{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h1><center>Analyzing Data-Science community Stack Exchange Data</center></h1>\n",
    "<h3><center>Determine the most popular content on the Data-science community of Stack exchange</center></h3>\n",
    "\n",
    "Stack Exchange is a network of question-and-answer websites on topics in diverse fields, each site covering a specific topic, where questions, answers, and users are subject to a reputation award process. The reputation system allows the sites to be self-moderating. <br>\n",
    "Data Science Stack Exchange is a question-answer forum, to connect the Data Science community. Posts regarding Data Science - Machine Learning, Statistical Mathematics, Visualizations etc are mainly found here. Each post usually contains a question asked pertaining to above topics and attached are a string of answers.\n",
    "\n",
    "\n",
    "The aim of this project is to analyze the following:-\n",
    "\n",
    "* Questions posted on the data science community of Stack Exchange server in the 2019 year to determine the most popular topics among these post. \n",
    "* Trend in the popularity of deep learning topic over the years. \n",
    "    * Would it benefit the community to post on Deep learning?\n",
    "    * What other topics can be paired with Deep learning for the posts?\n",
    "\n",
    "The purpose of the analysis is to determine and write posts, to help the community, on topics that are most sought after in the data science community of Stack Exchange<br>\n",
    "Stack Exchange hosts an open Database - <a href=\"https://data.stackexchange.com/datascience/query/new\">Link</a>, the community can utilize. After analyzing the database, the table *Posts* is found relevant for the end goal. This is downloaded into the 2019_questions.csv file. The posts are of the year 2019 only.\n",
    "\n",
    "The *Posts* table has the following attributes(columns) :<br>\n",
    "\n",
    "    PostTypeId : id\n",
    "    CreationDate : date the post was uploaded\n",
    "    Score : the score of the post\n",
    "    ViewCount : number of views\n",
    "    Tags : tags associated with the posts\n",
    "    AnswerCount : number of replies to the post\n",
    "    FavoriteCount : number of likes recieved by the post\n",
    "\n",
    "Out of the type of posts, we will mainly focus on the Questions and Answers. The other kind of posts are inconsequential.<br>\n",
    "\n",
    "The database is queried for all questions & answers in 2019 with above columns, and stored in the csv file - 2019_questions.csv\n",
    "For the second part of the analysis, all questions are downloaded into the all_questions.csv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from itertools import combinations"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The dataset is read into the environment with the `parse_dates` paramter so that the column *CreationDate* is automatically read in as a datetime object."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('2019_questions.csv',parse_dates=['CreationDate'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 8839 entries, 0 to 8838\n",
      "Data columns (total 7 columns):\n",
      " #   Column         Non-Null Count  Dtype         \n",
      "---  ------         --------------  -----         \n",
      " 0   Id             8839 non-null   int64         \n",
      " 1   CreationDate   8839 non-null   datetime64[ns]\n",
      " 2   Score          8839 non-null   int64         \n",
      " 3   ViewCount      8839 non-null   int64         \n",
      " 4   Tags           8839 non-null   object        \n",
      " 5   AnswerCount    8839 non-null   int64         \n",
      " 6   FavoriteCount  1407 non-null   float64       \n",
      "dtypes: datetime64[ns](1), float64(1), int64(4), object(1)\n",
      "memory usage: 483.5+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Id                  0\n",
       "CreationDate        0\n",
       "Score               0\n",
       "ViewCount           0\n",
       "Tags                0\n",
       "AnswerCount         0\n",
       "FavoriteCount    7432\n",
       "dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.fillna(0,inplace=True)\n",
    "df.FavoriteCount = df.FavoriteCount.astype(int)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The *FavoriteCount* column has Null values. Since the focus of the project is on finding the most popular post, it is imperative that popular in this context has to be defined. FavoriteCount gives the number of likes recieved by a post, while this gives a peak into which post might be popular, it does not show the entire picture. A lot of posts amass many viewers, but not every viewer *Likes* the post. Hence *ViewCount* gives a better statistic as to how popular a post is. By this logic, the definition of popular for the project, A post accumulating alot of views relatively, is considered to be popular.<br>\n",
