NLP_Preprocessing.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "06c7e00f",
   "metadata": {},
   "source": [
    "# Preprocessing Using spaCy\n",
    "### Overview\n",
    "After previously cleaning the review text using AWS Glue, this notebook will complete the preprocessing by using spaCy to tokenize the text so that the data can be used for model training and analysis."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7ecd98b9",
   "metadata": {},
   "source": [
    "## Import libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "51b558e4",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import altair as alt\n",
    "import re\n",
    "import spacy\n",
    "from spacy import displacy\n",
    "import awswrangler as wr\n",
    "import warnings\n",
    "warnings.simplefilter(action='ignore', category=FutureWarning)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "26a79910",
   "metadata": {},
   "source": [
    "Verify installed version of spaCy."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "00c5bee4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'3.5.0'"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spacy.__version__"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2683acc9",
   "metadata": {},
   "source": [
    "## 1. Import data from S3\n",
    "Data will be imported from S3 which has been previously cleaned in AWS Glue. Prior cleaning involved removing whitespace, lowercasing the text, and removing early access reviews since these reviews provided very little information."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "92889f92",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = wr.s3.read_csv(path='s3://siads696-wi23-steam-data/write_down_dyf_to_s3/')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "36f91420",
   "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>app_id</th>\n",
       "      <th>app_name</th>\n",
       "      <th>review_text</th>\n",
       "      <th>review_score</th>\n",
       "      <th>review_votes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>this is not a game for people who want fast ac...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>oh man. where to even begin with this one. it ...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>my first game on a3 brought me the most horrif...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>i spent several years in the army and i gotta ...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>having accumulated over 13400 hours in-game wi...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   app_id app_name                                        review_text  \\\n",
       "0  107410   Arma 3  this is not a game for people who want fast ac...   \n",
       "1  107410   Arma 3  oh man. where to even begin with this one. it ...   \n",
       "2  107410   Arma 3  my first game on a3 brought me the most horrif...   \n",
       "3  107410   Arma 3  i spent several years in the army and i gotta ...   \n",
       "4  107410   Arma 3  having accumulated over 13400 hours in-game wi...   \n",
       "\n",
       "   review_score  review_votes  \n",
       "0             1             1  \n",
       "1             1             1  \n",
       "2             1             0  \n",
       "3             1             1  \n",
       "4             1             1  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "af9fd198",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(846712, 5)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "bc7e0444",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "app_id           int64\n",
       "app_name        object\n",
       "review_text     object\n",
       "review_score     int64\n",
       "review_votes     int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "326063aa",
   "metadata": {},
   "source": [
    "## 2. Preprocessing and Data Exploration\n",
    "### Check for missing values\n",
    "As part of the preprocessing done in AWS Glue we dropped records with missing text but as a sanity check we will check once more if there are any records that have a missing review."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "fa0b88cf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "app_id           0\n",
       "app_name         0\n",
       "review_text     16\n",
       "review_score     0\n",
       "review_votes     0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "03646ff3",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.dropna(inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "2b66bdc6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "app_id          0\n",
       "app_name        0\n",
       "review_text     0\n",
       "review_score    0\n",
       "review_votes    0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eee6b8c2",
   "metadata": {},
   "source": [
    "### 2.2 Clean text using ReGex\n",
    "Here we create a function that removes puncuation and only keep words."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "2abcf69f",
   "metadata": {},
   "outputs": [],
   "source": [
    "def cleaner(df):\n",
    "    \"\"\"\n",
    "        Applies regex to review_text column to clean text\n",
    "        \n",
    "        Input:\n",
    "            dataframe\n",
    "        \n",
    "        Returns:\n",
    "            dataframe with clean_text column\n",
    "    \"\"\"\n",
    "\n",
    "    pattern = re.compile(r\"\\b\\w*[']\\w*\\b|[a-zA-Z]+\")\n",
    "    \n",
    "    df['clean_text'] = df['review_text'].str.findall(pattern).str.join(' ')\n",
    "    \n",
    "    return df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ef83215e",
   "metadata": {},
   "source": [
    "Produce a new dataframe after removing punctuation and retaining only words for each review."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "d0ebd80e",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 18.2 s, sys: 318 ms, total: 18.5 s\n",
      "Wall time: 18.4 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "df_clean = cleaner(df)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fd9569ec",
   "metadata": {},
   "source": [
    "### 2.3 Drop rows with empty text after applying regex\n",
    "After using the regex to clean the text, we will now identify if there are any rows with empty text and drop them."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "29a8896b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Rows to drop: 10422\n"
     ]
    }
   ],
   "source": [
    "print('Rows to drop:', len(df_clean[df_clean['clean_text'] == '']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "624a9c4b",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_clean = df_clean[df_clean['clean_text'] != '']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "7afb0cbc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(836274, 6)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_clean.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5a3874cf",
   "metadata": {},
   "source": [
    "### 2.4 Check for duplicate reviews and drop\n",
    "Let's see if any reviews are duplicates and spot check the reviews."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "5ac8dcfb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "183821"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_clean[df_clean.duplicated(subset=['app_name', 'clean_text'])].sort_values(['app_name','clean_text']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "79f5b083",
   "metadata": {
    "scrolled": true
   },
   "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>app_id</th>\n",
       "      <th>app_name</th>\n",
       "      <th>review_text</th>\n",
       "      <th>review_score</th>\n",
       "      <th>review_votes</th>\n",
       "      <th>clean_text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>817414</th>\n",
       "      <td>230410</td>\n",
       "      <td>Warframe</td>\n",
       "      <td>you wanna know how good i think this game is? ...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>you wanna know how good i think this game is i...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>817425</th>\n",
       "      <td>230410</td>\n",
       "      <td>Warframe</td>\n",
       "      <td>you wanna know how good i think this game is? ...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>you wanna know how good i think this game is i...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>817426</th>\n",
       "      <td>230410</td>\n",
       "      <td>Warframe</td>\n",
       "      <td>you wanna know how good i think this game is? ...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>you wanna know how good i think this game is i...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        app_id  app_name                                        review_text  \\\n",
       "817414  230410  Warframe  you wanna know how good i think this game is? ...   \n",
       "817425  230410  Warframe  you wanna know how good i think this game is? ...   \n",
       "817426  230410  Warframe  you wanna know how good i think this game is? ...   \n",
       "\n",
       "        review_score  review_votes  \\\n",
       "817414             1             0   \n",
       "817425             1             0   \n",
       "817426             1             0   \n",
       "\n",
       "                                               clean_text  \n",
       "817414  you wanna know how good i think this game is i...  \n",
       "817425  you wanna know how good i think this game is i...  \n",
       "817426  you wanna know how good i think this game is i...  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_clean[(df_clean['app_name'] == 'Warframe') &  (df_clean['clean_text'].str.startswith('you wanna know how good i think'))]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "64c5ce3e",
   "metadata": {},
   "source": [
    "We can see that for the game Warframe there are 3 reviews that start with the text 'you wanna know how good i think'. Let's see if they are identical to each other."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "80a77c01",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "you wanna know how good i think this game is it's free to play and i spend money on it v\n",
      "you wanna know how good i think this game is it's free to play and i spend money on it v\n",
      "you wanna know how good i think this game is it's free to play and i spend money on it v\n"
     ]
    }
   ],
   "source": [
    "print(df_clean.loc[817414]['clean_text'])\n",
    "print(df_clean.loc[817425]['clean_text'])\n",
