Kaggle-Outbrain-PageViews_EventsAnalytics.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Kaggle - Outbrain Click Prediction\n",
    "## Unveiling the secrets of page_views.csv and its relationship with events.csv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "By [Gabriel S. P. Moreira](https://about.me/gspmoreira)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this notebook, I leverage the power of a **Spark cluster** to explore the large (~88GB) page_views.csv dataset and analyze its relationshiop with events.csv in the [Kaggle - Outbrain Click Prediction challenge](https://www.kaggle.com/c/outbrain-click-prediction/) .  \n",
    "We answer questions like: \n",
    "\n",
    "* **How to join page_views.csv and events.csv?**\n",
    "* **Is events.csv a subset of page_views.csv?**\n",
    "* **Are there additional page views for users in events.csv?**  \n",
    "\n",
    "\n",
    "*If this Kaggle kernel helps you, don't forget to thumbs it up! ;)*"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dependencies"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* PySpark 2.0\n",
    "* Pandas and Matplotlib - Only on master node\n",
    "\n",
    "Just for reference, I am running this notebook on a Spark cluster with 1 master and 4 worker nodes (deployed on [Google Dataproc](http://cloud.google.com/dataproc/) ), each server with 4 CPUs and 16GB RAM.  \n",
    "For some slower Spark transformations and actions, I've used the %time Jupyter magic command to log elapsed processing time."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Analyzing page views"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading page_views.csv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from pyspark.sql.types import *\n",
    "import pyspark.sql.functions as F"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "page_views_schema = StructType(\n",
    "                    [StructField(\"uuid\", StringType(), True),\n",
    "                    StructField(\"document_id\", IntegerType(), True),\n",
    "                    StructField(\"timestamp\", IntegerType(), True),\n",
    "                    StructField(\"platform\", IntegerType(), True),\n",
    "                    StructField(\"geo_location\", StringType(), True),\n",
    "                    StructField(\"traffic_source\", IntegerType(), True)]\n",
    "                    )\n",
    "#Data is read from Google Cloud Storage, a distributed filesystem like HDFS stored externally to the cluster\n",
    "#If you wanna use your own HDFS, just prefix the file path with the absolute path 'hdfs:' or \n",
    "#with the relative path if HDFS is your cluster default filesystem\n",
    "page_views_df = spark.read.schema(page_views_schema).options(header='true', inferschema='false', nullValue='\\\\N') \\\n",
    "                .csv(\"gs://sc-ml-hdfs-prd/tmp/outbrain/page_views.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Spark sets the number of partitions based on file size, in this case, 708 partitions, each with about ~125MB. During job processing each partition becomes a task, which are processed in parallel by the available executors on Spark worker nodes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "708"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "page_views_df.rdd.getNumPartitions()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+--------------+-----------+---------+--------+------------+--------------+\n",
      "|          uuid|document_id|timestamp|platform|geo_location|traffic_source|\n",
      "+--------------+-----------+---------+--------+------------+--------------+\n",
      "|1fd5f051fba643|        120| 31905835|       1|          RS|             2|\n",
      "|8557aa9004be3b|        120| 32053104|       1|       VN>44|             2|\n",
      "|c351b277a358f0|        120| 54013023|       1|       KR>12|             1|\n",
      "|8205775c5387f9|        120| 44196592|       1|       IN>16|             2|\n",
      "|9cb0ccd8458371|        120| 65817371|       1|   US>CA>807|             2|\n",
      "|2aa611f32875c7|        120| 71495491|       1|       CA>ON|             2|\n",
      "|f55a6eaf2b34ab|        120| 73309199|       1|       BR>27|             2|\n",
      "|cc01b582c8cbff|        120| 50033577|       1|       CA>BC|             2|\n",
      "|6c802978b8dd4d|        120| 66590306|       1|       CA>ON|             2|\n",
      "|f4e423314303ff|        120| 48314254|       1|   US>LA>622|             1|\n",
      "|3937372ca2709b|        120| 24360074|       1|          NO|             2|\n",
      "|31f8d101c6a851|        120| 13847456|       1|       PH>D9|             2|\n",
      "|67606983fe1acf|        984| 82129416|       2|   US>OK>671|             2|\n",
      "|3f9d5b09ac4a0a|        984| 82583638|       2|          US|             2|\n",
      "|3dd8a359aa8699|        984| 59439284|       2|   US>MD>511|             2|\n",
      "|95e966c81b9316|        984|   983791|       3|   US>AL>698|             2|\n",
      "|483cf1feb3e47c|        984| 17944009|       2|   US>MI>505|             1|\n",
      "|f7baf0caf201ca|        984| 37096677|       2|       CA>AB|             2|\n",
      "|7a408fe90c02b3|        984| 30457643|       1|          PH|             1|\n",
      "|9ee2877617838a|        984| 86093121|       3|   US>MN>613|             2|\n",
      "+--------------+-----------+---------+--------+------------+--------------+\n",
      "only showing top 20 rows\n",
      "\n",
      "CPU times: user 0 ns, sys: 0 ns, total: 0 ns\n",
      "Wall time: 516 ms\n"
     ]
    }
   ],
   "source": [
    "%time page_views_df.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Page views timestamp stats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 60 ms, sys: 20 ms, total: 80 ms\n",
      "Wall time: 5min 24s\n"
     ]
    }
   ],
   "source": [
    "%time page_views_timestamp_stats_df = page_views_df.describe('timestamp')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'count': '2034275448',\n",
       " 'max': '1296000000',\n",
       " 'mean': '6.51464604736012E8',\n",
       " 'min': '0',\n",
       " 'stddev': '3.809745454636914E8'}"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "page_views_timestamp_stats = page_views_timestamp_stats_df.rdd.collectAsMap()\n",
    "page_views_timestamp_stats"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Total of page views available in this dataset**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "# Page Views: 2034275448\n"
     ]
    }
   ],
   "source": [
    "page_views_count = int(page_views_timestamp_stats['count'])\n",
    "print('# Page Views:',page_views_count)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "** Page Views time range **"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see below that dataset contains only page views between 2016-06-14 and 2016-06-28. As min and max timestamp in dataset are at about 4am, it suggests that dates are in UTC and the dataset days were filtered according to UTC-4 (Eastern Time - EST).   \n",
    "To further model users preferences, it would be nice if we had in this dataset the full log of previous users pages visits, who had at least an event between that period (2016-06-14 and 2016-06-28)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from datetime import datetime\n",
    "def convert_odd_timestamp(timestamp_ms_relative):\n",
    "    TIMESTAMP_DELTA=1465876799998\n",
    "    return datetime.fromtimestamp((int(timestamp_ms_relative)+TIMESTAMP_DELTA)//1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2016-06-14 03:59:59\n"
     ]
    }
   ],
   "source": [
    "print(convert_odd_timestamp(page_views_timestamp_stats['min']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2016-06-29 03:59:59\n"
     ]
    }
   ],
   "source": [
    "print(convert_odd_timestamp(page_views_timestamp_stats['max']))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Page views by user"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#Grouping page views by user id and aggregating the count by user\n",
    "page_views_by_user_df = page_views_df.groupBy('uuid').count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#Grouping page views user counts by counts (to build a histogram)\n",
    "page_views_user_distribution_df = page_views_by_user_df.select('count').groupby('count').count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 412 ms, sys: 88 ms, total: 500 ms\n",
      "Wall time: 16min 4s\n"
     ]
    }
   ],
   "source": [
    "#Converting the PySpark Dataframe to a Pandas Dataframe (for plotting)\n",
    "%time page_views_by_user_pandas = page_views_user_distribution_df.toPandas()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "563"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(page_views_by_user_pandas)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This table shows that 458234809 users have only 1 page view in the dataset, 94245789 have exactly 2 page views (94245789), etc."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>usersCount</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pageViewsByUser</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>458234809</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>94245789</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>44977389</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>26869230</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>17942534</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>12878964</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>9691732</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>7573614</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>6036047</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>5003689</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 usersCount\n",
       "pageViewsByUser            \n",
       "1                 458234809\n",
       "2                  94245789\n",
       "3                  44977389\n",
       "4                  26869230\n",
       "5                  17942534\n",
       "6                  12878964\n",
       "7                   9691732\n",
       "8                   7573614\n",
       "9                   6036047\n",
       "10                  5003689"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "page_views_by_user_pandas.columns = ['pageViewsByUser', 'usersCount']\n",
    "page_views_by_user_pandas = page_views_by_user_pandas.set_index('pageViewsByUser').sort_index()\n",
    "page_views_by_user_pandas.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Total number of distinct users in page views**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "717506889"
      ]
     },
     "execution_count": 167,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distinct_users_count = page_views_by_user_pandas['usersCount'].sum()\n",
    "distinct_users_count"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The **Average page views by user** indicates that we have higher average (**2.835**) in **page views** dataset, when compared to the **events dataset** (**1.167**)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.8351998833561027"
      ]
     },
     "execution_count": 171,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "avg_page_views_by_user = page_views_count / float(distinct_users_count)\n",
    "avg_page_views_by_user"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Plotting the number of users (y axis) by the count of their page views (x axis). We can see that most users (458234809 ~ 5*10^8) have only 1 page view during the period."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fe6ec45a358>"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfMAAAGFCAYAAADpdBqXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe6ec462240>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "page_views_by_user_pandas.head(20).plot.bar(y='usersCount')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Calculating the cumulative percentual of users with at most N page views"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "page_views_by_user_pandas['cum_perc'] = (page_views_by_user_pandas['usersCount'].cumsum() / distinct_users_count) * 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>usersCount</th>\n",
       "      <th>cum_perc</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pageViewsByUser</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>458234809</td>\n",
       "      <td>63.864865</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>94245789</td>\n",
       "      <td>77.000041</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>44977389</td>\n",
       "      <td>83.268606</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>26869230</td>\n",
       "      <td>87.013411</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>17942534</td>\n",
       "      <td>89.514088</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>12878964</td>\n",
       "      <td>91.309049</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>9691732</td>\n",
       "      <td>92.659800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>7573614</td>\n",
       "      <td>93.715346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>6036047</td>\n",
       "      <td>94.556598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>5003689</td>\n",
       "      <td>95.253970</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 usersCount   cum_perc\n",
       "pageViewsByUser                       \n",
       "1                 458234809  63.864865\n",
       "2                  94245789  77.000041\n",
       "3                  44977389  83.268606\n",
       "4                  26869230  87.013411\n",
       "5                  17942534  89.514088\n",
       "6                  12878964  91.309049\n",
       "7                   9691732  92.659800\n",
       "8                   7573614  93.715346\n",
       "9                   6036047  94.556598\n",
       "10                  5003689  95.253970"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "page_views_by_user_pandas.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In cumulative chart, we can see that 65% of the users have only 1 page view, 77%, have at most 2 page views and 89% of the users have at most 5 page views. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fe6ec6444a8>"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAg4AAAF5CAYAAAD3dKLdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAAPYQAAD2EBqD+naQAAIABJREFUeJzt3Xt8FNX9//HXBxAQkKCggCCCgIBWkAQFFC0KamuLt9ZL\nrAKiP7DWttLvt/WGipeq1Sr1gtavFxTUWMQqVlGKiCYiohAURe4XUW4agSBErjm/P86GbJYQNmQn\nk919Px+PeSQ7c3bmkyxh33vmzBlzziEiIiISj1phFyAiIiLJQ8FBRERE4qbgICIiInFTcBAREZG4\nKTiIiIhI3BQcREREJG4KDiIiIhI3BQcRERGJm4KDiIiIxE3BQUREROJW6eBgZqeY2etmtsrMis3s\nnHLa3GFmq82syMymmFmHmO31zGy0mRWY2Q9mNsHMDqvKDyIiIiLB258eh4bAp8A1wB43ujCz64Fr\ngaHAicAWYLKZ1Y1q9g/gF8CvgFOBw4FX9qMWERERqUZWlZtcmVkxcJ5z7vWodauB+51zoyKPGwPr\ngEHOufGRx98BlzjnXo206QTMB3o55z7e74JEREQkUAkd42Bm7YAWwNSSdc65TcBMoHdkVQ+gTkyb\nhcDKqDYiIiJSAyV6cGQL/OmLdTHr10W2ATQHtkcCxd7aiIiISA1UJ+wC4mFmTYGzgBXA1nCrERER\nSSr1gbbAZOfc91XdWaKDw1rA8L0K0b0OzYE5UW3qmlnjmF6H5pFt5TkLeCHBtYqIiKST3wAvVnUn\nCQ0OzrnlZrYW6AfMhd2DI3sCoyPNZgM7I22iB0e2AWbsZdcrAJ5//nm6dOmSyJKlkoYPH86oUaPC\nLkMCptc5fei1Tn3z58/nsssug8h7aVVVOjiYWUOgA75nAeAoM+sGrHfOfY2/1HKEmS2JFHkn8A0w\nEfxgSTN7GnjQzDYAPwAPA9MruKJiK0CXLl3IzMysbMmSQBkZGXoN0oBe5/Sh1zqtJORU//70OPQA\npuEHQTrggcj654Ahzrn7zKwB8ATQBMgDfu6c2x61j+HALmACUA94G/jdfv0EIiIiUm0qHRycc++z\nj6sxnHMjgZEVbN8G/D6yiIiISJLQvSpEREQkbgoOUinZ2dlhlyDVQK9z+tBrLZWVFPM4xGvlypUU\nFBSEXUZK69SpE/n5+WGXkTDNmjWjTZs2YZdR4+jNJH3otZbKSpngsHLlSrp06UJRUVHYpUgSadCg\nAfPnz1d4EBGJU8oEh4KCAoqKijTXg8St5NrmgoICBQcRkTilTHAoobkeREREgpNywUFERCTd7NwJ\nmzeXLj/8UPr9558n9lgKDiIiItXIOdi6FTZt8m/wJV/Le9OP/r6ibVur8faPCg4iIiJx2LWr9I0+\netmfdTt3Vnyshg2hUSO/HHRQ6fdNmkDr1uVvi/4++vGSJdCnT+J+DwoOIiKS8pyDoiIoLISNG8su\nsevKe7xpk39+RRo1gsaN/Rt248aly2GH7bkuul3J15I3/IYNoVYCZ1latSpx+wIFBxERSRI7d8KG\nDbB+fdkl3hCwt0/5der4T/IZGf5rydKqVen62Df92IDQqBHUrl29v4+wKDiIiEi12rat/ADw/fd7\nroteNm0qf3/16pV9w8/IgKZNoX37suti25R836ABmJW/b9mTgoOIiOy3rVuhoAC++650iX5cXhjY\nsqX8fR10EBxySNmlffs910UvBx8MBx5YvT9zulNwkKRVVFREgwYNwi5DJGU45z/VlxcA9hYONm/e\ncz8HHgiHHuqXpk2hRQs45ph9B4ADDqj+n1kqTze5SiKrV6/myiuvpFWrVtSvX5+jjjqKa665hp07\ndzJy5EhqlTOa5tlnn6VWrVqsXLly97q2bdtyzjnn8O6779KjRw8aNGhA165dee+99wB45ZVXOO64\n4zjwwAPp0aMHn376aaXqfP/996lVqxbjx4/npptuomXLljRq1Ihzzz2Xb775Zo/2M2fO5Gc/+xlN\nmjShYcOG9O3blw8//LBMm5Kfb/78+Vx66aUccsghnHLKKbu3L1y4kIsuuojDDjuMBg0a0LlzZ0aM\nGFGpukVSUXExfPstzJ0L//0vjB0L990Hf/oTXHop9O8P3brB4YeXdvl37AgnnQTnnANXXgl/+xtM\nnAiLF/s3927d4OKL4fbb/f4mTYJPPoEVK3xvQlERfPUVzJoFkyfDiy/Co4/CHXfAddfBwIHwy1/6\nY3Tu7AcPKjQkD/U4JIk1a9ZwwgknsGnTJoYNG0anTp1YtWoVEyZMoKioCDPDyjlJV956M2PRokVc\nfvnlXH311QwcOJD777+fc889l4ceeoiRI0dy7bXX4pzj7rvv5uKLL2bhwoWVrvmvf/0rtWrV4oYb\nbuDbb79l1KhRnHHGGXz66afUq1cPgHfffZezzz6bHj167A4HY8aM4fTTT+eDDz6gR48eu2sGuPDC\nCzn66KO55557cM4BMHfuXE455RTq1avHsGHDOPLII1m6dClvvPEGd911V6XrFkkGmzfD2rV+WbOm\n9PvYZd06fxlhtCZNfC9Aixb+TbtzZ9870KxZaU9ByeOmTfWmLmUpOCSJkjffjz/+mO7du+9eP3Lk\nyP3a3+LFi5kxYwYnnngi4KfqPuuss/jd737HokWLaNWqFQBNmjTh6quvJjc3l1NPPbVSx9iwYQML\nFizYfTqhe/fuXHTRRTz55JNce+21APz2t7+lX79+vPnmm7ufN2zYMI455hhGjBjB22+/XWaf3bt3\nZ9y4cWXW/f73v8fMmDNnzu66Ae65555K1SsSNuf8VQDffOOXVav2Hg5ixwnUrQstW5YGghNOKP0+\nen3z5lC/fjg/n6SGtAwORUWwYEHwx+nc2Y/WrSrnHBMnTuScc84pExqq4phjjtkdGgB69uwJQP/+\n/cu8+fbs2RPnHMuWLat0cBg0aFCZMQi//vWvadmyJZMmTeLaa69lzpw5LF68mFtuuYXvv/9+dzvn\nHP369eP5558vsz8zY9iwYWXWFRQUkJeXx/Dhw8vULVLTOOfHBZSEgvKWVav2DASHHlr6pn/UUb57\nPzoIlCxNmujKAKkeaRkcFiyArKzgjzN7NiTiflvfffcdmzZt4thjj636ziJi7wbZuHFjAFq3bl1m\nfUZGBuB7DyqrQ4cO5a5bsWIFAEuWLAFg4MCB5T6/Vq1aFBYW7q4BoF27dmXaLFu2DCChvxuRytq1\ny58S2Fco2L699Dl16vhxBa1b++X44/28ASWPW7f2AUGnCaSmScvg0Lmzf1OvjuNUl/LGNwDsij25\nGVF7LzOV7G19yXiCRCouLgbggQceoFu3buW2adSoUZnHB+q6KwnB1q1+sN+KFbB8uV9WrICVK30o\nWLOm7DiCevVK3/zbtPG9BCWPS8LBYYelz4RBklrSMjg0aJCYnoDqcuihh9K4cWO++OKLvbY5+OCD\nAdi0adPu3gNg96f7MCxevHiPdUuWLNkdEtq3bw/AQQcdxOmnn75fxzjqqKMAKvzdiOzLzp3w9ddl\ng0FJOFi+HFavLm1bu7YPA23b+g8H/fvvGQqaNtVpA0ldaRkcko2Zcd555/HCCy+Qn59PZjmpp337\n9jjnyM3N5Ze//CUAW7ZsYezYsdVd7m5jx47lhhtu2N1r8PLLL7NmzRpuvPFGALKysmjfvj1///vf\nyc7OpmHDhmWeX1BQQLNmzSo8RrNmzTj11FN55plnGD58OEcccUQwP4wkteJi3yuwt2Dw9delPQZm\n/hRCu3Z+Of300u/btvXBoI7+55Q0pn/+SeLuu+9mypQpnHrqqQwdOpQuXbqwevVqJkyYwPTp0znz\nzDNp06YNQ4YM4c9//vPuyxoPO+wwvv7661BqPuSQQ+jTpw9XXHEFa9eu5aGHHuLoo4/mqquuAnwg\neuqppzj77LM59thjueKKK2jVqhWrVq1i2rRpZGRkMHHixH0e5+GHH+aUU04hMzOToUOH0q5dO5Yv\nX86kSZOYM2dO0D+m1BA7d/ogsHChXxYvLg0IX33lpzkuceihpUHgxBPLBoMjj/SnGkSkfAoOSeLw\nww9n5syZ3HLLLbz44ots2rSJVq1acfbZZ9OgQQPq1KnDa6+9xjXXXMOtt95KixYtGD58OBkZGQwZ\nMqTMvioz50NF6ytiZtx0003MnTuXe++9lx9++IEzzjiD0aNHUz/qWrCf/vSnzJgxgzvvvJPRo0ez\nefNmWrRoQc+ePfe4gmJvunbtykcffcQtt9zCP//5T7Zu3cqRRx7JxRdfXKmapeYruTKhJBwsWlT6\n/dKlsGOHb9egAXTo4K9C+MUvygaDtm39DYlEZP9YEIPeEs3MMoHZs2fPLrebHiA/P5+srCwqaiPV\n4/333+e0005jwoQJXHDBBWGXs1f6N1Nzbd0KS5aUhoLokFBygY+Z7x3o1MkvRx9d+n2rVom9LbFI\nMiv5vw7Ics7lV3V/6nEQkVA4569IiO05WLjQn1oo+UzTpElpIBgwoDQkdOigmxuJhEHBQeK2Y8cO\n1q9fX2Gb6DkXREp8/72/V8Jnn/mvc+fC/Pl+Mjbwgw07dPCB4MILS4NCp05+2mNdoSBScyg4SNw+\n/PBDTjvttL1uNzPGjBnDkUceWekxEZIadu70vQfRIeGzz/zkR+AHHf7kJ/4mSdnZpeGgXTtdqSCS\nLAL5UzWzRsBdwHnAYUA+cJ1zblZk+xhgUMzT3nbOnR1EPZIYxx9/PO+8806FbY499liaN2++14mn\nJHXE9iJ89hnMm1d69ULr1j4gDBzov3bt6u+6qIAgktyC+hN+GjgG+A2wBrgceMfMujjn1kTavAUM\nBko+mm6L3YnULBkZGfs9UZMkr331ItSv73sRjj8eBg3yAaFrVzjkkHDrFpFgJDw4mFl94AJggHNu\nemT17WY2APgtcGtk3Tbn3HeJPr6I7L+tW2HOHPj4Yx8O9taLUBIQunXzYxPUiyCSPoL4c68D1GbP\nHoQfgT5Rj/ua2TpgA/AuMMI5V/HIOxFJGOf8vRY++ghmzPBf58zxN2JSL4KI7E3Cg4NzbrOZzQBu\nMbMFwDrgUqA3UHLzgreAV4DlQHvgHmCSmfV2yTCxhEgS+vFHf3O3GTNKg8KayInD9u2hVy+4/HL/\ntWtX3ZVRRMoXVAfjZcAzwCpgJ35w5ItAFoBzbnxU23lm9jmwFOgLTNvbTktmQoyWnZ1Ndnb27sfz\n589PyA8gqS+V/60456daju5N+PRTP16hYUM/zfLgwdC7N/Ts6e/UKCLJLycnh5ycnDLrCgsLE3qM\nQGeONLMDgcbOuXVm9hLQ0Dk3YC9tvwVuds49Wc62fc4cuXLlSrp06UJRyYXhInFo0KAB8+fPp02b\nNmGXUiVbtsAnn5QNCt9+67cdfbTvRejd2y/HHqsxCSLpJKlmjnTO/Qj8aGYHA2cB/1teOzNrDTTF\nX4GxX9q0acP8+fMpKCjY311IGmrWrFnShQbn/HTMJQFhxgz4/HN/d8eDDvI9CEOHlvYmNG0adsUi\nkkqCmsfhTPxllguBjsB9wJfAs2bWELgNP8ZhLdAB+BuwCJhcleO2adMm6d4EROLx9dcwdSpMmeK/\nrlvn13fp4gPCNdf4r126QO3a4dYqIqktqB6HDPyAx1bAemAC/qqJXWa2C+gKDASaAKvxgeFW59yO\ngOoRSSobN8J778E77/hl4UI/7XJWFlxxBfTt63sTmjQJu1IRSTeBBAfn3MvAy3vZthX4WRDHFUlW\n27b50w7vvON7FT75BIqL/dUOZ5wBf/0rnHaaLocUkfBpiJRICIqL4YsvfEh45x3IzfU3fGraFPr3\nh6uugn79/D0cRERqEgUHkWqycmXpqYepU/1VD/Xrw6mnwsiRPjB06wa1aoVdqYjI3ik4iARk40aY\nNq00LCxa5Mcp9OgBV17pg8JJJ/nwICKSLBQcRBLEOcjPh1df9acgZs3ypyQ6dPAh4Z57/KBGjVMQ\nkWSm4CBSBc75+zu8/DKMHw/LlvlgcMYZfi6Ffv2gbduwqxQRSRwFB5FKcs5P3zx+vA8MS5f6QY0X\nXAD//KfvVdB9HkQkVSk4iMTBOX+L6ZKwsGSJ71m44AJ47DF/qaTCgoikAwUHkb0oCQslpyFKwsL5\n58Po0QoLIpKeFBxEojgHc+eWhoXFi+Hgg33PwqOPwumnKyyISHpTcJC055y/SVRJWFi0yIeF88+H\nhx/2AxwVFkREPAUHSUvO+ZkbS8YsLFzow8J558FDD/mehbp1w65SRKTmUXCQtLJoETz/vA8LCxb4\nm0Sdfz6MGuV7FhQWREQqpuAgKa+4GN56Cx55BCZPhowMHxYeeMBPzKSwICISPwUHSVmFhTBmjL8C\nYskSP9Xz2LFw0UVQr17Y1YmIJCcFB0k58+f7KyCee87frvqii2DcOOjZ098rQkRE9p+Cg6SEXbv8\n6YiHH/b3iWjeHP73f2HYMGjZMuzqRERSh4KDJLWNG+GZZ/zpiGXL4MQT/eDHX/9apyNERIKg4CBJ\n6csv/WDHsWNhxw5/OuLFF/3pCBERCY6CgySNXbvgzTf96YipU6FFC/jLX/zpiBYtwq5ORCQ9KDhI\njbdhQ+npiOXLoVcveOEFfzpCl1KKiFQvBQepsebN86cjxo3zpyMuuQT+9S844YSwKxMRSV8KDlKj\n7NoF//mPDwzvvuuviLjhBhg61F8pISIi4VJwkBrBOcjJgREj/OmI3r394wsu0OkIEZGaRMFBQvfR\nR3DddTBzpr/J1PjxfpZHERGpeWqFXYCkr5Ur4dJLfe/Ctm0wbRq8+qpCg4hITabgINVu82a45Rbo\n1MmPY3j6aZg1C/r2DbsyERHZF52qkGpTXOwnbLrpJli/Hv7nf/zAx4MOCrsyERGJVyA9DmbWyMz+\nYWYrzKzIzD4wsx4xbe4ws9WR7VPMrEMQtUjNkJfnp4O+4go49VRYsAD++leFBhGRZBPUqYqngX7A\nb4CfAFOAd8ysJYCZXQ9cCwwFTgS2AJPNTOPnU8zy5XDhhT4s1KoFH3wAL70EbduGXZmIiOyPhAcH\nM6sPXAD82Tk33Tm3zDl3O7AE+G2k2R+BO51zbzjnvgAGAocD5yW6HgnHpk1w/fXQuTN8+KE/RfHR\nR3DyyWFXJiIiVRFEj0MdoDawLWb9j0AfM2sHtACmlmxwzm0CZgK9A6hHqtGuXfDkk9Cxo5/E6cYb\nYdEiuPxy3+MgIiLJLeH/lTvnNgMzgFvMrKWZ1TKzy/ChoCU+NDhgXcxT10W2SZKaOhUyM/0sj2ee\n6QPDyJHQsGHYlYmISKIE9RnwMsCAVcBW/HiGF4HigI4nIVq8GM49F/r39yFh5kx/f4nWrcOuTERE\nEi2QyzGdc8uB08zsQKCxc26dmb0ELAPW4kNFc8r2OjQH5lS03+HDh5ORkVFmXXZ2NtnZ2YksX+K0\nYQPceSc8+qi/p0RODlx8MZiFXZmISHrKyckhJyenzLrCwsKEHsOccwndYbkHMTsYHxr+1zn3tJmt\nBu53zo2KbG+MDxEDnXMvl/P8TGD27NmzyczMDLxeqdjOnfDEE3DbbbB1qx/H8Kc/wYEHhl2ZiIjE\nys/PJysrCyDLOZdf1f0F0uNgZmfiexUWAh2B+4AvgWcjTf4BjDCzJcAK4E7gG2BiEPVI4rz9tp+4\naf58GDzYz8XQsmXYVYmISHUJaubIDOAeoBWwHpgAjHDO7QJwzt1nZg2AJ4AmQB7wc+fc9oDqkSra\nvBmuugr+9S8/J8OsWX4gpIiIpJegxji8DOxxyiGmzUhgZBDHl8RavNjftXLlSnjxRbjkEo1jEBFJ\nV7qyXir0xhv+bpW7dsHHH0N2tkKDiEg6U3CQchUX+zkYBgyA007zoaFLl7CrEhGRsOnumLKHjRvh\nsstg0iS46y5/1YRmfRQREVBwkBhffAHnnw8FBfDmm/Dzn4ddkYiI1CT6HCm7jR8PvXr5+RhmzVJo\nEBGRPSk4CDt3wp//7Gd9POccmDED2rcPuyoREamJdKoizRUU+Msr33sPHnwQrrtOV02IiMjeKTik\nsdmz4YIL4Mcf4Z13oG/fsCsSEZGaTqcq0tRzz8HJJ0Pz5j5AKDSIiEg8FBzSzPbt8Lvf+ftMXHYZ\n5ObCEUeEXZWIiCQLnapII2vWwIUX+smcnngChg4NuyIREUk2Cg5pYvp0+PWv/UROubn+sksREZHK\n0qmKFOccPPaYH8PQsaMfz6DQICIi+0vBIYX9+CMMGeLHNFxzDUydCi1ahF2ViIgkM52qSFFffQW/\n+hXMmwfjxvmBkCIiIlWl4JCCpk71kzo1agQffgjdu4ddkYiIpAqdqkghzsH998OZZ0Jmpr/fhEKD\niIgkkoJDiti+HbKz4S9/8cukSdC0adhViYhIqtGpihTgnB/8+O9/w8sv+8suRUREgqDgkAIefBCe