diff --git a/notebooks/images/add_the_repository.jpg b/notebooks/images/add_the_repository.jpg
new file mode 100644
index 0000000..98dcb0e
Binary files /dev/null and b/notebooks/images/add_the_repository.jpg differ
diff --git a/notebooks/images/governance.jpg b/notebooks/images/governance.jpg
new file mode 100644
index 0000000..7863cfe
Binary files /dev/null and b/notebooks/images/governance.jpg differ
diff --git a/notebooks/images/monitoring_job.jpg b/notebooks/images/monitoring_job.jpg
new file mode 100644
index 0000000..06818b3
Binary files /dev/null and b/notebooks/images/monitoring_job.jpg differ
diff --git a/notebooks/images/monitoring_result.jpg b/notebooks/images/monitoring_result.jpg
new file mode 100644
index 0000000..bf1f337
Binary files /dev/null and b/notebooks/images/monitoring_result.jpg differ
diff --git a/notebooks/images/repository.jpg b/notebooks/images/repository.jpg
new file mode 100644
index 0000000..0c552ec
Binary files /dev/null and b/notebooks/images/repository.jpg differ
diff --git a/notebooks/images/runtime.jpg b/notebooks/images/runtime.jpg
new file mode 100644
index 0000000..f50d219
Binary files /dev/null and b/notebooks/images/runtime.jpg differ
diff --git a/notebooks/images/studio_model.jpg b/notebooks/images/studio_model.jpg
new file mode 100644
index 0000000..cbd0d82
Binary files /dev/null and b/notebooks/images/studio_model.jpg differ
diff --git a/notebooks/images/violations.jpg b/notebooks/images/violations.jpg
new file mode 100644
index 0000000..01a172e
Binary files /dev/null and b/notebooks/images/violations.jpg differ
diff --git a/notebooks/model_bias_monitor.ipynb b/notebooks/model_bias_monitor.ipynb
new file mode 100644
index 0000000..c307aa6
--- /dev/null
+++ b/notebooks/model_bias_monitor.ipynb
@@ -0,0 +1,2319 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "d6eccd85-1c0e-4f7b-9da0-dd473f8a9d95",
+   "metadata": {},
+   "source": [
+    "# Model Bias Monitoring with AWS SageMaker Clarity"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "67557cb8-fe01-473a-a256-442ab3131197",
+   "metadata": {},
+   "source": [
+    "This Jupyter notebook shows how to perform model bias observability with AWS SageMaker (based on [docs](https://sagemaker-examples.readthedocs.io/en/latest/sagemaker_model_monitor/fairness_and_explainability/SageMaker-Model-Monitor-Fairness-and-Explainability.html))\n",
+    "\n",
+    "1. <a href=\"#1.-Configuration\">Configuration</a>\n",
+    "1. <a href=\"#2.-Dataset\">Dataset</a> description\n",
+    "1. <a href=\"#3.-Deploying-model\">Deploying model</a> for ML Observability\n",
+    "1. <a href=\"#4.-Setting-up-monitoring-job\">Setting up monitoring job</a> - Creates Monitoring tasks by creating baseline and scheduling regular monitoring\n",
+    "1. <a href=\"#5.-Generating-traffic\">Generate traffic</a> - Provides traffic (examples and ground truth) to the endpoint based on which bias metrics will be calculated\n",
+    "1. <a href=\"#6.-Cleaning-up\">Cleaning up</a> - Removes all the created resources.\n",
+    "\n",
+    "Prerequisites:\n",
+    "\n",
+    "- Existing Roles with all needed permissions (S3, SageMaker, etc.);\n",
+    "- Configured SageMaker Domain;\n",
+    "- SageMaker Studio user.\n",
+    "\n",
+    "One can use SageMaker Studio (Jupyter-like environment) to run this notebook. To do that, follow the next steps:\n",
+    "\n",
+    "1. Run the SageMaker Studio\n",
+    "2. Clone this repository (https://github.com/griddynamics/gd-ml-observability.git)\n",
+    "\n",
+    "\n",
+    "<img alt=\"Add the repository.png\" src=\"images/add_the_repository.jpg\" style=\"height:100px\"/>\n",
+    "\n",
+    "<img alt=\"repository.png\" src=\"images/repository.jpg\" style=\"height:300px; margin: auto\"/>\n",
+    "\n",
+    "3. Run this notebook cell by cell paying attention to comments"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "9096958f-3dbf-41b6-96bd-dce1fe2bb299",
+   "metadata": {},
+   "source": [
+    "### Background"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "30192a0e-66ad-4f5f-b1fa-10296809ef81",
+   "metadata": {},
+   "source": [
+    "A computer system powered by machine learning model might contain bias, i.e., discriminate against certain individuals or groups of individuals. The models learn from data, so it might memorize bias that appears in that data. Biased judgements impact on people badly and unfairly, and automation of such process could lead to disasterous consequences. Hence, people developing such systems must be aware of it, and be able to address biased models. [\\*](https://docs.aws.amazon.com/sagemaker/latest/dg/clarify-detect-data-bias.html)\n",
+    "\n",
+    "AWS provides SageMaker Clarity services to identify and measure bias level by calculating [pre-training](https://docs.aws.amazon.com/sagemaker/latest/dg/clarify-measure-data-bias.html) and [post-training](https://docs.aws.amazon.com/sagemaker/latest/dg/clarify-measure-post-training-bias.html) bias metrics based on model input, output, and ground truth information gathered and added during monitoring process and comparing it to calculated in advance baseline metrics."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "4d6d91b3-4a1a-4384-b836-cdf69c6382f4",
+   "metadata": {},
+   "source": [
+    "## 1. Configuration"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "db0aa694-f45f-4051-b253-1870d150a9ab",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Requirement already satisfied: boto3 in /opt/conda/lib/python3.7/site-packages (1.26.80)\n",
+      "Requirement already satisfied: s3transfer<0.7.0,>=0.6.0 in /opt/conda/lib/python3.7/site-packages (from boto3) (0.6.0)\n",
+      "Requirement already satisfied: jmespath<2.0.0,>=0.7.1 in /opt/conda/lib/python3.7/site-packages (from boto3) (1.0.1)\n",
+      "Requirement already satisfied: botocore<1.30.0,>=1.29.80 in /opt/conda/lib/python3.7/site-packages (from boto3) (1.29.80)\n",
+      "Requirement already satisfied: urllib3<1.27,>=1.25.4 in /opt/conda/lib/python3.7/site-packages (from botocore<1.30.0,>=1.29.80->boto3) (1.26.14)\n",
+      "Requirement already satisfied: python-dateutil<3.0.0,>=2.1 in /opt/conda/lib/python3.7/site-packages (from botocore<1.30.0,>=1.29.80->boto3) (2.8.2)\n",
+      "Requirement already satisfied: six>=1.5 in /opt/conda/lib/python3.7/site-packages (from python-dateutil<3.0.0,>=2.1->botocore<1.30.0,>=1.29.80->boto3) (1.14.0)\n",
+      "\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\u001b[33m\n",
+      "\u001b[0m\n",
+      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m A new release of pip is available: \u001b[0m\u001b[31;49m23.0\u001b[0m\u001b[39;49m -> \u001b[0m\u001b[32;49m23.0.1\u001b[0m\n",
+      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m To update, run: \u001b[0m\u001b[32;49mpip install --upgrade pip\u001b[0m\n"
+     ]
+    }
+   ],
+   "source": [
+    "!pip install --upgrade boto3"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "3dccd774-9ef4-477a-a409-4e8a4606ee94",
+   "metadata": {},
+   "source": [
+    "### Importing necessary libraries"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "23465867-bed7-4cef-8699-3eeab0903008",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import copy\n",
+    "import json\n",
+    "import random\n",
+    "import time\n",
+    "import matplotlib.pyplot as plt\n",
+    "import numpy as np\n",
+    "import pandas as pd\n",
+    "import sagemaker\n",
+    "\n",
+    "from datetime import datetime\n",
+    "\n",
+    "from IPython.display import display\n",
+    "from sagemaker import get_execution_role, image_uris\n",
+    "from sagemaker.clarify import (\n",
+    "    BiasConfig,\n",
+    "    DataConfig,\n",
+    "    ModelConfig,\n",
+    "    ModelPredictedLabelConfig,\n",
+    ")\n",
+    "from sagemaker.model import Model\n",
+    "from sagemaker.model_monitor import (\n",
+    "    BiasAnalysisConfig,\n",
+    "    CronExpressionGenerator,\n",
+    "    DataCaptureConfig,\n",
+    "    EndpointInput,\n",
+    "    ModelBiasMonitor,\n",
+    ")\n",
+    "from sagemaker.s3 import S3Downloader, S3Uploader\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "735c2543-a393-467c-a0f8-c7c6d6c3a9a1",
+   "metadata": {},
+   "source": [
+    "### Setting up necessary variables and constants"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "be24abcc-7a7f-429b-b635-a33c30a39625",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "sagemaker_session = sagemaker.session.Session()\n",
+    "sagemaker_client = sagemaker_session.sagemaker_client\n",
+    "sagemaker_runtime_client = sagemaker_session.sagemaker_runtime_client\n",
+    "role =  get_execution_role()\n",
+    "\n",
+    "DATA_BUCKET = 'adp-rnd-ml-datasets'\n",
+    "# Change the following bucket names if you want to run this code outside GridDynamics to the ones you have accees to.\n",
+    "STAGE_BUCKET = 'adp-rnd-ml-stage'\n",
+    "MODEL_BUCKET = 'adp-rnd-ml-models'\n",
+    "MODEL_PATH = f\"s3://{MODEL_BUCKET}/xgboost-for-loan-default-data/output/xgboost-for-loan-default-data-2023-02-24-08-18-30-629/output/model.tar.gz\"\n",
+    "DATASET_PATH = f\"s3://{DATA_BUCKET}/loan_default/dataset.zip\"\n",
+    "TEST_PATH = f\"s3://{DATA_BUCKET}/loan_default/test.csv\"\n",
+    "VALIDATION_PATH = f\"s3://{DATA_BUCKET}/loan_default/eval.csv\"\n",
+    "\n",
+    "NOW = datetime.now()\n",
+    "\n",
+    "%matplotlib inline\n",
+    "plt.style.use(\"seaborn-pastel\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "f06b3c0a-117f-415f-b30b-a52938c0c623",
+   "metadata": {},
+   "source": [
+    "## 2. Dataset"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "6414debb-84e5-4f6a-9914-fe83ad09300b",
+   "metadata": {},
+   "source": [
+    "<a href=\"https://www.kaggle.com/datasets/yasserh/loan-default-dataset\">Loan Default Dataset</a> was used for this work. It's published under CC0: Public Domain license, hence both non-profit, and commercial usage is permitted. It consists of 33 features of different types (categorical/binary, continuous/decimal) including the binary target feature **Status** which is used for the classification task. The dataset has set of columns with missing values, which needs to be handled before using with the machine learning models. It has outliers, which needs to be handled, it is likely to have data leakege, which solves the task for 100% accuracy, so such features are to remove."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "6c5acb59-42b8-49d9-986d-1790c3fe0c9d",
+   "metadata": {},
+   "source": [
+    "### Basic EDA"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "957b1e1e-004a-40fb-b96c-300e12153d00",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "<class 'pandas.core.frame.DataFrame'>\n",
+      "Int64Index: 148670 entries, 24890 to 173559\n",
+      "Data columns (total 33 columns):\n",
+      " #   Column                     Non-Null Count   Dtype  \n",
+      "---  ------                     --------------   -----  \n",
+      " 0   year                       148670 non-null  int64  \n",
+      " 1   loan_limit                 145326 non-null  object \n",
+      " 2   Gender                     148670 non-null  object \n",
+      " 3   approv_in_adv              147762 non-null  object \n",
+      " 4   loan_type                  148670 non-null  object \n",
+      " 5   loan_purpose               148536 non-null  object \n",
+      " 6   Credit_Worthiness          148670 non-null  object \n",
+      " 7   open_credit                148670 non-null  object \n",
+      " 8   business_or_commercial     148670 non-null  object \n",
+      " 9   loan_amount                148670 non-null  int64  \n",
+      " 10  rate_of_interest           112231 non-null  float64\n",
+      " 11  Interest_rate_spread       112031 non-null  float64\n",
+      " 12  Upfront_charges            109028 non-null  float64\n",
+      " 13  term                       148629 non-null  float64\n",
+      " 14  Neg_ammortization          148549 non-null  object \n",
+      " 15  interest_only              148670 non-null  object \n",
+      " 16  lump_sum_payment           148670 non-null  object \n",
+      " 17  property_value             133572 non-null  float64\n",
