project_neuralNet.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "project_neuralNet.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "Vnx2pOP5yR86",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!pip install alibi"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kaufSjwSxqzK",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "1a5aa520-e1c5-4063-98a0-65f6d8c19b06"
      },
      "source": [
        "import tensorflow as tf\n",
        "tf.get_logger().setLevel(40) # suppress deprecation messages\n",
        "tf.compat.v1.disable_v2_behavior() # disable TF2 behaviour as alibi code still relies on TF1 constructs\n",
        "from tensorflow.keras.layers import Dense, Input\n",
        "from tensorflow.keras.models import Model, load_model\n",
        "from tensorflow.keras.utils import to_categorical\n",
        "import matplotlib\n",
        "%matplotlib inline\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "import os\n",
        "import pandas as pd\n",
        "from alibi.explainers import CounterFactualProto\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "print('TF version: ', tf.__version__)\n",
        "print('Eager execution enabled: ', tf.executing_eagerly()) # False"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "TF version:  2.3.0\n",
            "Eager execution enabled:  False\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "g4WE_PdeA8Eg",
        "colab_type": "text"
      },
      "source": [
        "## Data set\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AedCcZLaygOh",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 359
        },
        "outputId": "fc14d557-6402-480a-c5c3-b1cb3268e914"
      },
      "source": [
        "names = [\n",
        "    'id', \n",
        "    'clump_thickness', \n",
        "    'cell_size_uniformity', \n",
        "    'cell_shape_uniformity',\n",
        "    'marginal_adhesion',\n",
        "    'epithelial_cell_size',\n",
        "    'bare_nuclei',\n",
        "    'bland_chromatin',\n",
        "    'normal_nucleoli',\n",
        "    'mitoses',\n",
        "    'class'\n",
        "]\n",
        "\n",
        "url = 'https://archive.ics.uci.edu/ml/machine-learning-databases/breast-cancer-wisconsin/breast-cancer-wisconsin.data'\n",
        "df1 = pd.read_csv(url, names=names)\n",
        "df1.head(10)\n",
        "# Dataset is now stored in a Pandas Dataframe"
      ],
      "execution_count": 11,
      "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>clump_thickness</th>\n",
              "      <th>cell_size_uniformity</th>\n",
              "      <th>cell_shape_uniformity</th>\n",
              "      <th>marginal_adhesion</th>\n",
              "      <th>epithelial_cell_size</th>\n",
              "      <th>bare_nuclei</th>\n",
              "      <th>bland_chromatin</th>\n",
              "      <th>normal_nucleoli</th>\n",
              "      <th>mitoses</th>\n",
              "      <th>class</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1000025</td>\n",
              "      <td>5</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1002945</td>\n",
              "      <td>5</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>5</td>\n",
              "      <td>7</td>\n",
              "      <td>10</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1015425</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1016277</td>\n",
              "      <td>6</td>\n",
              "      <td>8</td>\n",
              "      <td>8</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>7</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1017023</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>1017122</td>\n",
              "      <td>8</td>\n",
              "      <td>10</td>\n",
              "      <td>10</td>\n",
              "      <td>8</td>\n",
              "      <td>7</td>\n",
              "      <td>10</td>\n",
              "      <td>9</td>\n",
              "      <td>7</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>1018099</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>10</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>1018561</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>1033078</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>5</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>1033078</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "        id  clump_thickness  ...  mitoses  class\n",
              "0  1000025                5  ...        1      2\n",
              "1  1002945                5  ...        1      2\n",
              "2  1015425                3  ...        1      2\n",
              "3  1016277                6  ...        1      2\n",
              "4  1017023                4  ...        1      2\n",
              "5  1017122                8  ...        1      4\n",
              "6  1018099                1  ...        1      2\n",
              "7  1018561                2  ...        1      2\n",
              "8  1033078                2  ...        5      2\n",
              "9  1033078                4  ...        1      2\n",
              "\n",
