{
 "cells": [
  {
   "attachments": {
    "image.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![image.png](attachment:image.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predict fraudulent providers in Medicare using Neptune ML\n",
    "RUN in Sagemaker Notebooks (not Studio): This notebook uses the medicare datasets provided by CMS. In this Notebook  we'll walk through how Neptune ML can predict identification of fraudulent providers. To demonstrate this we'll predict the \"Fraud\" property of providers. We'll walk through each step of loading and exporting the data, configuring and training the model, and finally we'll show how to use that model to infer Fraud using Gremlin traversals."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## If Neptune has not been setup, please run the cloudformation script from the link below\n",
    "https://docs.aws.amazon.com/neptune/latest/userguide/machine-learning-quick-start.html"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Checking that we are ready to run Neptune ML \n",
    "Run the code below to check that this cluster is configured to run Neptune ML."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install sagemaker\n",
    "!pip install neptune_ml_utils"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "import neptune_ml_utils as neptune_ml\n",
    "import sagemaker\n",
    "\n",
    "neptune_ml.check_ml_enabled()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Get the endpoint of the Neptune Cluster"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "cluster_endpoint= neptune_ml.get_host() + \":8182\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Please check that service limits of the following are available. If not, pls raise a ticket with AWS Support\n",
    "\n",
    "ml.r5.xlarge for endpoint usage - 2\n",
    "ml.p3.2xlarge for training job usage - 2\n",
    "ml.r5.large for processing job usage - 2\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Delete existing graph"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%gremlin\n",
    "g.V().drop()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ensure you saved the \"vertex_input.csv\" and \"edge_input.csv\" files in S3  \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load the nodes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create an IAM role that provides full access to Amazon Neptune and Amazon S3 as indicated in the instructions below\n",
    "https://docs.aws.amazon.com/neptune/latest/userguide/bulk-load-tutorial-IAM.html#bulk-load-tutorial-IAM-CreateRole"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "s3_nodes_uri = \"REPLACE WITH THE S3URI WHERE VERTEX_INPUT.CSV HAS BEEN UPLOADED\"\n",
    "load_role = \"REPLACE WITH ARN OF IAM ROLE FROM ABOVE\"\n",
    "region = sagemaker.Session().boto_region_name"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import requests\n",
    "url = 'https://{0}/loader'.format(cluster_endpoint)\n",
    "headers = {'Content-Type': 'application/json'}\n",
    "data = {\n",
    "    'source': s3_nodes_uri,\n",
    "    'format': 'csv',\n",
    "    'iamRoleArn': load_role,\n",
    "    'region': region,\n",
    "    'failOnError': 'true',\n",
    "    'parallelism': 'MEDIUM',\n",
    "    'updateSingleCardinalityProperties': 'FALSE',\n",
    "    'queueRequest': 'TRUE'\n",
    "}\n",
    "\n",
    "response = requests.post(url, headers=headers, json=data)\n",
    "print(\"Status Code\", response.status_code)\n",
    "print(\"JSON Response \", response.json())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Check to see if the load is complete - you should have a status 200 for a succesful execution."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load the edges"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "s3_edges_uri = \"REPLACE WITH S3 URI WHERE EDGE_INPUT.CSV HAS BEEN UPLOADED\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "url = 'https://{0}/loader'.format(cluster_endpoint)\n",
    "headers = {'Content-Type': 'application/json'}\n",
    "data = {\n",
    "    'source': s3_edges_uri,\n",
    "    'format': 'csv',\n",
    "    'iamRoleArn': load_role,\n",
    "    'region': region,\n",
    "    'failOnError': 'true',\n",
    "    'parallelism': 'MEDIUM',\n",
    "    'updateSingleCardinalityProperties': 'FALSE',\n",
    "    'queueRequest': 'TRUE'\n",
    "}\n",
    "\n",
    "response = requests.post(url, headers=headers, json=data)\n",
    "print(\"Status Code\", response.status_code)\n",
    "print(\"JSON Response \", response.json())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Check to see if the load is complete - you should have a status 200 for a succesful execution."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Provide an S3 bucket to store outputs of the data export job needed to train the model in SageMaker"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "s3_bucket_uri=\"REPLACE WITH S3URI OF OUTPUT BUCKET\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Check to make sure the data is loaded"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%gremlin\n",
    "g.V().groupCount().by(label).unfold().order().by(keys)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If our nodes loaded correctly then the output is:\n",