+      " 18  construction_type          148670 non-null  object \n",
+      " 19  occupancy_type             148670 non-null  object \n",
+      " 20  Secured_by                 148670 non-null  object \n",
+      " 21  total_units                148670 non-null  object \n",
+      " 22  income                     139520 non-null  float64\n",
+      " 23  credit_type                148670 non-null  object \n",
+      " 24  Credit_Score               148670 non-null  int64  \n",
+      " 25  co-applicant_credit_type   148670 non-null  object \n",
+      " 26  age                        148470 non-null  object \n",
+      " 27  submission_of_application  148470 non-null  object \n",
+      " 28  LTV                        133572 non-null  float64\n",
+      " 29  Region                     148670 non-null  object \n",
+      " 30  Security_Type              148670 non-null  object \n",
+      " 31  Status                     148670 non-null  int64  \n",
+      " 32  dtir1                      124549 non-null  float64\n",
+      "dtypes: float64(8), int64(4), object(21)\n",
+      "memory usage: 38.6+ MB\n"
+     ]
+    }
+   ],
+   "source": [
+    "dataset = pd.read_csv(DATASET_PATH, compression='zip', index_col='ID')\n",
+    "dataset.info()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "f770040e-9539-4709-b744-72b522d2dc73",
+   "metadata": {
+    "tags": []
+   },
+   "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>year</th>\n",
+       "      <th>loan_limit</th>\n",
+       "      <th>Gender</th>\n",
+       "      <th>approv_in_adv</th>\n",
+       "      <th>loan_type</th>\n",
+       "      <th>loan_purpose</th>\n",
+       "      <th>Credit_Worthiness</th>\n",
+       "      <th>open_credit</th>\n",
+       "      <th>business_or_commercial</th>\n",
+       "      <th>loan_amount</th>\n",
+       "      <th>...</th>\n",
+       "      <th>credit_type</th>\n",
+       "      <th>Credit_Score</th>\n",
+       "      <th>co-applicant_credit_type</th>\n",
+       "      <th>age</th>\n",
+       "      <th>submission_of_application</th>\n",
+       "      <th>LTV</th>\n",
+       "      <th>Region</th>\n",
+       "      <th>Security_Type</th>\n",
+       "      <th>Status</th>\n",
+       "      <th>dtir1</th>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>ID</th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>24890</th>\n",
+       "      <td>2019</td>\n",
+       "      <td>cf</td>\n",
+       "      <td>Sex Not Available</td>\n",
+       "      <td>nopre</td>\n",
+       "      <td>type1</td>\n",
+       "      <td>p1</td>\n",
+       "      <td>l1</td>\n",
+       "      <td>nopc</td>\n",
+       "      <td>nob/c</td>\n",
+       "      <td>116500</td>\n",
+       "      <td>...</td>\n",
+       "      <td>EXP</td>\n",
+       "      <td>758</td>\n",
+       "      <td>CIB</td>\n",
+       "      <td>25-34</td>\n",
+       "      <td>to_inst</td>\n",
+       "      <td>98.728814</td>\n",
+       "      <td>south</td>\n",
+       "      <td>direct</td>\n",
+       "      <td>1</td>\n",
+       "      <td>45.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>24891</th>\n",
+       "      <td>2019</td>\n",
+       "      <td>cf</td>\n",
+       "      <td>Male</td>\n",
+       "      <td>nopre</td>\n",
+       "      <td>type2</td>\n",
+       "      <td>p1</td>\n",
+       "      <td>l1</td>\n",
+       "      <td>nopc</td>\n",
+       "      <td>b/c</td>\n",
+       "      <td>206500</td>\n",
+       "      <td>...</td>\n",
+       "      <td>EQUI</td>\n",
+       "      <td>552</td>\n",
+       "      <td>EXP</td>\n",
+       "      <td>55-64</td>\n",
+       "      <td>to_inst</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>North</td>\n",
+       "      <td>direct</td>\n",
+       "      <td>1</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>24892</th>\n",
+       "      <td>2019</td>\n",
+       "      <td>cf</td>\n",
+       "      <td>Male</td>\n",
+       "      <td>pre</td>\n",
+       "      <td>type1</td>\n",
+       "      <td>p1</td>\n",
+       "      <td>l1</td>\n",
+       "      <td>nopc</td>\n",
+       "      <td>nob/c</td>\n",
+       "      <td>406500</td>\n",
+       "      <td>...</td>\n",
+       "      <td>EXP</td>\n",
+       "      <td>834</td>\n",
+       "      <td>CIB</td>\n",
+       "      <td>35-44</td>\n",
+       "      <td>to_inst</td>\n",
+       "      <td>80.019685</td>\n",
+       "      <td>south</td>\n",
+       "      <td>direct</td>\n",
+       "      <td>0</td>\n",
+       "      <td>46.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>24893</th>\n",
+       "      <td>2019</td>\n",
+       "      <td>cf</td>\n",
+       "      <td>Male</td>\n",
+       "      <td>nopre</td>\n",
+       "      <td>type1</td>\n",
+       "      <td>p4</td>\n",
+       "      <td>l1</td>\n",
+       "      <td>nopc</td>\n",
+       "      <td>nob/c</td>\n",
+       "      <td>456500</td>\n",
+       "      <td>...</td>\n",
+       "      <td>EXP</td>\n",
+       "      <td>587</td>\n",
+       "      <td>CIB</td>\n",
+       "      <td>45-54</td>\n",
+       "      <td>not_inst</td>\n",
+       "      <td>69.376900</td>\n",
+       "      <td>North</td>\n",
+       "      <td>direct</td>\n",
+       "      <td>0</td>\n",
+       "      <td>42.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>24894</th>\n",
+       "      <td>2019</td>\n",
+       "      <td>cf</td>\n",
+       "      <td>Joint</td>\n",
+       "      <td>pre</td>\n",
+       "      <td>type1</td>\n",
+       "      <td>p1</td>\n",
+       "      <td>l1</td>\n",
+       "      <td>nopc</td>\n",
+       "      <td>nob/c</td>\n",
+       "      <td>696500</td>\n",
+       "      <td>...</td>\n",
+       "      <td>CRIF</td>\n",
+       "      <td>602</td>\n",
+       "      <td>EXP</td>\n",
+       "      <td>25-34</td>\n",
+       "      <td>not_inst</td>\n",
+       "      <td>91.886544</td>\n",
+       "      <td>North</td>\n",
+       "      <td>direct</td>\n",
+       "      <td>0</td>\n",
+       "      <td>39.0</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>5 rows × 33 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "       year loan_limit             Gender approv_in_adv loan_type  \\\n",
+       "ID                                                                  \n",
+       "24890  2019         cf  Sex Not Available         nopre     type1   \n",
+       "24891  2019         cf               Male         nopre     type2   \n",
+       "24892  2019         cf               Male           pre     type1   \n",
+       "24893  2019         cf               Male         nopre     type1   \n",
+       "24894  2019         cf              Joint           pre     type1   \n",
+       "\n",
+       "      loan_purpose Credit_Worthiness open_credit business_or_commercial  \\\n",
+       "ID                                                                        \n",
+       "24890           p1                l1        nopc                  nob/c   \n",
+       "24891           p1                l1        nopc                    b/c   \n",
+       "24892           p1                l1        nopc                  nob/c   \n",
+       "24893           p4                l1        nopc                  nob/c   \n",
+       "24894           p1                l1        nopc                  nob/c   \n",
+       "\n",
+       "       loan_amount  ...  credit_type  Credit_Score  co-applicant_credit_type  \\\n",
+       "ID                  ...                                                        \n",
+       "24890       116500  ...          EXP           758                       CIB   \n",
+       "24891       206500  ...         EQUI           552                       EXP   \n",
+       "24892       406500  ...          EXP           834                       CIB   \n",
+       "24893       456500  ...          EXP           587                       CIB   \n",
+       "24894       696500  ...         CRIF           602                       EXP   \n",
+       "\n",
+       "         age submission_of_application        LTV Region  Security_Type  \\\n",
+       "ID                                                                        \n",
+       "24890  25-34                   to_inst  98.728814  south         direct   \n",
+       "24891  55-64                   to_inst        NaN  North         direct   \n",
+       "24892  35-44                   to_inst  80.019685  south         direct   \n",
+       "24893  45-54                  not_inst  69.376900  North         direct   \n",
+       "24894  25-34                  not_inst  91.886544  North         direct   \n",
+       "\n",
+       "      Status dtir1  \n",
+       "ID                  \n",
+       "24890      1  45.0  \n",
+       "24891      1   NaN  \n",
+       "24892      0  46.0  \n",
+       "24893      0  42.0  \n",
+       "24894      0  39.0  \n",
+       "\n",
+       "[5 rows x 33 columns]"
+      ]
+     },
+     "execution_count": 6,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# Few instances of dataset\n",
+    "dataset.head()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "1e542bba-9c2e-4df8-aaad-0ae5d8c085dd",
+   "metadata": {
+    "tags": []
+   },
+   "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>year</th>\n",
+       "      <th>loan_amount</th>\n",
+       "      <th>rate_of_interest</th>\n",
+       "      <th>Interest_rate_spread</th>\n",
+       "      <th>Upfront_charges</th>\n",
+       "      <th>term</th>\n",
+       "      <th>property_value</th>\n",
+       "      <th>income</th>\n",
+       "      <th>Credit_Score</th>\n",
+       "      <th>LTV</th>\n",
+       "      <th>Status</th>\n",
+       "      <th>dtir1</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>count</th>\n",
+       "      <td>148670.0</td>\n",
+       "      <td>1.486700e+05</td>\n",
+       "      <td>112231.000000</td>\n",
+       "      <td>112031.000000</td>\n",
+       "      <td>109028.000000</td>\n",
+       "      <td>148629.000000</td>\n",
+       "      <td>1.335720e+05</td>\n",
+       "      <td>139520.000000</td>\n",
+       "      <td>148670.000000</td>\n",
+       "      <td>133572.000000</td>\n",
+       "      <td>148670.000000</td>\n",
+       "      <td>124549.000000</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>mean</th>\n",
+       "      <td>2019.0</td>\n",
+       "      <td>3.311177e+05</td>\n",
+       "      <td>4.045476</td>\n",
+       "      <td>0.441656</td>\n",
+       "      <td>3224.996127</td>\n",
+       "      <td>335.136582</td>\n",
+       "      <td>4.978935e+05</td>\n",
+       "      <td>6957.338876</td>\n",
+       "      <td>699.789103</td>\n",
+       "      <td>72.746457</td>\n",
+       "      <td>0.246445</td>\n",
+       "      <td>37.732932</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>std</th>\n",
+       "      <td>0.0</td>\n",
+       "      <td>1.839093e+05</td>\n",
+       "      <td>0.561391</td>\n",
+       "      <td>0.513043</td>\n",
+       "      <td>3251.121510</td>\n",
+       "      <td>58.409084</td>\n",
+       "      <td>3.599353e+05</td>\n",
+       "      <td>6496.586382</td>\n",
+       "      <td>115.875857</td>\n",
+       "      <td>39.967603</td>\n",
+       "      <td>0.430942</td>\n",
+       "      <td>10.545435</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>min</th>\n",
+       "      <td>2019.0</td>\n",
+       "      <td>1.650000e+04</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>-3.638000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>96.000000</td>\n",
+       "      <td>8.000000e+03</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>500.000000</td>\n",
+       "      <td>0.967478</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>5.000000</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>25%</th>\n",
+       "      <td>2019.0</td>\n",
+       "      <td>1.965000e+05</td>\n",
+       "      <td>3.625000</td>\n",
+       "      <td>0.076000</td>\n",
+       "      <td>581.490000</td>\n",
+       "      <td>360.000000</td>\n",
+       "      <td>2.680000e+05</td>\n",
+       "      <td>3720.000000</td>\n",
+       "      <td>599.000000</td>\n",
