 7764 samples\n",
            "Epoch 1/5\n",
            "18115/18115 [==============================] - 10s 553us/sample - loss: 0.0596 - accuracy: 0.9875 - val_loss: 0.0240 - val_accuracy: 0.9933\n",
            "Epoch 2/5\n",
            "18115/18115 [==============================] - 9s 501us/sample - loss: 0.0170 - accuracy: 0.9947 - val_loss: 0.0173 - val_accuracy: 0.9946\n",
            "Epoch 3/5\n",
            "18115/18115 [==============================] - 9s 497us/sample - loss: 0.0104 - accuracy: 0.9966 - val_loss: 0.0154 - val_accuracy: 0.9951\n",
            "Epoch 4/5\n",
            "18115/18115 [==============================] - 9s 503us/sample - loss: 0.0069 - accuracy: 0.9978 - val_loss: 0.0148 - val_accuracy: 0.9954\n",
            "Epoch 5/5\n",
            "18115/18115 [==============================] - 9s 505us/sample - loss: 0.0048 - accuracy: 0.9985 - val_loss: 0.0151 - val_accuracy: 0.9955\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7fab0a3c1a90>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 23
        }
      ]
    },
    {
      "metadata": {
        "id": "yqyusPvdQ0lr",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "log_rec = pd.read_csv('training.log')\n",
        "log_rec['accuracy'] = round(log_rec.accuracy * 100, 2)\n",
        "log_rec['loss'] = round(log_rec.loss, 3)\n",
        "log_rec['val_accuracy'] = round(log_rec.val_accuracy * 100, 2)\n",
        "log_rec['val_loss'] = round(log_rec.val_loss, 3)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "QVe9A63Ob1OF",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 404
        },
        "outputId": "2f9deb88-5493-4f64-e772-993e5b577bc1"
      },
      "cell_type": "code",
      "source": [
        "plt.figure(figsize=(19,6))\n",
        "plt.subplot(1,2,1)\n",
        "plt.plot(log_rec['epoch'], log_rec['loss'], 'g', label='Training Loss', marker='h')\n",
        "plt.plot(log_rec['epoch'], log_rec['val_loss'], 'bo', label='Validation Loss')\n",
        "plt.xlabel('Number of Epochs')\n",
        "plt.ylabel('Cost')\n",
        "plt.xticks(log_rec['epoch'])\n",
        "plt.title('Cost Function During Optimization')\n",
        "plt.legend()\n",
        "plt.subplot(1,2,2)\n",
        "plt.plot(log_rec['epoch'], log_rec['accuracy'], 'r', label='Training Accuracy', marker='h')\n",
        "plt.plot(log_rec['epoch'], log_rec['val_accuracy'], 'bo', label='Validation Accuracy')\n",
        "plt.xlabel('Number of Epochs')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.xticks(log_rec['epoch'])\n",
        "#plt.yticks(log_rec['accuracy'])\n",
        "plt.title('The Process of Model Accuracy During Optimization')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABGEAAAGDCAYAAACLJA0cAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 1368x432 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "f0FYaC785IlE",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "822915fb-350b-4d50-8335-e1d00b30371e"
      },
      "cell_type": "code",
      "source": [
        "print(\"\\nThe accuracy of Machine Learning model is ------> %.1f%%\" % (100 * accuracyMLmodel))\n",
        "print(\"\\nThe accuracy of Deep Learning model is ---------> %.1f%%\" % (log_rec['val_accuracy'][n_epoch-1]))"
      ],
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "The accuracy of Machine Learning model is ------> 19.5%\n",
            "\n",
            "The accuracy of Deep Learning model is ---------> 99.5%\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}