              "[10 rows x 11 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cVhkPmPQ0TyZ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 359
        },
        "outputId": "2067ead2-a395-4163-ce72-0b5e012ba3ec"
      },
      "source": [
        "# remove NaN values\n",
        "dataset1 = df1.replace('?', np.NaN)\n",
        "dataset1.dropna(inplace=True)\n",
        "\n",
        "# remove non-feature col\n",
        "dataset1.drop(columns=['id'], inplace=True)\n",
        "\n",
        "\n",
        "# convert labels to M=1 and B=0\n",
        "dataset1['class'] = (dataset1['class'] == 4).astype(int)\n",
        "dataset1 = dataset1.astype(np.float64)\n",
        "dataset1['class'] = dataset1['class'].astype('int32')\n",
        "dataset1.head(10)"
      ],
      "execution_count": 8,
      "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>clump_thickness</th>\n",
              "      <th>cell_size_uniformity</th>\n",
              "      <th>cell_shape_uniformity</th>\n",
              "      <th>marginal_adhesion</th>\n",
              "      <th>epithelial_cell_size</th>\n",
              "      <th>bare_nuclei</th>\n",
              "      <th>bland_chromatin</th>\n",
              "      <th>normal_nucleoli</th>\n",
              "      <th>mitoses</th>\n",
              "      <th>class</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>5.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>10.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>3.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>6.0</td>\n",
              "      <td>8.0</td>\n",
              "      <td>8.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>8.0</td>\n",
              "      <td>10.0</td>\n",
              "      <td>10.0</td>\n",
              "      <td>8.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>10.0</td>\n",
              "      <td>9.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>10.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>4.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   clump_thickness  cell_size_uniformity  ...  mitoses  class\n",
              "0              5.0                   1.0  ...      1.0      0\n",
              "1              5.0                   4.0  ...      1.0      0\n",
              "2              3.0                   1.0  ...      1.0      0\n",
              "3              6.0                   8.0  ...      1.0      0\n",
              "4              4.0                   1.0  ...      1.0      0\n",
              "5              8.0                  10.0  ...      1.0      1\n",
              "6              1.0                   1.0  ...      1.0      0\n",
              "7              2.0                   1.0  ...      1.0      0\n",
              "8              2.0                   1.0  ...      5.0      0\n",
              "9              4.0                   2.0  ...      1.0      0\n",
              "\n",
              "[10 rows x 10 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cCz0G82TzQvx",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Split Train and Test datasets\n",
        "# Shuffle the data and define the train and test set:\n",
        "X = dataset1.iloc[:,:-1].values\n",
        "y = dataset1.iloc[:,9:].values\n",
        "\n",
        "#Standardise\n",
        "mu = X.mean(axis=0)\n",
        "sigma = X.std(axis=0)\n",
        "X = (X - mu) / sigma\n",
        "\n",
        "#Split Train and Test datasets\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25, random_state=42 )"
      ],
      "execution_count": 23,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mR1E9ZnT1NVP",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Modeling a ligistic Regression\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.metrics import accuracy_score"
      ],
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "N0-HlMoU0vbA",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "935fb389-e2e4-4e2d-92e3-250e6b61bca4"
      },
      "source": [
        "#Modeling a ligistic Regression\n",
        "lr = LogisticRegression(penalty='l2', max_iter=200, fit_intercept=True)\n",
        "lr.fit(X_train, np.ravel(y_train))\n",
        "accuracy_score(np.ravel(y_test), lr.predict(X_test))"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9532163742690059"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 19
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gEoDihVg3FZo",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y_train = to_categorical(y_train)\n",
        "y_test = to_categorical(y_test)"
      ],
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "E66cAD_V2K5_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def nn_model():\n",
        "    x_in = Input(shape=(9,))\n",
        "    x = Dense(40, activation='relu')(x_in)\n",
        "    x = Dense(40, activation='relu')(x)\n",
        "    x_out = Dense(2, activation='softmax')(x)\n",
        "    nn = Model(inputs=x_in, outputs=x_out)\n",
        "    nn.compile(loss='categorical_crossentropy', optimizer='sgd', metrics=['accuracy'])\n",
        "    return nn"
      ],
