Bias Evaluation & Explainability.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import sklearn\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import classification_report, mean_squared_error, roc_curve, auc\n",
    "import seaborn as sn\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import shap\n",
    "from keras.layers import Input, Dense, Flatten, \\\n",
    "    Concatenate, concatenate, Dropout, Lambda\n",
    "from keras.models import Model, Sequential\n",
    "from keras.layers.embeddings import Embedding\n",
    "import keras\n",
    "from livelossplot import PlotLossesKeras\n",
    "import eli5\n",
    "from eli5.sklearn import PermutationImportance\n",
    "import scipy\n",
    "from scipy.cluster import hierarchy as hc\n",
    "from lime.lime_tabular import LimeTabularExplainer\n",
    "import math\n",
    "import xai\n",
    "import alibi\n",
    "\n",
    "params = {\n",
    "          \"ytick.color\" : \"w\",\n",
    "          \"xtick.color\" : \"w\",\n",
    "          \"text.color\": \"white\",\n",
    "          'figure.facecolor': \"#384151\",\n",
    "          'legend.facecolor': \"#384151\",\n",
    "          \"axes.labelcolor\" : \"w\",\n",
    "          \"axes.edgecolor\" : \"w\",\n",
    "          'font.size': '20.0',\n",
    "          'figure.figsize': [20, 7],\n",
    "    }\n",
    "\n",
    "plt.rcParams.update(params)\n",
    "\n",
    "\n",
    "shap.initjs()\n",
    "  \n",
    "\n",
    "label_column = \"loan\"\n",
    "csv_path = 'data/adult.data'\n",
    "csv_columns = [\"age\", \"workclass\", \"fnlwgt\", \"education\", \"education-num\", \"marital-status\",\n",
    "                   \"occupation\", \"relationship\", \"ethnicity\", \"gender\", \"capital-gain\", \"capital-loss\",\n",
    "                   \"hours-per-week\", \"native-country\", \"loan\"]\n",
    "input_columns = [\"age\", \"workclass\", \"education\", \"education-num\", \"marital-status\",\n",
    "                   \"occupation\", \"relationship\", \"ethnicity\", \"gender\", \"capital-gain\", \"capital-loss\",\n",
    "                   \"hours-per-week\", \"native-country\"]\n",
    "categorical_features = [\"workclass\", \"education\", \"marital-status\",\n",
    "                       \"occupation\", \"relationship\", \"ethnicity\", \"gender\",\n",
    "                       \"native-country\"]\n",
    "\n",
    "def prepare_data(df):\n",
    "    \n",
    "    if \"fnlwgt\" in df: del df[\"fnlwgt\"]\n",
    "    \n",
    "    tmp_df = df.copy()\n",
    "\n",
    "    # normalize data (this is important for model convergence)\n",
    "    dtypes = list(zip(tmp_df.dtypes.index, map(str, tmp_df.dtypes)))\n",
    "    for k,dtype in dtypes:\n",
    "        if dtype == \"int64\":\n",
    "            tmp_df[k] = tmp_df[k].astype(np.float32)\n",
    "            tmp_df[k] -= tmp_df[k].mean()\n",
    "            tmp_df[k] /= tmp_df[k].std()\n",
    "\n",
    "    cat_columns = tmp_df.select_dtypes(['object']).columns\n",
    "    tmp_df[cat_columns] = tmp_df[cat_columns].astype('category')\n",
    "    tmp_df[cat_columns] = tmp_df[cat_columns].apply(lambda x: x.cat.codes)\n",
    "    tmp_df[cat_columns] = tmp_df[cat_columns].astype('int8')\n",
    "    \n",
    "    return tmp_df\n",
    "\n",
    "def get_dataset_1():\n",
    "    tmp_df = df.copy()\n",
    "    tmp_df = tmp_df.groupby('loan') \\\n",
    "                .apply(lambda x: x.sample(100) if x[\"loan\"].iloc[0] else x.sample(7_900)) \\\n",
    "                .reset_index(drop=True)\n",
    "    \n",
    "    X = tmp_df.drop(label_column, axis=1).copy()\n",
    "    y = tmp_df[label_column].astype(int).values.copy()\n",
    "    \n",
    "    return tmp_df, df_display.copy()\n",
    "\n",
    "def get_production_dataset():\n",
    "    tmp_df = df.copy()\n",
    "    tmp_df = tmp_df.groupby('loan') \\\n",
    "                .apply(lambda x: x.sample(50) if x[\"loan\"].iloc[0] else x.sample(60)) \\\n",
    "                .reset_index(drop=True)\n",
    "    \n",
    "    X = tmp_df.drop(label_column, axis=1).copy()\n",
    "    y = tmp_df[label_column].astype(int).values.copy()\n",
    "    \n",
    "    return X, y\n",
    "\n",
    "\n",
    "def get_dataset_2():\n",
    "    tmp_df = df.copy()\n",
    "    tmp_df_display = df_display.copy()\n",
    "#     tmp_df_display[label_column] = tmp_df_display[label_column].astype(int).values\n",
    "    \n",
    "    X = tmp_df.drop(label_column, axis=1).copy()\n",
    "    y = tmp_df[label_column].astype(int).values.copy()\n",
    "    \n",
    "    X_display = tmp_df_display.drop(label_column, axis=1).copy()\n",
    "    y_display = tmp_df_display[label_column].astype(int).values.copy()\n",
    "    \n",
    "    X_train, X_valid, y_train, y_valid = \\\n",
    "        train_test_split(X, y, test_size=0.2, random_state=7)\n",
    "\n",
    "    return X, y, X_train, X_valid, y_train, y_valid, X_display, y_display, tmp_df, tmp_df_display\n",
    "    \n",
    "df_display = pd.read_csv(csv_path, names=csv_columns)\n",
    "df_display[label_column] = df_display[label_column].apply(lambda x: \">50K\" in x)\n",
    "df = prepare_data(df_display)\n",
    "\n",
    "def build_model(X):\n",
    "    input_els = []\n",
    "    encoded_els = []\n",
    "    dtypes = list(zip(X.dtypes.index, map(str, X.dtypes)))\n",
    "    for k,dtype in dtypes:\n",
    "        input_els.append(Input(shape=(1,)))\n",
    "        if dtype == \"int8\":\n",
    "            e = Flatten()(Embedding(df[k].max()+1, 1)(input_els[-1]))\n",
    "        else:\n",
    "            e = input_els[-1]\n",
    "        encoded_els.append(e)\n",
    "    encoded_els = concatenate(encoded_els)\n",
    "\n",
    "    layer1 = Dropout(0.5)(Dense(100, activation=\"relu\")(encoded_els))\n",
    "    out = Dense(1, activation='sigmoid')(layer1)\n",
    "\n",
    "    # train model\n",
    "    model = Model(inputs=input_els, outputs=[out])\n",
    "    model.compile(optimizer=\"adam\", loss='binary_crossentropy', metrics=['accuracy'])\n",
    "    return model\n",
    "\n",
    "def f_in(X, m=None):\n",
    "    \"\"\"Preprocess input so it can be provided to a function\"\"\"\n",
    "    if m:\n",
    "        return [X.iloc[:m,i] for i in range(X.shape[1])]\n",
    "    else:\n",
    "        return [X.iloc[:,i] for i in range(X.shape[1])]\n",
    "\n",
    "def f_out(probs):\n",
    "    \"\"\"Convert probabilities into classes\"\"\"\n",
    "    return list((probs >= 0.5).astype(int).T[0])\n",
    "\n",
    "def plot_roc(y, probs):\n",
    "    \n",
    "    fpr, tpr, _ = roc_curve(y, probs)\n",
    "\n",
    "    roc_auc = auc(fpr, tpr)\n",
    "    print(roc_auc)\n",
    "\n",
    "    plt.figure()\n",
    "    plt.plot(fpr, tpr, label='ROC curve (area = %0.2f)' % roc_auc)\n",
    "    plt.plot([0, 1], [0, 1], 'k--')\n",
    "    plt.xlim([0.0, 1.0])\n",
    "    plt.ylim([0.0, 1.05])\n",
    "    plt.xlabel('False Positive Rate')\n",
    "    plt.ylabel('True Positive Rate')\n",
    "    plt.title('ROC curve (area = %0.2f)' % roc_auc)\n",
    "    plt.legend(loc=\"lower right\")\n",
    "    plt.rcParams.update(params)\n",
    "    plt.show()\n",
    "    \n",
    "def plot_learning_curves(model, X, y):\n",
    "    X_train, X_val, y_train, y_val = train_test_split(X, y, test_size=0.2)\n",
    "    train_errors, val_errors = [], []\n",
    "    for m in list(np.logspace(0.6,4,dtype='int')):\n",
    "        if m >= len(X_train): break\n",
    "        model.fit(f_in(X_train,m), y_train[:m], epochs=50, batch_size=512, verbose=0)\n",
    "        y_train_predict = model.predict(f_in(X_train,m))\n",
    "        y_val_predict = model.predict(f_in(X_val))\n",
    "        y_train_predict = f_out(y_train_predict)\n",
    "        y_val_predict = f_out(y_val_predict)\n",
    "        train_errors.append(mean_squared_error(y_train[:m], y_train_predict))\n",
    "        val_errors.append(mean_squared_error(y_val, y_val_predict))\n",
    "    plt.plot(np.sqrt(train_errors), \"r-+\", linewidth=2, label=\"train\")\n",
    "    plt.plot(np.sqrt(val_errors), \"b-\", linewidth=3, label=\"val\")\n",
    "    \n",
    "def keras_score(self, X, y, **kwargs):\n",
    "    \"\"\" Scorer class for eli5 library on feature importance\"\"\"\n",
    "    input_test = [X[:,i] for i in range(X.shape[1])]\n",
    "    loss = self.evaluate(input_test, y)\n",
    "    if type(loss) is list:\n",
    "        # The first one is the error, the rest are metrics\n",
    "        return -loss[0]\n",
    "    return -loss\n",
    "\n",
    "class ModelWrapper():\n",
    "    \"\"\" Keras model wrapper to override the predict function\"\"\"\n",
    "    def __init__(self, model):\n",
    "        self.model = model\n",
    "    \n",
    "    def predict(self, X, **kwargs):\n",
    "        return self.model.predict([X.iloc[:,i] for i in range(X.shape[1])])\n",
    "\n",
    "def plot_all_features(X, plot_numeric=True, hist=True, dropna=False):\n",
    "    fig = plt.figure(figsize=(20,15))\n",
    "    cols = 5\n",
    "    rows = math.ceil(float(X.shape[1]) / cols)\n",
    "    for i, column in enumerate(X.columns):\n",
    "        ax = fig.add_subplot(rows, cols, i + 1)\n",
    "        ax.set_title(column)\n",
    "        if X.dtypes[column] == np.object:\n",
    "            X[column].value_counts().plot(kind=\"bar\", axes=ax)\n",
    "        elif plot_numeric:\n",
    "            if hist:\n",
    "                X[column].hist(axes=ax)\n",
    "                plt.xticks(rotation=\"vertical\")\n",
    "            else:\n",
    "                if dropna:\n",
    "                    X[column].dropna().plot()\n",
    "                else:\n",
    "                    X[column].plot()\n",
    "                    \n",
    "    plt.subplots_adjust(hspace=0.7, wspace=0.2)\n",
    "    \n",
    "def plot_dendogram(corr, X):\n",
    "    corr_condensed = hc.distance.squareform(1-corr)\n",
    "    z = hc.linkage(corr_condensed, method=\"average\")\n",
    "    fig = plt.figure(figsize=(16,5))\n",
    "    dendrogram = hc.dendrogram(\n",
    "        z, labels=X.columns, orientation=\"left\", leaf_font_size=16)\n",
    "    plt.show()\n",
    "    \n",
    "def shap_predict(X):\n",
    "    values = model.predict([X[:,i] for i in range(X.shape[1])]).flatten()\n",
    "    return values\n",
    "\n",
    "def lime_predict_proba(X):\n",
    "    values = model.predict([X[:,i] for i in range(X.shape[1])]).flatten()\n",
    "    prob_pairs = np.array([1-values, values]).T\n",
    "    return prob_pairs    \n",
    "\n",
    "\n",
    "import alibi\n",
    "import numpy as np\n",
    "from sklearn.preprocessing import LabelEncoder, StandardScaler, OneHotEncoder\n",
    "from sklearn.impute import SimpleImputer\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.compose import ColumnTransformer\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "alibi_data, alibi_labels, alibi_feature_names, alibi_category_map = alibi.datasets.adult()\n",
    "\n",
    "def get_alibi_data():\n",
    "    \n",
    "    # define train and test set\n",
    "    np.random.seed(0)\n",
    "    data_perm = np.random.permutation(np.c_[alibi_data, alibi_labels])\n",
    "    data = data_perm[:, :-1]\n",
    "    labels = data_perm[:, -1]\n",
    "\n",
    "    idx = 30000\n",
    "    X_train, y_train = data[:idx, :], labels[:idx]\n",
    "    X_test, y_test = data[idx + 1:, :], labels[idx + 1:]\n",
    "\n",
    "    # feature transformation pipeline\n",
    "    ordinal_features = [x for x in range(len(alibi_feature_names)) if x not in list(alibi_category_map.keys())]\n",
    "    ordinal_transformer = Pipeline(steps=[('imputer', SimpleImputer(strategy='median')),\n",
    "                                          ('scaler', StandardScaler())])\n",
    "\n",
    "    categorical_features = list(alibi_category_map.keys())\n",
    "    categorical_transformer = Pipeline(steps=[('imputer', SimpleImputer(strategy='median')),\n",
    "                                              ('onehot', OneHotEncoder(handle_unknown='ignore'))])\n",
    "\n",
    "    preprocessor = ColumnTransformer(transformers=[('num', ordinal_transformer, ordinal_features),\n",
    "                                                   ('cat', categorical_transformer, categorical_features)])\n",
    "\n",
    "\n",
    "    np.random.seed(0)\n",
