Train_Reviews_Comprehend.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Setup Comprehend Through CLI/API:\n",
    "https://docs.aws.amazon.com/comprehend/latest/dg/get-started-customclass.html"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Setup Comprehend Through AWS Console \n",
    "\n",
    "https://docs.aws.amazon.com/comprehend/latest/dg/getting-started-document-classification.html"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Good example of using Comprehend for Positive/Negative Sentiment:  https://github.com/aws-samples/amazon-comprehend-custom-entity/blob/master/3-AWS-Comprehend-Negative-Custom-Classifier.ipynb"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Make sure this SageMakerNotebookExecutionRole has access to Comprehend"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import boto3\n",
    "import sagemaker\n",
    "import pandas as pd\n",
    "\n",
    "sess   = sagemaker.Session()\n",
    "bucket = sess.default_bucket()\n",
    "role = sagemaker.get_execution_role()\n",
    "region = boto3.Session().region_name\n",
    "\n",
    "sm = boto3.Session().client(service_name='sagemaker', region_name=region)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "comprehend = boto3.client('comprehend')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "s3://sagemaker-us-east-2-533787958253/data/amazon_reviews_us_Digital_Software_v1_00_noheader.csv\n"
     ]
    }
   ],
   "source": [
    "%store -r noheader_train_s3_uri\n",
    "\n",
    "print(noheader_train_s3_uri)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2020-04-29 14:19:18   15164581 amazon_reviews_us_Digital_Software_v1_00_noheader.csv\r\n"
     ]
    }
   ],
   "source": [
    "!aws s3 ls $noheader_train_s3_uri"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create Data Access Role for Comprehend"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create Policy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "assume_role_policy_doc = {\n",
    "  \"Version\": \"2012-10-17\",\n",
    "  \"Statement\": [\n",
    "    {\n",
    "      \"Effect\": \"Allow\",\n",
    "      \"Principal\": {\n",
    "        \"Service\": \"comprehend.amazonaws.com\"\n",
    "      },\n",
    "      \"Action\": \"sts:AssumeRole\"\n",
    "    }\n",
    "  ]\n",
    "} "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create Role and Attach Policies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "#iam_comprehend_role_name = 'DSOAWS_Comprehend'\n",
    "iam_comprehend_role_name = 'COEAWS_Comprehend'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "import boto3\n",
    "from botocore.exceptions import ClientError\n",
    "\n",
    "try:\n",
    "    iam = boto3.client('iam')\n",
    "\n",
    "    iam_role_comprehend = iam.create_role(\n",
    "        RoleName=iam_comprehend_role_name,\n",
    "        AssumeRolePolicyDocument=json.dumps(assume_role_policy_doc),\n",
    "        Description='COEAWS Comprehend Role'\n",
    "    )\n",
    "except ClientError as e:\n",
    "    if e.response['Error']['Code'] == 'EntityAlreadyExists':\n",
    "        iam_role_comprehend = iam.get_role(RoleName=iam_comprehend_role_name)\n",
    "        print(\"Role already exists\")\n",
    "    else:\n",
    "        print(\"Unexpected error: %s\" % e)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If you have and Unexpected error above you need to create an IAM inline access policy and inline attach\n",
    "\n",
    "```json\n",
    "{\n",
    "    \"Version\": \"2012-10-17\",\n",
    "    \"Statement\": [\n",
    "        {\n",
    "            \"Sid\": \"jsr1\",\n",
    "            \"Effect\": \"Allow\",\n",
    "            \"Action\": [\n",