    "print(df_clean.loc[817426]['clean_text'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "81d44eb4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check for string match\n",
    "df_clean.loc[817414]['clean_text'] == df_clean.loc[817425]['clean_text'] == df_clean.loc[817426]['clean_text']"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b9adf993",
   "metadata": {},
   "source": [
    "The 3 reviews are an exact match and we are going to drop all duplicated reviews except for the first occurrence."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "200e76fe",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_clean = df_clean.drop_duplicates(subset=['app_name', 'clean_text'], keep='first')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "5ace7614",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(652453, 6)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_clean.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "033aecaa",
   "metadata": {},
   "source": [
    "### 2.4 Data exploration\n",
    "### Plot histogram of number of words per review\n",
    "The dataset currently sits at 651K reviews.  Before moving forward, let's get the length of words in each review and the distribution."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "95fe07c1",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_clean['num_words'] = df_clean['clean_text'].apply(lambda x: len(str(x).split()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "29118919",
   "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>app_id</th>\n",
       "      <th>app_name</th>\n",
       "      <th>review_text</th>\n",
       "      <th>review_score</th>\n",
       "      <th>review_votes</th>\n",
       "      <th>clean_text</th>\n",
       "      <th>num_words</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>this is not a game for people who want fast ac...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>this is not a game for people who want fast ac...</td>\n",
       "      <td>457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>oh man. where to even begin with this one. it ...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>oh man where to even begin with this one it wo...</td>\n",
       "      <td>674</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>my first game on a3 brought me the most horrif...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>my first game on a brought me the most horrifi...</td>\n",
       "      <td>198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>i spent several years in the army and i gotta ...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>i spent several years in the army and i gotta ...</td>\n",
       "      <td>179</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>having accumulated over 13400 hours in-game wi...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>having accumulated over hours in game with arm...</td>\n",
       "      <td>40</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   app_id app_name                                        review_text  \\\n",
       "0  107410   Arma 3  this is not a game for people who want fast ac...   \n",
       "1  107410   Arma 3  oh man. where to even begin with this one. it ...   \n",
       "2  107410   Arma 3  my first game on a3 brought me the most horrif...   \n",
       "3  107410   Arma 3  i spent several years in the army and i gotta ...   \n",
       "4  107410   Arma 3  having accumulated over 13400 hours in-game wi...   \n",
       "\n",
       "   review_score  review_votes  \\\n",
       "0             1             1   \n",
       "1             1             1   \n",
       "2             1             0   \n",
       "3             1             1   \n",
       "4             1             1   \n",
       "\n",
       "                                          clean_text  num_words  \n",
       "0  this is not a game for people who want fast ac...        457  \n",
       "1  oh man where to even begin with this one it wo...        674  \n",
       "2  my first game on a brought me the most horrifi...        198  \n",
       "3  i spent several years in the army and i gotta ...        179  \n",
       "4  having accumulated over hours in game with arm...         40  "
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_clean.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "47f8903e",
   "metadata": {},
   "source": [
    "Plot a histram of bins containing 25 words."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "383aab72",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlUAAAHHCAYAAACWQK1nAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy88F64QAAAACXBIWXMAAA9hAAAPYQGoP6dpAABPSElEQVR4nO3de1xVVf7/8fcBAa/n4A2QvKGWijdGFCQvaZJozEymTZhNmeMlDW2USc0ytdvo2JQ6eZvqmzR9tdSZsWa0MENRU9IkyTuZWVgKmgpHTUFh/f7ox/56BBVtG6Cv5+OxHw/2Wp+z99oLkLd777OPwxhjBAAAgJ/Fq6wHAAAAcCMgVAEAANiAUAUAAGADQhUAAIANCFUAAAA2IFQBAADYgFAFAABgA0IVAACADQhVAAAANiBUASgmJSVFDodD//znP8t6KKWSnZ2t++67T7Vr15bD4dCsWbPKekil1r17d3Xv3r2sh1HhPPLII2rcuHFZDwPwQKgCykhiYqIcDocqV66s77//vlh/9+7d1bp16zIYWcUzduxYrVq1ShMnTtTbb7+t3r17l/WQANyEKpX1AICbXV5enqZPn65XX321rIdSYa1Zs0b33HOPnnjiibIeCn4hr7/+ugoLC8t6GIAHzlQBZSwsLEyvv/66Dh06VNZD+cWdPn3alu0cOXJE/v7+tmzLbnYdY3lxtcdzvY7fx8dHfn5+12XbwLUiVAFl7KmnnlJBQYGmT59+2bpvvvlGDodDiYmJxfocDoemTp1qrU+dOlUOh0Nffvmlfv/738vlcqlu3bp65plnZIzRwYMHdc8998jpdCooKEgvv/xyifssKCjQU089paCgIFWrVk2//e1vdfDgwWJ1mzdvVu/eveVyuVS1alXdcccd2rhxo0dN0Zh2796tgQMHqmbNmurSpctlj/nrr7/W7373O9WqVUtVq1ZVp06dtHLlSqu/6BKqMUZz586Vw+GQw+G45Pbat2+vfv36ebS1adNGDodD27dvt9qWLFkih8OhPXv2WG3btm1Tnz595HQ6Vb16dfXs2VOffvqpx7aKxrNu3To99thjCggIUP369a3+1157TU2bNlWVKlUUERGhDRs2lDjOV199Va1atVLVqlVVs2ZNdejQQYsXL77sXBXdB7dkyZIy+55d6fg//PBDde3aVdWqVVONGjUUGxurXbt2Wf1//etf5XA49O233xbb9sSJE+Xr66sTJ05IKvmeqsLCQs2aNUutWrVS5cqVFRgYqEcffdR6jSQlJCSodu3aMsZYbaNHj5bD4dDf/vY3qy07O1sOh0Pz58+/5PECFyNUAWUsJCREDz/88HU5WxUXF6fCwkJNnz5dkZGReuGFFzRr1izddddduuWWW/SXv/xFzZo10xNPPKH169cXe/2LL76olStXasKECXr88ce1evVqRUdH68yZM1bNmjVr1K1bN7ndbk2ZMkV//vOflZOTozvvvFNbtmwpts3f/e53+vHHH/XnP/9Zw4YNu+TYs7Ozdfvtt2vVqlV67LHH9OKLL+rs2bP67W9/q+XLl0uSunXrprfffluSdNddd+ntt9+21kvStWtXffLJJ9b68ePHtWvXLnl5eXkEnA0bNqhu3bpq2bKlJGnXrl3q2rWrvvjiC40fP17PPPOMDhw4oO7du2vz5s3F9vPYY49p9+7dmjx5sp588klJ0v/8z//o0UcfVVBQkGbMmKHOnTuXGHhef/11Pf744woNDdWsWbP07LPPKiwsrMT9lKQsv2eXO/63335bsbGxql69uv7yl7/omWee0e7du9WlSxd98803kqT7779fDodDS5cuLbbNpUuXqlevXqpZs+Yl9/voo49q3Lhx6ty5s2bPnq3Bgwdr0aJFiomJ0blz5yT99DNQ9H0vsmHDhhJ/BqSffsaAUjMAysTChQuNJPPZZ5+Z/fv3m0qVKpnHH3/c6r/jjjtMq1atrPUDBw4YSWbhwoXFtiXJTJkyxVqfMmWKkWSGDx9utZ0/f97Ur1/fOBwOM336dKv9xIkTpkqVKmbQoEFW29q1a40kc8sttxi32221L1261Egys2fPNsYYU1hYaG699VYTExNjCgsLrboff/zRhISEmLvuuqvYmB544IFSzc+YMWOMJLNhwwar7eTJkyYkJMQ0btzYFBQUeBx/fHz8Fbe5bNkyI8ns3r3bGGPMf/7zH+Pn52d++9vfmri4OKuubdu25t5777XW+/bta3x9fc3+/futtkOHDpkaNWqYbt26WW1F39MuXbqY8+fPW+35+fkmICDAhIWFmby8PKv9tddeM5LMHXfcYbXdc889Ht/30ioP37NLHf/JkyeNv7+/GTZsmEd9VlaWcblcHu1RUVEmPDzco27Lli1GkvnHP/5htQ0aNMg0atTIWt+wYYORZBYtWuTx2qSkJI/2I0eOGElm3rx5xhhjcnJyjJeXl/nd735nAgMDrdc9/vjjplatWh5zBFwJZ6qAcqBJkyZ66KGH9Nprr+nw4cO2bXfo0KHW197e3urQoYOMMRoyZIjV7u/vr+bNm+vrr78u9vqHH35YNWrUsNbvu+8+1atXTx988IEkKT09Xfv27dPAgQN17Ngx/fDDD/rhhx90+vRp9ezZU+vXry92M/GIESNKNfYPPvhAERERHpebqlevruHDh+ubb77R7t27SzcJF+jataskWWflNmzYoI4dO+quu+6yzkzk5ORo586dVm1BQYE++ugj9e3bV02aNLG2Va9ePQ0cOFCffPKJ3G63x36GDRsmb29va33r1q06cuSIRowYIV9fX6v9kUcekcvl8nitv7+/vvvuO3322WdXfXxS2X7Pilx8/KtXr1ZOTo4eeOABa38//PCDvL29FRkZqbVr11q1cXFxSktL0/79+622JUuWyM/PT/fcc88l97ls2TK5XC7dddddHvsIDw9X9erVrX3UrVtXLVq0sH4GNm7cKG9vb40bN07Z2dnat2+fpJ9+Nrp06XLZy8nAxQhVQDkxadIknT9//or3Vl2Nhg0beqy7XC5VrlxZderUKdZ+4X0nRW699VaPdYfDoWbNmlmXa4r+AA0aNEh169b1WN544w3l5eUpNzfXYxshISGlGvu3336r5s2bF2svuiRX0n03VxIYGKhbb73VClAbNmxQ165d1a1bNx06dEhff/21Nm7cqMLCQitUHT16VD/++OMlx1JYWFjsEt7Fx1g01ovn08fHxyOoSdKECRNUvXp1RURE6NZbb1V8fHyxe50upyy/Z5eqL9rnnXfeWWyfH330kY4cOWLV/u53v5OXl5eWLFkiSTLGaNmyZdb9bJeyb98+5ebmKiAgoNg+Tp065bGPrl27evwMdOjQQR06dFCtWrW0YcMGud1uffHFF9bPAFBaPFIBKCeaNGmi3//+93rttdes+1AudKn/MRcUFFxymxeeLbhcmySPG3dLq+iMxksvvaSwsLASa6pXr+6xXqVKlavej526dOmi5ORknTlzRmlpaZo8ebJat24tf39/bdiwQXv27FH16tX1q1/96pr38XOOsWXLlsrIyNCKFSuUlJSkf/3rX5o3b54mT56sZ5999pq3W+SX+J5dXF+0z7fffltBQUHF6itV+r8/RcHBweratauWLl2qp556Sp9++qkyMzP1l7/85bL7LCwsVEBAgBYtWlRif926da2vu3Tpotdff11ff/21FawdDoe6dOmiDRs2KDg42CNYA6VFqALKkUmTJul///d/S/wDUnSDbk5Ojkf7tZyxKa2iMwxFjDH66quv1LZtW0lS06ZNJUlOp1PR0dG27rtRo0bKyMgo1r53716r/1p07dpVCxcu1LvvvquCggLdfvvt8vLysv6g7tmzR7fffrsVPuvWrauqVatecixeXl5q0KDBFY9F+mk+77zzTqv93LlzOnDggNq1a+dRX61aNcXFxSkuLk75+fnq16+fXnzxRU2cOFGVK1e+7L7K8nt2KUX7DAgIKNU+4+Li9NhjjykjI0NLlixR1apV9Zvf/OaK+/j444/VuXPnK4bAorC0evVqffbZZ9Z/Yrp166b58+crODhY1apVU3h4eGkOD7Bw+Q8oR5o2barf//73+vvf/66srCyPPqfTqTp16hR7l968efOu23j+8Y9/6OTJk9b6P//5Tx0+fFh9+vSRJIWHh6tp06b661//qlOnThV7/dGjR69533fffbe2bNmi1NRUq+306dN67bXX1LhxY4WGhl7Tdov+oP7lL39R27ZtrXuaunbtquTkZG3dutXjDIW3t7d69eql999/37qEJv307sTFixerS5cul70sJUkdOnRQ3bp1tWDBAuXn51vtiYmJxULysWPHPNZ9fX0VGhoqY4z1DrbLKcvv2aXExMTI6XTqz3/+c4nHcPE++/fvL29vb73zzjtatmyZfv3rX6tatWqX3cf999+vgoICPf/888X6zp8/7zHPISEhuuWWWzRz5kydO3dOnTt3lvTTz8D+/fv1z3/+U506dfI4gwaUBj8xQDnz9NNP6+2331ZGRoZatWrl0Td06FBNnz5dQ4cOVYcOHbR+/Xp9+eWX120stWrVUpcuXTR48GBlZ2dr1qxZatasmfW2ei8vL73xxhvq06ePWrVqpcGDB+uWW27R999/r7Vr18rpdOq///3vNe37ySef1DvvvKM+ffro8ccfV61atfTWW2/pwIED+te//iUvr2v7P2GzZs0UFBSkjIwMjR492mrv1q2bJkyYIEnFLvu88MILWr16tbp06aLHHntMlSpV0t///nfl5eVpxowZV9ynj4+PXnjhBT366KO68847FRcXpwMHDmjhwoXF7qnq1auXgoKC1LlzZwUGBmrPnj2aM2eOYmNjPW5Av5Sy/J5ditPp1Pz58/XQQw+pffv2GjBggOrWravMzEytXLlSnTt31pw5c6z6gIAA9ejRQ6+88opOnjypuLi4K+7jjjvu0KOPPqpp06YpPT1dvXr1ko+Pj/bt26dly5Zp9uzZuu+++6z6rl276t1331WbNm2ss8Dt27dXtWrV9OWXX2rgwIG2zgFuEmX51kPgZnbhIxUuNmjQICOp2Fvrf/zxRzNkyBDjcrlMjRo1zP3332+9RbykRyocPXq02HarVatWbH8XP76h6O3577zzjpk4caIJCAgwVapUMbGxsebbb78t9vpt27aZfv36mdq1axs/Pz/TqFEjc//995vk5OQrjuly9u/fb+677z7j7+9vKleubCIiIsyKFSuK1amUj1Qo8rvf/c5IMkuWLLHa8vPzTdWqVY2vr685c+ZMsdd8/vnnJiYmxlSvXt1UrVrV9OjRw2zatMmj5nLfU2OMmTdvngkJCTF+fn6mQ4cOZv369eaOO+7weKTC3//+d9OtWzdrLps2bWrGjRtncnNzL3tM5eF7dqXjX7t2rYmJiTEul8tUrlzZNG3a1DzyyCNm69atxWpff/11I8nUqFGjxO/HxY9UKPLaa6+Z8PBwU6VKFVOjRg3Tpk0bM378eHPo0CGPurlz5xpJZuTIkR7t0dHRRpLHPACl5TDmGu5OBQCUKykpKerRo4eWLVvmcUYGwC+He6oAAABsQKgCAACwAaEKAADABtxTBQAAYAPOVAEAANiAUAUAAGADHv75CyosLNShQ4dUo0YNPvkcAIAKwhijkydPKjg4+LIPHiZU/YIOHTp0xc8IAwAA5dPBgwdVv379S/YTqn5BRR8xcfDgwSt+VhgAACgf3G63GjRocMWPiiJU/YKKLvk5nU5CFQAAFcyVbt3hRnUAAAAbEKoAAABsQKgCAACwAaEKAADABoQqAAAAGxCqAAAAbECoAgAAsAGhCgAAwAaEKgAAABsQqgAAAGxAqAIAALABoQoAAMAGhCoAAAAbEKoAAABsUKmsB4CfLz8/X+np6cXaw8LC5Ovr+8sPCACAmxCh6gaQnp6uUXPflyu4idWWe+hrzYmXIiIiynBkAADcPAhVNwhXcBPVDgkt62EAAHDT4p4qAAAAGxCqAAAAbECoAgAAsAGhCgAAwAaEKgAAABsQqgAAAGxAqAIAALABoQoAAMAGZRqq5s+fr7Zt28rpdMrpdCoqKkoffvih1d+9e3c5HA6PZcSIER7byMzMVGxsrKpWraqAgACNGzdO58+f96hJSUlR+/bt5efnp2bNmikxMbHYWObOnavGjRurcuXKioyM1JYtWzz6z549q/j4eNWuXVvVq1dX//79lZ2dbd9kAACACq1MQ1X9+vU1ffp0paWlaevWrbrzzjt1zz33aNeuXVbNsGHDdPjwYWuZMWOG1VdQUKDY2Fjl5+dr06ZNeuutt5SYmKjJkydbNQcOHFBsbKx69Oih9PR0jRkzRkOHDtWqVausmiVLlighIUFTpkzR559/rnbt2ikmJkZHjhyxasaOHav//ve/WrZsmdatW6dDhw6pX79+13mGAABAReEwxpiyHsSFatWqpZdeeklDhgxR9+7dFRYWplmzZpVY++GHH+rXv/61Dh06pMDAQEnSggULNGHCBB09elS+vr6aMGGCVq5cqZ07d1qvGzBggHJycpSUlCRJioyMVMeOHTVnzhxJUmFhoRo0aKDRo0frySefVG5ururWravFixfrvvvukyTt3btXLVu2VGpqqjp16lSqY3O73XK5XMrNzZXT6bzWKSpmy5Ytenr5Do+PqTl2YLdevLcNn/0HAMDPVNq/3+XmnqqCggK9++67On36tKKioqz2RYsWqU6dOmrdurUmTpyoH3/80epLTU1VmzZtrEAlSTExMXK73dbZrtTUVEVHR3vsKyYmRqmpqZKk/Px8paWledR4eXkpOjraqklLS9O5c+c8alq0aKGGDRtaNSXJy8uT2+32WAAAwI2pzD9QeceOHYqKitLZs2dVvXp1LV++XKGhP51xGThwoBo1aqTg4GBt375dEyZMUEZGhv79739LkrKysjwClSRrPSsr67I1brdbZ86c0YkTJ1RQUFBizd69e61t+Pr6yt/fv1hN0X5KMm3aND377LNXOSMAAKAiKvNQ1bx5c6Wnpys3N1f//Oc/NWjQIK1bt06hoaEaPny4VdemTRvVq1dPPXv21P79+9W0adMyHHXpTJw4UQkJCda62+1WgwYNynBEAADgeinzy3++vr5q1qyZwsPDNW3aNLVr106zZ88usTYyMlKS9NVXX0mSgoKCir0Dr2g9KCjosjVOp1NVqlRRnTp15O3tXWLNhdvIz89XTk7OJWtK4ufnZ72zsWgBAAA3pjIPVRcrLCxUXl5eiX3p6emSpHr16kmSoqKitGPHDo936a1evVpOp9O6hBgVFaXk5GSP7axevdq6b8vX11fh4eEeNYWFhUpOTrZqwsPD5ePj41GTkZGhzMxMj/u/AADAzatML/9NnDhRffr0UcOGDXXy5EktXrxYKSkpWrVqlfbv36/Fixfr7rvvVu3atbV9+3aNHTtW3bp1U9u2bSVJvXr1UmhoqB566CHNmDFDWVlZmjRpkuLj4+Xn5ydJGjFihObMmaPx48frD3/4g9asWaOlS5dq5cqV1jgSEhI0aNAgdejQQREREZo1a5ZOnz6twYMHS5JcLpeGDBmihIQE1apVS06nU6NHj1ZUVFSp3/kHAABubGUaqo4cOaKHH35Yhw8flsvlUtu2bbVq1SrdddddOnjwoD7++GMr4DRo0ED9+/fXpEmTrNd7e3trxYoVGjlypKKiolStWjUNGjRIzz33nFUTEhKilStXauzYsZo9e7bq16+vN954QzExMVZNXFycjh49qsmTJysrK0thYWFKSkryuHl95syZ8vLyUv/+/ZWXl6eYmBjNmzfvl5koAABQ7pW751TdyHhOFQAAFU+Fe04VAABARUaoAgAAsAGhCgAAwAaEKgAAABsQqgAAAGxAqAIAALABoQoAAMAGhCoAAAAbEKoAAABsQKgCAACwAaEKAADABoQqAAAAGxCqAAAAbECoAgAAsAGhCgAAwAaEKgAAABsQqgAAAGxAqAIAALABoQoAAMAGhCoAAAAbEKoAAABsQKgCAACwAaEKAADABoQqAAAAGxCqAAAAbECoAgAAsAGhCgAAwAaEKgAAABsQqgAAAGxAqAIAALABoQoAAMAGhCoAAAAbEKoAAABsQKgCAACwAaEKAADABoQqAAAAGxCqAAAAbFCmoWr+/Plq27atnE6nnE6noqKi9OGHH1r9Z8+eVXx8vGrXrq3q1aurf//+ys7O9thGZmamYmNjVbVqVQUEBGjcuHE6f/68R01KSorat28vPz8/NWvWTImJicXGMnfuXDVu3FiVK1dWZGSktmzZ4tFfmrEAAICbV5mGqvr162v69OlKS0vT1q1bdeedd+qee+7Rrl27JEljx47Vf//7Xy1btkzr1q3ToUOH1K9fP+v1BQUFio2NVX5+vjZt2qS33npLiYmJmjx5slVz4MABxcbGqkePHkpPT9eYMWM0dOhQrVq1yqpZsmSJEhISNGXKFH3++edq166dYmJidOTIEavmSmMBAAA3OVPO1KxZ07zxxhsmJyfH+Pj4mGXLlll9e/bsMZJMamqqMcaYDz74wHh5eZmsrCyrZv78+cbpdJq8vDxjjDHjx483rVq18thHXFyciYmJsdYjIiJMfHy8tV5QUGCCg4PNtGnTjDGmVGMpjdzcXCPJ5Obmlvo1pbF582YT/eQbJu7vm6wl+sk3zObNm23dDwAAN6PS/v0uN/dUFRQU6N1339Xp06cVFRWltLQ0nTt3TtHR0VZNixYt1LBhQ6WmpkqSUlNT1aZNGwUGBlo1MTExcrvd1tmu1NRUj20U1RRtIz8/X2lpaR41Xl5eio6OtmpKM5aS5OXlye12eywAAODGVOahaseOHapevbr8/Pw0YsQILV++XKGhocrKypKvr6/8/f096gMDA5WVlSVJysrK8ghURf1FfZercbvdOnPmjH744QcVFBSUWHPhNq40lpJMmzZNLpfLWho0aFC6SQEAABVOmYeq5s2bKz09XZs3b9bIkSM1aNAg7d69u6yHZYuJEycqNzfXWg4ePFjWQwIAANdJpbIegK+vr5o1ayZJCg8P12effabZs2crLi5O+fn5ysnJ8ThDlJ2draCgIElSUFBQsXfpFb0j78Kai9+ll52dLafTqSpVqsjb21ve3t4l1ly4jSuNpSR+fn7y8/O7itkAAAAVVZmfqbpYYWGh8vLyFB4eLh8fHyUnJ1t9GRkZyszMVFRUlCQpKipKO3bs8HiX3urVq+V0OhUaGmrVXLiNopqibfj6+io8PNyjprCwUMnJyVZNacYCAABubmV6pmrixInq06ePGjZsqJMnT2rx4sVKSUnRqlWr5HK5NGTIECUkJKhWrVpyOp0aPXq0oqKi1KlTJ0lSr169FBoaqoceekgzZsxQVlaWJk2apPj4eOsM0YgRIzRnzhyNHz9ef/jDH7RmzRotXbpUK1eutMaRkJCgQYMGqUOHDoqIiNCsWbN0+vRpDR48WJJKNRYAAHBzK9NQdeTIET388MM6fPiwXC6X2rZtq1WrVumuu+6SJM2cOVNeXl7q37+/8vLyFBMTo3nz5lmv9/b21ooVKzRy5EhFRUWpWrVqGjRokJ577jmrJiQkRCtXrtTYsWM1e/Zs1a9fX2+88YZiYmKsmri4OB09elSTJ09WVlaWwsLClJSU5HHz+pXGAgAAbm4OY4wp60HcLNxut1wul3Jzc+V0Om3b7pYtW/T08h2qHRJqtR07sFsv3ttGERERtu0HAICbUWn/fpe7e6oAAAAqIkIVAACADQhVAAAANiBUAQAA2IBQBQAAYANCFQAAgA0IVQAAADYgVAEAANiAUAUAAGADQhUAAIANCFUAAAA2IFQBAADYgFAFAABgA0IVAACADQhVAAAANiBUAQAA2IBQBQAAYANCFQAAgA0IVQAAADYgVAEAANiAUAUAAGADQhUAAIANCFUAAAA2IFQBAADYgFAFAABgA0IVAACADQhVAAAANiBUAQAA2IBQBQAAYANCFQAAgA0IVQAAADYgVAEAANiAUAUAAGADQhUAAIANCFUAAAA2IFQBAADYgFAFAABggzINVdOmTVPHjh1Vo0YNBQQEqG/fvsrIyPCo6d69uxwOh8cyYsQIj5rMzEzFxsaqatWqCggI0Lhx43T+/HmPmpSUFLVv315+fn5q1qyZEhMTi41n7ty5aty4sSpXrqzIyEht2bLFo//s2bOKj49X7dq1Vb16dfXv31/Z2dn2TAYAAKjQyjRUrVu3TvHx8fr000+1evVqnTt3Tr169dLp06c96oYNG6bDhw9by4wZM6y+goICxcbGKj8/X5s2bdJbb72lxMRETZ482ao5cOCAYmNj1aNHD6Wnp2vMmDEaOnSoVq1aZdUsWbJECQkJmjJlij7//HO1a9dOMTExOnLkiFUzduxY/fe//9WyZcu0bt06HTp0SP369buOMwQAACoKhzHGlPUgihw9elQBAQFat26dunXrJumnM1VhYWGaNWtWia/58MMP9etf/1qHDh1SYGCgJGnBggWaMGGCjh49Kl9fX02YMEErV67Uzp07rdcNGDBAOTk5SkpKkiRFRkaqY8eOmjNnjiSpsLBQDRo00OjRo/Xkk08qNzdXdevW1eLFi3XfffdJkvbu3auWLVsqNTVVnTp1uuLxud1uuVwu5ebmyul0XvM8XWzLli16evkO1Q4JtdqOHditF+9to4iICNv2AwDAzai0f7/L1T1Vubm5kqRatWp5tC9atEh16tRR69atNXHiRP34449WX2pqqtq0aWMFKkmKiYmR2+3Wrl27rJro6GiPbcbExCg1NVWSlJ+fr7S0NI8aLy8vRUdHWzVpaWk6d+6cR02LFi3UsGFDq+ZieXl5crvdHgsAALgxVSrrARQpLCzUmDFj1LlzZ7Vu3dpqHzhwoBo1aqTg4GBt375dEyZMUEZGhv79739LkrKysjwClSRrPSsr67I1brdbZ86c0YkTJ1RQUFBizd69e61t+Pr6yt/fv1hN0X4uNm3aND377LNXORMAAKAiKjehKj4+Xjt37tQnn3zi0T58+HDr6zZt2qhevXrq2bOn9u/fr6ZNm/7Sw7wqEydOVEJCgrXudrvVoEGDMhwRAAC4XsrF5b9Ro0ZpxYoVWrt2rerXr3/Z2sjISEnSV199JUkKCgoq9g68ovWgoKDL1jidTlWpUkV16tSRt7d3iTUXbiM/P185OTmXrLmYn5+fnE6nxwIAAG5MZRqqjDEaNWqUli9frjVr1igkJOSKr0lPT5ck1atXT5IUFRWlHTt2eLxLb/Xq1XI6nQoNDbVqkpOTPbazevVqRUVFSZJ8fX0VHh7uUVNYWKjk5GSrJjw8XD4+Ph41GRkZyszMtGoAAMDNq0wv/8XHx2vx4sV6//33VaNGDeveJJfLpSpVqmj//v1avHix7r77btWuXVvbt2/X2LFj1a1bN7Vt21aS1KtXL4WGhuqhhx7SjBkzlJWVpUmTJik+Pl5+fn6SpBEjRmjOnDkaP368/vCHP2jNmjVaunSpVq5caY0lISFBgwYNUocOHRQREaFZs2bp9OnTGjx4sDWmIUOGKCEhQbVq1ZLT6dTo0aMVFRVVqnf+AQCAG1uZhqr58+dL+umxCRdauHChHnnkEfn6+urjjz+2Ak6DBg3Uv39/TZo0yar19vbWihUrNHLkSEVFRalatWoaNGiQnnvuOasmJCREK1eu1NixYzV79mzVr19fb7zxhmJiYqyauLg4HT16VJMnT1ZWVpbCwsKUlJTkcfP6zJkz5eXlpf79+ysvL08xMTGaN2/edZodAABQkZSr51Td6HhOFQAAFU+FfE4VAABARUWoAgAAsAGhCgAAwAaEKgAAABsQqgAAAGxAqAIAALABoQoAAMAGhCoAAAAbEKoAAABsQKgCAACwAaEKAADABoQqAAAAGxCqAAAAbECoAgAAsAGhCgAAwAaEKgAAABsQqgAAAGxAqAIAALABoQoAAMAGhCoAAAAbEKoAAABsQKgCAACwAaEKAADABoQqAAAAGxCqAAAAbECoAgAAsAGhCgAAwAaEKgAAABsQqgAAAGxAqAIAALABoQoAAMAGhCoAAAAbXFOoatKkiY4dO1asPScnR02aNPnZgwIAAKhorilUffPNNyooKCjWnpeXp++///5nDwoAAKCiqXQ1xf/5z3+sr1etWiWXy2WtFxQUKDk5WY0bN7ZtcAAAABXFVYWqvn37SpIcDocGDRrk0efj46PGjRvr5Zdftm1wAAAAFcVVharCwkJJUkhIiD777DPVqVPnugwKAACgormme6oOHDhgS6CaNm2aOnbsqBo1aiggIEB9+/ZVRkaGR83Zs2cVHx+v2rVrq3r16urfv7+ys7M9ajIzMxUbG6uqVasqICBA48aN0/nz5z1qUlJS1L59e/n5+alZs2ZKTEwsNp65c+eqcePGqly5siIjI7Vly5arHgsAALg5XdWZqgslJycrOTlZR44csc5gFXnzzTdLtY1169YpPj5eHTt21Pnz5/XUU0+pV69e2r17t6pVqyZJGjt2rFauXKlly5bJ5XJp1KhR6tevnzZu3Cjpp3u5YmNjFRQUpE2bNunw4cN6+OGH5ePjoz//+c+SfgqBsbGxGjFihBYtWqTk5GQNHTpU9erVU0xMjCRpyZIlSkhI0IIFCxQZGalZs2YpJiZGGRkZCggIKNVYAADAzcthjDFX+6Jnn31Wzz33nDp06KB69erJ4XB49C9fvvyaBnP06FEFBARo3bp16tatm3Jzc1W3bl0tXrxY9913nyRp7969atmypVJTU9WpUyd9+OGH+vWvf61Dhw4pMDBQkrRgwQJNmDBBR48ela+vryZMmKCVK1dq586d1r4GDBignJwcJSUlSZIiIyPVsWNHzZkzR9JPlzobNGig0aNH68knnyzVWK7E7XbL5XIpNzdXTqfzmuaoJFu2bNHTy3eodkio1XbswG69eG8bRURE2LYfAABuRqX9+31NZ6oWLFigxMREPfTQQ9c8wJLk5uZKkmrVqiVJSktL07lz5xQdHW3VtGjRQg0bNrSCTGpqqtq0aWMFKkmKiYnRyJEjtWvXLv3qV79SamqqxzaKasaMGSNJys/PV1pamiZOnGj1e3l5KTo6WqmpqaUeCwAAuHldU6jKz8/X7bffbutACgsLNWbMGHXu3FmtW7eWJGVlZcnX11f+/v4etYGBgcrKyrJqLgxURf1FfZercbvdOnPmjE6cOKGCgoISa/bu3VvqsVwsLy9PeXl51rrb7b7SNAAAgArqmm5UHzp0qBYvXmzrQOLj47Vz5069++67tm63LE2bNk0ul8taGjRoUNZDAgAA18k1nak6e/asXnvtNX388cdq27atfHx8PPpfeeWVq9reqFGjtGLFCq1fv17169e32oOCgpSfn6+cnByPM0TZ2dkKCgqyai5+l17RO/IurLn4XXrZ2dlyOp2qUqWKvL295e3tXWLNhdu40lguNnHiRCUkJFjrbrebYAUAwA3qms5Ubd++XWFhYfLy8tLOnTu1bds2a0lPTy/1dowxGjVqlJYvX641a9YoJCTEoz88PFw+Pj5KTk622jIyMpSZmamoqChJUlRUlHbs2KEjR45YNatXr5bT6VRoaKhVc+E2imqKtuHr66vw8HCPmsLCQiUnJ1s1pRnLxfz8/OR0Oj0WAABwY7qmM1Vr1661Zefx8fFavHix3n//fdWoUcO6N8nlcqlKlSpyuVwaMmSIEhISVKtWLTmdTo0ePVpRUVHWjeG9evVSaGioHnroIc2YMUNZWVmaNGmS4uPj5efnJ0kaMWKE5syZo/Hjx+sPf/iD1qxZo6VLl2rlypXWWBISEjRo0CB16NBBERERmjVrlk6fPq3BgwdbY7rSWAAAwM3rmp9TZYf58+dLkrp37+7RvnDhQj3yyCOSpJkzZ8rLy0v9+/dXXl6eYmJiNG/ePKvW29tbK1as0MiRIxUVFaVq1app0KBBeu6556yakJAQrVy5UmPHjtXs2bNVv359vfHGG9YzqiQpLi5OR48e1eTJk5WVlaWwsDAlJSV53Lx+pbEAAICb1zU9p6pHjx7Fnk11oTVr1vysQd2oeE4VAAAVz3V9TlVYWJjH+rlz55Senq6dO3cW+6BlAACAm8E1haqZM2eW2D516lSdOnXqZw0IAACgIrqmd/9dyu9///tSf+4fAADAjcTWUJWamqrKlSvbuUkAAIAK4Zou//Xr189j3Rijw4cPa+vWrXrmmWdsGRgAAEBFck2hyuVyeax7eXmpefPmeu6559SrVy9bBgYAAFCRXFOoWrhwod3jAAAAqNB+1sM/09LStGfPHklSq1at9Ktf/cqWQQEAAFQ01xSqjhw5ogEDBiglJcX6cOGcnBz16NFD7777rurWrWvnGAEAAMq9a3r33+jRo3Xy5Ent2rVLx48f1/Hjx7Vz50653W49/vjjdo8RAACg3LumM1VJSUn6+OOP1bJlS6stNDRUc+fO5UZ1AABwU7qmM1WFhYXy8fEp1u7j46PCwsKfPSgAAICK5ppC1Z133qk//vGPOnTokNX2/fffa+zYserZs6dtgwMAAKgorilUzZkzR263W40bN1bTpk3VtGlThYSEyO1269VXX7V7jAAAAOXeNd1T1aBBA33++ef6+OOPtXfvXklSy5YtFR0dbevgAAAAKoqrOlO1Zs0ahYaGyu12y+Fw6K677tLo0aM1evRodezYUa1atdKGDRuu11gBAADKrasKVbNmzdKwYcPkdDqL9blcLj366KN65ZVXbBscAABARXFVoeqLL75Q7969L9nfq1cvpaWl/exBAQAAVDRXFaqys7NLfJRCkUqVKuno0aM/e1AAAAAVzVWFqltuuUU7d+68ZP/27dtVr169nz0oAACAiuaqQtXdd9+tZ555RmfPni3Wd+bMGU2ZMkW//vWvbRscAABARXFVj1SYNGmS/v3vf+u2227TqFGj1Lx5c0nS3r17NXfuXBUUFOjpp5++LgMFAAAoz64qVAUGBmrTpk0aOXKkJk6cKGOMJMnhcCgmJkZz585VYGDgdRkoAABAeXbVD/9s1KiRPvjgA504cUJfffWVjDG69dZbVbNmzesxPgAAgArhmp6oLkk1a9ZUx44d7RwLAABAhXVNn/0HAAAAT4QqAAAAGxCqAAAAbECoAgAAsAGhCgAAwAaEKgAAABsQqgAAAGxAqAIAALABoQoAAMAGhCoAAAAbEKoAAABsQKgCAACwAaEKAADABmUaqtavX6/f/OY3Cg4OlsPh0HvvvefR/8gjj8jhcHgsvXv39qg5fvy4HnzwQTmdTvn7+2vIkCE6deqUR8327dvVtWtXVa5cWQ0aNNCMGTOKjWXZsmVq0aKFKleurDZt2uiDDz7w6DfGaPLkyapXr56qVKmi6Oho7du3z56JAAAAFV6ZhqrTp0+rXbt2mjt37iVrevfurcOHD1vLO++849H/4IMPateuXVq9erVWrFih9evXa/jw4Va/2+1Wr1691KhRI6Wlpemll17S1KlT9dprr1k1mzZt0gMPPKAhQ4Zo27Zt6tu3r/r27audO3daNTNmzNDf/vY3LViwQJs3b1a1atUUExOjs2fP2jgjAACgoqpUljvv06eP+vTpc9kaPz8/BQUFldi3Z88eJSUl6bPPPlOHDh0kSa+++qruvvtu/fWvf1VwcLAWLVqk/Px8vfnmm/L19VWrVq2Unp6uV155xQpfs2fPVu/evTVu3DhJ0vPPP6/Vq1drzpw5WrBggYwxmjVrliZNmqR77rlHkvSPf/xDgYGBeu+99zRgwAC7pgQAAFRQ5f6eqpSUFAUEBKh58+YaOXKkjh07ZvWlpqbK39/fClSSFB0dLS8vL23evNmq6datm3x9fa2amJgYZWRk6MSJE1ZNdHS0x35jYmKUmpoqSTpw4ICysrI8alwulyIjI62akuTl5cntdnssAADgxlSuQ1Xv3r31j3/8Q8nJyfrLX/6idevWqU+fPiooKJAkZWVlKSAgwOM1lSpVUq1atZSVlWXVBAYGetQUrV+p5sL+C19XUk1Jpk2bJpfLZS0NGjS4quMHAAAVR5le/ruSCy+rtWnTRm3btlXTpk2VkpKinj17luHISmfixIlKSEiw1t1uN8EKAIAbVLk+U3WxJk2aqE6dOvrqq68kSUFBQTpy5IhHzfnz53X8+HHrPqygoCBlZ2d71BStX6nmwv4LX1dSTUn8/PzkdDo9FgAAcGOqUKHqu+++07Fjx1SvXj1JUlRUlHJycpSWlmbVrFmzRoWFhYqMjLRq1q9fr3Pnzlk1q1evVvPmzVWzZk2rJjk52WNfq1evVlRUlCQpJCREQUFBHjVut1ubN2+2agAAwM2tTEPVqVOnlJ6ervT0dEk/3RCenp6uzMxMnTp1SuPGjdOnn36qb775RsnJybrnnnvUrFkzxcTESJJatmyp3r17a9iwYdqyZYs2btyoUaNGacCAAQoODpYkDRw4UL6+vhoyZIh27dqlJUuWaPbs2R6X5f74xz8qKSlJL7/8svbu3aupU6dq69atGjVqlCTJ4XBozJgxeuGFF/Sf//xHO3bs0MMPP6zg4GD17dv3F50zAABQPpXpPVVbt25Vjx49rPWioDNo0CDNnz9f27dv11tvvaWcnBwFBwerV69eev755+Xn52e9ZtGiRRo1apR69uwpLy8v9e/fX3/729+sfpfLpY8++kjx8fEKDw9XnTp1NHnyZI