    "\n",
    "* 1084 Drugs\n",
    "* 1836 HCPCS code\n",
    "* 5983 Providers \n",
    "\n",
    "To check that our edges loaded correctly we check the edge counts:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%gremlin\n",
    "g.E().groupCount().by(label).unfold().order().by(keys)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If our edges loaded correctly then the output is:\n",
    "\n",
    "* 57807 conducts_procedure\n",
    "* 62193 prescribes_drug\n",
    "\n",
    "\n",
    "## Prepare data for Training - This data is used for training the model in Sagemaker \n",
    "First, let's simulate new providers being added into our product knowledge graph by randomly removing the `fraud` property from 1% of all  fraudulent providers (approx. 55)  nodes.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%gremlin \n",
    "g.V().has('Provider', 'Fraud', 'YES').properties('Fraud').sample(55).drop()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's store these nodes to a variable \"test_fraud\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "%%gremlin --store-to test_fraud\n",
    "g.V().hasLabel('Provider').hasNot('Fraud').valueMap('NPI')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's also randomly removing the `fraud` property from a few  non-fraudulent providers (approx. 55)  nodes.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%gremlin \n",
    "g.V().has('Provider', 'Fraud', 'NO').properties('Fraud').sample(55).drop()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lets store all the simulated \"new\" nodes  to a variable \"test_all\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%gremlin --store-to test_all\n",
    "g.V().hasLabel('Provider').hasNot('Fraud').valueMap('NPI', 'Fraud')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now lets prepare a pandas dataframes test_df that contains information on which of the simulated \"new\" nodes are classified as fraud versus. This will be necessary to evaluate the performance of our model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "test_fraud = pd.DataFrame.from_dict(test_fraud,dtype='string')\n",
    "test_fraud['Fraud'] = 'YES'\n",
    "test_all = pd.DataFrame.from_dict(test_all,dtype='string')\n",
    "test_all['NPI'] = test_all['NPI'].str.replace(']','', regex=True)\n",
    "test_all['NPI'] = test_all['NPI'].str.replace('[','', regex=True)\n",
    "test_fraud['NPI'] = test_fraud['NPI'].str.replace(']','', regex=True)\n",
    "test_fraud['NPI'] = test_fraud['NPI'].str.replace('[','', regex=True)\n",
    "test_df = test_all.merge(test_fraud,on='NPI',how = 'left')\n",
    "test_df = test_df.fillna('NO')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_df['Fraud'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lets pick one of the providers"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " Double check if the providers in test_df no longer have `Fraud` values by picking a value randomly and using the query below "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "NPI = test_df['NPI'].head(1).to_string(index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "NPI"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Query the graph using the above NPI "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%gremlin\n",
    "\n",
    "g.V().has('Provider', 'NPI', ${NPI}).valueMap('NPI','Fraud')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# With these properties removed we're ready to build our Neptune ML Node Classification model to predict the Fraud indicator for these providers in our provider knowledge graph.  \n",
    "\n",
    "\n",
    "# Export the data and model configuration\n",
    "\n",
    "<div style=\"background-color:#eeeeee; padding:10px; text-align:left; border-radius:10px; margin-top:10px; margin-bottom:10px; \"><b>Note</b>: Before exporting data ensure that Neptune Export has been configured as described here: <a href=\"https://docs.aws.amazon.com/neptune/latest/userguide/export-service.html\">Neptune Export Service</a></div>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The export process is triggered by calling to the [Neptune Export service endpoint](https://docs.aws.amazon.com/neptune/latest/userguide/machine-learning-data-export-service.html).  \n",
    "The configuration options provided to the export service are broken into two main sections, selecting the target and configuring features. \n",
    "\n",
    "## Selecting the target\n",
    "\n",
    "In the first section, selecting the target, we specify what type of machine learning task will be run, and in the case of node classification, what the target node label and property we want to predict.  To run a node classification task we are only required to specify the node label and target property to infer. \n",
    "\n",
    "In this example below we specify the `Fraud` property on the `Provider` node as our target for prediction by including these values in the `targets` sub-parameter of the `additionalParams` object as shown below. \n",
    "\n",
    "```\n",
    "\"additionalParams\": {\n",
    "        \"neptune_ml\": {\n",
    "          \"targets\": [\n",
    "            {\n",
    "              \"node\": \"Provider\",\n",