    "                \"iam:CreateRole\"\n",
    "            ],\n",
    "            \"Resource\": [\n",
    "                \"*\"\n",
    "            ]\n",
    "        }\n",
    "    ]\n",
    "}\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Version': '2012-10-17', 'Statement': [{'Action': ['s3:GetObject'], 'Resource': ['arn:aws:s3:::sagemaker-us-east-2-533787958253/*'], 'Effect': 'Allow'}, {'Action': ['s3:ListBucket'], 'Resource': ['arn:aws:s3:::sagemaker-us-east-2-533787958253'], 'Effect': 'Allow'}, {'Action': ['s3:PutObject'], 'Resource': ['arn:aws:s3:::sagemaker-us-east-2-533787958253/*'], 'Effect': 'Allow'}]}\n"
     ]
    }
   ],
   "source": [
    "comprehend_s3_policy_doc = {\n",
    "    \"Version\": \"2012-10-17\",\n",
    "    \"Statement\": [\n",
    "        {\n",
    "            \"Action\": [\n",
    "                \"s3:GetObject\"\n",
    "            ],\n",
    "            \"Resource\": [\n",
    "                \"arn:aws:s3:::{}/*\".format(bucket)\n",
    "            ],\n",
    "            \"Effect\": \"Allow\"\n",
    "        },\n",
    "        {\n",
    "            \"Action\": [\n",
    "                \"s3:ListBucket\"\n",
    "            ],\n",
    "            \"Resource\": [\n",
    "                \"arn:aws:s3:::{}\".format(bucket)\n",
    "            ],\n",
    "            \"Effect\": \"Allow\"\n",
    "        },\n",
    "        {\n",
    "            \"Action\": [\n",
    "                \"s3:PutObject\"\n",
    "            ],\n",
    "            \"Resource\": [\n",
    "                \"arn:aws:s3:::{}/*\".format(bucket)\n",
    "            ],\n",
    "            \"Effect\": \"Allow\"\n",
    "        }\n",
    "    ]\n",
    "}\n",
    "\n",
    "print(comprehend_s3_policy_doc)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "response = iam.put_role_policy(\n",
    "    RoleName=iam_comprehend_role_name,\n",
    "    PolicyName='COEAWS_ComprehendPolicyToS3',\n",
    "    PolicyDocument=json.dumps(comprehend_s3_policy_doc)\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If you have an error you'll need to add PutRolePolicy\n",
    "\n",
    "```json\n",
    "{\n",
    "    \"Version\": \"2012-10-17\",\n",
    "    \"Statement\": [\n",
    "        {\n",
    "            \"Sid\": \"jsr2\",\n",
    "            \"Effect\": \"Allow\",\n",
    "            \"Action\": [\n",
    "                \"iam:PutRolePolicy\"\n",
    "            ],\n",
    "            \"Resource\": [\n",
    "                \"*\"\n",
    "            ]\n",
    "        }\n",
    "    ]\n",
    "}\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train the Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "s3://sagemaker-us-east-2-533787958253/models/comprehend/output\n"
     ]
    }
   ],
   "source": [
    "prefix = 'models'\n",
    "\n",
    "s3_output_job = 's3://{}/{}/{}'.format(bucket, prefix, 'comprehend/output')\n",
    "print(s3_output_job)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "iam_role_comprehend_arn = iam_role_comprehend['Role']['Arn']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "import datetime\n",
    "\n",
    "id = str(datetime.datetime.now().strftime(\"%s\"))\n",
    "\n",
    "training_job = comprehend.create_document_classifier(\n",
    "    DocumentClassifierName='Amazon-Customer-Reviews-Classifier-'+ id,\n",
    "    DataAccessRoleArn=iam_role_comprehend_arn,\n",
    "    InputDataConfig={\n",
    "        'S3Uri': noheader_train_s3_uri\n",
    "    },\n",
    "    OutputDataConfig={\n",
    "        'S3Uri': s3_output_job\n",
    "    },\n",
    "    LanguageCode='en'\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If the above has errors you'll need to create a PassRole policy\n",