+       "      <td>60.474860</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>31.000000</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>50%</th>\n",
+       "      <td>2019.0</td>\n",
+       "      <td>2.965000e+05</td>\n",
+       "      <td>3.990000</td>\n",
+       "      <td>0.390400</td>\n",
+       "      <td>2596.450000</td>\n",
+       "      <td>360.000000</td>\n",
+       "      <td>4.180000e+05</td>\n",
+       "      <td>5760.000000</td>\n",
+       "      <td>699.000000</td>\n",
+       "      <td>75.135870</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>39.000000</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>75%</th>\n",
+       "      <td>2019.0</td>\n",
+       "      <td>4.365000e+05</td>\n",
+       "      <td>4.375000</td>\n",
+       "      <td>0.775400</td>\n",
+       "      <td>4812.500000</td>\n",
+       "      <td>360.000000</td>\n",
+       "      <td>6.280000e+05</td>\n",
+       "      <td>8520.000000</td>\n",
+       "      <td>800.000000</td>\n",
+       "      <td>86.184211</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>45.000000</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>max</th>\n",
+       "      <td>2019.0</td>\n",
+       "      <td>3.576500e+06</td>\n",
+       "      <td>8.000000</td>\n",
+       "      <td>3.357000</td>\n",
+       "      <td>60000.000000</td>\n",
+       "      <td>360.000000</td>\n",
+       "      <td>1.650800e+07</td>\n",
+       "      <td>578580.000000</td>\n",
+       "      <td>900.000000</td>\n",
+       "      <td>7831.250000</td>\n",
+       "      <td>1.000000</td>\n",
+       "      <td>61.000000</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "           year   loan_amount  rate_of_interest  Interest_rate_spread  \\\n",
+       "count  148670.0  1.486700e+05     112231.000000         112031.000000   \n",
+       "mean     2019.0  3.311177e+05          4.045476              0.441656   \n",
+       "std         0.0  1.839093e+05          0.561391              0.513043   \n",
+       "min      2019.0  1.650000e+04          0.000000             -3.638000   \n",
+       "25%      2019.0  1.965000e+05          3.625000              0.076000   \n",
+       "50%      2019.0  2.965000e+05          3.990000              0.390400   \n",
+       "75%      2019.0  4.365000e+05          4.375000              0.775400   \n",
+       "max      2019.0  3.576500e+06          8.000000              3.357000   \n",
+       "\n",
+       "       Upfront_charges           term  property_value         income  \\\n",
+       "count    109028.000000  148629.000000    1.335720e+05  139520.000000   \n",
+       "mean       3224.996127     335.136582    4.978935e+05    6957.338876   \n",
+       "std        3251.121510      58.409084    3.599353e+05    6496.586382   \n",
+       "min           0.000000      96.000000    8.000000e+03       0.000000   \n",
+       "25%         581.490000     360.000000    2.680000e+05    3720.000000   \n",
+       "50%        2596.450000     360.000000    4.180000e+05    5760.000000   \n",
+       "75%        4812.500000     360.000000    6.280000e+05    8520.000000   \n",
+       "max       60000.000000     360.000000    1.650800e+07  578580.000000   \n",
+       "\n",
+       "        Credit_Score            LTV         Status          dtir1  \n",
+       "count  148670.000000  133572.000000  148670.000000  124549.000000  \n",
+       "mean      699.789103      72.746457       0.246445      37.732932  \n",
+       "std       115.875857      39.967603       0.430942      10.545435  \n",
+       "min       500.000000       0.967478       0.000000       5.000000  \n",
+       "25%       599.000000      60.474860       0.000000      31.000000  \n",
+       "50%       699.000000      75.135870       0.000000      39.000000  \n",
+       "75%       800.000000      86.184211       0.000000      45.000000  \n",
+       "max       900.000000    7831.250000       1.000000      61.000000  "
+      ]
+     },
+     "execution_count": 7,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# Shows statistics of numerical features\n",
+    "dataset.describe() "
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "f2b8a2dd-5ddd-4a76-b6a1-1f1e9c0e21a3",
+   "metadata": {},
+   "source": [
+    "Here one can get intuition of sensitive features distribution before preprocessing (handling missing values, and one-hot encoding):"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "id": "2e2da775-d508-4583-8212-5884bd0193e5",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 900x1200 with 4 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "SENSITIVE_COLUMNS = ['Gender', 'occupancy_type', 'age'] # + \"income\", but processed separately, as different type\n",
+    "\n",
+    "figs, axs = plt.subplots(4, 1, figsize=(9, 12))\n",
+    "for i, column in enumerate(SENSITIVE_COLUMNS):\n",
+    "    data = dataset[column].value_counts()\n",
+    "    axs[i].barh(data.index, data.values)\n",
+    "    axs[i].set_title(column, pad=3)\n",
+    "\n",
+    "axs[-1].hist(dataset['income'], bins=100, log=True)\n",
+    "axs[-1].set_title('income', pad=3)\n",
+    "plt.tight_layout()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "6d2b35a3-6e7d-4850-b091-9a1c006a18ce",
+   "metadata": {},
+   "source": [
+    "## 3. Deploying model"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "99413c71-e1e1-4a76-ae4b-582a25c0344c",
+   "metadata": {},
+   "source": [
+    "Setting up data capture config to be able to monitor model bias based on the stored data"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "id": "7176b64d-d8cc-4c80-83fa-977a9c9b0c83",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "S3 key: s3://adp-rnd-ml-stage/xgboost-for-loan-default-data-2023-02-28\n"
+     ]
+    }
+   ],
+   "source": [
+    "container = sagemaker.image_uris.retrieve(\"xgboost\", sagemaker_session.boto_region_name, \"1.5-1\")\n",
+    "\n",
+    "prefix = f'xgboost-for-loan-default-data-{NOW:%Y-%m-%d}'\n",
+    "s3_key = f\"s3://{STAGE_BUCKET}/{prefix}\"\n",
+    "print(f\"S3 key: {s3_key}\")\n",
+    "s3_capture_upload_path = f\"{s3_key}/datacapture\"\n",
+    "\n",
+    "data_capture_config = DataCaptureConfig(\n",
+    "    enable_capture=True,\n",
+    "    sampling_percentage=100,\n",
+    "    destination_s3_uri=s3_capture_upload_path,\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "e509f304-701d-435e-9062-86564f893d86",
+   "metadata": {},
+   "source": [
+    "Creating and deploying Model instance fox XGBoost classifier model which then will be used for ML Observability"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "id": "aa52d932-3ec3-479b-8c62-24eadb74945b",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "-------!"
+     ]
+    }
+   ],
+   "source": [
+    "xgb_model_name = f'loan-default-xgboost-{NOW:%Y-%m-%d}'\n",
+    "xgb_endpoint_name = f'{xgb_model_name}-endpoint'\n",
+    "\n",
+    "xgb_model = Model(\n",
+    "    image_uri=container,\n",
+    "    model_data=MODEL_PATH,\n",
+    "    role=role,\n",
+    "    name=xgb_model_name,\n",
+    "    sagemaker_session=sagemaker_session\n",
+    ")\n",
+    "\n",
+    "xgb_model.deploy(\n",
+    "    endpoint_name=xgb_endpoint_name,\n",
+    "    model_name=xgb_model_name,\n",
+    "    initial_instance_count=1, \n",
+    "    instance_type=\"ml.m4.xlarge\",\n",
+    "    data_capture_config=data_capture_config,\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "fef6ee74-b0fe-4b4b-a463-06eacd2c30dd",
+   "metadata": {},
+   "source": [
+    "### Results"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "9bbec4fa-af38-4985-ac47-d35f9798e850",
+   "metadata": {
+    "tags": []
+   },
+   "source": [
+    "As a result, one will get a Load Default model and an endpoint. You can find them in the AWS Console -> SageMaker -> Governance -> Model cards. The new model should appear with specified endpoint (and that's all for now)\n",
+    "\n",
+    "<img alt=\"governance.png\" src=\"images/governance.jpg\">\n",
+    "\n",
+    "\\* here and later all screenshots are given for illustration. You can have different names/values"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "5bcab79a-8deb-4734-8838-37ce22bccc53",
+   "metadata": {},
+   "source": [
+    "## 4. Setting up monitoring job"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "id": "653cb277-1a13-47c0-9030-b390dc52c6be",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Capture path: s3://adp-rnd-ml-stage/xgboost-for-loan-default-data-2023-02-28/datacapture\n",
+      "Ground truth path: s3://adp-rnd-ml-stage/xgboost-for-loan-default-data-2023-02-28/ground_truth_data\n",
+      "Report path: s3://adp-rnd-ml-stage/xgboost-for-loan-default-data-2023-02-28/reports\n",
+      "Baseline results uri: s3://adp-rnd-ml-stage/xgboost-for-loan-default-data-2023-02-28/baselining\n"
+     ]
+    }
+   ],
+   "source": [
+    "monitoring_prefix = f\"{prefix}/ClarifyModelMonitor-{NOW:%Y-%m-%d}\"\n",
+    "\n",
+    "ground_truth_upload_path = f\"{s3_key}/ground_truth_data\"\n",
+    "s3_report_path = f\"{s3_key}/reports\"\n",
+    "\n",
+    "print(f\"Capture path: {s3_capture_upload_path}\")\n",
+    "print(f\"Ground truth path: {ground_truth_upload_path}\")\n",
+    "print(f\"Report path: {s3_report_path}\")\n",
+    "\n",
+    "baseline_results_uri = f\"{s3_key}/baselining\"\n",
+    "print(f\"Baseline results uri: {baseline_results_uri}\")\n",
+    "model_bias_baselining_job_result_uri = f\"{baseline_results_uri}/model_bias\"\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "d1a9ee97-66cb-44a2-82eb-6ba9fb258759",
+   "metadata": {},
+   "source": [
+    "After preprocessing some columns were dropped, some were transformed into numerical by converting binary columns to 0/1 columns, and some were one-hot encoded:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 12,
+   "id": "9c456e85-3d8c-4e6d-b740-e4bf7177f117",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "TARGET_COLUMN = 'Status'\n",
+    "ALL_INPUT_COLUMNS = ['Status', 'loan_amount', 'term', 'income', 'Credit_Score', 'loan_limit',\n",
+    "       'approv_in_adv', 'Credit_Worthiness', 'open_credit',\n",
+    "       'business_or_commercial', 'Neg_ammortization', 'interest_only',\n",
+    "       'lump_sum_payment', 'Secured_by', 'co-applicant_credit_type',\n",
+    "       'submission_of_application', 'Gender_Female', 'Gender_Joint',\n",
+    "       'Gender_Male', 'Gender_Sex Not Available', 'loan_purpose_p1',\n",
+    "       'loan_purpose_p2', 'loan_purpose_p3', 'loan_purpose_p4',\n",
+    "       'occupancy_type_ir', 'occupancy_type_pr', 'occupancy_type_sr',\n",
+    "       'total_units_1U', 'total_units_2U', 'total_units_3U', 'total_units_4U',\n",
+    "       'age_25-34', 'age_35-44', 'age_45-54', 'age_55-64', 'age_65-74',\n",
+    "       'age_gt74', 'age_lt25']\n",
+    "DATASET_TYPE = \"text/csv\""
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "a92fd3c2-48e9-4c2f-9906-854d1875a480",
+   "metadata": {},
+   "source": [
+    "Creating and configuring monitor:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "id": "1978e0a4-2324-4ee0-901d-4a3fcf5591dd",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "model_bias_monitor = ModelBiasMonitor(\n",
+    "    role=role,\n",
+    "    sagemaker_session=sagemaker_session,\n",
+    "    max_runtime_in_seconds=1800,\n",
+    ")\n",
+    "\n",
+    "model_bias_data_config = DataConfig(\n",
+    "    s3_data_input_path=VALIDATION_PATH,\n",
+    "    s3_output_path=model_bias_baselining_job_result_uri,\n",
+    "    label=TARGET_COLUMN,\n",
+    "    headers=ALL_INPUT_COLUMNS,\n",
+    "    dataset_type=DATASET_TYPE,\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "eb2a0667-9f49-4254-9d99-c3dfa299239f",
+   "metadata": {},
+   "source": [
+    "Setting up facets - sensitive features based on which model bias will be calculated and later observed"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 14,
+   "id": "2398faa8-73c9-411f-add7-9eea111978bd",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "FACET_COLUMNS_VALUES = {\n",
+    "    facet: values \n",
+    "    for facet, values\n",
+    "    in zip(\n",
+    "        [\n",