\nfhqefVahQUREgqVTFUnu9df9LbFvvBEGDQq7GhERSXUKDkns00/h0kvh/PPhrrvCrkZERNKBgkOS\nWrMGBgyAzp1h7Fg/lbSIiEjQ9HaThIqK4Nxz/aDI11+Hhg3DrkhERNKFBkcmmeJiP5Zh3jzIy4PD\nDw+7IhERSScKDknmtttgwgR/6WVmZtjViIhIutGpiiTy/PN+EOS99/oBkSIiItVNwSFJTJ8OV14J\ngwf7mSFFRETCkPDgYGa1zeweM1tuZkVmtsTMRsS0GWNmxTHLpETXkiqWL/c9DL16wRNPgFnYFYmI\nSLoKYozDzcCVwEDgS6AH8KyZbXTOPRrV7i1gMFDyNrgtgFqSXmEh/PKX0LgxvPIK1K0bdkUiIpLO\ngggOJwATnXNvRx6vNLNLgRNj2m1zzn0XwPFTxs6d/vbYq1bBRx9Bs2ZhVyQiIukuiDEObwH9zKwj\ngJl1A04GYk9F9DWzdWa2wMweM7NDAqglqf3pTzBlir+KonPnsKsREREJoMfBOfeYmR0BLDSznfhw\ncrNz7qWoZm8BrwDLgfbAPcAkM+vtnHOJrikZjR4NjzwCjz8O/fuHXY2IiIiX8OBgZn8ABgEX48c4\nHA88ZGarnXPjAJxz46OeMs/MPgeWAn2BaYmuKdlMngx//KNfrr467GpERERKBTHG4Sbgdufcy5HH\n88ysLXAjMK68JzjnlptZAdCBCoLD8OHDycjIKLMuOzub7OzsBJRdM3z5JVx0EZx1FjzwQNjViIhI\nMsnJySEnJ6fMusLCwoQeI4jgUAvYFbOumArGU5hZa6ApsKaiHY8aNYrMFJ4u8bvv/BUUbdpATg7U\nrh12RSIikkzK+zCdn59PVlZWwo4RRHB4DRhhZt8A84BMYDjwFICZNQRuw49xWIvvZfgbsAiYHEA9\nSWHbNrjgAtiyBd59119+KSIiUtMEERyuA24HHgWaA6uBx4E7I9t3AV3x8zw0iWyfDNzqnNsRQD01\nnnMwdCh88glMmwZt24ZdkYiISPmCuKqiCPhzZClv+1bgZ4k+bjK7914YOxZefBF69w67GhERkb3T\nvSpC9sorcNNN/q6XKTTGU0REUpSCQ4hmzYLLL/ezQ952W9jViIiI7JuCQ0i++QbOOQe6doVnntGN\nq0REJDkoOIRg82YYMADq1IHXXoMDDwy7IhERkfgEcVWFVKC4GC67DJYsgenToUWLsCsSERGJn4JD\nNbvxRnj9db907Rp2NSIiIpWj4FCNxoyB++6DBx/0M0SKiIgkG41xqCYffADDhvmJnq67LuxqRERE\n9o+CQzUoLobf/Q569IBHH9UVFCIikrx0qqIavPoqzJ0LeXlwwAFhVyMiIrL/1OMQsOJiGDkS+veH\nPn3CrkZERKRq1OMQsFdegS++gCeeCLsSERGRqlOPQ4CKi+H22+HMM+Gkk8KuRkREpOrU4xCgl1+G\nefPgqafCrkRERCQx1OMQkF27fG/Dz34GvXqFXY2IiEhiqMchIOPHw/z58OyzYVciIiKSOOpxCMCu\nXXDHHXD22XDiiWFXIyIikjjqcQjASy/BggUwdmzYlYiIiCSWehwSbOdO39vwi1/ACSeEXY2IiEhi\nqcchwXJyYNEieOGFsCsRERFJPPU4JFBJb8OAAf6+FCIiIqlGPQ4J9MILsGQJ/OtfYVciIiISDPU4\nJMjOnXDnnXDuuZCZGXY1IiIiwVCPQ4I8/zwsXQoTJoRdiYiISHDU45AAO3b43obzz4fjjw+7GhER\nkeCoxyEBxo2DZcvg1VfDrkRERCRY6nGooh074K674Fe/gq5dw65GREQkWOpxqKLnnoPly2HixLAr\nERERCV7CexzMrLaZ3WNmy82syMyWmNmIctrdYWarI22mmFmHRNcStO3bfW/DhRfCcceFXY2IiEjw\nguhxuBm4EhgIfAn0AJ41s43OuUcBzOx64NpImxXAXcBkM+vinNseQE2BePZZWLkS3nwz7EpERESq\nRxDB4QRgonPu7cjjlWZ2KRB9n8g/Anc6594AMLOBwDrgPGB8ADUl3Pbt8Ne/wkUXwbHHhl2NiIhI\n9QhicORbQD8z6whgZt2Ak4FJkcftgBbA1JInOOc2ATOB3gHUE4hnnoGvv4Zbbw27EhERkeqT8B4H\n59xjZnYEsNDMduLDyc3OuZciTVoADt/DEG1dZFuNt22b72245BI45piwqxEREak+CQ8OZvYHYBBw\nMX6Mw/HAQ2a22jk3LtHHC8PTT8Pq1eptEBGR9BPEGIebgNudcy9HHs8zs7bAjcA4YC1gQHPK9jo0\nB+ZUtOPhw4eTkZFRZl12djbZ2dkJKTweW7fC3XdDdjZ07lxthxUREdmnnJwccnJyyqwrLCxM6DGC\nCA61gF0x64oj63HOLTeztUA/YC6AmTUGegKjK9rxqFGjyAz5DlJPPQVr1qi3QUREap7yPkzn5+eT\nlZWVsGMEERxeA0aY2TfAPCATGA48FdXmH5E2S/CXY94JfAPU6GmUtm6Fe+6B3/wGjj467GpERESq\nXxDB4TrgduBR/OmH1cDj+HAAgHPuPjNrADwBNAHygJ/X9Dkc/u//YO1auOWWsCsREREJRxBXVRQB\nf44sFbUbCYxM9PGD8uOPvrfhssugY8ewqxEREQmHbnIVpyeegO++U2+DiIikNwWHOBQVwb33wuWX\nQ4eku6NehTUvAAAXZklEQVSGiIhI4ig4xOGf/4SCAhixx626RERE0ouCwz5s2QJ/+xsMGgTt24dd\njYiISLgUHPbh8cdh/Xr1NoiIiICCQ4W2bIH77oPBg6Fdu7CrERERCZ+CQwVGj4YNG+Dmm8OuRERE\npGZQcNiLzZvh/vthyBBo2zbsakRERGoGBYe9GD0aCgvV2yAiIhJNwaEcP/zgexuuvBLatAm7GhER\nkZpDwaEcjz7qw8NNN4VdiYiISM2i4BBj0yb4+9/hqqvgiCPCrkZERKRmUXCI8cgjfmDkjTeGXYmI\niEjNo+AQpbAQHngAhg6F1q3DrkZERKTmUXCI8vDD/oZW6m0QEREpn4JDxMaN8OCDMGwYHH542NWI\niIjUTAoOEQ89BFu3wg03hF2JiIhIzaXggA8Mo0bB1VdDy5ZhVyMiIlJzKTgAH3/sB0YOGhR2JSIi\nIjWbggOQmwuNG8Nxx4VdiYiISM2m4ADk5UGfPlC7dtiViIiI1GxpHxx27oQPP4RTTgm7EhERkZov\n7YPDp5/6mSJPPTXsSkRERGq+tA8OublQvz706BF2JSIiIjVf2geHvDzo1Qvq1g27EhERkZovrYND\ncbEPDhrfICIiEp+0Dg4LFsD332t8g4iISLzSOjjk5vpLMHv1CrsSERGR5JDw4GBmy82suJzlkcj2\nZ8vZNinRdcQjLw+ysqBRozCOLiIiknzqBLDPHkD0VErHAf8FxkceO+AtYDBgkXXbAqijQs7B++9D\ndnZ1H1lERCR5JTw4OOe+j35sZgOApc65vKjV25xz3yX62JWxYgWsWqWBkSIiIpUR6BgHMzsA+A3w\ndMymvma2zswWmNljZnZIkHWUJy8SY/r0qe4ji4iIJK8gTlVEOx/IAJ6LWvcW8AqwHGgP3ANMMrPe\nzjkXcD275ebCT34Ch1R7ZBEREUleQQeHIcBbzrm1JSucc+Ojts8zs8+BpUBfYFpFOxs+fDgZGRll\n1mVnZ5O9HwMV8vKgf/9KP01ERKTGysnJIScnp8y6wsLChB7DgvqQb2ZtgGXAec65N/bR9lvgZufc\nk3vZngnMnj17NpmZmVWube1aaNkScnLgkkuqvDsREZEaKz8/n6ysLIAs51x+VfcX5BiHIcA6oMJL\nLc2sNdAUWBNgLWV88IH/qoGRIiIilRNIcDAzw19u+axzrjhqfUMzu8/MeprZkWbWD3gNWARMDqKW\n8uTmwlFHQatW1XVEERGR1BDUGIf+wBHAmJj1u4CuwECgCbAaHxhudc7tCKiWPeTlaZppERGR/RFI\ncHDOTaHsJFAl67cCPwvimPHauBE++wx+//swqxAREUlOaXeviunT/ayR6nEQERGpvLQLDnl5/oqK\n9u3DrkRERCT5pF1wyM31V1OY7butiIiIlJVWwaGoCGbN0mkKERGR/ZVWwWHmTNixQ/M3iIiI7K+0\nCg55edCkib9HhYiIiFReWgWH3Fx/N8xaafVTi4iIJE7avIXu2AEzZmh8g4iISFWkTXDIz/eDIzW+\nQUREZP+lTXDIy4MGDSABN9cUERFJW2kTHHJzoVcvqFs37EpERESSV1oEh+JifyttjW8QERGpmrQI\nDvPmwYYNCg4iIiJVlRbBITcXDjgAevYMuxIREZHklhbBIS8PevTwgyNFRERk/6V8cHCu9MZWIiIi\nUjUpHxyWLYM1azS+QUREJBFSPjjk5vpbaJ98ctiViIiIJL+UDw55edC1q7+5lYiIiFRNygcHjW8Q\nERFJnJQODqtXw9KlGt8gIiKSKCkdHPLy/Ff1OIiIiCRGSgeH3Fzo2BFatAi7EhERkdSQ0sEhL0+n\nKURERBIpZYPD+vXw+ec6TSEiIpJIKRscpk/3X9XjICIikjgpGxxyc6FVK2jbNuxKREREUkfKBoeS\n8Q1mYVciIiKSOhIeHMxsuZkVl7M8EtXmDjNbbWZFZjbFzDoksoYtW2D2bI1vEBERSbQgehx6AC2i\nljMAB4wHMLPrgWuBocCJwBZgspnVTVQBH30EO3dqfIOIiEii1Un0Dp1z30c/NrMBwFLnXGQ6Jv4I\n3OmceyOyfSCwDjiPSLioqtxcOOQQ6NIlEXsTERGREoGOcTCzA4DfAE9HHrfD90JMLWnjnNsEzAR6\nJ+q4JfenqJWyIzhERETCEfRb6/lABvBc5HEL/GmLdTHt1kW2Vdn27f5UhcY3iIiIJF7CT1XEGAK8\n5Zxbm4idDR8+nIyMjDLrsrOzyc7O3v141izYulXjG0REJP3k5OSQk5NTZl1hYWFCjxFYcDCzNkB/\n/NiFEmsBA5pTttehOTBnX/scNWoUmZmZFbbJy4OGDaF790qXLCIiktRiP0wD5Ofnk5WVlbBjBHmq\nYgg+HEwqWeGcW44PD/1K1plZY6An8GEiDpqbCyedBHWC7ksRERFJQ4EEBzMzYDDwrHOuOGbzP4AR\nZjbAzI4DxgLfABOretxdu/xU0zpNISIiEoygPpf3B44AxsRucM7dZ2YNgCeAJkAe8HPn3PaqHvTz\nz6GwUAMjRUREghJIcHDOTQFqV7B9JDAy0cfNy4O6deHEExO9ZxEREYEUu1dFbi6ccAIceGDYlYiI\niKSmlAkOzpXe2EpERESCkTLBYfFiWLdO4xtERESClDLBITfXTzF90klhVyIiIpK6UiY45OVBt24Q\nM7GkiIiIJFDKBIfcXI1vEBERCVpKBIevv4YVKxQcREREgpYSwSEvz3/t0yfcOkRERFJdygSHzp3h\nsMPCrkRERCS1pURwyM3VZZgiIiLVIemDQ0EBfPmlxjeIiIhUh6QPDh984L+qx0FERCR4SR8c8vKg\nTRs48siwKxEREUl9SR8cNL5BRESk+iR1cPjhB8jP1/gGERGR6pLUwWHGDCguVo+DiIhIdUnq4JCb\nC82a+TkcREREJHhJHRzy8vxpCrOwKxEREUkPSRsctm2DmTN1mkJERKQ6JW1w+OQTHx40MFJERKT6\nJG1wyM2Fgw6Cbt3CrkRERCR9JG1wyMuDk0+G2rXDrkRERCR9JGVw2LkTpk/X+AYREZHqlpTB4bPP\n/ORPGt8gIiJSvZIyOOTlQb16cMIJYVciIiKSXpIyOOTmQs+ePjyIiIhI9Um64OBc6cRPIiIiUr0C\nCQ5mdriZjTOzAjMrMrPPzCwzavsYMyuOWSbFs+8FC6CgQAMjRUREwlAn0Ts0sybAdGAqcBZQAHQE\nNsQ0fQsYDJRMGL0tnv3n5flLMHv3Tki5IiIiUgkJDw7ADcBK59xVUeu+KqfdNufcd5XdeW4udO/u\nJ38SERGR6hXEqYoBwCwzG29m68ws38yuKqdd38j2BWb2mJkdEs/ONb5BREQkPEEEh6OA3wILgTOB\nx4GHzezyqDZvAQOB04G/AD8FJplVfJ/LNWtg5UqNbxAREQlLEKcqagEfO+duiTz+zMx+AlwNjANw\nzo2Paj/PzD4HlgJ9gWl72/GcOf5rnz4Jr1lERETiEERwWAPMj1k3H7hgb09wzi03swKgAxUEh9Gj\nh9OoUQZDhpSuy87OJjs7u0oFi4iIpIKcnBxycnLKrCssLEzoMYIIDtOBTjHrOlH+AEkAzKw10BQf\nOvaqXr1RnHdeJo8/XuUaRUREUk55H6bz8/PJyspK2DGCGOMwCuhlZjeaWXszuxS4CngUwMwamtl9\nZtbTzI40s37Aa8AiYHJFO/7qK41vEBERCVPCg4NzbhZwPpANfA7cDPzROfdSpMkuoCswET+A8kng\nE+BU59yOfe1fwUFERCQ8QZyqwDk3CSh3Jkjn3FbgZ/uz38MPhyOOqEplIiIiUhVJda+K7t3DrkBE\nRCS9JVVwyMzcdxsREREJTlIFB/U4iIiIhCupgkObNmFXICIikt6SKjhUPCG1iIiIBC2pgoOIiIiE\nS8FBRERE4qbgICIiInFTcBAREZG4KTiIiIhI3BQcREREJG4KDiIiIhI3BQcRERGJm4KDiIiIxE3B\nQUREROKm4CAiIiJxU3AQERGRuCk4iIiISNwUHERERCRuCg4iIiISNwUHERERiZuCg4iIiMRNwUFE\nRETipuAgIiIicVNwEBERkbgpOIiIiEjcFBxEREQkbgoOIiIiErdAgoOZHW5m48yswMyKzOwzM8uM\naXOHma2ObJ9iZh2CqEUSKycnJ+wSpBrodU4feq2lshIeHMysCTAd2AacBXQB/gfYENXmeuBaYChw\nIrAFmGxmdRNdjySW/pNJD3qd04dea6msOgHs8wZgpXPuqqh1X8W0+SNwp3PuDQAzGwisA84DxgdQ\nk4iIiCRAEKcqBgCzzGy8ma0zs3wz2x0izKwd0AKYWrLOObcJmAn0DqAeERERSZAggsNRwG+BhcCZ\nwOPAw2Z2eWR7C8DhexiirYtsExERkRoqiFMVtYCPnXO3RB5/ZmY/Aa4Gxu3nPusDzJ8/PwHlSVUU\nFhaSn58fdhkSML3O6UOvdeqLeu+sn4j9BREc1gCx7/DzgQsi368FDGhO2V6H5sCcveyzLcBll12W\nsCJl/2VlZYVdglQDvc7pQ6912mgLfFjVnQQRHKYDnWLWdSIyQNI5t9zM1gL9gLkAZtYY6AmM3ss+\nJwO/AVYAWxNfsoiISMqqjw8NkxOxM3POJWI/pTs064EPDyPxV0j0BJ4A/p9z7qVIm78A1wOD8WHg\nTuBY4Fjn3PaEFiQiIiIJk/DgAGBmZwP3Ah2A5cADzrlnYtqMxM/j0ATIA37nnFuS8GJEREQkYQIJ\nDiIiIpKadK8KERERiZuCg4iIiMRNwUH2ycxuM7PimOXLsOuSqjGzU8zsdTNbFXlNzymnjW5Gl+T2\n9Tqb2Zhy/r4nhVWv7B8zu9HMPjazTZFZm181s6PLaVflv2kFB4nXF/i5NlpElj7hliMJ0BD4FLgG\nP5trGboZXcqo8HWOeIuyf9/Z1VOaJNApwCP4Kxn7AwcA/zWzA0saJOpvOoh5HCQ17XTOfRd2EZI4\nzrm3gbcBzMzKaaKb0aWAOF5ngG36+05uzrmzox+b2WDgWyAL+CCyOiF/0+pxkHh1jHR1LjWz583s\niLALkuDoZnRpp2+ke3uBmT1mZoeEXZBUWRN8D9N6SOzftIKDxOMj/GRdZ+HvOdIOyDWzhmEWJYHS\nzejSx1vAQOB04C/AT4FJFfROSA0Xee3+AXzgnCsZj5awv2mdqpB9cs5FT1P6hZl9jJ9C/CJgTDhV\niUgiOOeiu6jnmdnnwFKgLzAtlKKkqh4DjgFODmLn6nGQSnPOFQKL8DODSmqKvhldtOaRbZKinHPL\ngQL0952UzOxR4Gygr3NuTdSmhP1NKzhIpZlZI/x/Kmv21VaSU+TNo+RmdECZm9FV+e56UnOZWWug\nKfr7TjqR0HAucJpzbmX0tkT+TetUheyTmd0P/Ad/eqIVcDuwA8gJsy6pmsgYlQ74TyEAR5lZN2C9\nc+5r/DnSEWa2hNKb0X0DTAyhXNlPFb3OkeU24BX8m0oH4G/4HsWE3ElRqoeZPYa/jPYcYIuZlfQs\nFDrnSu4qnZC/ad2rQvbJzHLw1wg3Bb7DX9pzcyTBSpIys5/iz2HH/ifwnHNuSKTNSHQzuqRW0euM\nn9vhNeB4/Gu8Gh8YbtXlmcnFzIopf56OK5xzY6PajaSKf9MKDiIiIhI3jXEQERGRuCk4iIiISNwU\nHERERCRuCg4iIiISNwUHERERiZuCg4iIiMRNwUFERETipuAgIiIicVNwEBERkbgpOIikOTMbZGbr\nw65DRJKDgoNIijKzuZEb35S37XIz22pmhwAvAUdXb3W76xhkZsVmtivy9Qczm2Vm51dyP2PM7N/l\nrP9pZL+NE1e1SHpTcBBJXU8Dl5hZvXK2DQZec86td85tc84VVG9pZRQCLSLL8fibLI03s44J2n9g\nN+QxswOC2rdITaXgIBIAM5tmZo9Elo1m9p2Z3RG1/TIz+8TMNpnZGjN7wcwOjdnHOWa2yMyKzOy/\nkV6CMp+ezayPmeVG2nxlZg+ZWYPI5ueBBsCvYvbbDuiLDxaY2WAz2xDT5lwzm21mP5rZEjO71cxq\nRbbdb2b/iWp7XaSuM6PWLTazkjts9jWzmWa22cw2mFmemR0RdTjnnPvOOfetc24pMAIoBrpGnn+L\nmX1ezu/4UzO7fV+vRcxzuprZu5Hfe2HkNciM8/eJmS03sxFm9pyZFQJPVOb4IqlAwUEkOAOBHcAJ\nwB+AP5nZlZFtdfBvkF2Bc4EjgTElTzSztsDLwL+BbsBTwN1EfXo2s/bAW5F2PwEuBk4GHgFwzn0P\nTASGxNQ1GPjaOTcl8tjF7PcU/C2XRwGdgWHAIODmSJP3gZPNzCKPT8Xfbr1v5PmtgKOAaWZWG3gV\nf1vnnwC9gP9jL70AkXAyOLI9P7L6GaCzmWVFtese2d8z5e2nAi8AXwNZQCZwL/412ufvM8r/AJ/i\ne0furOTxRZKfc06LFi0JXvBvlF/ErLsndl3Uth7ALqBBVNvPYtrcGWnTOPL4SeDxmDZ9gJ1A3cjj\nMyOPj4xqsxwYGfV4ELA+6vEU4PqY/f4GWBX5PiOyz8zI4wLgL8CHUW1XRr4/OFLzKXv5uQfhexc2\nAT9E9lsEDIxp9ybwaNTjh4GpUY/HAP8uZ/8/jfmdFQKX76WWeH6fy4EJYf/70qIlzEU9DiLB+Sjm\n8Qygo3lZZvZ6pDt8E/BepE2byNdOwCcxz/845nE3YHBkQOEPZvYD8HZkW7vI1ynAKuAKADPrDxwB\nPFtB3d2AW2P2+yTQ3MzqO+cKgc+AvmZ2HLAN34vQPdKtfyq+VwLn3AZ878V/Iz/vH8ysRczxNkWO\n2Q3/Kf4m4Akz+0VUmyeBbDOrGxlXkE3kVEslPQg8bWZTzOx6Mzsq5ufe1+8TYPZ+HFckZSg4iFS/\nA/FvSBuBS/G9DSVXEdStxH4a4c+xd6X0jbcr/gqJpeAHD+BDwqDIc64ApjnnVuxjv7dF7bMbvuv+\naOfc1kib94DT8J/o33fObQTmA6eUrCvZmXNuCP4UxXR89/8iMzsx6njFzrnlzrllzrkvnHP/wPfY\nXB/V5j/4gHI+MAB/queVqO2b8D0hsZrgexy2RGq5HTgGeAM4HfjSzM6N+rkr/H1GbCnvlyaSLuqE\nXYBICusZ87g3sBg/bqApcKNzbhVAzBspwELg5zHrYtvkA8c455bvo44xwM1m9ivgPPYc8xArH+jk\nnFtWQZv3I/vZQemn8vfxPQEdKe1BAcA59xm+l+JvZvYhPjDF9qBEK8YHrJLn7zKzsZFjbgdecs5t\ni2q/ELjYzA5wzu2IWp8FLHfO7Yra1xLgIeAhM3sRH6YmEv/vUyStqcdBJDhtzOzvZna0mWUD1wL/\nAFbi3/z+YGbtzOwc/EDJaE/gBwTea2YdzewiSnsNSgYW/g04KXLlRjcz6xC5GqLMYL5I78I0/OmE\nrfjBihW5AxgYuZLiGDPrbGYXm1n0QMBc4CDgl5SGhPfw4xvWRN6cMbO2Zna3mfUyszaRKy86Al9G\n7cvMrHlkaWtmQ4GzgNdi6noK30twFnsOinwh8nsZa2aZZtY+clXHH4C/Rw5SP/K7+mmklpPxA1dL\naonr9ymS9sIeZKFFSyou+DfqR4DR+FMSBcAdUdsvxnd/FwEfAL/Ad6l3jWrzS/wn6SJgKv7qhl1E\nBupF2mThP/EX4rvr5wA3lFPPJZHnPlzOtjKDIyPrzgDygM3ABvz4jCtj2swBvol6fDB+IOHzUesO\nw18Z8g3wI7AMuDXm2LuiliL8KY/rASun1veBuXv5nXcAJuCvmtiE70G4Imr7AcCLwIpILV/jg1zc\nv89I/X8I+9+XFi1hLuZcYHOjiKQtM5sGzHHO/SmB+7wZGOqcOzJR+0w2ZrYYf3XFQ2HXIpKuNMZB\npIYys9/ir6z4Hn9Z4P/iL0NMO2bWDD9+ojkVXxEiIgFTcBAJRiK68jrixz4cjB8XcT9+wqJ09C1+\nkqn/5/zloCISEp2qEBERkbjpqgoRERGJm4KDiIiIxE3BQUREROKm4CAiIiJxU3AQERGRuCk4iIiI\nSNwUHERERCRuCg4iIiISt/8P3Rn7mYm3zbMAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe6ec682f60>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "page_views_by_user_pandas.head(20).plot.line(y='cum_perc')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the long tail, we can see there are users with more than 600 page views during that period!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>usersCount</th>\n",
       "      <th>cum_perc</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pageViewsByUser</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>668</th>\n",
       "      <td>2</td>\n",
       "      <td>99.999999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>675</th>\n",
       "      <td>1</td>\n",
       "      <td>99.999999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>677</th>\n",
       "      <td>1</td>\n",
       "      <td>99.999999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>684</th>\n",
       "      <td>1</td>\n",
       "      <td>99.999999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>696</th>\n",
       "      <td>1</td>\n",
       "      <td>99.999999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>703</th>\n",
       "      <td>1</td>\n",
       "      <td>99.999999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>704</th>\n",
       "      <td>1</td>\n",
       "      <td>100.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>718</th>\n",
       "      <td>1</td>\n",
       "      <td>100.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>811</th>\n",
       "      <td>1</td>\n",
       "      <td>100.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>829</th>\n",
       "      <td>1</td>\n",
       "      <td>100.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 usersCount    cum_perc\n",
       "pageViewsByUser                        \n",
       "668                       2   99.999999\n",
       "675                       1   99.999999\n",
       "677                       1   99.999999\n",
       "684                       1   99.999999\n",
       "696                       1   99.999999\n",
       "703                       1   99.999999\n",
       "704                       1  100.000000\n",
       "718                       1  100.000000\n",
       "811                       1  100.000000\n",
       "829                       1  100.000000"
      ]
     },
     "execution_count": 168,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "page_views_by_user_pandas.tail(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Page views by Platform"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 52 ms, sys: 16 ms, total: 68 ms\n",
      "Wall time: 5min 26s\n"
     ]
    }
   ],
   "source": [
    "%time page_views_by_platform_pandas = page_views_df.groupBy('platform').count().toPandas()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>platform</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>790149483</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1053365944</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>190760021</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               count\n",
       "platform            \n",
       "1          790149483\n",
       "2         1053365944\n",
       "3          190760021"
      ]
     },
     "execution_count": 182,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "page_views_by_platform_pandas = page_views_by_platform_pandas.set_index('platform').sort_index()\n",
    "page_views_by_platform_pandas"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Distribution of page views by platform**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fe6ec22bda0>"
      ]
     },
     "execution_count": 183,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjsAAAH5CAYAAABnHv95AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAAPYQAAD2EBqD+naQAAIABJREFUeJzs3XmczfX+wPHX+5xjZszYDUWWsYSxRIMbQsaSLbRIiHCv\nImnRqkWhX8tVWW6LlEqSriVdEkKWVG5lSCXGcm0phUiWwZn5/P74nBlnZs7MnGFmvrO8n4/H9zHO\n53yX9zkz5rzn83l/Pl8xxqCUUkopVVi5nA5AKaWUUio3abKjlFJKqUJNkx2llFJKFWqa7CillFKq\nUNNkRymllFKFmiY7SimllCrUNNlRSimlVKGmyY5SSimlCjVNdpRSSilVqGmyo5SDRGSNiKx2Oo6s\niMggEUkSkRinYwEQkeq+eG7L4fO6RWSCiOwTkUQRWZCT51dKOUOTHZUv+X24Jm+nRSReRF4WkYpO\nxxeIiIzyxdo+k31u9+1zna/JAEl5E+FFKxT3lhGRliLylIiUCvD0P4AHgbnAbcCkPA1OKZUrPE4H\noFQmDDAG2AOEAa2BO4GuItLQGJPgYGyB/Bt4AegPrMpgn/7AYWCp73GnPIhLpdYKeBJ4Bzie5rlY\n4GdjzIN5HpVSKtdoz47K75YZY2YbY942xvwdmAzUAHo5HFc6xphfgdXAjSJSLO3zIlIZaAPMNcYk\n+o7xGmO8eRtpkSeZPFcROJZjF7JCc+p8SqkLo8mOKmhWYT+sagCISFkReVFEvheRv0TkTxFZIiJX\npD1QRKqJyCIROSEiv4nIRBG51jes1DbNvleJyDIROSYiJ321Na2CiG8WUBroHuC5fr7YZ/tdZ42I\npOoFEpEQERknIjtEJMFXP/JPEQnx2+dDEYlLc9zHaYbIEJG/+do6+x57fEM4231Dg4dFZJ2IdAji\ntQFEiMg033F/isi7IlLG73ozROSQiLjTHigiy0Vka2Yn970f34tIjIh8KSKnROR/IjIsq8BEpJGI\nvCMiu3yv7VcReUtEyvnt8xQwwfdwj++9SUyuAQLaAQ392tv6jgsXkZd834sEEdkmIg8EiCFJRP4l\nIv1F5EcgAejsV2N0v4iM9L2mk773pIovKRojIvt9r/kj//dVKXVxdBhLFTS1fV+P+L7WBHoC84Dd\nwCXAMGCNiNQ3xhwE+2GF7XW5BNs79Bt2SCmWNLUovpqbJcAGYCy2pmYIsEpEWhtjNmQS3wJgqu/c\n/0nzXD9grzHmK7+2tNcW4GPsUMs0YBvQCBgFXA7c6Nt1HdBTREoYY0742loBidjeo8W+tra+ti99\nj8cBo4E3gG+BUkAzIAb4LJPXBTZRewU4CjwF1AVGANWw7yPAe8BAoDP2PUx+XZf49nkqi2sYoBzw\nCbZuZjbQB5gqImeMMTMyObYTNgl+GzgINMD+LNQHWvr2+RCoA/QF7uX8z9HvwADgCSAC+x4JkJyc\nfQxcA0wHNvte3wsiUtkYkzbp6eCL+RXskOUev+cGYn/vTvG9zod9r/MLX4zPY3/G7wFeBIZm8nqV\nUsEyxuimW77bgEHYD+lYoDxwGXALcAg4AVTy7VcswLHVgNPA435t9/vOd51fWwjwk6+9rV97PPBJ\nmnOGAruww2pZxT4HOAmU8Gurg02ank6z72pgld/jAcA5oGWa/e7wxdnC97ip73ydfY8b+h7/G/jK\n77j/ABv8Hm8CFl3g9yMJ+Bpw+7U/6P++YhOEfcDsNMePArxA9Syus9p3vnv92ooBG4Ffk68NVPfF\nc5v/9yjA+W7xne9qv7YHfG3VMrj+92naevmuNTpN+1zfa6rh15bk+/7VTbNvcrwH0/xcPONr3wS4\n/Nrf9/0Mp/v51k033bK/6TCWys8E29twCNiP/Sv/OHC9sfUxGGPOpews4vINWZzCJiz+06Q7AweM\nMck9HhhjzgJvprqgSBNsD8oHIlI+eQNK+mJJNdyVgVlAcc73wgDciu21mB3wiPN6Y3sTtqe5/mrf\n+5Hcg7IJm/Qlx9MG+x7NBJqKSJivvTW2FyjZMaCBiNTmwrxhfPVGPlOxiUM3AGOMwX5Q9xSRCL/9\n+mOTsL1BXMOL7XnCd85z2F6uitgkLyBjzJnkf4tIqO99+xr7vl3MlPmuvpheTtP+ErYUoGua9jXG\nmPgMzjXXnO+JwxcfwExjTFKa9hBskq+Uukia7Kj8zGBnX3XE1lLUN8bUMsasTN7BV+swSkS2A2ew\nwwa/Y4d+Svudqzq2ZyatnWkeX+77OhObZCVvv2OHFEJEpDSZWwr8gf2AT9YX2GyMybRmxXf9Bmmu\nfQibvBnsBz6+D8b12CQH39d12OEqN9BCROpjh0r8k50ngTLYZOp7sWvKNMoipmSGNO+XMeYktscl\nyq95JhAO3AAgInWxScrMIK/zizHmdJq27dikJSr97pbY+q0pInIQ2ytyCPifL+6svmeZqe6L6WSa\n9q1+z/vbk8m59qd5/Kfv688ZtJcNJkClVOa0Zkfld98aYzZm8vzjwHhsLcUT2CQjCVsTcSHJfPIx\nD2BrMwI5kUE7YGdYichcYKiIVMB+QF+OHfIJ5vo/YId9As0a8v+w/AJ4TOxsnzbYIbI/fYWxbbAJ\nmsEv2THGrBORWtihmWux68qMEpFhxpi3g4gvS8aYrb7i6QHYXq4B2ER0Xk6cPxPzgBbYAuTN2O+T\nC/iUvP3DLm2i5i8xm+2ZzRxTSgVJkx1V0N2ErXm5w7/RN5PlkF/TXiA6wPGXp3mc3PvzlzEmo7Vy\ngvE+MBxbM1KT8/U0WdkFXGGMCWZV5XXYoY5+QGXOJzWfY4e3fgO2G2P83weMMceAd4F3fYXb67CF\n2FklO4J9v9amNNihqkrYgmJ/M4GXRORSX3yfGGP+JDiVRaR4mt6dutjEbU/AwOz3uz0wxhjzjF97\noOG67C6OuBfoICIRaXp3ov2eV0rlYzqMpQq6RNL89SsiN5O+1uFT4DIR6eG3XxjpZ7vEYROOB9PU\nnCQfExlMUMaYL7EfzAOxCc9aY8wvQRw6F6giIrcHuHaYLzlJ9jW2luQR4A+/IbJ12B6OtqQewsJ/\nGrYvzlPYoalg14K5Q0T8/0gagR02W5Jmvw98X6dgZ0i9F+T5wf4RNtwv5mLYWVWHsN+fQJJ7RtL+\nThtF+uQmOWEJdmr3El9MIwOcO4nzC0QqpfIp7dlR+VkwXfiLgTEi8jbwFbZW51bS1+dMw35Y/VtE\npmDrTG7l/JCDAVtgKyJDsR9wW0TkHeAANnmKxdZSBLug4WzgMd+5nwjymPc4P9U6lvM1ONHAzdih\np42+WE/7hotaAIv8zvE5dvp0cq+Nv59EZA02afgDaI4tiv5XkPGFAJ/5hunqYWuq1vkXfvtiOywi\ny3wxHyV9MpSZX4CHRSQKW6vTF7gCuD1NcbT/9f4Skc99x4Vgv2fXYocQ0/4cxfnanhWRf2NnTy0K\nUCeU7GNsgfgzIlKD81PPewCTjDG7s/HaskOHsJTKIZrsqPwsmOGGZ7Ef6v2xSUIcdmbQ8/7HG2NO\n+pKHl7FrmJzEJhZfYWs9Evz2XSsiLbG3qrgLKIGdMvw1NmkK1vvAo75zf5jJfv5xGhHphe01uA24\nHju77H/Y+zRtT3PsOuAqUtfl/CYiO7HDZ2mTnSnYdYk6YXtz9mITsheDeD0GmzDeil2vp5jvNd6b\nwf4zgeuAOf6z5oJwFDvV/RVsz9tvwF0BaorS/nz0w35/R2AThU+xM6V+IfV7vEFEnsD2HnXG9gbV\nwE6ZT3de3/ekB7Y27BZgMLbX7kFjTNp7Z5kAcWX1XGb7K6VygNiZokoVTSJyH3YKcZXk6ewqZ4hI\nT+AjoI1JvZBiZsesBsobY9KtgK2UUhdKa3ZUkeG39oz/42HADk10csUdwP+CTXSUUiq36DCWKkoW\niMg+4DtsceoA7MrG/TM9SmWLiCTX2HTFDhkqpZSjNNlRRckybA1If2zR70/ALcaY+Y5GVfjMBv7C\nrn009QKO17F1pVSO0podpZRSShVqWrOjlFJKqUJNkx2llFJKFWqa7CillFKqUNNkRymllFKFmiY7\nSimllCrUNNlRSimlVKGmyY5SSimlCjVNdpRSSilVqOkKykoplQURqQZEOh2HKnIOG2P2OR1EYaDJ\njlJKZUJEqrlcrvikpKSwrPdWKue4XK4EEamrCc/F02RHKaUyF5mUlBQ2a9YsoqOjnY5FFRFbt25l\nwIABYdgeRU12LpImO0opFYTo6GhiYmKcDkMpdQG0QFkppZRShZomO0oppZQq1DTZUUoppVShpsmO\nUkoppQo1TXaUUkopVajpbCyllLoI+/bt4/Dhw47GEBkZSbVq1RyN4UIMHjyYDz/8kL/++svpUIIy\nYcIEZsyYwU8//ZTtY6dNm8azzz7Lzp07KVasWC5EpzKjyY5SSl2gffv2UbduNAkJpxyNIywsnPj4\nrdlOeN59912GDBmS8jg0NJRy5crRqFEjunfvzpAhQyhRokROh5tCRBCRLPebOnUq4eHhDBo0KNdi\nycpff/3FhAkTmDhxYqr2uXPn8vHHH/P111+zc+dO2rVrx6pVq9IdP3jwYMaOHcu0adMYOXJkXoWt\nfDTZUUqpC3T48GFfojMLcGrBwa0kJAzg8OHDF9S7IyI8/fTTREVFce7cOQ4ePMiaNWu47777mDhx\nIosWLaJRo0a5EHfwXnvtNSpUqOBosvPWW2+RmJhI3759U7VPnTqVjRs30rx5c/74448Mjw8NDWXQ\noEFMnDhRkx0HaLKjlFIXLRoouAsOdunSJdWCiY888ghr1qyhe/fu9OrVi61btxIaGupghM6bMWMG\nPXv2JCQkJFX7rFmzuOyyywCyTAr79OnDhAkTWLNmDe3atcutUFUAWqCslFIqnXbt2jFmzBj27t3L\nrFmzUj0XHx9P7969KV++PMWLF6d58+Z8/PHHqfbxer2MGzeOOnXqULx4cSIjI2nTpg2fffZZptf9\n7rvvqFChAu3bt+fkyZPUqFGDLVu2sGbNGlwuFy6Xi/bt26fsv3v3bm6++WbKly9PREQELVu2ZMmS\nJanOuXbtWlwuF3PnzuWxxx6jUqVKlChRgl69evHzzz9n+V7s2bOH77//no4dO6Z7LjnRCUZMTAzl\nypVj4cKFQR+jcoYmO0oppQIaOHAgxhiWL1+e0rZlyxZatGhBfHw8jz76KBMnTqREiRJcf/31qT7E\nn3rqKcaPH0+HDh149dVXeeKJJ6hevTobN27M8HrffvstHTp0oFmzZixbtoyIiAimTJlClSpViI6O\n5v3332fWrFk8/vjjAPz++++0bNmSFStWMHLkSJ599lnOnDlDz549AyYUzzzzDEuXLmX06NHce++9\nrFixgk6dOnHmzJlM34evvvoKEcmR24XExMTw5ZdfXvR5VDYZY3TTTTfddMtgw45Pmbi4OJNWXFyc\nAQzEGTAObTaGQPFlZcaMGcblcmV6bJkyZUzTpk1THnfo0ME0adLEnDt3LtV+V199talbt27K4yZN\nmpgePXpkev3BgwebkiVLGmOM+eKLL0zp0qVNz549zdmzZ1Pt17BhQxMbG5vu+Pvuu8+4XC7z1Vdf\npbSdOHHC1KxZ09SsWTOlbc2aNUZETNWqVc3JkydT2ufNm2dExLz88suZxjlmzBjjcrlSHRtIRnH6\nGzZsmImIiMh0H2P8f7aIMfng/0FB37RnRymlVIZKlCiRMjX86NGjrF69mptvvpk///yTI0eOpGzX\nXnstO3bs4NdffwWgTJkybNmyhZ07d2Z5jTVr1tClSxc6derEhx9+GPTU7KVLl/K3v/2Nli1bprRF\nRERwxx13sGfPnnRTxAcNGkR4eHjK4969e1OpUqV0w15pHTlyBI/Hk+rYC1W2bFlOnz5NQkLCRZ9L\nBU+THaWUUhk6ceIEJUuWBGDnzp0YYxgzZgwVKlRItY0dOxawQ0sA48eP59ixY9SpU4crrriChx9+\nmB9++CHd+U+fPk337t2JiYlhzpw5eDzBz5vZu3cvdevWTdceHR2d8ry/2rVrp9u3du3a7NmzJ+hr\nXixjewuDmnKvco7OxlJKKRXQgQMH+PPPP1OShKSkJAAefPBBOnfuHPCY5H3btGnDrl27WLhwIcuX\nL+ett95i0qRJTJs2jb///e8p+4eFhdGtWzcWLlzI0qVL6d69ey6/quwrX748Xq+XkydPEhERcVHn\nOnr0KOHh4UV+dlte02RHKaVUQDNnzkRE6NKlCwA1a9YEoFixYqlmRGWkTJkyDBo0iEGDBnHq1Cna\ntGnD2LFjUyU7IsL7779Pz549ufnmm1m2bBlt27ZNdZ6MekGqV69OfHx8uvatW7emPO9vx44d6fbd\nuXMnjRs3zvR11KtXD7Azvxo2bJjpvlnZvXt3Ss+Tyjs6jKWUUiqdVatW8X//93/UrFmT/v37A1Ch\nQgXatWvHtGnTOHjwYLpj/G+bkXaBvfDwcGrXrh1w5pPH42HBggU0b96c6667jg0bNqR6PiIigmPH\njqU7rlu3bnzzzTd8/fXXKW0nT57kjTfeoEaNGtSvXz/V/jNnzuTEiRMpj+fNm8evv/5Kt27dMnsr\naNmyJcaYdHFdiI0bN9KqVauLPo/KHu3ZUUqpi7a1wF7bGMOSJUvYunUrXq+X3377jVWrVrFixQpq\n1KjBokWLUi2k9+qrr9KmTRsaNWrE7bffTs2aNfntt99Yv349Bw4cYNOmTQDUr1+fdu3a0bRpU8qV\nK8e3337L/PnzueeeewLGERYWxuLFi2nfvj1dunRh7dq1NGjQAICmTZvy+uuv88wzz1C7dm0qVqxI\nbGwso0eP5oMPPqBLly7cc889lCtXjhkzZrB3714WLFiQ7hrlypWjdevWDBkyhIMHDzJlyhTq1KnD\n0KFDM32PatSoQcOGDVm5ciWDBw9O9dy6dev4/PPPMcZw6NAhTp06xTPPPANA27ZtadOmTcq+cXFx\n/PHHH1x//fVZf2NUznJ6OphuuummW37eyGTq+d69e01YWHjyFGHHtrCwcLN379508WUleep58hYW\nFmYqV65sOnfubF555RVz4sSJgMft3r3bDB482FSuXNmEhoaaqlWrmp49e5oFCxak7PPss8+aFi1a\nmHLlypmIiAhTv3598/zzzxuv15uyz+DBg02pUqVSnfvIkSOmYcOGpnLlymbXrl3GGGN+++0306NH\nD1O6dGnjcrlSTe/evXu36dOnjylXrpwJDw83LVq0MEuXLk11zjVr1hiXy2XmzJljHn/8cXPppZea\niIgI07NnT7N///6g3qtJkyaZUqVKmYSEhFTtY8eOTfUe+m/jxo1Lte8jjzxioqKigrqeTj3P2U2M\n/c+slFIqABGJAeLi4uICLiqndz3P/9auXUtsbCzz58/nxhtvvKBzHD9+nFq1ajFhwoRUN08N1tmz\nZ4mKiuKxxx4L6t5YGzdupGnTpgBNjTEZr8SogqLDWEopdRGqVaumiUYRUKpUKR566CFeeOGFC0p2\n3nnnHUJCQhg2bFguRKeyogXKSimlVBAefvjhdAsVBmvYsGHs2bMn6AUTVc7SZEcppVShp4v4FW06\njKWUUqpQu+aaa0hMTHQ6DOUg7dlRSimlVKGmyY5SSimlCjVNdpRSSilVqGmyo5RSSqlCTZMdpZRS\nShVqmuwopZRSqlDTZEcppZRShZomO0oppXKNy+XK8E7n/mbMmIHL5WLfvn0pbe3ataN9+/Y5HlO3\nbt0u+LYN/fr145ZbbsnhiFRu00UFlVLqIhTkG4G+++67Kfd5+uKLL2jVqlW6fapWrcqBAwe47rrr\nWLRo0UXHmhERSbfKcW6sevzll1+ycuVK4uPjU9ri4+N56623WLFiBbt27aJEiRLExMQwbty45Jtx\npnjkkUdo1qwZP/zwA40aNcrx+FTu0GRHKaUu0L59+6hbry4JpxMcjSOseBjx2+Iv+IakxYsXZ/bs\n2emSnbVr13LgwAHCwsJyIsxM3XbbbfTr14+QkJBcvc6LL75Ihw4dqFGjRkrb9OnTefvtt7npppu4\n6667+PPPP5k2bRotWrTg008/TdW71KRJE5o1a8ZLL73EjBkzcjVWlXM02VGqiBMRNxDmt4WmeXwO\nOAGc9H09YYw560y0+cvhw4dtonMjEOlUEJCwIIHDhw9fcLLTrVs35s2bx7/+9S9