Pytorch/(Pytorch) Sentiment Analysis with Transformer BERT.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Google-play App Review Sentiment Analysis with BERT "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "I0424 14:12:08.299082 14500 file_utils.py:35] PyTorch version 1.3.1 available.\n",
      "I0424 14:12:10.643192 14500 file_utils.py:48] TensorFlow version 2.0.0 available.\n"
     ]
    }
   ],
   "source": [
    "import transformers\n",
    "from transformers import BertModel, BertTokenizer, AdamW, get_linear_schedule_with_warmup\n",
    "\n",
    "import torch\n",
    "from torch import nn, optim\n",
    "from torch.utils.data import Dataset, DataLoader\n",
    "\n",
    "import json\n",
    "from tqdm import tqdm\n",
    "\n",
    "import random\n",
    "import os\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "from pylab import rcParams\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib import rc\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import confusion_matrix, classification_report\n",
    "from collections import defaultdict\n",
    "from textwrap import wrap\n",
    "\n",
    "from pygments import highlight\n",
    "from pygments.lexers import JsonLexer\n",
    "from pygments.formatters import TerminalFormatter\n",
    "\n",
    "# Google-Play-Scraper provides APIs to easily crawl the Google Play Store for Python without any external dependencies!\n",
    "from google_play_scraper import Sort, reviews, app\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "%config InlineBackend.figure_format='retina'\n",
    "sns.set(style='whitegrid', palette='muted', font_scale=1.2)\n",
    "HAPPY_COLORS_PALETTE = [\"#01BEFE\", \"#FFDD00\", \"#FF7D00\", \"#FF006D\", \"#ADFF02\", \"#8F00FF\"]\n",
    "sns.set_palette(sns.color_palette(HAPPY_COLORS_PALETTE))\n",
    "rcParams['figure.figsize'] = 12, 8"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def seed_everything(seed = 623):\n",
    "    random.seed(seed)\n",
    "    os.environ['PYTHONHASHSEED'] = str(seed)\n",
    "    np.random.seed(seed)\n",
    "    torch.manual_seed(seed)\n",
    "    torch.cuda.manual_seed(seed)\n",
    "    torch.backends.cudnn.deterministic = True\n",
    "seed_everything()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Get top US APPs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "app_packages = [\n",
    "  'com.anydo',\n",
    "  'com.todoist',\n",
    "  'com.ticktick.task',\n",
    "  'com.habitrpg.android.habitica',\n",
    "  'cc.forestapp',\n",
    "  'com.oristats.habitbull',\n",
    "  'com.levor.liferpgtasks',\n",
    "  'com.habitnow',\n",
    "  'com.microsoft.todos',\n",
    "  'prox.lab.calclock',\n",
    "  'com.gmail.jmartindev.timetune',\n",
    "  'com.artfulagenda.app',\n",
    "  'com.tasks.android',\n",
    "  'com.appgenix.bizcal',\n",
    "  'com.appxy.planner'\n",
    "]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Get App Information"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 15/15 [00:06<00:00,  2.48it/s]\n"
     ]
    }
   ],
   "source": [
    "# Code from https://github.com/JoMingyu/google-play-scraper\n",
    "app_infos = []\n",
    "\n",
    "for ap in tqdm(app_packages):\n",
    "    info = app(ap, \n",
    "               lang='en', # defaults to 'en'\n",
    "               country='us') # defaults to 'us'\n",
    "    del info['comments']\n",
    "    app_infos.append(info)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Get information for all 15 APPs above\n",
    "def print_json(json_object):\n",
    "    json_str = json.dumps(\n",
    "    json_object,\n",
    "    indent=2,\n",
    "    sort_keys=True,\n",
    "    default=str\n",
    "  )\n",
    "    print(highlight(json_str, JsonLexer(), TerminalFormatter()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\n",
      "  \u001b[94m\"adSupported\"\u001b[39;49;00m: \u001b[34mnull\u001b[39;49;00m,\n",
      "  \u001b[94m\"androidVersion\"\u001b[39;49;00m: \u001b[33m\"Varies\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"androidVersionText\"\u001b[39;49;00m: \u001b[33m\"Varies with device\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"appId\"\u001b[39;49;00m: \u001b[33m\"com.anydo\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"containsAds\"\u001b[39;49;00m: \u001b[34mnull\u001b[39;49;00m,\n",
      "  \u001b[94m\"contentRating\"\u001b[39;49;00m: \u001b[33m\"Everyone\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"contentRatingDescription\"\u001b[39;49;00m: \u001b[34mnull\u001b[39;49;00m,\n",
      "  \u001b[94m\"currency\"\u001b[39;49;00m: \u001b[33m\"USD\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"description\"\u001b[39;49;00m: \u001b[33m\"<b>\\ud83c\\udfc6 Editor's Choice by Google</b>\\r\\n\\r\\nAny.do is a To Do List, Calendar, Planner, Tasks & Reminders App That Helps Over 25M People Stay Organized and Get More Done.\\r\\n\\r\\n<b>\\ud83e\\udd47 \\\"It\\u2019s A MUST HAVE PLANNER & TO DO LIST APP\\\" (NYTimes, USA TODAY, WSJ & Lifehacker).</b>\\r\\n\\r\\nAny.do is a free to-do list, planner & calendar app for managing and organizing your daily tasks, to-do lists, notes, reminders, checklists, calendar events, grocery lists and more.\\r\\n\\r\\n\\ud83d\\udcc5 Organize Your Tasks & To-Do List in Seconds\\r\\n\\r\\n\\u2022 ADVANCED CALENDAR & DAILY PLANNER - Keep your to-do list and calendar events always at hand with our calendar widget. Any.do to-do list & planner support daily calendar view, 3-day Calendar view, Weekly calendar view & agenda view, with built-in reminders. Review and organize your calendar events and to do list side by side.\\r\\n\\r\\n\\u2022 SYNCS SEAMLESSLY - Keeps all your to do list, tasks, reminders, notes, calendar & agenda always in sync so you\\u2019ll never forget a thing. Sync your phone\\u2019s calendar, google calendar, Facebook events, outlook calendar or any other calendar so you don\\u2019t forget an important event.\\r\\n\\r\\n\\u2022 SET REMINDERS - One time reminders, recurring reminders, Location reminders & voice reminders. NEW! Easily create tasks and get reminders in WhatsApp.