      "execution_count": 25,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n3Iplx_62TZj",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 289
        },
        "outputId": "7d154cb7-119a-4099-cada-867d19d7ab28"
      },
      "source": [
        "nn = nn_model()\n",
        "nn.summary()\n",
        "nn.fit(X_train, y_train, batch_size=64, epochs=500, verbose=0)\n",
        "nn.save('nn_breastCancer.h5', save_format='h5')"
      ],
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"functional_3\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "input_2 (InputLayer)         [(None, 9)]               0         \n",
            "_________________________________________________________________\n",
            "dense_3 (Dense)              (None, 40)                400       \n",
            "_________________________________________________________________\n",
            "dense_4 (Dense)              (None, 40)                1640      \n",
            "_________________________________________________________________\n",
            "dense_5 (Dense)              (None, 2)                 82        \n",
            "=================================================================\n",
            "Total params: 2,122\n",
            "Trainable params: 2,122\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pvt-_oEe3fJR",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "2e2b6443-64fc-4073-84fe-6e91c1851737"
      },
      "source": [
        "nn = load_model('nn_breastCancer.h5')\n",
        "score = nn.evaluate(X_test, y_test, verbose=0)\n",
        "print('Test accuracy: ', score[1])"
      ],
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Test accuracy:  0.9590643\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ASf9Ftz_3qox",
        "colab_type": "text"
      },
      "source": [
        "Generate counterfactual guided by the nearest class prototype"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "s6ZIi7Hq3g7R",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        },
        "outputId": "9a4530a7-7c00-4c5c-dac0-62a316a664ac"
      },
      "source": [
        "# Original instance:\n",
        "X = X_test[1].reshape((1,) + X_test[1].shape)\n",
        "shape = X.shape\n",
        "\n",
        "# Run counterfactual:\n",
        "\n",
        "# define model\n",
        "nn = load_model('nn_breastCancer.h5')\n",
        "\n",
        "# initialize explainer, fit and generate counterfactual\n",
        "cf = CounterFactualProto(nn, shape, use_kdtree=True, theta=10., max_iterations=1000,\n",
        "                         feature_range=(X_train.min(axis=0), X_train.max(axis=0)),\n",
        "                         c_init=1., c_steps=10)\n",
        "\n",
        "cf.fit(X_train)\n",
        "explanation = cf.explain(X)\n",
        "\n",
        "print('Original prediction: {}'.format(explanation.orig_class))\n",
        "print('Counterfactual prediction: {}'.format(explanation.cf['class']))"
      ],
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "No encoder specified. Using k-d trees to represent class prototypes.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Original prediction: 1\n",
            "Counterfactual prediction: 0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PGeQczsAARyY",
        "colab_type": "text"
      },
      "source": [
        "The explainer outputs We a counterfactual to see which variables need to be changed to increase or decrease to get the cancer diagnostic. The output below shows the result of undoing the pre-processing step and then checks where the counterfactual differs from the original instance:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XDYfsFbs4tAN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 170
        },
        "outputId": "7868c347-c8b9-4464-d854-961d9879d0c9"
      },
      "source": [
        "feature_names = names[1:-1]\n",
        "orig = X * sigma + mu\n",
        "counterfactual = explanation.cf['X'] * sigma + mu\n",
        "delta = counterfactual - orig\n",
        "for ii, ff in enumerate(feature_names):\n",
        "    if np.abs(delta[0][ii]) > 1e-4:\n",
        "        print('{}: {}'.format(ff, delta[0][ii]))"
      ],
      "execution_count": 37,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "clump_thickness: -4.526571275874357\n",
            "cell_size_uniformity: -2.010276286841414\n",
            "cell_shape_uniformity: -0.3861853841621876\n",
            "marginal_adhesion: -1.634759570244614\n",
            "epithelial_cell_size: -1.3719380826456486\n",
            "bare_nuclei: -6.525691631301381\n",
            "bland_chromatin: -3.389368207491753\n",
            "normal_nucleoli: -0.543684742930715\n",
            "mitoses: -5.531326410432158\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hgMwvrszDk-0",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "7885a639-9709-4e0a-9175-a58cabc1c994"
      },
      "source": [
        "print('Proportion of the observation %{} mitoses'.format(orig[0][8]))"
      ],
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Proportion of the observation %7.0 mitoses\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}