    "\n",
    "```json\n",
    "{\n",
    "    \"Version\": \"2012-10-17\",\n",
    "    \"Statement\": [\n",
    "        {\n",
    "            \"Sid\": \"jsr3\",\n",
    "            \"Effect\": \"Allow\",\n",
    "            \"Action\": [\n",
    "                \"iam:PassRole\"\n",
    "            ],\n",
    "            \"Resource\": [\n",
    "                \"*\"\n",
    "            ]\n",
    "        }\n",
    "    ]\n",
    "}\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: SUBMITTED\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINING\n",
      "Custom classifier: TRAINED\n",
      "\n",
      "Status TRAINED\n",
      "\n",
      "{'DocumentClassifierArn': 'arn:aws:comprehend:us-east-2:533787958253:document-classifier/Amazon-Customer-Reviews-Classifier-1588622802', 'LanguageCode': 'en', 'Status': 'TRAINED', 'SubmitTime': datetime.datetime(2020, 5, 4, 20, 6, 43, 57000, tzinfo=tzlocal()), 'EndTime': datetime.datetime(2020, 5, 4, 20, 44, 59, 324000, tzinfo=tzlocal()), 'TrainingStartTime': datetime.datetime(2020, 5, 4, 20, 10, 18, 836000, tzinfo=tzlocal()), 'TrainingEndTime': datetime.datetime(2020, 5, 4, 20, 44, 8, 222000, tzinfo=tzlocal()), 'InputDataConfig': {'S3Uri': 's3://sagemaker-us-east-2-533787958253/data/amazon_reviews_us_Digital_Software_v1_00_noheader.csv'}, 'OutputDataConfig': {'S3Uri': 's3://sagemaker-us-east-2-533787958253/models/comprehend/output/533787958253-CLR-2de1788aef00df3cb08c64d47834cfd1/output/output.tar.gz'}, 'ClassifierMetadata': {'NumberOfLabels': 5, 'NumberOfTrainedDocuments': 31005, 'NumberOfTestDocuments': 3445, 'EvaluationMetrics': {'Accuracy': 0.5213, 'Precision': 0.5223, 'Recall': 0.5213, 'F1Score': 0.5214, 'MicroPrecision': 0.5213, 'MicroRecall': 0.5213, 'MicroF1Score': 0.5213, 'HammingLoss': 0.4787}}, 'DataAccessRoleArn': 'arn:aws:iam::533787958253:role/COEAWS_Comprehend', 'Mode': 'MULTI_CLASS'}\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "\n",
    "jobArn = training_job['DocumentClassifierArn']\n",
    "\n",
    "max_time = time.time() + 3 * 60 * 60 # 3 hours\n",
    "while time.time() < max_time:\n",
    "    describe_custom_classifier = comprehend.describe_document_classifier(\n",
    "        DocumentClassifierArn = jobArn\n",
    "    )\n",
    "    status = describe_custom_classifier[\"DocumentClassifierProperties\"][\"Status\"]\n",
    "    print(\"Custom classifier: {}\".format(status))\n",
    "    \n",
    "    if status == \"TRAINED\" or status == \"IN_ERROR\":\n",
    "        print('')\n",
    "        print('Status {}'.format(status))\n",
    "        print('')\n",
    "        print(describe_custom_classifier[\"DocumentClassifierProperties\"])\n",
    "        break\n",
    "        \n",
    "    time.sleep(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2020-03-04 22:00:41 coe-ai-video\r\n",
      "2020-03-04 19:39:23 coe-cal-ai-polly-anil\r\n",
      "2020-03-04 17:56:16 coe-faas\r\n",
      "2020-03-26 17:28:18 coe-opm-faces\r\n",
      "2020-03-09 16:58:45 rekognition-video-console-demo-cmh-533787958253-tubndls8jejq0n\r\n",
      "2020-03-03 18:41:08 sagemaker-studio-0qnahwofwbs\r\n",
      "2020-03-03 15:56:05 sagemaker-studio-3rk9wz4j1uh\r\n",
      "2020-03-03 18:32:07 sagemaker-studio-fbc9rb6owm\r\n",
      "2020-03-03 17:45:10 sagemaker-studio-k06y0dzfxg\r\n",
      "2020-03-03 18:51:13 sagemaker-studio-k0n6398mj8\r\n",
      "2020-03-03 17:35:55 sagemaker-studio-mwp36qinptr\r\n",
      "2020-03-03 18:41:30 sagemaker-studio-qlfop46x0ji\r\n",
      "2020-03-03 17:48:24 sagemaker-studio-wa13v0wmlme\r\n",
      "2020-04-27 14:05:08 sagemaker-us-east-2-533787958253\r\n"
     ]
    }
   ],
   "source": [