\\r\\n\\r\\n\\u2022 WORK TOGETHER - Share your to do list and assign tasks with your friends, family & colleagues from your task list to collaborate and get more done. \\r\\n\\r\\n---\\r\\n\\r\\nALL-IN-ONE PLANNER & CALENDAR APP FOR GETTING THINGS DONE\\r\\nCreate and set reminders with voice to your to do list. \\r\\nFor better task management flow we added a calendar integration to keep your agenda always up to date. \\r\\nFor better productivity, we added recurring reminders, location reminders, one-time reminder, sub-tasks, notes & file attachments. \\r\\nTo keep your to do list up to date, we\\u2019ve added a daily planner and focus mode.\\r\\n\\r\\nINTEGRATIONS\\r\\nAny.do To do list, Calendar, planner & Reminders Integrates with Google Calendar, Outlook, WhatsApp, Slack, Gmail, Google Tasks, Evernote, Trello, Wunderlist, Todoist, Zapier, Asana, Microsoft to-do, Salesforce, OneNote, Google Assistant, Amazon Alexa, Office 365, Exchange, Jira & More.\\r\\n\\r\\nTO DO LIST, CALENDAR, PLANNER & REMINDERS MADE SIMPLE\\r\\nDesigned to keep you on top of your to do list, tasks and calendar events with no hassle. With intuitive drag and drop of tasks, swiping to mark to-do's as complete, and shaking your device to remove completed from your to do list - you can stay organized and enjoy every minute of it.\\r\\n\\r\\nPOWERFUL TO DO LIST TASK MANAGEMENT\\r\\nAdd a to do list item straight from your email / Gmail / Outlook inbox by forwarding do@Any.do. Attach files from your computer, Dropbox, or Google Drive to your to- tasks.\\r\\n\\r\\nDAILY PLANNER & LIFE ORGANIZER\\r\\nAny.do is a to do list, a calendar, an inbox, a notepad, a checklist, task list, a board for post its or sticky notes, a task & project management tool, a reminder app, a daily planner, a family organizer, an agenda, a bill planner and overall the simplest productivity tool you will ever have. \\r\\n\\r\\nSHARE LISTS, ASSIGN & ORGANIZE TASKS\\r\\nTo plan & organize projects has never been easier. Now you can share lists between family members, assign tasks to each other, chat and much more. Any.do will help you and the people around you stay in-sync and get reminders so that you can focus on what matters, knowing you had a productive day and crossed off your to do list.\\r\\n\\r\\nGROCERY LIST & SHOPPING LIST\\r\\nAny.do task list, calendar, agenda, reminders & planner is also great for shopping lists at the grocery store. Simply create a list on Any.do, share it with your loved ones and see them adding their shopping items in real-time.\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"descriptionHTML\"\u001b[39;49;00m: \u001b[33m\"<b>\\ud83c\\udfc6 Editor&#39;s Choice by Google</b><br><br>Any.do is a To Do List, Calendar, Planner, Tasks &amp; Reminders App That Helps Over 25M People Stay Organized and Get More Done.<br><br><b>\\ud83e\\udd47 &quot;It\\u2019s A MUST HAVE PLANNER &amp; TO DO LIST APP&quot; (NYTimes, USA TODAY, WSJ &amp; Lifehacker).</b><br><br>Any.do is a free to-do list, planner &amp; calendar app for managing and organizing your daily tasks, to-do lists, notes, reminders, checklists, calendar events, grocery lists and more.<br><br>\\ud83d\\udcc5 Organize Your Tasks &amp; To-Do List in Seconds<br><br>\\u2022 ADVANCED CALENDAR &amp; DAILY PLANNER - Keep your to-do list and calendar events always at hand with our calendar widget. Any.do to-do list &amp; planner support daily calendar view, 3-day Calendar view, Weekly calendar view &amp; agenda view, with built-in reminders. Review and organize your calendar events and to do list side by side.<br><br>\\u2022 SYNCS SEAMLESSLY - Keeps all your to do list, tasks, reminders, notes, calendar &amp; agenda always in sync so you\\u2019ll never forget a thing. Sync your phone\\u2019s calendar, google calendar, Facebook events, outlook calendar or any other calendar so you don\\u2019t forget an important event.<br><br>\\u2022 SET REMINDERS - One time reminders, recurring reminders, Location reminders &amp; voice reminders. NEW! Easily create tasks and get reminders in WhatsApp.<br><br>\\u2022 WORK TOGETHER - Share your to do list and assign tasks with your friends, family &amp; colleagues from your task list to collaborate and get more done. <br><br>---<br><br>ALL-IN-ONE PLANNER &amp; CALENDAR APP FOR GETTING THINGS DONE<br>Create and set reminders with voice to your to do list. <br>For better task management flow we added a calendar integration to keep your agenda always up to date. <br>For better productivity, we added recurring reminders, location reminders, one-time reminder, sub-tasks, notes &amp; file attachments. <br>To keep your to do list up to date, we\\u2019ve added a daily planner and focus mode.<br><br>INTEGRATIONS<br>Any.do To do list, Calendar, planner &amp; Reminders Integrates with Google Calendar, Outlook, WhatsApp, Slack, Gmail, Google Tasks, Evernote, Trello, Wunderlist, Todoist, Zapier, Asana, Microsoft to-do, Salesforce, OneNote, Google Assistant, Amazon Alexa, Office 365, Exchange, Jira &amp; More.<br><br>TO DO LIST, CALENDAR, PLANNER &amp; REMINDERS MADE SIMPLE<br>Designed to keep you on top of your to do list, tasks and calendar events with no hassle. With intuitive drag and drop of tasks, swiping to mark to-do&#39;s as complete, and shaking your device to remove completed from your to do list - you can stay organized and enjoy every minute of it.<br><br>POWERFUL TO DO LIST TASK MANAGEMENT<br>Add a to do list item straight from your email / Gmail / Outlook inbox by forwarding do@Any.do. Attach files from your computer, Dropbox, or Google Drive to your to- tasks.<br><br>DAILY PLANNER &amp; LIFE ORGANIZER<br>Any.do is a to do list, a calendar, an inbox, a notepad, a checklist, task list, a board for post its or sticky notes, a task &amp; project management tool, a reminder app, a daily planner, a family organizer, an agenda, a bill planner and overall the simplest productivity tool you will ever have. <br><br>SHARE LISTS, ASSIGN &amp; ORGANIZE TASKS<br>To plan &amp; organize projects has never been easier. Now you can share lists between family members, assign tasks to each other, chat and much more. Any.do will help you and the people around you stay in-sync and get reminders so that you can focus on what matters, knowing you had a productive day and crossed off your to do list.<br><br>GROCERY LIST &amp; SHOPPING LIST<br>Any.do task list, calendar, agenda, reminders &amp; planner is also great for shopping lists at the grocery store. Simply create a list on Any.do, share it with your loved ones and see them adding their shopping items in real-time.