    "This means, the *FavoriteCount* column is no longer of importance and hence rows containing Null values are filled with 0s.<br>\n",
    "\n",
    "The dataset as such doesnot contain the text of the questions posted. This makes it difficult to learn the topic of the question. An alternative is to consider the tags given to the question by the author. These tags summarize the topic of the question. The *Tags* column holds the tags for the question. Each question can have multiple tags."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0                         <machine-learning><data-mining>\n",
       "1       <machine-learning><regression><linear-regressi...\n",
       "2            <python><time-series><forecast><forecasting>\n",
       "3                   <machine-learning><scikit-learn><pca>\n",
       "4                <dataset><bigdata><data><speech-to-text>\n",
       "                              ...                        \n",
       "8834      <pca><dimensionality-reduction><linear-algebra>\n",
       "8835                       <keras><weight-initialization>\n",
       "8836                     <python><visualization><seaborn>\n",
       "8837                                        <time-series>\n",
       "8838                                               <k-nn>\n",
       "Name: Tags, Length: 8839, dtype: object"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.Tags"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The *Tags* column is the center of interest to the project. The tags have to be cleaned and made uniform for further analysis. \n",
    "\n",
    "* All types of chevrons are removed\n",
    "* The string is split on ',' to convert it to list format\n",
    "\n",
    "A list format makes it easier to carry out any modifications/inferences to the *Tags* column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0                         [machine-learning, data-mining]\n",
       "1       [machine-learning, regression, linear-regressi...\n",
       "2            [python, time-series, forecast, forecasting]\n",
       "3                   [machine-learning, scikit-learn, pca]\n",
       "4                [dataset, bigdata, data, speech-to-text]\n",
       "                              ...                        \n",
       "8834      [pca, dimensionality-reduction, linear-algebra]\n",
       "8835                       [keras, weight-initialization]\n",
       "8836                     [python, visualization, seaborn]\n",
       "8837                                        [time-series]\n",
       "8838                                               [k-nn]\n",
       "Name: Tags, Length: 8839, dtype: object"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.Tags = df.Tags.str.replace(\"><\",\",\")\n",
    "df.Tags = df.Tags.str.replace(\"[<>]\",\"\")\n",
    "df.Tags = df.Tags.str.split(\",\")\n",
    "df.Tags"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The popularity as defined above, is the relative amount of views on a post. Since the actual post is missing from the data. The popularity will now be based on the tags. The rectified definition - Popularity of a topic (tag) is the number of posts of that topic and the total views amassed by the topic overall.<br>\n",
    "\n",
    "For this purpose, two functions are coded. The `count_tags()` function returns the number of posts tagged with the topic (tag) passed. The `count_views()` function returns the number of views accumulated by the topic (tag) across all posts. A new dataframe is created that holds the returned value for each unique tag in the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Tags</th>\n",
       "      <th>TagCount</th>\n",
       "      <th>ViewCount</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>.net</td>\n",
       "      <td>1</td>\n",
       "      <td>438</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3d-object-detection</td>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3d-reconstruction</td>\n",
       "      <td>9</td>\n",
       "      <td>1129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ab-test</td>\n",
       "      <td>6</td>\n",
       "      <td>153</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>accuracy</td>\n",
       "      <td>89</td>\n",
       "      <td>15233</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  Tags  TagCount  ViewCount\n",
       "0                 .net         1        438\n",
       "1  3d-object-detection         1          7\n",
       "2    3d-reconstruction         9       1129\n",
       "3              ab-test         6        153\n",
       "4             accuracy        89      15233"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tags = np.unique(np.concatenate(df.Tags,axis=None))\n",