    "!aws s3 ls "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Show Results of the Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'DocumentClassifierArn': 'arn:aws:comprehend:us-east-2:533787958253:document-classifier/Amazon-Customer-Reviews-Classifier-1588350035', 'LanguageCode': 'en', 'Status': 'TRAINED', 'SubmitTime': datetime.datetime(2020, 5, 1, 16, 20, 35, 361000, tzinfo=tzlocal()), 'EndTime': datetime.datetime(2020, 5, 1, 16, 59, 40, 62000, tzinfo=tzlocal()), 'TrainingStartTime': datetime.datetime(2020, 5, 1, 16, 22, 50, 848000, tzinfo=tzlocal()), 'TrainingEndTime': datetime.datetime(2020, 5, 1, 16, 57, 28, 837000, tzinfo=tzlocal()), 'InputDataConfig': {'S3Uri': 's3://sagemaker-us-east-2-533787958253/data/amazon_reviews_us_Digital_Software_v1_00_noheader.csv'}, 'OutputDataConfig': {'S3Uri': 's3://sagemaker-us-east-2-533787958253/models/comprehend/output/533787958253-CLR-c38b32ce24d337245e9009550ff06cae/output/output.tar.gz'}, 'ClassifierMetadata': {'NumberOfLabels': 5, 'NumberOfTrainedDocuments': 31005, 'NumberOfTestDocuments': 3445, 'EvaluationMetrics': {'Accuracy': 0.5202, 'Precision': 0.5304, 'Recall': 0.5202, 'F1Score': 0.5239, 'MicroPrecision': 0.5202, 'MicroRecall': 0.5202, 'MicroF1Score': 0.5202, 'HammingLoss': 0.4798}}, 'DataAccessRoleArn': 'arn:aws:iam::533787958253:role/DSOAWS_Comprehend', 'Mode': 'MULTI_CLASS'}\n"
     ]
    }
   ],
   "source": [
    "print(describe_custom_classifier[\"DocumentClassifierProperties\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "arn:aws:comprehend:us-east-2:533787958253:document-classifier/Amazon-Customer-Reviews-Classifier-1588350035\n"
     ]
    }
   ],
   "source": [
    "model_arn = describe_custom_classifier[\"DocumentClassifierProperties\"][\"DocumentClassifierArn\"]\n",
    "print(model_arn)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "s3://sagemaker-us-east-2-533787958253/models/comprehend/output/533787958253-CLR-c38b32ce24d337245e9009550ff06cae/output/output.tar.gz\n",
      "models/comprehend/output/533787958253-CLR-c38b32ce24d337245e9009550ff06cae/output/output.tar.gz\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "#Retrieve the S3URI from the model output and create jobkey variable.\n",
    "job_output = describe_custom_classifier[\"DocumentClassifierProperties\"][\"OutputDataConfig\"][\"S3Uri\"]\n",
    "print(job_output)\n",
    "\n",
    "path_prefix = 's3://{}/'.format(bucket)\n",
    "\n",
    "job_key = os.path.relpath(job_output, path_prefix)\n",
    "\n",
    "print(job_key)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Completed 291 Bytes/291 Bytes (3.1 KiB/s) with 1 file(s) remaining\r",
      "download: s3://sagemaker-us-east-2-533787958253/models/comprehend/output/533787958253-CLR-c38b32ce24d337245e9009550ff06cae/output/output.tar.gz to ./output.tar.gz\r\n"
     ]
    }
   ],
   "source": [
    "!aws s3 cp $job_output ./output.tar.gz"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Download the model metrics\n",
    "s3 = boto3.resource('s3')\n",
    "\n",
    "s3.Bucket(bucket).download_file(job_key, './output.tar.gz')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "output/\r\n",
      "output/confusion_matrix.json\r\n"
     ]
    }
   ],
   "source": [
    "#Unpack the gzip file\n",
    "!tar xvzf ./output.tar.gz"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\n",
      "  \"confusion_matrix\": [\n",
      "    [\n",
      "      411,\n",
      "      208,\n",
      "      50,\n",
      "      11,\n",
      "      9\n",
      "    ],\n",
      "    [\n",
      "      125,\n",
      "      340,\n",
      "      177,\n",
      "      34,\n",
      "      13\n",
      "    ],\n",
      "    [\n",
      "      56,\n",
      "      182,\n",