crvPVDbNnz6ZZ\ns2Z50nMlIrme6Bw6dIhPPvmEN954I1V7//79GTduHOHh4SltQ4YMITo6mrFjx6YbSuvTpw9jx47l\ntddeS3WMyr+0ZkepQkJEwkXkchGJFZEBIjJaRF4WkQUiIRtEQuNFQveKhPwmUuyYiOeUiMsLeLFJ\nzGHgZ2AXsAWIA74EvgF+AvYCR4AzIi6vSLHj9lyhu0RCN4sU+0JEPhGROSLytoj8U0TuFJFuItJA\nREo48sbkhUigskPbRSZZIkK/fv04cuQIK1asSGk/d+4c8+fPp3///hhj0h136tQpHnjgAapVq0ZY\nWBj16tXjpZdeyvA6s2fPpl69ehQvXpxmzZqxbt26VM8HqtkJ5OzZszz11FNcfvnlhIWFUa1aNR55\n5BHOns06/168eDGJiYl06NAhVfuVV16ZLmkpV64cbdq0YevWrenO06lTJ06cOJHq/VL5m/bsKFUA\niC1eqAw0AqoBlwFVQKpAseqQVBkomfqokolQOQmqe+AygVKk7rAJ1ImTti2U87lQqs4dN5wsCSdK\nwomKfu3AX0nwZxIcNHDQA4kphRcixf4E1144uwvYg02g9vr9+5gJ9MmqclVUVBQtWrTggw8+oHPn\nzgAsWbKE48eP07dvX6ZMmZLumB49erB27VqGDh1K48aN+fTTT3nooYf45Zdf0iU9a9asYc6cOdxz\nzz2Ehoby2muv0bVrV7755hvq168PBK7ZScsYQ48ePfjqq68YNmwY9erV44cffmDSpEns2LGDBQsW\nZHr8+vXrKV++PFWrVg3qfTl48CCRkemzyfr161O8eHG+/PJLevXqFdS5lLM02VEqnxGRYkA00Bho\nAu4rwRMD3tK+PYDIc1BFbCJTBZv7+PKflH9HuAG3Ay/BRUqvcSLwC355TWnYewXsaQg7E+GAG876\n9TC7T4mEbIdz/8X2LMUBW3TYLPf179+fxx57jDNnzhAaGsrs2bO55ppruPTSS9Ptu3DhQlavXs2z\nzz7L6NGjAbjzzjvp06cPU6ZMYeTIkalqYrZs2UJcXBxNmjQB4JZbbqFu3bo8+eSTzJ8/P+gY33//\nfVatWsXnn39Oy5YtU9obNGjAnXfeyX//+19atGiR4fHbtm0jKioqqGutW7eO9evX8+STT6Z7zu12\nU7VqVX766aegY1fO0mRHKQeJSDlsUuNLbIo1BakHxvd/s+o5iPHAlWJ3uQKoChQr5lTM2ePGxlsV\naOP/hC8hMsDv+HXuhMOPTeDrBrBjOCQBrnMiIVs0Acpdffr04b777mPx4sV07tyZxYsX88orrwTc\nd+nSpXg8Hu6+++5U7Q888ADz589n6dKljBgxIqW9VatWKYkO2BlevXr1YvHixRhjgp51NX/+fKKj\no6lTpw5HjhxJaY+NjcUYw+rVqzNNdo4cOUKVKlWyvM6hQ4fo378/tWrV4qGHHgq4T9myZR2fhaeC\np8mOUnnEVwjcGIgFuQY8zQHfn80hSdAgySY2TTif2JQuIEnNhRLgEt/2N/8nitmhsc3AhmIQF0wC\n9IMxxpvHL6DQiIyMpGPHjsyePZuTJ0+SlJRE7969A+67d+9eKleuTERERKr26OjolOf91a5dO905\n6tSpw6lTpzh06BAVK1YMKsYdO3awbds2KlSokO45EeH333/P8hxZjZKeOnWK7t27c/LkSZYvX55h\nAXJ2kjTlPE12lMolIuLCZi3twBULnnbgLWkTm1bAVa7znTp1XODRCQOpRGDfp5Tp0FkkQO4TIq5P\nwSwDPjXG7Hcm7oKrf//+3H777fz666907dqVkiVLZn1QHkpKSqJRo0ZMmjQpYNKSVS1O+fLlOXr0\naIbPnzt3jhtuuIEff/yR5cuXpyRvgRw9epQ6deoEH7xylCY7SuUgEakMdAbpAp4u4C0FxZKglYH2\nbogF/uayhb8q+zJKgL4DVpWAT66Hb2+CJERC4uHcYuBTYJ0xxtnFcAqAG264gWHDhvH1118zZ86c\nDPerXr06n332GSdPnkzVu5M8c6l69eqp9t+xY0e6c8THxxMeHh6wlyYjtWrV4vvvvyc2NjboY/zV\nq1cvwyJmYwwDBw5k9erVzJs3j9atW2d4nsTERPbv36/FyQWI/iWp1EUQkVAR6SAiE0RCfgIOgLwN\nV94Io0vBGuC4C9a44Uls3YomOjkrArgaGAP8121n0M8FBtaFivcCy8F1TMS9VETu9k3P1/GHACIi\nInj99dcZO3YsPXr0yHC/bt264fV609X0TJo0CZfLRdeuXVO1r1+/nk2bNqU83r9/P4sWLaJz587Z\nGgrq06cPP//8M2+++Wa65xISEjh16lSmx7ds2ZKjR4+yZ8+edM+NHDmSefPmMXXq1CyTmJ9++omE\nhASuvvrqoGNXztKeHaWySUQ8QHugP7h7Q2IERHqhuwc6A52ASP2/5ZiywM2+zXjgR2BZKCzpBF90\nBq9AsX0ishhYBqy46F4fJ+tUL/LaaYeDBg4cmOUxPXr0IDY2lscff5zdu3enTD3/+OOPGTVqVKqZ\nWAANGzakS5cu3H333YSEhDB16lREhLFjx2Yr1oEDBzJ37lzuvPNOVq9ezdVXX01iYiJbt25l3rx5\nLF++nJiYmAyP7969O263m5UrVzJ06NCU9smTJzN16lRatWpFWFgY77//fqrjbrzxRooXL57yePny\n5URERNCxY8dsxa+co7+QlQqCryegJdAPPP3BWw5qeGGgB24AGntssa3KXwS7NFEj4CG3XQdoNbCs\nGnx8B+wfAe6TIjIXmA2sNsYkBnv2yMhIwoqHkbDA+dtFBFoPJhjB9KykXQNHRPj444958sknmTNn\nDjNmzCAqKooXX3yRUaNGpTu2Xbt2tGjRgrFjx7J//34aNGjAzJkzadiwYbbiExEWLlzIpEmTmDlz\nJv/5z38WFJraAAAgAElEQVQIDw+nZs2ajBo1KssamooVK9KtWzfmzp2bKtnZvHkzIsL69etZv359\nuuPatGmTanXq+fPnc9NNN6Ur0Fb5l+j6XUplTEQaAf2h2AA4VwUqemGAB/oDMWiCU5AZYBvwb2Cm\nF/Z4wHMEvO9jE59vjDFGRGKAuLi4uIC9BgX5RqBF0RdffEFsbCzbtm2jVq1a2T7+u+++o1mzZmza\ntClXbwS6cePG5JuQNjXGbMy1CxURmuwolYaI1AT6QrHb4FxdKJUIfd3QD1tz48Q6fSp3GWADNseZ\n5YXDHvDsBe907LT2JRklO6rg6d69O1WqVGHatGnZPrZfv34AfPDBBzkdViqa7OQsTXaUAkSkLDAA\nPAPB2xzCkuAGgf4C1wK5e4NClZ8kYgvL3wPmJEGCAKLJjspLmuzkLK3ZUUWaiNQF7gH330FCoYuB\nW4EeLjvLRxU9bqCDb3vZBS8C450NSSl1UTTZUUWOr9i4I7jvB7rYmVR3e2AYcIkW4Sg/JYFeaLKj\nVMGmyY4qMkSkOHao6kHw1oH6XngQuMWja98opVThpcmOKvRE5DJgBHhGQGIZ6G5gFNBWp4srpVQR\noMmOKrREpDnIKJA+UBy4ww13AzU1w1FKqSJEkx1VqPjqcTqBZzxwFVT1wv1uGAKUcjg6pZRSTtBk\nRxUaInIVuP8JiddA00R4FLjOo+viKKVU0abJjirwRCQaXM8C10MdL0wAuru1HkcppRToXc9VASYi\n1URcb4Nsgcuus4vA/eCB69BERymlVDLt2VEFjohEAo+C624oIzBW4A6dPq6UUiogTXZUgSEiJYBR\n4B4NoaEw2g33YRd+U8oZeiPQ8+Lj44mOjuaVV15hxIgRme47evRopkyZwunTp/MktuPHj1OjRg2m\nT5/ODTfckK1jz5w5Q1RUFM899xyDBw/OnQBVrtJkR+V7IhIK3AGesUAZuMdli48jnQxLKfbt20d0\n3bqcSkhwNI7wsDC2xsdnK+FxubKuYhARVq9eTdu2bS8mvAzPbSdPZt97773H8ePHueuuu4I+5oUX\nXuDSSy9NleisXr2aiRMnsnnzZg4dOkTp0qVp1KgRDz/8MJ06dUrZLzQ0lHvvvZfx48czYMAAPB79\n6Cxo9Dum8jURiQXPW5AYBQOxQ1bO/wWrFMDhw4c5lZDALCDaoRi2AgMSEjh8+HC2kp1Zs2alevzu\nu++ycuVKZs2ahf8NoqOjnXplGZs5cyYHDhwIOtk5c+YMr776KmPHjk3VvnXrVkJDQxkxYgSXXHIJ\nR44c4b333qNz587MmzePm266KWXfoUOHMmbMGObPn0/fvn1z8uWoPKDJjsqXRKQ8yAvAELgqEaYJ\nNHA6LKUCigYK2v3Q+/fvn+rx+vXrWblyJf369XMootyzYMECjh8/Tu/evVO1jxgxIt1w2/Dhw4mK\niuLNN99MlexERkYSGxvLjBkzNNkpgHQ2lspXxOoPnu1Q4jZ4E/jcrYmOUs5JSEjgiSeeoGnTppQu\nXZqSJUsSGxvLl19+meExEyZMoFq1aoSHh9OxY0fi4+ODutbbb79NTEwM4eHhREZGMnDgQA4ePJjy\nfMuWLfnss8/Ytm0bLpcLl8tF/fr1Mz3nwoULqVu3LpUrV87y+iVKlKBcuXIBh6o6derEqlWrOHXq\nVFCvReUf2rOj8g0RqQHuaZDYCW4w8C+BS50OS6ki78iRI8ycOZO+ffsyfPhwjh07xvTp0+nUqRMb\nN26kXr16qfafNm0ap0+f5t577+XkyZNMmjSJ9u3b8+OPP1K2bNkMrzNmzBiee+45br31VoYPH87B\ngweZMmUK33zzDZs2bSI8PJzx48dz//33c/ToUV544QWMMZQuXTrT+L/66iuuueaaDJ8/fvw4586d\n49ChQ0yfPp39+/fz+uuvp9uvadOmJCYm8t///pf27dtn8a6pfMUYo5tujm7YpPtBcCVA5XPwsQGj\nm275ZIszgImLizNpxcX5nnMwwDjIML7sGDlypHG5XAGfS0xMNF6vN1XbH3/8YcqXL29GjhyZ0rZt\n2zYjIqZ06dLm0KFDKe3r1q0zImIef/zxlLbRo0eb4sWLpzzevn27cbvdZvLkyamus2nTJuN2u82k\nSZNS2jp27Giio6ODel0nT540ImLGjBmT4T7t2rUzIpIS+5IlSwLut3v3biMi5uWXXw7q2hcj+WcL\niDHG+d/TBX3TYSzlKBFpCp44kAlwTyjE+xYFVErlFy6XC7fb3nbFGMPRo0dJTEwkJiaGjRs3ptv/\n5ptvJjLy/GzJ1q1b07hxY5YsWZLhNebPn4/L5eLGG2/kyJEjKVuVKlWIiopi9erVFxT7kSNHADLt\nUZo4cSLLly9n+vTpXH755fTu3ZvPP/883X7J53B6qQGVfTqMpRzhWzNnnL0reb1EeFugudNhKaUy\nMH36dCZPnsz27dvxer0p7YHqZWrXrp2urU6dOqxYsSLD8+/cuROv10v16tXTPSciVKhQ4QIjt4wx\nGT535ZVXpvz71ltvpXHjxtx9991s3rw54DkudMq8co4mOyrPici14Hkb3JXgaYH7PFDM6bCUUhmY\nPn06d9xxB3369OHxxx8nMjISt9vNuHHjOHToUI5cIykpiZCQEJYuXRowMSlVqtQFnbd8+fIAHD16\nNKj9Q0ND6d69O5MnT+b06dMUL1485bnkc/j3WqmCQZMdlWdEpBjwNPAItEmCN11Qy+mwlFJZ+PDD\nD2nQoAH//ve/U7U//PDDAfffsWNHurbt27cTFRWV4TVq1arFuXPnuPzyy6lSpUqm8WSnZyU8PJwq\nVaqwe/fuoI9Jnm2VduHF5HPkx7WHVOa0ZkflCRGpCp514HoY/gms1ERHqQLC7Xan6235/PPPA9br\ngK2/+f3331Mer1u3js2bN9OtW7cMr5G8Bs64cePSPZdcJ5QsIiKCY8eOBR1/y5Yt2bBhQ7r2QL1S\nR44cYdGiRTRv3pzQ0NT329uwYQMej4errroq6Gur/EF7dlSuE5HrwD0LKkbAPIFWToekVI7aWsiv\nfd111zFixAh69+5N586d2blzJ2+88Qb169cnKSkp3f5RUVFcffXVDB8+nBMnTjB58mQqVarE/fff\nn+E16tWrx5NPPsn48ePZsWMHPXr0ICIigl27dvHRRx9x//33pywA2LRpUxYtWsTo0aNp0qQJpUuX\npmvXrhmeu1evXsyfP5/9+/dTtWrVlPb27dtTp04dmjdvToUKFdi9ezfvvPMOf/zxB3Pnzk13npUr\nVxIbG0tERER23j6VHzg9HUy3wrsBIcBLgIHuiXAkH0wj1k237G4ZTz3fu3evCQ8LM/Zn3LktPCzM\n7N27N1182TFy5EjjdrsDPpeUlGSefvppU716dRMeHm6aN29uVqxYYfr27Wvq16+fst+2bduMy+Uy\nr7zyipkwYYKpVq2aCQ8PNx06dDBbt25Ndc7Ro0eb8PDwdNeaO3euad26tSlZsqQpVaqUadCggRk1\napT53//+l7LP8ePHTd++fU3ZsmWNy+XKchr66dOnTZkyZcxLL72Uqn3y5MmmdevWpkKFCiYkJMRU\nqlTJ9O3b12zevDndOQ4fPmw8Ho/54IMPMr1WTtGp5zm7iTEmF1MpVVSJSBR45oFpCi+IvTu5zmBQ\nBdFGoClxcXHExKS/KYTe9bxgeOKJJ/jwww/ZuvXC+sKef/553njjDbZv354nNwLduHEjTZs2BWhq\njAk8XqiCpsNYKseJyPXgngmVisN8gb85HZJSuaZatWqaaBQADz30EK+//jofffRRqjufB+Ps2bO8\n/PLLPPvss3rH8wJKv2sqx4hIKDABuAd6GHhHoIzTYSmlFKVLl77gHriQkBAOHDiQwxGpvKTJjsoR\nIlITPB8CV8Ak4C7RYSullFL5gSY76qKJSAdw/weqhMF8FzR1OiSllFIqha6zoy6KiAwC+RTah8N3\nHk10lFJK5Tea7KgLItZTwAz4uws+cUFpp8NSSiml0tFhLJVtIhICMg3MYHgGeFTrc5RSSuVbmuyo\nbBGR0uD+CKQdvAv0dzokpfLEha7PotSF0J+3nKWLCqqg+e5vtRyKXw6L3NDO6ZCUygP7cLmiSUo6\n5XQgqohxuVwJSUlJdY0x+5yOpaDTZEcFRUSuBM+nUKksfOoBveuvKkr2Ac6ukpx7fgBGJsLpnyHx\nTuA3pyNSKQ5ropMzNNlRWRKRruD+EK4IgSVuuNTpkJRSOSoe6OiFg0fA29EY86PTESmVk3Q2lsqU\niNwBshi6hMI6TXSUKpTqAl97oF55cH8lIm2djkipnKTJjgrIN7V8PDAN7nLBQhdEOB2WUirXVAa+\n8ECbcHCtFJEbnY5IqZyiyY5KR0QEGA+MgX8C/wLczgallMoDpYFlbujtAeaLyJ1OR6RUTtBkRwXy\nJPAEvAA8jK6ho1RREgp8IHCvAK+JyBNOR6TUxdJkR6Xi+8U2Fp4DHnQ4GqWUM1zYG/o+DfC0iOgv\nA1Wg6WwslUJERgPP2V9w+secUgrs74JnAO4yxrzmcDBKXRBNdhQAvr/cXoCngLEOR6OUyj8McD8w\nGWCwMeZdZ+NRKvs02VGIyH3AJHgc26ujNTpKKX8GGA68acD0NcbMdToipbJDk50iTkRGAi/DI9g6\nHU10lFKBJAK3GfggCcwNxpiPnY5IqWBpslOE+aaVvgYPYGdeaaKjlMqMF+hjYGEiJHU1xqx0OiKl\ngqHJThElIrcDb8B9wEQ00VFKBecs0DMJVpyFpE7GmC+cjkiprGiyUwSJyGDgHRiJXTBQEx2lVHac\nBromwhdnILGdMeZbpyNSKjOa7BQxItIZZAnc7oLX0URHKXVhTgAdE2HDSUhsY4z53umIlMqIJjtF\niIhcAe710DnM3uvK43RISqkC7RjQLhG2HANvK2PMdqcjUioQTXaKCBG5DDwboH4F+NINJZwOSSlV\nKBwGWnlh917wNjPGHHM6IqXS0ttFFAEiUhI8S6FiJCzVREcplYMigU88UDwK3HNERO8arPIdTXYK\nORHxgGsehNSHZR6o7HRISqlC53LgQzckdQKedzoapdLSZKfwewdJ6gwt3NDI6ViUUoVWJ2CSAA+K\nyG1OR6OUP63ZKcREZBjwOtWBvQDNgC+BECfDUkoVWgb4h4F3EyGpjTHmv05HpBRoslNoiUgssIK/\n4aYr8DWwDKAisBm41MHolFKF1xnsDK0NR8HbxBhzwOmIlNJkpxASkVq4iKM6JRmAi+Rywe3AXCAx\nFMxa4CrnglRKFWK/AVd64dD34G1tjDntdESqaNOanUJGRIrj4iNKEUEfv0QHoA5wO1DiDEhL4F1n\nglRKFXKXAIs94G4CMl1EdPVS5ShNdgqfKQgN6IuH4gGevQQYBlQyIIOxNwFVSqmcFgO86wLTH3jI\n6WhU0abDWIWIiAwEZtIT+3smM+eA/wBbADoCn6K5r1Iq5z0BPGOAHsaYT5yORhVN+ulWSIhIA4Q3\naIzhyiAOKAb0BtoBrARqAcdzL0BVBEwEbsKOl5YBQoHqwEDguwyO2QcMB2oDYUAE0BgYi733Unb9\n6LteNd/1SwF/AyZjM/xg/AP7q9EFfHUBMajUxgM9Dbjnikg9p6NRRZP27BQCIlICFxspRw3uwJPt\nmeU/AB8BpgSYb4DonA9SFQEVgFPAFcBlvrYt2Mp4FzAHuNFv/x+AttgkOwrbHZmATTCOAg18/y4Z\n5PVXAD2wSU000BD4E1jni6sdNrHP7G+81UAH3z7Gd2yrIK+vMvYX0MwL/9sG3ubGmASnI1JFi/bs\nFHC+wr83cFGLWy4g0QG71uAQIOwEuBoBi3I0RlVULMImKeuB+b5tK/AqkAjcQerelQexic5dwE5g\nHvAxsAdoAfyE7S0K1kjf+Z/H9vD8G1jqO3dNYC3wXibHn8H2MjX0XV/lnJLAXA9QH3jO4WBUEaTJ\nTsF3B9CPXriocBFnqYotXC6XCNILeDZHglNFSUsCL1g5HDtMehSbwCT71vf1CcB/sk5J4GFsz8q3\nBOc4sAMIJ30t7KXYRCir840HdgGvY8d5Vc5qDLzoAu4TkS5OR6OKFk12CjARaYTwMs3ImTtBlAGG\nYj+XeBzomwMnVQrOJw+l/drKB3FcMPuAvbltaCbPJw/XZ3S+H4AXsfU6OmyVe+4BOieBZ5aIXOJ0\nNKro0GSngBKREFzMpjxC5xw8cRjQH18v/hzsX2M6vK4uxntAPNAaW5uTbIjv69NAkl/7n8A/sb09\n/wjyGi5sYfIpYEKa534BXsP2Og0McKzBdpCWDXCsylmCnY5eujS439X1d1Re8TgdgLpgT2JowE1I\njve4u4Au2D+Cl3wPVAGzCTvWpVRWXsQWJp/E1uz8BHQH3k6z32PYGVf/Aj4BmmIT6y+B4sD72ALm\nYL2CTVzGYBfMbIgd3vocqAwswc76CnTcN8BMUvc8qdxxCTDLA107Y8cXX3Y4IFUEaLJTAInIVcBj\ntEOolIsXag6UA+YcgXO1wawE2uTiBVXh8Cmwyu9xRWw9T5kA+16NnSEVh+9utT4dIKg1FPyFYJOj\ndcA23wa2NyGWwLMMf8bWDMUCt2bzeurCdQHuBl59UUSWG2PinY5IFW46jFXAiEg4Lt6nEkm0zoML\n1sLeYqLUWZBrgGl5cFFVsK3Azr46iu1VuQJbA9YrzX6vYKeKH+b8TK6fsWviLMEOe+3IxnUfAgZh\nx2JXY3t1dvuu/bbvfEfSHHMXcBaYmo3rqJzxPBDl8tXv6B/eKldpslPwPAvU4Ebcqe57lZsqYEsa\nqhiQ4dgPCKWyUgrbc5M8RLUMWOx77g/sMJYbu8hTd9/+lbB/8f8fNjF5MshrbQMmYQuVl2F7eCKw\niwuOw/7M7sEOsSX7EDvV/RHg8uy/PHWRwoH3PZDYFL2dhMplmuwUICISC9xLp4ucZn4hIrB/NF8B\nttizLfavd6Wy4gFuwdbTfOFr24Ct16mBLYJP62bf18+DvMbn2CLnFtip5oHOZ9KcLznxWo4dxvLf\nkld8Hul7PDPIOFT2tABGC8h4EbnC6WhU4aVdhwWEiJTCxXtUIZGr8qxPJzUPcD0QCXy2DjuzZjO2\nsEepzET6vibfAuIP39eMCoKT248Gef4LPZ8AX2dy3s2+r7FBxqGy7ylgIbD9fRFpaow563REqvDR\nnp2C43lcXMoNuB39rgm2RrkP4P4ZpArwvYMBqYJhDfaHJ/nWSFG+r/HYWVtpfZNmv6wk77eJ82vq\nZHW+d7C9k4G25FlgX/geBzucprIvFDucZRpgC6yUynGa7BQAIvI3YDgdcVPW6Wh86mOXQCl+GuRK\n7K0BVNH1FXYWVtpEw4udWTwLW5Nzi6+9Obae5iTni4ST/QKMwiZHN5Pat9hZVZ3StHfG1uvsxs6u\n8o8jHpusBDpfVvTegXmjCfCogOtREanpdDSq8NEbgeZzIuLGRRyRNGAYHocGsDL2J3Y5lEOAeRJb\nDKqKnnexiwRGYouRy2NnWf0A/Iot+poDdPM7ZhV2NlYCdh2cZsBp7L21TvjOswZbyJpsLXZIKQr4\nX5oYZgF/xyZYNbFT14/4zncWWwT9H4L7Gy8WW9+jNwLNO6eAy73w2zJjvD2cjkYVLtqzk/8NJ4nG\n9MiHiQ7YUoh/AHXA3lvoelKvhquKhmuwIxD1sAnOfGxvTyRwP/bGnN3SHNMeu77OIGxB2FLfMbWx\n94pcR+pEJ5mQ+l5ayQZgh51uxiZQi7DDWjHYovqFZO9Xni7um7fCgckeSLxORLo6HY0qXLRnJx8T\nkUsRdhJDBPn975wk4DPs4rfUw36IBfqgUkqpjBigfSJ8sQ+80caYM05HpAoH7dnJ314ilFA6OB1G\nEFzYMoqegGwDuQxbP6GUUsES4BU3JEUB9zkcjCpENNnJp0SkPdCfzngKVAdJDHAbEHIMpA62u0cp\npYLVALhHwD1WRC5zOhpVOGiykw/57mj+BlVJDLjeWn5XA7vichkvSEdgisMBKaUKlrFAGQ/IC05H\nogoHTXbyp3sw1OQ6h9fUuRjlsQlPNYztjf67wwEppQqO0sALHjD9RKRtlrsrlQUtUM5nRKQcwh6a\nUZLuTkeTAxKxt0baCHAVdraMLtytlMpKEnBVIny3DbxNjDFepyNSBVdB7TcozB7FTTjXOB1GDnFj\nl1K5Fuyy/FWB352MSClVILiAqW5IbAAMczoaVbBpz04+IiLVEHZyDcVo53Q0uSAemAckhoFZh11E\nTimlMjPUwIwTkFjLGHPI6WhUwaQ9O/nL0xRHaOl0GLmkLjAUiEgA+Rt2xVullMrMcwLh4cD/OR2J\nKrg02cknRKQxMJBYPIQ6HU0uuhTbIX2pARkIPOJwQEqp/K0CMM4NMlTvm6UulCY7+YUwgbIkEuN0\nIHmgJHZyVjTABKAreosJpVTGhgHlk4BHnY5EFUya7OQDItIew7V0yqf3v8oNxYDeQFuAZdiba51w\nMiKlVL4VDoz2gAwRkepOR6MKHk12HCYigosXuYxE29NRhLiw94K8EXDtAqmMrWJWSqm0hgNlDDDa\n6UhUwaPJjvO6kMSVdMBdZG+yfAUwGAj9C1wNgCXOxqOUyocigEc8vtqdqk5HowoWTXYc5OvVeZLL\nSKSG09E4rBp2WL5sIkh3bC2PUkr5GwGUEuBhpyNRBYsmO85qSxItaFuEe3X8lQVuB2qCnaV1q6Ph\nKKXym5LAg25wDRORyk5HowoOTXacJDxBBbzUcTqQfCQM6A80B5iNvY36WScjUkrlK3cD4S7gIacj\nUQWHJjsOEZHmGDpyDR7t1UnDDXTHzkhnE3AZ8IuTESml8o3SwANucI0QkUucjkYVDJrsOEV4nLJ4\nqe90IPnYVdiRrGKHQWoAXzockFIqf7gXCHMDDzgdiSoYNNlxgIg0xNCLNnj0O5CFy7F1PCXPgrQB\n3nI4IKWU88oCo9zgultEKjgdjcr/9KPWGY9SEi9XOB1GAVERO1OrsgEZiv2rTilVtI0CQkN8/1Aq\nU3rX8zzmWx9iD11w0cLpaAqYc8Ai4AeAWGAlmq8rVZSNBl48DYmVjTHHnI5G5V/6SZH3hlEMw5VO\nh1EAFcOuttweYDVQA9Dfb0oVXfcCJgxdp0JlQZOdPCQiobi4kya4C/WdzXOTYO+ndTPg3gdSBfjR\n2ZiUUg6pBPQEPHeJiM5rVRnSZCdv3UQS5ewaMuqiNACGAGEnQRoDHzkckFLKGcMFvNHY+ZtKBaTJ\nTl5ycQ/VSaSi04EUElWwhcsVkkBuBMY7HJBSKu91Aqp4sb8NlApIk508IiJNSOIqrsLtdCyFShng\nH0BtgKeAm4AkJyNSSuUpF3CnB1z9RKSM09Go/EmTnbxzFyXwUtfpMAqhUKAf0BJgAfY26glORqSU\nylNDAAkBBjgdicqfNNnJAyJSFmEgzfFov04ucQGdgR6AbAGpDOx1NialVB5JKVQeoYXKKhBNdvLG\nYIQQYpwOowhoCgwEih0FuRxY42w8Sqk8Miy5UFlXMFPpaLKTF1zcTj2gpNOBFBE1gTuA0udA2gOv\nORyQUir3dQKqerH/+5VKRZOdXCYiV5BENI313uZ5KhL7K6+qAbkL/f2nVGHnAoZrobIKSJOd3DeA\nMBLtbCGVp8KB24DGAG8CVwNeJyNSSuWqv6OFyioQTXZykYi4cXEbjXBrYbJDPEAvbA83XwHVgcNO\nRqSUyjWXoisqq0A02cld15DEJXp3c4cJtlPnFsDzC0g14DtnY1JK5ZLhAt56aKGy8qPJTu4aQGm8\nVHE6DAVANHYBwvDT4GoKzHE4IKVUzusIVPRi76CnFKDJTq4RkeIIt9AEj5Ym5yOVsIvKV0wC6Qs8\n5nBASqmc5QJu8kCxm3UoSyXTZCf39MQQrkNY+VApbB1jXYDngOvQW0woVZhcD5yrgm96glKa7OQW\nYSCV8VLe6UBUQCFAH6ANwCdAPeCEkxEppXJMO6BEIjbrUUqTndwgIqWAzlyBx+lYVCZcQAfsr0PX\nDpDLDOxyOCil1MULAXq6oZjW7ShAk53c0gWDh3pOh6GC0gQYBIQeF1z1gE8dDkgpdfFuAM7VF5Ga\nTkeinKfJTu64nop40TU8C47q2EWWy3hBugATHQ5IKXVxugDFktChLIUmOzlOREIQelJfh7AKnHLA\n7UAUwAPY7h6lVMFUArgW8PR2OhLlPE12ct41GCLsTB9V4BTHLjTfFGAm0Bw462RESqkLdqMLvC1E\npKLTkShnabKT866jBF4udToMdcHc2NnoXQA2AFWBg05GpJS6ID2wS6jT0+FAlMM02clBIiK4uJ66\nupBggSfYxeb7A8V+B4kCvnYwIKVU9lUArk4C101OR6KcpclOzqpLEtWo43QYKsfUAYYCJc6AtATe\ndTggpVT23OQGOopISacjUc7RZCdndcdNEjWcDkPlqEuwt5ioZEAGY4uXlVIFw/VAkgfo6nQkyjma\n7OQkoRPVsetZqcKlBDAEaAB2Wnon9BYTShUEUUADL9DN4UCUgzTZySEi4gHaUkPf00KrGHATdiV6\nVgK1gOMOBqSUCk4HDxSLdToK5Rz9YM45MRiK2zVaVKEl2GTnJsC1B+QyYKuTESmlstQaOFdNRHSe\nbBGlyU7OaYeHJCo7HYbKE42ww1phJ8DVCFjkcEBKqYxdne4fqmjRZCenCO2oil2jRRUNVbGFy+US\nQXoBzzgckFIqsMpA1XPYLh5VBGmykwN89TrXaL1OEVQGOzW9FsATQF9Hw1Hq/9u77zC5yvL/4+97\ndlMghBZKqKEECC30JqFXQZSmNEGliYiiPwW+iohioQgKCiIiTapSFQkQJJAACZEOIYSS0JIQSE9I\nITtz7t8fz1my2WydnZ1n5szndV1zzWZmzpk7C9n57H2eIq3Zuwf02Dt2FRKHPpxLYzuc5TVep0b1\nJiw+uCvAP4BtgEUxKxKRZewO5Aeb2QqxK5HyU9gpjb2o03idmpYjbC9xKGCvgq0LfBi3JhFpYgjg\nOesymdsAACAASURBVGDn2JVI+SnslIKxD+uB9jkXdiJsJNpjBthAYGTkgkQk2BxYMY/G7dQkhZ0u\nMjPDGMIG+l5KamPgNGDFxWB7A9fFrUdECB93e+Sgbs/YlUj56QO66waQsKIuYclSVgdOB9Z1sDOA\n70YuSERgSA74QjqpRGqIwk7XbQ/AWpGrkMrTB/gGMBjgz8CeQD5mRSI1bghQWI6wUpbUEIWdrtue\n5cmj/XSlJfWEfQj3A3gK2BCYGbMikRq2I1CfoHE7NUdhp6uMHVhb30dpgwF7AF8D6ialM7VejVuT\nSE3qDezgYFpJucboQ7qrjJ0UdqRDtgBOAZZfCLYdcE/kgkRq0RfqoMeusauQ8tKHdBeY2Vok9NN4\nHemwtQlbTKyegH0V+HnkgkRqzSCgYT0z6xW7EikfhZ2u0eBk6bwVCVtMbArwK8KgniRmRSI1ZBDp\n4oIbx65Eykdhp2u2pxd5VopdhlSdnsAxpHsw/wvYElgQsyKRGrHZMl9I9insdM22rEUOi12GVKUc\ncADwFSA3Hmwd4N24NYlk3hpA3wKhxSM1QmGnK+rYktX1PZQu2g44Ceg5G2xT4PHIBYlkmQGbOers\n1BR9UBfJzHIkbMiqsSuRTNiAsOLyynmw/YGr4tYjkmlb1kOPrWJXIeWjsFO8tXF6KuxIyfQjBJ71\ncfgBcHLkgkSyajPANzUzDUKoEQo7xdsECB9QIqWyHHASFub53QTsiraYECm1QUC+L2EAj9QAhZ3i\nDcSAlWOXIZlTBxwGHAQwBlgP+CRmRSIZoxlZtUZhp3ib0JcGtHeudAcDdgOOA+qngg0Ano9bk0hm\nbAzkHM3IqhkKO8UbyGrUxS5CMm4zwgKEfRaB7QzcFrkgkSzoBayXR52dmqGwU6w6tqCfvn9SBv0J\nW0z0d7ATgfMiFySSBVvVQ27z2FVIeejDughmZiRsoJlYUjZ9CZOzNge4DPgi2mJCpCsGGdRvGbsK\nKQ+FneKsitNL20RIWfUAjgb2BHiEsLnWpzErEqliGwEN62j6eW1Q2CnOmgD0iVyF1J4csC9wJJCb\nALY28GbcmkSq0uqA1xG25pWMU9gpTn8AVohchdSuwcA3gV7zILclMDRuPSJV5/NF0laLWYWUh8JO\ncUJnR2FHYlqfMHB5lQLYoYSxPCLSMast84Vkl8JOcdakjoSescuQmrcKcBph+AHnASdELUekenye\ncbQOfg0oKuyY2XAzW2btYDNb0cyGd72sitefPhTQsDapBL2B44GdAe4AtgcWx6xIpAroMlYtKbaz\nsze02NfoDexRdDXVY01WUNSRClIHHJLe7CVgHWBK1JJEKlsvYLkCCjs1oVObHZjZ4CZ/3MLM+jf5\ncx1wMDC5FIVVuP701UYRUoF2BlYF/jEd8huCDwd2j1yUSKVaNYHJuoxVAzr7gf0y4OmtpctVC4Hv\ndbWoilfHOhqcLBVrIGEcz22LYd4e4NcDp0QuSqQSrQZMVmenBnQ27GxI2KJwIuF3yGlNnlsMfOLu\nhRLVVslWY/nYJYi0YQ3CTK07HSafCv4qcFXkokQqzZr16DJWTehU2HH399Mva3sWl7O8ZmJJxetD\nWIvnX8BrfwReBR6n1v/5iiyxukH9GrGrkO5X9LgTM9sE2IfwO+RSPz3d/aIu1lXZnN70iF2ESAfU\nE1ZbXh0Y/iShOfsKsMxkSpEa1A/IKezUgKLCjpmdBlwLTAemEsbwNHIgs2HHzHJAL4UdqRpG2E+r\nH3DfB5CsC/4ssFXcukSiWw1wDVCuAcV2dn4GnO/ul5aymCqxHIDCjlSdLYGVgDvmw8JtwO8Bjohc\nlEhMqwH5lc3M3N3bfblUrWIv3q8C3F3KQqpIGJqsMTtSjdYlDFxePQE7kgw3YUU6YGXSzUC1rXPG\nFRt27gYOLGUhVSSEHXV2pFqtRJiJPhDgQuAoIIlZkUgkn/8g17ppGVfsf+B3gF+Z2a7Aa0BD0yfd\n/Y9dLayChd8AFHakmvUCjgMeA0bfR9hG/Tkar9KK1Ia6Zb6QbCo27JwOfArsld6aciDLYUedHcmG\nHHAQYdjCf14H1gF/CRgQtSyR8lHYqRVFhR1337DUhVQRhR3Jlh0Io/DumgUNm4API2x/J5J19ct8\nIdmk1cU6TxuASvZsROjXrtQAtg/w58gFiZSDOju1oth1dm5s63l3P7m4cqpC2A5DkxQla1YjBJ67\ngA+/C/4EcHjcmkS61djGL9TZybhi/wOv0uzPPQgrlK1MyxuEZonCjmTX8sBJwH+gx8v30MA9sSsS\n6VZ1QEEDEzKv2DE7y6xElq4sfC0woatFVbgQdjRTV7KqHvgKNOSg7sVw3fYC4EuRyxIptTHAmeHL\nhjZfKFWvZK07d0/M7PfAk8BlpTpvBQoxR50dyTIDvgyFHSB3G1y4MKw3cRXLtnVFqtW0JV8W4lUh\n5VDqAcobk/1rn7qMJbVjHUjOAbaFO4BBwEORSxIplSYJJx+vCimHYgco/775Q8BawKHALV0tqsLp\nMpbUlhxwOBR2hOm3w5cWwjeAK9He6VLdmiQcdXYyrtjOznbNboPTx38E/KAEdVUyXcaS2rTuki7P\nbYQuz8ORSxLpikKLX0oWFTtAeZ9SF1JFdBlLaleTLs+02+CQRfBN4A+oyyPVR2GndnRpzI6ZrW5m\nQ9Lb6qUqqsJ9BugKr9S2dSE5F9gGbiV0eR6JXJJIZ3225MvF8aqQcigq7JhZn3RhwY+Akeltipnd\nYGbLl7LACjQbgEWRqxCJLQccAYVTYFpv+CJwMjAnclkiHTUTsDDtfEHsWqR7FdvZ+T1hA9DDCN3r\nlYGvpI9dUZrSKtZcQGFHpNF6aZdnMPyd0OV5NHJJIh0xA6iHOe6ugQkZV2zYOQo4xd0fdve56W0o\ncBpwdOnKqzzunsdY2LT/KVLzcsCRUDgZPukNBwOn0vibgUhlmgFYuJOMKzbsLA983MLjn9C4K3iW\n5Zirzo5IC9ZPuzxbw82ELs+wuBWJtGoGkCy1tqBkVbFhZzTwSzPr3fiAmS0HXJg+l3WzFXZEWpED\njgpdno97w0GElq+6PFJppoPnwy/pknHFhp0fALsDk8zscTN7HPgwfezsUhVXsVxhR6RdTbo8NwGb\nA49FLkmkqWlhXq0uY9WAosKOu78GbAL8BHg5vf0fMNDdXy9deRUqYbrCjkgHNHZ5vgVTe8GBwOnA\nvMhliQBMD3cz41Yh5VDsdhE/Aaa6+/XNHj/ZzFZ390tLUl3lms1CCkBd7EJEqsIASM4D7oMbxob9\ntW4B9o9cltS2WSGOq7NTA4q9jPVtYFwLj78OnFF8OVVjNgu1O5ZIp+SAoyH5JkzthR9A+GGhLo/E\n0ADMD7+wKuzUgGLDTn9aHtQ1jbAhaNZNZR4WuwiRqrQBJOdibAnXE8byDI9cktSeJteuFHZqQLFh\np3EwcnO7A1OKL6dqTGYR9doyQqRIdcBXl3R59gO+A3watyqpITNa/FKyqtiwcz1wpZl9y8wGpLeT\nCfsBXt/OsVkQAp1+Mot0zQZQOBdjC/grocvzROSSpDYo7NSWogYoA78D+gF/Bnqmjy0CLnX3i0tR\nWIWbDISFQ7TVs0jX1AFfg2QiTLkL9l0cujyXAStELk2yS2GntlhXtgQxsxUIv4wtBN5295rYRMHM\nVgFmcjSwVexqRDKkANwDuTdgHcJeW3vHrUgy6k/A2dDgsJy7F2LXI92r2MtYALj7p+7+nLuPrZWg\nk5qNsUDbO4uUWB1wDCQnweSesA9wFjA/clmSPW8BPeA9BZ3a0KWwU6vc3cnxIbNjVyKSURul6/IM\ngmuBLYARkUuSbHkTksVhuRSpAQo7xSowgTkUfw1QRNpWBxwLyYmhy7M38H2y0eUZD5wArA30BjYE\nvkfnBo8UgF8AXwI2BlYElgMGEtYvmtjGsWOBE4H1gV7psTsDVxLWn6kF48K38K3YdUh5dGnMTi0z\ns2tYjdM4ix6xaxHJvAJwN+TGw3qEsTx7Ri6pWMOBLxMGOg4idK3GAm8C6wLPEkJQe+YDfdPb4PSY\nxYS9e94H+gCPAl9odtxjwGGEULM5YdjhHOApYAEhVP6XbP8mvAhYHnA41d1viF2PdL8s///c3SYw\nizqtoyxSBk26PJPSLs/ZVF+XZyFwfHp/IeEayt3AG8A5wCTglA6eqzfwDDCLEFT+AdxP6Oj8hPC9\n+XYLx51FCDqXEELWXcDDwDvARoTLhbd2+m9WXSZAY1tenZ0aobBTvHEUyGmQskgZbQyFc8E3g6uB\nLQkf9NXiPsLS85sBP2/23G+ADYBhwGsdOFcdsBvL/hA34CJCGBoHS/2Imgu8TehqnNPsuP6EIOTA\ncx14/2r29pIvFXZqhMJO8cLAtmmRqxCpNfXAcZB8HSb1gL2AHxIuwVS6F9L7li7B1bNkWfp/dfF9\njBCG6gmXsxqtQBij05rGQQ39uvj+le4toC40v1ra9kgySGGneJMwFirsiEQyEArngW8KfySMPXkm\ndk3tmE8IIqu08nw/QuB4pYvvcwkh/B3D0ivH5ggDkxcQFm1sagpLVok9sYvvX+neAnJhbTgNWq0R\nxa6gXPPc3a3O3mAa28euRaRm1QPHQ/I2fPBP2KMBfgD8mnCpptKsTggz77fy/LvpfWvPt+Y8Qoti\nLvBqevyJhGn7zV2d1nABcAshJM4FRhIGOQ8lzOjKsjeg0BCu8kmNUGenKxJe5WNtByoS3SZLujxX\nET7AR8WuqQWNl68eYqldt4GwB81jhM7PvE6e9z7CDLUHCAOU1wd2IUxFb65nWseGhCnw9xDGCX1G\nWMRx806+dzV6M+Q9jdepIQo7XTOOaZhW2xGpAI1dnhPggx4wBPgRYeZTpTgQ2J4QZg4mDASeD4wG\nvkiYYe90/gfz2+mx04BHgFWB7wJntvDac4BvEAYwP0Ho6rwLnA/cSPi+ZXmzqDnAjPB/i8JODVHY\n6Zpx5KnTjCyRCtLY5dkkLJK3FSFMVIr7CDW9QOi+9CUMTJ5OWCQQWh/T055VgQMI6+SsB1zH0jO7\nxgN/IAxUfoTQ4elD6AT9khCQ3gMuL/L9q4FmYtUmhZ2uCdd8NUhZpLLUAydAchy83yOEiXOojC7P\n+oSF//5JmEV2BmEczThgjfQ1W3bxPVYkLBwISw/aHgkkwK6EqebNfZXQWRrZxfevZG8s+fLt1l8l\nWaOw0zXvYyzk49hliEiLNku7PAPh98DWhBWKY8sBRwFXEGZAnQmsTAgmRml2el8tvf+0yWON44RW\nauWYxsdnleD9K9WzQA+Y6O5zY9ci5aOw0wXunmA8z2SN2hGpWPXA10OX570eYfuEcwlbBlSSqcC9\nhJByZAnO92R6P6jJYxuk9y9Biz+0/tfsdVk0EvIN2le25ijsdFXCM3xAIXYZItKOzdLVlweGjsrW\nwJgIZbxOmPnU1CTgK4QuzBUsvfDfNYQZUuc3O2YoLY9FWpi+diQwgDAoutFBhPE67wI/Y+nA8yZh\nVWcjXM7Kok+BcSH+VtIwLikDrbPTdWOYTz1zaL03LCKVoQehyzMe3r0HdsuHsTy/JMxOKofLCXtY\nbQ+sRVgf52nCJp4/D+UtZTohiHzU7PHnCHWvA2xL+PEzlTAeaCZhTZ+7CVPNG61CuGx2MnAxYT+t\n7Qizr0anNRxKmK2VRf+Dxu0MK3FlAulG6ux0XfjlcFLkKkSk4walY3k2DuFjMEsu4XS3IwiX0t4k\nXLYaCxxCuOx0YSvHWHpr6ijC1Pp1gOcJweZ5wvo5Pyd0kHZo4VxfJ4SrrxIu5f2bcFlre0IQ+hfZ\n/WAYDdSFBs8b7b1WssW0WnbXWZ1NZhfW5qDYlYhIp70BdfdCkg8rEf+CtvePkup1CCTD4PG8+4Ht\nv1qyJKsBvrwSnuZDjdsRqUqbp12ejeBSQpcn67t+16IEGAVJofK3UJNuoLBTGs/yEaa4I1KlegAn\ngX8NJtSHdWh+yrIDiaV6vQXMCeNUNV6nBinslMazFMhpvR2RKrdFmLGVbBh2Dt+GMA5Gql86/cqJ\nMwlPIlPYKY2XgDwfxi5DRLqsJ/CN0OV5pz5s6XA+6vJUu1FADxivxQRrk8JOCbj7IoxnmaDFBUUy\no0mX52LC9O4XYtckRRsJDQ3wVOw6JA6FnVJxHuFdEo3bEcmQxi7PV+HtOtgZuICwHo1Uj9nAW2Fk\nlsbr1CiFndJ5jAbqtN6OSAZtGWZsJRvAbwhdnhcjlyQd12Q/NK2cXKMUdkrnBXLMZWLsMkSkW/QE\nvgl+NLxVBzsRFu9Tl6fyDQN6hMWqtdN5jVLYKRF3L5AwjLfJx65FRLrRVmmXZwD8mrDdwkuxa5I2\n/Tts/vlv1yq6NUthp7SG8RH1LIxdhoh0q57At8CPgjfTLs+FqMtTiSaEWz1h71SpUQo7pfUYDrwX\nuwwRKYutw4ytwgD4FWF/qZdj1yRLeRgwyAP/7cp5zCzp5K3oQQ1mdnl6jiM7ccyW6TH/LvZ9S8nM\ntjazvJld0OzxHczsz2Y2xsymmNlnZjbbzJ4xs++aWV0L5zoh/bsdXGw92vW8hNz9Pauz95jABmwe\nuxoRKYtehC7PazD+AdixEGZs/ZQw/UfiehCSHDyVd5/XxVPd3MJjQ4CNgVdYNudO68J7eXqLwswu\nB/4fcLS731fkaS4D5gBXNnt8f+DbwLvAq8AMYE1gd2A34EgzO9Ddm85tvpMwRO5S4JFiilHYKbWE\nobzF6Tj1y2xTLCLZtTUUNgVuh19+APcAtxFWYZY4FgBPAAX4T1fP5e4nN3/MzG4ihJ0H3P2irr5H\nBelS2DKzPYCDgF/7siHzHuAOd/+w2TFrASOAvYFTges+L8Y9MbPLgOvN7AR3v72zNekyVuk9yFzq\ntXWESA3qBZwMfgS8UQc7ABcBDZHLqlWPAw3hc07jdTqnq7+qn0kIS7c2f8LdJzQPOunjHwFXpO89\npIVz3k1YyPw7xRSksFN6wzHmMS52GSISzTZQOAcK68EvCKHn1cgl1aL7gR4wEXgzVg1mljOzE83s\nn2b2tpnNN7M5ZjbKzJbpFrVw/J5m9nh6zGwze8jMti2ijj3M7H4zm5qOk3k/HTvTv9nrphMuYQHc\n02wc0vYdeJ9VgMOBl9y9s1P9G38vmNP8iXSbj6HAbmbW6YEiCjsl5u6Lce5jrKagi9S03sApocsz\nLu3y/Ap1ecolD9wfppz/M/KU81WAW4A9gA+BBwibkW4D/M3M/tjGsQcQGlSrEi7FTQC+CDxjZrt3\ntAAzO4dwiehQ4J20hjnAGcDzZrZBk5ffDoxNv36CMFbpZuAmYHoH3m5/Qo/zyY7Wl9a4OvADQkfo\ntlZe1njOQztzblDY6S73MpN6PoldhohEl3Z58uuF6ek7Aa/FrqkGPA3MDuNS749cygLgS8Da7r6v\nu5/g7gcCGxFCxXdb6dQYcDpwvrtvlx63A3AOsBxwS0szl5Y5idm+wCWEoLStuw9x92PcfTBwNrA2\n8NfG17v72cCj6R+vcfeT09sp7v5BB/6+exICy3Pt1LWVmd1sZreY2aOEecyrAMe6+7OtHPY/wvdl\nrw7UsRSFne7xGMYC3ohdhohUhMYuz+EwNhemqP8G1P7tRvcD9fAx8HzMOtx9obsPbd5dcvePCRP3\nDPhKK4e/4e6XNTvuCkJI2pCOdTh+mt6f7O5LDbBw9z8Bw4H9zGzjDpyrIwan92+187p1gBOBrxO6\nQb0JHa+2jhuf3nf6Mp7CTjdw90U4D+hSlogsZduwLk9+3fAptyNLrhdI6ThwN+TzcI+7J7HrATCz\nHc3sPDO7xsxuTGdyfSt9epMWDkn/Gi26ixCS9mjnPXsRBvvOdPfWdnxvfHznNv8CHbdGej+zrRe5\n+6PuXkdYoWEg8BPCDK6nzazFWtx9DlAAVutsUZp63n3uZRrHMwPoF7sUEakYvYFTwV+CsQ/Cdgn8\nEjgX/UAulReAjyrjEhZm1puwTsxXaHk6twN9Wzn8/VYefy+9X7udt1+bsN73qmbWVuhziggQrVgp\nve/QukZpGH0XuMzMPiKMb7qMMAW9JZ8CK5pZb3df1NGi9G+r+zyCsYhx9G47e4tITdoOCoOA2+Fn\nk8LiI7cCW0YuKwtuBephZh5Gxq4F+Bkh6DxH6F68DMxO147ZjpDNumtVtsarN7OBf7Xz2lLNWJtD\nWCSwL+10d1pwF/A3YFczs1YGlvcFFncm6IDCTrdx9wVm9h/Gcjh76PssIi1YjtDleRFe/Q9sm4R1\nec5BP5yL9RlwCxTycKO7V8Lkt8MJnZOj3H1Ss+c2aufYAe08PqWd4z8iXPZZ1NKiiN3kE2BTwgyy\n1jpTLXL3BjObRxio3AtYKtCY2UqEANeRWWFL0Zid7nUbH1PP1NhliEhF2z6dsbUOnE8YPKGluorz\nL2AO1AE3xK4ltQqQQIufBMe0cZwBX23luWMJAerptt7Y3RcAo4H+rY2DaUXjnrbFZO5X0vvNOnug\nmW1FCEkfttK5aVxfp9Nb0CnsdK+h5JjBS7HLEJGKtxxwGviX4bVcWITlUjRjq7Ouh6Qexrj7+PZf\nXRZvET5rT2v6oJmdBBxF29sybJ6ukdP0uP8HbE0Yt9ORbTB+TQhOd5jZrs2fNLMVzez0Zg9PSY/p\ndGAhDHg2wioLyzCzH5jZMuOD0qBzG+H7cX0r524MbCM6W5TCTjdy9wYSbuRlCvqJJSIdsj3kfwz5\ntcMAj11Bq1h00PvA45DLN1k3pgzaG29zCeED/BozG21mt5vZi4RF+hq3R2jNdcBvzeyl9LjngcuB\nhcA3mm2W2SJ3Hwb8iHDpa5SZvWxmd5vZXWY2hrBh6bVm1jQPPExY//ICM3vQzK5Pb+u1936E3eU/\no/UBxucDU8zsubSGu83sOUJHaGtCc+7SVo7dh/C9fKgDdSxFYaf73cRn1MVbrFxEqs7ywOngX4JX\n0i7PZYTBF9K6m4FcGOfxzzK+bZurM7v7o4QP6RGEsSyHEAbxHkqYedTapptOCA77AbPS1w8kBJEh\n7t7SJawWz+XufyDk5tuAlQmLHDaubXMjcFDTKfruPpHQdXqOsIDfyelt9bb+rumxs4D7gG3NrKXO\n0I8J4/H7AgcDhwFrAQ8CX3P3I919mfZAOl7nYGCUu3c6/1vcVbRrg9XZGDZkR05UuBSRTlpA+Iia\nEtbluRUYFLeiipQA60N+Mvzd3U+JXU8tS7eyeAq42N3PL9E5TwP+Apzo7nd0+niFne5nZqcC1/ND\nlqxAICLSGc9D/VCwJKy+/P8Io3Al+C9hIylgd3cfFbUYwcweAnYDNkg38ezKuXKE1ZMXANsVs9eZ\nOg3l8U+MRZ0fPy4iktoxjOVpWCssQPgFIm7lXYH+Bt4jbHI5OnYtAoT/TfsSNvfsquOAjYHzit3U\nVZ2dMjGzm1mJEzibekVMEemS56D+Ycgl8FvCp0ktd3lmAv0haQgfhpfHrkcqjz52y+cG5lDPu7HL\nEJGqtxPkfwSL+4fRnrvT/q6LWXY7kA8Dc/8euxapTAo75fM0OV5nNBWxKZ2IVLk+wBnAIfCChTm7\nv6c2Z2xdD3mDB939k9i1SGVS2CkTd3cSfsc75JgWuxoRyYydw1iexf3DYipDqK0uz/PAa1CfVM6K\nyVKBFHbK6y5yzODZ2GWISKY0dnm+CM+nXZ4/UBtdnkvB6+ED4JHYtUjlUtgpI3f/jIQreZmE+bGr\nEZHM2SUdy7NmmJq+B2F6Ula9BdwLlofftrQQnUgjhZ3y+wsJeV6IXYaIZNIKwHeAg+E5g62AqyCT\ngwUvBerCDti3xK5FKpvCTpm5+3Scm3mWvPbLEpFus2vo8ny2ZpianrUuzyTg7+B5uLSVHbJFPqew\nE8eVLKCesbHLEJFMa+zyHAT/S7s8fyQbXZ4rwt08wmaZIm1S2InA3d/AeJRR5NveQk5EpAR2S7s8\na8DZwJ7AhNg1dcF04C+Q5OEqd58Xux6pfAo7sTiX8Qn1vB27EBGpCSsAZwIHwpi0y/MnqrPL8yeg\nARYTGlUi7dJ2EZGYmZHjGdZgZ75NHRa7ogp2E/B+G89/HRjY5M9zCNM0JhMu7M8grK36TWCDTr73\nbODKDrxuO+ArzR5bQNj39+30PBA2gt2EMICiTydrESmVeYS1hqeFdXluATaKW1GHzQPWhcJcuNrd\nS7HvktSA+tgF1Cp3dzM7n6kM501gUOyKKpilt82Bni08t2Kzx94grLhhzV5XjJ7Atm08P5awmMmA\nZo/PBf6W3vdlSRibDDwLvA6cSgg/IuXWF/guMAqefQy2dLicMLyn0tv9fwXmhV9frohdi1QPhZ2I\n3P0Jy9nTDGdXNqNe3Z12HAis3IHXrQLsBqwDrA08BEws8j2XBw5v5blpwMtAD0IQa2oEIegMAo5m\nyb+0PHAvMB54oo1zi5TDFyC/FeRvhbOmwT8IXZ4NY9fVis+AyyDvcKu7fxi7HqkelR7is8/5GZ9Q\nz/jYhWTIZsBBhEEJq1J8V6c9r6b3g4BezZ6bnL7vEJb+laKeMDrU09eIxLYioctzAIw22AL4M5U5\nlufvwCdhg/dLY9ci1UVhJzJ3H4HxJMPJV+RPF2nda4RAM7iF55bvwPEdeY1IuewO+R/CotVC9tkH\neC9ySU3lgd9CPgf3ufubseuR6qKwUwmcC5im7k67XiRckhoKjCEMRI7lfcKg4z7Axi083zjO52lY\navHIPOESlwHbd2eBIkVYETgL2B9GWbg6ey2V0eW5AXgvbPj5m9i1SPXRmJ0K4O5PW84eZzh7MYh6\nRdBWPNXkaweGES4J7RWhllcJgWUrWv6VYTBhNtYIwmyuddPHJxEGNB8IbNP9ZYoUZQjkB0P+73Dm\ndPgncDPLjsMvlznAT6FgcGfi/lKkMqSKKexUCucCpjOKsbR8WaSWDSB0QdYjzCKZA4wDRgJPAr2B\nXcpYTz59f2j7v9XawFrAu0DTpvsGwPrdUplI6TR2eZ6Cp4fD5h52Uj+d7hsG15rfArPDwOSf64H6\n3AAAGRxJREFUlPmtJSPUQ6gQ7j4a4988Rp6G2NVUmH0IoWIVQjzvR1in5lhCh+dJKOs+Y28DC4HV\nCIGmJW8QprVMBY4Azklvh6eP3QR80O2VinTdHpD/ASzsB2cA+9H2slelNhH4PSQJXOLuk8r41pIh\nCjuVxDmHeRjPxi6kSmxMCBuLCJeHyuUVwq+2rV2GSoCH0/uvEILa8ultG+DLhHA2rNsrFSmNlYDv\nAfvCU+mMrb9CWXa7OTe8zXTgd2V4O8kohZ0K4u5vAdcwkgKfxq6mSvRL78v1/VrEkq2jt27lNTMJ\na+zUAZu28Pxm6XOTKW9HSqSr9oT82bCgH3wb2J/ubVA+BdwLlodz3H1+N76VZJzCTuW5iDwLeCJ2\nGVViYXrfo0zvN5YQUAbQ+urHjTX1pOXBDTmWrAS9qKTViXS/lfm8yzMynbF1PaXv8iTA96FQDy8B\nt5X49FJjFHYqjLvPwPk5L+B8FLuaCjefJb9WrlWm92ychdXWTKrGVZ4XsmRPrKZmps/1QGvtSPVq\n7PKsGgYtHwiUcknj24CXoS4PZ7t7Jcx+lyqmsFOZriHHOwylUJaL4pXsQ8LWCs1/1M0C7iLse7wZ\ny+6P1VmTgasJS7S2ZnZaTz1h0EJr+hJmjjnwIEt3bxalj0H4lVj/AqWarQx8H9gbnky7PDfQ9S7P\nfODcsIDgve7+VLsHiLRDU88rkLs3mNlZfMijvE5Yy6VWzQAeAFYgdG96E6aeTyGsV7MmcFizY+YR\nglDjJaTphJ++D7FkW4dNWHp9nob0dW2NoXk1Pc9mLLs9RHOHERYmmQj8kaXX2VlImFm2fzvnEKkW\ne0N+m7Auz6mzwro8N7Dkf/vO+h1h67kEzi1ViVLb9HtlhXL3YRgP8ih5PotdTUTrADsROjdTCOvb\nfEIIPgcBpxFWMW6qkL52cnr7jBB8pjd5rKXLS427q7emre0hmluDME93J2A5Quh5lxDahhD6/n07\ncB6RarEKcDawFwwnbBl3I53v8kwCLglTza9w92K38BVZirnX+nWSymVmG2K8wc704ouxqxER6aBZ\nhEvCs8LvJH+j412eE8Hvgll52NDd53ZbjVJT1NmpYO7+Ls75jKG868iIiHRFky7PfwljeW6i/S7P\nf4HbwlTz8xR0pJTU2alwZlZPjudYla04g3qNshKRqjIL7Bbw2XAwocuzTgsv+xTYAvJT4JkC7OP6\ncJISUmenwrl7noRvMZ0cz8SuRkSkk1YB/wGwZ+jcDCLspNI8yfwMmAyFApyioCOlprBTBdz9ZeBS\nRpAwLXY1IiJF2Bfy34f5K8M3gUMJ8wgARhEmLSbwU3efEKtEyS5dxqoSZrYcOV5nHdbnW9QppopI\n1Xoc6p8KExWvBC6B/LvwSh52cfdC7PIkexR2qoiZ7QU8yaGEKc0iItVqBtitYSyPQeIw2N1fj12W\nZJP6A1XE3UcA1zOMQovrxIiIVIt+4EcD4A7XKuhId1Jnp8qY2crp5aw1dTlLRKrWYuBa8szhVRJ2\ncfe21i8X6RJ9VFYZd59NwrF8SA7tGCMi1Wo4MJuEhBMUdKS7KexUoXRjvN/wJF7SbYZFRMphIvAs\n4Pyfu4+PXY5kny5jVSkz60GOUfRlW75DPb1jVyQi0gGfAn8mz0Kewtnf3ZPYJUn2qbNTpdy9gYRj\nmUsDD3d6rz0RkfJLgHtJWMQcnOMUdKRcFHaqmLtPwDmDVzBei12NiEg7ngbexUg4xt0/jl2O1A6F\nnep3K8Y/eFDT0UWkgr0HDMeBX7v745GrkRqjMTsZkE5HH0t/+nMyddosVEQqynzCOJ0FPIuzt1ZJ\nlnJTZycD0unoR/ARCUNjVyMi0kQC3E/CQubhHKOgIzEo7GSEuz+H821eBJ6PXY2ISGoU8A45Eo53\n9yntvl6kGyjsZIi73wRcy1CtvyMiFWAC8DgOXOLuj8QuR2qXxuxkjJn1JMdIlmMHzqCevrErEpGa\nNB34KwUaeBznEF2+kpjU2ckYd19MwhEsZBb/pIAWYReRclsI3EGePO/ifE1BR2JT2Mkgd/+IhMP5\nEGdY7GpEpKYUgLtJmMUCEg5x9zmxSxJR2Mkodx8FfI//AS/GrkZEasZjhL2vnCPd/e3Y5YiAwk7W\nXQdcx4M478QuRUQy7wXCBp/wfS0cKJVEYSfDPIw+Pwt4hLso8FHsikQks94HHsKBvwB/jlyNyFI0\nG6sGmNkK6QytrTmNelaOXZGIZMp04G/k+YxR6U7mDbFLEmlKnZ0a4O6fkvBFFvIRt5JnYeyKRCQz\n5gC3kGcxE9NxOgo6UnEUdmqEu39MwgHMZD53akq6iJTAAuDv5JnPJyTs5+4zYpck0hKFnRri7m/i\nHMKHFLgfJ4ldkYhUrcXA7RSYxTwS9nX3SbFLEmmNwk6NcfdROMfxOmGKqIZsiUhn5YF/kDCFxSQc\n6O5vxi5JpC0KOzXI3e8Dvsdo4InY1YhIVUmAB3AmkuAc5u7aelgqnsJOjXL3a4AfMxIYEbsaEakK\nDjwCjAWc47SWjlQLhZ0a5u5XAD/lCeCp2NWISEVzYDjwPwC+4+73RK1HpBMUdmqcu18M/ILHgVGx\nqxGRiuTAf2n8pejH7n5d1HpEOklhRwAuAi5mGDAmdikiUlEcGAY8A8AP046wSFXRCsoCgJkZ8Dvg\nR3wJ2DFyQSISX+MYnfBL0Pfc/eqo9YgUSWFHPpcGnquA73EosFPkgkQkHgcepnGMzpnufm3UekS6\nQGFHlpIGnj8AZ7MvsAdgcWsSkTJLgKFAmFT+bXf/a9R6RLpIYUeWkQaenwEXsStwIBrdJVIrEuA/\nwIs4cKq73xi5IpEuU9iRVpnZmcDVbAN8GaMudkUi0q0agPtw3gDgZHe/OWo9IiWisCNtMrPjgFvZ\nDONocvSIXZGIdIsFwJ0UmEQe52vu/u/YJYmUisKOtMvMDsG4j/Wp5zjq6B27IhEpqdnAreSZxack\nHOLuo2OXJFJKCjvSIWY2BONh1mA5TqSOFWJXJCIlMZUQdBYylYT9tamnZJHCjnSYmW1Djv/Sh5U5\ngXr6x65IRLpkIuHSVYHXSTjI3afGLkmkOyjsSKeY2brkeIgcW3E0OQbFrkhEivIacD+O8zjOke4+\nL3ZJIt1FYUc6zcz6YNyKcwT7A7ujtXhEqkUCjEhvcCthevnimCWJdDetniKd5u7zcY4GfsN/gQdw\n8rGrEpF2LQLuImEEDpwPfENBR2qBOjvSJWZ2AsZNrEOO46ijT+yKRKRF04E7yDOLRTjHuPvQ2CWJ\nlIvCjnSZme1GjgdZgZU4XgOXRSrOm8A9FCgwkYRD3f3t2CWJlJPCjpSEmQ0gx38wtuAwcmwbuyIR\nIQGeAp4AjH/hnOTucyNXJVJ2CjtSMma2PHAN8E22B74IWnFZJJLPCLOtxmPAhcCv3T2JXJVIFAo7\nUnJmdjLGtaxBHV+jjn6xKxKpMZOBu8kzh8U4x2nrB6l1CjvSLcxsW3LcTx3r8WXq2Dp2RSI1IAFG\nAY/jGC+T8DV3fyd2WSKxKexItzGzFTGuwzmW7XEOxugZuyqRjJoH3EfCu+SAy4ALNK1cJFDYkW5l\nZgacjHENq1LHUdSzduyqRDLmTeB+CixmFgnHuft/Y5ckUkkUdqQszGxLctyFsyV7YewB1MWuSqTK\nNQCPAf8DjIdxvuHu0yJXJVJxFHakbMysJ/Az4HzWxDmSOtaMXZVIlZoK3Eue6TjOj4CrXT/QRVqk\nsCNlZ2Y7kuN2YCD7kOMLqMsj0lF5wto5I3GM8ellq1dilyVSySp2bywzS5rcdmnjdV9r8rqJJXjf\nX6TnOqkTxwxIjxne7PG90sdv7GpdHazjpPT9vtjkMTOz/c3sGjN7zsw+MbNFZvaOmV1rZhu0cq4/\nmNkCM1u31HW6+/MkbEPC5TyOcwMF1HgXad8U4DryjKCA8ysStlXQEWlfxYadlKe3E9p4zQlNXlfK\n9yzXcSVhZr2AXwFj3P3hJk9tBAwDzgDWBp4B/gP0BL4NvGJmX2jhlJem97/ujnrdfZG7nwfszlQ+\n4C8kjAIK3fFuIlVuMeFf8fU4MxgH7OjuF2q2lUjHVHrYKQCvAceY2TK1mtmqwMHAi4CVubamJgOb\nA9+IWMOZwLrAxc0ed8KPyX3dfR13P8LdjwY2Bm4C+gK3m9lSF5LcfSpwC/B1MxvUXUW7+2gStqLA\nnxiG8xfyfNBd7yZShSYA15BnNA0455Owo7u/HLsskWpS6WEH4HZgdeCgFp47FqgHbitrRc24e97d\n33L3SRHL+A4wE3io6YPuPtHdD3b3Ec0ebwC+C8wB1gda6u7cRvh/5IxuqXhJLQvc/QfALszgNW4E\nHsD5tDvfVaTCzSNs93ArMJdncLZ094vTf7si0gnVEHbuIHQnvt7Cc18HPgX+1dYJzOwQM3vMzGaa\n2UIzG29mF5vZSu0ct7OZPWpms8xsjpkNa2n8UGtjdtpjZgeb2UNNxtFMMLMr0o5VZ86zJzAQuMfd\nO3whyN0XAW+lf1xm9Rt3fwb4gNDd6fblAN39ORJ2As7gFebxRwo8R1gVVqRW5AkXm/9IgVeZC5yC\ns492KhcpXsWHHXefDIwEvpxuNAmAmW0I7ArcByxs7Xgz+wlhjMqewPPA/cBywHnAs2a2eiuH7p6+\n79rAUGA8sB8wwsz27+JfCzO7JD3vvum5/0VYNeOHwJg26mrJoYRAOKK9FzarwYAB6R+ntvKyEcAq\ntNz5KTl3L7j7dTgDWczNPAT8lQIxe2Yi5eCExQGvJs9jJDTwZ5yN3P1GTSkX6ZqKDzup24DlgSOb\nPPZ1wo+HVi9hmdmOhEG784Dd3f1Adz+e0AW5G9iMsEv3MocCpwKXuvvW7n6Cu+9CGBfTE7g5HRBc\nFDP7KnAu8Cqwhbvv6e7HuPsg4CLCeJqrOnHKPdP75zpZyvHAGsA0wo46LQnLlcFenTx3l7j7NHc/\nFfgCnzCOvwH/JvyXFMmaT4BbSbgTmMMIYLC7f9/dZ0auTCQTqiXs3EOYj9B0VtbxwEdAW5eOvkf4\noP6juz/f+GB6zfssQkfoCDNbp4Vj3wd+2fQBd78OGAOsBRzV+b/G584nBLXj3P3dZu/xC+Bl4OhO\nXM4aDCTuPqGjBZjZesAf0jouaGMcwPj0ftuOnruU0gHM2wHf4yXmchUJw4FFMaoRKbEFhP7utTjv\n8QHwFZwD3P31yJWJZEpVhB13n0MYeLufma1hZjsRujJ3ttPeHZLe39HCOacRZinlCJesmrvX3Vsa\nLXInIUDt0Ym/wufSy1ODgbfd/Y1WXvYMYZm9HTpwvj6Ey3KzO1HD8oTLf/2A+939+jZe3vibZWcu\nq5VUemnrapwNyPM7nmIxV1JgNGF8g0i1WUzopYZxaQtwziNhkLv/W5esREqvPnYBnXAbcARhBtZG\nhI7E7e0c0zjo9r1Wnm98vHlnxwmdnbaOKXY7yw3S+03MrK2htw6s1oHzNQ6y7tAFHjOrJ3TKdiCM\nSWprDSOAuen9yh05f3dy91nA/5nZn1jEhTzKqYyiwH7UM5gqie5S0xqAF4CR5FmAATcAF6ZLPYhI\nN6mmsDOUME36G4Sg8UYJ1pqI8RtU40fyVODRdl7bWuBqak5637e9F6YDkv/OkrWJvuzun7VzWGOY\n6nDnqLulg9ZPN7Pf8ym/5QGO4GnyHEA9mxJ3xSWRluSBl4AR5PmUOsIvahe5e5dXfReR9lVN2HH3\nxWZ2N3AaIaRc2YHDphA6KQNYMvakqQ3S+8nNHm86S6m5xsendOD9W9I4r2i6u59c5Dk+5+7zzWwh\nHeu8XE3ojI0HDnb3ue28HsJMLKDyNnRw9/HAkWa2CzO4nDsZwloU2JM6NkOdHomvALwCPEmeudQR\nuqq/cPc34xYmUluq7ePgVmA64YN3mXE4LXgqvT+u+RNmthphoUInjJFp7si0E9LccekxT3ek4ObS\nrsR4YAszG1jMOVrwCpBr63xm9mvCwoPvAQe4+/QOnnvz9L5iV2x19zE4ewIHMZXR/IOw4uzLaPsJ\niSNP+BfzJ/LpLMJ/E2ZYHaegI1J+VRV23P1pd1/D3fu7+4cdOOQaQjD5vpl9PtjXzHoQuhzLEQYi\nN+/sQOj6/KLpA2Z2OmFtn6nAvUX9JYJfEQYg32dm2zR/0sxWNbNTO3G+xlC3U0tPmtkPgZ8SZq8d\n0MrftzU7U8QaPuXmwTBPfA/gC8zkUR4AriTPGMKAUJHutpDwr/EP5HkAmMOjwPae+FHuPjZucSK1\nq2ouYxXD3Z8zswsIm1mONrMnCZ2h3YH1CEt4ndXK4dcD55nZkYT1cAYSwsRi4Jvp6sPF1nWnmW1B\nCCAvmNnLhB1wjLDGzmDCgOO/dfCUDwHnAHsTZot9Lg1TlxMCy3vAz1puWHG9u7e01s5ehPE6ra3D\nU3HcfTTwJTPbmnn8Hw9zLE+SsBv17ESIuCKlNBN4FniRAgUc5xbg9574uMiViQiVH3Y6M4C4xV3H\n3f3iNEz8ENiR8FH3AXAJYdHAOc2PSc8zirBR5q8IKxQb8Bjwc3cf09H3b6OuC8zsUULY2h3YijDz\naTKhI3VPW3/ZZucaaWZvAUeZ2XfdvemE7KZjeXZNby15gmaBxsyGEELhVdW4u7K7vwacYGYXsJAf\n8wSnMpIc21DHTkD/2BVK1fsQGIXzBpBjDgl/Aq5x948jVyYiTZiWdMgGM/s+YZHAo939/hKd8zrg\nFGDrNtYEqhpm1h/4NjnOJGEN1qXAztSxBZUf+6VyNADjgP9RYDJ15JhIwu+Av7v7gsjViUgLFHYy\nIt2+4k3g43Rri66eby3CpbV/uPu3unq+SpKO2foyxndx9qE3eXagnh1ZMvdMpLmPCAs2vEKBxdSR\nYyQJVwD/aWUBUhGpEAo7GWJmJwI3A4e5+9AunusPhGn+g9w9s9twmtlmwBkYp+D0ZSAJO5BjE9Tt\nkTDgeCzwPAU+po4c00m4HrjR3d+JXJ2IdJDCjgifb6FxLDnOImE7elJgS+oYTFhZqarmLUqXJIRR\nfS8Cr5NQAIyHcP4GDG02Jk5EqoDCjkgzZrY5cAI5TiJhPfqQZ3C6JUV/tEJzFiWEwcbjgLHkmU89\nOd4n4TrgFncvdhFREakACjsirUgXldwVOJ4cJ5CwCv3S4LMFYecyBZ/q1RhwXgde/zzgfELCXYTZ\nkM9oLI5INijsiHRAOqh5P+AEjKNxerMyeQal+3Gtj8b4VIMCS3dwFiwVcP4JjFbAEckehR2RTkrH\n9+wDHEaOw0lYkx4U2IQcm2JsAvSJXKQsMYswr/AdnAkkNFBHjo/TgHM3CjgimaewI9IF6aWubYAv\npcEnbEuyDgU2oY4NgHVR16ecPiOsFT4BeIsGZtMDSMgxhoSHgUeB5xVwRGqHwo5ICZnZmsAhwGEY\n++P0pY6E9YANyTEAWBvoGbXMbFkATEpv71PgQ3IkGDkmkfAQIdwMb2W1dBGpAQo7It3EzOoI+5zt\ng7EPsDfOCuRw+lNgAPWsR5jhtQoa7NwRCTCNMO4mhJsGZtEDgByzSHiasK3Lo8Dbrh9wIoLCjkjZ\npOFnS2AIMIQc+5CkO3T1oMBawFrUsRYhAK0O1MWqtgLkgRnAx+ltCgmTcBqoI1yWGkfCSGB0epuo\ncCMiLVHYEYko3a9rO2BbYDvq2JkCAwDI4axOgbWopx+h+7Nqep+lndsbCLuGTwOmp/cf08BM6knS\nfleOj0h4gSXB5jl3/zRSxSJSZRR2RCqMmfUlDHoOASjH9sDGJPT9/EW9yLMq0I/6zwPQioRZYH2A\n5amMrpATxtTMBuY0u80mz2xgYZPh2zlmAeNIGAu8CrwGvObus8tcuYhkiMKOSJUws1WBjZe65dgU\n2ISENZY5oBd5+qSjhFagjj6EjlA90KPJfY9mjzVGj6Sd22JgEWH/qKb3C8izAGcRsIg6kiabbRiL\nyTGZhIk47xE2ZviAMHdqnLvPKMX3SkSkKYUdkQwws+WANYE1mt3CY0Z/cqyFsyrQG6cXTi+6Oiza\nWIwxF2MWzgwSphNWtmm8zaRxKHEINdM1rkZEyk1hR6RGpWsE9ST0e5reehN6N/l2bovcfWH5KxcR\n6RyFHREREcm0XPsvEREREaleCjsiIiKSaQo7IiIikmkKOyIiIpJpCjsiIiKSaQo7IiIikmkKOyIi\nIpJpCjsiIiKSaQo7IiIikmkKOyIiIpJpCjsiIiKSaQo7IiIikmkKOyIiIpJpCjsiIiKSaQo7IiIi\nkmkKOyIiIpJpCjsiIiKSaQo7IiIikmkKOyIiIpJpCjsiIiKSaQo7IiIikmkKOyIiIpJpCjsiIiKS\naQo7IiIikmkKOyIiIpJpCjsiIiKSaQo7IiIikmkKOyIiIpJpCjsiIiKSaQo7IiIikmkKOyIiIpJp\nCjsiIiKSaQo7IiIikmkKOyIiIpJpCjsiIiKSaQo7IiIikmkKOyIiIpJpCjsiIiKSaQo7IiIikmkK\nOyIiIpJpCjsiIiKSaQo7IiIikmkKOyIiIpJpCjsiIiKSaQo7IiIikmkKOyIiIpJpCjsiIiKSaQo7\nIiIikmkKOyIiIpJpCjsiIiKSaQo7IiIikmkKOyIiIpJpCjsiIiKSaQo7IiIikmn/H701xN0N3keE\nAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe6ec26bb38>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "page_views_by_platform_pandas.plot.pie(y='count', figsize=(6, 6), labels=['Desktop (1)', 'Mobile (2)', 'Tablet (3)'], \n",
    "              title='Page Views by platform', autopct='%.2f', fontsize=15)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Page views by Traffic Source"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 40 ms, sys: 124 ms, total: 164 ms\n",
      "Wall time: 5min 26s\n"
     ]
    }
   ],
   "source": [
    "%time page_views_by_traffic_source_pandas = page_views_df.groupBy('traffic_source').count().toPandas()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>traffic_source</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1293550863</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>343133832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>397590753</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     count\n",
       "traffic_source            \n",
       "1               1293550863\n",
       "2                343133832\n",
       "3                397590753"
      ]
     },
     "execution_count": 185,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "page_views_by_traffic_source_pandas = page_views_by_traffic_source_pandas.set_index('traffic_source').sort_index()\n",
    "page_views_by_traffic_source_pandas"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Distribution of page views by traffic source**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fe6ec237b38>"
      ]
     },
     "execution_count": 186,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAhoAAAH5CAYAAADHgZeXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAAPYQAAD2EBqD+naQAAIABJREFUeJzs3Xd4VNXWwOHfmkwghE6oKr1XkaACShERroqgKCDI1aAg\nyLUjV1AQuKLYQexeUUAsiCh+6EUUFRtFE4qoCFIEsQBBiggIk9nfH/tMmEwmpM3kpKz3ec4zySn7\nrJlM5qzZ7YgxBqWUUkqpaPC4HYBSSimlii9NNJRSSikVNZpoKKWUUipqNNFQSimlVNRooqGUUkqp\nqNFEQymllFJRo4mGUkoppaJGEw2llFJKRY0mGkoppZSKGk00lMoBEVkmIp+4HUd2ROQaEfGLSDu3\nY4kkEekoIitF5JCIpIlIE2f9JSKyTkSOOutLicjrIrLB7ZiVUpYmGipfgi5sgeWIiGwUkSdEpLrb\n8YUjIrc5sXY/yT7DnX16O6sM4C+YCPMtavcVEJGuIX/vrJa0CJ6zFPAmEA/cAvwT+FVEagCvA/uB\nkcAQ4DgR/luJyGUi8pmI7HYSnc0i8pqInB+pcyhVnHndDkAVCwaYAPwExAHnAjcAF4pIK2PMURdj\nC+d14GFgMPBxFvsMBlKBxc7vFxRAXEXBBuwFPdgDwJ/AFECicM7mQC3gDmPMa4GVzoU+DhhrjFkR\ntH5IpOIQkfHAf4Cl2Od3FGgE9AQuBz6KxHmUKs400VCR8r4xZrXz84si8gdwG9AXmOdeWJkZY35z\nmkH6icgNxpjjwdtF5BSgM/CsMSbNOcbnQqiFjjFmN/Bq8DoRGQfsCU4CTkZE4o0xh3Nx2hrYZPZA\nmPWErg/8zfJLREoDY4F3jDGXhWz+t4hUjcR5chlTGWPMkYI+r1L5oU0nKlo+xn6rrA8gIpVF5BER\n+UZE/hSRAyLyPxFpE3qgiNQRkf9zqql3ichjItLTqZLvErLv2SLyvojsF5G/nL4UnXIQ31ygInBx\nmG2DnNjTL6hOuRlqP5z+AJNF5Eenj8AOEXnQqeoP7LNARFJCjlsU0iyDiJzlrOvl/O4VkYkisslp\njkoVkc9zUV1fVkSec447ICKzRaRS0PlmicgeEYkJPVBEPohUHwcRecB5Xo1E5A0R2Qd86Gw7Q0Tm\niMhW5zn+6sRcMej414D3nV/fdcpaLCIrgGed9d866592jsnUR0NEPCJyh4isd861S0TeC/f+C1IT\n21yzPNxGY0xqyDlqOK/rbucca0RkUMg+vZxYzwpZ39RZPyBo3evO36iJiCwRkT+BmUHbz3HW73P+\nV9aIyMiQcluJyNsisldEDovIKhH5x0mes1IRpzUaKloaOY97nccGQB9gPrAN+210BLBMRFoYY34H\n+20X+MTZPh3YhW3GOI+Qvgdi+1j8D0gGJmHb5YcCH4vIucaY5JPE9xbwjFP2wpBtg4DtxpjgC0zo\nuQVYBHQCngN+AFpja3EaA/2cXT8H+ohIOWPMIWddJyANW2vyrrOui7PuS+f3ydhv088DXwMVgPZA\nO7KvrhfgSWAfMBFoCowC6mBfR4CXsX0demFfw8DzquHsMzGbc+RU4HVbCHwP3Bm07kLgFOAF7N+5\nNfY90RTo5uzzBLAdGAM8AqwFfgVigSuBJKfMX4BNQecM7afyCjAQeAf79yoFdAXOBL7JIvZfAR/Q\nV0SeN8aE1qikE5GywBfAqcAMYKdzvlecv/1/w7wm2TFAaeADZ5mPbaLCSVLfwr42j2Ffv5bYxPlZ\nZ5+2wGfAVuB+4Aj2vf2uiPQ2xryPUgXBGKOLLnlegGuwF8jzgATsB+1AYA9wCKjl7Bcb5tg62A+/\nu4PW3e6U1ztoXSnsRSoN6BK0fiPwXkiZpYEt2Kac7GKfB/wFlAta1wSbsNwbsu8nwMdBvwc6HnYM\n2e96J84Ozu+JTnm9nN9bOb+/DiwPOm4hkBz0+xrg//L49/ADq4CYoPV3BL+u2GRkB/BqyPG3YS+u\ndXNxzvXBr03ItqlOPC+E2Vb6JO+nxKB1vZwyLgrZd4Szb4uQ9a8B3wf9fqFz/P15eD0fcM5xEJtY\njgXahNnvTme/S4PWebFJ8F4gLui5pAFnhRzf1IlxQMjzSAPGh+zrxSYyG4CyJ4n9C+d94AlaJ9jE\ndW1uXwtddMnrok0nKhIE+y17D/AztsnhIPZD9zcAE9QPwqnGrgIcxiYLwUMxewG/GGMC3/QxxhwD\ngr8RBr6tNQZeE5GEwAKUd2LJ0MSShblAGU7UPgBchf0m+WrYI064AvtBvynk/J84r0eg5mANNuEK\nxNMZ+xrNARJFJM5Zfy629iNgP9BSRBqRN8+bjH0VnsFetC4CMMYY7Lf8Ps638YDB2ARoex7PG47h\nRDPHiZXG/B34WUTinNdvFfb1i+Tw3MuBY8B9uT3QGDMWm/ysA/7hlLHWaYII/ttciK0FWxh0rA9b\nI1MJW4uVV6Gv3dnYmqDHjDF/hTtARGo655wHVA56fyZga0dai0jlfMSkVI5poqEiwWBHmfTAVnm3\nMMY0NMYsDewg1m0isgn4GzuiYze2urxiUFl1sTUSoTaH/N7YeZyDTXACy25gGFAquK0/C4uBP7AX\n14ArgXXGmOz6KDTGVlXvCVk2Yl+P6gDGGD+wAptg4Dx+jm0iiQE6iEgLoAoZE417sBeoTWL7tTwk\nIq2ziSnAEPJ6ORek34B6QavnYPsgXAa2nwC2BmZODs+TG9tCV4hIVRF5SkR2YZPOPdiaK0PG90R+\nNQB2ZHVRzo4xZq4xpjP2b3Qh9uJ9JrAwqI9LXU403QTbgE2c6ubl3MBhE9IXBGiIfY2+O8lxgf+P\nh8n8/zHO2VYoh5+r4kf7aKhI+dqcGHUSzt3YYYIvAOOxF3g/8Dh5S3gDx4zGftsM51AW6wH7jVNE\n3gCGiUg17EW4MbaZISfnX49tagg3lPLnoJ+/AO4SO4qhM7ZZ5oCIfOv8vht74UhPNIwxn4tIQ+yo\nnZ7AdcBtIjLCGPNiDuLLljFmg9iOqkOwtTtDsEng/EiUHyLcSImF2ETzIexr+Rd2uOoiCuGXIGPM\nnzj9JWwXHQZga16+zk0xWazP1CnXkdcRJoHX735sLVs4O/JYtlK5oomGKiiXY9vxrw9e6YyE2BO0\najt23oRQjUN+D9R6/GmMyWoujJx4BTvZ00DsN99A/4nsbMG21edkttDPsf1MBmGrvAMJxWfYJpVd\nwCZjTPDrgDFmPzAbmO10kv0c2+k1u0RDsK/Xp+krbPNILeC9kH3nAI86Ve2DsH1esuz0GClOp9NO\nwBhjzKNB61tF4XRbgE4hHXLzKxmbaNRyft+O7d8Tqjk2uQg0Re3D/n0qhexXLxfn3uKU0YosRsRw\n4v/j73z+fyiVb4XuW4MqttII+eYvIv2xnUeDLQFOFZFLgvaLwzaHBEvBfpjeEdLHIHBMjuY4MMZ8\niZ1o7J/YZONTY8yvOTj0DeA0ERke5txxTmIQsArbwfJO4I+gZpnPgQ7YZOPzkDKqhMR5GNscUjoH\nsQFcLyLBXyRGYb81/y9kv8DcF49jhyK/nMPy8yvQfyT0M+g2wn/rz89spwuwid7duTlIRMqLSPss\nNl/oPG50Hv8H1BWRvkHHe4Ebsf1tAqOJtmGfS2gfohvI+XNchR1lM1pEyofbwRizE1gJ/Cvc/0JO\n/z+UigSt0VCRkJNZGN8FJojIi9hvYa2xHS9D+2M8h/1wfl1EHsf2K7iKE1XIBmxnRhEZhv2A/05E\nXsJ++J6K7Yh5ANvskBOvAnc5ZY/P4TEvY7/RPiMi53Giz0VzoD+2uWO1E+sRp4miA/B/QWV8BpTF\n9pPIkGgA34vIMmxC9Qe2T8AV2KGTOVEK+MhpGmqGvZB9HtzJ1oktVUTed2LeR+ZEJCqc834FjHcS\nxV3Yi/dphH8/5XmmT2PM+yIyHzvJVgvsPB5e7PDWd40xM7M4tALwlYh8iW0y+RnbT6Mf0BF4zRgT\nSDSewibDr4rIk86+V2KbVkYGOr46z/sdYIyTiOzAvk9z3DHTafIbhU2g1ojIbOzr1xxoYIwJvO9H\nYmu1vhWRF7BJTi3gHOd8HXJ6TqXyxe1hL7oU7YUTwxHbZbNfKWxb/E5s34lPgbOwE3t9FLJvXewF\n+RD2A/QR7Id7GnBmyL5tsH0KdmM7FG7Ffkvvlovn0Nwp+y+gQhb7fBImzhhsf45vnHOnAl9hvzmX\nC9n3Qecco0PWb8LWdtQLWT8O24l0r/M6fIetEYnJ5rkE/h7nYkeapGKTrtlApSyOuQLbZPR0Ht8D\n60Nfm6BtU53nFx9m22nA29hEai82eTvNiX9M0H6BIaG5Gd76Xcg6D/BvbOfMI8Dv2Dk1Wp3kecUC\nw7F9SbY6xx10/sY3EzRs1Nm/BvCS8148gk00rwxTbnXsHBiHsM2G04HTnecSOrx110ni64xNmg44\nSwowLGSfhtjmsd+w06dvd17z3lmVq4sukV7EmKjdf0mpiBGRW4FHgdOMM2RWRYaI9MFefDqbjJOU\nKaVUvmmioQodEYkzQTdic/porMG+X5u5F1nxJCLvAk2NMaEdbpVSKt+0j4YqjN4SkR3Y6aYrYYdd\nNiHjfBcqn0TkSmzT04XYpgCllIo4rdFQhY6I3IztWFcP2w/ie+BBY8ybbsZV3IiIH3vvjNeBG4yd\nXEwppSJKEw2llFJKRY3Oo6GUUkqpqNFEQymllFJRo4mGUkoppaJGEw2llFJKRY0mGkoppZSKGk00\nlFJKKRU1mmgopZRSKmo00VBKKaVU1OgU5EoplQciUgeo6nYcqlhLNcbscDuI/NJEQymlcklE6ng8\nno1+vz/O7VhU8eXxeI6KSNOinmxooqGUUrlX1e/3x82dO5fmzZu7HYsqhjZs2MCQIUPisLVmmmgo\npVRJ1Lx5c9q1a+d2GEoVatoZVCmllFJRo4mGUkoppaJGEw2llFJKRY0mGkoppZSKGk00lFJKKRU1\nOupEKaUiaMeOHaSmprodBlWrVqVOnTpuh1HodevWDY/Hw8cff5ztvn/99RcNGjRg+vTpDBo0KFfn\n8fl8NGjQgLvuuouRI0fmNdwiSRMNpZSKkB07dtC0aXOOHj3sdijExcWzceOGXCcbs2fPZujQoSQn\nJ+d66O6RI0d46KGHOO+88+jSpUuujnWLiOR43+nTp1OhQgWuvPLK9HW///4706dP56uvviI5OZlD\nhw6xbNmyTM/f6/Vy++23M2XKFK699lpKlSoVsedQ2GmioZRSEZKamuokGXMBNyfy2sDRo0NITU3N\nU61Gbi6+wQ4fPszkyZMRkSKTaOSUz+djxowZjB49OsPrs3HjRh5++GEaN25MmzZtWLFiRZZlDB06\nlLFjx/Lqq6+SlJRUAFEXDppoKKVUxDUHSt5EXsaYqJR7+PBh4uPjo1J2Ti1atIjU1FT69++fYX37\n9u3Zu3cvlSpVYsGCBSdNNCpWrEjPnj2ZNWtWiUo0tDOoUkqpk0pKSqJ8+fL8+uuvXHrppZQvX57q\n1aszZsyY9ORi+/btVK9eHRFh0qRJeDwePB4P//nPf9LL2bhxI1dccQUJCQmUKVOGM888k0WLFmU4\n1+zZs/F4PHz22WeMGjWKGjVqULt2bYD0crds2UJSUhKVK1emUqVKXHvttRw9ejRDOS+99BLnn38+\nNWrUIC4ujpYtW/Lss8/m+TV45513qFevHvXr18+wvmzZslSqVCnH5VxwwQV88cUX7N+/P8+xFDVa\no6GUUuqkRAS/30+vXr3o0KEDjz76KEuXLuWxxx6jUaNGjBgxgmrVqvHss88ycuRI+vXrR79+/QBo\n06YNAN999x3nnnsup512GuPGjaNs2bK88cYbXHrppbz11lv07ds3wzlHjRpF9erVmThxIn/99Vd6\nHAADBgygQYMGPPDAA6xevZoXXniBGjVqMHXq1PTjn332WVq1akXfvn3xer0sWrSIUaNGYYzhhhtu\nyPVrsHz58ohMN5+YmIjf72f58uVcdNFF+S6vKNBEQymlVLaOHj3KoEGDuOuuuwC4/vrrSUxMZObM\nmYwYMYL4+Hguv/xyRo4cSZs2bRg8eHCG42+55Rbq1avH119/jddrLz033HAD5557LnfeeWemRKNq\n1ap89NFHYfuLJCYm8vzzz6f/npqaysyZMzMkGp999hmlS5dO/33UqFFceOGFPPbYY7lONNLS0tiy\nZQuXXnppro4Lp0GDBgB8//33JSbR0KYTpZRSOTJixIgMv3fu3JmtW7dme9y+ffv45JNP6N+/PwcO\nHGDv3r3pS8+ePfnxxx/57bff0vcXEYYPHx42yRCRsHHs3buXQ4cOpa8LTjIOHjzI3r176dKlC1u3\nbuXPP//M8XMG+OOPPzDGULly5VwdF06gjMIwBLqgaI2GUkqpbMXFxZGQkJBhXeXKldm3b1+2x27e\nvBljDBMmTGD8+PGZtosIu3fvplatWunr6tWrl2V5oSNpAhfvffv2Ua5cOQC+/PJLJk6cyMqVKzl8\n+MRwYxHhwIEDlC9fPtu4Q0Wis2ugjLyO7CmKNNFQSimVrZiYmDwf6/f7Abjjjjvo1atX2H0aNWqU\n4fcyZcrkOpbARXzr1q306NGD5s2bM23aNGrXrk2pUqV47733mD59eno8OVWlShVEJEdJVXYCZVSt\nWjXfZRUVmmgopZSKiKy+pQf6JcTGxtK9e/eox7Fo0SKOHTvGokWLOPXUU9PXf/TRR3kqLyYmhoYN\nG7Jt27Z8xxYoo3lzN+dZKVjaR0MppVREBOa6CB26Wa1aNbp168Zzzz3H77//num4SPdXCNR4BNdc\nHDhwgFmzZuW5zI4dO5KcnJzf0EhOTsbj8dCxY8d8l1VUaI2GUkpF3IYiff689kWIi4ujRYsWzJs3\nj8aNG1OlShVatWpFy5Yteeqpp+jcuTOtW7dm+PDhNGjQgF27drFixQp++eUX1qxZk+/zB/Ts2ZPY\n2Fh69+7NiBEj+PPPP9OHwIZLdHKib9++zJ07l82bN2dq5pkyZQoiwnfffYcxhjlz5vD5558DcPfd\nd2fYd+nSpZxzzjkR6VhaVGiioZRSEVK1alXi4uI5enSI26EQFxef534AWY32yMm+M2fO5KabbuL2\n22/n2LFjTJw4kZYtW9K8eXOSk5OZPHkys2fPZu/evVSvXp0zzjiDe+65J0fnyqkmTZqwYMECxo8f\nz5gxY6hZsyajRo0iISGB6667LsfPLdgll1xCQkICb7zxRvoQ34B77rknvQwR4aWXXkr/OTjROHjw\nIB988EG+Jg4riiRaU8YqpVRxJSLtgJSUlJRMkzjp3VuLrylTpvDSSy+xefPmPCVD06dP55FHHmHL\nli0Zht+Gs3r1ahITEwESjTGr8xZx4aA1GkopFUF16tTRC3wxddttt/Hkk0/y+uuv5+k28dOnT2fC\nhAnZJhnFjSYaSimlVA6ULVs2z308vF4vP/30U2QDKiJ01IlSSimlokYTDaWUUkpFjSYaSimllIoa\nTTSUUkopFTUlPtEQEb+IpLkdR2EnItc4r9U92e+d6dirnWMvDFl/mojcICIvicj3IpLm7NflJGVN\nE5HDInJaXp6HUkqpglXiE41Ici6S2d8zuejK9aQrIlIauBdYZYxZHLL5cuAp4GqgaQ7P8aDzOCW3\nsSillCp4mmio3MjLdH2jgNOAqWG2bQWmAYOBJsCH2RVmjPkdmA0MEZFmeYhHKaVUAdJEQ0XbDcAf\nwHuhG4wxi4wxo40x84wxW8h5jclc7Ht3ZOTCVEopFQ2aaIQhInWdZpCPRSRORB4QkZ9E5KiI/Cgi\n/w7Z/xoR8WMvlPWcYwPLxyH7lhGRcSKyWkT+dJYVInJ1FrH4RWSriMSKyD0issGJ4y1ne1Kg74SI\n1BaRV0Vkt9OP4WsR6Z1FuYNE5DUR2Sgih0TkoIiscvpM5O9GAyfO0QVoBLxpjIlYPxhjzJfADmyt\nRqlIlauUUiryNNE4uVLAB8B1wNfAx8ApwAMi8p+g/TYDs7BNC4ecnwPL+4GdRKQasBK4D6gBLAM+\nxfZPmCUij2cRhwdYCNzhnGsh8JuzzThLfSfG9sBSYDXQDnhbRHoEF+b0m3gFON8p5/+AFUALbJ+J\nmdm+MjlzsRPbpxEqL9inQGWgUxTKVkoVQUlJSZQvXz5fZRhjaN26NVOnhmvtzV7Hjh2588478xVD\ncaNTkJ9cR2wyUM8Y8xek30xpFXCbiDxgjDnsfMP+UkSSgFRjzLVZlDcLaIXtlzDWGHPcKbMatmnh\nRhF5zxjzQchxtYG/gSZOH4VQgu1Q+YgxJv0dLiK3OOcaj00+AnzApcB7wTUNIpIALAauEZEXjTFf\nnPTVyV5g9MjX+SwnnK+AfwJdsX8jpQqF4nBTtfXr1zN58mSSk5PZtWsXCQkJtGjRgj59+nDjjTdG\nONLIEZF83/n11VdfZefOnRmeZ3JyMrNmzWLZsmX89NNPJCQk0KFDB6ZMmULjxo0zHH/nnXcyZMgQ\nRo8eTfXq1fMVS3GhicbJ+YERgSQDwBizWkQWY7+ttwc+y0lBInI6cCF29MXo4G3GmD0icj22FuIG\nbC1KqLFZJBkB24C7Q9Y9CdwDdBARrzHG55wvDVuLkYExZq+IjMN2yuwL5DfRaAP4nf4XkfaD89g2\nCmUrlSc7duygabOmHD1y1O1QiCsTx8YfNuY62Vi+fDndu3enbt26XH/99dSsWZOff/6ZlStXMmPG\njEKdaETCI488wqBBgzLUjDz44IMsX76c/v3706ZNG37//XeeeOIJ2rVrx6pVq2jRokX6vn379qVC\nhQo8/fTTTJo0yYVnUPhoonFy240xm8Os34RNNGrloqye2GaEd8JtNMasFZFDwFnhNgPvZlP+skAi\nEVRmmohsA84AEoBdwdud5KcnUBeIx9aMBP67MqbpuSQiZYEywN78lHMSfziP1aJUfrHn9MUpjf2b\nVwDKYt9rfiDNWfxZPKZha9n+MsbkethzcZWammqTjH5AVTcDgaNvHSU1NTXXicZ9991HpUqVSE5O\nztQM4UZNzeHDh4mPjy+Qc61Zs4Z169Yxbdq0DOtHjx7Na6+9htd74pI5YMAAWrduzQMPPMCcOXPS\n14sIV1xxBXPmzNFEw6F9NE5uZxbr/3Qec3Ov33rYC/n9IZ1F0xegHOE/nnYHmlkiEavTsfRVYA3w\nALYW5Rps88tl2ItN/ho6oWLI+SPtoPNYKUrlFyliJYjIGSLSR0RuFJEHReQFEZkn4nlfpNRKkdIb\nRErvFIk9AHIMOALsxvb9WQd8A3wLbMAm1JuxtWXbse+xX7EJayrwJ4hPJHa/SOntIqXXiXg/FpE3\nROQpEZkoIqNEpL+IdBWR+iJSMr7cVMX25nJryUeSs3XrVlq2bBm2r0PVqpkLnjt3Lu3btyc+Pp6E\nhAQGDRrEzp0ZP46++OILBgwYQN26dYmLi6NOnTrcfvvtHD2aseYn0Mdi69atXHTRRVSoUIEhQ4ak\nb1+1ahUXXXQRVapUoVy5cpx++unMmDEjU0y//vorl156KeXLl6d69eqMGTOGnOTDCxcupHTp0nTu\n3DnD+g4dOmRIMgAaNWpEy5Yt2bBhQ6ZyLrjgArZv3866deuyPWdJUDL+6fPOH8GyAknd58DJmhLC\n/TfkpB42N7GOBq7EXlj+jU049jk1II2BjeRtzoxgB5zH/CYsWQkkMvujVH6h4tQ+1MJ22K2D7bdT\nBzz1wFsPPKeCPyjxjTVQ0wfVgYoxUMlj/xSByotwj2WdY09WkRH8eAQ44IH9FZ2lDuwD9vhhV5rN\nX/Z5IS3ovSRpIqV3gm8j+Ldg/xcCy9bgZkrljrp167Jy5Uq+++47WrZsedJ977vvPu655x6uvPJK\nhg8fzp49e5gxYwZdu3ZlzZo1VKhQAYD58+dz5MgRRo0aRUJCAl999RVPPPEEv/zyC/PmzUsvT0Tw\n+Xz06tWLzp078+ijj6bXZnz44YdccsklnHLKKdx6663UrFmTDRs28N5773HzzTenlxE4vkOHDjz6\n6KMsXbqUxx57jEaNGjFixIiTPp8VK1bQqlUrYmJicvRa7dq1i1atWmVan5iYiDGGL7/8ktNPPz1H\nZRVnmmgUnECKv9AYM+2ke0bfpdiEZpAx5oeQbQ0icQJjzF8icoTo1ThUdh73RKl814hIPNAS28el\nDcScDt624Kt4Yq+qx6GuQD1vhrwj/bG6gCe24KMHbFLtJNYGmwvuBn4CtsTAlrp22XQctsXA0fSa\nVZFSqWC2gG8DkOIs64wxhwv4OZRYd9xxBxdddBFt27blrLPOonPnzpx//vmcd955Gb7V79ixg0mT\nJnH//fdnGGXRr18/2rZty9NPP83YsWMBeOihhyhd+kQePGzYMBo2bMjdd9/Nzp07Oe20E3cUOHbs\nGAMHDmTKlBOT//r9fkaMGMGpp57K2rVrTzqy5O+//2bQoEHcddddAFx//fUkJiYyc+bMbBONH374\ngQ4dOuTodZo7dy6//PJLhjgDTjnlFEqVKsX333+fo7KKO000Ius4Wb+mH2Kn4r4MOxLETYGL9C9h\ntg2M4HnWAWeLSKMs+rrkR3PncW2Eyy1Qzj1bEklPKkolkt7MJkD949AuNn0zLbHJRGm3kohcEuzb\nrTInZplPF2sTkV0EVWxUha1V4dtE+PYa8AmIX6T0j3BsBZp8RF2PHj1YsWIFU6dOZcmSJaxcuZKH\nHnqIatWq8cILL3DJJZcAsGDBAowx9O/fn717T3TFql69Oo0bN+aTTz5JTzSCk4zDhw9z5MgROnbs\niN/vZ82aNRkSDYCRIzPOxbdmzRp++uknHn/88RwNXw1NKDp37szcuXOzPW7v3r1Urlw52/1++OEH\nbrzxRs455xyuvjrsFEhUrly5UIw+Kgw00YisX4FTRKSCMeZg8AZjzFci8iFwgYg8CYwzxmTovyAi\nbYBaxpglUY5zE3YirZHAw0HnvwI7ZDRSnfs+B84GzsS29UfSWURvjo6ocJo/6mOH5HaB2B7Y6dmB\nij5oK9A2xiYUrbFJRXwRSSjySoCaznJO8Aav7Wv6LZDigZSm8FXDjMlHqU1wfCU28fgU+FY7pkZG\nYmIib77lSp5aAAAgAElEQVT5Jj6fj3Xr1vH2228zbdo0+vfvz9q1a2nWrBmbN2/G7/fTqFGjTMeL\nCKVKnZhL7+eff2bChAksWrSIffv2ZdjvwIEDGY71er2ZEo8tW7YgItk25QDExcWRkJCQYV3lypUz\nnPdksnsL7dq1i4svvpjKlSszf/78LIfTGmPyPdS2uNBEI+/CvYP+D7gRWCMiy7F9KzYaYx5xtg/B\nzlNxAzBYRNZik5OK2KtLbWA6EO1E4yGgF/CgiAzAJh6Nsd+sHwHGROg87zlldQNeC90oIjWxk48F\n/rObYV/XZ0QkkKi9Z4wJdwO1rtg6+eURijXinMSiKScSi/PheA37FFv6oLvXbjobOMWb/24xxU1p\n7FsyMbAiNPloBqsawXdO8uHdKyLvY+eM+cgY87M7cRcfXq+XxMREEhMTady4MUOHDmX+/PlMmDAB\nv9+Px+Ph/fffx+PJPK6gXLlygG326NGjB/v372fcuHE0bdqUsmXL8ssvv3DNNdfg92fsXhZc+5EX\nOe1fEU5CQsJJE5KDBw/yj3/8g4MHD/LFF19Qs2bNLPfdv39/2M6zJZEmGla4FNZksf5kx4xzHvsC\nA7Cv76fYi3dgvoxOwHBsZ8y22EnBAnXH04F5ZJbdt7RcxWqM+VxEzsXOUHoGNslYjx2UtxY7A2le\nXpOMOxvzmYhsAi4XkX+FDr/FXknODHOO4JulZWrkdGKvDTxujDmW03gKgojUBi4BOQ+83cFXBTwG\nzkiD85zE4hygsv7v5Um45OMI8CWwNAGWDIR1V4FBpNQWOB5IPJYZY0pEx+Foad++PQC//WYnJW7Y\nsCHGGOrVqxe2ViNg/fr1/Pjjj7z88stcddVV6euXLl2a5TGhAuf69ttv6d69ex6fQfaaNWvGtm3b\nwm77+++/6d27N5s3b+ajjz6iadNMTYHpfv31V44dO0bz5s2z3KckKfHDW40xHmOMN2TddmNMjDHm\n/CyOmexsnxOy/rAx5hZjTD1jTGlnn+4h+xwzxjxljOlsjKlijCnj7N/dGDPNGPNrmPganiT+2c55\n7s1i+3nGGK8xZkfI+q+MMRcYY6oaYyo58byT1XPP7jwn8TS2gf6SMLEFznWy5bowZf4TO+zh+VzG\nEnHOsNI2IjJBpNQ6YAfEPAkdL4N/V7GVU/sFkr22lao3J7rIqMgoA/TAjtRe47X9g+cDQxtCnRHA\n28BekVJfi8i9InJ2pO7nUxwtW7Ys7Pr33rP3RWzWzH4P6NevHx6Ph8mTJ4fd/48/7FQ3gRqG0JqL\n6dOn57hpoV27dtSvX5/p06dnamqJpI4dO/Ltt99y/HjG2QT8fj8DBgxg1apVvPnmm5x1Vrjpjk5I\nSUlBROjUSe+QAFqjoaLvOeB2YCz2Az9fRKQWNtF42RiTeQB7AXDmgjgH6AuxV8Dx2hCfBr09tq/v\nhWKHlCp3JABXOAte2Ap85IEPE+HDtrB/PMT+KiKvYTOSryLet8PtPoD5OP9NN93E4cOHueyyy2jW\nrBnHjh3jyy+/5I033qBBgwYkJSUB0KBBA6ZMmcJdd93Ftm3b0uet2Lp1KwsXLmTEiBHcfvvtNGvW\njIYNGzJ69Gh27txJhQoVWLBgAfv357yCSUR45pln6NOnD23btmXo0KHUqlWLH374ge+//57Fixfn\n/QkH6du3L1OmTOHTTz+lR48Tt4i6/fbbWbRoEX369CE1NZVXXnklw3HBNTUAH3zwAXXq1KFtW524\nGLAdVnTRJZoLNjFIAy6KQFnTsDeuO62An0M8dljwS+DdDxiofhxGGnjfwFEDRpdCv/gMfGzgBgMJ\nx+3fMXYntnnzbEBy+H5oB5iUlBQTbPv27SauTFygidHVJa5MnNm+fbvJrSVLlphhw4aZFi1amAoV\nKpi4uDjTpEkTc+utt5o9e/Zk2v/tt982Xbp0MeXLlzfly5c3LVq0MDfffLP58ccf0/f54YcfTM+e\nPU2FChVM9erVzciRI8369euNx+Mxs2fPTt8vKSnJVKhQIcvYli9fbnr16mUqVqxoypcvb9q2bWue\nfvrpbI+fNGmSiYmJydHzP/30083w4cMzrOvWrZvxeDxZLsH8fr855ZRTzMSJE3N0vqykpKQE/pbt\nTB4/twrLIsb+0yilQjjV652Ba8EzEPxx0MQHV3htN5z2aOtjUZaGvVXRfGCeD/7wQuwvcPx1sqnp\ncG6umJKSkkK7du0ybCsON1UryebOncuNN97Ijh070iccy42FCxcyZMgQtmzZQo0aNfIcx+rVq0lM\nTARINMasznNBhYAmGkqFcDp0Xg3e4eCrC3V9MMxr+/c2cTs8FRVZJh2zgP8aY7YH732yREMVbcYY\nTj/9dAYNGsS4ceOyPyBEp06d6Nq1a55vMx+giYZSxYyIxAF9IeY6SOsBcQYGeuBabKWG9h0sOQJJ\nx+vA3DQ47AHP++B/BvifsVP1a6Khoqo4JRraGVSVWE7TyBnAteD9J/gqQPs0GC4wQKJ3mxZVuMUA\n5znLozE24XjqAlh7IXh/E5FngFWuhqhUEaKJhipxRKQ0cCV4bwdfG6jqg+u8MBRoqqNFVJBywDBs\n01ky8FwtmDspZ/c5VEqB9mRTJYiIVBORCeDdCcyCHq1gEfCb187BkPUEPErZzr//BX73wJ362alU\nDuk/iyr2RKSpiDwPnp1QehJcV9VOOLrYYyfQ0oo9lRsVsR2DlVI5oZ+wqtgSkTPBMw64FKqlwS1e\nGAHo/QeUUqqgaKKhihWng+cFEDMO6Ab1fDBW4J9eiHM5OqWUKnk00VDFhoh0A+/D4GsPrdPgbuAy\nrx1FoJRSyg2aaKgiT0TOgJgHgQugTZrt2NkjRue+UEop92lnUFVkiUgjEc/rwGqofx68CSTHwAVo\nkqGUUoWD1mioIse5g+sEkOuhuoEpQJJX385KKVX46CezKjJEpBLwb/DcDuW8MD4GbgTKuB2aUun0\npmp5k5SUxKeffsq2bdtyfWy9evXo3r07L774Yrb7/vzzzzRp0oSPP/6Yjh075uo8f/zxB3Xq1OHN\nN9/kH//4R67jLKk00VCFnojEAjdCzCTwloPbPfBvoJLLkSmV0Y4dO2jetCmHj7o/c2h8XBwbNm7M\nU7Kxfv16Jk+eTHJyMrt27SIhIYEWLVrQp08fbrzxxihECyKCx5O31nw72Cxn/vOf/9ChQ4cMScbC\nhQt59tlnWb9+PXv37iUhIYHExETGjRuXYb8qVaowbNgwJkyYoIlGLmiioQo1EekE3v9CWnMYJjAR\nqOV2WEqFlZqayuGjR5kLNHcxjg3AkKNHSU1NzXWisXz5crp3707dunW5/vrrqVmzJj///DMrV65k\nxowZUUs0XnjhBfx+f1TKDkhNTWXOnDm8/PLLGdavX7+eKlWqcOutt1K1alV+//13XnzxRbp06cJH\nH31Ely5d0vcdOXIkM2bMYNmyZXTr1i2q8RYXmmioQklEErDDR4bZkST/FdC7ZKqioTlF99163333\nUalSJZKTkylfPuONBaPZJBQTE0NMTHSHor/88svExsbSu3fvDOsnTJiQad/rrruO0047LT3hCGjW\nrBmtWrVi1qxZmmjkkI46UYWKWNeA90coOxSeAr6KKbof20oVLVu3bqVly5aZkgyw/T6CpaWlce+9\n99KoUSPi4uKoX78+d999N8eOHct07OLFi+natSsVKlSgYsWKnHXWWbz22mvp25OSkqhfv36GYx55\n5BHOOeccqlatSnx8PO3bt2fBggV5fm7vvPMOZ599NvHx8dnuW61aNeLj4/F6M38fv+CCC1i0aFGe\n4yhpNNFQhYaItICYz4BZ0L8SbI6BUeiEW0oVnLp165KSksJ3332X7b7XXXcdEydOpH379kyfPp1u\n3boxdepUBg0alGG/WbNm0bt3b/bv389dd93Fgw8+yBlnnMGSJUvS9xGRTH0tZsyYQbt27bj33nuZ\nOnUqsbGxDBgwgMWLF+f6efl8Pr7++mvatcv6S8uBAwdITU3l22+/ZdiwYfj9/rBNRYmJiezfv5/v\nv/8+13GUSMYYXXRxdQHigftBfNDgOHxowOiiSyFeUgxgUlJSTLCUFGe9ywGmQNj4cuLDDz80sbGx\nxuv1mk6dOpk777zTfPDBB+b48eMZ9lu3bp0RETNixIgM68eMGWM8Ho9ZtmyZMcaYAwcOmAoVKphO\nnTqZv//+O8vzJiUlmfr162dYd/To0Qy/+3w+07p1a9OjR48M6+vVq2eGDh160ue1ZcsWIyLmqaee\nynKfZs2aGRExImJq1Khhli9fHna/FStWGBEx8+fPP+k58yPwXgLaGeP+53R+Fq3RUK4SkV7g3Qix\nd8KkGPjOCz3cDkupEqtHjx6sWLGCvn378s033/Dwww/Tq1cvTj311AzNBf/73/8QEW677bYMx48e\nPRpjDO+99x4AH3zwAYcOHWLs2LGUKlUqV7GULl06/ef9+/ezb98+OnfuzOrVq3P9vPbu3QtA5cqV\ns9xn1qxZLFmyhGeeeYYqVarQu3dvvvnmm0z7BcooDMOYiwLtDKpcISJlgIeAG6GLH57zQCO3w1JK\nYZsG3nzzTXw+H+vWrePtt99m2rRp9O/fn7Vr19KsWTO2b9+Ox+OhUaOM/7c1atSgUqVKbN++HbB9\nPgBatmyZ6zjeffdd7rvvPtauXcvff/+dvj6vw2ABjDFZbjv77LPTfx44cCDNmzdnzJgxGZp4gsvI\nzbDakkxrNFSBE5G24F0LpUbBk8BSTTKUKoS8Xi+JiYlMmTKFp59+mmPHjjF//vwM+0TrYvv555/T\nt29f4uPjeeaZZ1i8eDFLly5l8ODBJ00WspKQkADAvn37crR/pUqV6N69O6tWrcq0LVBGaOdYFZ7W\naKgCIyIeYDTI/XYA4OseaOF2WEqpHGjfvj0Av/32G2A7jfr9fn788UeaNm2avt/u3bvZv38/devW\nBaBhw4YYY/j2229p0KBBjs/31ltvUaZMGZYsWZJh5MfMmTPzFH+dOnUoU6ZMrmYePXLkSNjak23b\ntiEiNG/u5mwpRYfWaKgCISK1IeYTkIdgjBe+9mqSoVThs2zZsrDrA30umjVrBsBFF12EMYbp06dn\n2O/RRx9FRLj44osB6NmzJ+XLl2fq1KkZmj+yExMTg4jg8/nS1/3000+88847uXk66bxeL+3btyc5\nOTnTtj179mRa99NPP2WarCsgJSWFihUr0qKFfoblhNZoqKgTkSsh5nmoXgZeAc5zOySlompDET7/\nTTfdxOHDh7nsssto1qwZx44d48svv+SNN96gQYMGJCUlAdCmTRuuueYann/+efbt20fXrl1ZtWoV\nc+bMoV+/fnTt2hWA8uXLM23aNIYPH86ZZ57J4MGDqVy5MuvWrePIkSO89NJLYeO4+OKLeeyxx+jV\nqxeDBw9m165dPP300zRu3DhsB82c6Nu3L+PHj+fQoUOUK1cufX3r1q05//zzadu2LZUrV2bTpk28\n+OKLeDwe7rvvvkzlfPjhh1xyySV5iqFEcnvYiy7FdwEqgswFDAz0wx+FYFiiLrpEYgk/vHX79u0m\nPi7O2Pe8u0t8XJzZvn27ya0lS5aYYcOGmRYtWpgKFSqYuLg406RJE3PrrbeaPXv2ZNg3LS3N3Hvv\nvaZhw4amdOnSpm7dumb8+PHm2LFjmcp99913zbnnnmvKli1rKlWqZDp06GDmzZuXvj0pKck0aNAg\nwzEvvfSSadq0qSlTpoxp0aKFmT17tpk0aZLxeDwZ9qtfv7659tprs31uu3fvNrGxseaVV17JsH7y\n5MnmrLPOMgkJCaZUqVKmdu3aJikpyWzevDlTGRs2bDAiYj755JNsz5cfxWl4qxhjopjGqJJKRBLB\n+w6UrgnPxcBgQHtoq+JiNZBISkpKpgmg9O6thduwYcPYtGkTn332WZ6Ov/XWW/niiy/CNsFE0urV\nq0lMTARINMbkfjxvIaJNJyriRORq8LwAbTywIAbquR2SUgWmTp06eoEvxCZOnEjTpk1ZsWJFnm4T\n/+KLL2YaeaNOThMNFTHO7dwfBm6Bqw08IxDndlhKKZWudu3aHD58OE/HVqlShYMHD0Y4ouJPEw0V\nESJSDWLmA13gcWCUaFOJUkopTTRUvonIGeB9FyrUgLcFMg8HU0opVTLpPBoqX0RkMHhWQKsasDZG\nkwyllFLBNNFQeSIiXhF5BHgFrioFy2OgttthKaWUKmS06UTlmohUgpgFYM6Dx4CbtT+GUkqpsDTR\nULkiIqeA90Mo29T2x9BZPlXJtWGD23OAquKqOL23dMIulWMi0gS8H0G1mvCR194YTamSaAceT3P8\n/rwNk1QqJzwez1G/39/UGLPD7VjyQxMNlSPOTJ8fQP0KNsnQ/hiqpNsBuD8DaP5tA27ywe6/IG0k\nsMntiFS61KKeZIAmGioHRKQHxPwftCsFi2Mgwe2QlFIRtRe4IA2+OQxpvYwxK9yOSBUfOupEnZSI\nDABZDOeXhk80yVCqWEoAlsVAh3iI+VhEursdkSo+NNFQWRKRfwGvw6AYeNcDZd0OSSkVNRWAD2Lg\nvFLgeV9EersdkSoeNNFQmYg1GXgSbhN4WSDW7bCUUlEXj/1S0ScGZKGIDHQ7IlX0aaKhMhARAZ4A\n7oEHgEfRt4lSJUlpYL4HBntAXhORa92OSBVtegVR6ZwkYxrwL3gOuBOdiEupksgLzBEYIcBMEbnZ\n7YhU0aWJhgLSk4wHgFvgaeB6lyNSSrnLg/0suAPgcRG52/mcUCpXNNFQAROBf9sKjRvcjkUpVSgI\n8BAwGWAKMN7VcFSRpPNoKERkHHC/rdC40+1wlFKF0r3APQA3GmOecjkYVYRoolHCichtwGMwCVup\noZRS4RhgNLbWkyHGmFfcjUcVFZpolGAiMgp4CsYC96MdP5VSJ+cHrjUwx4Dpa4x51+2IVOGniUYJ\n5QxZmwm3Ym/1rkmGUionfMAVfljkA/8FxpjP3I5IFW6aaJRAInIV8DLcIPAUmmQopXLnKHChHz4/\nAmldjDGr3Y5IFV6aaJQw9h4G8gFc7YEXRQceKaXy5k/gvDRYdwB8nYwxG92OSBVOmmiUICLSDGK+\ngvPKwv88Oq24Uip/9gLn+GDLHvCdbYz52e2IVOGjiUYJISLVwJsCjWrBSi9UdDskpVSx8AvQ0Qe/\nbXeSjb1uR6QKF603LwFEJA68i6BSLXhfkwylVASdCnzshQp1IWaBiGhVqcpAE41iTkQ84JkFnjPh\nPS/UdTskpVSx0whY6AW64Ey0oVSAJhrF3yTwD4RXPHCW27EopYqtzsAzAvxLREa4HY0qPLSPRjEm\nIlcDs3VqcaVUwbkJeCoNTHedY0OBJhrFloh0BVkKQ2PgBdG5MpRSBeM40DMNvjgIvnbGmJ/cjki5\nSxONYkhEGkNMMnQuBx/oMFalVAHbCyT64JeN4OtgjDnkdkTKPdpHo5gRkXjwLoR68fCWJhlKKRck\nYDufxzYDz8u2U7oqqfSPX4yIiIA8A95mtgd4ZbdDUkqVWC2BeTFgLsW5v7wqmTTRKF6GgbkanvdA\nK7djUUqVeJcA9wFMFJF+LgejXKJ9NIoJEWkHnpUwPBaedTscpZRyGGCAgbf/grRWxpjtbkekCpYm\nGsWAiFQE7zfQ6lRYEQNxboeklFJB9gOtfLArGXydjTE+tyNSBUebToo42y/D818ofSos0CRDKVUI\nVQLmeSHtbOBut6NRBUsTjaJvGPj7w0sx0MDtWIqgVOAOoBkQD1QFzgbGhuy3CLgGaANUA0oB1YEL\ngYV5OO9Q7L9fVsvzWRx3BJiC7YMTiPci4NM8xKBUQToHmCQgE0XkXLejUQVHm06KMBFpBZ4UGFYK\nnnM7nCIoBegF7MP2kG8FHAS+B3YCx4L27Q+87exXBygP/ASswrZB30LubvEwFJjjnL9mmO3XAF1D\n1v0FdHPiTgA6AQeAldhJkmYCSbmIQamC5gO6psFXu8DX0hiz3+2IVPRpolFEOfNlrIHGDSDFC2Xc\nDqmISQWaA0eB14GLQ7YnA+2Dfl+HTTBChwx/DZyPTQK+AhJzeP5AovEJ9j5UOXET8BRwJrAYqOKs\nXwn0xCYbm4DaOSxPKTfsAFqmweG3wD/Q6EWo2NOmk6JrKngawXxNMvLkHuAP4BEyJxmQMckAOJ3w\n85KcCQx0fv4iYtFldhx4CTuV/AxOJBkAHYCbsTUw06MYg1KRUAeYGWObfBnqdjQq+jTRKIKc9s2b\n4AGPrcpXuXMUeAUoS2SaGgKzr1aMQFlZ2QAcBkpj+5CEOg/bhPNOFGNQKlIGAEMNeJ4SkaZuR6Oi\ny+t2ACp3RKQMeOdAoh9ujnE7nqIpGfgTe1vr0thmiKXYBKQJ9kOwVg7LWg/Mw/aqvyQPsSwA3gTS\ngPpOGeE+d/9yHrNKZhKcx23AIaBcHmJRqiDNEPjMC9vnicjZxpi/3Y5IRYcmGkXPZJC6MMsDmmfk\nzffOY3XgMmwtQODutga4C9ux8sowx76LTQ6OY9ualwMtsM0aCWH2z86TQT8b4E7gBmzzSHCFYzXn\ncQ/wNzZBCrYt6OcdTkxKFWblgDe8cHYb7JBXnaa8mNKmkyJERM4GuQP+47HDMVXe7HMe3wGWAM8A\nu7GjSMZgh5AmAd+EOXYdthPna8CX2Fy9C3BKLmNoh53BdRO2SWQr8DS2H8jTThzBGmFrWQwwO0x5\nM4N+/jOXsSjllnbAOAEZJyKaHRdTOuqkiBCR0s7snw3h6xitjMqPqdgvUAI8BIwO2T4Q25wxGHg5\nizKOAZuxScpTQF3sqJNqWeyfU99jP3wNNvk4NWjbE9hhtOWwNR59scNxH3O2ebFNMCuxnVSVKgqO\nYmcN/SkZ0s4xxvjdjkhFltZoFB3jgcYwR5OMfAvuv5AUZvtQ7IX+ZJNglcI2TzyBHXa6A3gwArG1\nAPpg5xv4KGTbTdjRJX8B12KbaupjE537ODEqRu/aq4qSOOAFL6R1AIa5HY2KPE00igAROQPkLpgg\n0NrtcIqBus5jPOH7VdRzHnfnsLx/YhOTSA1vbew8/hZm23RgDTAZuB6bf67G1sr8gR3qrDPEqqKm\nG5BkIOZREQk3g50qwvSrcSEnIrF2lElTP4zVxDAiznAej2A7dcaGbP/DeczpyI2qzuOhfMYVEOhD\nUjaL7W2cJdhn2GaTc9HvD6poekTgnTJw4HFOTE6jigH9RCr87gR/S5jjtdX1Kv9qYyfgyqp5ZJnz\n2C6H5QX2b56vqKxjwHu5PD/YJhzB1nIoVRQlADNiwD9ARC5yOxoVOZpoFGIiUhc8E2CM5O6io7L3\nb2yicQfwe9D6tdjOlQKMdNalAi9ga0BCfYgdkipkbl7+Gpt8XBCyfiMwl4z3UgmcZyDwM9AWey+T\nYHucbcHSgInYIbfdgX5hYlSqqLgKOD8NvM+LSFZVeqqI0VEnhZhIzDyo2g+2eHUCpmi4FjtUtCL2\non4EOy/GMWzNwDPOftuxnS7jsfcyOQ3bIXMT8AM2yRiP7TcR7FPsjJ31sCNIQtdXxk51Xg34FXuz\ntEPYGpePgIZhyuuObfqpj02UVmATpXbYpKdSrl8FpQqXLUALPxybZoy5w+1oVP5polFIicg5wBcw\nC3snTxUdM7F3vt2ATRhOB0YAQ4L2OYKd22IZ8B22k6gfO6/FOdiaj9DaBziRGNTDfngG/Ia9x8pK\n7Nwde7ETcDXBjji5mfAzgO7E1l58CfyCrZBsiv0W+C+0y5UqPh4A7jJg2htjVrsdjcofTTQKIRHx\ngDcFWrWGlBht4VJKlSzHgdN98GMK+DrqHV6LNr2CFU7/BF9beEKTDKVUCRQLzPCC72zgcrejUfmj\nNRqFjIiUA+9W6FcV5kn2RyilVHF1oR+W7gRfE73pWtGlX5cLn7EgCfCQJhlKqRLuUQ/4a2M7Iaki\nSms0ChFnOOuPMC4WprgdjlJKFQI3AP/9E9LqG2P2uh2Nyj2t0ShUPA9DVYGxbgeilFKFxGSgdDz2\nToiqCNJEo5AQkXPB3x8e0jkzlFIqXXVgbAx4bhKROm5Ho3JPm04KCZHYL6BlB1itI02UUiqDQ0Dd\nNNg3xxj/tW5Ho3JHr2iFgIh0Bd85cK8mGUoplUk5YFIMkCQiLdyORuWO1mgUAiLeT6DFubDOa2en\nVEopldHfQGMf/PKuMWmXuR2Nyjn9+uwy2zcjrRtM1iRDKaWyVBq4zwv+S0Uk0e1oVM5pjYbLRLwf\nQtNusN6reZ9SSp1MGtDQBzsWGuPv73Y0Kmf0yuYiEekAaT1goiYZSimVrRhgnBe4XEQaux2Nyhm9\nurkqZiI08elU/koplVPXAAlpgN5CvojQRMMlItIe0v5hazNi3A5HKaWKiDhgtBc814pILbejUdnT\nRMM1ngm2rXGg24EopVQRcwNQRoBb3I5EZU8TDReISFvw94F7tDZDKaVyrSLwrxiIuVFEKrodjTo5\nTTRc4RkHdX0w2O1AlFKqiLoVkDLASLcjUSeniUYBE5FTwFwOt3vB63Y4SilVRNUChnrAO0ZE4tyO\nRmVNE42CN9xOPHON23EopVQRNwZISwCudjsSlTVNNAqQiMSCdxRcE2PbGJVSSuVdY+ByA967REQ7\nvBVSmmgUrL7gqw6j3I5DKaWKiTECvrpAT7cjUeFpolGgvDdDpzRo43YgSilVTJwJtPKBZ7jbkajw\nNNEoIPbWxr7OcJNW7ymlVMQIMMILpo+I1HA7GpWZJhoFZxRU9UE/t+NQSqli5iogVoAklwNRYWii\nUQBEpDzEDIWRXijldjhKKVXMVAb6C3hHioi4HY3KSBONgjEETBkY4XYcSilVTF0v4KsHdHU7EpWR\nJhpRZrNr7y3QBzjN7XCUUqqY6gw08IFop9BCRhON6DsLfE1hlFbnKaVU1Ai2eVr6i0gVt6NRJ2ii\nEX1X2k6g3d2OQymlirmrAfECQ9yORJ2giUYU2ZnqvFfBYL1Lq1JKRV0NoC/gvUE7hRYemmhEV2fw\nVYfmR7sAACAASURBVINBbsehlFIlxHABXzPsTF6qENBEI7quhNN8cLbbcSilVAlxAba5moFuR6Is\nTTSixLmB2pUwxGs7KSmllIq+GOAKL8QO0OaTwkETjejpAb6KcKXbcSilVAlzOXD8NOAMtyNRmmhE\n0yBo7NMbqCmlVEHrClRIQ+/5UChoohEFIlIGYi7XZhOllHJDLNAvBmK1n0YhoIlGdFwEafHabKKU\nUm7pBxxvJCLN3Y6kpNNEIypkEJzugyZuB6KUUiXUBUCZNGyHDeUiTTQiTERKg1wMV3rdjkUppUqu\nOOASz/+3d9/xbtfVH8dfJ/fe0rJaRtkbBARayiqz0CKCgKgsERCcuPdCFEFFBVEBFQRZKggCykYQ\n+LH3lC17lUILhUJLS+lNcn5/nO8t6e29bW5ukk/G+/l4hORmfHN6yU1OPuMc6Pp46kjanRKN6tsO\nikNht9RxiIi0uX0MukeZ2ZqpI2lnSjSqbxcYmYdRqeMQEWlzuwJDimj3SVJKNKqua3fYTbtNRESS\nWxz4kEHnvqkjaWdKNKrIzFaE7g1gl9ShiIgIAHsY5Lcws+GpI2lXSjSqa+cYydgpdRwiIgLABIjP\nunGJA2lbSjSqa5fY1joydRwiIgLAWsBK3cCOqSNpV0o0qsTMctC5a6zPEBGRxmDAB7ug64OpI2lX\nSjSqZxPIj9D6DBGRRjMB6N7QzJZJHUk7UqJRPbvAogXYOnUcIiIyjwkQQxs7JA6kLSnRqJrO3WCn\nXDTzERGRxrEasHo3WcYh9aVEowqiW2thK9hFxTNERBqS1mmkokSjOjYB79C0iYhIo9oR6F7PzJZP\nHUm7UaJRHWOjzO1GqeMQEZE+jZ/vgtSHEo3qGAubuNZniIg0qhWBdbROIwElGlUxZFvYuiN1FCIi\nsiA7dkHX9qmjaDdKNAYp9mXPWQ3Gpg5FREQWaBMgv66ZDU0dSTtRojF4W8SZEg0RkcY2hli4z4ap\nI2knSjQGbywMz0c9fRERaVyjyT72xiQOpK0o0Ri03JawVS6KzomISONaFFirm5hDkTpRojEIZmaQ\n2xq21O9RRKQpbNEFnZuljqKd6ANycFaH/FJanyEi0izGAL5xdNyWetAvenCyDGOLtFGIiEiZNgEK\nw4B1UkfSLpRoDM5GMLIblksdh4iIlGXMfBektpRoDM76sIF+hyIiTWMksLwWhNaRPiQHZcgo2EAV\nQUVEmspmnZBTolEnSjQqZGYdkF8b1k8dioiIDMgmBh2bpo6iXSjRqNzqUOxSoiEi0mzWAbpHmtmw\n1JG0AyUalcsyjPXSRiEiIgO0Rs+F1RIG0TaUaFRubehyWDV1HCIiMiBrzHdBakeJRuXWhtXy+hWK\niDSbVYCco0SjLvQpWTFbC9btTB2FiIgMVCewYh5YM3Uk7UCJRsW61od11ElNRKQprZ1DIxp1oUSj\nAlEjP786rJ06FBERqchaHdClMuR1oESjMiOhOETJsIhIs1pj7n+ktpRoVCZrbrJ82ihERKRCawLd\ny6iWRu0p0ajMyHnORESkyazRc2H1hEG0BSUalVGiISLS1NbouaCdJzWmRKMyI6NY1xKp4xARkYqs\n2HNhhZRRtIOKEg0zu97MRvRx/ZJmdv3gw2p4y8HSedDuVhGR5tQFDCsA832WSXVVOqIxHhjSx/VD\ngXEVR9M8Rs5dDyoiIk1qySJKNGpuQJUtzWx0yY8bmFnpkFMH8CFgUjUCa3AjYQVVBRURaWojgClK\nNGpsoB+WDwCenfqaInkH+Ppgg2p8nSvAcpo3ERFpakvl0IhGzQ000ViTWJjwLDAWeK3ktjnAq+5e\nqFJsDSy3vHaciIg0u2U6UKJRcwNKNNz9hexim+9WKS6rRENEpNmNADqXSR1Fq6t4nYGZvQ+YQKyK\nnCfxcPefDzKuhmVmHcBwJRoiIs1uBJBbOnUUra6iRMPMDgFOBqYCk4k1Gz0caNlEA1gUMNXQEBFp\ndiPm/kdqp9IRjcOBH7v7r6sZTJPoirO+dveKiEjzGAEUlkwdRaurdK3FUsA/qxlIE+ma50xERJrU\nCKCwWDYlLjVSaaLxT2DnagbSRJRoiIi0hOE9FzSqUUOVTp08DRxlZlsBDwPdpTe6+x8GG1gDU6Ih\nItIS5k6B6w29hipNNL4AvA3skJ1KOdDKiUb2ytTrUkSkuXXMd0Gqr6JEw93bua2uRjRERFqCEo16\naPPCWxVRoiEi0hLm5hf6LKyhSutonLmg2939s5WF0xSUaEgbcOAqYFrqQERq6JGeCxrRqKFK12gs\n1evnLmAjYq9QX83WWokSDWkD4+ngZtqgcZG0uQ6goDf0mqp0jcaeva8zsxxRLfSZwQbV4FSwS1rc\nV4CbAdgP+EHSWERq5y7i1U6vnZNSXRX3OunN3YtmdhxwI3BstY7bgLLfmUbapBUdD5wMG0PhZbjp\nNTgXTWBLayppP55PF0Xrq/b7x9pUMXlpULPnORNpGZeBfQfWBD4CjI1GRjcnjkqkVkqyC80S1lCl\ni0GP630VsCKwO/C3wQbV4GbOcybSEh6C3J6wNDFf0gFsBp1Xwt8dxqcNTqQmCn1elGqrdPRhk14/\nF4lRqO8CC9yR0gKUaEiLmQy2JQwrwkHA0OzqHORXhfNfhBNLrhZpFSXZhaZOaqjSxaATqh1IE5k1\nz5lIU5sNjIaO2fBJSlo/ZLaDt8+FK4G96h+cSE1p6qQ+BrVGw8xGmtl22WlktYJqcBrRkBZRBDYH\new0+QUx+9rYudHTCWXWOTKQe8n1elGqrKNEws8Wyol2vEGvFbgZeNrMzzGzRagbYgJRoSIv4CPAo\nfBhYp/97FdaGf6PSXdJ63gQsMm69oddQpSMaxxHN1PYginSNAD6aXfe76oTWmNx9DlhBr0tpbt8G\n/g3jgM0WctfxMa58Yc1jEqmvqUAnvOnunjqWVlZporE38Dl3v8rdp2enK4FDgH2qF16jyr2jREOa\n18lgJ0Qt3x3LuPuKYENbfzuZtJ+pgMWZ1FClicaiwJQ+rn81u63FKdGQZnU12FdgVeBjxMb0MhQ3\ngFuBF2sYmUi9TQWK89TtklqoNNG4A/iZmc3d8WZmw4Ajs9tanM1UoiHN5zHI7R6divZnYHvOdog3\ni3/UJjCRJKaC5/v+0ixVVGkdjW8B/wFeMrMHs+s2Bt4Fdq5GYA1ulra3SnOZCrYFLFKIWhnDBvjw\n4eBLwt+mw6G1CE8kgSmx20RTJzVW0YiGuz8MvA84DHggO/0QWMfdH61eeI2q8Dq8kToIkTLNAUZB\nx6yoldG793KZfAz8D3ioipGJpJTNmSjRqLFKS5AfBkx299N6Xf9ZMxvp7r+uSnQNqzARXiygzmrS\nFLYCJsO+wMqDOMy20HEznAOMrk5gIklNiy/bSjRqrNI1Gl8EHuvj+keBL1UeTtOYBC8VUwchsnB7\nA/+F3YD1BnmoRaAwEs4mCg+INLPZwDvxZVGJRo1VmmisQOww6e01+q4v2GomweQO0NZraWSHARfB\nNsDYKh1ybFTpu6VKhxNJ5fU+L0otVJpoTAS27eP6bYGXKw+nabwMc3KqlSiN6wywY2ADYKcqHnYz\n6DT4exUPKZLC1D4vSi1UmmicBpxgZp8xs9Wz02eB47PbWt2kec5EGsr1YIfASsCeDLKjUS85yK8C\n5xNbzESalRKN+ql0e+tvgGWAPwFDsutmA79296OrEViDK0k0RiUNRGReT4HtAsMdDgC6avAU28GM\nf0RH1z1rcHiReiiZL1GiUWOVbm91dz8UGEksad8YWNrdf17N4BrYK3GmEQ1pJG+CbQqL5KNWxmI1\nepr11NFVml9WfjwPzEgdS6sb1KCqu7/t7ve4+yPu3jYjqdFYresNJRrSOPLAKMi9DQcS4401VFgL\nriC6X4o0o+eBrijToFX9NVbN2ds2Yy8r0ZDGsR3wUuxmXbUOTzc+Uht1dJVm9b8oP/5w6jjagRKN\ninU/Dy8pE5YGcABwF+xC7DKph5Ugtwj8TXu8pUk9AvkiPJ46jnagRKNiPgleyKeOQtrdT4F/RJ2M\nrer7zFlHV3upvk8rMmhzgBdjM4QSjTpQolG5J+DpDiikjkPa1t/BfgbrAh+i7JbvVTM+nlIdXaXZ\nPAMU4+WrRKMOlGhU7lF4NwfPpY5D2tJtkDs4avTuQ5q/5OHgS8BfEzy1yGCUZBdPpIuifSjRqNyj\n85yJ1M0LYBNgcY8dJkMW+oCa8THR9Egr6qSZPA50xrbWvlppSJUp0ajcy9DxthINqa+3wTZ2hnTD\nwcDiicPZLrpSnZM4DJGBeBwweEJbW+tDiUaF4gVqjyrRkPopAqPA3jIOAJZNHQ9zO7qehTq6SvN4\nDPLd8EjqONqFEo1ByT8ED3anjkLaxXjg+aj7vXraSOaxRZTKvTV1HCJlcODxWAiq9Rl1okRjcB6B\nJ7XzROrgM8At0Ym10drrqKOrNJEpwNsx46cdJ3WiRGNwHoXuXGyWEqmVo4G/wmbAtolD6UtHdHQ9\nD3V0lcb3eJ8XpZaUaAyOdp5IjV0A9iNYB9iN+tfKKNe2sYT/qtRxiCzEE4DFkqJnU8fSLpRoDM4U\n6JyuRENq4x7I7R89kvclBnsb1frQ0aGOrtL4Hge64IVojin1oERjELKtUY8o0ZDqewlsHCxWhE8C\ni6SOZ+EKa0dH17dSByKyAHdDYQ7clzqOdqJEY9Dy98Gd2nkiVTQLbDR0vQsHAUumjqdMO0A36ugq\njetd4J6YgLwldSztRInG4N0Kz3fFWmaRwSoCY8CmwSeA5VLHMwArq6OrNLZ7ge743NNu7DpSojF4\nt81zJjIoOwNPwUeAtVLHMnDF98MtYJNSByLSh1uBDngHeCh1LO1EicYgufsk6JqkBFkG78vAdVGX\na0ziUCo1Xh1dpXHdEqNtt7t7PnUs7USJRlV03wA36YUrg3AccEokGDukjmUQRkBRHV2lARWBW6BY\ngJtTx9JulGhUx23wYAfMTB2HNKXLwL4LawJ70Li1Msq1cezD0l4saST/A6bHJnENP9eZEo3quBUK\nBnenjkOazgOQ2xOWAfajsWtllGucOrpK47mVuYW67kodS7tRolEdj0HHDCXKMjCTwbaGYcXYxjo0\ndTxVsggUloW/oY6u0jhuBTrhQXfX0HOdKdGoAncvgt8KN+t9Vco0GxgFHbOjINfw1PFU2RbwMtqL\nJY3jRujuhhtTx9GOlGhUTfEWuN3VyVUWrghsBjY1amWsmDqeGthcHV2lcbwUpy407JyEEo3quRVm\ndcDDqeOQhrcH8FicrZM6lhrpgPzKsc1VDSUktdv6vCj1okSjeu6BXDdcnzoOaWjfAq6EccCmqWOp\nse3U0VUaw61AFzzv7irhnIASjSpx99nA9XCZ5k6kH38C+z2MAnZMHUsdrA+d6ugqDeAGyHfDDanj\naFdKNKqqeBncmoM3UwciDecqsK/CqsBHaf5aGWXKrwWXo46uks6LwKPQCfxf6ljalRKN6roi6mlc\nnToOaSiPQm4PWArYn3jLaxfjo6PrRanjkLZ1CWCQB/6dOpZ2pUSjitz9Reh6NL7DiQBMBRsLixSi\nVsaw1PHUWXR0dXV0lVQujG9/17m7BtYSUaJRdd0Xw+WFSKClvc0hamXMiloZS6WOJ43i+7Gb1dFV\nEpgK3AK5ogbVklKiUX2Xw/QOuD11HJLclsBk2BdYOXUsCY2PJSnnpY5D2s7lzB1KuzRpIG1OiUb1\n3Qudr2v6pN3tBTwAuwHrpY4lsRHg6ugqCVwExU64U9ta01KiUWVRjjx/KVysuZO2dShwMWwDjE0d\nS2Pw0fAI8FjqQAbgfuDXwN7EZqEcC+95Nw34PvA+onXN8sSA1oMVPP9nsufs73Rqmcf5XMlj2mmc\ndQaxLD8P/0odS7trp/Xv9XQ5PPNZeIp4y5H2cQbYsfB+YKfUsTSQcdBxW3R0/WXqWMp0FDHeXu5O\n5MnAtsBzRFX53bLrLibGN69g4C8JA3YBVujjtnIGym4A/kIkGe22Gvc/QHf80y9OHUu7U6JRG9dG\nldAruuDbqWORurke7JBYj7EnGi8sNTTr6Do1PsCb4VezDbAxMSi1ObA6Cy6nfgiRZOwOXMB7G4wu\nIybSDgSeBRYbYBw/BLYf4GMA3gW+BGwELAncUcExmtlF4F3w2Bz351LH0u6a4e+96WRtiK+DC1Ul\ntG08BbYLDHc4gGjfJPPaHCbRPMP33wd+SoxMLLeQ+75EFGnoAk5m3l3MHyF6500Fzqx6lP37OfAM\ncArt93KcA1wOxW74Z+pYRIlGDRX/Abd1RF06aW1vgm0Ki+ThYGDR1PE0qC1at6Pr/dn5msAqfdw+\ngZi6qNfWh4eB3xLrM7ap03M2kuuBmbGkRttaG4ASjdq5GHLvwrmp45CaygOjIPd2jI0vnTqeBtbC\nHV1nZuf9lUpZJjuvZFHohcA3gK8SycMTC7m/A1/IYjm2gudrBRcBXfACsQZZElOiUSPuPgP8Ivhb\nvv2WYbWTbYGXYB9ia4Is2DYwnVio10pGZucv9HN7zyKBN4BZAzz2icBJxBTID4ANgK8BxQXc/27g\nd8DwAT5XKygAF0YTtX+6u958G4ASjZryc+DxTngodSBSE/sDd8e2gPenjqVJbAAdLdjRdSywCDAF\nuKaP2/9acnlGmcfclEguniSSk2eBPxEjFX8i1pD09hJwODFVc2CZz9NqbgLeiI0OmjZpEEo0ausa\n6HwTzk4dh1TdEcB5Ufxz69SxNJdC1tF1eupAqmhJ4CvE2OXBRCOv6cQ0x37A4yX3LfdN9+vETpa1\niSRmdWIXyc3AEGLkondZ968S01InV/KPaBGnxG6Tp4A7U8ciQYlGDbl7N+TPhr/mo4eltIa/gx0V\nhQx2SR1LE9ohPgxb7evm0URxrteI7awjiIGuy4Dfl9xvxCCfZwNiJ0seuK7k+guJBO5Q2rd6zxTi\nddUNf9K0SeNQHY3aOxNe/3rMSu+ROhYZtNsgd3CUfNwbpeqVWCXr6PoufLr8elgNbwhwPnAb8df+\nGrFs5xMl91mH6mw17UkkXim57ors/Bpi+qDUA9n514h1G58hRl5azV8Aj291rTY719SUaNSYuz9g\n1vUwnLkR7NEyb6rt6XmwCbCExwT4kNTxNK/i+thND8LLwEqpg6mybbNTqZ5PvQlVeo5p2Xnv4l8G\n3LWAx/XseqlWHI2kCJwMeYfz3P2N1PHIe/R9rC7yp8ag5qupA5GKzQAb4wzphoOAxVPH0+TGt1dH\n15OIf+/nq3CsOURxMIgFoz3+Quy46OvUU1n01uznI6oQR6O5FngROh3+nDoWmZcSjfo4F7zQmqWK\n2kERGA32lnEAsGzqeFrAUuCLt1ZH14nEdEmp2cSCznuI6YrNe91+D7GO44O9rn+CeLfoXW9kKrG4\ndCIwhoEX42rlRQungHfGutt2q7be8JRo1EEM4/lFcHK+/93v0rh2AJ6PFX6rJw6lhfjGUcHyf6kD\n6ceVwFbEpqKtiQ99L/l5a+CqkvtfT7S52YZYl/FRokromcCuxJbU3mYRScUzva6fTKyhWJFYb/xJ\nYEdiB8plwGpEPxUJLxO/lzycqEWgjUdrNOrG/wBPfzyWie2WOhgp22eAW6Pt5kapY2kxJR1df5E6\nlj68Row4lDKiGFbpfXpsRuw6uZNYC7EIMAr4LPCpBTyPMf+K2HWJdox3EqUtX8+Oty6x4+QbDLwY\nVysvEDsTsOgjp2HjBmRK/urDzAw674VxY+B6jSQ1hV8BP45PkA/T2u/UqfwRVnk9OgLp1yuVKACr\nQf5lOMvdP5c6HpmfPvDqJIbz8r+FG3KqFNoMLgD7cexH3A19CtbKFlHNslk6ukrj+Q/wcozOaxFo\ng1KiUV//gs7JcELqOGSB7oHc/tEbfF+iB6TURtbR9ZzUcUjTOgWKnbHcp/dMlzQIJRp1lFUKPR7O\nLsZyL2k8L4GNg8WKsQJvkdTxtLgOyK8UiUardXSV2psI/BssDydpEWjjUqJRf6eBz2nvbgSNahbY\naOh6N2plLJE6njaxbfQFuTp1HNJ0TgFysYv43NSxSP+UaNSZu0+DwunwxwK8kzocmasIbAw2LZqy\nLpc6njayAXR2qPWgDMwbwAlQKERfk3Kb4koCSjTS+D1My2lmupF8EHg6ih+smTqW9pNfEy6ltTq6\nSm39HpgdveV+kzoWWTAlGgm4+9OQuxx+m2/tWn3N4kvA9dEAYuPUsbSp8bFG4+LUcUhTeBM4DgrF\nGM2YkjoeWTAlGskUj4MnOjUzndrvgD9HPeftF3ZfqZno6MrfUschTeH3wKwooXFs6lhk4ZRopHMz\ndN4PRxY0qpHKpWDfi6mSPVCtjMSK68GNzNv6XKS3t4DfxWjGye6u7XtNQIlGIlkBrx/B3R3RVUHq\n6wHI7RUN0vZDtTIawYT26ugqlfkjMDNWb2s0o0ko0UjrGui8HQ5Ts7W6mgy2NQzLamUMTR2PAC3Z\n0VWqazrwmxjNOMXdX04dj5RHiUZC2ajGD+HhTrgwdThtYjYwCjpnR62MgXamkpry0VGg//HUgUhD\nOhF4O+aaf506FimfEo3E3P0W6LgWfpyPtU1SO0VgM7CpMV2yQup4ZD7jYhZLG7+ltxnAsTGacaq7\nT0odj5RPiUZDKPwInupUh+Na+zDwWCz8XCd1LNKnYVBYJqZPtERaSp0ETI+XxTGpY5GBUaLRANz9\nXshdAj/Jq+NDrXwTuCq2sG6aOhZZoM2jo+sdqeOQhvE2cCzkHU5394mp45GBUaLRMIqHw8QOOCN1\nIC3oJLA/wCiiKJc0trHq6CrzOpko0gUcnTQQqYip4V3jMMudAyM/Ds93wrDU4bSIq8B2g1WBg4HO\n1PFIWU6D4ZPgNaArdSyS1BvA2lB4E85w9y+mjkcGTiMaDcWPhNdyMRspg/co2B6wFNEoTUlG89gm\nCjOpbq4cDsyI7WJHpI5FKqNEo4FEDxQ/HY4qwNTU4TS5qWBbwNBCbGPVAFFz2TA6ump5dHu7n2gF\nX4Afq6dJ81Ki0Xh+AjNnwQ81p1WxOcAo6HgnkoylUscjlcivCZcQ2xql/RSBL0OhI8qqaJi3iSnR\naDDu/ioUfghnGNyZOpwmtSUwGfYFVkodi1Rse3gXdXRtV2cDd0NHHr7k7vnU8UjllGg0pj9D50Pw\npYKKeA3UnsADsDuwXupYZFBWg44h6ujajt4Evgt5g/Pc/abU8cjgKNFoQO5egPwX4cEOODV1OE3k\nB8AlsA2wRepYpBoK68MNgFp0tpefAm9Ct8P3Uscig6dEo0G5+51gZ8ChBXg1dThN4HSw38AGwE6p\nY5GqUUfXtvMw8EfwAvxUpcZbgxKNhuaHwayZcKgWhi7Q9WBfgJWJmRO9qluHOrq2FQe+AoUcPAuc\nkDoeqQ69JTcwd38NCofCXw1uTx1Og3oScrvACIcDUHWnFuSj4UHgidSBSM2dB9waC0C/4u7qx9Ai\nlGg0vtOg87/wpTxo4fW83gTbDIbkYxvroqnjkZpQR9e2MAP4FuRzcLG7X5M6HqkeJRoN7r2FoQ93\nRMV/CXlgI8i9DQcCS6eOR2pGHV3bwlHA1GgD/+3UsUh1qddJkzCzU2DY5+HRDlgzdTgNYCxwD+wH\nvD91LP14mZhpnpSdphMrG49cyOMKwN3AI0SBWAeGE/1aJgBLDCCGKcBtwPPATGJoYFmiwdzY7OdS\nxxO1vxdkKaIZbj3dDlwTZ1vX+aml9h4GNokFoEe4+y9SxyPVpe4PzeMH0P1hOHgFuLFj/k+IdvIJ\n4B74EI2bZADcTNQ0tAE85h3gLOAVIqFYK7v+DeC/wBjKTzSeAc4lSiwuSyQq7wIvEE1EniAazZWO\na24IzOrneM8TScjqZT5/NW0JndfCOa5Eo9XMAQ6AgsFTwG9TxyPVp0SjSbj7dDM7EG69Ib52tuv2\n8iOA86P451apY1mIVYHlid0wKxFr6BdWf+18omjEBGAc8yYB04BFBvD8VxJJxk7AtiXXzwDOJBKO\nh4jkpcfO/RzLgeOyy6MHEEO1dEB+RTjn5Xj1a81v6/gZ8CjgcIC7z04dj1Sf1mg0kaxC3nFwWDHG\n1dvNWWBHRcXPXVLHUoZtiYRhXWDxMu7/CDFqsCGwA/P/dS5F+QteZwOvE5/I2/a6bQli2sSJKZ1y\nPEskKKWjLPW2TVSM1CrB1nEncDS4w5Hu/t/U8UhtKNFoPocDT8H+hRh0bBe3gn0aVgT2pjVfufcT\n0yxjq3CsIZQ3Xllu4vIQEVuK0YweG6ijayuZCRwI+Q64D/h16nikdlrx7bqlxdBifv8YbPxZ6nDq\n5DmwHWHJrFbGkNTx1EABeJH4i1yZmD65HrgcuImB1+DOEUlBN3Brr9umA/cQy3zKSRy6ibUmEItI\nU8lBfo1osqaOrs3vh8DzUMjDJ9U0rbVpjUYTcvf/mtmRcPRRsLtFc49WNR1sjDOk2ziI8qYgmtE0\nYsfu4sAdwHW9br+RWJMykCmj3bLzG4iKV8vx3mLQJYhtwcuUcZzHs8etQKw5SWkHePeZaB9/UOJQ\npHJXAyfGxe+5u2qxtTiNaDSvX0PHvXBgHt5OHUuNFIHRYNONA4idE62qZwncLCLJ2AL4BtEn7iPE\nV4I7iZGIcnUSO0RGENtkHyN2ouSBNYCRZR7nQWLaZOMBPHetZB1dz0odh1RsCnOnTK4F/pQ6Hqk9\nJRpNKoYa8wfCi3n4fupwamR74AXYizRbKuvJS87fR4xGLAUMAzYhdoM4cMsAjnkNMc/QCXwKOIyo\nf7E9sVX2TPrfytpjJrEQ1ICNBvDcNVRYL3IxdXRtPkXgYCi+BW8V4CB3L1bjuGY2wcwuNLOXzOxd\nM3vDzB43swvM7KtmNpDqM9WI56dmVjSzg6twrOfNbGH71fp6XIeZPWFmd/a6foSZHW1m12bHnmlm\ns8zsMTP7rZkt18exVsjuc2Il/wYlGk3M3Z+C4nfgFGIyv5V8CrgttmY2yAdcTZWuOxnTx+096THA\nBAAAH0tJREFU180gamoszGvEFMwQ4JPECMYQYnRjArHg9E0W3kLnYeLTYS0GViislibE2flpo5AK\nnABcA7k8HOjuU6pxTDM7gsg9P0a8qi8nZmdmEW0W/0D9K+441StkW+lxvgSsw/wlAlcGDiXeVZ4H\nLiV+f0sB3wEeMrP3zROA+2TgVOAQM1tnoIEo0Wh+p0DucjiwEOPireCXwFmwGfNvzWxVw0suj+jj\n9i5gsezyzDKO9wLx9rQKfScIG2S3v7CQ4zTCbpPelgYWU0fXZnM/cGhWkcXdr67GMc1sU+KDdA7w\nMXffyN33cff93X1T4kP1+0QCUk9/JJKbi+v8vACY2RDgJ8BDffyuJwKbuftIdx/v7ge4+x7EuPHZ\nxGqu45jfscQS8qMGGo8SjSbn7g7Fg+CdF+Gj+YWPhTe688EOjzx8NwZWVbOZDSW+T0BUB+3NeW8d\nRzm7bnqOMXQBz9ffc/V4nSij3gWsX8Zz1pGPhgeAJ1MHImV5G9g3Vgc9BPyoiofei3iXuMDd5xvW\ndfdX3f04d6/rS8Xd33D3J9091QapfYmE4W+9b3D36X3VLMm65fb8v9muj9tfJpaW72lm5a7wApRo\ntAR3fwvyH4H/5eHz3rytp+6G3AHx57Ev7VdlfT3if93zfdw2kdgC20l5i2J7RkVeoe+Xw6Re9+vL\ng9n5+2m8LcXbq6Nrs3DgM+AvQHcePu7u71bx8COzp3htoA80s1XM7M/ZOoXZZjYlW+ex+UIe84ds\n7cMsM3vdzO4xsyPMbPGS+/W5RsPM1s5uu93MXsnWk0w0s7/1nq4YpM8Tk54DnWHszs7763h0LvFu\n8OmBHFSJRotw90eg+Gn4h8WoXbN5EWwcLFaMNQUDKbXdKrYiPj3vBl4quX4m8B/ie9smzJuATSL2\nCfbehrEO8XbwJlGPozTZmEp8LzGiCml/Hqbxpk16DIPC0uro2gyOAv4FVoh6GdUeWZhIvEr3Hsi3\nbDMbRSyJ/jwxDHwhMUD2MeB2M9u7j8eMI/4qvkqk/JcRVWqWJKZvSmvm9rdG4/NE0cVFib/0S4kP\n9U8C95jZoFekZQnPdsDT7v7KAB7XSRRncvrf2HVjdr77QGJSHY0W4u7nm9mW8J1vwabWx+hXg5oF\nNga65kRxhEZZdDhYTxLFtnqmfwrEn/DpJffZgdhlAjG68GHi7esvxPqKocRb6TtEVdSdej1HN5E4\n9C53NIx4K7iU2KnySPb4WUQSU8iet78tqy8StT1SlhxfmM3gxWvhLhq/7U27upC5KxGPcPeLavAU\n5xD7qVYDnjazi4gP//uI9Qn97Wo5h1jt82t3nzuVY2Z7Av8EzjSzW3sWrJrZUtk/Z0mi9sfxpQeL\n911eLiPei4FT3H2e1VFm9inir/4E5v8rH6htia8jC90Mb2ZnEAMOSxGr4pYm1mf8tK/7u/tzZjYV\nGGtmQ7LploXSiEbrORS4DfbKl/e6T60IbAw2DfYnpk1axSzif0FPm3iIpGNSyan3ws5NiA03awKv\nEut7FwPGA5+h7ykMo++1LBsDnyVGLfJEt9bJRMKxO/H77u8doGc0Y1Q/x24EW8Y3JU2fNKb/Ap+E\nYg4uAGrS+t3dnyPS8xeJcncHA6cRa0+nmtlJZrZC6WPMbDyxl+1FYsFk6fEuJurBLU789fQ4hJi0\nvKp3kpE97i53n1pGvHf3TjKy6/8G3AaMr8JW3J4xyHJGjw7OTnsQrR+fAu5ZyLbjJ4gx57J38mhE\no8W4e7eZ7QPTHoK9l4GbOhpvgr3UB4Gn4aPEh2srGUPfW1UXZo3sVO59e29eK7UKsd5loHZngIOj\nCXRCfiX4+8vxFUwdXRvHZODDkM/DI0X4TCxarw13vyHbcrk7UXFmLPFhOxz4MjGtMi7KAQDRFxli\nAWlf9SnOJhaZjgOOzq77ADEeeepg4zWzxYgP9jHECELPS3dFIq1fm1jrXKnliFgXuhHe3buymJYn\nRkKOAc4zszXc/dh+HtZz3LKnqjSi0YJiuC//UbjL4bupw1mALwLXR12ERqg6Kc1n61iGcm3qOGSu\nd4GPQuFVmJaHD7t7zbfCuXve3S9196+6+xbEh+CXiQ/FkcyteA7EN/f+ll1Tcv3KJdetmp0PqoaA\nme0IPEcsqvw+8DneG1Xo+ao12BGNns3yZe94cfcp2dTWTtnjjjSz/padT8/OF7SUfB5KNFqUu98J\n/o34+/pr6nD68Dvg1Jgq2D51LNK0NoTOnDq6NgoHDgG/F4pZkjFpoQ+qRRyxhfNUYsLRgAlm1t9m\n7/keXouYspGMC4hRjJ8S1WwWc/cOd+8Azuu56yCfqmfHyIATFnd/kVjVNRTob/dNTyJTdm0SJRqt\n7RSw02PLayN957sU7HuxyPDDNO4aAGl8OcivCRfRuh1/msnvgLPBsumSu1PHQ+y5glgc2fMN/GXi\nXae/xgZrZOelSdLE7HztQcQyjkgy/uXuP3f3J6Ib91zVWnb9KvHvW7rCx/esNemvhWVPxZ+ytxQr\n0WhhMS/qXwG/Gj5aiPVRqT0Aub1iWdXHab9aGVJ9O8Rw/SWp42hz/yZ6AAJHu3ujrNHt2dM1h/c+\nQHs6Bu1rZn19zTmIGNW4ueS6/yM+vL8wiFh6PqDnG+XJ1phsOohjl+qpgLPeQB9oZjne2674eD93\nW5/4k/tfucdVotHi3L0bivtA94OwSz6mB1OZDLY1DMtqZZQ7kCmyIOromtxjwMehYLE5+/B6Pa+Z\nHWVmx5rZfKMBZrYy8Gciabg0GlGCu99I7Ktag17ltLPtrXsS6xT+UnLT6USisquZfbOP59pyAWsa\nevTsAtmr9L5mNgI4g+ptzrid2MC+RV83mtl+fdXryLbwnkaMrNwVtZnmu89awDLA3eVubQUlGm3B\n3WdCfld48yXYKf9eYl9Ps8FGQefs+L4wfKEPEClbYb34ylmVLl0yIK8Du0F+DjxRjKJcVenIWqbF\nge8RNTQeN7OLzOxcM7uZ6Du8ObFl89u9HndgFvphZvaomZ1jZrcQtTIKwOdKm765+zRi/9YM4Hgz\ne8bMzjOzS83sSeLDfaVezzHPaIm730fMYa8GPJnFelEW5wpEkjZo7v42MWqzjpn1jgngQ0TjtKfN\n7OLs334j0fno08S30QP7OXzW0pArBhKTEo024e6vQn4nePEt2LVQ354oRWJUcCrsR/xJiVTT+DhT\nR9f6mgHsCoVJMCMPuyfo7XEUMT56NtENaDtgb6LGw13Ezo5NelfIzL6tb0p8g18se8y6xHKfbd39\nwt5P5O43EfvjTs6u+iiwDbEo8ifMvyOlr0WlHyG6Rr5KfOBvSuxA2YoFL64c6ALV04hEZ/9+bjuJ\n2D2yDbAPUTHnIaIO0+isPklfDiCmoebrobIgVsPtzdKAoo5/x83woUXgklx9SqnsBlwVf2LVmoUU\n6cV+A5vMjJKQUnuzgF2gcAfMLsB4d783dUwSsu6tLwBT3L2Saj59HXPl7JgXuPsBA3msRjTaTLwZ\nFPaCKx2+XIcGbN8ArootrEoypIZ8VCx3fmqh95TBmg3sAcU7YE4BdlaS0Viy9RM/B0ab2W5VOuwP\niGmlIwb6QCUabcjd/wP+OTjd4rVYKyeC/TFq9E1Y6J1FBkcdXetiDrA3FG+EfAF2c/fbU8ckfTqV\nKBe+oNrBZcnKuH8eONXdnx7w4zV10r7M7EfAL6OPz3wLqQfpKrDdYtnTQajYvdTH72H1abGaTeVZ\nqi8P7Ad+CRSKsSbjmtQxSePTiEZ7Oxo4Fr5FdIuolkfA9ohyMZ9ASYbUz+YxidwIlaJaTQE4GPxi\nKBZhLyUZUi4lGm0sa3T0Q+Do6InymyocdSrYWBhWiJGMYVU4pEi51NG1JorA58HPAxw+4e6Xp45J\nmocSjTaXJRs/Bn4Ra32OGcTR5gAbQcc7seGs7JY7IlXSCfkVo/dJd+pYWoQDXyM6Jjkc5O7/ShuR\nNBslGtKTbBwB/AwOI7Z5D/goRCG6KVFavK8yMSL1sA1MIwp4yeA4MdaZFY44pIFKi0sTUaIhQCQb\n7v5T4MioIDzQ3Sh7Ag/B7kTZG5FU1NG1ag4Hjo+LX3P3M5IGI01LiYbMw91/DhweO6J+Snl1Nr4P\nXArb0k91fZE6ykF+DXV0HQwHfgT8Kn78nruflDIeaW5KNGQ+7v5L4DD4GTGjsqBk4zSw38KGwAfq\nEp7Iwu0QRaUuTR1HE+oGPgV+dPz4PXf/XdKApOkp0ZA+ufsxwA/gF8SIRV99kq4D+yKsDHwMvZqk\ncaweHV3PTh1Hk8l6lxT/HrtZD1CSIdWggl2yQGb2deD30ZvnrwZDsluegNxGMDwPhwCLpotRpE//\ngtwj8AqwXOpYmsArwIeg8Gj0LvmIu1+fOiZpDfoOKgvk7n8E9oPzC/ChYjT8mwa2OQzJR60MJRnS\niCbEpJ86ui7c48BYyD8GUwuwjZIMqSYlGrJQ7v5PKO4EN8+CbQqwIeTejloZS6eOTqQfywCLDbCf\ndRu6DdgKCpPh6Txs4e4PpY5JWosSDSmLu98Eha3hf2/BK7Hwc5XUUYksmG8UbePV0bVvFwM7QvFt\nuCMfIxkTU8ckrUeJhpTN3R+B4jbkeJIbKOjdWxreDtHR9dzUcTSgk4C9gTxcWIAPuvu01DFJa1Ki\nIQPi7k9QZDPyXM05uLpXSUNbFApLzS2fLcTv4TCirLjD8cXoXTI7cVjSwpRoyIC5+9s4HwF+z5XA\nVfS9+1WkEWwGzwP3pI6jAcwGDgLPOhp9192/4+7665WaUqIhFXH3grt/G/gqd+GcQ5GZqaMS6cNW\n6ugK8AywJRT+AXlgf3c/LnVM0h5UR0MGzcw+SI7zWZQl2I9OVk0dkUgvf4alX4EpRNLRbi4GDobC\nbHgpDx9z9wdSxyTtQyMaMmjufi1FRjOL+zgT5w40IS6NZWt4g/br6DoH+A6wFzALLs3DxkoypN6U\naEhVuPtLFBmHcxxXA+fjaHmZNIqN2q+j60RgHBROiHLi3yrCPu7+Vuq4pP1o6kSqzsw+hnE2wxnK\nfnSyYuqIRICzYOizMBVYLHUsNfYfYH8ovA2v5mEvd78zdUzSvjSiIVXn7pfgjGE6j3EaRe5DUymS\n3vat39G1APwE2A2YAdfmYZSSDElNiYbUhLs/Q5EtKXIalwMX48xJHZW0tTWgo6t1O7pOBj4AhV+C\nO/yoALu7++up4xLR1InUnJl9EuM0lqGTfelk+dQRSdtq0Y6uNwIfh/w0eDMfazFuSh2TSA+NaEjN\nufvfcTbjDZ7mzzg3EWO8IvU2PmbxLkgdR5XMBg4HdgR/A27Nw0ZKMqTRKNGQunD3xyiyCUWO4QaK\nnEqByamjkrazLC3T0fU2YDTkfwV5hyMLsJO7T0kdl0hvmjqRujOzzclxFs76bI8xjvasoiRpXAXc\nFR1d10kdSwVmEL1K/gR0wD15+Iy7P5o4LJF+aURD6s7d76XIGJyjuIkCfybPy6mjkraxffN2dL0S\nWB/yJ8Nsh2/mYWslGdLoNKIhSZnZGHKcjbMh22HsgEY3pPZ+D2tNg6cBSx1LGaYC3wQ/F6wD/q8A\nh7j786njEimHRjQkKXd/gCKb4vyUW8hzCnkmpY5KWt6m8Cxwb+o4FsKBfwDrQuH8mDX5VAF2VpIh\nzUSJhiTn7t3u/nNgU97gUU7DuRpUwlxqZuvG7+g6EfgwFA8A3oKLCrCuu5/lGoaWJqOpE2koZtYJ\nfA/jSIbSyQfoZBNiUl2kmk6BZSZHoatGmq0rAn8GvgeFOfB6Hr7g7q1c0FRanEY0pKG4e97dj8FZ\nh3c4hyuAU8jzbOrIpOVsDa8D16WOo8S1wCZQ+AowC87IxyiGkgxpako0pCG5+yR3/zQwlte5l7OA\ncymigspSLaMap6Prf4GdoLAz8FgsHdnW3b+obqvSCjR1Ig3PzAzYlxzHASsxNtudMixxYNL8/gZD\nn0vX0fU54Mfg/wDrhKfz8H3gUq3DkFaiEQ1peB4uoMj7KHI4d/EOv6fA3aiUuQxO1tH1sjo/7VTg\nW8C64P+MH7+Qh/e7+yVKMqTVaERDmo6ZrQj8Evg0y1BgJzpZD6XNUpGOX8LO3VEMq9ZmAScAv4LC\nbHi3AL8CTnD3mXV4epEklGhI0zKzzchxHEW2Z1ny7EAnG6KEQwbmn5B7NHafjKzRU+SBvwCHQ/41\nwOEk4Jfu/lqNnlKkYegtWZqWu9/nBd8BGMfr3MCFwB/J8wCaUpHyja9dR9cicDGwIeS/ALwG/3JY\nz92/pSRD2oVGNKRlmNkWGIfjfIQlybM9nYyhsYokSEOyY2HzWXB3lY43i+gQ+1vIPwudHXBDAb7v\n7vdV6SlEmoYSDWk5ZrYxxo9x9mFxCmxHJ5sCQ1JHJg3rSuDu6H2y9iAO8woxJ3IiFKbHiPGFDse5\n+x3VCFOkGSnRkJZlZu8nOmofyDCKbEcnmwFDEwcmjWcmdPwGjgR+UsHDHwJ+F03PcJhdiOKef3D3\n56oap0gTUqIhLc/M1gZ+CHyaTozRdLA5sFLiwKSxDLCjaxG4GvgtFK+HXCe8koffAaer0JbIe5Ro\nSNsws5WBz5HjyxRZgRXJswWdbISmVQRuAa6De4DNF3C32UQ10d9A/kno7IT/5uFY4EJ3765HqCLN\nRImGtJ2scdtuGF/B2ZkhFBmTjXIslzo6SSYPnb+ArwHH93Hz88QW1RMh/wZ05OCyIvwWuE1FtkT6\np0RD2pqZrQkcQo4vUmRpVqXAFnSwAdqt0o5OgWUnx6LOTmAG8C/gL1C4BTo64J0CnEkU2Xo6aawi\nTUKJhghgZkOAj5HjqxTZnqHkGZMVAFuF8ibtpfk9CFwc5TofBb8QfDZYB9yYJRgXq4qnyMAo0RDp\nxczWB75AjoMpsgyLk2cjOtmASDpU5q71ODAJeBg674pKnl3wbDecAZzt7hPTBijSvJRoiPTDzDqA\n7YjOsftRZFkWJ8+GJSMdSjqalwNTgEeAh8gznU5yvE6RfxEFPa/R2guRwVOiIVKGLOnYlveSjpEs\nVjLSsSpKOprBHGJV59PAE+R5i05yTKfI+cB5wE3urgL2IlWkRENkgMwsB2xDJB2foMhyLEqe9ehk\nLWANYImkIUoPJ5qwPw08SZEXgCI5ckyiyOXAFcC17j4nZZgirUyJhsggZEnH1sA+5NiVIusBsAx5\n1qaTNYnEY1i6GNvOu8BzvDdqMYNOjG7gJpx/A/8BntC0iEh9KNEQqSIzWwGYAOxIjl0osiqGszyF\nuYnHaqhAWDXNIhZyTgKeo8BEchQxcjxPkSuAq4Ab3X1W0jhF2pQSDZEayup0TAA+QI6dKbIsOZyV\nKbIaHaxElEIfgbbQlqMbmMx7icVEunmTLgByTMe5Fec/wFWqcyHSGJRoiNSJmRmwPvABYsRjG4os\nD8Ai5FmZHCuRYwVgeWBpoCNZuOkVgTeIhOIl4CXyTKGDIobRjfEgRW4H7iI6vD+j6RCRxqNEQyQh\nM1se2Cw7bUGOLSlmhdBzFFmWIivSyfLAMsBwYvSjVTrQFonym28Ar5ecT6WbaXRQzPbydPAsBW7j\nvaTiQS3gFGkOSjREGoyZLQOMAkYDo8ixKc4GeEl6MYQCwymyNF1zk4/S88VojKmYbmINRc/pLd5L\nKF7LkonC3I3BRTqYRJH/4TwJPAX8D7jX3acliF5EqkCJhkgTyHa3rAisPs/JWIMca1Nk1XkSkQ6K\nLEaBocBQcgylg0VggachxHZQJ0YaFnZeJHZ4lCYSM3FmkmcmMJsO8vNVF3FyTMZ5HOcJIpnoSSqe\n0yiFSOtRoiHSArL1H0szbyKyIrDk3FOOERgjgCVxlsBZHM8WUlauSI43MabivEqRKUTliv5OU9z9\nnUE+p4g0ESUaIm3MzBYhyostCSxOjFcUSk75Xj/3Ps3SAkwRWRAlGiIiIlIz6s4gIiIiNaNEQ0RE\nRGpGiYaIiIjUjBKNFmdmE8zsQjN7yczeNbM3zOxxM7vAzL5qZk3fZ9TMVjezopldX4NjX29mL5pZ\nV8l1S5nZIWb2LzP7n5m9bWbTzexOM/uGmXX2cZyhZvaymV1R7RhFRBqZEo0WZmZHANcBHwPeBC4H\nriYqHuwJ/AF4f7IAG5yZ7Q6MB37p7t0lN30H+DPxe50JXEZUrBwNnABcZ2bz1O5099nAscCuZja+\n5sGLiDQI7TppUWa2KXAPUZtxX3e/vNftywGfBK5w9ycThFg1ZrY60Rj8RnffsYrHfZDoOrKKu+dL\nrv8hUX/zJHd/qeT6tYnEblXgaHc/vNfxhgIvEy3Kt65WnCIijUwjGq1rL6II9QW9kwwAd3/V3Y9r\n9iSjVsxsW6IM+HmlSQaAux/j7oeVJhnZ9c8APyR+7/v3PmY2qnEhMNbMNq5Z8CIiDUSJRusaSRRf\nem2gDzSzYWZ2mJndb2YzstMdZnZwP/ffzsxONLMHszUgs7K1C0eb2fA+7r9DtqbiTDNb3sxON7OJ\nZtZtZt/odd/9zexaM5tqZu+Y2XNmdr6Z9Tlyka2FOMbMnjez2Wb2lJn9YKC/A+DzxO/vvAE+7sHs\nfKV+bj+XSES+UEFMIiJNZ75Fa9IyJhIfaHub2THuXlbCYWYjgf8jvs2/AtyYHWcb4K9mtpm7f7PX\nw35DrE94KHvsUGBT4FBgdzPbyt1n9fF0I4npnQ7g1uxxs7I4csSH/D5ER43bgCnEtMRuQBfQe/Hn\nEOAaYt3JjURrsR2AY8xscXc/opzfQWZX4B2iU+hArJWdT+7n9tuJ6azdga8O8NgiIs3H3XVqwROw\nJrFQsUj0zPwL8DlgDJBbwOP+TZSW/h3QVXL9SOJDtwDs3OsxuwBL9LquCzg5u//hvW7bIYurAPwT\nGNJHHIdn93kIWK3XbUsA40p+Xr3keNcBi5XctinxwT4DWLTM39162fFuqeD3fm0Wx/ELuM892X1W\nT/060UknnXSq9UlTJy3K3Z8DPgy8SPSwOBg4DbgfmGpmJ5nZCqWPydYN7Arc7e7f9ZKdFh4jIl8g\nRje+3Ou5rnb3Gb2u6wa+TXygfrSfMN8FvuG9OnZmW0m/Q0xdfNbdX+x17BnufksfxysCX3T3mSX3\nvR+4ClgU2LyfOHobnZ0PaP2KmX0J+AAwDfj1Au76eHY+ZiDHFxFpRpo6aWHufoOZrUMM0+8MjCU+\nRIcTycLeZjbO3Z/KHrIz8eF+aT/He8DM3s6OMw8zWwnYA1ifaNCVI5KSOcD7+gnxfnd/pY/rNwdG\nAA+4+73l/FszL7j7031c/yTxO1ixzOMsl52/Ue4Tm9k4YmtrkUiO+ps6KT3uyHKPLyLSrJRotDiP\nHROXZifMbEngE8CviA+6E4mpD4A1iOTgV2b2qwUcdpHSH8zsO8DR9P16MiJ56cuL/Vy/anb+zAJi\n6MtL/VzfM9qySD+399azgHXGAu+VMbONiN9vF/B1d79sIQ+Znp2PKDMeEZGmpUSjzbj7dOBUM3uF\n+HCcYGZDPbZe9kyl3cKCP+TnJg5mtiXwW6Ig2DeAm4DJPdMuZjYJWKGvgwCzB/Nv6UOxSsd5Kztf\naNVUM1uTWIA6HDjS3f9UxvF7Epk3KwtPRKR5KNFoXz07NjqIb9aTeW9E4BJ3P77M4+xJJB4/cve/\nl96QFajqL8lYkInZ+doVPLYaXs3Ol17QncxsRWLx5/LACe7+izKPv1R2PuCtxyIizUaLQdtXz7qJ\nOcDU7PK12fmeAzhOz4fmpD5u+zgxdTJQ9xHf9jc2s3IXcFZTTy2M9fq7g5mNIMq5rwmc6e7fHcDx\ne8q+P1BZeCIizUOJRosys6PM7FgzW6uP21YmenU4cGm2jgN3v5tINnoKcM03dWBmo81sl5KrniSS\nic+VNhMzsw2AYyqJPduFcnx23DPMbLVeMSxpZttXcuwyn/9JYlRjTFbPYx5mNgy4EtgQuIABFN8y\ns0WIGiUT3f2F6kQsItK4NHXSuhYHvgl8z8yeBB4j1kSsAmxJ/L9/itiCWuqTxHbQLwMHmNkDRH+O\n4cSOlVWJ3RVXZ/f/C/BdYsfJE2Z2DzHlsANwcfZc8yQKZfoVsf3zY8CTZnYL8eG/KlEb4xrg5gEc\nb6AjK1cCnyLiv6OP2LYC8sT23TPN5j+8u3+mj+NuSywaVRdXEWkLSjRa11FEYahdgI2B7YhkYTrR\nafQS4GR3f6f0Qe7+mpltAxxC7E4ZA2xNVOV8hkgyzi+5/xvZ9MavieRiD6LB2Y+Jol/P0veuE+/n\n+p7jFojttwcBnwU2IyqHvkJ0Sz1lIMdbyG19OQ34NHAA8ycaI7LjddBHT5OS5+sr0Tgwu+30AcYj\nItKU1L1VpB9mdj+wMtG9tXth9y/jeD3dWx93920GezwRkWagNRoi/fsxUWvkkCod78vEqNJhVTqe\niEjD04iGyAKY2XXEDp21BzOqkY1mPAvc5+57VCs+EZFGp0RDREREakZTJyIiIlIzSjRERESkZpRo\niIiISM0o0RAREZGaUaIhIiIiNaNEQ0RERGpGiYaIiIjUjBINERERqRklGiIiIlIzSjRERESkZpRo\niIiISM0o0RAREZGaUaIhIiIiNaNEQ0RERGpGiYaIiIjUjBINERERqRklGiIiIlIzSjRERESkZpRo\niIiISM0o0RAREZGaUaIhIiIiNaNEQ0RERGpGiYaIiIjUjBINERERqRklGiIiIlIzSjRERESkZpRo\niIiISM0o0RAREZGaUaIhIiIiNaNEQ0RERGpGiYaIiIjUzP8Dwbq8F6+1qj0AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe6ec2bdf60>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "page_views_by_traffic_source_pandas.plot.pie(y='count', figsize=(6, 6), labels=['Internal (1)', 'Search (2)', 'Social (3)'], \n",
    "              title='Page Views by Traffic Source', autopct='%.2f', fontsize=15)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Users visiting a page more than once"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Surprisingly, there are **47937415 repeated combinations of users visiting a page more than once**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "47937415"
      ]
     },
     "execution_count": 187,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "page_views_by_user_doc_df = page_views_df.groupBy(['uuid','document_id']).count().alias('count').filter('count > 1')\n",
    "page_views_by_user_doc_df.count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+--------------+-----------+-----+\n",
      "|          uuid|document_id|count|\n",
      "+--------------+-----------+-----+\n",
      "|6a5618eba26990|      25529|    2|\n",
      "|571016386ffee7|      25792|    6|\n",
      "|6a91157d820e37|      25792|    6|\n",
      "|2aec87c28309c5|      25792|    2|\n",
      "|b97961739d2f29|      25792|    2|\n",
      "|a743b03f2b8ddc|      25792|    3|\n",
      "|8d87becfb35857|      25792|    4|\n",
      "|6d82e412aa0f0d|      25792|    8|\n",
      "|ad45fc764587b0|      25792|    6|\n",
      "|29ebac4412da91|      25792|    2|\n",
      "+--------------+-----------+-----+\n",
      "only showing top 10 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "page_views_by_user_doc_df.show(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If we sample one of those repeated combinations, we observe a user which has visited the same page in 6 different days..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+--------------+-----------+---------+--------+------------+--------------+\n",
      "|          uuid|document_id|timestamp|platform|geo_location|traffic_source|\n",
      "+--------------+-----------+---------+--------+------------+--------------+\n",
      "|571016386ffee7|      25792| 49576389|       1|   US>MO>616|             1|\n",
      "|571016386ffee7|      25792|120316675|       1|   US>MO>616|             1|\n",
      "|571016386ffee7|      25792|576124810|       1|   US>MO>616|             1|\n",
      "|571016386ffee7|      25792|639792203|       1|   US>MO>616|             1|\n",
      "|571016386ffee7|      25792|724651242|       1|   US>MO>616|             1|\n",
      "|571016386ffee7|      25792|817713829|       1|   US>MO>616|             1|\n",
      "+--------------+-----------+---------+--------+------------+--------------+\n",
      "\n"
     ]
    }
   ],
   "source": [
    "page_views_df.filter('uuid=\"571016386ffee7\" and document_id=25792').show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2016-06-14 17:46:16\n",
      "2016-06-15 13:25:16\n",
      "2016-06-20 20:02:04\n",
      "2016-06-21 13:43:12\n",
      "2016-06-22 13:17:31\n",
      "2016-06-23 15:08:33\n"
     ]
    }
   ],
   "source": [
    "for ts in [49576389, 120316675, 576124810, 639792203, 724651242, 817713829]:\n",
    "    print(convert_odd_timestamp(ts))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Exploring relationship with Events"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading events.csv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this section we explore Page Views and Events relationship."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "events_schema = StructType(\n",
    "                    [StructField(\"display_id\", IntegerType(), True),\n",
    "                    StructField(\"uuid\", StringType(), True),\n",
    "                    StructField(\"document_id\", IntegerType(), True),\n",
    "                    StructField(\"timestamp\", IntegerType(), True),\n",
    "                    StructField(\"platform\", IntegerType(), True),\n",
    "                    StructField(\"geo_location\", StringType(), True)]\n",
    "                    )\n",
    "\n",
    "events_df = spark.read \\\n",
    "            .schema(events_schema) \\\n",
    "            .options(header='true', inferschema='false', nullValue='\\\\N') \\\n",
    "            .csv(\"gs://sc-ml-hdfs-prd/tmp/outbrain/events.csv\").cache()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "events_df.rdd.getNumPartitions()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+----------+--------------+-----------+---------+--------+------------+\n",
      "|display_id|          uuid|document_id|timestamp|platform|geo_location|\n",
      "+----------+--------------+-----------+---------+--------+------------+\n",
      "|         1|cb8c55702adb93|     379743|       61|       3|   US>SC>519|\n",
      "|         2|79a85fa78311b9|    1794259|       81|       2|   US>CA>807|\n",
      "|         3|822932ce3d8757|    1179111|      182|       2|   US>MI>505|\n",
      "|         4|85281d0a49f7ac|    1777797|      234|       2|   US>WV>564|\n",
      "|         5|8d0daef4bf5b56|     252458|      338|       2|       SG>00|\n",
      "|         6|7765b4faae4ad4|    1773517|      395|       3|   US>OH>510|\n",
      "|         7|2cc3f6457d16da|    1149661|      602|       2|   US>MT>762|\n",
      "|         8|166fc654d73c98|    1330329|      638|       2|   US>PA>566|\n",
      "|         9|9dddccf70f6067|    1772126|      667|       1|   US>FL>528|\n",
      "|        10|b09a0e92aa4d17|     157455|      693|       1|          US|\n",
      "|        11|602e210c5831e5|    1773230|      710|       3|   US>IL>675|\n",
      "|        12|6fa993bd0e0157|     892701|      718|       1|   US>TX>612|\n",
      "|        13|7355615832b3af|    1778533|      739|       1|   US>AZ>753|\n",
      "|        14|daef797fc210a2|    1759953|      798|       3|   US>NC>560|\n",
      "|        15|24c64dc30891c0|    1777136|     1000|       2|          GB|\n",
      "|        16|30c0ad12b36375|    1727882|     1033|       1|   US>FL>561|\n",
      "|        17|c80c06d718ba65|    1667725|     1044|       2|   US>MI>540|\n",
      "|        18|eb58e66b4f6bb0|     429183|     1175|       2|   US>OR>820|\n",
      "|        19|c419799a427c72|    1155107|     1202|       3|   US>HI>744|\n",
      "|        20|650e3b5699738b|    1642724|     1282|       2|   US>MI>505|\n",
      "+----------+--------------+-----------+---------+--------+------------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "events_df.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Events count**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 4 ms, sys: 4 ms, total: 8 ms\n",
      "Wall time: 27.8 s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "23120126"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%time events_count = events_df.count()\n",
    "events_count"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Distinct users in events**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 12 ms, sys: 0 ns, total: 12 ms\n",
      "Wall time: 5.97 s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "19794967"
      ]
     },
     "execution_count": 220,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%time distinct_events_users = events_df.agg(F.countDistinct(\"uuid\").alias('count')).collect()[0]['count']\n",
    "distinct_events_users"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Average events by user**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.1679800223966021"
      ]
     },
     "execution_count": 191,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "avg_events_by_user = events_count / float(distinct_events_users)\n",
    "avg_events_by_user"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Creating a dummy columns on Page Views and Events datasets, which will be useful to filter matches in outer joins**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "When joining DataFrames, the resulting DataFrame keeps only one column for each of the join fields. Thus, we've included dummy columns (dummyPageView, dummyEvent) on each dataset to be able to identify rows where there is a match (both dummy columns filled). Otherwise, only one of dummy columns will be filled and the other will be None (because its an outer join)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "page_views_df = page_views_df.withColumn('dummyPageView', F.lit(1))\n",
    "events_df = events_df.withColumn('dummyEvent', F.lit(1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Performing a join between page views and events, on fields uuid and document_id"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here we perform an outer join between page_views.csv and events.csv datasets, on uuid and document_id fields. The objective is to explore rows matches, and see if events dataset subset is a subset of page views.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "page_views_events_joined_by_uuid_document_df = page_views_df.alias('page_views').join(events_df.alias('events'), on=['uuid','document_id'], how='outer')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Count of events without a match on page views (0). We can see that all combinations of (uuid, document_id) events are represented on page views dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 164 ms, sys: 16 ms, total: 180 ms\n",
      "Wall time: 16min 20s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 194,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%time page_views_events_joined_by_uuid_document_df.filter('dummyEvent is not null and dummyPageView is null').count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the matches (uuid, document_id) table below, we can see that for (uuid=10012c47f7317b, document_id=2897953) there are two records on Page Views (left side), but only one record in Events (right side), repeated with display_id=21485757.  \n",
    "As the fields platform (3) and geo_location (US>CO>751) are also equal, this suggests that **timestamp is important when joining page views and events, as the user might visit a page more than once, at different times**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+--------------+-----------+----------+--------+------------+--------------+----------+----------+--------+------------+\n",
      "|          uuid|document_id| timestamp|platform|geo_location|traffic_source|display_id| timestamp|platform|geo_location|\n",
      "+--------------+-----------+----------+--------+------------+--------------+----------+----------+--------+------------+\n",
      "|100025359c1e4b|     542474| 239288958|       3|          ZA|             1|   3749002| 239293893|       3|          ZA|\n",
      "|10003a00054aa3|     573399| 252875434|       2|   US>MA>506|             1|   3972594| 252875434|       2|   US>MA>506|\n",
      "|10012c47f7317b|    2897953|1212693052|       3|   US>CO>751|             1|  21485757|1207434141|       3|   US>CO>751|\n",
      "|10012c47f7317b|    2897953|1207434141|       3|   US>CO>751|             1|  21485757|1207434141|       3|   US>CO>751|\n",
      "|10013cf2933dd1|    2287664| 636931394|       1|   US>MO>609|             1|   9359025| 636931394|       1|   US>MO>609|\n",
      "|10031b0b19b6a6|    1984874| 339728683|       2|   US>FL>539|             3|   5248084| 339728683|       2|   US>FL>539|\n",