9nWd1+++1avHixJk2apKeeekq33nqr3nvvPbVu3dqqGT9+vE6fPq3hw4crJydHXbp0UVJSkipXrnw9pwgAAFQQDmOMKetB3CzcbrdcLpdyc3Ntvb9qy5Ytenr5DtUOCbXajh3YrRfvbaOIiAjb9gMAwM2otH+/K9Q9VQAAAOUVoQoAAMAGhCoAAAAbEKoAAABsQKgCAACwAaEKAADABoQqAAAAGxCqAAAAbECoAgAAsAGhCgAAwAaEKgAAABsQqgAAAGxAqAIAALABoQoAAMAGhCoAAAAbEKoAAABsQKgCAACwAaEKAADABoQqAAAAGxCqAAAAbECoAgAAsAGhCgAAwAaVynoAuD4KC85r165dHm1hYWHy9fUtoxEBAHBjI1TdoE5mZ2rmt2cV+GWhJCn30NeaEy9FRESU8cgAALgxEapuYDUCG6l2SGhZDwMAgJsC91QBAADYgFAFAABgA0IVAACADQhVAAAANiBUAQAA2IBQBQAAYANCFQAAgA0IVQAAADYgVAEAANiAUAUAAGADQhUAAIANCFUAAAA2KNNQtX79ev3mN79RcHCwHA6H3nvvPY9+Y4wmT56sevXqqUqVKoqOjta+ffs8ao4fP64HH3xQTqdT/v7+GjJkiE6dOuVRs337dnXt2lWVK1dWgwYNNGPGjGJjWbZsmVq0aKHKlSurTZs2+uCDD656LAAA4OZVpqHq9OnTateunebOnVti/4wZM/S3v/1NCxYs0ObNm1WtWjXFxMTo7NmzVs2DDz6oXbt2afXq1VqxYoXWr1+v4cOHW/1ut1u9evVSo0aNlJaWppdeeklTp07Va6+9ZtVs2rRJDzzwgIYMGaJt27apb9++6tu3r3bu3HlVYwEAADcvhzHGlPUgJMnhcGj58uXq27evpJ/ODAUHB+tPf/qTnnjiCUlSbm6uAgMDlZiYqAEDBmjPnj0KDQ3VZ599pg4dOkiSkpKSdPfdd+u7775TcHCw5s+fr6efflpZWVny9fWVJD355JN67733tHfvXklSXFycTp8+rRUrVljj6dSpk8LCwrRgwYJSjaU03G63XC6XcnNz5XQ6bZk3SdqyZYueXr5DtUNCrbavN66QjytQDVp3lCQdO7BbL97bRhEREbbtFwCAm0Fp/36X23uqDhw4oKysLEVHR1ttLpdLkZGRSk1NlSSlpqbK39/fClSSFB0dLS8vL23evNmq6datmxWoJCkmJkYZGRk6ceKEVXPhfopqivZTmrGUJC8vT26322MBAAA3pnIbqrKysiRJgYGBHu2BgYFWX1ZWlgICAjz6K1WqpFq1annUlLSNC/dxqZoL+680lpJMmzZNLpfLWho0aHCFowYAABVVuQ1VN4KJEycqNzfXWg4ePFjWQwIAANdJuQ1VQUFBkqTs7GyP9uzsbKsvKChIR44c8eg/f/68jh8/7lFT0jYu3Melai7sv9JYSuLn5yen0+mxAACAG1O5DVUhISEKCgpScnKy1eZ2u7V582ZFRUVJkqKiopSTk6O0tDSrZs2aNSosLFRkZKRVs379ep07d86qWb16tZo3b66aNWtaNRfup6imaD+lGQsAALi5lWmoOnXqlNLT05Weni7ppxvC09PTlZmZKYfDoTFjxuiFF17Qf/7zH+3YsUMPP/ywgoODrXcItmzZUr1799awYcO0ZcsWbdy4UaNGjdKAAQMUHBwsSRo4cKB8fX01ZMgQ7dq1S0uWLNHs2bOVkJBgjeOPf/yjkpKS9PLLL2vv3r2aOnWqtm7dqlGjRklSqcYCAABubpXKcudbt25Vjx49rPWioDNo0CAlJiZq/PjxOn36tIYPH66cnBx16dJFSUlJqly5svWaRYsWadSoUerZs6e8vLzUv39//e1vf7P6XS6XPvroI8XHxys8PFx16tTR5MmTPZ5ldfvtt2vx4sWaNGmSnnrqKd16661677331Lp1a6umNGMBAAA3r3LznKqbAc+pAgCg4qnwz6kCAACoSAhVAAAANiBUAQAA2IBQBQAAYANCFQAAgA0IVQAAADYgVAEAANiAUAUAAGADQhUAAIANCFUAAAA2IFQBAADYgFAFAABgA0IVAACADQhVAAAANiBUAQAA2IBQBQAAYANCFQAAgA0IVQAAADYgVAEAANiAUAUAAGADQhUAAIANCFUAAAA2IFQBAADYgFAFAABgA0IVAACADQhVAAAANiBUAQAA2IBQBQAAYANCFQAAgA0IVQAAADYgVAEAANiAUAUAAGADQhUAAIANCFUAAAA2IFQBAADYoFJZDwC/jMKC89q1a5dHW1hYmHx9fctoRAAA3FgIVTeJk9mZmvntWQV+WShJyj30tebESxEREWU8MgAAbgzl+vLf1KlT5XA4PJYWLVpY/WfPnlV8fLxq166t6tWrq3///srOzvbYRmZmpmJjY1W1alUFBARo3LhxOn/+vEdNSkqK2rdvLz8/PzVr1kyJiYnFxjJ37lw1btxYlStXVmRkpLZs2XJdjvl6qhHYSLVDQlU7JFSu4CZlPRwAAG4o5TpUSVKrVq10+PBha/nkk0+svrFjx+q///2vli1bpnXr1unQoUPq16+f1V9QUKDY2Fjl5+dr06ZNeuutt5SYmKjJkydbNQcOHFBsbKx69Oih9PR0jRkzRkOHDtWqVausmiVLlighIUFTpkzR559/rnbt2ikmJkZHjhz5ZSYBAACUe+U+VFWqVElBQUHWUqdOHUlSbm6u/ud//kevvPKK7rzzToWHh2vhwoXatGmTPv30U0nSRx99pN27d+t///d/FRYWpj59+uj555/X3LlzlZ+fL0lasGCBQkJC9PLLL6tly5YaNWqU7rvvPs2cOdMawyuvvKJhw4Zp8ODBCg0N1YIFC1S1alW9+eabv/yEAACAcqnch6p9+/YpODhYTZo00YMPPqjMzExJUlpams6dO6fo6GirtkWLFmrYsKFSU1MlSampqWrTpo0CAwOtmpiYGLndbuum7dTUVI9tFNUUbSM/P19paWkeNV5eXoqOjrZqLiUvL09ut9tjAQAAN6ZyHaoiIyOVmJiopKQkzZ8/XwcOHFDXrl118uRJZWVlydfXV/7+/h6vCQwMVFZWliQpKyvLI1AV9Rf1Xa7G7XbrzJkz+uGHH1RQUFBiTdE2LmXatGlyuVzW0qBBg6ueAwAAUDGU63f/9enTx/q6bdu2ioyMVKNGjbR06VJVqVKlDEdWOhMnTlRCQoK17na7CVYAANygyvWZqov5+/vrtttu01dffaWgoCDl5+crJyfHoyY7O1tBQUGSpKCgoGLvBixav1KN0+lUlSpVVKdOHXl7e5dYU7SNS/Hz85PT6fRYAADAjalChapTp05p//79qlevnsLDw+Xj46Pk5GSrPyMjQ5mZmYqKipIkRUVFaceOHR7v0lu9erWcTqdCQ0Otmgu3UVRTtA1fX1+Fh4d71BQWFio5OdmqAQAAKNeh6oknntC6dev0zTffaNOmTbr33nvl7e2tBx54QC6XS0OGDFFCQoLWrl2rtLQ0DR48WFFRUerUqZMkqVevXgoNDdVDDz2kL774QqtWrdKkSZMUHx8vPz8/SdKIESP09ddfa/z48dq7d6/mzZunpUuXauzYsdY4EhIS9Prrr+utt97Snj17NHLkSJ0+fVqDBw8uk3kBAADlT7m+p+q7777TAw88oGPHjqlu3brq0qWLPv30U9WtW1eSNHPmTHl5eal///7Ky8tTTEyM5s2bZ73e29tbK1as0MiRIxUVFaVq1app0KBBeu6556yakJAQrVy5UmPHjtXs2bNVv359vfHGG4qJibFq4uLidPToUU2ePFlZWVkKCwtTUlJSsZvXAQDAzcthjDFlPYibhdvtlsvlUm5urq33V23ZskVPL9+h2iGhVtvXG1fIxxWoBq07lrh+7MBuvXhvGz6mBgCAKyjt3+9yffkPAACgoiBUAQAA2IBQBQAAYANCFQAAgA0IVQAAADYgVAEAANiAUAUAAGADQhUAAIANCFUAAAA2IFQBAADYoFx/9h+un8KC89q1a5dHW1hYmHx9fctoRAAAVGyEqpvUyexMzfz2rAK/LJQk5R76WnPixWcBAgBwjQhVN7EagY08PoQZAABcO+6pAgAAsAGhCgAAwAaEKgAAABsQqgAAAGxAqAIAALABoQoAAMAGhCoAAAAbEKoAAABswMM/Iankj62R+OgaAABKi1AFScU/tkbio2sAALgahCpY+NgaAACuHfdUAQAA2IBQBQAAYANCFQAAgA0IVQAAADbgRnVcUkmPWeARCwAAlIxQhUu6+DELPGIBAIBLI1ThsnjMAgAApUOoQqlxORAAgEsjVKHUuBwIAMClEapwVbgcCABAyQhVuGZcDgQA4P8QqnDNLr4ceOLgPsX33KVWrVpZNYQsAMDNglCFn+XCy4G5h77WzKSd3HMFALgpEaqu0ty5c/XSSy8pKytL7dq106uvvkpouMCFIauky4Pnzp2TJPn4+FhtnM0CANwICFVXYcmSJUpISNCCBQsUGRmpWbNmKSYmRhkZGQoICCjr4ZU7F18elKTvt3+iStVrKbDJT8GLS4YAgBsFoeoqvPLKKxo2bJgGDx4sSVqwYIFWrlypN998U08++WQZj658uvjdgrmHvpaPK/CSlwwvDlkXn9kq6UwXZ78AAOUBoaqU8vPzlZaWpokTJ1ptXl5eio6OVmpqahmOrOK73H1ZF5/Zuni9pDY7gtnPXS+v27xe45I8g2x+fr7S09Mv2Q8ANyJCVSn98MMPKigoUGBgoEd7YGCg9u7dW+Jr8vLylJeXZ63n5uZKktxut61jO3XqlI5/s0fn8878374Of6NKJ3PlW8nrmtbt2MY1b7N6TetYCs7ly3Eu75LrJbWdPHJQz76xR/710iRJx77ZI2+/6vKv16DE9dLUXO16ed3m9RrX6ePZGntfD7Vo0UKStHfvXs3851pVqxVYYj8AXA8dOnS4Ltst+rttjLl8oUGpfP/990aS2bRpk0f7uHHjTERERImvmTJlipHEwsLCwsLCcgMsBw8evGxW4ExVKdWpU0fe3t7Kzs72aM/OzlZQUFCJr5k4caISEhKs9cLCQh0/fly1a9eWw+GwbWxut1sNGjTQwYMH5XQ6bdvuzYw5tRfzaS/m017Mp/1utDk1xujkyZMKDg6+bB2hqpR8fX0VHh6u5ORk9e3bV9JPISk5OVmjRo0q8TV+fn7y8/PzaPP3979uY3Q6nTfED295wpzai/m0F/NpL+bTfjfSnLpcrivWEKquQkJCggYNGqQOHTooIiJCs2bN0unTp613AwIAgJsXoeoqxMXF6ejRo5o8ebKysrIUFhampKSkYjevAwCAmw+h6iqNGjXqkpf7yoqfn5+mTJlS7FIjrh1zai/m017Mp72YT/vdrHPqMOZK7w8EAADAlXhduQQAAABXQqgCAACwAaEKAADABoQqAAAAGxCqbgBz585V48aNVblyZUVGRmrLli1lPaRyZ+rUqXI4HB7LhZ9Dd/bsWcXHx6t27dqqXr26+vfvX+zp+ZmZmYqNjVXVqlUVEBCgcePG6fz587/0oZSZ9evX6ze/+Y2Cg4PlcDj03nvvefQbYzR58mTVq1dPVapUUXR0tPbt2+dRc/z4cT344INyOp3y9/fXkCFDdOrUKY+a7du3q2vXrqpcubIaNGigGTNmXO9DKxNXms9HHnmk2M9s7969PWqYz/8zbdo0dezYUTVq1FBAQID69u2rjIwMjxq7fs9TUlLUvn17+fn5qVmzZkpMTLzeh/eLK818du/evdjP6IgRIzxqbrr5tOWD8VBm3n33XePr62vefPNNs2vXLjNs2DDj7+9vsrOzy3po5cqUKVNMq1atzOHDh63l6NGjVv+IESNMgwYNTHJystm6davp1KmTuf32263+8+fPm9atW5vo6Gizbds288EHH5g6deqYiRMnlsXhlIkPPvjAPP300+bf//63kWSWL1/u0T99+nTjcrnMe++9Z7744gvz29/+1oSEhJgzZ85YNb179zbt2rUzn376qdmwYYNp1qyZeeCBB6z+3NxcExgYaB588EGzc+dO884775gqVaqYv//977/UYf5irjSfgwYNMr179/b4mT1+/LhHDfP5f2JiYszChQvNzp07TXp6urn77rtNw4YNzalTp6waO37Pv/76a1O1alWTkJBgdu/ebV599VXj7e1tkpKSftHjvd5KM5933HGHGTZsmMfPaG5urtV/M84noaqCi4iIMPHx8dZ6QUGBCQ4ONtOmTSvDUZU/U6ZMMe3atSuxLycnx/j4+Jhly5ZZbXv27DGSTGpqqjHmpz+AXl5eJisry6qZP3++cTqdJi8v77qOvTy6OAQUFhaaoKAg89JLL1ltOTk5xs/Pz7zzzjvGGGN2795tJJnPPvvMqvnwww+Nw+Ew33//vTHGmHnz5pmaNWt6zOmECRNM8+bNr/MRla1Lhap77rnnkq9hPi/vyJEjRpJZt26dMca+3/Px48ebVq1aeewrLi7OxMTEXO9DKlMXz6cxP4WqP/7xj5d8zc04n1z+q8Dy8/OVlpam6Ohoq83Ly0vR0dFKTU0tw5GVT/v27VNwcLCaNGmiBx98UJmZmZKktLQ0nTt3zmMeW7RooYYNG1rzmJqaqjZt2ng8PT8mJkZut1u7du36ZQ+kHDpw4ICysrI85tDlcikyMtJjDv39/dWhQwerJjo6Wl5eXtq8ebNV061bN/n6+lo1MTExysjI0IkTJ36hoyk/UlJSFBAQoObNm2vkyJE6duyY1cd8Xl5ubq4kqVatWpLs+z1PTU312EZRzY3+b+7F81lk0aJFqlOnjlq3bq2JEyfqxx9/tPpuxvnkieoV2A8//KCCgoJiH5MTGBiovXv3ltGoyqfIyEglJiaqefPmOnz4sJ599ll17dpVO3fuVFZWlnx9fYt92HVgYKCysrIkSVlZWSXOc1Hfza5oDkqaowvnMCAgwKO/UqVKqlWrlkdNSEhIsW0U9dWsWfO6jL886t27t/r166eQkBDt379fTz31lPr06aPU1FR5e3szn5dRWFioMWPGqHPnzmrdurUk2fZ7fqkat9utM2fOqEqVKtfjkMpUSfMpSQMHDlSjRo0UHBys7du3a8KECcrIyNC///1vSTfnfBKqcFPo06eP9XXbtm0VGRmpRo0aaenSpRXulxY3hwEDBlhft2nTRm3btlXTpk2VkpKinj17luHIyr/4+Hjt3LlTn3zySVkP5YZwqfkcPny49XWbNm1Ur1499ezZU/v371fTpk1/6WGWC1z+q8Dq1Kkjb2/vYu9eyc7OVlBQUBmNqmLw9/fXbbfdpq+++kpBQUHKz89XTk6OR82F8xgUFFTiPBf13eyK5uByP4tBQUE6cuSIR//58+d1/Phx5rkUmjRpojp16uirr76SxHxeyqhRo7RixQqtXbtW9evXt9rt+j2/VI3T6bwh/4N2qfksSWRkpCR5/IzebPNJqKrAfH19FR4eruTkZKutsLBQycnJioqKKsORlX+nTp3S/v37Va9ePYWHh8vHx8djHjMyMpSZmWnNY1RUlHbs2OHxR2z16tVyOp0KDQ39xcdf3oSEhCgoKMhjDt1utzZv3uwxhzk5OUpLS7Nq1qxZo8LCQusf46ioKK1fv17nzp2zalavXq3mzZvfsJeqSuu7777TsWPHVK9ePUnM58WMMRo1apSWL1+uNWvWFLvsadfveVRUlMc2imputH9zrzSfJUlPT5ckj5/Rm24+y/pOefw87777rvHz8zOJiYlm9+7dZvjw4cbf39/j3RYw5k9/+pNJSUkxBw4cMBs3bjTR0dGmTp065siRI8aYn95q3bBhQ7NmzRqzdetWExUVZaKioqzXF701uFevXiY9Pd0kJSWZunXr3lSPVDh58qTZtm2b2bZtm5FkXnnlFbNt2zbz7bffGmN+eqSCv7+/ef/998327dvNPffcU+IjFX71q1+ZzZs3m08++cTceuutHo8AyMnJMYGBgeahhx4yO3fuNO+++66pWrXqDfkIgMvN58mTJ80TTzxhUlNTzYEDB8zHH39s2rdvb2699VZz9uxZaxvM5/8ZOXKkcblcJiUlxeMt/j/++KNVY8fvedEjAMaNG2f27Nlj5s6dW6EfAXApV5rPr776yjz33HNm69at5sCBA+b99983TZo0Md26dbO2cTPOJ6HqBvDqq6+ahg0bGl9fXxMREWE+/fTTsh5SuRMXF2fq1atnfH19zS233GLi4uLMV199ZfWfOXPGPPbYY6ZmzZqmatWq5t577zWHDx/22MY333xj+vTpY6pUqWLq1Klj/vSnP5lz58790odSZtauXWskFVsGDRpkjPnpsQrPPPOMCQwMNH5+fqZnz54mIyPDYxvHjh0zDzzwgKlevbpxOp1m8ODB5uTJkx41X3zxhenSpYvx8/Mzt9xyi5k+ffovdYi/qMvN548//mh69epl6tata3x8fEyjRo3MsGHDiv1nifn8PyXNpSSzcOFCq8au3/O1a9easLAw4+vra5o0aeKxjxvFleYzMzPTdOvWzdSqVcv4+fmZZs2amXHjxnk8p8qYm28+HcYY88udFwMAALgxcU8VAACADQhVAAAANiBUAQAA2IBQBQAAYANCFQAAgA0IVQAAADYgVAEAANiAUAUAFVRiYqL8/f3LehgA/j9CFQAAgA0IVQBQzuXn55f1EACUAqEKQIXSvXt3Pf744xo/frxq1aqloKAgTZ06VZL0zTffyOFwKD093arPycmRw+FQSkqKJCklJUUOh0OrVq3Sr371K1WpUkV33nmnjhw5og8//FAtW7aU0+nUwIED9eOPP15xPCtWrJC/v78KCgokSenp6XI4HHryySetmqFDh+r3v/+9tf6vf/1LrVq1kp+fnxo3bqyXX37ZY5uNGzfW888/r4cfflhOp1PDhw+X9NPlvoYNG6pq1aq69957dezYMY/XffHFF+rRo4dq1Kghp9Op8PBwbd26tdRzC+DnIVQBqHDeeustVatWTZs3b9aMGTP03HPPafXq1Ve1jalTp2rOnDnatGmTDh48qPvvv1+zZs3S4sWLtXLlSn300Ud69dVXr7idrl276uTJk9q2bZskad26dapTp44V4oraunfvLklKS0vT/fffrwEDBmjHjh2aOnWqnnnmGSUmJnps969//avatWunbdu26ZlnntHmzZs1ZMgQjRo1Sunp6erRo4deeOEFj9c8+OCDql+/vj777DOlpaXpySeflI+Pz1XNC4Cfoaw/0RkArsYdd9xhunTp4tHWsWNHM2HCBHPgwAEjyWzbts3qO3HihJFk1q5da4wxZu3atUaS+fjjj62aadOmGUlm//79Vtujjz5qYmJiSjWm9u3bm5deeskYY0zfvn3Niy++aHx9fc3JkyfNd999ZySZL7/80hhjzMCBA81dd93l8fpx48aZ0NBQa71Ro0amb9++HjUPPPCAufvuuz3a4uLijMvlstZr1KhhEhMTSzVmAPbjTBWACqdt27Ye6/Xq1dORI0eueRuBgYGqWrWqmjRp4tFW2m3ecccdSklJkTFGGzZsUL9+/dSyZUt98sknWrdunYKDg3XrrbdKkvbs2aPOnTt7vL5z587at2+fdQlRkjp06OBRs2fPHkVGRnq0RUVFeawnJCRo6NChio6O1vTp07V///5SjR+APQhVACqciy9pORwOFRYWysvrp3/SjDFW37lz5664DYfDccltlkb37t31ySef6IsvvpCPj49atGih7t27KyUlRevWrdMdd9xRqu1cqFq1alf9mqlTp2rXrl2KjY3VmjVrFBoaquXLl1/1dgBcG0IVgBtG3bp1JUmHDx+22i68af16KbqvaubMmVaAKgpVKSkp1v1UktSyZUtt3LjR4/UbN27UbbfdJm9v70vuo2XLltq8ebNH26efflqs7rbbbtPYsWP10UcfqV+/flq4cOHPODIAV4NQBeCGUaVKFXXq1EnTp0/Xnj17tG7dOk2aNOm677dmzZpq27atFi1aZAWobt266fPPP9eXX37pcabqT3/6k5KTk/X888/ryy+/1FtvvaU5c+boiSeeuOw+Hn/8cSUlJemvf/2r9u3bpzlz5igpKcnqP3PmjEaNGqWUlBR9++232rhxoz777DO1bNnyuhwzgOIIVQBuKG+++abOnz+v8PBwjRkzptg75K6XO+64QwUFBVaoqlWrlkJDQxUUFKTmzZtbde3bt9fSpUv17rvvqnXr1po8ebKee+45PfLII5fdfqdOnfT6669r9uzZateunT766COPwOjt7a1jx47p4Ycf1m233ab7779fffr00bPPPns9DhdACRzmwpsPAAAAcE04UwUAAGADQhUAXEZmZqaqV69+ySUzM7OshwignODyHwBcxvnz5/XNN99csr9x48aqVKnSLzcgAOUWoQoAAMAGXP4DAACwAaEKAADABoQqAAAAGxCqAAAAbECoAgAAsAGhCgAAwAaEKgAAABsQqgAAAGzw/wAecOfcF7DTowAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.histplot(data = df_clean, x = 'num_words', binwidth = 25)\n",