    "    clf = RandomForestClassifier(n_estimators=50)\n",
    "\n",
    "    alibi_loan_model = Pipeline(steps=[\n",
    "        ('preprocessor', preprocessor),\n",
    "        ('clf', clf)])\n",
    "\n",
    "    alibi_loan_model.fit(X_train, y_train)\n",
    "    \n",
    "    return X_train, X_test, y_train, y_test, \\\n",
    "        alibi_category_map, alibi_feature_names, alibi_loan_model\n",
    "\n",
    "X_train_alibi, X_test_alibi, y_train_alibi, y_test_alibi, \\\n",
    "    category_map_alibi, feature_names_alibi, loan_model_alibi = get_alibi_data()\n",
    "\n",
    "\n",
    "import tensorflow as tf\n",
    "tf.logging.set_verbosity(tf.logging.ERROR)  # suppress deprecation messages\n",
    "from tensorflow.keras.models import load_model\n",
    "from tensorflow.keras.utils import to_categorical\n",
    "\n",
    "from alibi.explainers import CounterFactual\n",
    "\n",
    "# def cnn_model():\n",
    "#     from tensorflow.keras import backend as K\n",
    "#     from tensorflow.keras.layers import Conv2D, Dense, Dropout, Flatten, MaxPooling2D, Input, UpSampling2D\n",
    "#     from tensorflow.keras.models import Model\n",
    "#     x_in = Input(shape=(28, 28, 1))\n",
    "#     x = Conv2D(filters=64, kernel_size=2, padding='same', activation='relu')(x_in)\n",
    "#     x = MaxPooling2D(pool_size=2)(x)\n",
    "#     x = Dropout(0.3)(x)\n",
    "#     x = Conv2D(filters=32, kernel_size=2, padding='same', activation='relu')(x)\n",
    "#     x = MaxPooling2D(pool_size=2)(x)\n",
    "#     x = Dropout(0.3)(x)\n",
    "#     x = Flatten()(x)\n",
    "#     x = Dense(256, activation='relu')(x)\n",
    "#     x = Dropout(0.5)(x)\n",
    "#     x_out = Dense(10, activation='softmax')(x)\n",
    "#     cnn = Model(inputs=x_in, outputs=x_out)\n",
    "#     cnn.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])\n",
    "#     return cnn\n",
    "\n",
    "# cnn = cnn_model()\n",
    "# cnn.summary()\n",
    "# cnn.fit(x_train, y_train, batch_size=64, epochs=3, verbose=0)\n",
    "# cnn.save('mnist_cnn.h5')\n",
    "\n",
    "def show_iterations(explanation, max_lam_steps=10):\n",
    "    n_cfs = np.array([len(explanation['all'][iter_cf]) for iter_cf in range(max_lam_steps)])\n",
    "    examples = {}\n",
    "    for ix, n in enumerate(n_cfs):\n",
    "        if n>0:\n",
    "            examples[ix] = {'ix': ix, 'lambda': explanation['all'][ix][0]['lambda'],\n",
    "                           'X': explanation['all'][ix][0]['X']}\n",
    "    columns = len(examples) + 1\n",
    "    rows = 1\n",
    "\n",
    "    fig = plt.figure(figsize=(16,6))\n",
    "\n",
    "    for i, key in enumerate(examples.keys()):\n",
    "        ax = plt.subplot(rows, columns, i+1)\n",
    "        ax.get_xaxis().set_visible(False)\n",
    "        ax.get_yaxis().set_visible(False)\n",
    "        plt.imshow(examples[key]['X'].reshape(28,28))\n",
    "        plt.title(f'Iter {key}')\n",
    "\n",
    "(cf_x_train, cf_y_train), (cf_x_test, cf_y_test) = tf.keras.datasets.mnist.load_data()\n",
    "print('x_train shape:', cf_x_train.shape, 'y_train shape:', cf_y_train.shape)\n",
    "plt.gray()\n",
    "\n",
    "cf_x_train = cf_x_train.astype('float32') / 255\n",
    "cf_x_test = cf_x_test.astype('float32') / 255\n",
    "cf_x_train = np.reshape(cf_x_train, cf_x_train.shape + (1,))\n",
    "cf_x_test = np.reshape(cf_x_test, cf_x_test.shape + (1,))\n",
    "cf_y_train = to_categorical(cf_y_train)\n",
    "cf_y_test = to_categorical(cf_y_test)\n",
    "\n",
    "cf_xmin, cf_xmax = -.5, .5\n",
    "cf_x_train = ((cf_x_train - cf_x_train.min()) / (cf_x_train.max() - cf_x_train.min())) * (cf_xmax - cf_xmin) + cf_xmin\n",
    "cf_x_test = ((cf_x_test - cf_x_test.min()) / (cf_x_test.max() - cf_x_test.min())) * (cf_xmax - cf_xmin) + cf_xmin\n",
    "        "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# A practical guide towards explainability\n",
    "# and bias evaluation in machine learning\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "<br>\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "<br>\n",
    "\n",
    "<br>\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "<h2>Alejandro Saucedo</h2>\n",
    "<br>\n",
    "Chief Scientist, The Institute for Ethical AI & Machine Learning\n",
    "<br>\n",
    "Director of ML Engineering, Seldon Technologies\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "<h3>Twitter: AxSaucedo</h3>\n",
    "<br>\n",
    "<h3>Slides: github.com/EthicalML/explainability-and-bias</h3>\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "<hr>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Today\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "## 1) Hands on example: Company \"Hype-ML\" automating loan approval process \n",
    "\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "## 2) Terminology + why it's not about \"just removing bias\"\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "## 3) Data Analysis\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "## 4) Model Evaluation\n",
    "\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "## 5) Production Monitoring\n",
    "\n",
    "<br>\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "<hr>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Hype-ML - A NEW Project has come in!\n",
    "\n",
    "<br>\n",
    "\n",
    "* Insurance company has a process where domain expert approves/rejects loan applications\n",
    "\n",
    "<br>\n",
    "\n",
    "* They receive over 1m applications and want to automate the process\n",
    "\n",
    "<br>\n",
    "\n",
    "# Business wants it NOW!\n",
    "\n",
    "<br>\n",
    "\n",
    "* They heard their competitor is using \"Machine Learning\" and business says we need to use that\n",
    "\n",
    "\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "<hr>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# The team had a look at how this worked\n",
    "\n",
    "<br>\n",
    "\n",
    "<img src=\"images/mlall.png\" style=\"width:70vw\">\n",
    "\n",
    "\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "<hr>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# The team asked for DATA\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "## Business gave them an excel sheet with 25 rows\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "The team pushed back, and after a while they finally got a dataset with ~8000 rows\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "<hr>\n",
    "\n",
    "<br><br><br>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# And so it begun...\n",
    "\n",
    "<br>\n",
    "\n",
    "<img src=\"images/copypasta.jpg\" style=\"height:50vh\">\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "## The Hype-ML journey towards greatness...\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "<hr>\n",
    "\n",
    "<br><br><br>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "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>age</th>\n",
       "      <th>workclass</th>\n",
       "      <th>education</th>\n",
       "      <th>education-num</th>\n",
       "      <th>marital-status</th>\n",
       "      <th>occupation</th>\n",
       "      <th>relationship</th>\n",
       "      <th>ethnicity</th>\n",
       "      <th>gender</th>\n",
       "      <th>capital-gain</th>\n",
       "      <th>capital-loss</th>\n",
       "      <th>hours-per-week</th>\n",
       "      <th>native-country</th>\n",
       "      <th>loan</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>39</td>\n",
       "      <td>State-gov</td>\n",
       "      <td>Bachelors</td>\n",
       "      <td>13</td>\n",
       "      <td>Never-married</td>\n",
       "      <td>Adm-clerical</td>\n",
       "      <td>Not-in-family</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>2174</td>\n",
       "      <td>0</td>\n",
       "      <td>40</td>\n",
       "      <td>United-States</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>50</td>\n",
       "      <td>Self-emp-not-inc</td>\n",
       "      <td>Bachelors</td>\n",
       "      <td>13</td>\n",
       "      <td>Married-civ-spouse</td>\n",
       "      <td>Exec-managerial</td>\n",
       "      <td>Husband</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13</td>\n",
       "      <td>United-States</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>38</td>\n",
       "      <td>Private</td>\n",
       "      <td>HS-grad</td>\n",
       "      <td>9</td>\n",
       "      <td>Divorced</td>\n",
       "      <td>Handlers-cleaners</td>\n",
       "      <td>Not-in-family</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>40</td>\n",
       "      <td>United-States</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>53</td>\n",
       "      <td>Private</td>\n",
       "      <td>11th</td>\n",
       "      <td>7</td>\n",
       "      <td>Married-civ-spouse</td>\n",
       "      <td>Handlers-cleaners</td>\n",
       "      <td>Husband</td>\n",
       "      <td>Black</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>40</td>\n",
       "      <td>United-States</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>28</td>\n",
       "      <td>Private</td>\n",
       "      <td>Bachelors</td>\n",
       "      <td>13</td>\n",
       "      <td>Married-civ-spouse</td>\n",
       "      <td>Prof-specialty</td>\n",
       "      <td>Wife</td>\n",
       "      <td>Black</td>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>40</td>\n",
       "      <td>Cuba</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   age          workclass   education  education-num       marital-status  \\\n",
       "0   39          State-gov   Bachelors             13        Never-married   \n",
       "1   50   Self-emp-not-inc   Bachelors             13   Married-civ-spouse   \n",
       "2   38            Private     HS-grad              9             Divorced   \n",
       "3   53            Private        11th              7   Married-civ-spouse   \n",
       "4   28            Private   Bachelors             13   Married-civ-spouse   \n",
       "\n",
       "           occupation    relationship ethnicity   gender  capital-gain  \\\n",
       "0        Adm-clerical   Not-in-family     White     Male          2174   \n",
       "1     Exec-managerial         Husband     White     Male             0   \n",
       "2   Handlers-cleaners   Not-in-family     White     Male             0   \n",
       "3   Handlers-cleaners         Husband     Black     Male             0   \n",
       "4      Prof-specialty            Wife     Black   Female             0   \n",
       "\n",
       "   capital-loss  hours-per-week  native-country   loan  \n",
       "0             0              40   United-States  False  \n",
       "1             0              13   United-States  False  \n",
       "2             0              40   United-States  False  \n",
       "3             0              40   United-States  False  \n",
       "4             0              40            Cuba  False  "
      ]
     },
     "execution_count": 160,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_data, df_display = get_dataset_1()\n",
    "\n",
    "df_display.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training size: (6400,) Testing size:  (1600,)\n"
     ]
    }
   ],
   "source": [
    "X = df_data.drop(label_column, axis=1).copy()\n",
    "y = df_data[label_column].astype(int).values.copy()\n",
    "\n",
    "X_train, X_valid, y_train, y_valid = \\\n",
    "        train_test_split(X, y, test_size=0.2, random_state=7)\n",
    "\n",
    "print(\"Training size:\", y_train.shape, \"Testing size: \", y_valid.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "<img src=\"images/train-test.png\" style=\"width:70vw\">"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Log-loss (cost function):\n",
      "training   (min:    0.068, max:    0.553, cur:    0.068)\n",
      "validation (min:    0.062, max:    0.466, cur:    0.062)\n",
      "\n",
      "Accuracy:\n",
      "training   (min:    0.888, max:    0.988, cur:    0.988)\n",
      "validation (min:    0.987, max:    0.987, cur:    0.987)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x7fae564f3898>"
      ]
     },
     "execution_count": 162,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 1 layer, 100 neurons model, with softmax (0-1 probabilities)\n",
    "model = build_model(X)\n",