    "              \"property\": \"Fraud\"\n",
    "            }\n",
    "          ],\n",
    "          ....\n",
    "```\n",
    "\n",
    "## Configuring features\n",
    "The second section of the configuration, configuring features, is where we specify details about the types of data stored in our graph and how the machine learning model should interpret that data.  \n",
    "When data is exported from Neptune all properties of all nodes are included.  Each property is treated as a separate feature for the ML model.  Neptune ML does its best to infer the correct type of feature for a property, in many cases, the accuracy of the model can be improved by specifying information about the property used for a feature.  By default Neptune ML puts features into one of two categories:\n",
    "\n",
    "* If the feature represents a numerical property (float, double, int) then it is treated as a `numerical` feature type. In this feature type data is represented as a continuous set of numbers.  For example, 'total_claim_count_mean'\n",
    "* All other property types are represented as `category` features.  In this feature type, each unique value of data is represented as a unique value in the set of classifications used by the model. For our data, we have already one-hot encoded the only category feature \"Female\". \n",
    "\n",
    "If all of the properties fit into these two feature types then no configuration changes are needed at the time of export. However, in many scenarios these defaults are not always the best choice. In these cases, additional configuration options should be specified to better define how the property should be represented as a feature. For example, to handle text in the field, we could use techniques such as Word2vec from natural language processing to create a vector of data that represents a string of text.. \n",
    "\n",
    "\n",
    "Running the cell below we set the export configuration and run the export process.  Neptune export is capable of automatically creating a clone of the cluster by setting `cloneCluster=True` which takes about 20 minutes to complete and will incur additional costs while the cloned cluster is running.  Exporting from the existing cluster takes about 5 minutes but requires that the `neptune_query_timeout` parameter in the [parameter group](https://docs.aws.amazon.com/neptune/latest/userguide/parameters.html) is set to a large enough value (>72000) to prevent timeout errors."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "export_params={ \n",
    "\"command\": \"export-pg\", \n",
    "\"params\": { \"endpoint\": neptune_ml.get_host(),\n",
    "            \"profile\": \"neptune_ml\",\n",
    "            \"useIamAuth\": neptune_ml.get_iam(),\n",
    "            \"cloneCluster\": False\n",
    "            }, \n",
    "\"outputS3Path\": f'{s3_bucket_uri}/neptune-export',\n",
    "\"additionalParams\": {\n",
    "        \"neptune_ml\": {\n",
    "          \"version\": \"v2.0\",\n",
    "          \"split_rate\": [0.7,0.1,0.2],\n",
    "          \"targets\": [\n",
    "            {\n",
    "              \"node\": \"Provider\",\n",
    "              \"property\": \"Fraud\",\n",
    "              \"type\": \"classification\"\n",
    "            }\n",
    "          ]\n",
    "        }\n",
    "      },\n",
    "\"jobSize\": \"medium\"}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%neptune_ml export start --export-url {neptune_ml.get_export_service_host()} --export-iam --wait --store-to export_results\n",
    "${export_params}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ML data processing, model training, and endpoint creation\n",
    "\n",
    "Once the export job is completed we are now ready to train our machine learning model and create the inference endpoint. Training our Neptune ML model requires three steps.  \n",
    "  \n",
    "<div style=\"background-color:#eeeeee; padding:10px; text-align:left; border-radius:10px; margin-top:10px; margin-bottom:10px; \"><b>Note</b>: The cells below only configure a minimal set of parameters required to run a model training.</div>\n",
    "\n",
    "## Data processing \n",
    "The first step (data processing) processes the exported graph dataset using standard feature preprocessing techniques to prepare it for use by DGL. This step performs functions such as feature normalization for numeric data and encoding text features using word2vec. At the conclusion of this step the dataset is formatted for model training. \n",
    "\n",
    "This step is implemented using a SageMaker Processing Job and data artifacts are stored in a pre-specified S3 location once the job is complete.\n",
    "\n",
    "Additional options and configuration parameters for the data processing job can be found using the links below:\n",
    "\n",
    "* [Data Processing](https://docs.aws.amazon.com/neptune/latest/userguide/machine-learning-on-graphs-processing.html)\n",
    "* [dataprocessing command](https://docs.aws.amazon.com/neptune/latest/userguide/machine-learning-api-dataprocessing.html)\n",
    "\n",
    "Run the cells below to create the data processing configuration and to begin the processing job."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The training_job_name can be set to a unique value below, otherwise one will be auto generated\n",