    "tag_count = {}\n",
    "\n",
    "def count_tags(tag,data):\n",
    "    ctr=0\n",
    "    for row in data['Tags']:\n",
    "        if tag in row:\n",
    "            ctr+=1\n",
    "    return ctr\n",
    "\n",
    "def count_views(tag,data):\n",
    "    ctr=0\n",
    "    for row in data[['Tags','ViewCount']].iterrows():\n",
    "        if tag in row[1]['Tags']:\n",
    "            ctr+=row[1]['ViewCount']\n",
    "    return ctr\n",
    "\n",
    "for tag in tags:\n",
    "    tag_count[tag] = [count_tags(tag,df),count_views(tag,df)]\n",
    "\n",
    "df_tags = pd.DataFrame.from_dict(tag_count,orient='index').reset_index()\n",
    "df_tags.columns = ['Tags','TagCount','ViewCount']\n",
    "df_tags.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The dataframe summarizes the popularity of each tag for the purpose of the project. The *TagCount* shows the most popular tags in terms of the number of questions posted on that tag. The plot below gives a relative comparision among the Top 10 tags with highest number of posts on the Data science Stack Exchange server."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1728x864 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_tags.sort_values(by='TagCount',ascending=False,inplace=True)\n",
    "most_used = df_tags.Tags.iloc[:20]\n",
    "\n",
    "plt.style.use('fivethirtyeight')\n",
    "plt.figure(figsize=(24,12))\n",
    "sns.barplot(y=df_tags.Tags.iloc[0:10],x=df_tags.TagCount.iloc[0:10])\n",
    "plt.ylabel(\"Popular Tags\")\n",
    "plt.xlabel(\"Number of Posts\")\n",
    "plt.xticks([],rotation=15)\n",
    "plt.title(\"Top 10 most posted topics on Data science Stack Exchange\",fontsize=32,y=1.07)\n",
    "plt.suptitle('Top 10 tags with the highest number of posts',y=0.93, fontsize = 25)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The *ViewCount* similar shows the most popular tags with respect to the number of views recieved to the posts tagged. The plot below gives a relative comparision between the top 10 most viewed topics (tags) on the Data science Stack Exchange server."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1728x864 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_tags.sort_values(by='ViewCount',ascending=False,inplace=True)\n",
    "most_viewed = df_tags.Tags.iloc[:20]\n",
    "\n",
    "plt.figure(figsize=(24,12))\n",
    "sns.barplot(y=df_tags.Tags.iloc[0:10],x=df_tags.ViewCount.iloc[0:10])\n",
    "plt.ylabel(\"Popular Tags\")\n",
    "plt.xlabel(\"Number of Views\")\n",
    "plt.xticks([],rotation=45)\n",
    "plt.title(\"Top 10 most viewed topics on Data science Stack Exchange\",fontsize=32,y=1.07)\n",
    "plt.suptitle('Top 10 tags with the highest number of views',y=0.93, fontsize = 25)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The two plots above give the most popular tags in terms of number of views amassed or the number of questions posted with that tag. Between the two groups, there are common tags. These common tags are ultimately the most popular tags on the Data science Stack Exchange server. They have high views and high number of posts.\n",
    "\n",
    "NOTE : The plot shows the top 10 popular tags for both criteria, but for the intersection, the top 20 tags are considered, to increase the resulting set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['classification' 'clustering' 'cnn' 'dataset' 'deep-learning' 'keras'\n",
      " 'lstm' 'machine-learning' 'neural-network' 'nlp' 'pandas' 'python'\n",
      " 'regression' 'scikit-learn' 'tensorflow' 'time-series']\n"
     ]
    }
   ],
   "source": [
    "popular_tags = np.intersect1d(most_used,most_viewed)\n",
    "print(popular_tags)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Between the two segments there are 16 common tags. These tags are the most popular tags :-<br>\n",
    "\n",
    "    * classification      * clustering       * cnn            * dataset\n",
    "    * deep-learning       * keras            * lstm           * machine-learning\n",
    "    * neural-network      * nlp              * pandas         * python\n",
    "    * regression          * scikit-learn     * tensorflow     * time-series\n",
    "\n",
    "The popular tags mostly belong to a common topic. Based on the dataset, relations between these tags can be discovered to better understand which pairs or sets of tags are widely popular and thus which segment of data science is widely popular in terms of questions on the server.<br>\n",
    "\n",
    "One method is to make all possible pairs from the popular tags and make a table with the pair's Tag counts and View counts. This results in a popularity index as done for the single tags.<br>\n",