      "      312,\n",
      "      108,\n",
      "      31\n",
      "    ],\n",
      "    [\n",
      "      13,\n",
      "      42,\n",
      "      163,\n",
      "      320,\n",
      "      151\n",
      "    ],\n",
      "    [\n",
      "      16,\n",
      "      20,\n",
      "      43,\n",
      "      201,\n",
      "      409\n",
      "    ]\n",
      "  ],\n",
      "  \"labels\": [\n",
      "    \"1\",\n",
      "    \"2\",\n",
      "    \"3\",\n",
      "    \"4\",\n",
      "    \"5\"\n",
      "  ],\n",
      "  \"type\": \"multi_class\",\n",
      "  \"all_labels\": [\n",
      "    \"1\",\n",
      "    \"2\",\n",
      "    \"3\",\n",
      "    \"4\",\n",
      "    \"5\"\n",
      "  ]\n",
      "}\n"
     ]
    }
   ],
   "source": [
    "import json\n",
    "\n",
    "with open('./output/confusion_matrix.json') as json_file:\n",
    "    data = json.load(json_file)\n",
    "print(json.dumps(data, indent=2, default=str))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: tabulate in /home/ec2-user/anaconda3/envs/python3/lib/python3.6/site-packages (0.8.7)\n",
      "\u001b[33mYou are using pip version 10.0.1, however version 20.1 is available.\n",
      "You should consider upgrading via the 'pip install --upgrade pip' command.\u001b[0m\n"
     ]
    }
   ],
   "source": [
    "!pip install tabulate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<tbody>\n",
       "<tr><td>        </td><td style=\"text-align: right;\">  1</td><td style=\"text-align: right;\">  2</td><td style=\"text-align: right;\">  3</td><td style=\"text-align: right;\">  4</td><td style=\"text-align: right;\">  5</td><td>(Predicted)</td></tr>\n",
       "<tr><td>1       </td><td style=\"text-align: right;\">411</td><td style=\"text-align: right;\">208</td><td style=\"text-align: right;\"> 50</td><td style=\"text-align: right;\"> 11</td><td style=\"text-align: right;\">  9</td><td>           </td></tr>\n",
       "<tr><td>2       </td><td style=\"text-align: right;\">125</td><td style=\"text-align: right;\">340</td><td style=\"text-align: right;\">177</td><td style=\"text-align: right;\"> 34</td><td style=\"text-align: right;\"> 13</td><td>           </td></tr>\n",
       "<tr><td>3       </td><td style=\"text-align: right;\"> 56</td><td style=\"text-align: right;\">182</td><td style=\"text-align: right;\">312</td><td style=\"text-align: right;\">108</td><td style=\"text-align: right;\"> 31</td><td>           </td></tr>\n",
       "<tr><td>4       </td><td style=\"text-align: right;\"> 13</td><td style=\"text-align: right;\"> 42</td><td style=\"text-align: right;\">163</td><td style=\"text-align: right;\">320</td><td style=\"text-align: right;\">151</td><td>           </td></tr>\n",
       "<tr><td>5       </td><td style=\"text-align: right;\"> 16</td><td style=\"text-align: right;\"> 20</td><td style=\"text-align: right;\"> 43</td><td style=\"text-align: right;\">201</td><td style=\"text-align: right;\">409</td><td>           </td></tr>\n",
       "<tr><td>(Actual)</td><td style=\"text-align: right;\">   </td><td style=\"text-align: right;\">   </td><td style=\"text-align: right;\">   </td><td style=\"text-align: right;\">   </td><td style=\"text-align: right;\">   </td><td>           </td></tr>\n",
       "</tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from IPython.display import HTML, display\n",
    "import tabulate\n",
    "table = [['', '1', '2', '3', '4', '5', '(Predicted)'],\n",
    "         ['1', data['confusion_matrix'][0][0], data['confusion_matrix'][0][1], data['confusion_matrix'][0][2], data['confusion_matrix'][0][3], data['confusion_matrix'][0][4]],\n",