\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"developer\"\u001b[39;49;00m: \u001b[33m\"Any.do Calendar & To-Do List\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"developerAddress\"\u001b[39;49;00m: \u001b[33m\"Any.do Inc.\\n\\n6  Agripas Street, Tel Aviv\\n6249106 ISRAEL\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"developerEmail\"\u001b[39;49;00m: \u001b[33m\"feedback+androidtodo@any.do\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"developerId\"\u001b[39;49;00m: \u001b[33m\"5304780265295461149\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"developerInternalID\"\u001b[39;49;00m: \u001b[33m\"5304780265295461149\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"developerWebsite\"\u001b[39;49;00m: \u001b[33m\"https://www.any.do\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"free\"\u001b[39;49;00m: \u001b[34mtrue\u001b[39;49;00m,\n",
      "  \u001b[94m\"genre\"\u001b[39;49;00m: \u001b[33m\"Productivity\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"genreId\"\u001b[39;49;00m: \u001b[33m\"PRODUCTIVITY\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"headerImage\"\u001b[39;49;00m: \u001b[33m\"https://lh3.googleusercontent.com/dZknnlk1LM8fYS3wjOvVHOmWKOGH1HAe691Yuh7LAeBj6a730A1CQqZnXxjNahAYUFFw\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"histogram\"\u001b[39;49;00m: [\n",
      "    \u001b[34m27745\u001b[39;49;00m,\n",
      "    \u001b[34m9356\u001b[39;49;00m,\n",
      "    \u001b[34m13990\u001b[39;49;00m,\n",
      "    \u001b[34m30068\u001b[39;49;00m,\n",
      "    \u001b[34m263021\u001b[39;49;00m\n",
      "  ],\n",
      "  \u001b[94m\"icon\"\u001b[39;49;00m: \u001b[33m\"https://lh3.googleusercontent.com/zgOLUXCHkF91H8xuMTMLT17smwgLPwSBjUlKVWF-cZRFjlv-Uvtman7DiHEii54fbEE\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"installs\"\u001b[39;49;00m: \u001b[33m\"10,000,000+\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"minInstalls\"\u001b[39;49;00m: \u001b[34m10000000\u001b[39;49;00m,\n",
      "  \u001b[94m\"offersIAP\"\u001b[39;49;00m: \u001b[34mtrue\u001b[39;49;00m,\n",
      "  \u001b[94m\"price\"\u001b[39;49;00m: \u001b[34m0\u001b[39;49;00m,\n",
      "  \u001b[94m\"privacyPolicy\"\u001b[39;49;00m: \u001b[33m\"https://www.any.do/privacy\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"ratings\"\u001b[39;49;00m: \u001b[34m344182\u001b[39;49;00m,\n",
      "  \u001b[94m\"recentChanges\"\u001b[39;49;00m: \u001b[33m\"Faster and smoother for better user experience!\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"recentChangesHTML\"\u001b[39;49;00m: \u001b[33m\"Faster and smoother for better user experience!\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"released\"\u001b[39;49;00m: \u001b[33m\"Nov 10, 2011\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"reviews\"\u001b[39;49;00m: \u001b[34m122500\u001b[39;49;00m,\n",
      "  \u001b[94m\"score\"\u001b[39;49;00m: \u001b[34m4.427334\u001b[39;49;00m,\n",
      "  \u001b[94m\"screenshots\"\u001b[39;49;00m: [\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/VDIbHpEFPRDikEOFXnZFwuMi43EYChzO-YNdMjYalFChBnvJlwBAqIKqbuWZsCNBpg\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/f5OJsEyavJ-PKeoCx1wD3fcUKHyBtyIVyg4WufaNx9GS-2lkhOCySFCacNdLO31k9BE\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/cH4gwsqvMNp_VkVM_yrcfzbsPmAb8NEDAJpBEHOFvOuY_esaCEaN5e3EMiwKHNnMfZDD\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/dJjLoehztK13hzUgNxxgk4-i5tOwaZyxAETnMW9tRfMd7j5__BHawx8ooxyw1yczgFc\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/gQdicuWBaVrQVhhJTT3cVy5V8TqvzT8C0gCKVGaUSk2T_cBL4dLipm4WBvAOigaa3hg\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/RJWUuf4ZzEyC0ujZsQyYbJxENkwtj5iyTp3mgPgqQSFhjAuUS-J0LXTUlHb_akVHP98\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/URRiwpP8lp1ZHvFWBxw4azWhD9oT8plJEmEYLFNdfnrOOyYjEfCZmXT793Cw23C-AQ\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/QVI0FjzMYdCF-QHySMsg81Ycr1pAwMP8vaqB-7fBCuF3n0ba2cri5MmQ4AUaC-R5EQ\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/Zb20ffh_Qhx1w1F3p_DItDQIGoCKv3HhOzux9XS3WWaLvBXEJfohXtqvw1SSBye3ng\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/RfAJwvQ3irLGtJZ9TUB5vJFve6gNU5DbNoUXXLO7lkcMRuNflnvJ7Y0gkSBk73EW56A\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/QOq7txDjRrcejlt1ygComlUlwO1MFMvoYa6agD7FuHWXMwiN5Zi9EXplia5ZB5md-Wg\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/X-VVVquWXnXaudqjRae6n9ggtqkE8o7e8S4fowih0Inw0nvYmpbAKDMtvgTp3B_T-w\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/MLbtJlgRsxk_EimwHqDBBOdjtxVW8yeJGNtI1AlezfE7x-J-M0asXjYhRpDhivaor8M\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/gbOH7ZFWr9GjRRC_OMoZA_-CwP75O7dl0pWiTaniPeaKlA8XiqUoKCZZ6P8QbOuOQvM\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/wMQSvw6ztw2kBtPnxicqHEiij40OPHG1Z0lSPGg0quaVJZbD2dmjCWD4JS0M2b9sAVE\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/_tU_tw29je3QXi8H1Pe6j8hU0s5tfrr4jW4eoZA3lITzHKwPiYPr4c5YbH5QksYSl8Q\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/9CRfpxsFloacjtItE8shJ3E9VY_GZvw0e_WKqKk68t4MZKlT3dLQoP4x30AIQOzqgkUg\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/ya4Zu4Dym4dPxAjA2sLIwvyAYnvsRrT4ZmtTF0DMVeLIY3wyVJOTKKC_yqhOmoT4LQsB\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/zPp-qbajG3CiYxGsCuQdYseVFTcZADiGSusW_JPzRy0Khd9bOCay-9mTo-GA3yJMWg\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/kDhjPpW5gFZguG9vy4zpx5Iu4rWqLlaKFBk9ny7OX28tZ62rXitCyh-zGbBtKewMDA\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/H_JGmZ1aF2WvtmywgmAHOQouMfigvep8iw7b3l6zzX7bAH2w_AggKIY2-Gml4b87tQ\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/GPp18BeNaxqK3I_4_W431v9AgIhzmP0Y9pMy-RQdr0ZBoRi0pI9L3XpeS8eLChTYyw\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/NUFVywNmd0AACASsHdWmAZ11gll5wyKyT_a21vSzH_onqYEnmU54pVTZ-k5cAupnMQ\"\u001b[39;49;00m,\n",
      "    \u001b[33m\"https://lh3.googleusercontent.com/dwrByRBYhI5MLo6mR1pGOOB2RWt1-oWFPYVAa7vTAjsrfCJ2e23Acx9WNdrfaoDMUA\"\u001b[39;49;00m\n",
      "  ],\n",