    "training_job_name=neptune_ml.get_training_job_name('node-classification')\n",
    "\n",
    "processing_params = f\"\"\"\n",
    "--config-file-name training-data-configuration.json\n",
    "--job-id {training_job_name} \n",
    "--s3-input-uri {export_results['outputS3Uri']} \n",
    "--s3-processed-uri {str(s3_bucket_uri)}/preloading \"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%neptune_ml dataprocessing start --wait --store-to processing_results {processing_params}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model training\n",
    "The second step (model training) trains the ML model that will be used for predictions. The model training is done in two stages. The first stage uses a SageMaker Processing job to generate a model training strategy. A model training strategy is a configuration set that specifies what type of model and model hyperparameter ranges will be used for the model training. Once the first stage is complete, the SageMaker Processing job launches a SageMaker Hyperparameter tuning job. The SageMaker Hyperparameter tuning job runs a pre-specified number of model training job trials on the processed data, and stores the model artifacts generated by the training in the output S3 location. Once all the training jobs are complete, the Hyperparameter tuning job also notes the training job that produced the best performing model.\n",
    "\n",
    "Additional options and configuration parameters for the data processing job can be found using the links below:\n",
    "\n",
    "* [Model Training](https://docs.aws.amazon.com/neptune/latest/userguide/machine-learning-on-graphs-model-training.html)\n",
    "* [modeltraining command](https://docs.aws.amazon.com/neptune/latest/userguide/machine-learning-api-modeltraining.html)\n",
    "\n",
    "<div style=\"background-color:#eeeeee; padding:10px; text-align:left; border-radius:10px; margin-top:10px; margin-bottom:10px; \"><b>Information</b>: The model training process takes ~20 minutes</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "training_params=f\"\"\"\n",
    "--job-id {training_job_name}-1\n",
    "--data-processing-id {training_job_name}\n",
    "--instance-type ml.p3.2xlarge\n",
    "--s3-output-uri {str(s3_bucket_uri)}/training\n",
    "--max-hpo-number 2\n",
    "--max-hpo-parallel 2 \"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%neptune_ml training start --wait --store-to training_results {training_params}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Endpoint creation\n",
    "\n",
    "The final step is to create the inference endpoint which is an Amazon SageMaker endpoint instance that is launched with the model artifacts produced by the best training job. This endpoint will be used by our graph queries to  return the model predictions for the inputs in the request. The endpoint once created stays active until it is manually deleted. Each model is tied to a single endpoint.\n",
    "\n",
    "Additional options and configuration parameters for the data processing job can be found using the links below:\n",
    "\n",
    "* [Inference Endpoint](https://docs.aws.amazon.com/neptune/latest/userguide/machine-learning-on-graphs-inference-endpoint.html)\n",
    "* [Endpoint command](https://docs.aws.amazon.com/neptune/latest/userguide/machine-learning-api-endpoints.html)\n",
    "\n",
    "<div style=\"background-color:#eeeeee; padding:10px; text-align:left; border-radius:10px; margin-top:10px; margin-bottom:10px; \"><b>Information</b>: The endpoint creation process takes ~5-10 minutes</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "endpoint_params=f\"\"\"\n",
    "--id {training_job_name}\n",
    "--model-training-job-id {training_job_name} \"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%neptune_ml endpoint create --wait --store-to endpoint_results {endpoint_params}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Once this has completed we get the endpoint name for our newly created inference endpoint.  The cell below will set the endpoint name which will be used in the Gremlin queries below. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "endpoint=endpoint_results['endpoint']['name']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "endpoint"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predicting Fraud using Gremlin queries\n",
    "\n",
    "Now that we have our inference endpoint setup let's query our provider knowledge graph to predict Fraud indicators for the providers that we updated earlier.  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lets predict the Fraud indicator for `1003809195`.  To accomplish this we need to add two steps:\n",
    "\n",
    "- First, we add the `with()` step to specify the inference endpoint we want to use with our Gremlin query like this\n",
    "`g.with(\"Neptune#ml.endpoint\",\"<INSERT ENDPOINT NAME>\")`.  \n",
    "\n",
    "<div style=\"background-color:#eeeeee; padding:10px; text-align:left; border-radius:10px; margin-top:10px; margin-bottom:10px; \"><b>Note</b>: The endpoint values are automatically passed into the queries below</div>\n",
    "\n",