+    "            'Gender_Female', 'Gender_Joint', 'Gender_Male', 'Gender_Sex Not Available', \n",
+    "            'income',\n",
+    "            'occupancy_type_ir', 'occupancy_type_pr', 'occupancy_type_sr',\n",
+    "            'age_25-34', 'age_35-44', 'age_45-54', 'age_55-64', 'age_65-74', 'age_gt74', 'age_lt25'\n",
+    "        ],\n",
+    "        [\n",
+    "            [1], [1], [1], [1], # Gender - Binary data, assessing only positive values of one-hot-encoded features\n",
+    "            [25000.], # income - threshold splitting into 2 groups with income level less and more then 30000.0 \n",
+    "            [1], [1], [1], # occupancy_pyte - Binary data, assessing only positive values of one-hot-encoded features \n",
+    "            [1], [1], [1], [1], [1], [1], [1] # age -  Binary data, assessing only positive values of one-hot-encoded features\n",
+    "        ]\n",
+    "    )\n",
+    "}"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 15,
+   "id": "3740e90b-cb2a-4b12-912c-c4f1b2ab7a99",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "model_bias_config = BiasConfig(\n",
+    "    # Loan default value of target feature Status is 1. \n",
+    "    # Thus, we'll choose 0 as positive outcome, so Bias metrics could make sense\n",
+    "    label_values_or_threshold=[0],\n",
+    "    facet_name=list(FACET_COLUMNS_VALUES.keys()),\n",
+    "    facet_values_or_threshold=list(FACET_COLUMNS_VALUES.values()),\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 16,
+   "id": "35d4de0a-0df7-47d3-8197-f5693f7cecef",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "model_predicted_label_config = ModelPredictedLabelConfig(probability_threshold=.5)\n",
+    "model_config = ModelConfig(\n",
+    "    instance_count=1,\n",
+    "    instance_type=\"ml.m4.xlarge\",\n",
+    "    content_type=DATASET_TYPE,\n",
+    "    accept_type=DATASET_TYPE,\n",
+    "    endpoint_name=xgb_endpoint_name,\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "3738f518-c0d6-4515-ba22-3e52a8bcba25",
+   "metadata": {},
+   "source": [
+    "### Creating baseline"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 17,
+   "id": "222ad0c1-ef20-479c-8bf2-038ddea23425",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "INFO:sagemaker:Creating processing-job with name baseline-suggestion-job-2023-02-28-15-46-10-999\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      ".............................................................................!\n"
+     ]
+    }
+   ],
+   "source": [
+    "model_bias_monitor.suggest_baseline(\n",
+    "    model_config=model_config,\n",
+    "    data_config=model_bias_data_config,\n",
+    "    bias_config=model_bias_config,\n",
+    "    model_predicted_label_config=model_predicted_label_config,\n",
+    ")\n",
+    "\n",
+    "model_bias_monitor.latest_baselining_job.wait(logs=False)\n",
+    "model_bias_constraints = model_bias_monitor.suggested_constraints()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 18,
+   "id": "6a70d948-f54a-4643-9495-564c7f80e28b",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "ModelBiasMonitor suggested constraints: s3://adp-rnd-ml-stage/xgboost-for-loan-default-data-2023-02-28/baselining/model_bias/analysis.json\n",
+      "Visualizing part of suggested constrains:\n",
+      "{\n",
+      "    \"version\": \"1.0\",\n",
+      "    \"post_training_bias_metrics\": {\n",
+      "        \"label\": \"Status\",\n",
+      "        \"facets\": {\n",
+      "            \"Gender_Female\": [\n",
+      "                {\n",
+      "                    \"value_or_threshold\": \"1\",\n",
+      "                    \"metrics\": [\n",
+      "                        {\n",
+      "                            \"name\": \"AD\",\n",
+      "                            \"description\": \"Accuracy Difference (AD)\",\n",
+      "                            \"value\": -0.012274862880233273\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"CDDPL\",\n",
+      "                            \"description\": \"Conditional Demographic Disparity in Predicted Labels (CDDPL)\",\n",
+      "                            \"value\": null,\n",
+      "                            \"error\": \"Group variable is empty or not provided\"\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"DAR\",\n",
+      "                            \"description\": \"Difference in Acceptance Rates (DAR)\",\n",
+      "                            \"value\": -0.019687953899706012\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"DCA\",\n",
+      "                            \"description\": \"Difference in Conditional Acceptance (DCA)\",\n",
+      "                            \"value\": -0.035234696157777856\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"DCR\",\n",
+      "                            \"description\": \"Difference in Conditional Rejection (DCR)\",\n",
+      "                            \"value\": -0.5736707206241256\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"DI\",\n",
+      "                            \"description\": \"Disparate Impact (DI)\",\n",
+      "                            \"value\": 0.9544013145312146\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"DPPL\",\n",
+      "                            \"description\": \"Difference in Positive Proportions in Predicted Labels (DPPL)\",\n",
+      "                            \"value\": 0.040580765079404446\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"DRR\",\n",
+      "                            \"description\": \"Difference in Rejection Rates (DRR)\",\n",
+      "                            \"value\": -0.010994514578744008\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"FT\",\n",
+      "                            \"description\": \"Flip Test (FT)\",\n",
+      "                            \"value\": -0.11717435993272285\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"GE\",\n",
+      "                            \"description\": \"Generalized Entropy (GE)\",\n",
+      "                            \"value\": 0.06776540220552933\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"RD\",\n",
+      "                            \"description\": \"Recall Difference (RD)\",\n",
+      "                            \"value\": 0.01614804633418354\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"SD\",\n",
+      "                            \"description\": \"Specificity Difference (SD)\",\n",
+      "                            \"value\": 0.10832164460419041\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"TE\",\n",
+      "                            \"description\": \"Treatment Equality (TE)\",\n",
+      "                            \"value\": 0.11461989439613904\n",
+      "                        }\n",
+      "                    ]\n",
+      "                }\n",
+      "            ],\n",
+      "            \"income\": \"...\"\n",
+      "        }\n",
+      "    },\n",
+      "    \"pre_training_bias_metrics\": {\n",
+      "        \"label\": \"Status\",\n",
+      "        \"facets\": {\n",
+      "            \"Gender_Female\": [\n",
+      "                {\n",
+      "                    \"value_or_threshold\": \"1\",\n",
+      "                    \"metrics\": [\n",
+      "                        {\n",
+      "                            \"name\": \"CDDL\",\n",
+      "                            \"description\": \"Conditional Demographic Disparity in Labels (CDDL)\",\n",
+      "                            \"value\": null,\n",
+      "                            \"error\": \"Group variable is empty or not provided\"\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"CI\",\n",
+      "                            \"description\": \"Class Imbalance (CI)\",\n",
+      "                            \"value\": 0.6389460544516042\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"DPL\",\n",
+      "                            \"description\": \"Difference in Positive Proportions in Labels (DPL)\",\n",
+      "                            \"value\": 0.004437718747468233\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"JS\",\n",
+      "                            \"description\": \"Jensen-Shannon Divergence (JS)\",\n",
+      "                            \"value\": 1.3179244474955353e-05\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"KL\",\n",
+      "                            \"description\": \"Kullback-Liebler Divergence (KL)\",\n",
+      "                            \"value\": 5.261255142023462e-05\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"KS\",\n",
+      "                            \"description\": \"Kolmogorov-Smirnov Distance (KS)\",\n",
+      "                            \"value\": 0.004437718747468261\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"LP\",\n",
+      "                            \"description\": \"L-p Norm (LP)\",\n",
+      "                            \"value\": 0.006275882038666939\n",
+      "                        },\n",
+      "                        {\n",
+      "                            \"name\": \"TVD\",\n",
+      "                            \"description\": \"Total Variation Distance (TVD)\",\n",
+      "                            \"value\": 0.004437718747468247\n",
+      "                        }\n",
+      "                    ]\n",
+      "                }\n",
+      "            ],\n",
+      "            \"income\": \"...\"\n",
+      "        }\n",
+      "    }\n",
+      "}\n"
+     ]
+    }
+   ],
+   "source": [
+    "print(f\"ModelBiasMonitor suggested constraints: {model_bias_constraints.file_s3_uri}\")\n",
+    "print(\"Visualizing part of suggested constrains:\")\n",
+    "suggested_constrains = json.loads(S3Downloader.read_file(model_bias_constraints.file_s3_uri))\n",
+    "res = {k: v for k, v in suggested_constrains.items() if 'metrics' not in k}\n",
+    "res['post_training_bias_metrics'] = dict(label='Status', facets=dict(Gender_Female=suggested_constrains['post_training_bias_metrics']['facets']['Gender_Female']))\n",
+    "res['pre_training_bias_metrics'] = dict(label='Status', facets=dict(Gender_Female=suggested_constrains['pre_training_bias_metrics']['facets']['Gender_Female']))\n",
+    "res['post_training_bias_metrics']['facets']['income'] = '...'\n",
+    "res['pre_training_bias_metrics']['facets']['income'] = '...'\n",
+    "print(json.dumps(res, indent=4))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "67cc455b-fed0-493f-8ce2-fa78aa8bd7c4",
+   "metadata": {},
+   "source": [
+    "### Setting up monitoring job"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 19,
+   "id": "9444f5ad-4a5a-4ea5-8677-32930a1cf101",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "INFO:sagemaker.model_monitor.clarify_model_monitoring:Uploading analysis config to {s3_uri}.\n",
+      "INFO:sagemaker.model_monitor.model_monitoring:Creating Monitoring Schedule with name: monitoring-schedule-2023-02-28-15-53-44-998\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model bias monitoring schedule: monitoring-schedule-2023-02-28-15-53-44-998\n"
+     ]
+    }
+   ],
+   "source": [
+    "model_bias_analysis_config = None\n",
+    "if not model_bias_monitor.latest_baselining_job:\n",
+    "    model_bias_analysis_config = BiasAnalysisConfig(\n",
+    "        model_bias_config,\n",
+    "        headers=all_headers,\n",
+    "        label=label_header,\n",
+    "    )\n",
+    "\n",
+    "model_bias_monitor.create_monitoring_schedule(\n",
+    "    analysis_config=model_bias_analysis_config,\n",
+    "    output_s3_uri=s3_report_path,\n",
+    "    endpoint_input=EndpointInput(\n",
+    "        endpoint_name=xgb_endpoint_name,\n",
+    "        destination=\"/opt/ml/processing/input/endpoint\",\n",
+    "        start_time_offset=\"-PT1H\",\n",
+    "        end_time_offset=\"-PT0H\",\n",
+    "        probability_threshold_attribute=0.5,\n",
+    "    ),\n",
+    "    ground_truth_input=ground_truth_upload_path,\n",
+    "    schedule_cron_expression=CronExpressionGenerator.hourly(),\n",
+    ")\n",
+    "print(f\"Model bias monitoring schedule: {model_bias_monitor.monitoring_schedule_name}\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "c01ea526-ef99-4adb-9b97-2ee3a3a1d90e",