    "plt.title('Number of words per review')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "346df664",
   "metadata": {},
   "source": [
    "### Calculate quantiles for number of words in a review"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "446c0c70",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1\n",
      "4.0\n",
      "10.0\n",
      "21.0\n",
      "197.0\n",
      "478.0\n",
      "2665\n"
     ]
    }
   ],
   "source": [
    "print(df_clean['num_words'].min())\n",
    "print(df_clean['num_words'].quantile(0.05))\n",
    "print(df_clean['num_words'].quantile(0.25))\n",
    "print(df_clean['num_words'].median())\n",
    "print(df_clean['num_words'].quantile(0.95))\n",
    "print(df_clean['num_words'].quantile(0.99))\n",
    "print(df_clean['num_words'].max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "92287e25",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "<div id=\"altair-viz-81d16b396ace463b8b4b42229802d29e\"></div>\n",
       "<script type=\"text/javascript\">\n",
       "  var VEGA_DEBUG = (typeof VEGA_DEBUG == \"undefined\") ? {} : VEGA_DEBUG;\n",
       "  (function(spec, embedOpt){\n",
       "    let outputDiv = document.currentScript.previousElementSibling;\n",
       "    if (outputDiv.id !== \"altair-viz-81d16b396ace463b8b4b42229802d29e\") {\n",
       "      outputDiv = document.getElementById(\"altair-viz-81d16b396ace463b8b4b42229802d29e\");\n",
       "    }\n",
       "    const paths = {\n",
       "      \"vega\": \"https://cdn.jsdelivr.net/npm//vega@5?noext\",\n",
       "      \"vega-lib\": \"https://cdn.jsdelivr.net/npm//vega-lib?noext\",\n",
       "      \"vega-lite\": \"https://cdn.jsdelivr.net/npm//vega-lite@4.17.0?noext\",\n",
       "      \"vega-embed\": \"https://cdn.jsdelivr.net/npm//vega-embed@6?noext\",\n",
       "    };\n",
       "\n",
       "    function maybeLoadScript(lib, version) {\n",
       "      var key = `${lib.replace(\"-\", \"\")}_version`;\n",
       "      return (VEGA_DEBUG[key] == version) ?\n",
       "        Promise.resolve(paths[lib]) :\n",
       "        new Promise(function(resolve, reject) {\n",
       "          var s = document.createElement('script');\n",
       "          document.getElementsByTagName(\"head\")[0].appendChild(s);\n",
       "          s.async = true;\n",
       "          s.onload = () => {\n",
       "            VEGA_DEBUG[key] = version;\n",
       "            return resolve(paths[lib]);\n",
       "          };\n",
       "          s.onerror = () => reject(`Error loading script: ${paths[lib]}`);\n",
       "          s.src = paths[lib];\n",
       "        });\n",
       "    }\n",
       "\n",
       "    function showError(err) {\n",
       "      outputDiv.innerHTML = `<div class=\"error\" style=\"color:red;\">${err}</div>`;\n",
       "      throw err;\n",
       "    }\n",
       "\n",
       "    function displayChart(vegaEmbed) {\n",
       "      vegaEmbed(outputDiv, spec, embedOpt)\n",
       "        .catch(err => showError(`Javascript Error: ${err.message}<br>This usually means there's a typo in your chart specification. See the javascript console for the full traceback.`));\n",
       "    }\n",
       "\n",
       "    if(typeof define === \"function\" && define.amd) {\n",
       "      requirejs.config({paths});\n",
       "      require([\"vega-embed\"], displayChart, err => showError(`Error loading script: ${err.message}`));\n",
       "    } else {\n",
       "      maybeLoadScript(\"vega\", \"5\")\n",
       "        .then(() => maybeLoadScript(\"vega-lite\", \"4.17.0\"))\n",
       "        .then(() => maybeLoadScript(\"vega-embed\", \"6\"))\n",
       "        .catch(showError)\n",
       "        .then(() => displayChart(vegaEmbed));\n",
       "    }\n",
       "  })({\"config\": {\"view\": {\"continuousWidth\": 400, \"continuousHeight\": 300}, \"axis\": {\"grid\": false}}, \"layer\": [{\"mark\": \"bar\", \"encoding\": {\"x\": {\"axis\": {\"labelAngle\": 0}, \"field\": \"Quantile\", \"type\": \"nominal\"}, \"y\": {\"field\": \"Count of Words\", \"scale\": {\"domain\": [0, 600]}, \"type\": \"quantitative\"}}}, {\"mark\": {\"type\": \"text\", \"align\": \"center\", \"baseline\": \"middle\", \"dy\": -10}, \"encoding\": {\"text\": {\"field\": \"Count of Words\", \"type\": \"quantitative\"}, \"x\": {\"axis\": {\"labelAngle\": 0}, \"field\": \"Quantile\", \"type\": \"nominal\"}, \"y\": {\"field\": \"Count of Words\", \"scale\": {\"domain\": [0, 600]}, \"type\": \"quantitative\"}}}], \"data\": {\"name\": \"data-557e19b436127791c70c6814ea255aff\"}, \"height\": 400, \"title\": \"Number of Words per Quantitle\", \"width\": 600, \"$schema\": \"https://vega.github.io/schema/vega-lite/v4.17.0.json\", \"datasets\": {\"data-557e19b436127791c70c6814ea255aff\": [{\"Quantile\": \".05\", \"Count of Words\": 4.0}, {\"Quantile\": \".25\", \"Count of Words\": 10.0}, {\"Quantile\": \".95\", \"Count of Words\": 197.0}, {\"Quantile\": \".99\", \"Count of Words\": 478.0}]}}, {\"mode\": \"vega-lite\"});\n",
       "</script>"
      ],
      "text/plain": [
       "alt.LayerChart(...)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "my_dict = {'.05':df_clean['num_words'].quantile(0.05),\n",
    "           '.25': df_clean['num_words'].quantile(0.25),\n",
    "           '.95': df_clean['num_words'].quantile(0.95),\n",
    "           '.99': df_clean['num_words'].quantile(0.99)}\n",
    "\n",
    "df_quantile = pd.DataFrame(my_dict.items(), columns=['Quantile','Count of Words'])\n",
    "\n",
    "\n",
    "bars= alt.Chart(df_quantile).mark_bar().encode(\n",
    "    x = alt.X(\"Quantile\", axis = alt.Axis(labelAngle = 0)),\n",
    "    y = alt.Y(\"Count of Words\", scale = alt.Scale(domain=[0,600]))\n",
    ")\n",
    "\n",
    "text = bars.mark_text(\n",
    "    align='center',\n",
    "    baseline='middle',\n",
    "    dy = -10    \n",
    ").encode(\n",
    "    text='Count of Words:Q'\n",
    ")\n",
    "\n",
    "(bars + text).properties(\n",
    "    width = 600,\n",
    "    height = 400,\n",
    "    title = \"Number of Words per Quantitle\"\n",
    ").configure_axis(grid=False)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6b9ee278",
   "metadata": {},
   "source": [
    "### Plot boxenplot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "0165a78a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1500x1200 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(15,12))\n",
    "sns.boxenplot(x = \"num_words\", y = \"app_name\", \n",
    "              data = df_clean)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ef7365b6",
   "metadata": {},
   "source": [
    "### 2.5 Reduce the dataset\n",
    "Since 5% of the reviews have 4 words or less and 95% of reviews have 197 words or less, we will reduce the dataset to only keep reviews than have more than 4 words and at most 197 words. This will also help to remove some of the noise within our reviews."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "f374bc7f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(575223, 7)"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_clean = df_clean[(df_clean['num_words'] >= 5) & (df_clean['num_words'] <= 197)]\n",
    "df_clean.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "13d28332",
   "metadata": {},
   "source": [
    "Let's review the distribution of the number of words in each review after reducing the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "c2a8e93a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.histplot(data = df_clean, x = 'num_words', binwidth = 25)\n",
    "plt.title('Number of words per review')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5038677e",
   "metadata": {},
   "source": [
    "## 3. Setup the language processing pipeline\n",
    "Load the small language library `'en_core_web_sm'` into spaCy."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "e064a1d5",
   "metadata": {},
   "outputs": [],
   "source": [
    "nlp = spacy.load('en_core_web_sm')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b8609998",
   "metadata": {},
   "source": [
    "spaCy takes in raw text and returns back a Doc object which is a sequence of token objects:\n",
    "- Doc - A container for accessing linguistic annotations.\n",
    "- Token - An individual token, i.e., a word, punctuation symbol, whitespace, etc."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0c8acbf0",
   "metadata": {},
   "source": [
    "<img src=\"pipeline.svg\" width=600 />"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d4fdc919",
   "metadata": {},
   "source": [
    "### Visualizing the dependency parse\n",
    "Depenency parsing is a language processing technique that allows us to better determine the meaning of a sentence by analyzing how it’s constructed to determine how the individual words relate to each other. Here we will visualize the syntactic structure of a sentence using spaCy's syntactic dependency visualizer. Let's start by converting a review into a spaCy document."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "1cfa41eb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"it's not a good 'game' for newbies. no, it's not...  i use to sit at a place and watch the sunset. i have no idea how to actually shoot, drive and complete missions.  it's like my personal sunset simulator. and it's damn beautiful.  10/10\""
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_clean['review_text'].iloc[8]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "87b7fb7a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<span class=\"tex2jax_ignore\"><svg xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\" xml:lang=\"en\" id=\"7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0\" class=\"displacy\" width=\"1010\" height=\"257.0\" direction=\"ltr\" style=\"max-width: none; height: 257.0px; color: #000000; background: #ffffff; font-family: Arial; direction: ltr\">\n",
       "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"167.0\">\n",
       "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"50\">children</tspan>\n",
       "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"50\">NOUN</tspan>\n",
       "</text>\n",
       "\n",
       "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"167.0\">\n",
       "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"130\">play</tspan>\n",
       "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"130\">VERB</tspan>\n",
       "</text>\n",
       "\n",
       "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"167.0\">\n",
       "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"210\">call</tspan>\n",