    "\n",
    "model.fit(f_in(X_train), y_train, epochs=10,\n",
    "    batch_size=512, shuffle=True, validation_data=(f_in(X_valid), y_valid),\n",
    "    callbacks=[PlotLossesKeras()], verbose=0, validation_split=0.05,)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1600/1600 [==============================] - 0s 47us/step\n",
      "Error 0.0623: \n",
      "Accuracy 98.6875: \n"
     ]
    }
   ],
   "source": [
    "score = model.evaluate(f_in(X_valid), y_valid, verbose=1)\n",
    "print(\"Error %.4f: \" % score[0])\n",
    "print(\"Accuracy %.4f: \" % (score[1]*100))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Accuracy is ~98%!\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "### What a better result on a Friday evening!\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "# Time for PROD?\n",
    "\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "<hr>\n",
    "\n",
    "<br><br><br>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# We push to PROD!\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "### A few weeks go by...\n",
    "\n",
    "<br>\n",
    "\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "<hr>\n",
    "\n",
    "<br><br><br>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# BUZZFEED NEWS: HYPE-ML DEPLOYS RACIST SEXIST AI!!!!\n",
    "\n",
    "<br>\n",
    "\n",
    "<img src=\"images/layer.jpg\" style=\"height: 50vh\">\n",
    "\n",
    "\n",
    "<br>\n",
    "\n",
    "\n",
    "# ...but... but we followed the instructions in the internet!\n",
    "\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "<hr>\n",
    "\n",
    "<br><br><br>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Time to diagnose!\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "## We ask business to gather the 110 applications in production...\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "## ...and label it to understand what went wrong\n",
    "\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "<hr>\n",
    "\n",
    "<br><br><br>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(110, 13)\n"
     ]
    },
    {
     "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>age</th>\n",
       "      <th>workclass</th>\n",
       "      <th>education</th>\n",
       "      <th>education-num</th>\n",
       "      <th>marital-status</th>\n",
       "      <th>occupation</th>\n",
       "      <th>relationship</th>\n",
       "      <th>ethnicity</th>\n",
       "      <th>gender</th>\n",
       "      <th>capital-gain</th>\n",
       "      <th>capital-loss</th>\n",
       "      <th>hours-per-week</th>\n",
       "      <th>native-country</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.177293</td>\n",
       "      <td>4</td>\n",
       "      <td>15</td>\n",
       "      <td>-0.031359</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0.237140</td>\n",
       "      <td>-0.216656</td>\n",
       "      <td>-0.035429</td>\n",
       "      <td>39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>-0.555822</td>\n",
       "      <td>4</td>\n",
       "      <td>9</td>\n",
       "      <td>1.134722</td>\n",
       "      <td>4</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.145918</td>\n",
       "      <td>-0.216656</td>\n",
       "      <td>0.774456</td>\n",
       "      <td>39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>-1.142313</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>0.357334</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.145918</td>\n",
       "      <td>-0.216656</td>\n",
       "      <td>-0.035429</td>\n",
       "      <td>39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>0.397227</td>\n",
       "      <td>4</td>\n",
       "      <td>12</td>\n",
       "      <td>1.523415</td>\n",
       "      <td>5</td>\n",
       "      <td>12</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.145918</td>\n",
       "      <td>-0.216656</td>\n",
       "      <td>0.612479</td>\n",
       "      <td>39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>-1.069002</td>\n",
       "      <td>4</td>\n",
       "      <td>15</td>\n",
       "      <td>-0.031359</td>\n",
       "      <td>4</td>\n",
       "      <td>13</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.145918</td>\n",
       "      <td>-0.216656</td>\n",
       "      <td>-1.979153</td>\n",
       "      <td>39</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        age  workclass  education  education-num  marital-status  occupation  \\\n",
       "0  0.177293          4         15      -0.031359               2           1   \n",
       "1 -0.555822          4          9       1.134722               4          12   \n",
       "2 -1.142313          4          8       0.357334               4           3   \n",
       "3  0.397227          4         12       1.523415               5          12   \n",
       "4 -1.069002          4         15      -0.031359               4          13   \n",
       "\n",
       "   relationship  ethnicity  gender  capital-gain  capital-loss  \\\n",
       "0             0          4       1      0.237140     -0.216656   \n",
       "1             1          2       0     -0.145918     -0.216656   \n",
       "2             4          2       0     -0.145918     -0.216656   \n",
       "3             4          4       0     -0.145918     -0.216656   \n",
       "4             1          4       0     -0.145918     -0.216656   \n",
       "\n",
       "   hours-per-week  native-country  \n",
       "0       -0.035429              39  \n",
       "1        0.774456              39  \n",
       "2       -0.035429              39  \n",
       "3        0.612479              39  \n",
       "4       -1.979153              39  "
      ]
     },
     "execution_count": 164,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_prod, y_prod = get_production_dataset()\n",
    "\n",
    "print(X_prod.shape)\n",
    "X_prod.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJQAAAGDCAYAAAB9QUrBAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzs3XlclVXix/HvBZV9EUFiE7XcUUBFcE+NXEutLK1mUjPLGm1v2pdpGmfay2ZSU8vfTKXjXm5Fmpam4oqSuyayiWzKDin39wdKgyDeh4QH9fP+C84599zv08sovj7PuZaofiOtAgAAAAAAAGxkZ3YAAAAAAAAAXFkolAAAAAAAAGAIhRIAAAAAAAAMoVACAAAAAACAIRRKAAAAAAAAMIRCCQAAAAAAAIZQKAEAAAAAAMAQCiUAAAAAAAAYQqEEAAAAAAAAQxqYHaCmNq1dbHYEAAAAAACAq8qpUzkafNvYS67jDiUAAAAAAABIklLTTtq0jkIJAAAAAAAAhlAoAQAAAAAAwBAKJQAAAAAAABhCoQQAAAAAAABDKJQAAAAAAABgSANbF/br013hoR3U6vrmanV9c7m4OGt1zHq9NvUDw2/q491EE8eNVmREuDzc3ZSZla0fNsZq9tz5ys3LN7wfAAAAAAAA6o7NhdLYe+9Q6xtaKL+gUOnpmXJxca7RGwb4+2rmh1Pl5eWp9Ru2KCExWe3bttJdtw9TVESYJk55Xjk5eTXaGwAAAAAAALXP5kLpg399qpPpmUpKTlV4aAf9673Xa/SGTz06UV5ennpn2iwtXLKyfHzKpLEaM+pWPTT+Hr35/owa7Q0AAAAAAIDaZ/MZSjt2xSspOfV3vVmAv6+iIsKVkpqmRUtXVZib9dk8FRQWalB0Xzk6Ovyu9wEAAAAAAEDtqdNDuTuHdZQkxW6Lk9VqrTBXUFik3fH75eTkqJB2resyFgAAAAAAAAyw+ZG3yyE4yF+SdDwppcr5pORUKSJcQUH+2rZzT7V7Hcg4oAeXP3jZMwIAAAAAAFyrPu4+y6Z1dVoonT/IOy+/oMr5vLyycTcXlyrnhw+N1vBh0ee+O3vZ8wEAAAAAAODS6rRQ+r2WrYjRshUxkqRNaxdr3dh15gYCAAAAAAC4iuw7cNimdXV6hlL+uTuTXM/dqXQhV9ey8dz8/DrLBAAAAAAAAGPqtFBKSCw7O6lZoH+V84EBfpKkxMSqz1gCAAAAAACA+eq0UNqxq+yg7W5dQ2WxWCrMOTs5qlNIWxUWFil+38G6jAUAAAAAAAADaqVQsre3V3BQgAL8fSuMJ6ekafPWnfL389XtIwZXmJswdrScnZy0Oma9ioqKayMWAAAAAAAALgObD+Xu07Ob+vTsJklq4tVYkhTSoY1efOZPkqTTObmaNn2uJMnH20vz5k5T6omTuu3uhyrs8/YHMzXzw6l6cvIEdQ3vqITjSWrfrnXZ14nJmj7n88tyYQAAAAAAAKgdNhdKrW5ooaGD+lcYC/S/ToH+10mSUk+cLC+UqpOckqZxk57WA+PGKCoiXD0iOysj65TmL1qu2XPnKzePA7kBAAAAAADqM0tUv5FWs0PUxKa1i82OAAAAAAAAcFXZd+Cwxk965pLr6vRQbgAAAAAAAFz5KJQAAAAAAABgCIUSAAAAAAAADKFQAgAAAAAAgCEUSgAAAAAAADCEQgkAAAAAAACGUCgBAAAAAADAEAolAAAAAAAAGEKhBAAAAAAAAEMolAAAAAAAAGAIhRIAAAAAAAAMoVACAAAAAACAIRRKAAAAAAAAMIRCCQAAAAAAAIZQKAEAAAAAAMAQCiUAAAAAAAAYQqEEAAAAAAAAQyiUAAAAAAAAYAiFEgAAAAAAAAyhUAIAAAAAAIAhFEoAAAAAAAAwhEIJAAAAAAAAhlAoAQAAAAAAwBAKJQAAAAAAABhCoQQAAAAAAABDKJQAAAAAAABgCIUSAAAAAAAADKFQAgAAAAAAgCEUSgAAAAAAADCEQgkAAAAAAACGUCgBAAAAAADAEAolAAAAAAAAGEKhBAAAAAAAAEMolAAAAAAAAGAIhRIAAAAAAAAMoVACAAAAAACAIRRKAAAAAAAAMKSBkcU+3k00cdxoRUaEy8PdTZlZ2fphY6xmz52v3Lx8m/fpFNJW99w1Qq2uby4vL09lZ5/W0WPHtWDxSm3eutPwRQAAAAAAAKDu2FwoBfj7auaHU+Xl5an1G7YoITFZ7du20l23D1NURJgmTnleOTl5l9xn5K0D9cxjD6qgsFDrN2xRenqmfHya6MZeUeoR2UXTZ3+uuZ8v+l0XBQAAAAAAgNpjc6H01KMT5eXlqXemzdLCJSvLx6dMGqsxo27VQ+Pv0Zvvz6h2D3t7e02acK+Ki4s17qGndTwxpXxubrNFmjvzHY295w598d9l+vXXMzW4HAAAAAAAANQ2m85QCvD3VVREuFJS07Ro6aoKc7M+m6eCwkINiu4rR0eHavdxd3eVm6uLjielViiTJCnheLISE1Pk6OggJydHg5cBAAAAAACAumJTodQ5rKMkKXZbnKxWa4W5gsIi7Y7fLycnR4W0a13tPtnZp5WVfVrNAv0UGOBXYS4o0E9BgX46eOioTY/OAQAAAAAAwBw2PfIWHOQvSTqelFLlfFJyqhQRrqAgf23buafavd75cKZeee4xfTb9La3fsEUZmVny8W6iPr0idfRYol5+/V2DlwAAAAAAAIC6ZFOh5OLiLEnKyy+ocj4vr2zczcXlknutXb9JGRnZeu3FxzVkYL/y8cysbK1YvVbJqWkXfe3wodEaPizalsgAAAAAAACoJTYfyn25DLypj557cpLW/bhFjz3zF51IS9d1vj4a94dReurRiQoP7aAX//JOla9dtiJGy1bESJI2rV1cl7EBAAAAAABwjk1nKOWfuzPJ9dydShdydS0bz83Pr3afoEA/vfD0Izp6LFGvTf1ACYnJKi4pUUJisl6b+oH2HTisATf2VHhoByPXAAAAAAAAgDpkU6GUcO4T2ZoF+lc5f/6A7cTEqs9YOi+ya5gaNmyonXF7Kx3ubbVatWv3XklS29bX2xILAAAAAAAAJrCpUNqxq+yg7W5dQ2WxWCrMOTs5qlNIWxUWFil+38Fq92nYsKEkqbGne5XznufGfz1zxpZYAAAAAAAAMIFNhVJySpo2b90pfz9f3T5icIW5CWNHy9nJSatj1quoqLh8PDgoQMFBARXWxu0puwOpX5/uur5lcIW5Vtc3V/8+3VVaWqrtl/ikOAAAAAAAAJjH5kO53/5gpmZ+OFVPTp6gruEdlXA8Se3btS77OjFZ0+d8XmH9vLnTJEnd+99WPrZ3/2EtX7VGwwYP0Jx/van1G7boRFq6/K5rqj49u6lRo4aat/Br/XIs8TJdHgAAAAAAAC43mwul5JQ0jZv0tB4YN0ZREeHqEdlZGVmnNH/Rcs2eO1+5edUfyH3eG2/9Uzt379XQgf0UGREmZ2cnFeQXaHf8Pi1bEaPvvt9Y44sBAAAAAABA7bNE9RtpvfSy+mfT2sVmRwAAAAAAALiq7DtwWOMnPXPJdTadoQQAAAAAAACcR6EEAAAAAAAAQyiUAAAAAAAAYAiFEgAAAAAAAAyhUAIAAAAAAIAhFEoAAAAAAAAwhEIJAAAAAAAAhlAoAQAAAAAAwBAKJQAAAAAAABhCoQQAAAAAAABDKJQAAAAAAABgCIUSAAAAAAAADKFQAgAAAAAAgCEUSgAAAAAAADCEQgkAAAAAAACGUCgBAAAAAADAEAolAAAAAAAAGEKhBAAAAAAAAEMolAAAAAAAAGAIhRIAAAAAAAAMoVACAAAAAACAIRRKAAAAAAAAMIRCCQAAAAAAAIZQKAEAAAAAAMAQCiUAAAAAAAAYQqEEAAAAAAAAQyiUAAAAAAAAYAiFEgAAAAAAAAyhUAIAAAAAAIAhFEoAAAAAAAAwhEIJAAAAAAAAhlAoAQAAAAAAwBAKJQAAAAAAABhCoQQAAAAAAABDKJQAAAAAAABgCIUSAAAAAAAADKFQAgAAAAAAgCENjCz28W6iieNGKzIiXB7ubsrMytYPG2M1e+585eblG3rj1q1a6p47b1VYpw7y9HBXXl6+EhKT9fXKNVoVs87QXgAAAAAAAKg7NhdKAf6+mvnhVHl5eWr9hi1KSExW+7atdNftwxQVEaaJU55XTk6eTXvdMWKwHntkvHLz8vXT5u1Kz8iSu5urWrZopu6RnSmUAAAAAAAA6jGbC6WnHp0oLy9PvTNtlhYuWVk+PmXSWI0ZdaseGn+P3nx/xiX36dY1VI//6X7Fbo/TC6++pYLCogrz9vb2BuIDAAAAAACgrtl0hlKAv6+iIsKVkpqmRUtXVZib9dk8FRQWalB0Xzk6Olxyrz89eJ+Ki0v0yhvvVSqTJOns2bM2RgcAAAAAAIAZbCqUOod1lCTFbouT1WqtMFdQWKTd8fvl5OSokHatq92nZfNmanV9c8Vuj1NOTp46h4Xo7juHa8yoW9U1vKMsFksNLwMAAAAAAAB1xaZH3oKD/CVJx5NSqpxPSk6VIsIVFOSvbTv3XHSfdm1vkCRlZ5/Wv957XeGhHSrMHz5yTM+98qaSUk7YFB4AAAAAAAB1z6ZCycXFWZKUl19Q5XxeXtm4m4tLtfs09vSQJA0bMkDpGVl64rm/Km7PPnk19tT4P47S4Ogb9fbUF3Tv/Y/rzJkzlV4/fGi0hg+LtiUyAAAAAAAAaonNh3JfDnbnHmlrYG+vl19/R/F7D0qSCgoK9ZepHyo4KFDt296gfn2iFLN2Q6XXL1sRo2UrYiRJm9YurrvgAAAAAAAAKGfTGUr55+5Mcj13p9KFXF3LxnPz86vd5/x8RmZ2eZn0v378KVaS1L5tK1tiAQAAAAAAwAQ2FUoJiWVnJzUL9K9yPjDAT5KUmFj1GUvnHU9MliTl5VVdPOXm5kmSHBo1siUWAAAAAAAATGBTobRjV9lB2926hlb6JDZnJ0d1CmmrwsIixe+rfNfR/4rfe1AFhYXyu85Hjo4OleZbtmgmSUo5cdKm8AAAAAAAAKh7NhVKySlp2rx1p/z9fHX7iMEV5iaMHS1nJyetjlmvoqLi8vHgoAAFBwVUWFtcXKKvV66Rg4ODHhx/d4W561s009CB/XTmzBl9/8NPNb0eAAAAAAAA1DKbD+V++4OZmvnhVD05eYK6hndUwvEktW/XuuzrxGRNn/N5hfXz5k6TJHXvf1uF8ZmffqnwTu01+o5bFNK+jXbH75dXYw/d2DtKDg4Oeu+j2UpOSbsMlwYAAAAAAIDaYHOhlJySpnGTntYD48YoKiJcPSI7KyPrlOYvWq7Zc+cr9yLnIl2ooKBQDz36ov54923q37eH7hgxWMUlJYqL36cv/rtMsdvianwxAAAAAAAAqH2WqH4jrWaHqIlNaxebHQEAAAAAAOCqsu/AYY2f9Mwl19l0hhIAAAAAAABwHoUSAAAAAAAADKFQAgAAAAAAgCEUSgAAAAAAADCEQgkAAAAAAACGNDA7AAAAAADAuObPrjA7AlDrjv19qNkRcBHcoQQAAAAAAABDKJQAAAAAAABgCIUSAAAAAAAADKFQAgAAAAAAgCEUSgAAAAAAADCEQgkAAAAAAACGUCgBAAAAAADAEAolAAAAAAAAGEKhBAAAAAAAAEMamB2gxg4ckB580OwUAAAAAGCKeUczzY4A1L7Nb5md4NozY5ZNy7hDCQAAAAAAAIZcuXcotWkjrVtndgoAAAAAMMXoZ1eYHQGodcf+PtTsCNeeA4dtWsYdSgAAAAAAADCEQgkAAAAAAACGUCgBAAAAAADAEAolAAAAAAAAGEKhBAAAAAAAAEMolAAAAAAAAGAIhRIAAAAAAAAMoVACAAAAAACAIRRKAAAAAAAAMIRCCQAAAAAAAIZQKAEAAAAAAMAQCiUAAAAAAAAYQqEEAAAAAAAAQyiUAAAAAAAAYAiFEgAAAAAAAAyhUAIAAAAAAIAhFEoAAAAAAAAwhEIJAAAAAAAAhjQwstjHu4kmjhutyIhwebi7KTMrWz9sjNXsufOVm5dfowBhndrro3dek729vT79zwLNnPNljfYBAAAAAABA3bC5UArw99XMD6fKy8tT6zdsUUJistq3baW7bh+mqIgwTZzyvHJy8gy9ubOTo17682QVFZfIxdnJcHgAAAAAAADUPZsfeXvq0Yny8vLUO9Nm6dmX/6GPP/mPJj/5ir5c8JWCmwXqofH3GH7zx/90v1xdnPXvLxYbfi0AAAAAAADMYVOhFODvq6iIcKWkpmnR0lUV5mZ9Nk8FhYUaFN1Xjo4ONr9x7x4RGjZ4gN79aLbSM7OMpQYAAAAAAIBpbCqUOod1lCTFbouT1WqtMFdQWKTd8fvl5OSokHatbXrTxp4eevbJh7V+wxZ9890PBiMDAAAAAADATDYVSsFB/pKk40kpVc4nJadKkoLOrbuUZ5+cJDs7i/7x3nSb1gMAAAAAAKD+sOlQbhcXZ0lSXn5BlfN5eWXjbi4ul9xr2KD+6tOzm1587W1lZ5+