+   "metadata": {},
+   "source": [
+    "### Results"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "54f4f8d1-5466-434e-9bee-2dc7600c73b2",
+   "metadata": {},
+   "source": [
+    "Model monitoring job will be created. It might be found in the model dashboard if we go to the card or in the SageMaker Studio:\n",
+    "SageMaker Studio -> Home -> Deployments -> Endpoint -> _Model name_ -> Model bias\n",
+    "\n",
+    "\n",
+    "![studio_model](images/studio_model.jpg)\n",
+    "\n",
+    "\n",
+    "<img alt=\"monitoring_job\" src=\"images/monitoring_job.jpg\" style=\"width: 800px; margin: auto\"/>\n",
+    "\n",
+    "After some time, when several reports will be generated (after the 5th chapter) one will be able to see bias metrics charts below"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "19261039-b5df-4a45-8f6d-7d6d5f3c8c87",
+   "metadata": {},
+   "source": [
+    "## 5. Generating traffic"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "0bf5d84b-4fcb-49bc-be27-155944a51cea",
+   "metadata": {},
+   "source": [
+    "Preparing data for checking for model bias. It can be any data supported by model."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 20,
+   "id": "1a7ed27b-94d2-4f09-8f84-099d3549032e",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "test_df = pd.read_csv(TEST_PATH)\n",
+    "for col in FACET_COLUMNS_VALUES:\n",
+    "    if col == 'income':\n",
+    "        continue\n",
+    "    test_df[col] = test_df[col].astype(int)\n",
+    "\n",
+    "gender_aspects = [col for col in FACET_COLUMNS_VALUES if 'Gender' in col]\n",
+    "descr_colnames = ['metric_kind', 'good', 'min', 'max', 'description',]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 21,
+   "id": "a3921c14-32da-4b89-871b-480d916e390a",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "# Serialization function which will make Clarity understand types of attributes\n",
+    "def cast_str(x):\n",
+    "    if x in [1.0, 1, 0, 0.0]:\n",
+    "        return(str(int(x)))\n",
+    "    return str(x)\n",
+    "\n",
+    "def serialize_example(example):\n",
+    "    return ','.join(map(cast_str, example))\n",
+    "\n",
+    "# For reproducibility\n",
+    "np.random.seed(42)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 22,
+   "id": "be41b804-1bfc-4c73-b356-089f7257b55d",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import threading\n",
+    "\n",
+    "class WorkerThread(threading.Thread):\n",
+    "    def __init__(self, do_run, *args, **kwargs):\n",
+    "        super(WorkerThread, self).__init__(*args, **kwargs)\n",
+    "        self.__do_run = do_run\n",
+    "        self.__terminate_event = threading.Event()\n",
+    "\n",
+    "    def terminate(self):\n",
+    "        self.__terminate_event.set()\n",
+    "\n",
+    "    def run(self):\n",
+    "        while not self.__terminate_event.is_set():\n",
+    "            self.__do_run(self.__terminate_event)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 23,
+   "id": "f0d9f9f1-3d80-483b-b33f-dcb1197de03b",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "def invoke(terminate_event):\n",
+    "    # Getting predictions from deployed model for test dataframe\n",
+    "    for i, row in test_df.drop(columns=[TARGET_COLUMN]).iterrows():\n",
+    "        payload = serialize_example(row)\n",
+    "        response = sagemaker_runtime_client.invoke_endpoint(\n",
+    "            EndpointName=xgb_endpoint_name,\n",
+    "            Body=payload,\n",
+    "            ContentType=DATASET_TYPE,\n",
+    "            InferenceId=str(i)\n",
+    "        )\n",
+    "        prediction = response[\"Body\"].read()\n",
+    "        time.sleep(0.01)\n",
+    "        if terminate_event.is_set():\n",
+    "            break\n",
+    "\n",
+    "def real_ground_truth_with_id(inference_id):\n",
+    "    # Ground truth requires following format:\n",
+    "    return {\n",
+    "        \"groundTruthData\": {\n",
+    "            \"data\": str(int(test_df.iloc[inference_id][TARGET_COLUMN])),\n",
+    "            \"encoding\": \"CSV\",\n",
+    "        },\n",
+    "        \"eventMetadata\": {\n",
+    "            \"eventId\": str(inference_id),\n",
+    "        },\n",
+    "        \"eventVersion\": \"0\",\n",
+    "    }\n",
+    "\n",
+    "\n",
+    "def upload_ground_truth(upload_time):\n",
+    "    records = [real_ground_truth_with_id(i) for i in range(len(test_df))]\n",
+    "    records = [json.dumps(r) for r in records]\n",
+    "    data_to_upload = \"\\n\".join(records)\n",
+    "    target_s3_uri = f\"{ground_truth_upload_path}/{upload_time:%Y/%m/%d/%H/%M%S}.jsonl\"\n",
+    "    S3Uploader.upload_string_as_file_body(data_to_upload, target_s3_uri)\n",
+    "\n",
+    "def generate_ground_truth(terminate_event):\n",
+    "    invoke(terminate_event)\n",
+    "    upload_ground_truth(datetime.utcnow())\n",
+    "    for _ in range(0, 60):\n",
+    "        time.sleep(60)\n",
+    "        if terminate_event.is_set():\n",
+    "            break"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 24,
+   "id": "28ba35fd-2c42-4799-8664-a671190dc0fd",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "ground_truth_thread = WorkerThread(do_run=generate_ground_truth)\n",
+    "ground_truth_thread.start()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "b7d0e8d9-f546-4a67-b67d-bc6305aa338a",
+   "metadata": {},
+   "source": [
+    "### Results"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "544a1b16-7f68-4ba7-9264-a4a238880b6c",
+   "metadata": {},
+   "source": [
+    "After generating some trafic we need to wait until it is processed, so we could visualize the results"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 25,
+   "id": "f3ab7723-75a3-42ee-a32b-22b084e164f5",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "def wait_for_execution_to_start(model_monitor):\n",
+    "    print(\n",
+    "        \"A hourly schedule will kick off executions ON the hour (plus 0 - 20 min buffer).\"\n",
+    "    )\n",
+    "\n",
+    "    print(\"Waiting for the first execution to happen\", end=\"\")\n",
+    "    schedule_desc = model_monitor.describe_schedule()\n",
+    "    while \"LastMonitoringExecutionSummary\" not in schedule_desc:\n",
+    "        schedule_desc = model_monitor.describe_schedule()\n",
+    "        print(\".\", end=\"\", flush=True)\n",
+    "        time.sleep(60)\n",
+    "    print()\n",
+    "    print(\"Done! Execution has been created\")\n",
+    "\n",
+    "    print(\"Now waiting for execution to start\", end=\"\")\n",
+    "    while schedule_desc[\"LastMonitoringExecutionSummary\"][\"MonitoringExecutionStatus\"] in \"Pending\":\n",
+    "        schedule_desc = model_monitor.describe_schedule()\n",
+    "        print(\".\", end=\"\", flush=True)\n",
+    "        time.sleep(10)\n",
+    "\n",
+    "    print()\n",
+    "    print(\"Done! Execution has started\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 26,
+   "id": "aca68c10-401d-42f9-9495-56dcc859fbc8",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "# Waits for the schedule to have last execution in a terminal status.\n",
+    "def wait_for_execution_to_finish(model_monitor):\n",
+    "    schedule_desc = model_monitor.describe_schedule()\n",
+    "    execution_summary = schedule_desc.get(\"LastMonitoringExecutionSummary\")\n",
+    "    if execution_summary is not None:\n",
+    "        print(\"Waiting for execution to finish\", end=\"\")\n",
+    "        while execution_summary[\"MonitoringExecutionStatus\"] not in [\n",
+    "            \"Completed\",\n",
+    "            \"CompletedWithViolations\",\n",
+    "            \"Failed\",\n",
+    "            \"Stopped\",\n",
+    "        ]:\n",
+    "            print(\".\", end=\"\", flush=True)\n",
+    "            time.sleep(60)\n",
+    "            schedule_desc = model_monitor.describe_schedule()\n",
+    "            execution_summary = schedule_desc[\"LastMonitoringExecutionSummary\"]\n",
+    "        print()\n",
+    "        print(\"Done! Execution has finished\")\n",
+    "    else:\n",
+    "        print(\"Last execution not found\")\n",
+    "    return schedule_desc"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "490a911a-0d69-4d63-b3eb-7e2e9cf0e0ed",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "A hourly schedule will kick off executions ON the hour (plus 0 - 20 min buffer).\n",
+      "Waiting for the first execution to happen........\n",
+      "Done! Execution has been created\n",
+      "Now waiting for execution to start\n",
+      "Done! Execution has started\n",
+      "Waiting for execution to finish...........\n",
+      "Done! Execution has finished\n"
+     ]
+    }
+   ],
+   "source": [
+    "wait_for_execution_to_start(model_bias_monitor)\n",
+    "schedule_desc = wait_for_execution_to_finish(model_bias_monitor)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "c7ceb6d1-2984-4942-8826-eed29041bf2e",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "metrix_values = pd.DataFrame.from_dict({\n",
+    "    # Post\n",
+    "    'AD': [0, -1, 1, 'Accuracy Difference (AD)'],\n",
+    "    'DAR': [0, -1, 1, 'Difference in Acceptance Rates (DAR)'],\n",
+    "    'DCA': [0, -np.Inf, np.Inf, 'Difference in Conditional Acceptance (DCA)'],\n",
+    "    'DCR': [0, -np.Inf, np.Inf, 'Difference in Conditional Rejection (DCR)'],\n",
+    "    'DI': [1, 0, np.Inf, 'Disparate Impact (DI)'],\n",
+    "    'DPPL': [0, -1, 1, 'Difference in Positive Proportions in Predictected Labels (DPPL)'],\n",
+    "    'DRR': [0, -1, 1, 'Difference in Rejection Rates (DRR)'], \n",
+    "    'FT': [0, -1, 1, 'Flip Test (FT)'],\n",
+    "    'GE': [0, 0, .5, 'Generalized Entropy (GE)'],\n",
+    "    'RD': [0, -1, 1, 'Recall Difference (RD)'],\n",
+    "    'SD': [0, -1, 1, 'Specificity Difference (SD)'],\n",
+    "    'TE': [0, -np.Inf, np.Inf, 'Treatment Equality (TE)'],\n",
+    "    # Pre\n",
+    "    'CI': [0, -1, 1, 'Class Imbalance (CI)'],\n",
+    "    'DPL': [0, -1, 1, 'Difference in Positive Proportions in Labels (DPL)'],\n",
+    "    'JS': [0, 0, np.Inf, 'Jensen-Shannon Divergence (JS)'],\n",
+    "    'KL': [0, 0, np.Inf, 'Kullback-Liebler Divergence (KL)'],\n",
+    "    'KS': [0, 0, 1, 'Kolmogorov-Smirnov Distance (KS)'],\n",
+    "    'LP':  [0, 0, np.Inf, 'L-p Norm (LP)'],\n",
+    "    'TVD':  [0, 0, np.Inf, 'Total Variation Distance (TVD)'],\n",
+    "}, orient='index', columns=['good', 'min', 'max', 'description'])\n",
+    "\n",
+    "def extract_metrix_from_analysis(analysis_dict):\n",
+    "    post_training = analysis_dict['post_training_bias_metrics']\n",
+    "    pre_training = analysis_dict['pre_training_bias_metrics']\n",
+    "    facets = post_training['facets'].keys()\n",
+    "    frames = dict()\n",
+    "    \n",
+    "    for metric_kind in ['post_training_bias_metrics', 'pre_training_bias_metrics']:\n",
+    "        metric = analysis_dict[metric_kind]\n",
+    "        frames[metric_kind] = dict()\n",
+    "        for facet in facets:\n",
+    "            frames[metric_kind][facet] = []\n",
+    "            for facet_analysis in metric['facets'][facet]:\n",
+    "                frames[metric_kind][facet].append(\n",
+    "                    pd.DataFrame(facet_analysis['metrics'])\n",
+    "                    .drop(columns=['error', 'description'])\n",
+    "                    .set_index('name')\n",
+    "                    # as all features are binary:\n",
+    "                    .rename(columns={'value': facet})\n",
+    "                )\n",
+    "            frames[metric_kind][facet] = pd.concat(frames[metric_kind][facet], axis=1)\n",
+    "        frames[metric_kind] = pd.concat(frames[metric_kind].values(), axis=1)\n",
+    "        frames[metric_kind]['metric_kind'] = metric_kind\n",
+    "    return pd.concat(frames.values()).join(metrix_values, how='inner')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 31,
+   "id": "305c2eae-6df8-48a0-8317-755e85ceee46",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Report URI: s3://adp-rnd-ml-stage/xgboost-for-loan-default-data-2023-02-28/reports/loan-default-xgboost-2023-02-28-endpoint/monitoring-schedule-2023-02-28-15-53-44-998/2023/02/28/16\n",