      "|1003a343464da3|     641298| 732411861|       1|   US>MI>563|             1|  18800263| 732411861|       1|   US>MI>563|\n",
      "|1003b0110e0e82|    2631247| 853805550|       1|   US>GA>503|             1|  19160961| 853806655|       1|   US>GA>503|\n",
      "|100412cf7c613f|    1305530| 191532005|       2|       GB>H7|             1|   2901780| 191532467|       2|       GB>H7|\n",
      "|100655b134ceb2|    1254803| 936285398|       2|   US>VA>511|             2|  14255514| 936285398|       2|   US>VA>511|\n",
      "|10068ce0a48328|      13319| 664331049|       3|       CA>BC|             2|   9910263| 664331472|       3|       CA>BC|\n",
      "|1006a35734679c|    2729051|1286419339|       3|       MY>12|             1|  22943569|1286419339|       3|       MY>12|\n",
      "|100708c823ce81|    2241722| 496443973|       3|   US>VA>511|             1|   7233639| 496444493|       3|   US>VA>511|\n",
      "|10083c097a3264|    1146123| 754677444|       1|       CM>05|             1|  11371820| 754677444|       1|       CM>05|\n",
      "|10087ce9350dd8|    1413545| 425323913|       1|   US>CA>868|             1|   6357644| 425324428|       1|   US>CA>868|\n",
      "|1009ae5c722da8|    1882802| 161254964|       3|   US>HI>744|             1|   2555949| 161254964|       3|   US>HI>744|\n",
      "|1009b2327213f2|    1654727| 440543174|       2|       GB>K2|             1|   6537695| 440543174|       2|       GB>K2|\n",
      "|100ac27d807d62|    2262848| 560815264|       2|   US>PA>504|             2|   8140881| 560815264|       2|   US>PA>504|\n",
      "|100ad94f7f9d33|    1775106|1059619119|       2|       GB>J8|             1|  15837873|1059619119|       2|       GB>J8|\n",
      "|100bb588e24a42|    1681151|1078448905|       2|   US>FL>534|             1|  19724264|1078448905|       2|   US>FL>534|\n",
      "|100cbbff0a5b1f|    2036709| 222468828|       1|   US>GA>524|             1|   3409755| 222468828|       1|   US>GA>524|\n",
      "|100dad661b3f71|    1816063|  53030746|       1|   US>NE>652|             1|    813537|  53030746|       1|   US>NE>652|\n",
      "|100df0d3f39d92|    1724507|1120872646|       2|   US>SD>764|             1|  16837394|1120872646|       2|   US>SD>764|\n",
      "|100e0331b91072|    2277971| 548662907|       1|       AU>01|             1|  18268023| 548662907|       1|       AU>01|\n",
      "|100f877c435da9|    2660718| 906243636|       2|       DE>01|             2|  19291766| 906243636|       2|       DE>01|\n",
      "|100f935eb09103|    1774580| 235625910|       3|   US>NY>501|             2|   3679555| 235625910|       3|   US>NY>501|\n",
      "|10112449d036f6|    2712026| 931010209|       2|   US>ME>537|             1|  14173605| 931010209|       2|   US>ME>537|\n",
      "|1011c325276473|    1288204|1274862827|       2|       GB>Z1|             1|  22729586|1274865496|       2|       GB>Z1|\n",
      "|1012ebf9e2cdb1|    2596490| 827460594|       2|   US>TX>618|             1|  12466705| 827460594|       2|   US>TX>618|\n",
      "|10133ad6e6c212|    1880985| 112799039|       3|   US>CA>803|             2|   1684481| 112799039|       3|   US>CA>803|\n",
      "|10147ea2e71076|    2368293| 656055289|       2|   US>TX>625|             1|   9735737| 656079705|       2|   US>TX>625|\n",
      "|101488b7e89106|     394689|  58421819|       1|   US>CA>803|             1|   3584446| 231051790|       1|   US>CA>803|\n",
      "|101488b7e89106|     394689| 230937874|       1|   US>CA>803|             1|   3584446| 231051790|       1|   US>CA>803|\n",
      "|1014e6d2391b91|      60164| 367627229|       1|          US|             1|  15847538|1060602761|       1|   US>PA>566|\n",
      "|1014e6d2391b91|      60164| 535909081|       1|          US|             1|  15847538|1060602761|       1|   US>PA>566|\n",
      "|1014e6d2391b91|      60164|1060602338|       1|   US>PA>566|             1|  15847538|1060602761|       1|   US>PA>566|\n",
      "|1014e6d2391b91|      60164| 884642103|       1|   US>PA>566|             1|  15847538|1060602761|       1|   US>PA>566|\n",
      "|1014e6d2391b91|      60164| 627297766|       1|          US|             1|  15847538|1060602761|       1|   US>PA>566|\n",
      "|1014e6d2391b91|      60164|1143583425|       1|   US>PA>566|             1|  15847538|1060602761|       1|   US>PA>566|\n",
      "|1014e6d2391b91|      60164| 456384939|       1|          US|             1|  15847538|1060602761|       1|   US>PA>566|\n",
      "|1014e6d2391b91|      60164| 803800862|       1|          US|             1|  15847538|1060602761|       1|   US>PA>566|\n",
      "|1014e6d2391b91|      60164| 715376948|       1|          US|             1|  15847538|1060602761|       1|   US>PA>566|\n",
      "|1014e6d2391b91|      60164| 973340936|       1|   US>PA>566|             1|  15847538|1060602761|       1|   US>PA>566|\n",
      "|1015a136bb89d5|    1179111| 902357297|       2|       NL>03|             3|  13615396| 902357297|       2|       NL>03|\n",
      "|1015c21198bc2e|    1937353| 135205207|       1|   US>WA>819|             1|   2086130| 135205207|       1|   US>WA>819|\n",
      "|101637a37acd62|    2108111| 784337687|       1|   US>CA>803|             1|  11819050| 784337687|       1|   US>CA>803|\n",
      "|10168eccef2fb1|    2821957|1129824000|       1|       CA>QC|             3|  19943552|1129824000|       1|       CA>QC|\n",
      "|1016ae461f7194|    1772623|   5703583|       2|   US>IN>529|             3|     69222|   5703583|       2|   US>IN>529|\n",
      "|10176291a86e41|      65310| 806301277|       2|   US>SC>507|             1|  12027683| 806301277|       2|   US>SC>507|\n",
      "|101784c0d6c934|      25626|1242588053|       1|       IN>25|             2|  21977209|1242597080|       1|       IN>25|\n",
      "|101952f1e3f865|    1555928| 510520372|       2|       CA>BC|             3|   7443278| 510520372|       2|       CA>BC|\n",
      "|1019b336bf1720|    1943384| 879315766|       3|   US>HI>744|             1|  13273317| 879315766|       3|   US>HI>744|\n",
      "|101ac58c714ad4|    2103573| 571927409|       1|   US>NY>532|             1|   8384045| 571927968|       1|   US>NY>532|\n",
      "|101b1908d81b7e|     199827| 596902886|       2|   US>FL>548|             1|   8881796| 596902886|       2|   US>FL>548|\n",
      "|101bc99bbae8ea|    2799582|1088277253|       1|   US>MA>506|             1|  19753569|1088277673|       1|   US>MA>506|\n",
      "|101c55f5e5907f|    2322438| 572692392|       1|   US>SD>725|             1|   8400704| 572692392|       1|   US>SD>725|\n",
      "|101d1e6fd80dc8|    1686705|  78570583|       2|   US>AZ>753|             1|   1311884|  78570583|       2|   US>AZ>753|\n",
      "|101ef60b51f46a|    2846658|1269471951|       2|   US>TX>618|             3|  22609092|1269471951|       2|   US>TX>618|\n",
      "|101f529df49818|    2282903| 693146050|       1|       CA>QC|             1|  10396979| 693146207|       1|       CA>QC|\n",
      "|101fa3da92ad08|    2330549|1074962816|       2|   US>NC>517|             1|  19712216|1074962816|       2|   US>NC>517|\n",
      "|10202311c8dd2b|    2520315| 759955771|       2|   US>CA>862|             1|  11463912| 759956029|       2|   US>CA>862|\n",
      "|10207c31492ec9|    2127063| 460032874|       3|       CA>QC|             3|   6687190| 460033125|       3|       CA>QC|\n",
      "|1020e22d365bc4|     782847| 805810041|       2|          GB|             1|  12019463| 805810041|       2|          GB|\n",
      "|102241b7bf0451|    2536007| 809978523|       1|   US>TX>618|             1|  12093379| 809979134|       1|   US>TX>618|\n",
      "|102242dfddf5bb|    2102553|1044872937|       2|   US>WI>658|             1|  15704941|1044872937|       2|   US>WI>658|\n",
      "|102256fa0a8a58|    2257328| 506161289|       3|   US>MN>613|             1|   7376382| 506161289|       3|   US>MN>613|\n",
      "|102256fa0a8a58|    2257328| 678105901|       3|   US>MN>613|             1|   7376382| 506161289|       3|   US>MN>613|\n",
      "|1022ecd82be4d8|    1475541|1240739646|       2|   US>NC>517|             1|  21937421|1240739646|       2|   US>NC>517|\n",
      "|10231e5b453f32|     759612| 231453078|       1|       GB>H9|             1|  14090843| 926182688|       1|       GB>H9|\n",
      "|10231e5b453f32|     759612| 926182688|       1|       GB>H9|             1|  14090843| 926182688|       1|       GB>H9|\n",
      "|1023dae1a4a022|    1882701|  81969736|       2|   US>CA>803|             1|  17117514|  81969736|       2|   US>CA>803|\n",
      "|1023de703442de|     408884|  45125771|       2|   US>VA>511|             2|    635292|  45125771|       2|   US>VA>511|\n",
      "|10268a4b15a372|    1268709| 625408689|       2|   US>CA>804|             1|   7236614| 496634431|       2|   US>CA>803|\n",
      "|10268a4b15a372|    1268709| 496634431|       2|   US>CA>803|             2|   7236614| 496634431|       2|   US>CA>803|\n",
      "|1028f346ff4754|    1062653| 766001791|       1|   US>OR>801|             1|  11561259| 766001791|       1|   US>OR>801|\n",
      "|102a70268a925c|    1446303| 175581786|       2|   US>AZ>753|             1|   2768538| 175581786|       2|   US>AZ>753|\n",
      "|102a75d0c348bb|    1503002| 528782699|       2|       ZA>11|             1|   7689771| 528783048|       2|       ZA>11|\n",
      "|102ad4aafb8985|    1828395|  72107570|       2|       SG>00|             3|  17087055|  72107570|       2|       SG>00|\n",
      "|102b8762e08042|    2029005| 287238154|       3|       US>TX|             1|   4352364| 287239533|       3|       US>TX|\n",
      "|102db96cee239d|    2083622| 366478863|       2|       US>FL|             3|   5523732| 366478863|       2|       US>FL|\n",
      "|102e3b3d1cfce4|    2608487| 891263688|       2|          US|             3|  19238717| 891263688|       2|          US|\n",
      "|102e7278ff22a2|    2487723| 747409325|       2|   US>PA>504|             1|  11226362| 747409325|       2|   US>PA>504|\n",
      "|102e9d5b3d4267|    1777210|  85440672|       1|       CA>BC|             1|   1447701|  86598556|       1|       CA>BC|\n",
      "|102e9d5b3d4267|    1777210|  86598556|       1|       CA>BC|             1|   1447701|  86598556|       1|       CA>BC|\n",
      "|102f1c6cecfb3c|    1783789| 499373705|       1|   US>CA>807|             1|   7332699| 503077441|       1|   US>CA>807|\n",
      "|102fc40b24a671|    1419298| 985943471|       1|   US>OH>535|             1|  19497722| 986010393|       1|   US>OH>535|\n",
      "|103034b1ef41eb|    1630584| 583219637|       3|   US>GA>524|             1|  18398359| 583219637|       3|   US>GA>524|\n",
      "|10306dc4746285|     945451| 989355484|       1|   US>CA>807|             1|  14926270| 989356748|       1|   US>CA>807|\n",
      "|103110882ce930|     206329| 745493448|       2|   US>NJ>501|             3|  11187593| 745493448|       2|   US>NJ>501|\n",
      "|103148119d285e|     813249|1087222105|       1|   US>CA>803|             3|  16298407|1087326936|       1|   US>CA>803|\n",
      "|1031662e9f3021|    2055122| 985561743|       3|   US>PA>577|             1|  14857339| 985561743|       3|   US>PA>577|\n",
      "|10322b3e63a65a|    2200917| 500793817|       3|   US>NY>501|             1|   7298905| 500798318|       3|   US>NY>501|\n",
      "|1033c6ea78eb1f|       2861| 559067297|       1|   US>MA>506|             2|    559664|  41587138|       1|   US>MA>506|\n",
      "|1033c6ea78eb1f|       2861|  41587138|       1|   US>MA>506|             2|    559664|  41587138|       1|   US>MA>506|\n",
      "|1033ccd32b661b|    1135368| 851745088|       2|   US>TX>623|             3|  12921921| 851745088|       2|   US>TX>623|\n",
      "|103441daf02843|      60948| 819015261|       1|   US>NC>567|             1|  11051943| 738975169|       1|   US>NC>567|\n",
      "|103441daf02843|      60948| 731193177|       1|   US>NC>567|             1|  11051943| 738975169|       1|   US>NC>567|\n",
      "|1034c4a761ab4d|     623991| 323638222|       3|       CA>AB|             1|  17759611| 323638222|       3|       CA>AB|\n",
      "|103722acace934|    1158222|1132594285|       1|   US>CA>803|             1|  19975986|1132594285|       1|   US>CA>803|\n",
      "|1037e104d1192b|    1860786|1232195740|       2|   US>CA>803|             1|  21779889|1232195740|       2|   US>CA>803|\n",
      "|1037fc273e1fec|    1154913|1142243709|       2|          GB|             1|  20074733|1142243709|       2|          GB|\n",
      "|103949dc495ffe|    2862978|1210137807|       1|   US>WA>819|             3|  21532341|1210137807|       1|   US>WA>819|\n",
      "|1039850430c8f0|    2526129| 849395755|       1|   US>IL>602|             1|  19148693| 849395755|       1|   US>IL>602|\n",
      "|1039b9229cb780|    2084885| 417668045|       1|   US>IN>515|             1|  17979796| 417668045|       1|   US>IN>515|\n",
      "|103b46e9326282|    2627156| 849338781|       2|   US>TX>625|             1|  12883729| 849338781|       2|   US>TX>625|\n",
      "|103b80db70f9e4|    2102878| 340522792|       2|   US>NY>538|             3|  17803603| 340522792|       2|   US>NY>538|\n",
      "|103bdee1f1bfc7|    1690147| 228741055|       1|   US>GA>524|             1|   3536061| 228810448|       1|   US>GA>524|\n",
      "|103c04f330d070|    2481914| 739349729|       1|   US>CA>803|             1|  11059970| 739349729|       1|   US>CA>803|\n",
      "|103c1c50fc4705|    1316528| 655904608|       1|       GB>C6|             1|   9732087| 655904608|       1|       GB>C6|\n",
      "|103c565d6cc919|    1644971|1055525901|       1|       RU>48|             2|  15801702|1055525901|       1|       RU>48|\n",
      "|103cbdcea7a6e8|    2121289| 603612177|       3|   US>MA>506|             3|  18463116| 603614357|       3|   US>MA>506|\n",
      "|103d69734ec849|    1179111| 682998385|       2|   US>CA>803|             3|   8434915| 574216744|       2|   US>CA>803|\n",
      "|103d69734ec849|    1179111| 574216744|       2|   US>CA>803|             3|   8434915| 574216744|       2|   US>CA>803|\n",
      "|103d69734ec849|    1179111|1212530201|       2|   US>CA>803|             1|   8434915| 574216744|       2|   US>CA>803|\n",
      "|103d6cec524813|     781946| 744586831|       1|          GB|             2|  11169356| 744590470|       1|          GB|\n",
      "|103d84992e4897|    2240621| 515857466|       2|   US>TX>618|             1|   7529336| 515857466|       2|   US>TX>618|\n",
      "|103d97daa51548|    2088091| 299109500|       2|   US>NY>501|             1|   4564068| 299109500|       2|   US>NY>501|\n",
      "|103e65710f1508|    2355210| 614241075|       3|   US>CA>807|             1|  18483300| 614241075|       3|   US>CA>807|\n",
      "|103ede5986c82a|       7054| 959814381|       2|       SA>10|             1|  19444199| 959814381|       2|       SA>10|\n",
      "|103fa9ffc61556|    2508923| 819460284|       1|   US>TX>618|             2|  19043916| 819460710|       1|   US>TX>618|\n",
      "|1040b1e87e1ee7|     468335| 244027315|       1|   US>NC>517|             3|   3830171| 244027315|       1|   US>NC>517|\n",
      "|1042c60a88f3e4|    1302979|1018648293|       3|       GB>F8|             3|  15367911|1018648293|       3|       GB>F8|\n",
      "|1043a2285fd489|     403505| 195736021|       2|          ZA|             1|  17393148| 195736021|       2|          ZA|\n",
      "|1043f9dc59eaa2|    1233145| 559498276|       2|   US>NC>560|             1|   8112340| 559498276|       2|   US>NC>560|\n",
      "|10441aac01b321|    2724733| 992654771|       1|   US>WA>819|             1|  14985228| 992654771|       1|   US>WA>819|\n",
      "|10448baa63e1aa|    2626826| 913525052|       1|   US>VA>511|             1|  13846573| 913525052|       1|   US>VA>511|\n",
      "|1045d1397e0f20|    2083509| 265608272|       2|   US>CA>803|             2|   4149368| 265608272|       2|   US>CA>803|\n",
      "|10461d4674b8df|    1449981| 276494807|       2|   US>CO>751|             3|   4239263| 276494807|       2|   US>CO>751|\n",
      "|1047cb09808a03|    2124489| 592986157|       1|   US>AR>734|             1|   8810339| 592987154|       1|   US>AR>734|\n",
      "|1047df411a722d|    1744653| 473391468|       3|   US>CA>803|             1|   5796801| 385708702|       3|   US>CA>803|\n",
      "|1047df411a722d|    1744653| 554526103|       3|   US>CA>803|             1|   5796801| 385708702|       3|   US>CA>803|\n",
      "|1047df411a722d|    1744653| 608270338|       3|   US>CA>803|             1|   5796801| 385708702|       3|   US>CA>803|\n",
      "|1047df411a722d|    1744653| 385707937|       3|   US>CA>803|             1|   5796801| 385708702|       3|   US>CA>803|\n",
      "|10484789495ee2|      77718|1224106570|       3|       GB>I2|             1|  21708890|1225803625|       3|       GB>I2|\n",
      "|1048c6f1b705d5|    1796200|   5811674|       1|       CA>BC|             1|     70321|   5811674|       1|       CA>BC|\n",
      "|104926b096f103|    2616889| 836418570|       2|   US>OH>547|             1|  12651900| 836418570|       2|   US>OH>547|\n",
      "|104953deea86f1|    1268736| 423174702|       1|   US>GA>520|             1|   6329427| 423259373|       1|   US>GA>520|\n",
      "|1049675f602579|    2915491|1213241537|       2|   US>WA>819|             3|  21575229|1213241537|       2|   US>WA>819|\n",
      "|1049b336d8ab79|    1367396| 706933542|       1|   US>CA>807|             1|  10521167| 706933542|       1|   US>CA>807|\n",
      "|104a3cdcc6da3f|    1692269| 317443061|       1|   US>MA>506|             1|   4905232| 317443061|       1|   US>MA>506|\n",
      "|104b1398e771a1|    2122235| 319263752|       2|   US>FL>534|             2|   4938802| 319263752|       2|   US>FL>534|\n",
      "|104b64b888ba4f|    1772591| 356218989|       2|   US>CA>803|             1|   5450350| 356335108|       2|   US>CA>803|\n",
      "|104bf6f3d13f6f|    1788295| 336349465|       1|   US>NY>514|             1|   5201191| 336483274|       1|   US>NY>514|\n",
      "|104ca864c6cd3d|    1129130| 307348412|       3|   US>AZ>753|             1|   4719647| 307348412|       3|   US>AZ>753|\n",
      "|104ca864c6cd3d|    1129130| 307348412|       3|   US>AZ>753|             1|   4723291| 307543218|       3|   US>AZ>753|\n",
      "|104ce536076840|    2203423|1023370423|       2|   US>NY>501|             1|  15431584|1023370423|       2|   US>NY>501|\n",
      "|104d4136f57ea8|    2742063|1107827298|       2|   US>NY>501|             3|  19808768|1107827298|       2|   US>NY>501|\n",
      "|104db112485838|    1061689| 454231897|       2|       GB>K2|             1|   6633541| 454231897|       2|       GB>K2|\n",
      "|104dbb3e717cce|     846063|1000701376|       1|   US>NY>501|             2|  15108914|1000702419|       1|   US>NY>501|\n",
      "|104e0cd1254834|     363404| 593135329|       3|       CA>ON|             1|   8813195| 593144041|       3|       CA>ON|\n",
      "|104e42384ab299|    1132102| 352314372|       2|       NZ>F2|             1|  17830560| 352314372|       2|       NZ>F2|\n",
      "|104e990a8f68fe|    1595305| 394304379|       2|   US>CA>862|             1|   5932310| 394304379|       2|   US>CA>862|\n",
      "|10500084d3907b|    1458170| 557055811|       1|       TT>08|             1|   8059991| 557056746|       1|       TT>08|\n",
      "|105069fe8ec36d|    1573953| 106869865|       2|   US>FL>561|             1|   1622694| 106869865|       2|   US>FL>561|\n",
      "|105124a93f0120|    1827718| 758856346|       1|   US>WA>819|             1|  11445533| 758856346|       1|   US>WA>819|\n",
      "|10519c6a459405|    2261206| 515796169|       3|   US>AL>691|             1|   7528262| 515796169|       3|   US>AL>691|\n",
      "|1051a119874c2b|    2816434|1123986511|       1|   US>AZ>753|             1|  19864412|1123986511|       1|   US>AZ>753|\n",
      "|1054c164d5b9b8|    2259992| 561394209|       2|   US>PA>504|             1|   8153246| 561394209|       2|   US>PA>504|\n",
      "|1055465bd75501|    1353588| 920675626|       2|   US>TX>618|             1|  13986103| 920675626|       2|   US>TX>618|\n",
      "|1056c57feaa73e|    2902856|1284628237|       2|          US|             3|  22993899|1289092669|       2|          US|\n",
      "|105737cc26a800|    2165306|1152525328|       2|   US>DC>511|             1|  20239482|1152525328|       2|   US>DC>511|\n",
      "|1057395e00192e|    2365555| 648316017|       1|   US>TX>618|             1|   9574901| 648316017|       1|   US>TX>618|\n",
      "|10575bc7adb446|    2640167| 865113544|       2|   US>IL>602|             1|  13134049| 865113544|       2|   US>IL>602|\n",
      "|1058cdc870a8e2|    1300777|1276330758|       2|       GB>A9|             1|  22758310|1276330758|       2|       GB>A9|\n",
      "|1058d99b3c3d60|    1573953| 551605641|       2|   US>MN>613|             1|  18277445| 551605641|       2|   US>MN>613|\n",
      "|10592552170dc9|    2121879| 316282357|       2|       US>TX|             3|   4883725| 316282357|       2|       US>TX|\n",
      "|105a502b15db2d|    2684642| 952137432|       1|   US>FL>539|             1|  14486910| 952293716|       1|   US>FL>539|\n",
      "|105a56fb6dac62|    2500262| 747302229|       1|   US>PA>508|             1|  11224072| 747302808|       1|   US>PA>508|\n",
      "|105afd4b9391b3|    1007252|1223990608|       3|       IE>29|             1|  21689628|1223992835|       3|       IE>29|\n",
      "|105bf8d16df006|    1861497|  65532674|       2|   US>PA>504|             3|   1083022|  65532674|       2|   US>PA>504|\n",
      "|105e2934186a1b|    2771922|1097091561|       2|   US>PA>504|             3|  16454517|1097091561|       2|   US>PA>504|\n",
      "|105fc0c04bcedf|    2687732| 924925025|       3|   US>MD>511|             1|  14068745| 924925025|       3|   US>MD>511|\n",
      "|10609364a2cbd0|     814485|1135321504|       1|   US>TX>635|             1|  20002831|1135321504|       1|   US>TX>635|\n",
      "|10621622f3dba7|    1859708|  63323696|       1|   US>TN>640|             1|   1039020|  63323696|       1|   US>TN>640|\n",
      "|106230d2b7b472|    1708834| 319407004|       2|       GB>F2|             1|  17746612| 319407004|       2|       GB>F2|\n",
      "|10625b1457d6db|    2696622| 998597052|       1|   US>IL>602|             1|  15078153| 998597244|       1|   US>IL>602|\n",
      "|1062bb8c3f8997|    1446279|  57193872|       2|          US|             1|    906187|  57193872|       2|          US|\n",
      "|1062e1c64e77bf|    1271809| 551975140|       1|   US>WA>819|             1|   7952947| 551975382|       1|   US>WA>819|\n",
      "|106360b36bb1b1|    1408800| 466694970|       3|   US>NC>560|             3|   6780830| 466805120|       3|   US>NC>560|\n",
      "|10637b9989307e|    2333442| 639954422|       1|   US>FL>528|             1|   9415768| 639955509|       1|   US>FL>528|\n",
      "|10639c9a76f52d|     902185| 575290395|       1|       GB>R3|             1|   8459757| 575290776|       1|       GB>R3|\n",