2NackafjQaA0fFm3oNQAAAAAAALi8bP6Ut8vhOl8fPfbIeK1Zt1Fr1v9k+PXLVsRo2YoYSdKmtRzkDQAAAAAAYAabHnnLP3dnkuu5O5Uu5OpaNp6bn1/tPi888ycVF5forfdnGskIAAAAAACAesSmO5QSEsvOTmoWWPUZSYEBfpKkxMSqz1g6r02rlnJzddHqpXOrnB937yiNu3eU1m/Yomdf/oct0QAAAAAAAFDHbCqUduzaI0nq1jVUFoulwie9OTs5qlNIWxUWFil+38Fq91n17To5OjhUGg8K9FN4aAcdPHRU+w8e1cHDR41cAwAAAAAAAOqQTYVSckqaNm/dqaiIcN0+YrAWLllZPjdh7Gg5OzlpyVffqKiouHw8OChAkpSQmFw+9t5Hs6vcf8jAfgoP7aCNW7Zr5pwva3QhAAAAAAAAqBs2H8r99gczNfPDqXpy8gR1De+ohONJat+uddnXicmaPufzCuvnzZ0mSere/7bLmxgAAAAAAACmsrlQSk5J07hJT+uBcWMUFRGuHpGdlZF1SvMXLdfsufOVm1f9gdwAAAAAAAC4OthcKEnSyfRMvfHmRzatNXJn0spvvtfKb743EgUAAAAAAAAmsTM7AAAAAAAAAK4sFEoAAAAAAAAwhEIJAAAAAAAAhlAoAQAAAAAAwBAKJQAAAAAAABhCoQQAAAAAAABDKJQAAAAAAABgCIUSAAAAAAAADKFQAgAAAAAAgCEUSgAAAAAAADCEQgkAAAAAAACGUCgBAAAAAADAEAolAAAAAAAAGEKhBAAAAAAAAEMolAAAAAAAAGAIhRIAAAAAAAAMoVACAAAAAACAIRRKAAAAAAAAMIRCCQAAAAAAAIZQKAEAAAAAAMAQCiUAAAAAAAAYQqEEAAAAAAAAQyiUAAAAAAAAYAiFEgAAAAAAAAyhUAIAAAAAAIAhFEoAAAAAAAAwhEIJAAAAAAAAhlAoAQAAAAAAwBAKJQAAAAAAABhCoQQAAAAAAABDKJQAAAAAAABgCIUSAAAAAAAADKFQAgAAAAAAgCEUSgAAAAAAADCEQgkAAAAAAACGUCgBAAAAAADAEAolAAAAAAAAGNLAyGIf7yaaOG60IiPC5eHupsysbP2wMVaz585Xbl7+JV/v6OigPj0j1TOqs9q0aqmmPt4qtVp1PDFZMWs3aMGSlTpz5kyNLwYAAAAAAAC1z+ZCKcDfVzM/nCovL0+t37BFCYnJat+2le66fZiiIsI0ccrzysnJq3aPsI7t9doLj+n06Vxt3xWv9Rtj5e7qql49IjRl0ljd2DtKk598RSW//vq7LwwAAAAAAAC1w+ZC6alHJ8rLy1PvTJulhUtWlo9PmTRWY0bdqofG36M3359R7R6ZWdl65Y33tXb9TxXuRJo2/TP9873X1SmkrW4fMVhfLviqBpcCAAAAAACAumDTGUoB/r6KighXSmqaFi1dVWFu1mfzVFBYqEHRfeXo6FDtPoeOHNO3a36o9FhbQWGRvvxvWYnUOayDkfwAAAAAAACoYzYVSp3DOkqSYrfFyWq1VpgrKCzS7vj9cnJyVEi71jUOcuZsWcl09mxpjfcAAAAAAABA7bOpUAoO8pckHU9KqXI+KTlVkhR0bl1NDBs0QJK0OXZnjfcAAAAAAABA7bPpDCUXF2dJUl5+QZXzeXll424uLjUKcceIweoe2VkHDx3V16vWXHTd8KHRGj4sukbvAQAAAAAAgMvD5kO5a0vf3pF69JHxysjM1nOvvqWzZ89edO2yFTFatiJGkrRp7eK6iggAAAAAAID/YdMjb/nn7kxyPXen0oVcXcvGc/PzDb15n57d9PqLTyg7+7QeeeIlpaSmGXo9AAAAAAAA6p5NhVJCYtnZSc0Cqz4jKTDAT5KUmFj1GUtV6d+3u9545SllZZ/Ww4+/pOMGXgsAAAAAAADz2PTI245deyRJ3bqGymKxVPikN2cnR3UKaavCwiLF7zto05vePKCPXnp2stIzsvSnJ17mziQAAAAAAIAriE13KCWnpGnz1p3y9/PV7SMGV5ibMHa0nJ2ctDpmvYqKisvHg4MCFBwUUGmvITffqJefnay0tHQ9/NiLlEkAAAAAAABXGJsP5X77g5ma+eFUPTl5grqGd1TC8SS1b9e67OvEZE2f83mF9fPmTpMkde9/W/lY57AQPf/0I7K3t9eOXfEaOqh/pffJy8vX/EXLa3o9AAAAAAAAqGU2F0rJKWkaN+lpPTBujKIiwtUjsrMysk5p/qLlmj13vnLzLn0g93W+PrK3t5ck3TLkpirXpJ44edkLpeLiYqWkpOpkerpKz5Ze1r1x5bCzt1NTHx/5+/vJwcHB7DgAAAAAAFyxbC6UJOlkeqbeePMjm9b+751J56385nut/OZ7I2/5uxUXFysubo/Wb9yqhMQU5RcU1en7o/5wcXZUcJC/+vaMUGhoR0olAAAAAABqyFChdCVKSUnV+o1btffAUbOjwGT5BUXlfw58fLzVokVzM+MAAAAAAHDFsulQ7ivZyfR0JSSmmB0D9UhCYopOpqebHQMAAAAAgCvWVV8olZ4t5TE3VJBfUKTSUs7SAgAAAACgpq76QgmoktXsAAAAAAAAXLkolAAAAAAAAGAIhRIAAAAAAAAMoVACAAAAAACAIQ3MDmC27z1vNDtCtfqdWndZ9mnSpLHefuN5bdi0TbPnzr8se5qtZ/eumnDfXZo1d742btpmdhwAAAAAAK4Z3KEEAAAAAAAAQyiUAAAAAAAAYMg1/8jbtWD4sGiNGHazJKlX967q1b1r+dysufO1OXanbuwdqU4hbeXvd5083N1UXFKihOPJ+ua79drz84FKe771xnOSpJdef1cjht2sLuEh8vT00PJVa7RseYwkycPdTbePGKzQkHZydHTQibR0fbvmB2VkZevZJyZp6fJvy9ee5+LspME336jw0A7ybuKlM2fP6FhCklZ+s04/7ztYvu7PTzyktq2vlyRNuO8uTbjvrvK5p174mzIzsy/TPz0AAAAAAHAhCqVrwP6DR/Ttmh9184DeOp6Yoh1x8eVzxxNT5OLirLvvHK7DR45p776Dys3Ll4eHu8I6ttMTkyfo038v0A8bYyvta29vr2cef1Auzs6K33tQRUXFysjIkiS5ubnoxT//Sd5NvLT/4BEdOZogd3c3/WHMbYr/n2LofzXx8tSfn5gkH28vHTh0VHv2HpBDo0YK7dhOT0y+X3O/WKQfNpTl2LBpmwoKCtU5LEQ7dsXreFJK+T4FBYWX8x8fAAAAAAC4AIXSNeDAwaPKyMwuK5SSUirdFdSggb2eev5vyj51usK4k6Ojnn/6EY26bag2xe7Qr7+eqTDf2NNDKakn9fd3PlZJya8V5u4YMUTeTby08pvvtWDJyvLxmDU/6qVnp1SZc8LY0Wri5amPZ/1Hsdvifsvh5Khnn5ike+4coV1xe5WTm1d+CHfnsBDtiPuZQ7kBAAAAAKhDnKEEnTlztlKZJEmFRUX68aetcnVxVovgoCpfO3/h15XKJHt7e0VGhKmgoFBfr1xTYS4xOVU/bdleaZ+gAD+1bX29tu/cU6FMkqTCwiItXf6tGjVqqC6dOxq9PAAAAAAAcJlxhxIkSf5+vhp8841qfUMLeXq4q1GjhhXmG3t6VHpNScmvSkxOrTTu5+sjh0aNdCDhqIqKiyvNHzr8i/r2iqwwdn3LYEmSk5OThg+LrvQaN1fXspzXNbX9ogAAAAAAQK2gUIJatmimZx5/UPZ2dtq7/7B27d6rwqIiWa1WNQv0V+ewEDVoWPmPSk5uXpX7OTk5ls3n5FY5fzqn8utcXZ0lSSHtWyukfeuLZnVwcLjk9QAAAAAAgNpFoQTdMniAHBo10t/f/VgHDh6tMDd0YD91Dgu5yCutVY4WFpXdleTu7lblvIe7a+XXFBZJkj6fv1Tffb/RxuQAAAAAAMAMnKF0jbCWlkqS7CyWSnO+Tb2Vl5dfqUySpDatrzf8XqknTqq4pERBAX5yrOKOolY3tKg0duTocUlS6yrmLqa0mmsCAAAAAAC1h0LpGpFfUKjS0lJ5eXlWmsvIzJarq4sCA/wqjPfuEaGOHdoYfq+zZ88qdlucnJ2ddMuQARXmggL81COyS6XXHDuepAOHjqpLeEf17hFR5b6B/tfJzc3lt2vKL5AkNfFqbDgjAAAAAACoOR55u0YUF5fo6LFEtb6hhSaOH6O0tAyVlpZq5+69+nbNj+rYoY2ef+phbd0ep4LCIjUPDlTrG1po6/Y4RXQJNfx+C5esVLs2N2jIwH5q2aKZDh85Jg8Pd3XrEqrdP+9Xl7AQWUsrPjI3Y/YXeubxBzX+j3fqpn69dPSX4yooLFTjxh4KCvBTYICfXv/HNOXm5kuSDh9NUHFxiaIH9JKrq7NOny47s+m77zeqsKjo9/9DAwAAAAAAVaJQuoZ88umXGjPqVnVs30aRXcNkZ2enrFOntXHTNr3/zzm6ZfAARXQNlbXUqqPHEvWPd6fLx6dJjQqlnNw8vfHWR7pj+GB1Cmmrls2b6URauv795WIVl5SoS1hIpdIn+9Rpvfa3DzSgX091De+oqG6BTIlTAAAfYElEQVThsrOz0+mcXKWkpum77zcq6X8+Va6goFAfzfg/DR8WrZ5RXeXoWPZ43U+xOyiUAAAAAACoRdd8odTv1DqzI9SZk+mZ+uBfn1Y5F7dnn+L27Ks0fvDwL9q4aVul8adfmHrJ9zt1Kkez5s6vNH7b8EGSys5aulBRcbFWrF6rFavXXnJ/SYrfe0Dxew/YtBYAAAAAAFwenKGEWuPp4V5pLND/Ot3Ur6fy8vK1v4pDwAEAAAAAQP13zd+hhNrz8nNTdDI9U8kpJ1RcXCLfpt7q1LGd7CwWffL5lzpz5ozZEQEAAAAAQA1QKKHWrP9xi8JDOyiya5gcHR1UUFik+J8PaPV363WAu5MAAAAAALhiUSih1ixbEaNlK2LMjgEAAAAAAC4zzlACAAAAAACAIRRKAAAAAAAAMIRCCQAAAAAAAIZQKOHaZDE7AAAAAAAAV66rvlCys7eTi7Oj2TFQj7g4O8rO7qr/ow8AAAAAQK256n+rburjo+Agf7NjoB4JDvJXUx8fs2MAAAAAAHDFamB2gNrm7++nvj0jJEkJiSnKLygyORHM4uLsqOAgf/XtGSF/fz+z4wAAAAAAcMW66gslBwcHhYZ2lI+Pt06mp6u0tFSymp0Kdc4i2dnZqamPj/z9/eTg4GB2IgAAAAAArlhXfaEklZVKLVo0V4sWzU3NAQAAAAAAcDW46s9QAgAAAAAAwOVFoQQAAAAAAABDDD3y5uPdRBPHjVZkRLg83N2UmZWtHzbGavbc+crNy7d5H3c3V43/453q07Obmng11umcXG3ZulMzP52n9IxMwxcBAAAAAACAumNzoRTg76uZH06Vl5en1m/YooTEZLVv20p33T5MURFhmjjleeXk5F1yH3d3V82cNlXBQQHatmO3Yr7foOCgAA0bPEA9IrvogcnPKSU17XddFAAAAAAAAGqPzYXSU49OlJeXp96ZNksLl6wsH58yaazGjLpVD42/R2++P+OS+0y6/14FBwXoi/9+pWnTPysfHzVyiJ6YPEFPPzpRjz/7urGrAAAAAAAAQJ2x6QylAH9fRUWEKyU1TYuWrqowN+uzeSooLNSg6L5ydKz+o9idHB01KLqvCgoLNXvuvApzC5euUuqJk4rqFi5/P1+DlwEAAAAAAIC6YlOh1DmsoyQpdlucrFZrhbmCwiLtjt8vJydHhbRrXe0+Ie1by9HRQbvj96ugsKjCnNVq1ZatuyRJXcJCbL4AAAAAAAAA1C2bCqXgIH9J0vGklCrnk5JTJUlB59ZdTLNz84lJqVXOJ57fJ7D6fQAAAAAAAGAem85QcnFxliTl5RdUOZ+XVzbu5uJS7T6u5+bz8qv+RLjz426uVe8zfGi0hg+LliQVFBQqITH5EsmBK5unh7tOnc4xOwYA4Hfi5zmA2rBqXBuzI1xz+Hle9/YdOGx2hGuOn29Tm9bZfCh3fbBsRYyWrYgxOwZQZ+Z8/KbGT3rG7BgAgN+Jn+cAcHXg5znwG5seecs/d2eS67k7lS7k6lo2nnuRO4/OO38HkutF7mQ6P56bV/0+AAAAAAAAMI9NhVJCYtnZSc0ucrZRYICfJCkxseozls47fm4+KNCvyvmg8/tc5KwmAAAAAAAAmM+mQmnHrj2SpG5dQ2WxWCrMOTs5qlNIWxUWFil+38Fq94nfe1BFRcXqFNJWzk6OFeYsFou6dQ2VJG3fFW/zBQBXs2XLecQTAK4G/DwHgKsDP8+B39hUKCWnpGnz1p3y9/PV7SMGV5ibMHa0nJ2ctDpmvYqKisvHg4MCFBwUUGFtYVGRVsesl7OTk+6/b3SFuTtGDJa/n682x+5USmpaTa8HuKpwZhgAXB34eQ4AVwd+ngO/sUT1G2m1ZWGAv69mfjhVXl6eWr9hixKOJ6l9u9bqGt5RCYnJmjj5OeXk5JWv37R2sSSpe//bKuzj7u6qmdOmKjgoQNt27Nbe/YcU3CxQfXtFKivrlCZOeU7JKRRKAAAAAAAA9ZXNhZIkNfVpogfGjVFURLg83F2VkXVKP2zYotlz51c6SPtihZIkubu5avwf71SfXpHy9vLU6Zw8bY7doZmfzlN6RubvvCQAAAAAAADUJkOFEgAAAAAAAGDTGUoAAAAAAADAeRRKAAAAAABcwN7e3uwIQL3WwOwAAKrm4NBI7m6usrOruvdNO5lRx4kAAACAa8dX/52lFavX6qsVMUpKOWF2HKDe4QwloJ4ZFN1X944eqebNAi66xiqpd/SougsFAKhWWKf2NX7trt17L2MSAMDlsmrJZ3J3c5XVatX2nfFa8vU3+mFjrEpLS82OBtQL3KEE1CNDBvbTC08/otLSUsXt2ae09AydPct/sACgvvvnu3+R1Vqzv6PrxV8QAEC9dMuoCerfp7uGD4tW184d1SU8RNmncrR81RotWxGjE2npZkcETMUdSkA98p/Z78vbq7EefPR5JRxPNjsOAMBG9993V40LpTn/99/LnAYAcLkFBfppxLCBGnxzX3m4u6nUalXstl1a+vW32rBpW43/GwBcySiUgHpk3ep5Wrn6e735/gyzowAAAAC4QIMGDdS/T3fdOvQmhYd2kNVqVWbWKX218jstW/6tMjKzzY4I1Bk+5Q2oR3Jy8lTy669mxwAAAABQhTNnzmjjlu1av2GLMjKzZbFY5N2kscb/YZQWfT5djz0yXg0bcrIMrg38SQfqkY2bt6lzWIjZMQAAAABcoEO71hpxy80a0LeHGjVqqPyCQi1YslJfr1qjNq1aaswdt+iOEYPVsGEDvfX+TLPjArWOR96AesTd3VUzP/ybdsbt1Ycff6bCoiKzIwEAbDDtnddq9kKrVZOfevWyZgEAXD7OTo4aFH2jRtxys1o2D5LFYtHBw79o8bLV+nbNjyouKSlfa2dnp/f+/pJa3dBcQ24bZ2JqoG5whxJQj7zxytMqKi7RLUMGKHpAbyUlpyo3L7/yQn4BAYB6pXNohxq9jkNcAaD+ev6ph9X/xp5ydGikX389o9Ux67X4q9Xau/9wletLS0u1Iy5eXcJ54gDXBgoloB75319InBwd1Or65lWu4xcQAKhfet50h9kRAACX2dBB/ZWcckJLvv5WK1avVU5u3iVfs3PXz5rz7wV1kA4wH4+8AQAAAABwgW5dQxW7Lc7sGEC9xae8AQAAAABwAcokoHo88gYAAFCLfLybyMfbS40aNaxyftfuvXWcCABgi67hHTXwpj6aMecLZWRmV5r3btJYD46/Wyu/XaedcT+bkBAwF4USUA/d1K+nbh1yk1rf0FIurs4qyC/Q/oNH9fWq7/Td9xvNjgcAsEG3rqF69OFxCg4KqHZdr+hRdZQIAGDEHSOHqHmzgCrLJEnKyMxWSPs2cnFxplDCNYlCCahnXn5uim7u31sWi0WlpaU6dSpHnh5u6tq5o7qEh6hX9wi9+rf3zY4JAKhGh3at9fYbz+vU6RwtXLpKo0YO0c64n3U8MUWhndqpebNA/fjTVh08/IvZUQEAF9GmVUtt27G72jVx8fvUrWtYHSUC6hcKJaAeGXnLzRo4oI/2Hzyqf33yf9oZt1elpaWys7NTeGh7PfzAH3RTv56K27NXS77+1uy4AICL+OPdt6mk5FeNn/SMMjKzNWrkEO3YFV/+yT8PjB2t0aNu0Yw5X5icFABwMY0beyg9M6vaNVnZp+Tl6V5HiYD6hUO5gXpk2OABSj1xUg8/9qK274xXaWmpJKm0tFTbd8br4cdeUmpauoYNvsnkpACA6oS0b6Mff9pa4TEJi91v/9v1yWfzdCwhWRPuG21GPACADfLzCuTr413tGl8fbxUWFddRIqB+oVAC6pHmwYFavyFWxSUlVc4Xl5Tohw2xah5c/XkcAABzubo4K+1kRvn3v545IydHhwpr9vy8X2Gd2td1NACAjfYeOKQ+PbvJq7FnlfPeTRqrT89u2rv/UB0nA+oHCiWgHrFaJYul+jWXmgcAmC/71Gm5ubn89n32aQX4X1dhTQN7ezk4NKrraAAAGy1YslLOzk6a/sFf1atHhBo2LDsxpmHDBurdI0Ifv/9XOTk5asHilSYnBczBGUpAPZJwPEk39o7SjNlfVHmXkkOjRurTs5uOJSSZkA4AYKvEpJQKBdLP+w4qqltnBQX6KTEpVV6NPXVjnyglJaeamBIAUJ3YbXH69N8LNO4Po/T3156RVVJubp7c3FxlkWSxWPTpvxdo89adZkcFTGEf2KLdq2aHAFDG3s5ONw/orciIcKWknFDayQxZrdZzh3J30CvPPaoWwYGa+/lC7T94xOy4AICL8GrsoWGDBmjxstUqLilRRka2bh0WrVuH3KQ+vSJ1/313ys3VRR9/8rkOHz1mdlwAwEXsiPtZ8XsPyN3dVe5ubnJzdVZubr527IrXu9Nma/mqNWZHBExjieo30mp2CAC/efX5xxTdv5esVqusVqtycvLk7u4qi8Uii8WiNet+0st/fdfsmACAajg7O6l5s0AdS0hUQWGRJKlvr0g9MG6MAv19lXoiXfMWfq1lK2JMTgoAAFAzFEpAPRTdv5eGDRqg1q1ayNXFWXn5BTp46BctX71GMWs3mB0PAAAAAHCNo1ACAAAAAACAIRzKDQAAcJm1adVSPaO6aMnyb5WdfbrSvFdjT40YFq0ff9qqQ0eO1X1AAIBNmng11rh771BkRJh8vL3UoEHlX6GtknpHj6r7cIDJ7MwOAAAAcLW5+85bdcuQm6oskyQpK/uUhg25SWNG3VrHyQAAtvLx9tKn09/U8FtuVmFRsRo2bKi0kxlKSk5VaWmpLBaLDh9NUNzuvWZHBUzBHUqAiTZ+t1ClVqvuHjdFiUmp2vjdQlmtl34Klb8FAYD6LaR9G+3YFV/tmu079yisU/s6SgQAMGr8H+6UV2NPPf7n17V1x25t/G6hVqxeqzn/XiAf7yZ67slJ8ruuqSa/8orZUQFTUCgBJtq1e6+sVquKikoqfA8AuLI18fLUyfTMatdkZGTJ26txHSUCABgVGRGmzVt3aeuO3ZXm0jMy9cJrb+vzOe9rwtjReu+j2SYkBMxFoQSY6JEnXq72ewDAlamoqESenu7VrvH0dFfJr7/WUSIAgFFeXp5as25j+felpaVycGhU/n1hUZFit8epd89uFEq4JnGGEgAAwGV26Mgv6tOjm5wcHaucd3Z2Up8e3XTo8C91nAwAYKv8/EI1bNiw/Puc3Hz5eDe5YE2BGntU/xcIwNWKQgm4Ari7u6pvr0hFdg2TnR3/2gJAfbd0eYw8Pd31wVuv6PqWwRXmbmjZXB+8+Yo8PNy0bEWMSQkBAJdyIi1dTX1+K5AOHzmmLuEdy+9Sslgs6tYl9JKPOANXKx55A+qRkbcO1NCB/fTEs39VTm6epLKPnn7vHy/J3c1VkrT/4BH96clXVFRUbGZUAEA11qzbqO7dwjX45hs1d8bbyso+rfSMLPl4e8mrsYcsFotWfbtOMWs3mB0VAHAR23bu1oih0bK3t9fZs2e18tvv9fKzUzRz2lRt3R6nTiHt1KJ5kP7vi8VmRwVMYR/Yot2rZocAUGbyg/fJ3t5eC5euLB979YXHFNwsQCtXf6/MrGx1Dg1Rbm6+4vceMDEpAOBSftgYq8zMbAX4+crfz1feTRrL2clRvxxL1IxPv9Ssz+aZHREAUI0TaenKyc1TUnKqCguLdORogtxcXdSze1d16tBWTX2aaM26jZr28VydLS01Oy5Q5yxR/UbykVJAPfHVfz/Rxs3b9Y93p0uSPNzdtGLRHH29co3+8V7Z2Kx//l329vYa99DTZkYFABjg4NBIbq4uys3LV3FxidlxAAC/g6eHu/z9fJWadlLZ2afNjgOYhsNYgHrE3d1N2ad++49Sp5C2kqT1G7aUj8Xt2afrfH3qPBsAoOaKi0uUkZlNmQQAV5DB0TcqsmtYpfFTp3O0d/8hyiRc8yiUgHokJydPnu6/fUpEeGgHWa1W7f55f/mY1WpVo0YNq3o5AKCe8fRw18hbbtZjj4zXc08+XGG8fdsb5NCoUTWvBgCY6flnHlFUt3CzYwD1FoUSUI8kHE9Srx5d5e7uKlcXZ93Ur5f2HTisgoLC8jV+1zVVVtYpE1MCAGxxy+ABWvzFdD055QGNGjlEQwf1K5/zauypmdOm6uYBvU1MCACoTlbWKVksFrNjAPUWhRJQj8xfvEJNvBrrq/mfaOn8T+TV2EOLlq2usKZDu9Y6dOSYOQEBADaJ6BKqPz/xkBKTUvTsK29qyVffVJg/euy4fjmWqD69upmUEABwKZtjd6pLWAilEnARFEpAPbLhp616870ZOnosUccTk/XRjLn65rsfyufDQzvIyclRW7buMjElAOBS/jB6hDKzTmnS4y9pw09blXWq8jkbh48mqHlwkAnpAAC2mD7nczk7O+n5px+Rh7ub2XGAeqeB2QEAVLRsRYyWrYipcm5n3M8aOPyPdZwIAGBU2zY36Lu1Gyo8snyhkxmZatLYsw5TAQCM+MuLTygvv0CDo/squl9PpZ5IV2ZWtip9TLrVqslPvWpCQsBcFEoAAACXWcMGDVRYVFztGjcXF5WWltZRIgCAUZ1DO5R/3bBhQzUL8lezIP9K66zWShUTcE2gUAJM5NvUW5KUnpGl0tLS8u9tkXYyo7ZiAQB+p9QTJ9W2dctq13Ro10oJiSl1lAgAYFTPm+4wOwJQr1EoASZa/MV0Wa1WjRk3RYlJqeXfX4pVUu/oUbUfEABQIz/+FKt77hqh/n27a+36TZXmhw7qr+tbBmvGnC9MSAcAAPD7USgBJlr17TpZJeXlF1T4HgBwZfvPvKW6qV8vvfbiE+rXZ7NcXZwlSXeMGKzQju11Y+9IJSWnasGSlSYnBQAAqBlLVL+R/P4KAABwmfk29dbLz05RWKf2lebi9uzTK2+8p/SMLBOSAQBsUdXP74vZtXtvLSYB6icKJQAAgMssrFN75ecX6NCRY7q+ZbA6tm8jD3c35eXnK37vQR04dNTsiACAS9j43UKbD9zuxXEUuAbxyBtQj3h6uKt5cKAOHjqqgsKiSvPOzk5qfUML/XIsUadzck1ICACwxUfvvKaly2P09gczdeRogo4cTTA7EgDAoDn/XlBloeTm6qJ2bW5Qxw5ttGHTNv6SANcsCiWgHhl77x0aOrCfho26v8r50tJSvfXX5/TVyjWaNv2zug0HALDZ6dO5Ki4uMTsGAOB3mD13frXzQwb20xOTJ2jGbD5gAdcmO7MDAPhNty6hit2++6K/hBQVFSt2e5wiI8LqOBkAwIgdcT+rY4c2ZscAANSild98r5/3HtRDE+4xOwpgCgoloB5p2tRbySknql2TnJImX58mdZQIAFATM+Z8oeAgf00cN0b29vZmxwEA1JJDR34xdHg3cDXhkTegPrFa1bBh9f9aNmzYQHb2dMEAUJ/dd/dtOnLsuP54920aNniADh85pszsU9IFZ3FYrVb97e1/mZQSAPB7NfXxlj3/b45rFIUSUI8kJKYosmv1j7NFdg1TUnL1dzEBAMw1ZGC/8q+beHmqiVfVP9splADgymRnZ6ehg/qrX9/u2r1nn9lxAFNQKAH1yPc/bNKkCffoySkT9NH0/1NxyW9nKTk0aqTJD92nZkH+mjnnSxNTAgAu5fZ7JpkdAQDwOy38T9WFv729vbwae8re3k5nzpzR9Nmf13EyoH6gUALqkQWLVyi6fy+NvGWg+vSM1K7dPys9I0s+3l4K69RB3k0a6/CRY5q/aLnZUQEA1TiRlm52BADA72RnZyfrBY8qS9KZs2d15JcE7d1/WAuWrFDC8WQT0gHms0T1G1n53xAApnF1cdbTj01U/xt7ys5iKR8vtVr13doNeufDT5SXX2BiQgAAAADAtY5CCainPD3c1a7tDXJzcVFufr727juk0zm5ZscCAAAAAIBCCcD/t3f/wTWdeRzHP5ckQhISSZBqSEKSIkRS6keRhkXHdpF2UGx0q2pr7U7Z6Y8dm93V1trubltmO8tO7XRpVVE1MjtV9aOySIsgaRLRhMYWI25+IjeoRM7+YYqIZqvT3ueM8379lXuf58y8/8z93nueAwAAAOBmbfz8FBzcXlXVZ9XQ0NBs3dfXRx1DglVTc06X6+sNFAJm8XxDwKa6R3ZVyrBBenB0iukUAAAAwHFmzpikdateV7u2/rdcb+vvr7UrX9dj0x/xchlgDwyUAJuJ7RGlN5f/Re+8uVR//MMzynjul9fWkvr11s7N72rYkAEGCwEAAIA73+D7kpVzKF/naz23XD9f61HOoXzdz//mcCgGSoCNRN4dob8veUndIrtq/cYPtHd/bpP13Pwina/1KHXEEEOFAAAAgDNEdOmkEyfLWtxz8tRpRXQO91IRYC8MlAAbeWLGFPn6+GjW3Of1t+UrVVR8rNmewqJi9YrvaaAOAAAAcA6f1q1lWY0t7rEsS35+fl4qAuyFgRJgIwOS+ypr917998tT37jHXV6psLCOXqwCAAAAnOd0mVtJiX1a3JOUmKAz7govFQH2wkAJsJGgwECVV1a1uMflcsnXx8dLRQAAAIAz7f4kR/GxMZo+ZeIt19Onpik+Nlq7svd5uQywBz6VAjZSXXNWd98V0eKe6KhIuSsqvVQEAAAAONOa9Zka+6MRmjNrukY9MFT7D3ymisoqhYeFatDA/ortESV3eaVWr91kOhUwgoESYCMHcws0euRwdYu8SydOnm623iu+pwYk9dX7mVsM1AEAAADOUeup0y/m/04vZsxXn15xio+NkWVZcrlckqSCw8VauHipaj11hksBMxgoATby1rsbNTJlqJYvWaR/vrVO4aFXz0qKjopU/3699cSMybpw8ZLWrM80XAoAAADc+c64KzT7VwsUFxujhF5xCgoMUK2nToVHSlRytNR0HmCUa3BqmmU6AsB1gwcm6YWM+Qpo11bS1TOTvv4mxOOp04IX/qqDuYWGKwEAAAAATsZACbChwIB2Gjc2VX16x6lD+yDV1V1QYVGJPtjysc7XekznAQAAAHe8Nn5+Cg5ur6rqs2poaGi27uvro44hwaqpOafL9fUGCgGzGCgBAAAAAHCTObOma8ojD2n85Cdv+aVu+6BAZa5boTXrM7Vi5VoDhYBZrUwHALhuz7b3tHDBPNMZAAAAgOMNvi9ZOYfyv/EOgfO1HuUcytf9QwZ4uQywBwZKgI1cuHhJ7vIK0xkAAACA40V06aQTJ8ta3HPy1GlFdA73UhFgLwyUABspOVqqqO6RpjMAAAAAx/Np3VqW1djiHsuy5Ofn56UiwF4YKAE2snrtJg0dlKyB9yaaTgEAAAAc7XSZW0mJfVrck5SYoDNu7jCAM/mYDgBwXUhwB+3NydVrL2do1579OlJ8TNXVNbJucXT+h9uyvN4HAAAAOMXuT3KUPjVN06dM1DvrNjVbT5+apvjY6FuuAU7AU94AG8nevkGWZcnlcjV537phouRyuWRZloaNnuTtPAAAAMAxggIDtOqNV9UpPFQlx45r/4HPVFFZpfCwUA0a2F+xPaLkLq/Uz37+jGo9daZzAa9joATYyLgxD3zrvZu3Zv1gHQAAAACkLp3D9WLGfPXpFSdJTb78LThcrIWLl3LLGxyLgRIAAAAAAC2Ii41RQq84BQUGqNZTp8IjJSo5Wmo6CzCKgRJgQ239/ZUyfJDiekYrMKCdPHUXVHy0VLv27NfFS5dM5wEAAAAAHI5DuQGbGZkyRM/Ne0qBge2anKVkWZY8cy/oz0v+oZ27PjVYCAAAADhHaMcQDUjuq/CwUPn5Nv8IbVnSv1a/Z6AMMItfKAE2MvDeRL32coasxkZ9tGO3cvMKVVV9VqEdg5XcP0FjRg2Xy+XSr3+zSDmH8k3nAgAAAHe0WY9NUfrUh9W6datr7339kJwb/+aBOXAifqEE2MjM9Emqv1yvp+ZlNLsne/PWLG3Y9KGWL12kx2dMYqAEAAAA/IDGjBqhx9Mn6WBugd7P3KLFC5/V5o92at+BPCUnJugn40ZpR1a2Nv17q+lUwAgGSoCNxMVGa0dW9jce8Pd5yRfakZWt1JQhXi4DAAAAnOXh8WNVXlGl+c+/pCuNjZKksjPl2r4zW9t3Zus/e/bplT/9Vts+3mO4FDCj1f/fAsBb6i83qLKqpsU9lVU1qr/c4KUiAAAAwJl6RHfTp/sOXRsmSVKrVtc/Qu87kKd9OXmaNmWCiTzAOAZKgI3kFRSpX8I9Le7pl3CP8gqKvFQEAAAAOJOPj4/Ona+99vqrry4rMDCgyZ7S4ycUGxPl5TLAHhgoATaybMXb6hnTXXOe/Kn8/ds0WfP3b6O5s9MVE9VNy1a8bagQAAAAcIbK6hqFhYZce+0ur1SPmO5N9oSFhejKlcabLwUcgTOUABtJfzRNx0q/1PTJEzTxx6NVfLRU1TXn1DGkg+JjYxQQ0E55+UVKfzStyXWWZWnxK8sMVQMAAAB3npKjxxUT3e3a64O5BZrw0Gg9ODpFWbv3KjkxQSNHDFF+4ecGKwFzXINT0yzTEQCuyt6+4Ttdx6NKAQAAgO/XuLGpevbp2Zo282mVnSlXp/BQrXrjVQXdcNtbw5Urmjv/9zp8pMRgKWAGAyXARrp0Dv/O155xV3yPJQAAAABuFtGlk6ZNHq+uEV1U5i7Xxswt+uL4CdNZgBEMlAAAAAAAAHBbOJQbAAAAAAAAt4WBEgAAAAAAAG4LAyUAAAAAAADcFgZKAAAAAAAAuC0MlAAAAAAAAHBb/geZxzj6vrsugAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "probabilities = model.predict(f_in(X_prod))\n",
    "pred = f_out(probabilities)\n",
    "_= xai.metrics_plot(pred, y_prod, exclude_metrics=[\"auc\", \"specificity\", \"f1\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x504 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "xai.confusion_matrix_plot(y_prod, pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "[Text(0, 0, 'Rejected'), Text(0, 0, 'Approved')]"
      ],
      "text/plain": [
       "[Text(0, 0, 'Rejected'), Text(0, 0, 'Approved')]"
      ]
     },
     "execution_count": 167,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x504 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(1,2)\n",
    "a = sn.countplot(y_valid, ax=ax[0]); a.set_title(\"TRAINING DATA\"); a.set_xticklabels([\"Rejected\", \"Approved\"])\n",
    "a = sn.countplot(y_prod, ax=ax[1]); a.set_title(\"PRODUCTION\"); a.set_xticklabels([\"Rejected\", \"Approved\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Undesired bias and explainability\n",
    "\n",
    "<br>\n",
    "\n",
    "<img src=\"https://i.imgflip.com/33o3on.gif\" style=\"height: 50vh\">\n",
    "\n",
    "<br>\n",
    "\n",
    "#### We have seen several examples where undesired biases have led to undesired results. In critical usecases, this may have impact of multiple generations.\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "\n",
    "<br><br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Undesired bias and explainability\n",
    "\n",
    "<br>\n",
    "\n",
    "* Has become popular due to several high profile incidents:\n",
    "\n",
    "    * Amazon's \"sexist\" recruitment tool\n",
    "    * Microsoft's \"racist\" chatbot\n",
    "    * Negative discrimination in automated sentencing\n",
    "    * Black box models + complex patterns can't be interpretable\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "\n",
    "## Organisations cannot take on unknown risks\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# This challenge goes beyond the algorithms\n",
    "\n",
    "<br>\n",
    "\n",
    "### A large ethical decision should not just fall on the shoulders of a single data scientist\n",
    "\n",
    "<br>\n",
    "\n",
    "<img src=\"images/chart.png\" style=\"height:30vw;margin-left: 10vw; float: left; background-color: transparent\">\n",
    "<img src=\"images/chart-more.png\" style=\"height:30vw; margin-left: 10vw; float: left;  background-color: transparent\">\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "\n",
    "<br><br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Who's ethics?\n",
    "<br>\n",
    "\n",
    "### An analysis of eastern and western philosophy, in respect to AI Ethics\n",
    "\n",
    "<br>\n",
    "\n",
    "<div style=\"float: left; width: 50%\">\n",
    "<img style=\"\" src=\"images/philosophers.jpg\">\n",
    "</div>\n",
    "\n",
    "<div style=\"float: left; width: 50%\">\n",
    "<img src=\"images/whos-ethics.jpg\">\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Stages where bias can appear\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "## 1) Statistical bias (During the project)\n",
    "\n",
    "##### Sub-optimal choices on decisions after project starts (models, metrics, human-in-the-loop, infrastructure design, etc)\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "## 2) A-priori bias (Before the project)\n",
    "\n",
    "##### Limitations around the project that constrain the best possible outcome (limited time, budget, data, societal shifts in perception, etc)\n",
    "\n",
    "<br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Not as easy as just \"removing bias\"\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "* Any non trivial decision holds a bias, without exceptions - unless you build a classifier that only predicts \"maybe\".\n",
    "* It's impossible to \"just remove bias\" (as the whole purpose of ML is to discriminate towards the right answer)\n",
    "* Societal bias carries an inherent bias - what may be \"racist\" for one person, may not be for another group or geography\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "#### Emphasis on last point: Societal bias is inherently biased\n",
    "\n",
    "<br><br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# What it's about: Mitigating undesired bias through process and explainability techniques\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "* Like in cybersecurity, it's impossible to build a system that will never be hacked\n",
    "* But it's possible to build a system and introduce processes that ensure a reasonable level of security\n",
    "* Similarly we want to introduce processes that allow us to remove a reasonable level of undesired biases\n",
    "* This is going 0 to 1, trying to introduce a foundation for undesired bias\n",
    "\n",
    "<br><br><br>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Principles, Standards & Regulation\n",
    "\n",
    "### github.com/EthicalML/awesome-artificial-intelligence-guidelines\n",
    "\n",
    "<img src=\"images/guidelines.jpg\">\n",
    "<br>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Remember our workflow? Let's add three new steps:\n",
    "\n",
    "<br>\n",
    "\n",
    "<img src=\"images/mlall.png\" style=\"width=100vw\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Augmenting the data science workflow\n",
    "\n",
    "<br>\n",
    "\n",
    "<img src=\"images/gml.png\" style=\"width=100vw\">\n",
    "\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "<hr>\n",
    "\n",
    "<br><br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## The explainability tradeoff\n",
    "\n",
    "\n",
    "<br>\n",
    "\n",
    "By making introducing processes that allow us to mitigate undesired bias and increase explainability, we face severall tradeoffs:\n",
    "\n",
    "\n",
    "<br>\n",
    "\n",
    "* More redtape introduced\n",
    "\n",
    "<br>\n",
    "\n",
    "* Potentially lower accuracy \n",
    "\n",
    "<br>\n",
    "\n",
    "* Constrains on models that can be used\n",
    "\n",
    "<br>\n",
    "\n",
    "* Increase in infrastructure complexity\n",
    "\n",
    "<br>\n",
    "\n",
    "* Requirement of domain expert knowledge intersection\n",
    "\n",
    "<br>\n",
    "\n",
    "#### The amount of explainability and process is proportionate to the impact of the project (prototype vs prod)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# 1) Data Analysis\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "<hr>\n",
    "\n",
    "<br><br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# XAI - eXplainable AI \n",
    "\n",
    "<br>\n",
    "\n",
    "## A set of tools to explain machine learning data\n",
    "\n",
    "<br>\n",
    "\n",
    "<img src=\"images/xai.jpg\" style=\"width=100vw\">\n",
    "\n",
    "<br>\n",
    "\n",
    "## https://github.com/EthicalML/XAI\n",
    "\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "<hr>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Let's get the new training dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "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>age</th>\n",
       "      <th>workclass</th>\n",
       "      <th>education</th>\n",
       "      <th>education-num</th>\n",
       "      <th>marital-status</th>\n",
       "      <th>occupation</th>\n",
       "      <th>relationship</th>\n",
       "      <th>ethnicity</th>\n",
       "      <th>gender</th>\n",
       "      <th>capital-gain</th>\n",
       "      <th>capital-loss</th>\n",
       "      <th>hours-per-week</th>\n",
       "      <th>native-country</th>\n",
       "      <th>loan</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>39</td>\n",
       "      <td>State-gov</td>\n",
       "      <td>Bachelors</td>\n",
       "      <td>13</td>\n",
       "      <td>Never-married</td>\n",
       "      <td>Adm-clerical</td>\n",
       "      <td>Not-in-family</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>2174</td>\n",
       "      <td>0</td>\n",
       "      <td>40</td>\n",
       "      <td>United-States</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>50</td>\n",
       "      <td>Self-emp-not-inc</td>\n",
       "      <td>Bachelors</td>\n",
       "      <td>13</td>\n",
       "      <td>Married-civ-spouse</td>\n",
       "      <td>Exec-managerial</td>\n",
       "      <td>Husband</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13</td>\n",
       "      <td>United-States</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>38</td>\n",
       "      <td>Private</td>\n",
       "      <td>HS-grad</td>\n",
       "      <td>9</td>\n",
       "      <td>Divorced</td>\n",
       "      <td>Handlers-cleaners</td>\n",
       "      <td>Not-in-family</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>40</td>\n",
       "      <td>United-States</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>53</td>\n",
       "      <td>Private</td>\n",
       "      <td>11th</td>\n",
       "      <td>7</td>\n",
       "      <td>Married-civ-spouse</td>\n",
       "      <td>Handlers-cleaners</td>\n",
       "      <td>Husband</td>\n",
       "      <td>Black</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>40</td>\n",
       "      <td>United-States</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>28</td>\n",
       "      <td>Private</td>\n",
       "      <td>Bachelors</td>\n",
       "      <td>13</td>\n",