    "         ['2', data['confusion_matrix'][1][0], data['confusion_matrix'][1][1], data['confusion_matrix'][1][2], data['confusion_matrix'][1][3], data['confusion_matrix'][1][4]],\n",
    "         ['3', data['confusion_matrix'][2][0], data['confusion_matrix'][2][1], data['confusion_matrix'][2][2], data['confusion_matrix'][2][3], data['confusion_matrix'][2][4]],\n",
    "         ['4', data['confusion_matrix'][3][0], data['confusion_matrix'][3][1], data['confusion_matrix'][3][2], data['confusion_matrix'][3][3], data['confusion_matrix'][3][4]],\n",
    "         ['5', data['confusion_matrix'][4][0], data['confusion_matrix'][4][1], data['confusion_matrix'][4][2], data['confusion_matrix'][4][3], data['confusion_matrix'][4][4]],\n",
    "         ['(Actual)']]\n",
    "display(HTML(tabulate.tabulate(table, tablefmt='html')))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Deploy Endpoint"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "inference_endpoint_response = comprehend.create_endpoint(\n",
    "    EndpointName='comprehend-inference-endpoint',\n",
    "    ModelArn = model_arn,\n",
    "    DesiredInferenceUnits = 1\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "arn:aws:comprehend:us-east-2:533787958253:document-classifier-endpoint/comprehend-inference-endpoint\n"
     ]
    }
   ],
   "source": [
    "endpoint_arn = inference_endpoint_response[\"EndpointArn\"]\n",
    "print(endpoint_arn)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Predict with Endpoint"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'EndpointProperties': {'EndpointArn': 'arn:aws:comprehend:us-east-2:533787958253:document-classifier-endpoint/comprehend-inference-endpoint', 'Status': 'CREATING', 'ModelArn': 'arn:aws:comprehend:us-east-2:533787958253:document-classifier/Amazon-Customer-Reviews-Classifier-1588350035', 'DesiredInferenceUnits': 1, 'CurrentInferenceUnits': 0, 'CreationTime': datetime.datetime(2020, 5, 1, 18, 12, 35, 261000, tzinfo=tzlocal()), 'LastModifiedTime': datetime.datetime(2020, 5, 1, 18, 12, 35, 261000, tzinfo=tzlocal())}, 'ResponseMetadata': {'RequestId': '4fd7303b-ead1-453e-add7-1ed80852ad9f', 'HTTPStatusCode': 200, 'HTTPHeaders': {'x-amzn-requestid': '4fd7303b-ead1-453e-add7-1ed80852ad9f', 'content-type': 'application/x-amz-json-1.1', 'content-length': '402', 'date': 'Fri, 01 May 2020 18:13:04 GMT'}, 'RetryAttempts': 0}}\n"
     ]
    }
   ],
   "source": [
    "describe_response = comprehend.describe_endpoint(\n",
    "    EndpointArn = endpoint_arn\n",
    ")\n",
    "print(describe_response)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: CREATING\n",
      "Endpoint: IN_SERVICE\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "\n",
    "max_time = time.time() + 3*60*60 # 3 hours\n",
    "while time.time() < max_time:\n",
    "    describe_response = comprehend.describe_endpoint(\n",
    "        EndpointArn = endpoint_arn\n",
    "    )\n",
    "    status = describe_response[\"EndpointProperties\"][\"Status\"]\n",
    "    print(\"Endpoint: {}\".format(status))\n",
    "    \n",
    "    if status == \"IN_SERVICE\" or status == \"IN_ERROR\":\n",
    "        break\n",
    "        \n",
    "    time.sleep(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\n",
      "  \"Classes\": [\n",
      "    {\n",
      "      \"Name\": \"5\",\n",
      "      \"Score\": 0.9981356263160706\n",
      "    },\n",
      "    {\n",
      "      \"Name\": \"4\",\n",
      "      \"Score\": 0.0018844201695173979\n",
      "    },\n",
      "    {\n",
      "      \"Name\": \"3\",\n",
      "      \"Score\": 1.0016706255555619e-05\n",
      "    }\n",
      "  ],\n",
      "  \"ResponseMetadata\": {\n",
      "    \"RequestId\": \"edd5d8c7-9f33-4b32-9d68-d19d2339dca3\",\n",