+      "Found Report Files:\n",
+      "s3://adp-rnd-ml-stage/xgboost-for-loan-default-data-2023-02-28/reports/loan-default-xgboost-2023-02-28-endpoint/monitoring-schedule-2023-02-28-15-53-44-998/2023/02/28/16/analysis.json\n",
+      " s3://adp-rnd-ml-stage/xgboost-for-loan-default-data-2023-02-28/reports/loan-default-xgboost-2023-02-28-endpoint/monitoring-schedule-2023-02-28-15-53-44-998/2023/02/28/16/constraint_violations.json\n",
+      " s3://adp-rnd-ml-stage/xgboost-for-loan-default-data-2023-02-28/reports/loan-default-xgboost-2023-02-28-endpoint/monitoring-schedule-2023-02-28-15-53-44-998/2023/02/28/16/report.html\n",
+      " s3://adp-rnd-ml-stage/xgboost-for-loan-default-data-2023-02-28/reports/loan-default-xgboost-2023-02-28-endpoint/monitoring-schedule-2023-02-28-15-53-44-998/2023/02/28/16/report.ipynb\n",
+      " s3://adp-rnd-ml-stage/xgboost-for-loan-default-data-2023-02-28/reports/loan-default-xgboost-2023-02-28-endpoint/monitoring-schedule-2023-02-28-15-53-44-998/2023/02/28/16/report.pdf\n",
+      "\n",
+      "Gender aspects pre-training bias metrics:\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<style type=\"text/css\">\n",
+       "#T_42a61_row1_col0, #T_42a61_row1_col1, #T_42a61_row1_col2, #T_42a61_row2_col0, #T_42a61_row2_col1, #T_42a61_row2_col2, #T_42a61_row2_col3, #T_42a61_row3_col0, #T_42a61_row3_col1, #T_42a61_row3_col2, #T_42a61_row3_col3, #T_42a61_row4_col0, #T_42a61_row4_col1, #T_42a61_row4_col2, #T_42a61_row5_col0, #T_42a61_row5_col2, #T_42a61_row6_col0, #T_42a61_row6_col1, #T_42a61_row6_col2 {\n",
+       "  opacity: 20%;\n",
+       "}\n",
+       "</style>\n",
+       "<table id=\"T_42a61_\">\n",
+       "  <thead>\n",
+       "    <tr>\n",
+       "      <th class=\"blank level0\" >&nbsp;</th>\n",
+       "      <th class=\"col_heading level0 col0\" >Gender_Female</th>\n",
+       "      <th class=\"col_heading level0 col1\" >Gender_Joint</th>\n",
+       "      <th class=\"col_heading level0 col2\" >Gender_Male</th>\n",
+       "      <th class=\"col_heading level0 col3\" >Gender_Sex Not Available</th>\n",
+       "      <th class=\"col_heading level0 col4\" >metric_kind</th>\n",
+       "      <th class=\"col_heading level0 col5\" >good</th>\n",
+       "      <th class=\"col_heading level0 col6\" >min</th>\n",
+       "      <th class=\"col_heading level0 col7\" >max</th>\n",
+       "      <th class=\"col_heading level0 col8\" >description</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th id=\"T_42a61_level0_row0\" class=\"row_heading level0 row0\" >CI</th>\n",
+       "      <td id=\"T_42a61_row0_col0\" class=\"data row0 col0\" >0.631107</td>\n",
+       "      <td id=\"T_42a61_row0_col1\" class=\"data row0 col1\" >0.439681</td>\n",
+       "      <td id=\"T_42a61_row0_col2\" class=\"data row0 col2\" >0.433699</td>\n",
+       "      <td id=\"T_42a61_row0_col3\" class=\"data row0 col3\" >0.495513</td>\n",
+       "      <td id=\"T_42a61_row0_col4\" class=\"data row0 col4\" >pre_training_bias_metrics</td>\n",
+       "      <td id=\"T_42a61_row0_col5\" class=\"data row0 col5\" >0</td>\n",
+       "      <td id=\"T_42a61_row0_col6\" class=\"data row0 col6\" >-1.000000</td>\n",
+       "      <td id=\"T_42a61_row0_col7\" class=\"data row0 col7\" >1.000000</td>\n",
+       "      <td id=\"T_42a61_row0_col8\" class=\"data row0 col8\" >Class Imbalance (CI)</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th id=\"T_42a61_level0_row1\" class=\"row_heading level0 row1\" >DPL</th>\n",
+       "      <td id=\"T_42a61_row1_col0\" class=\"data row1 col0\" >-0.029181</td>\n",
+       "      <td id=\"T_42a61_row1_col1\" class=\"data row1 col1\" >-0.098880</td>\n",
+       "      <td id=\"T_42a61_row1_col2\" class=\"data row1 col2\" >0.025691</td>\n",
+       "      <td id=\"T_42a61_row1_col3\" class=\"data row1 col3\" >0.101349</td>\n",
+       "      <td id=\"T_42a61_row1_col4\" class=\"data row1 col4\" >pre_training_bias_metrics</td>\n",
+       "      <td id=\"T_42a61_row1_col5\" class=\"data row1 col5\" >0</td>\n",
+       "      <td id=\"T_42a61_row1_col6\" class=\"data row1 col6\" >-1.000000</td>\n",
+       "      <td id=\"T_42a61_row1_col7\" class=\"data row1 col7\" >1.000000</td>\n",
+       "      <td id=\"T_42a61_row1_col8\" class=\"data row1 col8\" >Difference in Positive Proportions in Labels (DPL)</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th id=\"T_42a61_level0_row2\" class=\"row_heading level0 row2\" >JS</th>\n",
+       "      <td id=\"T_42a61_row2_col0\" class=\"data row2 col0\" >0.000573</td>\n",
+       "      <td id=\"T_42a61_row2_col1\" class=\"data row2 col1\" >0.006849</td>\n",
+       "      <td id=\"T_42a61_row2_col2\" class=\"data row2 col2\" >0.000427</td>\n",
+       "      <td id=\"T_42a61_row2_col3\" class=\"data row2 col3\" >0.006377</td>\n",
+       "      <td id=\"T_42a61_row2_col4\" class=\"data row2 col4\" >pre_training_bias_metrics</td>\n",
+       "      <td id=\"T_42a61_row2_col5\" class=\"data row2 col5\" >0</td>\n",
+       "      <td id=\"T_42a61_row2_col6\" class=\"data row2 col6\" >0.000000</td>\n",
+       "      <td id=\"T_42a61_row2_col7\" class=\"data row2 col7\" >inf</td>\n",
+       "      <td id=\"T_42a61_row2_col8\" class=\"data row2 col8\" >Jensen-Shannon Divergence (JS)</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th id=\"T_42a61_level0_row3\" class=\"row_heading level0 row3\" >KL</th>\n",
+       "      <td id=\"T_42a61_row3_col0\" class=\"data row3 col0\" >0.002322</td>\n",
+       "      <td id=\"T_42a61_row3_col1\" class=\"data row3 col1\" >0.028928</td>\n",
+       "      <td id=\"T_42a61_row3_col2\" class=\"data row3 col2\" >0.001691</td>\n",
+       "      <td id=\"T_42a61_row3_col3\" class=\"data row3 col3\" >0.024665</td>\n",
+       "      <td id=\"T_42a61_row3_col4\" class=\"data row3 col4\" >pre_training_bias_metrics</td>\n",
+       "      <td id=\"T_42a61_row3_col5\" class=\"data row3 col5\" >0</td>\n",
+       "      <td id=\"T_42a61_row3_col6\" class=\"data row3 col6\" >0.000000</td>\n",
+       "      <td id=\"T_42a61_row3_col7\" class=\"data row3 col7\" >inf</td>\n",
+       "      <td id=\"T_42a61_row3_col8\" class=\"data row3 col8\" >Kullback-Liebler Divergence (KL)</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th id=\"T_42a61_level0_row4\" class=\"row_heading level0 row4\" >KS</th>\n",
+       "      <td id=\"T_42a61_row4_col0\" class=\"data row4 col0\" >0.029181</td>\n",
+       "      <td id=\"T_42a61_row4_col1\" class=\"data row4 col1\" >0.098880</td>\n",
+       "      <td id=\"T_42a61_row4_col2\" class=\"data row4 col2\" >0.025691</td>\n",
+       "      <td id=\"T_42a61_row4_col3\" class=\"data row4 col3\" >0.101349</td>\n",
+       "      <td id=\"T_42a61_row4_col4\" class=\"data row4 col4\" >pre_training_bias_metrics</td>\n",
+       "      <td id=\"T_42a61_row4_col5\" class=\"data row4 col5\" >0</td>\n",
+       "      <td id=\"T_42a61_row4_col6\" class=\"data row4 col6\" >0.000000</td>\n",
+       "      <td id=\"T_42a61_row4_col7\" class=\"data row4 col7\" >1.000000</td>\n",
+       "      <td id=\"T_42a61_row4_col8\" class=\"data row4 col8\" >Kolmogorov-Smirnov Distance (KS)</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th id=\"T_42a61_level0_row5\" class=\"row_heading level0 row5\" >LP</th>\n",
+       "      <td id=\"T_42a61_row5_col0\" class=\"data row5 col0\" >0.041269</td>\n",
+       "      <td id=\"T_42a61_row5_col1\" class=\"data row5 col1\" >0.139838</td>\n",
+       "      <td id=\"T_42a61_row5_col2\" class=\"data row5 col2\" >0.036333</td>\n",
+       "      <td id=\"T_42a61_row5_col3\" class=\"data row5 col3\" >0.143329</td>\n",
+       "      <td id=\"T_42a61_row5_col4\" class=\"data row5 col4\" >pre_training_bias_metrics</td>\n",
+       "      <td id=\"T_42a61_row5_col5\" class=\"data row5 col5\" >0</td>\n",
+       "      <td id=\"T_42a61_row5_col6\" class=\"data row5 col6\" >0.000000</td>\n",
+       "      <td id=\"T_42a61_row5_col7\" class=\"data row5 col7\" >inf</td>\n",
+       "      <td id=\"T_42a61_row5_col8\" class=\"data row5 col8\" >L-p Norm (LP)</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th id=\"T_42a61_level0_row6\" class=\"row_heading level0 row6\" >TVD</th>\n",
+       "      <td id=\"T_42a61_row6_col0\" class=\"data row6 col0\" >0.029181</td>\n",
+       "      <td id=\"T_42a61_row6_col1\" class=\"data row6 col1\" >0.098880</td>\n",
+       "      <td id=\"T_42a61_row6_col2\" class=\"data row6 col2\" >0.025691</td>\n",
+       "      <td id=\"T_42a61_row6_col3\" class=\"data row6 col3\" >0.101349</td>\n",
+       "      <td id=\"T_42a61_row6_col4\" class=\"data row6 col4\" >pre_training_bias_metrics</td>\n",
+       "      <td id=\"T_42a61_row6_col5\" class=\"data row6 col5\" >0</td>\n",
+       "      <td id=\"T_42a61_row6_col6\" class=\"data row6 col6\" >0.000000</td>\n",
+       "      <td id=\"T_42a61_row6_col7\" class=\"data row6 col7\" >inf</td>\n",
+       "      <td id=\"T_42a61_row6_col8\" class=\"data row6 col8\" >Total Variation Distance (TVD)</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n"
+      ],
+      "text/plain": [
+       "<pandas.io.formats.style.Styler at 0x7f30363efb10>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Gender aspects post-training bias metrics:\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<style type=\"text/css\">\n",
+       "#T_aa147_row0_col0, #T_aa147_row0_col1, #T_aa147_row0_col2, #T_aa147_row1_col0, #T_aa147_row1_col1, #T_aa147_row1_col2, #T_aa147_row2_col0, #T_aa147_row2_col1, #T_aa147_row2_col2, #T_aa147_row4_col0, #T_aa147_row4_col2, #T_aa147_row4_col3, #T_aa147_row5_col0, #T_aa147_row5_col2, #T_aa147_row5_col3, #T_aa147_row6_col0, #T_aa147_row6_col1, #T_aa147_row6_col2, #T_aa147_row6_col3, #T_aa147_row7_col1, #T_aa147_row8_col0, #T_aa147_row8_col1, #T_aa147_row8_col2, #T_aa147_row8_col3, #T_aa147_row9_col0, #T_aa147_row9_col1, #T_aa147_row9_col2, #T_aa147_row9_col3 {\n",
+       "  opacity: 20%;\n",
+       "}\n",
+       "</style>\n",
+       "<table id=\"T_aa147_\">\n",
+       "  <thead>\n",
+       "    <tr>\n",
+       "      <th class=\"blank level0\" >&nbsp;</th>\n",
+       "      <th class=\"col_heading level0 col0\" >Gender_Female</th>\n",
+       "      <th class=\"col_heading level0 col1\" >Gender_Joint</th>\n",
+       "      <th class=\"col_heading level0 col2\" >Gender_Male</th>\n",
+       "      <th class=\"col_heading level0 col3\" >Gender_Sex Not Available</th>\n",
+       "      <th class=\"col_heading level0 col4\" >metric_kind</th>\n",
+       "      <th class=\"col_heading level0 col5\" >good</th>\n",
+       "      <th class=\"col_heading level0 col6\" >min</th>\n",
+       "      <th class=\"col_heading level0 col7\" >max</th>\n",
+       "      <th class=\"col_heading level0 col8\" >description</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th id=\"T_aa147_level0_row0\" class=\"row_heading level0 row0\" >AD</th>\n",
+       "      <td id=\"T_aa147_row0_col0\" class=\"data row0 col0\" >-0.036073</td>\n",
+       "      <td id=\"T_aa147_row0_col1\" class=\"data row0 col1\" >-0.053750</td>\n",
+       "      <td id=\"T_aa147_row0_col2\" class=\"data row0 col2\" >-0.022003</td>\n",
+       "      <td id=\"T_aa147_row0_col3\" class=\"data row0 col3\" >0.109918</td>\n",
+       "      <td id=\"T_aa147_row0_col4\" class=\"data row0 col4\" >post_training_bias_metrics</td>\n",
+       "      <td id=\"T_aa147_row0_col5\" class=\"data row0 col5\" >0</td>\n",
+       "      <td id=\"T_aa147_row0_col6\" class=\"data row0 col6\" >-1.000000</td>\n",
+       "      <td id=\"T_aa147_row0_col7\" class=\"data row0 col7\" >1.000000</td>\n",
+       "      <td id=\"T_aa147_row0_col8\" class=\"data row0 col8\" >Accuracy Difference (AD)</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th id=\"T_aa147_level0_row1\" class=\"row_heading level0 row1\" >DAR</th>\n",
+       "      <td id=\"T_aa147_row1_col0\" class=\"data row1 col0\" >-0.054765</td>\n",
+       "      <td id=\"T_aa147_row1_col1\" class=\"data row1 col1\" >-0.046989</td>\n",
+       "      <td id=\"T_aa147_row1_col2\" class=\"data row1 col2\" >-0.028326</td>\n",
+       "      <td id=\"T_aa147_row1_col3\" class=\"data row1 col3\" >0.123350</td>\n",