      "  \u001b[94m\"size\"\u001b[39;49;00m: \u001b[33m\"Varies with device\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"summary\"\u001b[39;49;00m: \u001b[33m\"Task Manager \\u2705 Organizer \\ud83d\\udcc5 Agenda \\ud83d\\udcdd Daily Reminders \\ud83d\\udd14 All-in-One Simple App.\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"summaryHTML\"\u001b[39;49;00m: \u001b[33m\"Task Manager \\u2705 Organizer \\ud83d\\udcc5 Agenda \\ud83d\\udcdd Daily Reminders \\ud83d\\udd14 All-in-One Simple App.\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"title\"\u001b[39;49;00m: \u001b[33m\"Any.do: To do list, Calendar, Planner & Reminders\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"updated\"\u001b[39;49;00m: \u001b[34m1587284863\u001b[39;49;00m,\n",
      "  \u001b[94m\"url\"\u001b[39;49;00m: \u001b[33m\"https://play.google.com/store/apps/details?id=com.anydo&hl=en&gl=us\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"version\"\u001b[39;49;00m: \u001b[33m\"Varies with device\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"video\"\u001b[39;49;00m: \u001b[33m\"https://www.youtube.com/embed/2nkllLD0x6o?ps=play&vq=large&rel=0&autohide=1&showinfo=0\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"videoImage\"\u001b[39;49;00m: \u001b[33m\"https://i.ytimg.com/vi/2nkllLD0x6o/hqdefault.jpg\"\u001b[39;49;00m\n",
      "}\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Example\n",
    "print_json(app_infos[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Convert the JSON objects to dataframe and to CSV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "app_infos_df = pd.DataFrame(app_infos)\n",
    "app_infos_df.to_csv('apps.csv', index=None, header=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Get App Reviews"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 15/15 [02:13<00:00,  8.87s/it]\n"
     ]
    }
   ],
   "source": [
    "# Code from https://github.com/JoMingyu/google-play-scraper\n",
    "# Filter the review score.\n",
    "# Then sort the reviews by their helpfulness\n",
    "# Just in case, get a subset from the newest, too:\n",
    "app_reviews = []\n",
    "\n",
    "for ap in tqdm(app_packages):\n",
    "    for score in list(range(1, 6)):\n",
    "        for sort_order in [Sort.MOST_RELEVANT, Sort.NEWEST]:\n",
    "            rvs, _ = reviews(\n",
    "                ap,\n",
    "                lang='en', # defaults to 'en'\n",
    "                country='us',# defaults to 'us'\n",
    "                sort=sort_order,\n",
    "                count= 200 if score == 3 else 100,\n",
    "                filter_score_with=score # defaults to None(means all score)\n",
    "              )\n",
    "            for r in rvs:\n",
    "                r['sortOrder'] = 'most_relevant' if sort_order == Sort.MOST_RELEVANT else 'newest'\n",
    "                r['appId'] = ap\n",
    "            app_reviews.extend(rvs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\n",
      "  \u001b[94m\"appId\"\u001b[39;49;00m: \u001b[33m\"com.anydo\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"at\"\u001b[39;49;00m: \u001b[33m\"2020-04-19 13:16:32\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"content\"\u001b[39;49;00m: \u001b[33m\"This application fails a lot and moreover it is complicated,weird and not intuitive. I don't renew my subscription.\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"repliedAt\"\u001b[39;49;00m: \u001b[34mnull\u001b[39;49;00m,\n",
      "  \u001b[94m\"replyContent\"\u001b[39;49;00m: \u001b[34mnull\u001b[39;49;00m,\n",
      "  \u001b[94m\"reviewCreatedVersion\"\u001b[39;49;00m: \u001b[33m\"4.19.0.6\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"score\"\u001b[39;49;00m: \u001b[34m1\u001b[39;49;00m,\n",
      "  \u001b[94m\"sortOrder\"\u001b[39;49;00m: \u001b[33m\"most_relevant\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"thumbsUpCount\"\u001b[39;49;00m: \u001b[34m1\u001b[39;49;00m,\n",
      "  \u001b[94m\"userImage\"\u001b[39;49;00m: \u001b[33m\"https://lh3.googleusercontent.com/a-/AOh14Gg1TnyLdy95wr0sxvIJoYknH7bWwzgA7RGmVXXSkFQ\"\u001b[39;49;00m,\n",
      "  \u001b[94m\"userName\"\u001b[39;49;00m: \u001b[33m\"Jo\\u00e3o Manuel\"\u001b[39;49;00m\n",
      "}\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Example\n",
    "print_json(app_reviews[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "15792"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(app_reviews)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Convert the JSON objects to dataframe and to CSV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "app_reviews_df = pd.DataFrame(app_reviews)\n",
    "app_reviews_df.to_csv('reviews.csv', index=None, header=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Sentiment Analysis with BERT and Transformers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "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>userName</th>\n",
       "      <th>userImage</th>\n",
       "      <th>content</th>\n",
       "      <th>score</th>\n",
       "      <th>thumbsUpCount</th>\n",
       "      <th>reviewCreatedVersion</th>\n",
       "      <th>at</th>\n",
       "      <th>replyContent</th>\n",
       "      <th>repliedAt</th>\n",
       "      <th>sortOrder</th>\n",
       "      <th>appId</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>João Manuel</td>\n",
       "      <td>https://lh3.googleusercontent.com/a-/AOh14Gg1T...</td>\n",
       "      <td>This application fails a lot and moreover it i...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4.19.0.6</td>\n",
       "      <td>2020-04-19 13:16:32</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>most_relevant</td>\n",
       "      <td>com.anydo</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Anshul Rawal</td>\n",
       "      <td>https://lh3.googleusercontent.com/a-/AOh14GgH3...</td>\n",
       "      <td>New update issue. In widget, when I tap to add...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4.19.0.6</td>\n",
       "      <td>2020-04-21 07:20:06</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>most_relevant</td>\n",
       "      <td>com.anydo</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Jason Thornton</td>\n",
       "      <td>https://lh3.googleusercontent.com/-B4TxqW_x-Jo...</td>\n",
       "      <td>This could be an awesome app. It has great pot...</td>\n",
       "      <td>1</td>\n",
       "      <td>68</td>\n",