      "|1064c725b1b6fb|    1872851| 154819337|       2|   US>WA>819|             2|   2459013| 154819337|       2|   US>WA>819|\n",
      "|1064cd2b530eb3|    2750195|1035166496|       1|   US>WA>819|             1|  15596653|1035166496|       1|   US>WA>819|\n",
      "|106570e6cf7bee|     814045| 747240041|       3|   US>TX>709|             3|  11222781| 747240041|       3|   US>TX>709|\n",
      "|10670bd9a09c5e|     428928| 657428335|       3|   US>TX>641|             1|   9764087| 657428335|       3|   US>TX>641|\n",
      "|106864ce212321|    2710435| 985392850|       2|   US>NJ>501|             1|  14854373| 985392850|       2|   US>NJ>501|\n",
      "|106918822649c5|    1361146| 322647339|       3|       CA>NB|             1|   4997138| 322647339|       3|       CA>NB|\n",
      "|1069ddfb1cbc80|    2223355| 454720050|       2|   US>FL>534|             1|   6635235| 454475653|       2|   US>FL>534|\n",
      "|1069feed9fa7aa|    2792500|1184612145|       1|       CA>ON|             1|  21014654|1184620079|       1|       CA>ON|\n",
      "|106a15bb2d86f4|    1794730| 424459661|       1|       HK>00|             1|   6346027| 424459661|       1|       HK>00|\n",
      "|106bb60ef1b9bf|    1297445|1087774766|       2|   US>MA>506|             1|  16284121|1086441750|       2|   US>MA>506|\n",
      "|106bfd24753570|      38915| 155726599|       2|   US>NC>518|             1|  11962813| 801753568|       2|   US>NC>518|\n",
      "|106bfd24753570|      38915|1004447303|       2|   US>NC>518|             1|  11962813| 801753568|       2|   US>NC>518|\n",
      "|106bfd24753570|      38915| 541483357|       2|   US>NC>518|             1|  11962813| 801753568|       2|   US>NC>518|\n",
      "|106bfd24753570|      38915|1073870601|       2|   US>NC>518|             1|  11962813| 801753568|       2|   US>NC>518|\n",
      "|106bfd24753570|      38915| 721711173|       2|   US>NC>517|             1|  11962813| 801753568|       2|   US>NC>518|\n",
      "|106bfd24753570|      38915| 801753568|       2|   US>NC>518|             1|  11962813| 801753568|       2|   US>NC>518|\n",
      "|106bfd24753570|      38915| 629270646|       2|   US>NC>518|             1|  11962813| 801753568|       2|   US>NC>518|\n",
      "|106bfd24753570|      38915| 499031513|       2|   US>NC>518|             1|  11962813| 801753568|       2|   US>NC>518|\n",
      "|106c93df657cfc|     178158|1158819519|       2|       QA>01|             2|  20384365|1158819519|       2|       QA>01|\n",
      "+--------------+-----------+----------+--------+------------+--------------+----------+----------+--------+------------+\n",
      "only showing top 200 rows\n",
      "\n",
      "CPU times: user 124 ms, sys: 56 ms, total: 180 ms\n",
      "Wall time: 11min 44s\n"
     ]
    }
   ],
   "source": [
    "page_view_events_matches_df = page_views_events_joined_by_uuid_document_df \n",
    "                        .filter('dummyPageView is not null and dummyEvent is not null')\n",
    "%time page_view_events_matches_df.show(200)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Performing a join between page views and events, on fields uuid, document_id, platform, and geo_location"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here we left join events and page views datasets on uuid, document_id, platform, and geo_location.  \n",
    "There are **26323644** rows in the joined dataset, but events dataset has fewer rows (**23120126**) suggesting that those four fields does not uniquely join the datasets, because some events rows are duplicated to match page views."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 176 ms, sys: 48 ms, total: 224 ms\n",
      "Wall time: 21min 45s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "26323644"
      ]
     },
     "execution_count": 142,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%time events_with_day_df.alias('events').join(page_views_with_day_df.alias('page_views'),\n",
    "                                              on=['uuid','document_id','platform','geo_location'], how='left') \\\n",
    "                                        .count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Performing a join between page views and events, on fields uuid, document_id and timestamp"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's try to join datasets now considering uuid, document_id and timestamp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "page_views_events_joined_by_uuid_document_timestamp_df  = events_df.alias('events') \\\n",
    "                                                            .join(page_views_df.alias('page_views'), \n",
    "                                                                   on=['uuid','document_id','timestamp'], \n",
    "                                                                   how='left')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that joining on those 3 fields we have **15854250** matches. This is only **68%** of the total **23120126** events, so **timestamps between page views and events appears not to be exactly aligned**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 168 ms, sys: 24 ms, total: 192 ms\n",
      "Wall time: 17min 47s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "15854250"
      ]
     },
     "execution_count": 198,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%time page_views_events_joined_by_uuid_document_timestamp_df.filter('dummyEvent is not null and dummyPageView is not null').count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Analyzing delta between page views and events timestamps"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We've built a User Defined Function (UDF) to calculate the delta seconds between page views and events timestamps"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "diff_between_page_views_events_timestamps_udf = F.udf(lambda event_ts, page_view_ts: round(abs((convert_odd_timestamp(event_ts) - convert_odd_timestamp(page_view_ts)).total_seconds())), IntegerType())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We use again the matching rows of joined dataset only on (uuid, document_id) fields, to analyze the statistics of  timestamp delta (in seconds) between the two datasets. We can see that the **average (arithmetic mean) delta between timestamps is 49323 seconds (13.7 hours)**. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 172 ms, sys: 48 ms, total: 220 ms\n",
      "Wall time: 17min 47s\n",
      "+-------+--------------------+\n",
      "|summary|timestamp_delta_secs|\n",
      "+-------+--------------------+\n",
      "|  count|            26750921|\n",
      "|   mean|   49323.24790279183|\n",
      "| stddev|   159902.2796012831|\n",
      "|    min|                   0|\n",
      "|    max|             1295724|\n",
      "+-------+--------------------+\n",
      "\n"
     ]
    }
   ],
   "source": [
    "page_view_events_matches_with_delta_ts_df = page_view_events_matches_df \\\n",
    "                                        .withColumn('timestamp_delta_secs', diff_between_page_views_events_timestamps_udf('events.timestamp', 'page_views.timestamp')) \\\n",
    "                                        .select('timestamp_delta_secs')\n",
    "%time page_view_events_matches_with_delta_ts_df.describe('timestamp_delta_secs').show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As arithmetic mean are very sensitive to large outliers, we analyze the distribution using quantiles.\n",
    "We can see that at least 50% (median) of matching rows on (uuid, document_id) have a delta of 0 seconds (identical), but at 75% quantile we have samples with a delta of around 211 seconds (3.5 min).  \n",
    "I guess one possible source of these differences might be that, for some platforms or providers, a page view is logged on page load, and then, after some time in that browser tab, the user clicks in one ad, and a display_id is logged in events dataset with a newer timestamp."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 208 ms, sys: 36 ms, total: 244 ms\n",
      "Wall time: 18min 15s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[(0.25, 0.0),\n",
       " (0.3, 0.0),\n",
       " (0.35, 0.0),\n",
       " (0.4, 0.0),\n",
       " (0.45, 0.0),\n",
       " (0.5, 0.0),\n",
       " (0.75, 211.0)]"
      ]
     },
     "execution_count": 195,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "quantiles_levels = [0.25, 0.3, 0.35, 0.4, 0.45, 0.5, 0.75]        \n",
    "%time quantiles = page_view_events_matches_with_delta_ts_df.approxQuantile(\"timestamp_delta_secs\", quantiles_levels, 0.05)\n",
    "list(zip(quantiles_levels, quantiles))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Performing the join on fields uuid, document_id, platform, geo_location and day(timestamp)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we have a delta between page views and events timestamps, we extract the dates from timestamps (ignoring hours) from both datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "truncate_day_from_timestamp_udf = F.udf(lambda ts: int(ts / 1000 / 60 / 60 / 24), IntegerType())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "page_views_with_day_df = page_views_df.withColumn('day', truncate_day_from_timestamp_udf('timestamp')).withColumn('dummyPageView', F.lit(1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "events_with_day_df = events_df.withColumn('day', truncate_day_from_timestamp_udf('timestamp')).withColumn('dummyEvent', F.lit(1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Now we join the datasets on fields uuid, document_id, platform, geo_location and day(timestamp).  \n",
    "This join approach appears to be reasonable, because there are 23120126 resulting rows (exactly the events count, with no duplication) and the count of rows where events have no corresponding page view is 55568 (only 0.24% of events)**  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "page_views_events_with_day_joined_df = events_with_day_df.alias('events').join(page_views_with_day_df.alias('page_views'), on=['uuid','document_id','platform','geo_location','day'], how='left')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 376 ms, sys: 104 ms, total: 480 ms\n",
      "Wall time: 32min 26s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "23120126"
      ]
     },
     "execution_count": 134,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%time page_views_events_with_day_joined_df.count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Probably this relatively few events (0.24% of events) not matching any page view in the same day are due to possible page view and events timestamps around 4am (midnight at Eastern time) being splits in different days, even if delta between them is just some minutes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 360 ms, sys: 128 ms, total: 488 ms\n",
      "Wall time: 33min 29s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "55568"
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "events_without_matching_page_views_on_same_day_df = page_views_events_with_day_joined_df \\\n",
    "                                            .filter('page_views.dummyPageView is null')\n",
    "%time events_without_matching_page_views_on_same_day_df.count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Analyzing users' additional page views not in events"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The main reason for the usage of page views dataset is that it might provide additional user logs when compaired to events dataset, that could help us to model users preferences or interests. Let's test this hypothesis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 229,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#Registering events dataset as a view, to use it in a SQL query\n",
    "events_with_day_df.createOrReplaceTempView(\"EVENTS\")    \n",
    "\n",
    "#Registering \"page_views left join events dataset\" as a view, to use it in a SQL query\n",
    "page_views_left_join_events_df = page_views_with_day_df.alias('page_views') \\\n",
    "                                        .join(events_with_day_df.alias('events'), \n",
    "                                              on=['uuid','document_id','platform','geo_location','day'], \n",
    "                                              how='left')\n",
    "page_views_left_join_events_df.createOrReplaceTempView(\"PAGE_VIEWS_LEFT_JOIN_EVENTS\")\n",
    "\n",
    "#Grouping how many users have exactly N additional page views (which are not in events dataset, \n",
    "#                                                              but only from users in events dataset)\n",
    "additional_page_views_by_user_df = spark.sql('''\n",
    "            SELECT additionalPageViewsByUser, COUNT(*) as usersCount FROM\n",
    "                (SELECT UUID, COUNT(*) as additionalPageViewsByUser \n",
    "                FROM PAGE_VIEWS_LEFT_JOIN_EVENTS p \n",
    "                WHERE dummyEvent IS NULL\n",
    "                  AND EXISTS (SELECT uuid FROM EVENTS where uuid = p.uuid) \n",
    "                GROUP BY UUID)\n",
    "            GROUP BY additionalPageViewsByUser''')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this table, we show that **2288031 users have only 1 additional page view** (which is not in events dataset), **1654220 users have exactly 2 additional page views**, etc."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 230,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 368 ms, sys: 88 ms, total: 456 ms\n",
      "Wall time: 27min 19s\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>additionalPageViewsByUser</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2288031</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1654220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1288044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1049242</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>872857</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>742933</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>641927</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>559609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>493910</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>439872</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             count\n",
       "additionalPageViewsByUser         \n",
       "1                          2288031\n",
       "2                          1654220\n",
       "3                          1288044\n",
       "4                          1049242\n",
       "5                           872857\n",
       "6                           742933\n",
       "7                           641927\n",
       "8                           559609\n",
       "9                           493910\n",
       "10                          439872"
      ]
     },
     "execution_count": 230,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%time additional_page_views_by_user_pandas = additional_page_views_by_user_df.toPandas()\n",
    "additional_page_views_by_user_pandas = additional_page_views_by_user_pandas.set_index('additionalPageViewsByUser').sort_index()\n",
    "additional_page_views_by_user_pandas.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "** There are 14951413 (about 75%) users in events dataset which have at least one additional page view (besides the ones logged as events), and 4843554 (about 25%) users with no additional page view. **"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 233,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "14951413"
      ]
     },
     "execution_count": 233,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "users_with_additional_page_views_count = additional_page_views_by_user_pandas['usersCount'].sum()\n",
    "users_with_additional_page_views_count"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4843554"
      ]
     },
     "execution_count": 234,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "users_without_additional_page_view = distinct_events_users - users_with_additional_page_views_count\n",
    "users_without_additional_page_view"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Below it can be seen an histogram drawing a beautiful Poisson distribution of the users count with at least one additional page view.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fe6ec11f080>"
      ]
     },
     "execution_count": 199,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAigAAAF9CAYAAAAuppNcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nUMzMzKxwXKCYmZlZ4fz/ggqxBAGGrysAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe6ec109908>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "additional_page_views_by_user_pandas.head(20).plot.bar(y='usersCount')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Users count cumulative percentage (users with at least one additional page view)**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>usersCount</th>\n",
       "      <th>cum_perc</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>additionalPageViewsByUser</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2288031</td>\n",
       "      <td>15.303109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1654220</td>\n",
       "      <td>26.367080</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1288044</td>\n",
       "      <td>34.981945</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1049242</td>\n",
       "      <td>41.999622</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>872857</td>\n",
       "      <td>47.837579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>742933</td>\n",
       "      <td>52.806561</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>641927</td>\n",
       "      <td>57.099981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>559609</td>\n",
       "      <td>60.842831</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>493910</td>\n",
       "      <td>64.146265</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>439872</td>\n",
       "      <td>67.088275</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                           usersCount   cum_perc\n",
       "additionalPageViewsByUser                       \n",
       "1                             2288031  15.303109\n",
       "2                             1654220  26.367080\n",
       "3                             1288044  34.981945\n",
       "4                             1049242  41.999622\n",
       "5                              872857  47.837579\n",
       "6                              742933  52.806561\n",
       "7                              641927  57.099981\n",
       "8                              559609  60.842831\n",
       "9                              493910  64.146265\n",
       "10                             439872  67.088275"
      ]
     },
     "execution_count": 208,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "additional_page_views_by_user_pandas['cum_perc'] = (additional_page_views_by_user_pandas['usersCount'].cumsum() / users_with_additional_page_views_count) * 100\n",
    "additional_page_views_by_user_pandas.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fe6ec043198>"
      ]
     },
     "execution_count": 209,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgUAAAF5CAYAAAAPlFkqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe6ec0bdc18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "additional_page_views_by_user_pandas.head(20).plot.line(y='cum_perc')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Some outlier users with more than 500 additional page views..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 228,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>usersCount</th>\n",
       "      <th>cum_perc</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>additionalPageViewsByUser</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>541</th>\n",
       "      <td>1</td>\n",
       "      <td>99.999940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>546</th>\n",
       "      <td>1</td>\n",
       "      <td>99.999946</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>553</th>\n",
       "      <td>1</td>\n",
       "      <td>99.999953</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>560</th>\n",
       "      <td>1</td>\n",
       "      <td>99.999960</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>618</th>\n",
       "      <td>1</td>\n",
       "      <td>99.999967</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>635</th>\n",
       "      <td>1</td>\n",
       "      <td>99.999973</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>654</th>\n",
       "      <td>1</td>\n",
       "      <td>99.999980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>656</th>\n",
       "      <td>1</td>\n",
       "      <td>99.999987</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>676</th>\n",
       "      <td>1</td>\n",
       "      <td>99.999993</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>717</th>\n",
       "      <td>1</td>\n",
       "      <td>100.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                           usersCount    cum_perc\n",
       "additionalPageViewsByUser                        \n",
       "541                                 1   99.999940\n",
       "546                                 1   99.999946\n",
       "553                                 1   99.999953\n",
       "560                                 1   99.999960\n",
       "618                                 1   99.999967\n",
       "635                                 1   99.999973\n",
       "654                                 1   99.999980\n",
       "656                                 1   99.999987\n",
       "676                                 1   99.999993\n",
       "717                                 1  100.000000"
      ]
     },
     "execution_count": 228,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "additional_page_views_by_user_pandas.tail(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Conclusion"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**In this notebook, we leveraged the processing power of a Spark cluster, and explored page_views.csv dataset and its relationship with events.csv dataset.  \n",
    "We discovered that fields (uuid, document_id, platform, geo_location and day(timestamp)) can be used to effectivelly join both datasets. And events dataset can be considered to be a subset of page_views dataset, as almost all (99.76%) its rows can be matched in page_views.csv using these fields.  \n",
    "We also discovered that 75% of users in events.csv have additional logged visits on page_views.csv, which hopefully might be used to model users preferences for personalized ads ranking.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.4.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}