+       "      <td id=\"T_aa147_row1_col4\" class=\"data row1 col4\" >post_training_bias_metrics</td>\n",
+       "      <td id=\"T_aa147_row1_col5\" class=\"data row1 col5\" >0</td>\n",
+       "      <td id=\"T_aa147_row1_col6\" class=\"data row1 col6\" >-1.000000</td>\n",
+       "      <td id=\"T_aa147_row1_col7\" class=\"data row1 col7\" >1.000000</td>\n",
+       "      <td id=\"T_aa147_row1_col8\" class=\"data row1 col8\" >Difference in Acceptance Rates (DAR)</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th id=\"T_aa147_level0_row2\" class=\"row_heading level0 row2\" >DCA</th>\n",
+       "      <td id=\"T_aa147_row2_col0\" class=\"data row2 col0\" >-0.079445</td>\n",
+       "      <td id=\"T_aa147_row2_col1\" class=\"data row2 col1\" >-0.009282</td>\n",
+       "      <td id=\"T_aa147_row2_col2\" class=\"data row2 col2\" >-0.053351</td>\n",
+       "      <td id=\"T_aa147_row2_col3\" class=\"data row2 col3\" >0.125789</td>\n",
+       "      <td id=\"T_aa147_row2_col4\" class=\"data row2 col4\" >post_training_bias_metrics</td>\n",
+       "      <td id=\"T_aa147_row2_col5\" class=\"data row2 col5\" >0</td>\n",
+       "      <td id=\"T_aa147_row2_col6\" class=\"data row2 col6\" >-inf</td>\n",
+       "      <td id=\"T_aa147_row2_col7\" class=\"data row2 col7\" >inf</td>\n",
+       "      <td id=\"T_aa147_row2_col8\" class=\"data row2 col8\" >Difference in Conditional Acceptance (DCA)</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th id=\"T_aa147_level0_row3\" class=\"row_heading level0 row3\" >DCR</th>\n",
+       "      <td id=\"T_aa147_row3_col0\" class=\"data row3 col0\" >-0.868890</td>\n",
+       "      <td id=\"T_aa147_row3_col1\" class=\"data row3 col1\" >2.435185</td>\n",
+       "      <td id=\"T_aa147_row3_col2\" class=\"data row3 col2\" >-1.064791</td>\n",
+       "      <td id=\"T_aa147_row3_col3\" class=\"data row3 col3\" >1.119565</td>\n",
+       "      <td id=\"T_aa147_row3_col4\" class=\"data row3 col4\" >post_training_bias_metrics</td>\n",
+       "      <td id=\"T_aa147_row3_col5\" class=\"data row3 col5\" >0</td>\n",
+       "      <td id=\"T_aa147_row3_col6\" class=\"data row3 col6\" >-inf</td>\n",
+       "      <td id=\"T_aa147_row3_col7\" class=\"data row3 col7\" >inf</td>\n",
+       "      <td id=\"T_aa147_row3_col8\" class=\"data row3 col8\" >Difference in Conditional Rejection (DCR)</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th id=\"T_aa147_level0_row4\" class=\"row_heading level0 row4\" >DI</th>\n",
+       "      <td id=\"T_aa147_row4_col0\" class=\"data row4 col0\" >0.948794</td>\n",
+       "      <td id=\"T_aa147_row4_col1\" class=\"data row4 col1\" >1.125680</td>\n",
+       "      <td id=\"T_aa147_row4_col2\" class=\"data row4 col2\" >0.907514</td>\n",
+       "      <td id=\"T_aa147_row4_col3\" class=\"data row4 col3\" >1.013672</td>\n",
+       "      <td id=\"T_aa147_row4_col4\" class=\"data row4 col4\" >post_training_bias_metrics</td>\n",
+       "      <td id=\"T_aa147_row4_col5\" class=\"data row4 col5\" >1</td>\n",
+       "      <td id=\"T_aa147_row4_col6\" class=\"data row4 col6\" >0.000000</td>\n",
+       "      <td id=\"T_aa147_row4_col7\" class=\"data row4 col7\" >inf</td>\n",
+       "      <td id=\"T_aa147_row4_col8\" class=\"data row4 col8\" >Disparate Impact (DI)</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th id=\"T_aa147_level0_row5\" class=\"row_heading level0 row5\" >DPPL</th>\n",
+       "      <td id=\"T_aa147_row5_col0\" class=\"data row5 col0\" >0.045510</td>\n",
+       "      <td id=\"T_aa147_row5_col1\" class=\"data row5 col1\" >-0.106880</td>\n",
+       "      <td id=\"T_aa147_row5_col2\" class=\"data row5 col2\" >0.083611</td>\n",
+       "      <td id=\"T_aa147_row5_col3\" class=\"data row5 col3\" >-0.011995</td>\n",
+       "      <td id=\"T_aa147_row5_col4\" class=\"data row5 col4\" >post_training_bias_metrics</td>\n",
+       "      <td id=\"T_aa147_row5_col5\" class=\"data row5 col5\" >0</td>\n",
+       "      <td id=\"T_aa147_row5_col6\" class=\"data row5 col6\" >-1.000000</td>\n",
+       "      <td id=\"T_aa147_row5_col7\" class=\"data row5 col7\" >1.000000</td>\n",
+       "      <td id=\"T_aa147_row5_col8\" class=\"data row5 col8\" >Difference in Positive Proportions in Predictected Labels (DPPL)</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th id=\"T_aa147_level0_row6\" class=\"row_heading level0 row6\" >DRR</th>\n",
+       "      <td id=\"T_aa147_row6_col0\" class=\"data row6 col0\" >-0.046609</td>\n",
+       "      <td id=\"T_aa147_row6_col1\" class=\"data row6 col1\" >0.027778</td>\n",
+       "      <td id=\"T_aa147_row6_col2\" class=\"data row6 col2\" >0.034953</td>\n",
+       "      <td id=\"T_aa147_row6_col3\" class=\"data row6 col3\" >-0.013975</td>\n",
+       "      <td id=\"T_aa147_row6_col4\" class=\"data row6 col4\" >post_training_bias_metrics</td>\n",
+       "      <td id=\"T_aa147_row6_col5\" class=\"data row6 col5\" >0</td>\n",
+       "      <td id=\"T_aa147_row6_col6\" class=\"data row6 col6\" >-1.000000</td>\n",
+       "      <td id=\"T_aa147_row6_col7\" class=\"data row6 col7\" >1.000000</td>\n",
+       "      <td id=\"T_aa147_row6_col8\" class=\"data row6 col8\" >Difference in Rejection Rates (DRR)</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th id=\"T_aa147_level0_row7\" class=\"row_heading level0 row7\" >FT</th>\n",
+       "      <td id=\"T_aa147_row7_col0\" class=\"data row7 col0\" >-0.145946</td>\n",
+       "      <td id=\"T_aa147_row7_col1\" class=\"data row7 col1\" >-0.017794</td>\n",
+       "      <td id=\"T_aa147_row7_col2\" class=\"data row7 col2\" >-0.158451</td>\n",
+       "      <td id=\"T_aa147_row7_col3\" class=\"data row7 col3\" >-0.102767</td>\n",
+       "      <td id=\"T_aa147_row7_col4\" class=\"data row7 col4\" >post_training_bias_metrics</td>\n",
+       "      <td id=\"T_aa147_row7_col5\" class=\"data row7 col5\" >0</td>\n",
+       "      <td id=\"T_aa147_row7_col6\" class=\"data row7 col6\" >-1.000000</td>\n",
+       "      <td id=\"T_aa147_row7_col7\" class=\"data row7 col7\" >1.000000</td>\n",
+       "      <td id=\"T_aa147_row7_col8\" class=\"data row7 col8\" >Flip Test (FT)</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th id=\"T_aa147_level0_row8\" class=\"row_heading level0 row8\" >GE</th>\n",
+       "      <td id=\"T_aa147_row8_col0\" class=\"data row8 col0\" >0.071114</td>\n",
+       "      <td id=\"T_aa147_row8_col1\" class=\"data row8 col1\" >0.071114</td>\n",
+       "      <td id=\"T_aa147_row8_col2\" class=\"data row8 col2\" >0.071114</td>\n",
+       "      <td id=\"T_aa147_row8_col3\" class=\"data row8 col3\" >0.071114</td>\n",
+       "      <td id=\"T_aa147_row8_col4\" class=\"data row8 col4\" >post_training_bias_metrics</td>\n",
+       "      <td id=\"T_aa147_row8_col5\" class=\"data row8 col5\" >0</td>\n",
+       "      <td id=\"T_aa147_row8_col6\" class=\"data row8 col6\" >0.000000</td>\n",
+       "      <td id=\"T_aa147_row8_col7\" class=\"data row8 col7\" >0.500000</td>\n",
+       "      <td id=\"T_aa147_row8_col8\" class=\"data row8 col8\" >Generalized Entropy (GE)</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th id=\"T_aa147_level0_row9\" class=\"row_heading level0 row9\" >RD</th>\n",
+       "      <td id=\"T_aa147_row9_col0\" class=\"data row9 col0\" >0.023645</td>\n",
+       "      <td id=\"T_aa147_row9_col1\" class=\"data row9 col1\" >-0.044927</td>\n",
+       "      <td id=\"T_aa147_row9_col2\" class=\"data row9 col2\" >0.026070</td>\n",
+       "      <td id=\"T_aa147_row9_col3\" class=\"data row9 col3\" >0.004010</td>\n",
+       "      <td id=\"T_aa147_row9_col4\" class=\"data row9 col4\" >post_training_bias_metrics</td>\n",
+       "      <td id=\"T_aa147_row9_col5\" class=\"data row9 col5\" >0</td>\n",
+       "      <td id=\"T_aa147_row9_col6\" class=\"data row9 col6\" >-1.000000</td>\n",
+       "      <td id=\"T_aa147_row9_col7\" class=\"data row9 col7\" >1.000000</td>\n",
+       "      <td id=\"T_aa147_row9_col8\" class=\"data row9 col8\" >Recall Difference (RD)</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th id=\"T_aa147_level0_row10\" class=\"row_heading level0 row10\" >SD</th>\n",
+       "      <td id=\"T_aa147_row10_col0\" class=\"data row10 col0\" >0.152032</td>\n",
+       "      <td id=\"T_aa147_row10_col1\" class=\"data row10 col1\" >-0.207411</td>\n",
+       "      <td id=\"T_aa147_row10_col2\" class=\"data row10 col2\" >0.213436</td>\n",
+       "      <td id=\"T_aa147_row10_col3\" class=\"data row10 col3\" >-0.149188</td>\n",
+       "      <td id=\"T_aa147_row10_col4\" class=\"data row10 col4\" >post_training_bias_metrics</td>\n",
+       "      <td id=\"T_aa147_row10_col5\" class=\"data row10 col5\" >0</td>\n",
+       "      <td id=\"T_aa147_row10_col6\" class=\"data row10 col6\" >-1.000000</td>\n",
+       "      <td id=\"T_aa147_row10_col7\" class=\"data row10 col7\" >1.000000</td>\n",
+       "      <td id=\"T_aa147_row10_col8\" class=\"data row10 col8\" >Specificity Difference (SD)</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th id=\"T_aa147_level0_row11\" class=\"row_heading level0 row11\" >TE</th>\n",
+       "      <td id=\"T_aa147_row11_col0\" class=\"data row11 col0\" >0.228039</td>\n",
+       "      <td id=\"T_aa147_row11_col1\" class=\"data row11 col1\" >-0.166453</td>\n",
+       "      <td id=\"T_aa147_row11_col2\" class=\"data row11 col2\" >0.171154</td>\n",
+       "      <td id=\"T_aa147_row11_col3\" class=\"data row11 col3\" >-0.110849</td>\n",
+       "      <td id=\"T_aa147_row11_col4\" class=\"data row11 col4\" >post_training_bias_metrics</td>\n",
+       "      <td id=\"T_aa147_row11_col5\" class=\"data row11 col5\" >0</td>\n",
+       "      <td id=\"T_aa147_row11_col6\" class=\"data row11 col6\" >-inf</td>\n",
+       "      <td id=\"T_aa147_row11_col7\" class=\"data row11 col7\" >inf</td>\n",
+       "      <td id=\"T_aa147_row11_col8\" class=\"data row11 col8\" >Treatment Equality (TE)</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n"
+      ],
+      "text/plain": [
+       "<pandas.io.formats.style.Styler at 0x7f303acc0910>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "analisys_df = None\n",
+    "\n",
+    "execution_summary = schedule_desc.get(\"LastMonitoringExecutionSummary\")\n",
+    "if execution_summary and execution_summary[\"MonitoringExecutionStatus\"] in [\n",
+    "    \"Completed\",\n",
+    "    \"CompletedWithViolations\",\n",
+    "]:\n",
+    "    last_model_bias_monitor_execution = model_bias_monitor.list_executions()[-1]\n",
+    "    last_model_bias_monitor_execution_report_uri = (\n",
+    "        last_model_bias_monitor_execution.output.destination\n",
+    "    )\n",
+    "    print(f\"Report URI: {last_model_bias_monitor_execution_report_uri}\")\n",
+    "    last_model_bias_monitor_execution_report_files = sorted(\n",
+    "        S3Downloader.list(last_model_bias_monitor_execution_report_uri)\n",
+    "    )\n",
+    "    print(\"Found Report Files:\")\n",
+    "    print(\"\\n \".join(last_model_bias_monitor_execution_report_files))\n",
+    "    if execution_summary[\"MonitoringExecutionStatus\"] == \"CompletedWithViolations\":\n",
+    "        file = [name for name in last_model_bias_monitor_execution_report_files if 'analysis.json' in name]\n",
+    "        assert len(file) == 1, 'Error: analysis file was not generated'\n",
+    "        file = file[0]\n",
+    "        analysis = json.loads(S3Downloader.read_file(file))\n",
+    "        analysis_df = extract_metrix_from_analysis(analysis)\n",
+    "        print('Gender aspects pre-training bias metrics:')\n",
+    "        display(analysis_df[gender_aspects + descr_colnames].query(\"metric_kind == 'pre_training_bias_metrics'\")\n",
+    "                   .style.applymap(lambda v: 'opacity: 20%;' if (v < 0.1) and (v > -0.1) else None, subset=gender_aspects))\n",
+    "        print('Gender aspects post-training bias metrics:')\n",
+    "        idx = pd.IndexSlice\n",
+    "        display(analysis_df[gender_aspects + descr_colnames].query(\"metric_kind == 'post_training_bias_metrics'\")\n",
+    "                   .style.applymap(lambda v: 'opacity: 20%;' if (v < 0.1) and (v > -0.1) else None, subset=gender_aspects)\n",