    "The pairs are combinations of the popular tags and not the entire tags list, because based on the assumption that the top 10 are the most frequent and rest are not, The rest of the pairs can be eliminated via Upward Closure property from Data Mining concepts.<br>\n",
    "\n",
    "Similar to previous steps, two functions are coded. The `count_pair()` function returns the number of times the pair of tags have appeared together in the questions posted. The `count_pair_views()` function returns the number of views amassed by posts tagged with the pair. The popular pair of tags will be the intersection of the two i.e. pair of tags having high number of posts as well as high number of views."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Tags</th>\n",
       "      <th>TagCount</th>\n",
       "      <th>ViewCount</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>(classification, clustering)</td>\n",
       "      <td>12</td>\n",
       "      <td>1532</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>(classification, cnn)</td>\n",
       "      <td>20</td>\n",
       "      <td>1096</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>(classification, dataset)</td>\n",
       "      <td>28</td>\n",
       "      <td>1991</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>(classification, deep-learning)</td>\n",
       "      <td>59</td>\n",
       "      <td>35105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>(classification, keras)</td>\n",
       "      <td>58</td>\n",
       "      <td>37610</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                              Tags  TagCount  ViewCount\n",
       "0     (classification, clustering)        12       1532\n",
       "1            (classification, cnn)        20       1096\n",
       "2        (classification, dataset)        28       1991\n",
       "3  (classification, deep-learning)        59      35105\n",
       "4          (classification, keras)        58      37610"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pair_tags = list(combinations(popular_tags,2))\n",
    "pair_tag_count = {}\n",
    "\n",
    "def count_pair(tags,data):\n",
    "    ctr=0\n",
    "    for row in data['Tags']:\n",
    "        if set(tags).issubset(row):\n",
    "            ctr+=1\n",
    "    return ctr\n",
    "\n",
    "def count_pair_views(tags,data):\n",
    "    ctr=0\n",
    "    for row in data[['Tags','ViewCount']].iterrows():\n",
    "        if set(tags).issubset(row[1]['Tags']):\n",
    "            ctr+=row[1]['ViewCount']\n",
    "    return ctr\n",
    "\n",
    "for tags in pair_tags:\n",
    "    pair_tag_count[tags] = [count_pair(tags,df),count_pair_views(tags,df)]\n",
    "    \n",
    "df_pair_tags = pd.DataFrame.from_dict(pair_tag_count,orient='index').reset_index()\n",
    "df_pair_tags.columns = ['Tags','TagCount','ViewCount']\n",
    "df_pair_tags.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_pair_tags.sort_values(by='TagCount',ascending=False,inplace=True)\n",
    "most_used_pair = df_pair_tags.Tags.iloc[:10]\n",
    "\n",
    "df_pair_tags.sort_values(by='ViewCount',ascending=False,inplace=True)\n",
    "most_viewed_pair = df_pair_tags.Tags.iloc[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The initial tags were the most popular individually. Making combinations from them, the resulting dataframe was created. For every pair of tags, the dataframe shows the number of posts tagged with that pair and total number of views amassed. Since the initial set was of popular tags, the resulting pairs of popular tags represent association between the two tags forming the pair."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([('deep-learning', 'keras'), ('deep-learning', 'machine-learning'),\n",
       "       ('deep-learning', 'neural-network'), ('keras', 'neural-network'),\n",
       "       ('keras', 'python'), ('keras', 'tensorflow'),\n",
       "       ('machine-learning', 'neural-network'),\n",
       "       ('machine-learning', 'python'), ('pandas', 'python')], dtype=object)"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "popular_tags_pair = np.intersect1d(most_used_pair,most_viewed_pair)\n",
    "popular_tags_pair"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The popular pair of tags as a result of the above analysis are :-\n",
    "\n",
    "    * deep-learning : keras                 \n",
    "    * deep-learning : machine-learning       \n",
    "    * deep-learning : neural-network\n",
    "    * keras : neural-network                \n",
    "    * keras : python                         \n",
    "    * keras : tensorflow\n",
    "    * machine-learning : neural-netowork    \n",
    "    * machine-learning : python              \n",
    "    * pandas : python\n",
    "\n",
    "The concept used above was nothing but finding frequent item sets of Data mining, or in this case popular item sets. The next step in the algorithm would be to make triplets from these popular pairs. This would represent associativity between three tags.<br>\n",