    "- Second, when we ask for the property within our query we use the `properties()` step with an additional `with()` step (`with(\"Neptune#ml.classification\")`) which specifies that we want to retrieve the predicted value for this property.\n",
    "\n",
    "Putting these items together we get the query below, which will predict `Fraud` for the same provider we checked earlier "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "%%gremlin\n",
    "g.with(\"Neptune#ml.endpoint\", '${endpoint}').\n",
    "  V().has('NPI', ${NPI}).properties('Fraud').with(\"Neptune#ml.classification\").value()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "## Comparing the accuracy of predicted and actual genres"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, lets predict the Fraud indicator for all providers without a `fraud` attribute and store the results in a variable called \"results\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%gremlin --store-to results\n",
    "g.with(\"Neptune#ml.endpoint\",\"${endpoint}\").\n",
    "    V().hasLabel('Provider').hasNot('Fraud').\n",
    "    project('NPI', 'predicted').\n",
    "        by('NPI').\n",
    "        by( properties(\"Fraud\").with(\"Neptune#ml.classification\").value()).\n",
    "    order(local).by(keys, desc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Prepare a dataframe results that stores the predicted as well as the original values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "results = pd.DataFrame.from_dict(results,dtype='string')\n",
    "results = results.merge(test_df,on='NPI',how = 'inner')\n",
    "results['predicted'] = results['predicted'].astype('str')\n",
    "results = pd.get_dummies(results, drop_first = True )\n",
    "results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Store the results in arrays that we will use to prepare confusion matrix and a classification report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_preds = results.predicted_YES.values.astype('float32')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_true = results.Fraud_YES.values.astype('float32')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Install seaborn and sklearn libraries that are needed for preparing the confusion matrix and a classification report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install seaborn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install sklearn"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import necessary libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import balanced_accuracy_score\n",
    "from sklearn.metrics import classification_report"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create confusion matrix and classification report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_confusion_matrix(y_true, y_predicted):\n",
    "\n",
    "    cm  = confusion_matrix(y_true, y_predicted)\n",
    "    # Get the per-class normalized value for each cell\n",
    "    cm_norm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "    \n",
    "    # We color each cell according to its normalized value, annotate with exact counts.\n",
    "    ax = sns.heatmap(cm_norm, annot=cm, fmt=\"d\")\n",
    "    ax.set(xticklabels=[\"non-fraud\", \"fraud\"], yticklabels=[\"non-fraud\", \"fraud\"])\n",
    "    ax.set_ylim([0,2])\n",
    "    plt.title('Confusion Matrix')\n",
    "    plt.ylabel('Real Classes')\n",
    "    plt.xlabel('Predicted Classes')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "print(\"Balanced accuracy = {:.3f}\".format(balanced_accuracy_score(y_true, y_preds)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_confusion_matrix(y_true, y_preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "print(classification_report(\n",
    "    y_true, y_preds, target_names=['non-fraud', 'fraud']))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Comparing the `original` versus the `predicted` results we see that our model did a good job of predicting whether the provider is fraudulent (46 out of 55) while also minimizing the number of falsely classified as fraud (3 out of 55)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Cleaning up \n",
    "Now that you have completed this walkthrough you have created a Sagemaker endpoint which is currently running and will incur the standard charges.  If you are done trying out Neptune ML and would like to avoid these recurring costs, run the cell below to delete the inference endpoint.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "neptune_ml.delete_endpoint(training_job_name)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In addition to the inference endpoint the CloudFormation script that you used has setup several additional resources.  If you are finished then we suggest you delete the CloudFormation stack to avoid any recurring charges. For instructions, see Deleting a Stack on the [Deleting a Stack on the Amazon Web Services CloudFormation Console](https://docs.aws.amazon.com/AWSCloudFormation/latest/UserGuide/cfn-console-delete-stack.html). Be sure to delete the root stack (the stack you created earlier). Deleting the root stack deletes any nested stacks."
   ]
  }
 ],
 "metadata": {
  "celltoolbar": "Raw Cell Format",
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}