       "      <td>Married-civ-spouse</td>\n",
       "      <td>Prof-specialty</td>\n",
       "      <td>Wife</td>\n",
       "      <td>Black</td>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>40</td>\n",
       "      <td>Cuba</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   age          workclass   education  education-num       marital-status  \\\n",
       "0   39          State-gov   Bachelors             13        Never-married   \n",
       "1   50   Self-emp-not-inc   Bachelors             13   Married-civ-spouse   \n",
       "2   38            Private     HS-grad              9             Divorced   \n",
       "3   53            Private        11th              7   Married-civ-spouse   \n",
       "4   28            Private   Bachelors             13   Married-civ-spouse   \n",
       "\n",
       "           occupation    relationship ethnicity   gender  capital-gain  \\\n",
       "0        Adm-clerical   Not-in-family     White     Male          2174   \n",
       "1     Exec-managerial         Husband     White     Male             0   \n",
       "2   Handlers-cleaners   Not-in-family     White     Male             0   \n",
       "3   Handlers-cleaners         Husband     Black     Male             0   \n",
       "4      Prof-specialty            Wife     Black   Female             0   \n",
       "\n",
       "   capital-loss  hours-per-week  native-country   loan  \n",
       "0             0              40   United-States  False  \n",
       "1             0              13   United-States  False  \n",
       "2             0              40   United-States  False  \n",
       "3             0              40   United-States  False  \n",
       "4             0              40            Cuba  False  "
      ]
     },
     "execution_count": 168,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X, y, X_train, X_valid, y_train, y_valid, X_display, y_display, df, df_display \\\n",
    "    = get_dataset_2()\n",
    "df_display.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## 1.1) Visualising complex data imbalances\n",
    "#### We may be able to identify an imbalance of examples for a specific class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "im = xai.imbalance_plot(df_display, \"gender\", \"loan\" , categorical_cols=[\"loan\", \"gender\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## 1.2) Upsampling / Downsampling of data imbalances\n",
    "#### We can re-balance the metrics for imbalanced datasets, but we need to understand what this actually means"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "im = xai.balance(df_display, \"gender\", \"loan\", categorical_cols=[\"gender\", \"loan\"],\n",
    "                upsample=0.5, downsample=0.8)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## 1.2) Balanced testing / validation datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_train_balanced, y_train_balanced, X_valid_balanced, y_valid_balanced, train_idx, test_idx = \\\n",
    "    xai.balanced_train_test_split(\n",
    "            X, y, \"gender\", \n",
    "            min_per_group=300,\n",
    "            max_per_group=300,\n",
    "            categorical_cols=[\"gender\", \"loan\"])\n",
    "\n",
    "X_valid_balanced[\"loan\"] = y_valid_balanced\n",
    "im = xai.imbalance_plot(X_valid_balanced, \"gender\", \"loan\", categorical_cols=[\"gender\", \"loan\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## 1.4 Correlations hidden in data\n",
    "#### We can identify potential correlations across variables through a dendogram visualiation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/alejandro/miniconda3/envs/reddit-classification/lib/python3.7/site-packages/scipy/stats/stats.py:248: RuntimeWarning: The input array could not be properly checked for nan values. nan values will be ignored.\n",
      "  \"values. nan values will be ignored.\", RuntimeWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "corr = xai.correlations(df_display, include_categorical=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## 1.5 Shoutout to other tools and techniques\n",
    "https://github.com/EthicalML/awesome-production-machine-learning#industrial-strength-visualisation-libraries\n",
    "![](images/dataviz.jpg)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# 2) Model evaluation\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "<hr>\n",
    "\n",
    "<br><br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Alibi - Black Box Model Explanations\n",
    "\n",
    "<br>\n",
    "\n",
    "## A set of proven scientific techniques to explain ML models as black boxes\n",
    "\n",
    "<br>\n",
    "\n",
    "<img src=\"images/alibi-repo-new.jpg\" style=\"width=100vw\">\n",
    "\n",
    "<br>\n",
    "\n",
    "## https://github.com/SeldonIO/Alibi\n",
    "\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "<hr>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Model Evaluation Metrics: White / Black Box\n",
    "\n",
    "![](images/whiteblackbox.jpg)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Model Evaluation Metrics: Global vs Local\n",
    "\n",
    "![](images/globallocal.jpg)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Local black box model evaluation metrics with Alibi\n",
    "\n",
    "### Overview of methods\n",
    "\n",
    "![](images/alibi-table.jpg)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Anchors \n",
    "\n",
    "<br>\n",
    "\n",
    "#### Consists of if-then rules, called the anchors, which sufficiently guarantee the explanation locally and try to maximize the area for which the explanation holds. (ArXiv: Anchors: High-Precision Model-Agnostic Explanations)\n",
    "\n",
    "<br>\n",
    "\n",
    "<div style=\"float: left; width: 50%\">\n",
    "<img src=\"images/textanchor.jpg\">\n",
    "</div>\n",
    "\n",
    "<div style=\"float: left; width: 50%\">\n",
    "<img src=\"images/anchorimage.jpg\">\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Let's first train our model with the new, more reasonable dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Log-loss (cost function):\n",
      "training   (min:    0.311, max:    0.587, cur:    0.311)\n",
      "validation (min:    0.312, max:    0.476, cur:    0.312)\n",
      "\n",
      "Accuracy:\n",
      "training   (min:    0.735, max:    0.857, cur:    0.855)\n",
      "validation (min:    0.794, max:    0.857, cur:    0.857)\n"
     ]
    }
   ],
   "source": [
    "# Let's start by building our model with our newly balanced dataset\n",
    "model = build_model(X)\n",
    "model.fit(f_in(X_train), y_train, epochs=20, batch_size=512, shuffle=True, validation_data=(f_in(X_valid), y_valid), callbacks=[PlotLossesKeras()], verbose=0, validation_split=0.05,)\n",
    "probabilities = model.predict(f_in(X_valid))\n",
    "pred = f_out(probabilities)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## We can now use the Tabular Anchor technique in Alibi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explainer built\n"
     ]
    }
   ],
   "source": [
    "from alibi.explainers import AnchorTabular\n",
    "\n",
    "explainer = AnchorTabular(\n",
    "    loan_model_alibi.predict, \n",
    "    feature_names_alibi, \n",
    "    categorical_names=category_map_alibi)\n",
    "\n",
    "explainer.fit(\n",
    "    X_train_alibi, \n",
    "    disc_perc=[25, 50, 75])\n",
    "\n",
    "print(\"Explainer built\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[52,  4,  0,  2,  8,  4,  2,  0,  0,  0, 60,  9]])"
      ]
     },
     "execution_count": 175,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test_alibi[:1] "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Anchor: Marital Status = Separated AND Sex = Female\n",
      "Precision: 0.97\n",
      "Coverage: 0.11\n"
     ]
    }
   ],
   "source": [
    "explanation = explainer.explain(X_test_alibi[:1], threshold=0.95)\n",
    "\n",
    "print('Anchor: %s' % (' AND '.join(explanation['names'])))\n",
    "print('Precision: %.2f' % explanation['precision'])\n",
    "print('Coverage: %.2f' % explanation['coverage'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Counterfactual Explanations\n",
    "\n",
    "### The counterfactual explanation of an outcome or a situation Y takes the form “If X had not occured, Y would not have occured” \n",
    "\n",
    "![](images/counterfactuals7.jpg)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "### We load a convolutional neural network\n",
    "#### Explainability techniques work with any classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cnn = load_model('mnist_cnn.h5')\n",
    "cf_X = cf_x_test[0].reshape((1,) + cf_x_test[0].shape)\n",
    "plt.imshow(cf_X.reshape(28, 28));"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "### Build a counterfactual explanation\n",
    "#### We can select the target class we want to aim for"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Counterfactual prediction: 9 with probability [2.8820663e-12 3.0502541e-11 5.0862967e-09 2.1951721e-06 3.9167953e-05\n",
      " 7.8527846e-06 2.6525455e-14 9.4558978e-05 7.3595431e-05 9.9978262e-01]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "shape = (1,) + cf_x_train.shape[1:]\n",
    "target_class = 9 # any class other than 7 will do\n",
    "\n",
    "cf = CounterFactual(cnn, shape=shape, target_class=target_class, target_proba=1.0, max_iter=20)\n",
    "explanation = cf.explain(cf_X)\n",
    "\n",
    "print(f\"Counterfactual prediction: {explanation['cf']['class']} with probability {explanation['cf']['proba'][0]}\")\n",
    "plt.imshow(explanation['cf']['X'].reshape(28, 28));"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x432 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "show_iterations(explanation)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Improving computational efficiency\n",
    "\n",
    "### Our Data Science team at Seldon published a paper that tackles the issue.\n",
    "\n",
    "<img src=\"images/with-prototypes.jpg\" style=\"height: 80vh\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "### Intersection with Adversarial Robustness\n",
    "\n",
    "<br>\n",
    "\n",
    "* Black box explainability techniques allow us to understand black boxes\n",
    "\n",
    "<br>\n",
    "\n",
    "* But also provide tools that could be used maliciously\n",
    "\n",
    "<br>\n",
    "\n",
    "* Extra considerations need to be taken into account (i.e. when is a model \"being explained\")\n",
    "\n",
    "<br>\n",
    "\n",
    "* Limited access to explainability (auditors, domain experts, etc)\n",
    "\n",
    "<br>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## 1.5 Shoutout to other tools and techniques\n",
    "https://github.com/EthicalML/awesome-production-machine-learning#explaining-black-box-models-and-datasets\n",
    "![](images/modevallibs.jpg)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# 3) Production Monitoring\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "<hr>\n",
    "\n",
    "<br><br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Seldon Core - Production ML in K8s\n",
    "\n",
    "<br>\n",
    "\n",
    "## A language agnostic ML serving & monitoring framework in Kubernetes\n",
    "\n",
    "<br>\n",
    "\n",
    "<img src=\"images/seldon-core-repo.jpg\" style=\"width=100vw\">\n",
    "\n",
    "<br>\n",
    "\n",
    "## https://github.com/SeldonIO/seldon-core\n",
    "\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "<hr>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Design patterns for explainers\n",
    "\n",
    "#### Using Seldon Core for deployment, orchestration and monitoring\n",
    "\n",
    "<img src=\"images/prodxai-1.jpg\" style=\"height: 80vh\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Deploy model\n",
    "\n",
    "#### We first deploy our model\n",
    "\n",
    "<img src=\"images/prodxai-2.jpg\" style=\"height: 80vh\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Request predictions\n",
    "\n",
    "#### We can send http request to get a prediction\n",
    "\n",
    "<img src=\"images/prodxai-3.jpg\" style=\"height: 80vh\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Deploy explainer\n",
    "\n",
    "#### We then would be able ot wrap and deploy the explainer\n",
    "\n",
    "<img src=\"images/prodxai-4a.jpg\" style=\"height: 80vh\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Request explanations\n",