       "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"210\">NOUN</tspan>\n",
       "</text>\n",
       "\n",
       "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"167.0\">\n",
       "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"290\">of</tspan>\n",
       "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"290\">ADP</tspan>\n",
       "</text>\n",
       "\n",
       "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"167.0\">\n",
       "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"370\">duty</tspan>\n",
       "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"370\">NOUN</tspan>\n",
       "</text>\n",
       "\n",
       "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"167.0\">\n",
       "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"450\">boys</tspan>\n",
       "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"450\">NOUN</tspan>\n",
       "</text>\n",
       "\n",
       "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"167.0\">\n",
       "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"530\">play</tspan>\n",
       "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"530\">VERB</tspan>\n",
       "</text>\n",
       "\n",
       "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"167.0\">\n",
       "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"610\">battlefield</tspan>\n",
       "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"610\">NOUN</tspan>\n",
       "</text>\n",
       "\n",
       "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"167.0\">\n",
       "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"690\"> </tspan>\n",
       "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"690\">SPACE</tspan>\n",
       "</text>\n",
       "\n",
       "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"167.0\">\n",
       "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"770\">men</tspan>\n",
       "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"770\">NOUN</tspan>\n",
       "</text>\n",
       "\n",
       "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"167.0\">\n",
       "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"850\">play</tspan>\n",
       "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"850\">VERB</tspan>\n",
       "</text>\n",
       "\n",
       "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"167.0\">\n",
       "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"930\">arma</tspan>\n",
       "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"930\">PROPN</tspan>\n",
       "</text>\n",
       "\n",
       "<g class=\"displacy-arrow\">\n",
       "    <path class=\"displacy-arc\" id=\"arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-0\" stroke-width=\"2px\" d=\"M70,122.0 C70,82.0 120.0,82.0 120.0,122.0\" fill=\"none\" stroke=\"currentColor\"/>\n",
       "    <text dy=\"1.25em\" style=\"font-size: 0.8em; letter-spacing: 1px\">\n",
       "        <textPath xlink:href=\"#arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-0\" class=\"displacy-label\" startOffset=\"50%\" side=\"left\" fill=\"currentColor\" text-anchor=\"middle\">nsubj</textPath>\n",
       "    </text>\n",
       "    <path class=\"displacy-arrowhead\" d=\"M70,124.0 L62,112.0 78,112.0\" fill=\"currentColor\"/>\n",
       "</g>\n",
       "\n",
       "<g class=\"displacy-arrow\">\n",
       "    <path class=\"displacy-arc\" id=\"arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-1\" stroke-width=\"2px\" d=\"M150,122.0 C150,82.0 200.0,82.0 200.0,122.0\" fill=\"none\" stroke=\"currentColor\"/>\n",
       "    <text dy=\"1.25em\" style=\"font-size: 0.8em; letter-spacing: 1px\">\n",
       "        <textPath xlink:href=\"#arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-1\" class=\"displacy-label\" startOffset=\"50%\" side=\"left\" fill=\"currentColor\" text-anchor=\"middle\">dobj</textPath>\n",
       "    </text>\n",
       "    <path class=\"displacy-arrowhead\" d=\"M200.0,124.0 L208.0,112.0 192.0,112.0\" fill=\"currentColor\"/>\n",
       "</g>\n",
       "\n",
       "<g class=\"displacy-arrow\">\n",
       "    <path class=\"displacy-arc\" id=\"arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-2\" stroke-width=\"2px\" d=\"M230,122.0 C230,82.0 280.0,82.0 280.0,122.0\" fill=\"none\" stroke=\"currentColor\"/>\n",
       "    <text dy=\"1.25em\" style=\"font-size: 0.8em; letter-spacing: 1px\">\n",
       "        <textPath xlink:href=\"#arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-2\" class=\"displacy-label\" startOffset=\"50%\" side=\"left\" fill=\"currentColor\" text-anchor=\"middle\">prep</textPath>\n",
       "    </text>\n",
       "    <path class=\"displacy-arrowhead\" d=\"M280.0,124.0 L288.0,112.0 272.0,112.0\" fill=\"currentColor\"/>\n",
       "</g>\n",
       "\n",
       "<g class=\"displacy-arrow\">\n",
       "    <path class=\"displacy-arc\" id=\"arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-3\" stroke-width=\"2px\" d=\"M390,122.0 C390,82.0 440.0,82.0 440.0,122.0\" fill=\"none\" stroke=\"currentColor\"/>\n",
       "    <text dy=\"1.25em\" style=\"font-size: 0.8em; letter-spacing: 1px\">\n",
       "        <textPath xlink:href=\"#arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-3\" class=\"displacy-label\" startOffset=\"50%\" side=\"left\" fill=\"currentColor\" text-anchor=\"middle\">compound</textPath>\n",
       "    </text>\n",
       "    <path class=\"displacy-arrowhead\" d=\"M390,124.0 L382,112.0 398,112.0\" fill=\"currentColor\"/>\n",
       "</g>\n",
       "\n",
       "<g class=\"displacy-arrow\">\n",
       "    <path class=\"displacy-arc\" id=\"arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-4\" stroke-width=\"2px\" d=\"M310,122.0 C310,42.0 445.0,42.0 445.0,122.0\" fill=\"none\" stroke=\"currentColor\"/>\n",
       "    <text dy=\"1.25em\" style=\"font-size: 0.8em; letter-spacing: 1px\">\n",
       "        <textPath xlink:href=\"#arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-4\" class=\"displacy-label\" startOffset=\"50%\" side=\"left\" fill=\"currentColor\" text-anchor=\"middle\">pobj</textPath>\n",
       "    </text>\n",
       "    <path class=\"displacy-arrowhead\" d=\"M445.0,124.0 L453.0,112.0 437.0,112.0\" fill=\"currentColor\"/>\n",
       "</g>\n",
       "\n",
       "<g class=\"displacy-arrow\">\n",
       "    <path class=\"displacy-arc\" id=\"arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-5\" stroke-width=\"2px\" d=\"M150,122.0 C150,2.0 530.0,2.0 530.0,122.0\" fill=\"none\" stroke=\"currentColor\"/>\n",
       "    <text dy=\"1.25em\" style=\"font-size: 0.8em; letter-spacing: 1px\">\n",
       "        <textPath xlink:href=\"#arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-5\" class=\"displacy-label\" startOffset=\"50%\" side=\"left\" fill=\"currentColor\" text-anchor=\"middle\">advcl</textPath>\n",
       "    </text>\n",
       "    <path class=\"displacy-arrowhead\" d=\"M530.0,124.0 L538.0,112.0 522.0,112.0\" fill=\"currentColor\"/>\n",
       "</g>\n",
       "\n",
       "<g class=\"displacy-arrow\">\n",
       "    <path class=\"displacy-arc\" id=\"arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-6\" stroke-width=\"2px\" d=\"M630,122.0 C630,42.0 765.0,42.0 765.0,122.0\" fill=\"none\" stroke=\"currentColor\"/>\n",
       "    <text dy=\"1.25em\" style=\"font-size: 0.8em; letter-spacing: 1px\">\n",
       "        <textPath xlink:href=\"#arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-6\" class=\"displacy-label\" startOffset=\"50%\" side=\"left\" fill=\"currentColor\" text-anchor=\"middle\">compound</textPath>\n",
       "    </text>\n",
       "    <path class=\"displacy-arrowhead\" d=\"M630,124.0 L622,112.0 638,112.0\" fill=\"currentColor\"/>\n",
       "</g>\n",
       "\n",
       "<g class=\"displacy-arrow\">\n",
       "    <path class=\"displacy-arc\" id=\"arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-7\" stroke-width=\"2px\" d=\"M630,122.0 C630,82.0 680.0,82.0 680.0,122.0\" fill=\"none\" stroke=\"currentColor\"/>\n",
       "    <text dy=\"1.25em\" style=\"font-size: 0.8em; letter-spacing: 1px\">\n",
       "        <textPath xlink:href=\"#arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-7\" class=\"displacy-label\" startOffset=\"50%\" side=\"left\" fill=\"currentColor\" text-anchor=\"middle\">dep</textPath>\n",
       "    </text>\n",
       "    <path class=\"displacy-arrowhead\" d=\"M680.0,124.0 L688.0,112.0 672.0,112.0\" fill=\"currentColor\"/>\n",
       "</g>\n",
       "\n",
       "<g class=\"displacy-arrow\">\n",
       "    <path class=\"displacy-arc\" id=\"arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-8\" stroke-width=\"2px\" d=\"M790,122.0 C790,82.0 840.0,82.0 840.0,122.0\" fill=\"none\" stroke=\"currentColor\"/>\n",
       "    <text dy=\"1.25em\" style=\"font-size: 0.8em; letter-spacing: 1px\">\n",
       "        <textPath xlink:href=\"#arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-8\" class=\"displacy-label\" startOffset=\"50%\" side=\"left\" fill=\"currentColor\" text-anchor=\"middle\">nsubj</textPath>\n",
       "    </text>\n",
       "    <path class=\"displacy-arrowhead\" d=\"M790,124.0 L782,112.0 798,112.0\" fill=\"currentColor\"/>\n",
       "</g>\n",
       "\n",
       "<g class=\"displacy-arrow\">\n",
       "    <path class=\"displacy-arc\" id=\"arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-9\" stroke-width=\"2px\" d=\"M550,122.0 C550,2.0 850.0,2.0 850.0,122.0\" fill=\"none\" stroke=\"currentColor\"/>\n",
       "    <text dy=\"1.25em\" style=\"font-size: 0.8em; letter-spacing: 1px\">\n",
       "        <textPath xlink:href=\"#arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-9\" class=\"displacy-label\" startOffset=\"50%\" side=\"left\" fill=\"currentColor\" text-anchor=\"middle\">ccomp</textPath>\n",
       "    </text>\n",
       "    <path class=\"displacy-arrowhead\" d=\"M850.0,124.0 L858.0,112.0 842.0,112.0\" fill=\"currentColor\"/>\n",
       "</g>\n",
       "\n",
       "<g class=\"displacy-arrow\">\n",
       "    <path class=\"displacy-arc\" id=\"arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-10\" stroke-width=\"2px\" d=\"M870,122.0 C870,82.0 920.0,82.0 920.0,122.0\" fill=\"none\" stroke=\"currentColor\"/>\n",
       "    <text dy=\"1.25em\" style=\"font-size: 0.8em; letter-spacing: 1px\">\n",
       "        <textPath xlink:href=\"#arrow-7e6d4bc8db8b4488b2ae4f8aa6af4b4f-0-10\" class=\"displacy-label\" startOffset=\"50%\" side=\"left\" fill=\"currentColor\" text-anchor=\"middle\">dobj</textPath>\n",
       "    </text>\n",