       "      <td>4.16.6.2</td>\n",
       "      <td>2020-02-07 11:01:02</td>\n",
       "      <td>Hi Jason, as far as we can see in your latest ...</td>\n",
       "      <td>2020-02-09 04:37:41</td>\n",
       "      <td>most_relevant</td>\n",
       "      <td>com.anydo</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Lars Panzerbjørn</td>\n",
       "      <td>https://lh3.googleusercontent.com/a-/AOh14Gg-h...</td>\n",
       "      <td>It seems OK, but very basic. Recurring tasks n...</td>\n",
       "      <td>1</td>\n",
       "      <td>195</td>\n",
       "      <td>4.17.0.2</td>\n",
       "      <td>2020-03-12 01:17:34</td>\n",
       "      <td>We do offer this option as part of the Advance...</td>\n",
       "      <td>2020-03-14 23:20:13</td>\n",
       "      <td>most_relevant</td>\n",
       "      <td>com.anydo</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Scott Prewitt</td>\n",
       "      <td>https://lh3.googleusercontent.com/-K-X1-YsVd6U...</td>\n",
       "      <td>Absolutely worthless. This app runs a prohibit...</td>\n",
       "      <td>1</td>\n",
       "      <td>51</td>\n",
       "      <td>4.17.0.2</td>\n",
       "      <td>2020-03-14 10:41:01</td>\n",
       "      <td>We're sorry you feel this way! 90% of the app ...</td>\n",
       "      <td>2020-03-15 16:45:51</td>\n",
       "      <td>most_relevant</td>\n",
       "      <td>com.anydo</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           userName                                          userImage  \\\n",
       "0       João Manuel  https://lh3.googleusercontent.com/a-/AOh14Gg1T...   \n",
       "1      Anshul Rawal  https://lh3.googleusercontent.com/a-/AOh14GgH3...   \n",
       "2    Jason Thornton  https://lh3.googleusercontent.com/-B4TxqW_x-Jo...   \n",
       "3  Lars Panzerbjørn  https://lh3.googleusercontent.com/a-/AOh14Gg-h...   \n",
       "4     Scott Prewitt  https://lh3.googleusercontent.com/-K-X1-YsVd6U...   \n",
       "\n",
       "                                             content  score  thumbsUpCount  \\\n",
       "0  This application fails a lot and moreover it i...      1              1   \n",
       "1  New update issue. In widget, when I tap to add...      1              0   \n",
       "2  This could be an awesome app. It has great pot...      1             68   \n",
       "3  It seems OK, but very basic. Recurring tasks n...      1            195   \n",
       "4  Absolutely worthless. This app runs a prohibit...      1             51   \n",
       "\n",
       "  reviewCreatedVersion                   at  \\\n",
       "0             4.19.0.6  2020-04-19 13:16:32   \n",
       "1             4.19.0.6  2020-04-21 07:20:06   \n",
       "2             4.16.6.2  2020-02-07 11:01:02   \n",
       "3             4.17.0.2  2020-03-12 01:17:34   \n",
       "4             4.17.0.2  2020-03-14 10:41:01   \n",
       "\n",
       "                                        replyContent            repliedAt  \\\n",
       "0                                                NaN                  NaN   \n",
       "1                                                NaN                  NaN   \n",
       "2  Hi Jason, as far as we can see in your latest ...  2020-02-09 04:37:41   \n",
       "3  We do offer this option as part of the Advance...  2020-03-14 23:20:13   \n",
       "4  We're sorry you feel this way! 90% of the app ...  2020-03-15 16:45:51   \n",
       "\n",
       "       sortOrder      appId  \n",
       "0  most_relevant  com.anydo  \n",
       "1  most_relevant  com.anydo  \n",
       "2  most_relevant  com.anydo  \n",
       "3  most_relevant  com.anydo  \n",
       "4  most_relevant  com.anydo  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv(\"reviews.csv\")\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 15792 entries, 0 to 15791\n",
      "Data columns (total 11 columns):\n",
      "userName                15792 non-null object\n",
      "userImage               15792 non-null object\n",
      "content                 15792 non-null object\n",
      "score                   15792 non-null int64\n",
      "thumbsUpCount           15792 non-null int64\n",
      "reviewCreatedVersion    13572 non-null object\n",
      "at                      15792 non-null object\n",
      "replyContent            7538 non-null object\n",
      "repliedAt               7538 non-null object\n",
      "sortOrder               15792 non-null object\n",
      "appId                   15792 non-null object\n",
      "dtypes: int64(2), object(9)\n",
      "memory usage: 1.3+ MB\n"
     ]
    }
   ],
   "source": [
    "# Check any missing value in review column ('content' here)\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1d499308f88>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 490,
       "width": 742
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.countplot(df.score)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Pre-processing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Convert the dataset into negative, neutral and positive sentiment to make it balacned:\n",
    "# Negative as 0\n",
    "# Neutral as 1\n",
    "# Positive as 2\n",
    "def to_sentiment(rating):\n",
    "    rating = int(rating)\n",
    "    if rating <= 2:\n",
    "        return 0\n",
    "    elif rating == 3:\n",
    "        return 1\n",
    "    else:\n",
    "        return 2\n",
    "\n",
    "df['sentiment'] = df.score.apply(to_sentiment)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 490,
       "width": 742
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "class_names = ['negative', 'neutral', 'positive']\n",
    "\n",
    "ax = sns.countplot(df.sentiment)\n",
    "plt.xlabel('review sentiment')\n",
    "ax.set_xticklabels(class_names);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\") # GPU running"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The cased version works better. Intuitively, that makes sense, since “BAD” might convey more sentiment than “bad”.\n",
    "PRE_TRAINED_MODEL_NAME = 'bert-base-cased'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "I0424 14:12:48.610908 14500 tokenization_utils.py:398] loading file https://s3.amazonaws.com/models.huggingface.co/bert/bert-base-cased-vocab.txt from cache at C:\\Users\\bokhy\\.cache\\torch\\transformers\\5e8a2b4893d13790ed4150ca1906be5f7a03d6c4ddf62296c383f6db42814db2.e13dbb970cb325137104fb2e5f36fe865f27746c6b526f6352861b1980eb80b1\n"