    "The possible triplets that can be formed as a combination of the above popular pairs are :-\n",
    "\n",
    "* deep-learning : keras : neural-network\n",
    "* deep-learning : machine-learning : neural-network\n",
    "* machine-learning : python : keras\n",
    "\n",
    "NOTE : The triplets formed are not all the possible triplets. They are pruned based on the logic - A triplet is formed if and only if all pairs formed from the triplets are part of the set of pairs from which the triplets are derived."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For the first part of the analysis the following conclusions are drawn :-\n",
    "\n",
    "* More generally, tags representing *predictive analysis* are more popular on the Data science community of the Stack Exchange server as compared to the Descriptive analysis or visualization part of Data science field.\n",
    "* The following tags are associated to each other and popular across the Data science community on Stack Exchange server,\n",
    "\n",
    "    * deep-learning\n",
    "    * machine-learning\n",
    "    * neural-networks\n",
    "    * keras"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The next part of the analysis is to discover the trend in popularity of deep-learning over the years and to answer the question :-\n",
    "\n",
    "* Would it benefit the community to post on Deep learning?\n",
    "* What other topics can be paired with Deep learning for the posts?\n",
    "\n",
    "A different dataset has been downloaded for this purpose. Instead of just having questions from the year 2019. All questions across all years are present."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Id</th>\n",
       "      <th>CreationDate</th>\n",
       "      <th>Tags</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>45416</td>\n",
       "      <td>2019-02-12 00:36:29</td>\n",
       "      <td>&lt;python&gt;&lt;keras&gt;&lt;tensorflow&gt;&lt;cnn&gt;&lt;probability&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>45418</td>\n",
       "      <td>2019-02-12 00:50:39</td>\n",
       "      <td>&lt;neural-network&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>45422</td>\n",
       "      <td>2019-02-12 04:40:51</td>\n",
       "      <td>&lt;python&gt;&lt;ibm-watson&gt;&lt;chatbot&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>45426</td>\n",
       "      <td>2019-02-12 04:51:49</td>\n",
       "      <td>&lt;keras&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>45427</td>\n",
       "      <td>2019-02-12 05:08:24</td>\n",
       "      <td>&lt;r&gt;&lt;predictive-modeling&gt;&lt;machine-learning-mode...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Id         CreationDate  \\\n",
       "0  45416  2019-02-12 00:36:29   \n",
       "1  45418  2019-02-12 00:50:39   \n",
       "2  45422  2019-02-12 04:40:51   \n",
       "3  45426  2019-02-12 04:51:49   \n",
       "4  45427  2019-02-12 05:08:24   \n",
       "\n",
       "                                                Tags  \n",
       "0      <python><keras><tensorflow><cnn><probability>  \n",
       "1                                   <neural-network>  \n",
       "2                      <python><ibm-watson><chatbot>  \n",
       "3                                            <keras>  \n",
       "4  <r><predictive-modeling><machine-learning-mode...  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('all_questions.csv',parse_dates=True)\n",
    "df.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The dataset contains only 3 columns as compared to the dataset used previously. Both the datasets are of the same Posts table from the open database provided by Stack Exchange.\n",
    "\n",
    "    * Id - id of the posted question\n",
    "    * CreationDate - date the question was posted\n",
    "    * Tags - tags associated with the post"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 21576 entries, 0 to 21575\n",
      "Data columns (total 3 columns):\n",
      " #   Column        Non-Null Count  Dtype \n",
      "---  ------        --------------  ----- \n",
      " 0   Id            21576 non-null  int64 \n",
      " 1   CreationDate  21576 non-null  object\n",
      " 2   Tags          21576 non-null  object\n",
      "dtypes: int64(1), object(2)\n",
      "memory usage: 505.8+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A similar approach, as done previously, to clean the tags column for ease of the analysis.\n",
    "* All chevrons are removed from the tags column.\n",
    "* The tags are converted to lists for ease of usage"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.Tags = df.Tags.str.replace(\"><\",\",\")\n",
    "df.Tags = df.Tags.str.replace(\"[<>]\",\"\")\n",