       "    <path class=\"displacy-arrowhead\" d=\"M920.0,124.0 L928.0,112.0 912.0,112.0\" fill=\"currentColor\"/>\n",
       "</g>\n",
       "</svg></span>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "doc = nlp(df_clean['review_text'].iloc[7])\n",
    "displacy.render(doc, style = 'dep', jupyter = True, options = {'distance':80})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "24c3082c",
   "metadata": {},
   "source": [
    "### Visualizing the entity recognizer\n",
    "Visualize entities in the text of the first review."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "8126c510",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<span class=\"tex2jax_ignore\"><div class=\"entities\" style=\"line-height: 2.5; direction: ltr\">drive in a tempest truck at \n",
       "<mark class=\"entity\" style=\"background: #e4e7d2; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
       "    100 km/h. friend desyncs\n",
       "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">QUANTITY</span>\n",
       "</mark>\n",
       ". slam into back of his truck. flip... sit there for a \n",
       "<mark class=\"entity\" style=\"background: #e4e7d2; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
       "    second\n",
       "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">ORDINAL</span>\n",
       "</mark>\n",
       "... 'you're still ok' a friend says... ...and bam! i blow up with a \n",
       "<mark class=\"entity\" style=\"background: #bfeeb7; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
       "    300k\n",
       "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">PRODUCT</span>\n",
       "</mark>\n",
       " truck with \n",
       "<mark class=\"entity\" style=\"background: #bfeeb7; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
       "    300k\n",
       "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">PRODUCT</span>\n",
       "</mark>\n",
       " worth of oil. scream alt+f4 while mashing keyboard. launch \n",
       "<mark class=\"entity\" style=\"background: #aa9cfc; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
       "    arma spawn\n",
       "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">PERSON</span>\n",
       "</mark>\n",
       " with all my gear and oil in truck. 10/10 would play again</div></span>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "doc = nlp(df_clean['review_text'].iloc[100])\n",
    "displacy.render(doc, style = 'ent', jupyter = True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "52f134c1",
   "metadata": {},
   "source": [
    "## 4. Tokenize the text using processing pipeline\n",
    "The `nlp.pipe` object will be used to tokenize the clean text, lemmatize each token, and return the token if it's not stopword. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "d7a908ea",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(575223, 7)"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_clean.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "79894c58",
   "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>app_id</th>\n",
       "      <th>app_name</th>\n",
       "      <th>review_text</th>\n",
       "      <th>review_score</th>\n",
       "      <th>review_votes</th>\n",
       "      <th>clean_text</th>\n",
       "      <th>num_words</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>i spent several years in the army and i gotta ...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>i spent several years in the army and i gotta ...</td>\n",
       "      <td>179</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>having accumulated over 13400 hours in-game wi...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>having accumulated over hours in game with arm...</td>\n",
       "      <td>40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>joined exile sever. found a 8 year old kid. ki...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>joined exile sever found a year old kid kid sa...</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>after some 1,400 hours on the game, i can say ...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>after some hours on the game i can say with fu...</td>\n",
       "      <td>34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>this game made me realize that 15fps is playable.</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>this game made me realize that fps is playable</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   app_id app_name                                        review_text  \\\n",
       "3  107410   Arma 3  i spent several years in the army and i gotta ...   \n",
       "4  107410   Arma 3  having accumulated over 13400 hours in-game wi...   \n",
       "6  107410   Arma 3  joined exile sever. found a 8 year old kid. ki...   \n",
       "7  107410   Arma 3  after some 1,400 hours on the game, i can say ...   \n",
       "8  107410   Arma 3  this game made me realize that 15fps is playable.   \n",
       "\n",
       "   review_score  review_votes  \\\n",
       "3             1             1   \n",
       "4             1             1   \n",
       "6             1             0   \n",
       "7             1             1   \n",
       "8             1             0   \n",
       "\n",
       "                                          clean_text  num_words  \n",
       "3  i spent several years in the army and i gotta ...        179  \n",
       "4  having accumulated over hours in game with arm...         40  \n",
       "6  joined exile sever found a year old kid kid sa...         59  \n",
       "7  after some hours on the game i can say with fu...         34  \n",
       "8     this game made me realize that fps is playable          9  "
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_clean.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "0cf10d6a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['tok2vec', 'tagger', 'parser', 'attribute_ruler', 'lemmatizer', 'ner']"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nlp.pipe_names"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "091945c4",
   "metadata": {},
   "source": [
    "### 4.1 Use parallelization job to preprocess the reviews\n",
    "We create a function to lemmatize each token in each doc and remove the token if it's a stopword.  Each doc is a review so we will be creating approximately 575K documents and generating tokens for each document since we have 575K reviews.\n",
    "\n",
    "When creating the `nlp.pip` object we are going to disable the parser and ner (named entity recognition) because they are computationally expensive and not necessary for our task."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "943d6585",
   "metadata": {},
   "outputs": [],
   "source": [
    "# get a set of stop words from spaCy nlp object\n",
    "stopwords = nlp.Defaults.stop_words\n",
    "\n",
    "def lemmatization(texts, allowed_postags=['NOUN','ADJ','VERB','ADV']):\n",
    "    '''\n",
    "        Returns a list of lemmatized texts\n",
    "        \n",
    "        Parameters:\n",
    "            texts: a list of strings\n",
    "        \n",
    "        Returns:\n",
    "            texts_out: a list of tokenized strings\n",
    "    '''\n",
    "    \n",
    "    texts_out = []\n",
    "    for doc in nlp.pipe(texts, disable=[\"parser\", \"ner\"], batch_size=1000, n_process=4):\n",
    "        new_text = []\n",
    "        for token in doc:\n",
    "            if token.pos_ in allowed_postags and token not in stopwords:\n",
    "                new_text.append(token.lemma_)\n",
    "        final = \" \".join(new_text)\n",
    "        texts_out.append(final)\n",
    "    return texts_out"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7621eb47",
   "metadata": {},
   "source": [
    "Test the lemmatization function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "03df1e18",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['spend several year army get say game spot training atmosphere military provide love difficult game love different feature game provide easy hit target assault rifle yard real life game seem match same difficulty fatigue real thing seriously hinder accuracy real life make stop catch breath run too far carry too much gear hard time hit target enemy stupid great vision engage enemy yard away use bush cover as well second rush move ai more advanced average taliban real life haha p give love military videogame one definitely favorite time']"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lemmatization(df_clean.iloc[:1]['clean_text'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8a14938c",
   "metadata": {},
   "source": [
    "Create a small dataset to test out parallelization job."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "177a78ee",
   "metadata": {},
   "outputs": [],
   "source": [
    "# create a copy to avoid SettingWithCopyWarning\n",
    "df_clean_small = df_clean.iloc[:8000].copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "dbd28483",
   "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>app_id</th>\n",
       "      <th>app_name</th>\n",
       "      <th>review_text</th>\n",
       "      <th>review_score</th>\n",
       "      <th>review_votes</th>\n",
       "      <th>clean_text</th>\n",
       "      <th>num_words</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>i spent several years in the army and i gotta ...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>i spent several years in the army and i gotta ...</td>\n",
       "      <td>179</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>having accumulated over 13400 hours in-game wi...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>having accumulated over hours in game with arm...</td>\n",
       "      <td>40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>joined exile sever. found a 8 year old kid. ki...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>joined exile sever found a year old kid kid sa...</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>after some 1,400 hours on the game, i can say ...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>after some hours on the game i can say with fu...</td>\n",
       "      <td>34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>this game made me realize that 15fps is playable.</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>this game made me realize that fps is playable</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   app_id app_name                                        review_text  \\\n",
       "3  107410   Arma 3  i spent several years in the army and i gotta ...   \n",