+    "                         .applymap(lambda v: 'opacity: 20%;' if (v < 1.1) and (v > -0.9) else None, subset=idx[idx['DI'], idx[gender_aspects]]))\n",
+    "        \n",
+    "else:\n",
+    "    last_model_bias_monitor_execution = None\n",
+    "    print(\n",
+    "        \"====STOP==== \\n No completed executions to inspect further. Please wait till an execution completes or investigate previously reported failures.\"\n",
+    "    )"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "9e78d581-bc25-4767-9af3-771f0e765b5d",
+   "metadata": {},
+   "source": [
+    "We can see some abnormal values (without opacity). Especially in **Difference in Conditional Rejection (DCR)** metrics. That means that model tends to give negative prediction differently across genders, i.e., we can see bias and need to handle that. At the same time, pre-training metrics, overall, looks normal."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 32,
+   "id": "8d1bd888-f68e-4c12-af36-9d1686a3a366",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Let's print the first few of violations:\n",
+      "{\n",
+      "    \"version\": \"1.0\",\n",
+      "    \"violations\": [\n",
+      "        {\n",
+      "            \"facet\": \"Gender_Female\",\n",
+      "            \"facet_value\": \"1\",\n",
+      "            \"metric_name\": \"DPL\",\n",
+      "            \"constraint_check_type\": \"bias_drift_check\",\n",
+      "            \"description\": \"Metric value -0.029181259499107992 doesn't meet the baseline constraint requirement 0.004437718747468233\"\n",
+      "        },\n",
+      "        {\n",
+      "            \"facet\": \"Gender_Female\",\n",
+      "            \"facet_value\": \"1\",\n",
+      "            \"metric_name\": \"JS\",\n",
+      "            \"constraint_check_type\": \"bias_drift_check\",\n",
+      "            \"description\": \"Metric value 0.0005725270156258485 doesn't meet the baseline constraint requirement 1.3179244474955353e-05\"\n",
+      "        },\n",
+      "        {\n",
+      "            \"facet\": \"Gender_Female\",\n",
+      "            \"facet_value\": \"1\",\n",
+      "            \"metric_name\": \"KL\",\n",
+      "            \"constraint_check_type\": \"bias_drift_check\",\n",
+      "            \"description\": \"Metric value 0.0023215021891724794 doesn't meet the baseline constraint requirement 5.261255142023462e-05\"\n",
+      "        },\n",
+      "        {\n",
+      "            \"facet\": \"Gender_Female\",\n",
+      "            \"facet_value\": \"1\",\n",
+      "            \"metric_name\": \"KS\",\n",
+      "            \"constraint_check_type\": \"bias_drift_check\",\n",
+      "            \"description\": \"Metric value 0.029181259499107992 doesn't meet the baseline constraint requirement 0.004437718747468261\"\n",
+      "        },\n",
+      "        {\n",
+      "            \"facet\": \"Gender_Female\",\n",
+      "            \"facet_value\": \"1\",\n",
+      "            \"metric_name\": \"LP\",\n",
+      "            \"constraint_check_type\": \"bias_drift_check\",\n",
+      "            \"description\": \"Metric value 0.041268532950767156 doesn't meet the baseline constraint requirement 0.006275882038666939\"\n",
+      "        },\n",
+      "        {\n",
+      "            \"facet\": \"Gender_Female\",\n",
+      "            \"facet_value\": \"1\",\n",
+      "            \"metric_name\": \"TVD\",\n",
+      "            \"constraint_check_type\": \"bias_drift_check\",\n",
+      "            \"description\": \"Metric value 0.029181259499107937 doesn't meet the baseline constraint requirement 0.004437718747468247\"\n",
+      "        },\n",
+      "        {\n",
+      "            \"facet\": \"Gender_Female\",\n",
+      "            \"facet_value\": \"1\",\n",
+      "            \"metric_name\": \"AD\",\n",
+      "            \"constraint_check_type\": \"bias_drift_check\",\n",
+      "            \"description\": \"Metric value -0.036073481794753226 doesn't meet the baseline constraint requirement -0.012274862880233273\"\n",
+      "        },\n",
+      "        {\n",
+      "            \"facet\": \"Gender_Female\",\n",
+      "            \"facet_value\": \"1\",\n",
+      "            \"metric_name\": \"DAR\",\n",
+      "            \"constraint_check_type\": \"bias_drift_check\",\n",
+      "            \"description\": \"Metric value -0.05476492787359355 doesn't meet the baseline constraint requirement -0.019687953899706012\"\n",
+      "        },\n",
+      "        {\n",
+      "            \"facet\": \"Gender_Female\",\n",
+      "            \"facet_value\": \"1\",\n",
+      "            \"metric_name\": \"DCA\",\n",
+      "            \"constraint_check_type\": \"bias_drift_check\",\n",
+      "            \"description\": \"Metric value -0.07944485592353545 doesn't meet the baseline constraint requirement -0.035234696157777856\"\n",
+      "        },\n",
+      "        {\n",
+      "            \"facet\": \"Gender_Female\",\n",
+      "            \"facet_value\": \"1\",\n",
+      "            \"metric_name\": \"DCR\",\n",
+      "            \"constraint_check_type\": \"bias_drift_check\",\n",
+      "            \"description\": \"Metric value -0.8688897309586965 doesn't meet the baseline constraint requirement -0.5736707206241256\"\n",
+      "        },\n",
+      "        {\n",
+      "            \"facet\": \"Gender_Female\",\n",
+      "            \"facet_value\": \"1\",\n",
+      "            \"metric_name\": \"DPPL\",\n",
+      "            \"constraint_check_type\": \"bias_drift_check\",\n",
+      "            \"description\": \"Metric value 0.045509812991475496 doesn't meet the baseline constraint requirement 0.040580765079404446\"\n",
+      "        },\n",
+      "        {\n",
+      "            \"facet\": \"Gender_Female\",\n",
+      "            \"facet_value\": \"1\",\n",
+      "            \"metric_name\": \"DRR\",\n",
+      "            \"constraint_check_type\": \"bias_drift_check\",\n",
+      "            \"description\": \"Metric value -0.04660856384994316 doesn't meet the baseline constraint requirement -0.010994514578744008\"\n",
+      "        },\n",
+      "        {\n",
+      "            \"facet\": \"Gender_Female\",\n",
+      "            \"facet_value\": \"1\",\n",
+      "            \"metric_name\": \"FT\",\n",
+      "            \"constraint_check_type\": \"bias_drift_check\",\n",
+      "            \"description\": \"Metric value -0.14594594594594595 doesn't meet the baseline constraint requirement -0.11717435993272285\"\n",
+      "        },\n",
+      "        {\n",
+      "            \"facet\": \"Gender_Female\",\n",
+      "            \"facet_value\": \"1\",\n",
+      "            \"metric_name\": \"GE\",\n",
+      "            \"constraint_check_type\": \"bias_drift_check\",\n",
+      "            \"description\": \"Metric value 0.07111418898122515 doesn't meet the baseline constraint requirement 0.06776540220552933\"\n",
+      "        },\n",
+      "        {\n",
+      "            \"facet\": \"Gender_Female\",\n",
+      "            \"facet_value\": \"1\",\n",
+      "            \"metric_name\": \"RD\",\n",
+      "            \"constraint_check_type\": \"bias_drift_check\",\n",
+      "            \"description\": \"Metric value 0.023645182352392546 doesn't meet the baseline constraint requirement 0.01614804633418354\"\n",
+      "        },\n",
+      "        {\n",
+      "            \"facet\": \"Gender_Female\",\n",
+      "            \"facet_value\": \"1\",\n",
+      "            \"metric_name\": \"SD\",\n",
+      "            \"constraint_check_type\": \"bias_drift_check\",\n",
+      "            \"description\": \"Metric value 0.15203216692023475 doesn't meet the baseline constraint requirement 0.10832164460419041\"\n",
+      "        },\n",
+      "        {\n",
+      "            \"facet\": \"Gender_Female\",\n",
+      "            \"facet_value\": \"1\",\n",
+      "            \"metric_name\": \"TE\",\n",
+      "            \"constraint_check_type\": \"bias_drift_check\",\n",
+      "            \"description\": \"Metric value 0.228039041703638 doesn't meet the baseline constraint requirement 0.11461989439613904\"\n",
+      "        }\n",
+      "    ]\n",
+      "}\n"
+     ]
+    }
+   ],
+   "source": [
+    "if last_model_bias_monitor_execution:\n",
+    "    model_bias_violations = last_model_bias_monitor_execution.constraint_violations()\n",
+    "    if model_bias_violations:\n",
+    "        print(\"Let's print the first few of violations:\")\n",
+    "        violations_part = copy.deepcopy(model_bias_violations.body_dict)\n",
+    "        violations_part['violations'] = [violation for violation in violations_part['violations'] if violation['facet'] == 'Gender_Female']\n",
+    "        print(json.dumps(violations_part, indent=4))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "7074f498-bb0a-408f-8c8d-469deb0dbd58",
+   "metadata": {},
+   "source": [
+    "After some time, we'll see the monitoring job results in the Monitoring Job History tab\n",
+    "\n",
+    "<img alt=\"monitoring_result\" src=\"images/monitoring_result.jpg\" style=\"width: 800px; margin: auto\"/>\n",
+    "\n",
+    "Clicking on them we can see the additional details: which metrics constrains were violated compared to the baseline \n",
+    "\n",
+    "<img alt=\"violations\" src=\"images/violations.jpg\" style=\"width: 800px; margin: auto\"/>"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "88b9969c-cc15-49af-8e57-c5af9e950775",
+   "metadata": {
+    "tags": []
+   },
+   "source": [
+    "## 6. Cleaning up "
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "cd0bf892-8bf7-431a-a9f1-843f4a9c776c",
+   "metadata": {},
+   "source": [
+    "Now we know how to set up model bias monitoring with AWS SageMaker tools. And we can remove all created resourses, in order not to recieve unexpected bills at the end of the month."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 33,
+   "id": "ebf8c362-688e-422c-8c37-ae07c0313231",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "\n",
+      "Stopping Monitoring Schedule with name: monitoring-schedule-2023-02-28-15-53-44-998\n"
+     ]
+    }
+   ],
+   "source": [
+    "# At first we must wait untill monitoring job finishes, \n",
+    "# as we can't delete monitoring schedule before\n",
+    "while model_bias_monitor.describe_schedule()['LastMonitoringExecutionSummary']['MonitoringExecutionStatus'] in ['Pending', 'InProgress']:\n",
+    "    time.sleep(60)\n",
+    "\n",
+    "model_bias_monitor.stop_monitoring_schedule()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 34,
+   "id": "4e373fc3-3fc1-4632-bdbe-00e6dea10d4c",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "\n",
+      "Deleting Monitoring Schedule with name: monitoring-schedule-2023-02-28-15-53-44-998\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "INFO:sagemaker.model_monitor.clarify_model_monitoring:Deleting Model Bias Job Definition with name: model-bias-job-definition-2023-02-28-15-53-44-998\n"
+     ]
+    }
+   ],
+   "source": [
+    "# Turning off traffic generation\n",
+    "ground_truth_thread.terminate()\n",
+    "# Deleting monitoring job\n",
+    "model_bias_monitor.delete_monitoring_schedule()\n",
+    "# Removing endpoint and model\n",
+    "sagemaker_client.delete_endpoint(EndpointName=xgb_endpoint_name)\n",
+    "sagemaker_client.delete_endpoint_config(EndpointConfigName=xgb_endpoint_name)\n",
+    "sagemaker_client.delete_model(ModelName=xgb_model_name);"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "8b9a2e0b-7454-4109-bdec-1dd00fa0a290",
+   "metadata": {},
+   "source": [
+    "Don't forget to stop the current runtime (if you work from the SageMaker Studio):\n",
+    "\n",
+    "<img src=\"images/runtime.jpg\" alt=\"runtime\" style=\"height: 300px; margin: left\" />"
+   ]
+  }
+ ],
+ "metadata": {
+  "instance_type": "ml.t3.medium",
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "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": "3.9.15"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}