    "\n",
    "#### We can then send http requsests to the explainer, which sends http requests to the model to reverse engineer it\n",
    "\n",
    "<img src=\"images/prodxai-5a.jpg\" style=\"height: 80vh\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Setup Seldon in your kubernetes cluster\n",
    "#### We already have a kubernetes cluster running in our localhost but you can set it up locally\n",
    "\n",
    "<img src=\"images/clusterm.jpg\">"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [],
   "source": [
    "%%bash\n",
    "# kubectl create clusterrolebinding kube-system-cluster-admin --clusterrole=cluster-admin --serviceaccount=kube-system:default\n",
    "# helm init\n",
    "# kubectl rollout status deploy/tiller-deploy -n kube-system\n",
    "# helm install seldon-core-operator --name seldon-core-operator --repo https://storage.googleapis.com/seldon-charts --set engine.image.tag=0.4.1 --set image.tag=0.4.1\n",
    "# helm install seldon-core-analytics --name seldon-core-analytics --repo https://storage.googleapis.com/seldon-charts\n",
    "# helm install stable/ambassador --name ambassador\n",
    "\n",
    "\n",
    "# kubectl patch svc ambassador --type='json' -p '[{\"op\":\"replace\",\"path\":\"/spec/type\",\"value\":\"NodePort\"}]'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import LabelEncoder, StandardScaler, OneHotEncoder\n",
    "from sklearn.impute import SimpleImputer\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.compose import ColumnTransformer\n",
    "\n",
    "# feature transformation pipeline\n",
    "ordinal_features = [x for x in range(len(alibi_feature_names)) if x not in list(alibi_category_map.keys())]\n",
    "ordinal_transformer = Pipeline(steps=[('imputer', SimpleImputer(strategy='median')),\n",
    "                                      ('scaler', StandardScaler())])\n",
    "\n",
    "categorical_features = list(alibi_category_map.keys())\n",
    "categorical_transformer = Pipeline(steps=[('imputer', SimpleImputer(strategy='median')),\n",
    "                                          ('onehot', OneHotEncoder(handle_unknown='ignore'))])\n",
    "\n",
    "preprocessor = ColumnTransformer(transformers=[('num', ordinal_transformer, ordinal_features),\n",
    "                                               ('cat', categorical_transformer, categorical_features)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [],
   "source": [
    "preprocessor.fit(alibi_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "np.random.seed(0)\n",
    "randomforest = RandomForestClassifier(n_estimators=50)\n",
    "randomforest.fit(preprocessor.transform(X_train_alibi), y_train_alibi)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [],
   "source": [
    "!mkdir -p pipeline/pipeline_steps/loanclassifier/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "### We have our loan classifier models\n",
    "#### Currently a random forest classifier and a preprocessor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: [[52  4  0  2  8  4  2  0  0  0 60  9]]\n",
      "Predicted class: [0]\n",
      "Probabilities: [[0.86 0.14]]\n"
     ]
    }
   ],
   "source": [
    "print(f\"Input: {X_test_alibi[:1]}\")\n",
    "print(f\"Predicted class: {randomforest.predict(preprocessor.transform(X_test_alibi[:1]))}\")\n",
    "print(f\"Probabilities: {randomforest.predict_proba(preprocessor.transform(X_test_alibi[:1]))}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### Save the model artefacts so we can deploy them"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [],
   "source": [
    "import dill\n",
    "\n",
    "with open(\"pipeline/pipeline_steps/loanclassifier/preprocessor.dill\", \"wb\") as prep_f:\n",
    "    dill.dump(preprocessor, prep_f)\n",
    "    \n",
    "with open(\"pipeline/pipeline_steps/loanclassifier/model.dill\", \"wb\") as model_f:\n",
    "    dill.dump(randomforest, model_f)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### Build a Model wrapper that uses the trained models through a predict function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%writefile pipeline/pipeline_steps/loanclassifier/Model.py\n",
    "import dill\n",
    "\n",
    "class Model:\n",
    "    def __init__(self, *args, **kwargs):\n",
    "        \n",
    "        with open(\"preprocessor.dill\", \"rb\") as prep_f:\n",
    "            self.preprocessor = dill.load(prep_f)\n",
    "        with open(\"model.dill\", \"rb\") as model_f:\n",
    "            self.clf = dill.load(model_f)\n",
    "        \n",
    "    def predict(self, X, feature_names=[]):\n",
    "        X_prep = self.preprocessor.transform(X)\n",
    "        proba = self.clf.predict_proba(X_prep)\n",
    "        return proba"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### Add the dependencies for the wrapper to work"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Overwriting pipeline/pipeline_steps/loanclassifier/requirements.txt\n"
     ]
    }
   ],
   "source": [
    "%%writefile pipeline/pipeline_steps/loanclassifier/requirements.txt\n",
    "dill==0.2.9\n",
    "scikit-image==0.15.0\n",
    "scikit-learn==0.20.1\n",
    "scipy==1.1.0\n",
    "numpy==1.17.1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [],
   "source": [
    "!mkdir pipeline/pipeline_steps/loanclassifier/.s2i"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [],
   "source": [
    "%%writefile pipeline/pipeline_steps/loanclassifier/.s2i/environment\n",
    "MODEL_NAME=Model\n",
    "API_TYPE=REST\n",
    "SERVICE_TYPE=MODEL\n",
    "PERSISTENCE=0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### Use the source2image command to containerize code"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true,
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [],
   "source": [
    "!s2i build pipeline/pipeline_steps/loanclassifier seldonio/seldon-core-s2i-python3:0.11 loanclassifier:0.1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### Define the graph of your pipeline with individual models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [],
   "source": [
    "%%writefile pipeline/pipeline_steps/loanclassifier/loanclassifiermodel.yaml\n",
    "apiVersion: machinelearning.seldon.io/v1alpha2\n",
    "kind: SeldonDeployment\n",
    "metadata:\n",
    "  labels:\n",
    "    app: seldon\n",
    "  name: loanclassifier\n",
    "spec:\n",
    "  name: loanclassifier\n",
    "  predictors:\n",
    "  - componentSpecs:\n",
    "    - spec:\n",
    "        containers:\n",
    "        - image: loanclassifier:0.1\n",
    "          name: model\n",
    "    graph:\n",
    "      children: []\n",
    "      name: model\n",
    "      type: MODEL\n",
    "      endpoint:\n",
    "        type: REST\n",
    "    name: loanclassifier\n",
    "    replicas: 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### Deploy your model!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "seldondeployment.machinelearning.seldon.io/loanclassifier configured\r\n"
     ]
    }
   ],
   "source": [
    "!kubectl apply -f pipeline/pipeline_steps/loanclassifier/loanclassifiermodel.yaml"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### We can now send a request through HTTP"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[52  4  0  2  8  4  2  0  0  0 60  9]]\n"
     ]
    }
   ],
   "source": [
    "batch = X_test_alibi[:1]\n",
    "print(batch)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "values {\n",
      "  list_value {\n",
      "    values {\n",
      "      number_value: 0.86\n",
      "    }\n",
      "    values {\n",
      "      number_value: 0.14\n",
      "    }\n",
      "  }\n",
      "}\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from seldon_core.seldon_client import SeldonClient\n",
    "\n",
    "sc = SeldonClient(\n",
    "    gateway=\"ambassador\", \n",
    "    gateway_endpoint=\"localhost:80\",\n",
    "    deployment_name=\"loanclassifier\",\n",
    "    payload_type=\"ndarray\",\n",
    "    namespace=\"default\",\n",
    "    transport=\"rest\")\n",
    "\n",
    "client_prediction = sc.predict(data=batch)\n",
    "\n",
    "print(client_prediction.response.data.ndarray)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### Now we can send data through the REST API"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\n",
      "  \"meta\": {\n",
      "    \"puid\": \"9naau11chptmvhgjapkng6a13r\",\n",
      "    \"tags\": {\n",
      "    },\n",
      "    \"routing\": {\n",
      "    },\n",
      "    \"requestPath\": {\n",
      "      \"model\": \"loanclassifier:0.1\"\n",
      "    },\n",
      "    \"metrics\": []\n",
      "  },\n",
      "  \"data\": {\n",
      "    \"names\": [\"t:0\", \"t:1\"],\n",
      "    \"ndarray\": [[0.86, 0.14]]\n",
      "  }\n",
      "}"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current\n",
      "                                 Dload  Upload   Total   Spent    Left  Speed\n",
      "\r",
      "  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0\r",
      "100   356  100   264  100    92  13200   4600 --:--:-- --:--:-- --:--:-- 18736\n"
     ]
    }
   ],
   "source": [
    "%%bash\n",
    "curl -X POST -H 'Content-Type: application/json' \\\n",
    "    -d \"{'data': {'names': ['text'], 'ndarray': [[52,  4,  0,  2,  8,  4,  2,  0,  0,  0, 60,  9]]}}\" \\\n",
    "    \"http://localhost:80/seldon/default/loanclassifier/api/v0.1/predictions\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### Now we can create an explainer for our model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Anchor: Marital Status = Separated AND Sex = Female AND Relationship = Unmarried\n",
      "Precision: 0.99\n",
      "Coverage: 0.05\n"
     ]
    }
   ],
   "source": [
    "from alibi.explainers import AnchorTabular\n",
    "\n",
    "predict_fn = lambda x: randomforest.predict(preprocessor.transform(x))\n",
    "explainer = AnchorTabular(predict_fn, alibi_feature_names, categorical_names=alibi_category_map)\n",
    "explainer.fit(X_train_alibi, disc_perc=[25, 50, 75])\n",
    "\n",
    "explanation = explainer.explain(X_test_alibi[0], threshold=0.95)\n",
    "\n",
    "print('Anchor: %s' % (' AND '.join(explanation['names'])))\n",
    "print('Precision: %.2f' % explanation['precision'])\n",
    "print('Coverage: %.2f' % explanation['coverage'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### But now we can use the remote model we have in production"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [],
   "source": [
    "def predict_remote_fn(X):\n",
    "    from seldon_core.seldon_client import SeldonClient\n",
    "    from seldon_core.utils import get_data_from_proto\n",
    "    \n",
    "    kwargs = {\n",
    "        \"gateway\": \"ambassador\", \n",
    "        \"deployment_name\": \"loanclassifier\",\n",
    "        \"payload_type\": \"ndarray\",\n",
    "        \"namespace\": \"default\",\n",
    "        \"transport\": \"rest\"\n",
    "    }\n",
    "    \n",
    "    try:\n",
    "        kwargs[\"gateway_endpoint\"] = \"localhost:80\"\n",
    "        sc = SeldonClient(**kwargs)\n",
    "        prediction = sc.predict(data=X)\n",
    "    except:\n",
    "        # If we are inside the container, we need to reach the ambassador service directly\n",
    "        kwargs[\"gateway_endpoint\"] = \"ambassador:80\"\n",
    "        sc = SeldonClient(**kwargs)\n",
    "        prediction = sc.predict(data=X)\n",
    "    \n",
    "    y = get_data_from_proto(prediction.response)\n",
    "    return y\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### And train our explainer to use the remote function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Anchor: Marital Status = Separated AND Sex = Female\n",
      "Precision: 0.98\n",
      "Coverage: 0.11\n"
     ]
    }
   ],
   "source": [
    "from seldon_core.utils import get_data_from_proto\n",
    "\n",
    "explainer = AnchorTabular(predict_remote_fn, alibi_feature_names, categorical_names=alibi_category_map)\n",
    "explainer.fit(X_train_alibi, disc_perc=[25, 50, 75])\n",
    "\n",
    "explanation = explainer.explain(batch, threshold=0.95)\n",
    "\n",
    "print('Anchor: %s' % (' AND '.join(explanation['names'])))\n",
    "print('Precision: %.2f' % explanation['precision'])\n",