     ]
    }
   ],
   "source": [
    "tokenizer = BertTokenizer.from_pretrained(PRE_TRAINED_MODEL_NAME)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Check the length of the token to optimize\n",
    "token_lens = []\n",
    "\n",
    "for txt in df.content:\n",
    "    tokens = tokenizer.encode(txt, max_length=512) # set the max length for now (don't know the lengh of the review)\n",
    "    token_lens.append(len(tokens))\n",
    "\n",
    "# Most of the reviews seem to contain less than 128 tokens, but we’ll be on the safe side and choose a maximum length of 200.\n",
    "sns.distplot(token_lens)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "MAX_LEN = 200"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Create Pytorch Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "class GPReviewDataset(Dataset):\n",
    "    def __init__(self, reviews, targets, tokenizer, max_len):\n",
    "        self.reviews = reviews\n",
    "        self.targets = targets\n",
    "        self.tokenizer = tokenizer\n",
    "        self.max_len = max_len\n",
    "        \n",
    "    def __len__(self):\n",
    "        return len(self.reviews)\n",
    "    \n",
    "    def __getitem__(self, item):\n",
    "        review = str(self.reviews[item])\n",
    "        target = self.targets[item]\n",
    "\n",
    "        encoding = self.tokenizer.encode_plus(\n",
    "          review,\n",
    "          add_special_tokens=True, # Add '[CLS]' and '[SEP]'\n",
    "          max_length=self.max_len,\n",
    "          return_token_type_ids=True, #False? \n",
    "          pad_to_max_length=True,\n",
    "          return_attention_mask=True,\n",
    "          return_tensors='pt', # Return PyTorch tensors\n",
    "        )\n",
    "\n",
    "        return {\n",
    "          'review_text': review,\n",
    "          'input_ids': encoding['input_ids'].flatten(),\n",
    "          'attention_mask': encoding['attention_mask'].flatten(),\n",
    "          'targets': torch.tensor(target, dtype=torch.long)\n",
    "        }"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Train/Test/Val Split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train, df_test = train_test_split(\n",
    "  df,\n",
    "  test_size=0.1\n",
    ")\n",
    "df_val, df_test = train_test_split(\n",
    "  df_test,\n",
    "  test_size=0.5\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((14212, 12), (790, 12), (790, 12))"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.shape, df_val.shape, df_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create DataLoader\n",
    "def create_data_loader(df, tokenizer, max_len, batch_size):\n",
    "    ds = GPReviewDataset(\n",
    "    reviews=df.content.to_numpy(),\n",
    "    targets=df.sentiment.to_numpy(),\n",
    "    tokenizer=tokenizer,\n",
    "    max_len=max_len\n",
    "  )\n",
    "\n",
    "    return torch.utils.data.DataLoader(\n",
    "    ds,\n",
    "    batch_size=batch_size,\n",
    "    num_workers=0 # In Windows, it should be 0 (no multi-processing)\n",
    "  )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create data Loader\n",
    "BATCH_SIZE = 4 # or 16 (if enough memory)\n",
    "\n",
    "train_data_loader = create_data_loader(df_train, tokenizer, MAX_LEN, BATCH_SIZE)\n",
    "val_data_loader = create_data_loader(df_val, tokenizer, MAX_LEN, BATCH_SIZE)\n",
    "test_data_loader = create_data_loader(df_test, tokenizer, MAX_LEN, BATCH_SIZE)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Use base BertModel and build our sentiment classifier on top of it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "I0424 14:13:02.420681 14500 configuration_utils.py:185] loading configuration file https://s3.amazonaws.com/models.huggingface.co/bert/bert-base-cased-config.json from cache at C:\\Users\\bokhy\\.cache\\torch\\transformers\\b945b69218e98b3e2c95acf911789741307dec43c698d35fad11c1ae28bda352.9da767be51e1327499df13488672789394e2ca38b877837e52618a67d7002391\n",
      "I0424 14:13:02.422730 14500 configuration_utils.py:199] Model config {\n",
      "  \"architectures\": [\n",
      "    \"BertForMaskedLM\"\n",
      "  ],\n",
      "  \"attention_probs_dropout_prob\": 0.1,\n",
      "  \"finetuning_task\": null,\n",
      "  \"hidden_act\": \"gelu\",\n",
      "  \"hidden_dropout_prob\": 0.1,\n",
      "  \"hidden_size\": 768,\n",
      "  \"id2label\": {\n",
      "    \"0\": \"LABEL_0\",\n",
      "    \"1\": \"LABEL_1\"\n",
      "  },\n",
      "  \"initializer_range\": 0.02,\n",
      "  \"intermediate_size\": 3072,\n",
      "  \"is_decoder\": false,\n",
      "  \"label2id\": {\n",
      "    \"LABEL_0\": 0,\n",
      "    \"LABEL_1\": 1\n",
      "  },\n",
      "  \"layer_norm_eps\": 1e-12,\n",
      "  \"max_position_embeddings\": 512,\n",
      "  \"model_type\": \"bert\",\n",
      "  \"num_attention_heads\": 12,\n",
      "  \"num_hidden_layers\": 12,\n",
      "  \"num_labels\": 2,\n",
      "  \"output_attentions\": false,\n",
      "  \"output_hidden_states\": false,\n",
      "  \"output_past\": true,\n",
      "  \"pad_token_id\": 0,\n",
      "  \"pruned_heads\": {},\n",
      "  \"torchscript\": false,\n",
      "  \"type_vocab_size\": 2,\n",
      "  \"use_bfloat16\": false,\n",
      "  \"vocab_size\": 28996\n",
      "}\n",
      "\n",
      "I0424 14:13:02.791183 14500 modeling_utils.py:406] loading weights file https://s3.amazonaws.com/models.huggingface.co/bert/bert-base-cased-pytorch_model.bin from cache at C:\\Users\\bokhy\\.cache\\torch\\transformers\\35d8b9d36faaf46728a0192d82bf7d00137490cd6074e8500778afed552a67e5.3fadbea36527ae472139fe84cddaa65454d7429f12d543d80bfc3ad70de55ac2\n"
     ]
    }
   ],
   "source": [
    "# Load the base BertModel\n",
    "bert_model = BertModel.from_pretrained(PRE_TRAINED_MODEL_NAME)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Sentiment Classifer on top of Bert Model\n",
    "# We use a dropout layer for some regularization and a fully-connected layer for our output\n",
    "class SentimentClassifier(nn.Module):\n",
    "    \n",
    "    def __init__(self, n_classes):\n",