       "4  107410   Arma 3  having accumulated over 13400 hours in-game wi...   \n",
       "6  107410   Arma 3  joined exile sever. found a 8 year old kid. ki...   \n",
       "7  107410   Arma 3  after some 1,400 hours on the game, i can say ...   \n",
       "8  107410   Arma 3  this game made me realize that 15fps is playable.   \n",
       "\n",
       "   review_score  review_votes  \\\n",
       "3             1             1   \n",
       "4             1             1   \n",
       "6             1             0   \n",
       "7             1             1   \n",
       "8             1             0   \n",
       "\n",
       "                                          clean_text  num_words  \n",
       "3  i spent several years in the army and i gotta ...        179  \n",
       "4  having accumulated over hours in game with arm...         40  \n",
       "6  joined exile sever found a year old kid kid sa...         59  \n",
       "7  after some hours on the game i can say with fu...         34  \n",
       "8     this game made me realize that fps is playable          9  "
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_clean_small.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "00ebf70f",
   "metadata": {},
   "source": [
    "The script below used to create a parallelization job can be attributed to Prashanth Rao who posted a blog on how to speed up nlp pipeline processing using spaCy.  The blog can be found at:  https://prrao87.github.io/blog/spacy/nlp/performance/2020/05/02/spacy-multiprocess.html"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "1c92b981",
   "metadata": {},
   "outputs": [],
   "source": [
    "from joblib import Parallel, delayed\n",
    "\n",
    "def lemmatize_pipe(doc, allowed_postags=['NOUN','ADJ','VERB','ADV']):\n",
    "    lemma_list = [token.lemma_ for token in doc\n",
    "                  if token.pos_ in allowed_postags and token not in stopwords] \n",
    "    return lemma_list\n",
    "\n",
    "def chunker(iterable, total_length, chunksize):\n",
    "    return (iterable[pos: pos + chunksize] for pos in range(0, total_length, chunksize))\n",
    "\n",
    "def flatten(list_of_lists):\n",
    "    \"Flatten a list of lists to a combined list\"\n",
    "    return [item for sublist in list_of_lists for item in sublist]\n",
    "\n",
    "def process_chunk(texts):\n",
    "    preproc_pipe = []\n",
    "    for doc in nlp.pipe(texts, disable=[\"parser\", \"ner\"], batch_size=20):\n",
    "        lemma_string = ' '.join(lemmatize_pipe(doc))\n",
    "        preproc_pipe.append(lemma_string)\n",
    "    return preproc_pipe\n",
    "\n",
    "def parallel_lemmatization_pipeline(texts, chunksize=100):\n",
    "    executor = Parallel(n_jobs=7, backend='multiprocessing', prefer=\"processes\")\n",
    "    do = delayed(process_chunk)\n",
    "    tasks = (do(chunk) for chunk in chunker(texts, len(df_clean), chunksize=chunksize))\n",
    "    result = executor(tasks)\n",
    "    return flatten(result)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "15c1e771",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 4.37 s, sys: 146 ms, total: 4.52 s\n",
      "Wall time: 8.04 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "df_clean_small['lemmatization'] = lemmatization(df_clean_small['clean_text'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "00bb44db",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 305 ms, sys: 202 ms, total: 507 ms\n",
      "Wall time: 4.61 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "df_clean_small['preproc_parallel'] = parallel_lemmatization_pipeline(df_clean_small['clean_text'], chunksize=1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "7eb49ac7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'spend several year army get say game spot training atmosphere military provide love difficult game love different feature game provide easy hit target assault rifle yard real life game seem match same difficulty fatigue real thing seriously hinder accuracy real life make stop catch breath run too far carry too much gear hard time hit target enemy stupid great vision engage enemy yard away use bush cover as well second rush move ai more advanced average taliban real life haha p give love military videogame one definitely favorite time'"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_clean_small['preproc_parallel'].iloc[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5c90cd71",
   "metadata": {},
   "source": [
    "### 4.2 Apply lemmatization function to clean text\n",
    "Apply the preprocessing nlp pipeline on the `clean_text` column and generate a new column of lemmatized tokens as a string."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "fef1a9f6",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# create a new column called clean_tokens and apply preprocess_pipe function to clean_text\n",
    "df_clean['clean_tokens'] = parallel_lemmatization_pipeline(df_clean['clean_text'], chunksize=1000)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "802f3dc9",
   "metadata": {},
   "source": [
    "Review processed token strings."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "73a87f1c",
   "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>app_id</th>\n",
       "      <th>app_name</th>\n",
       "      <th>review_text</th>\n",
       "      <th>review_score</th>\n",
       "      <th>review_votes</th>\n",
       "      <th>clean_text</th>\n",
       "      <th>num_words</th>\n",
       "      <th>clean_tokens</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>i spent several years in the army and i gotta ...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>i spent several years in the army and i gotta ...</td>\n",
       "      <td>179</td>\n",
       "      <td>spend several year army get say game spot trai...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>having accumulated over 13400 hours in-game wi...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>having accumulated over hours in game with arm...</td>\n",
       "      <td>40</td>\n",
       "      <td>have accumulate hour game certainly recommend ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>joined exile sever. found a 8 year old kid. ki...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>joined exile sever found a year old kid kid sa...</td>\n",
       "      <td>59</td>\n",
       "      <td>join exile sever find year old kid kid say m l...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>after some 1,400 hours on the game, i can say ...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>after some hours on the game i can say with fu...</td>\n",
       "      <td>34</td>\n",
       "      <td>hour game say full validity game ruin life lik...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>107410</td>\n",
       "      <td>Arma 3</td>\n",
       "      <td>this game made me realize that 15fps is playable.</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>this game made me realize that fps is playable</td>\n",
       "      <td>9</td>\n",
       "      <td>game make realize fps playable</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   app_id app_name                                        review_text  \\\n",
       "3  107410   Arma 3  i spent several years in the army and i gotta ...   \n",
       "4  107410   Arma 3  having accumulated over 13400 hours in-game wi...   \n",
       "6  107410   Arma 3  joined exile sever. found a 8 year old kid. ki...   \n",
       "7  107410   Arma 3  after some 1,400 hours on the game, i can say ...   \n",
       "8  107410   Arma 3  this game made me realize that 15fps is playable.   \n",
       "\n",
       "   review_score  review_votes  \\\n",
       "3             1             1   \n",
       "4             1             1   \n",
       "6             1             0   \n",
       "7             1             1   \n",
       "8             1             0   \n",
       "\n",
       "                                          clean_text  num_words  \\\n",
       "3  i spent several years in the army and i gotta ...        179   \n",
       "4  having accumulated over hours in game with arm...         40   \n",
       "6  joined exile sever found a year old kid kid sa...         59   \n",
       "7  after some hours on the game i can say with fu...         34   \n",
       "8     this game made me realize that fps is playable          9   \n",
       "\n",
       "                                        clean_tokens  \n",
       "3  spend several year army get say game spot trai...  \n",
       "4  have accumulate hour game certainly recommend ...  \n",
       "6  join exile sever find year old kid kid say m l...  \n",
       "7  hour game say full validity game ruin life lik...  \n",
       "8                     game make realize fps playable  "
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_clean.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b63dbe11",
   "metadata": {},
   "source": [
    "## 5. Upload data to S3 bucket"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3f41dabe",
   "metadata": {},
   "source": [
    "Write to csv and include original review text."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "4158bb7f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'paths': ['s3://siads696-wi23-steam-data/clean_data/reviews_clean_new_full.csv'],\n",
       " 'partitions_values': {}}"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cols = ['app_id','app_name','review_score','review_votes','num_words','clean_tokens','review_text']\n",
    "\n",
    "wr.s3.to_csv(\n",
    "    df=df_clean[cols],\n",
    "    path='s3://siads696-wi23-steam-data/clean_data/reviews_clean_new_full.csv',\n",
    "    index=False\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "283ac680",
   "metadata": {},
   "source": [
    "Write to csv that will be used for Redshift database."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "f674e526",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'paths': ['s3://siads696-wi23-steam-data/clean_data/reviews_clean_new.csv'],\n",
       " 'partitions_values': {}}"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cols = ['app_id','app_name','review_score','review_votes','num_words','clean_tokens']\n",
    "\n",
    "wr.s3.to_csv(\n",
    "    df=df_clean[cols],\n",
    "    path='s3://siads696-wi23-steam-data/clean_data/reviews_clean_new.csv',\n",
    "    index=False\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e8fa6fe9",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}