      "    \"HTTPStatusCode\": 200,\n",
      "    \"HTTPHeaders\": {\n",
      "      \"x-amzn-requestid\": \"edd5d8c7-9f33-4b32-9d68-d19d2339dca3\",\n",
      "      \"content-type\": \"application/x-amz-json-1.1\",\n",
      "      \"content-length\": \"139\",\n",
      "      \"date\": \"Fri, 01 May 2020 18:25:11 GMT\"\n",
      "    },\n",
      "    \"RetryAttempts\": 0\n",
      "  }\n",
      "}\n"
     ]
    }
   ],
   "source": [
    "txt = \"\"\"I loved it!  I will recommend this to everyone.\"\"\"\n",
    "\n",
    "response = comprehend.classify_document(\n",
    "    Text= txt,\n",
    "    EndpointArn = endpoint_arn\n",
    ")\n",
    "\n",
    "import json\n",
    "print(json.dumps(response, indent=2, default=str))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\n",
      "  \"Classes\": [\n",
      "    {\n",
      "      \"Name\": \"2\",\n",
      "      \"Score\": 0.8435407876968384\n",
      "    },\n",
      "    {\n",
      "      \"Name\": \"3\",\n",
      "      \"Score\": 0.08270315825939178\n",
      "    },\n",
      "    {\n",
      "      \"Name\": \"1\",\n",
      "      \"Score\": 0.07377045601606369\n",
      "    }\n",
      "  ],\n",
      "  \"ResponseMetadata\": {\n",
      "    \"RequestId\": \"51740b3b-09bd-4bbc-a0c1-0c024780133f\",\n",
      "    \"HTTPStatusCode\": 200,\n",
      "    \"HTTPHeaders\": {\n",
      "      \"x-amzn-requestid\": \"51740b3b-09bd-4bbc-a0c1-0c024780133f\",\n",
      "      \"content-type\": \"application/x-amz-json-1.1\",\n",
      "      \"content-length\": \"135\",\n",
      "      \"date\": \"Fri, 01 May 2020 18:25:42 GMT\"\n",
      "    },\n",
      "    \"RetryAttempts\": 0\n",
      "  }\n",
      "}\n"
     ]
    }
   ],
   "source": [
    "txt = \"\"\"Really bad.  I hope they don't make this anymore.\"\"\"\n",
    "\n",
    "response = comprehend.classify_document(\n",
    "    Text= txt,\n",
    "    EndpointArn = endpoint_arn\n",
    ")\n",
    "\n",
    "import json\n",
    "print(json.dumps(response, indent=2, default=str))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\n",
      "  \"Classes\": [\n",
      "    {\n",
      "      \"Name\": \"3\",\n",
      "      \"Score\": 0.9962722659111023\n",
      "    },\n",
      "    {\n",
      "      \"Name\": \"4\",\n",
      "      \"Score\": 0.0035591258201748133\n",
      "    },\n",
      "    {\n",
      "      \"Name\": \"5\",\n",
      "      \"Score\": 0.00018858080147765577\n",
      "    }\n",
      "  ],\n",
      "  \"ResponseMetadata\": {\n",
      "    \"RequestId\": \"d642110d-d8d9-41b1-9978-bff8cf20477e\",\n",
      "    \"HTTPStatusCode\": 200,\n",
      "    \"HTTPHeaders\": {\n",
      "      \"x-amzn-requestid\": \"d642110d-d8d9-41b1-9978-bff8cf20477e\",\n",
      "      \"content-type\": \"application/x-amz-json-1.1\",\n",
      "      \"content-length\": \"139\",\n",
      "      \"date\": \"Fri, 01 May 2020 18:28:57 GMT\"\n",
      "    },\n",
      "    \"RetryAttempts\": 0\n",
      "  }\n",
      "}\n"
     ]
    }
   ],
   "source": [
    "txt = \"\"\"It's OK.\"\"\"\n",
    "\n",
    "response = comprehend.classify_document(\n",
    "    Text= txt,\n",
    "    EndpointArn = endpoint_arn\n",
    ")\n",
    "\n",
    "import json\n",
    "print(json.dumps(response, indent=2, default=str))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Pretty printing has been turned OFF\n"
     ]
    }
   ],
   "source": [
    "pprint(json.dumps(response, indent=2, default=str))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "conda_python3",
   "language": "python",
   "name": "conda_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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}