    "        super(SentimentClassifier, self).__init__()\n",
    "        self.bert = BertModel.from_pretrained(PRE_TRAINED_MODEL_NAME)\n",
    "        self.drop = nn.Dropout(p=0.2)\n",
    "        self.out = nn.Linear(self.bert.config.hidden_size, n_classes)\n",
    "    \n",
    "    def forward(self, input_ids, attention_mask):\n",
    "        _, pooled_output = self.bert(                #pooled_output is a summary of the content\n",
    "            input_ids=input_ids,\n",
    "            attention_mask=attention_mask)\n",
    "        \n",
    "        output = self.drop(pooled_output)\n",
    "        \n",
    "        return self.out(output) # Note that we’re returning the raw output of the last layer since that is required for the cross-entropy loss function in PyTorch to work."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "I0424 14:14:45.014093 14500 configuration_utils.py:185] loading configuration file https://s3.amazonaws.com/models.huggingface.co/bert/bert-base-cased-config.json from cache at C:\\Users\\bokhy\\.cache\\torch\\transformers\\b945b69218e98b3e2c95acf911789741307dec43c698d35fad11c1ae28bda352.9da767be51e1327499df13488672789394e2ca38b877837e52618a67d7002391\n",
      "I0424 14:14:45.017090 14500 configuration_utils.py:199] Model config {\n",
      "  \"architectures\": [\n",
      "    \"BertForMaskedLM\"\n",
      "  ],\n",
      "  \"attention_probs_dropout_prob\": 0.1,\n",
      "  \"finetuning_task\": null,\n",
      "  \"hidden_act\": \"gelu\",\n",
      "  \"hidden_dropout_prob\": 0.1,\n",
      "  \"hidden_size\": 768,\n",
      "  \"id2label\": {\n",
      "    \"0\": \"LABEL_0\",\n",
      "    \"1\": \"LABEL_1\"\n",
      "  },\n",
      "  \"initializer_range\": 0.02,\n",
      "  \"intermediate_size\": 3072,\n",
      "  \"is_decoder\": false,\n",
      "  \"label2id\": {\n",
      "    \"LABEL_0\": 0,\n",
      "    \"LABEL_1\": 1\n",
      "  },\n",
      "  \"layer_norm_eps\": 1e-12,\n",
      "  \"max_position_embeddings\": 512,\n",
      "  \"model_type\": \"bert\",\n",
      "  \"num_attention_heads\": 12,\n",
      "  \"num_hidden_layers\": 12,\n",
      "  \"num_labels\": 2,\n",
      "  \"output_attentions\": false,\n",
      "  \"output_hidden_states\": false,\n",
      "  \"output_past\": true,\n",
      "  \"pad_token_id\": 0,\n",
      "  \"pruned_heads\": {},\n",
      "  \"torchscript\": false,\n",
      "  \"type_vocab_size\": 2,\n",
      "  \"use_bfloat16\": false,\n",
      "  \"vocab_size\": 28996\n",
      "}\n",
      "\n",
      "I0424 14:14:45.414847 14500 modeling_utils.py:406] loading weights file https://s3.amazonaws.com/models.huggingface.co/bert/bert-base-cased-pytorch_model.bin from cache at C:\\Users\\bokhy\\.cache\\torch\\transformers\\35d8b9d36faaf46728a0192d82bf7d00137490cd6074e8500778afed552a67e5.3fadbea36527ae472139fe84cddaa65454d7429f12d543d80bfc3ad70de55ac2\n"
     ]
    }
   ],
   "source": [
    "# Create an instance on GPU \n",
    "model = SentimentClassifier(len(class_names))\n",
    "model = model.to(device)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# BERT has some recommendations for fine-tuning:\n",
    "\n",
    "## Batch size: 16, 32\n",
    "## Learning rate (Adam): 5e-5, 3e-5, 2e-5\n",
    "## Number of epochs: 2, 3, 4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "EPOCHS = 3\n",
    "\n",
    "optimizer = AdamW(model.parameters(), lr = 2e-5, correct_bias = False)\n",
    "\n",
    "total_steps = len(train_data_loader) * EPOCHS\n",
    "\n",
    "scheduler = get_linear_schedule_with_warmup(\n",
    "  optimizer, \n",
    "  num_warmup_steps=0, # Linear scheduler with no warmup steps\n",
    "  num_training_steps=total_steps\n",
    ")\n",
    "\n",
    "loss_fn = nn.CrossEntropyLoss().to(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The scheduler gets called every time a batch is fed to the model. \n",
    "# We’re avoiding exploding gradients by clipping the gradients of the model using clipgrad_norm.\n",
    "def train_epoch(model,data_loader,loss_fn,optimizer,device,scheduler,n_examples):\n",
    "    model = model.train()\n",
    "    \n",
    "    losses = []\n",
    "    correct_predictions = 0\n",
    "\n",
    "    for d in data_loader:\n",
    "        input_ids = d[\"input_ids\"].to(device)\n",
    "        attention_mask = d[\"attention_mask\"].to(device)\n",
    "        targets = d[\"targets\"].to(device)\n",
    "\n",
    "        outputs = model(input_ids=input_ids,attention_mask=attention_mask)\n",
    "\n",
    "        _, preds = torch.max(outputs, dim=1)\n",
    "        loss = loss_fn(outputs, targets)\n",
    "\n",
    "        correct_predictions += torch.sum(preds == targets)\n",
    "        losses.append(loss.item())\n",
    "\n",
    "        loss.backward()\n",
    "        nn.utils.clip_grad_norm_(model.parameters(), max_norm=1.0)\n",
    "        optimizer.step()\n",
    "        scheduler.step()\n",
    "        optimizer.zero_grad()\n",
    "\n",
    "    return correct_predictions.double() / n_examples, np.mean(losses)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Evaluation of model on a given data loader"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "def eval_model(model, data_loader, loss_fn, device, n_examples):\n",
    "    model = model.eval()\n",
    "\n",
    "    losses = []\n",
    "    correct_predictions = 0\n",
    "\n",
    "    with torch.no_grad():\n",
    "        for d in data_loader:\n",
    "            input_ids = d[\"input_ids\"].to(device)\n",
    "            attention_mask = d[\"attention_mask\"].to(device)\n",
    "            targets = d[\"targets\"].to(device)\n",
    "\n",
    "            outputs = model(input_ids=input_ids,attention_mask=attention_mask)\n",
    "            _, preds = torch.max(outputs, dim=1)\n",
    "\n",
    "            loss = loss_fn(outputs, targets)\n",
    "\n",
    "            correct_predictions += torch.sum(preds == targets)\n",
    "            losses.append(loss.item())\n",
    "\n",
    "    return correct_predictions.double() / n_examples, np.mean(losses)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training Loop (Training + Evaluation Function)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/3\n",
      "----------\n",
      "Train loss 0.8248574028938145 accuracy 0.62278356318604\n",