    "print('Coverage: %.2f' % explanation['coverage'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### To containerise our explainer, save the trained binary\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [],
   "source": [
    "import dill\n",
    "\n",
    "with open(\"pipeline/pipeline_steps/loanclassifier-explainer/explainer.dill\", \"wb\") as x_f:\n",
    "    dill.dump(explainer, x_f)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### Expose it through a wrapper"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [],
   "source": [
    "%%writefile pipeline/pipeline_steps/loanclassifier-explainer/Explainer.py\n",
    "import dill\n",
    "import json\n",
    "import numpy as np\n",
    "\n",
    "class Explainer:\n",
    "    def __init__(self, *args, **kwargs):\n",
    "        \n",
    "        with open(\"explainer.dill\", \"rb\") as x_f:\n",
    "            self.explainer = dill.load(x_f)\n",
    "        \n",
    "    def predict(self, X, feature_names=[]):\n",
    "        print(\"Received: \" + str(X))\n",
    "        explanation = self.explainer.explain(X)\n",
    "        print(\"Predicted: \" + str(explanation))\n",
    "        return json.dumps(explanation, cls=NumpyEncoder)\n",
    "\n",
    "    \n",
    "    \n",
    "    \n",
    "class NumpyEncoder(json.JSONEncoder):\n",
    "    def default(self, obj):\n",
    "        if isinstance(obj, (\n",
    "        np.int_, np.intc, np.intp, np.int8, np.int16, np.int32, np.int64, np.uint8, np.uint16, np.uint32, np.uint64)):\n",
    "            return int(obj)\n",
    "        elif isinstance(obj, (np.float_, np.float16, np.float32, np.float64)):\n",
    "            return float(obj)\n",
    "        elif isinstance(obj, (np.ndarray,)):\n",
    "            return obj.tolist()\n",
    "        return json.JSONEncoder.default(self, obj)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [],
   "source": [
    "%%writefile pipeline/pipeline_steps/loanclassifier-explainer/requirements.txt\n",
    "scikit-learn==0.20.1\n",
    "alibi==0.3.0\n",
    "dill==0.2.9\n",
    "scikit-image==0.15.0\n",
    "scikit-learn==0.20.1\n",
    "scipy==1.1.0\n",
    "numpy==1.17.1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [],
   "source": [
    "!mkdir pipeline/pipeline_steps/loanclassifier-explainer/.s2i"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [],
   "source": [
    "%%writefile pipeline/pipeline_steps/loanclassifier-explainer/.s2i/environment\n",
    "MODEL_NAME=Explainer\n",
    "API_TYPE=REST\n",
    "SERVICE_TYPE=MODEL\n",
    "PERSISTENCE=0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "### Build the container for the explainer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true,
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [],
   "source": [
    "!s2i build pipeline/pipeline_steps/loanclassifier-explainer seldonio/seldon-core-s2i-python3:0.11 loanclassifier-explainer:0.1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### Add config files to build image with script"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [],
   "source": [
    "%%writefile pipeline/pipeline_steps/loanclassifier-explainer/loanclassifiermodel-explainer.yaml\n",
    "apiVersion: machinelearning.seldon.io/v1alpha2\n",
    "kind: SeldonDeployment\n",
    "metadata:\n",
    "  labels:\n",
    "    app: seldon\n",
    "  name: loanclassifier-explainer\n",
    "spec:\n",
    "  name: loanclassifier-explainer\n",
    "  annotations:             \n",
    "    seldon.io/rest-read-timeout: \"100000\"\n",
    "    seldon.io/rest-connection-timeout: \"100000\"\n",
    "    seldon.io/grpc-read-timeout: \"100000\"\n",
    "  predictors:\n",
    "  - componentSpecs:\n",
    "    - spec:\n",
    "        containers:\n",
    "        - image: loanclassifier-explainer:0.1\n",
    "          name: model-explainer\n",
    "    graph:\n",
    "      children: []\n",
    "      name: model-explainer\n",
    "      type: MODEL\n",
    "      endpoint:\n",
    "        type: REST\n",
    "    name: loanclassifier-explainer\n",
    "    replicas: 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### Deploy your remote explainer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "seldondeployment.machinelearning.seldon.io/loanclassifier-explainer configured\r\n"
     ]
    }
   ],
   "source": [
    "!kubectl apply -f pipeline/pipeline_steps/loanclassifier-explainer/loanclassifiermodel-explainer.yaml"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### Now we can request explanations throught the REST API"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[52  4  0  2  8  4  2  0  0  0 60  9]]\n",
      "['Marital Status = Separated', 'Sex = Female']\n"
     ]
    }
   ],
   "source": [
    "from seldon_core.seldon_client import SeldonClient\n",
    "import json\n",
    "\n",
    "batch = X_test_alibi[:1]\n",
    "print(batch)\n",
    "\n",
    "sc = SeldonClient(\n",
    "    gateway=\"ambassador\", \n",
    "    gateway_endpoint=\"localhost:80\",\n",
    "    deployment_name=\"loanclassifier-explainer\",\n",
    "    payload_type=\"ndarray\",\n",
    "    namespace=\"default\",\n",
    "    transport=\"rest\")\n",
    "\n",
    "client_prediction = json.loads(sc.predict(data=batch).response.strData)\n",
    "\n",
    "print(client_prediction[\"names\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\n",
      "  \"meta\": {\n",
      "    \"puid\": \"5q5bavp0qo1qg247lcml9864vl\",\n",
      "    \"tags\": {\n",
      "    },\n",
      "    \"routing\": {\n",
      "    },\n",
      "    \"requestPath\": {\n",
      "      \"model-explainer\": \"loanclassifier-explainer:0.1\"\n",
      "    },\n",
      "    \"metrics\": []\n",
      "  },\n",
      "  \"strData\": \"{\\\"names\\\": [\\\"Marital Status = Separated\\\", \\\"Sex = Female\\\", \\\"Education = Associates\\\"], \\\"precision\\\": 0.9739130434782609, \\\"coverage\\\": 0.0096, \\\"raw\\\": {\\\"feature\\\": [3, 7, 2], \\\"mean\\\": [0.8801571709233792, 0.9497206703910615, 0.9739130434782609], \\\"precision\\\": [0.8801571709233792, 0.9497206703910615, 0.9739130434782609], \\\"coverage\\\": [0.1782, 0.1029, 0.0096], \\\"examples\\\": [{\\\"covered\\\": [[41, 4, 1, 2, 2, 0, 4, 1, 0, 0, 40, 0], [23, 4, 4, 2, 7, 3, 4, 0, 0, 0, 20, 9], [33, 4, 0, 2, 1, 1, 4, 0, 0, 0, 40, 9], [50, 4, 4, 2, 2, 0, 4, 1, 4386, 0, 40, 0], [22, 4, 0, 2, 2, 1, 4, 1, 0, 0, 45, 9], [30, 4, 4, 2, 5, 4, 4, 0, 0, 0, 40, 9], [21, 4, 4, 2, 2, 3, 4, 1, 0, 0, 36, 9], [35, 4, 4, 2, 7, 1, 4, 0, 0, 0, 35, 9], [35, 4, 4, 2, 2, 0, 4, 1, 0, 0, 46, 9], [46, 4, 6, 2, 5, 0, 4, 1, 0, 0, 40, 9]], \\\"covered_true\\\": [[40, 4, 3, 2, 2, 0, 4, 1, 0, 0, 55, 9], [30, 4, 4, 2, 5, 4, 4, 0, 0, 0, 40, 9], [33, 4, 1, 2, 5, 1, 4, 1, 0, 0, 40, 1], [28, 4, 4, 2, 2, 0, 4, 1, 0, 0, 40, 9], [28, 4, 1, 2, 1, 1, 4, 1, 0, 0, 21, 9], [25, 0, 4, 2, 0, 3, 2, 1, 0, 0, 40, 0], [31, 4, 3, 2, 2, 0, 4, 1, 0, 0, 40, 9], [31, 0, 3, 2, 0, 1, 4, 0, 0, 0, 25, 9], [43, 4, 4, 2, 2, 0, 4, 1, 0, 0, 40, 9], [50, 4, 4, 2, 4, 0, 4, 1, 0, 0, 38, 9]], \\\"covered_false\\\": [[36, 4, 4, 2, 5, 3, 4, 0, 0, 0, 40, 9], [32, 4, 1, 2, 8, 0, 4, 1, 0, 0, 45, 9], [49, 4, 1, 2, 8, 1, 2, 0, 27828, 0, 60, 9], [31, 4, 1, 2, 5, 5, 4, 0, 7688, 0, 43, 9], [52, 4, 5, 2, 5, 0, 4, 1, 0, 0, 50, 9], [31, 4, 2, 2, 5, 0, 4, 1, 0, 1977, 99, 9], [50, 4, 5, 2, 5, 0, 4, 1, 0, 1977, 40, 9], [42, 5, 5, 2, 8, 0, 4, 1, 0, 0, 40, 9], [37, 4, 1, 2, 8, 0, 4, 1, 5178, 0, 40, 9], [57, 2, 5, 2, 8, 0, 4, 1, 0, 0, 50, 9]], \\\"uncovered_true\\\": [], \\\"uncovered_false\\\": []}, {\\\"covered\\\": [[29, 4, 1, 2, 1, 1, 4, 0, 0, 0, 35, 9], [50, 4, 1, 2, 1, 5, 4, 0, 0, 0, 40, 9], [19, 0, 3, 2, 0, 4, 4, 0, 0, 0, 30, 9], [22, 4, 4, 2, 1, 3, 4, 0, 0, 0, 40, 9], [37, 4, 0, 2, 8, 1, 4, 0, 0, 0, 40, 9], [38, 4, 4, 2, 2, 3, 4, 0, 0, 0, 25, 9], [59, 4, 1, 2, 1, 5, 4, 0, 2885, 0, 30, 9], [23, 4, 1, 2, 6, 1, 4, 0, 0, 0, 45, 9], [31, 4, 0, 2, 6, 3, 4, 0, 0, 0, 38, 9], [50, 4, 4, 2, 1, 2, 1, 0, 0, 0, 40, 7]], \\\"covered_true\\\": [[21, 4, 4, 2, 7, 3, 4, 0, 0, 0, 35, 9], [25, 0, 4, 2, 0, 4, 4, 0, 0, 0, 50, 9], [56, 4, 3, 2, 7, 4, 2, 0, 0, 0, 45, 9], [40, 0, 4, 2, 0, 5, 2, 0, 3464, 0, 20, 9], [52, 4, 4, 2, 6, 5, 4, 0, 0, 0, 52, 1], [25, 4, 4, 2, 4, 1, 3, 0, 0, 0, 40, 9], [21, 4, 4, 2, 6, 4, 0, 0, 0, 0, 38, 9], [21, 4, 4, 2, 6, 3, 4, 0, 0, 0, 20, 9], [26, 4, 4, 2, 1, 1, 4, 0, 0, 0, 40, 9], [25, 4, 0, 2, 5, 5, 4, 0, 0, 1887, 40, 9]], \\\"covered_false\\\": [[56, 6, 4, 2, 1, 4, 4, 0, 99999, 0, 40, 9], [33, 2, 1, 2, 5, 1, 4, 0, 0, 0, 50, 9], [63, 4, 1, 2, 5, 5, 4, 0, 7688, 0, 36, 9], [67, 4, 4, 2, 1, 1, 4, 0, 15831, 0, 16, 3], [43, 4, 5, 2, 5, 4, 4, 0, 0, 2547, 40, 9], [49, 4, 1, 2, 8, 1, 2, 0, 27828, 0, 60, 9], [62, 4, 4, 2, 2, 1, 4, 0, 8614, 0, 39, 9], [31, 4, 1, 2, 5, 5, 4, 0, 7688, 0, 43, 9], [36, 4, 4, 2, 5, 3, 4, 0, 0, 0, 40, 9], [53, 4, 2, 2, 5, 5, 4, 0, 99999, 0, 37, 9]], \\\"uncovered_true\\\": [], \\\"uncovered_false\\\": []}, {\\\"covered\\\": [[23, 7, 0, 2, 8, 1, 4, 0, 0, 0, 40, 9], [40, 4, 0, 2, 7, 5, 4, 0, 0, 0, 35, 9], [57, 4, 0, 2, 7, 4, 4, 0, 0, 0, 40, 9], [39, 4, 0, 2, 7, 0, 1, 0, 0, 0, 40, 7], [72, 4, 0, 2, 7, 4, 4, 0, 0, 0, 16, 9], [43, 4, 0, 2, 5, 0, 4, 0, 0, 0, 40, 9], [33, 4, 0, 2, 2, 4, 0, 0, 0, 0, 40, 9], [23, 4, 0, 2, 2, 4, 4, 0, 0, 0, 50, 9], [53, 6, 0, 2, 6, 0, 4, 0, 0, 0, 40, 1], [42, 4, 0, 2, 5, 4, 4, 0, 0, 0, 40, 9]], \\\"covered_true\\\": [[37, 4, 0, 2, 8, 0, 4, 0, 0, 1977, 60, 9], [18, 4, 0, 2, 6, 3, 4, 0, 0, 0, 25, 9], [47, 6, 0, 2, 8, 0, 4, 0, 0, 2415, 50, 9], [48, 7, 0, 2, 1, 0, 4, 0, 0, 0, 35, 9], [19, 4, 0, 2, 1, 3, 2, 0, 0, 0, 40, 9], [34, 4, 0, 2, 5, 0, 4, 0, 0, 0, 40, 6], [59, 4, 0, 2, 5, 4, 4, 0, 0, 0, 50, 9], [30, 4, 0, 2, 1, 4, 4, 0, 0, 0, 24, 9], [28, 2, 0, 2, 5, 1, 4, 0, 0, 0, 40, 9], [34, 4, 0, 2, 2, 0, 4, 0, 0, 0, 40, 3]], \\\"covered_false\\\": [[39, 4, 0, 2, 5, 5, 4, 0, 7688, 0, 20, 9], [40, 4, 0, 2, 8, 0, 4, 0, 0, 2415, 45, 9], [38, 4, 0, 2, 5, 0, 4, 0, 99999, 0, 65, 9], [70, 5, 0, 2, 8, 0, 4, 0, 0, 2377, 50, 9], [46, 4, 0, 2, 6, 0, 4, 0, 15024, 0, 40, 9], [42, 4, 0, 2, 6, 0, 4, 0, 7298, 0, 40, 9]], \\\"uncovered_true\\\": [], \\\"uncovered_false\\\": []}], \\\"all_precision\\\": 0, \\\"num_preds\\\": 1000101, \\\"names\\\": [\\\"Marital Status = Separated\\\", \\\"Sex = Female\\\", \\\"Education = Associates\\\"], \\\"instance\\\": [[52.0, 4.0, 0.0, 2.0, 8.0, 4.0, 2.0, 0.0, 0.0, 0.0, 60.0, 9.0]], \\\"prediction\\\": 0}}\"\n",
      "}"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current\n",
      "                                 Dload  Upload   Total   Spent    Left  Speed\n",
      "\r",
      "  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0\r",
      "100    92    0     0  100    92      0     76  0:00:01  0:00:01 --:--:--    76\r",
      "100  4784  100  4692  100    92   2329     45  0:00:02  0:00:02 --:--:--  2376\r",
      "100  4784  100  4692  100    92   2329     45  0:00:02  0:00:02 --:--:--  2375\n"
     ]
    }
   ],
   "source": [
    "%%bash\n",
    "curl -X POST -H 'Content-Type: application/json' \\\n",
    "    -d \"{'data': {'names': ['text'], 'ndarray': [[52,  4,  0,  2,  8,  4,  2,  0,  0,  0, 60, 9]] }}\" \\\n",
    "    http://localhost:80/seldon/default/loanclassifier-explainer/api/v0.1/predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "### Now we have an explainer deployed!\n",
    "\n",
    "![](images/xai-deployment.jpg)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# Visualise metrics and explanations \n",
    "\n",
    "![](images/deploy-expl.jpg)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Revisiting our workflow\n",
    "\n",
    "<img src=\"images/gml.png\" style=\"width=100vw\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# A practical guide towards explainability\n",
    "# and bias evaluation in machine learning\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "<br>\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "<br>\n",
    "\n",
    "<br>\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "\n",
    "<h2>Alejandro Saucedo</h2>\n",
    "<br>\n",
    "Chief Scientist, The Institute for Ethical AI & Machine Learning\n",
    "<br>\n",
    "Director of ML Engineering, Seldon Technologies\n",
    "\n",
    "<br>\n",
    "<br>\n",
    "<h3>Twitter: AxSaucedo</h3>\n",
    "<br>\n",
    "<h3>Slides: github.com/EthicalML/explainability-and-bias</h3>\n",
    "\n",
    "<br><br><br>\n",
    "\n",
    "<hr>\n"
   ]
  }
 ],
 "metadata": {
  "celltoolbar": "Slideshow",
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}