      "Val   loss 0.7498909547322928 accuracy 0.718987341772152\n",
      "\n",
      "Epoch 2/3\n",
      "----------\n",
      "Train loss 0.6245774182465055 accuracy 0.8004503236701379\n",
      "Val   loss 0.9060691505520031 accuracy 0.789873417721519\n",
      "\n",
      "Epoch 3/3\n",
      "----------\n",
      "Train loss 0.4542284175153856 accuracy 0.8909372361384745\n",
      "Val   loss 0.889203398742459 accuracy 0.8253164556962026\n",
      "\n",
      "Wall time: 2h 34min 40s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "history = defaultdict(list)\n",
    "best_accuracy = 0\n",
    "\n",
    "for epoch in range(EPOCHS):\n",
    "\n",
    "    print(f'Epoch {epoch + 1}/{EPOCHS}')\n",
    "    print('-' * 10)\n",
    "\n",
    "    train_acc, train_loss = train_epoch(\n",
    "        model,\n",
    "        train_data_loader,\n",
    "        loss_fn,\n",
    "        optimizer,\n",
    "        device,\n",
    "        scheduler,\n",
    "        len(df_train)\n",
    "      )\n",
    "\n",
    "    print(f'Train loss {train_loss} accuracy {train_acc}')\n",
    "\n",
    "    val_acc, val_loss = eval_model(\n",
    "        model,\n",
    "        val_data_loader,\n",
    "        loss_fn,\n",
    "        device,\n",
    "        len(df_val)\n",
    "      )\n",
    "\n",
    "    print(f'Val   loss {val_loss} accuracy {val_acc}')\n",
    "    print()\n",
    "    # Saving history for plotting\n",
    "    history['train_acc'].append(train_acc)\n",
    "    history['train_loss'].append(train_loss)\n",
    "    history['val_acc'].append(val_acc)\n",
    "    history['val_loss'].append(val_loss)\n",
    "\n",
    "    # Storing the state of the best model that has the highest validation accuracy.\n",
    "    if val_acc > best_accuracy:\n",
    "        torch.save(model.state_dict(), 'best_model_state.bin')\n",
    "        best_accuracy = val_acc"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training vs Validation Accuracy:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 503,
       "width": 731
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history['train_acc'], label='train accuracy')\n",
    "plt.plot(history['val_acc'], label='validation accuracy')\n",
    "\n",
    "plt.title('Training history')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend()\n",
    "plt.ylim([0, 1]);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Loading pre-trained model(from all the steps above)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# model = SentimentClassifier(len(class_names))\n",
    "# model.load_state_dict(torch.load('best_model_state.bin'))\n",
    "# model = model.to(device)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Evaluating Test-set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8316455696202532"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_acc, _ = eval_model(\n",
    "  model,\n",
    "  test_data_loader,\n",
    "  loss_fn,\n",
    "  device,\n",
    "  len(df_test)\n",
    ")\n",
    "\n",
    "test_acc.item()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Similar to the evaluation function, except that we’re storing the text of the reviews and the predicted probabilities:\n",
    "def get_predictions(model, data_loader):\n",
    "    model = model.eval()\n",
    "\n",
    "    review_texts = []\n",
    "    predictions = []\n",
    "    prediction_probs = []\n",
    "    real_values = []\n",
    "\n",
    "    with torch.no_grad():\n",
    "        for d in data_loader:\n",
    "            texts = d[\"review_text\"]\n",
    "            input_ids = d[\"input_ids\"].to(device)\n",
    "            attention_mask = d[\"attention_mask\"].to(device)\n",
    "            targets = d[\"targets\"].to(device)\n",
    "\n",
    "            outputs = model(input_ids=input_ids,attention_mask=attention_mask)\n",
    "            _, preds = torch.max(outputs, dim=1)\n",
    "\n",
    "            review_texts.extend(texts)\n",
    "            predictions.extend(preds)\n",
    "            prediction_probs.extend(outputs)\n",
    "            real_values.extend(targets)\n",
    "\n",
    "    predictions = torch.stack(predictions).cpu()\n",
    "    prediction_probs = torch.stack(prediction_probs).cpu()\n",
    "    real_values = torch.stack(real_values).cpu()\n",
    "    return review_texts, predictions, prediction_probs, real_values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_review_texts, y_pred, y_pred_probs, y_test = get_predictions(model,test_data_loader)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "    negative       0.84      0.83      0.83       258\n",
      "     neutral       0.74      0.79      0.77       235\n",
      "    positive       0.90      0.86      0.88       297\n",
      "\n",
      "    accuracy                           0.83       790\n",
      "   macro avg       0.83      0.83      0.83       790\n",
      "weighted avg       0.83      0.83      0.83       790\n",
      "\n"
     ]
    }
   ],
   "source": [
    "#  classification report\n",
    "print(classification_report(y_test, y_pred, target_names=class_names))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predicting on Raw Text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "review_text = \"I love completing my todos! Best app ever!!!\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "encoded_review = tokenizer.encode_plus(\n",
    "  review_text,\n",
    "  max_length=MAX_LEN,\n",
    "  add_special_tokens=True,\n",
    "  return_token_type_ids=True, #False?\n",
    "  pad_to_max_length=True,\n",
    "  return_attention_mask=True,\n",
    "  return_tensors='pt',\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Review text: I love completing my todos! Best app ever!!!\n",
      "Sentiment  : positive\n"
     ]
    }
   ],
   "source": [
    "input_ids = encoded_review['input_ids'].to(device)\n",
    "attention_mask = encoded_review['attention_mask'].to(device)\n",
    "\n",
    "output = model(input_ids, attention_mask)\n",
    "_, prediction = torch.max(output, dim=1)\n",
    "\n",
    "print(f'Review text: {review_text}')\n",
    "print(f'Sentiment  : {class_names[prediction]}')"
   ]
  }
 ],
 "metadata": {
  "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}