    "df.Tags = df.Tags.str.split(\",\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The purpose is to find the general trend deep learning has witnessed over time. The *CreationDate* holds granular information regarding the questions posted. Since the aim is to only find the trend, year extracted from *CreationDate* would suffice."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.CreationDate = pd.to_datetime(df.CreationDate)\n",
    "df['Year'] = df.CreationDate.dt.year"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Posts containing the deep-learning tag are relevant to the analysis. A new column *is_dl* is created to identify which posts are tagged as deep-learning."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['is_dl'] = df.Tags.apply(\n",
    "    lambda x: 1 if 'deep-learning' in x else 0\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The posts have been identified as relevant or not and the years have been extracted from *CreationDate*.<br>\n",
    "From these results, two groupings are made. Group 1, groups on the *Year* column and aggregates the size. In short group 1 indicates the number of questions posted on Data science Stack Exchange server for the years.\n",
    "Group 2, groups on the *Year* column, for only those questions that were tagged as deep-learning, and aggregated on size. Thus group 2 indicates the number of questions posted as deep-learning on Data science Stack Exchange for the years.<br>\n",
    "\n",
    "Both these groupings are comibined using the `pd.merge()` function. The resulting dataframe gives us the statistic on the number of questions posted as deep-learning vs number of questions in total posted for the years. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Year</th>\n",
       "      <th>DL_posts</th>\n",
       "      <th>Total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2014</td>\n",
       "      <td>8</td>\n",
       "      <td>562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015</td>\n",
       "      <td>30</td>\n",
       "      <td>1167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2016</td>\n",
       "      <td>157</td>\n",
       "      <td>2146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2017</td>\n",
       "      <td>425</td>\n",
       "      <td>2957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2018</td>\n",
       "      <td>902</td>\n",
       "      <td>5475</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2019</td>\n",
       "      <td>1216</td>\n",
       "      <td>8810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2020</td>\n",
       "      <td>67</td>\n",
       "      <td>459</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Year  DL_posts  Total\n",
       "0  2014         8    562\n",
       "1  2015        30   1167\n",
       "2  2016       157   2146\n",
       "3  2017       425   2957\n",
       "4  2018       902   5475\n",
       "5  2019      1216   8810\n",
       "6  2020        67    459"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grouped_1 = df.groupby(by=['Year'])\n",
    "grouped_2 = df[df.is_dl == 1].groupby(by=['Year'])\n",
    "\n",
    "temp_1 = pd.DataFrame(grouped_2.size(),columns=['DL_posts']).reset_index()\n",
    "temp_2 = pd.DataFrame(grouped_1.size(),columns=['Total']).reset_index()\n",
    "\n",
    "df_posts = pd.merge(temp_1,temp_2,left_on=['Year'],right_on=['Year'],how='inner')\n",
    "df_posts"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Dividing the columns *DL_posts* and *Total* gives the proportion of posts posted as deep-learning as compared to total posts for that year. These proportions can be used to visualize the trend in the number of posts tagged deep-learning through the years."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>DL_posts</th>\n",
       "      <th>Total</th>\n",
       "      <th>Percentage</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Year</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014</th>\n",
       "      <td>8</td>\n",
       "      <td>562</td>\n",
       "      <td>1.423488</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015</th>\n",
       "      <td>30</td>\n",
       "      <td>1167</td>\n",
       "      <td>2.570694</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016</th>\n",
       "      <td>157</td>\n",
       "      <td>2146</td>\n",
       "      <td>7.315937</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017</th>\n",
       "      <td>425</td>\n",
       "      <td>2957</td>\n",
       "      <td>14.372675</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018</th>\n",
       "      <td>902</td>\n",
       "      <td>5475</td>\n",
       "      <td>16.474886</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019</th>\n",
       "      <td>1216</td>\n",
       "      <td>8810</td>\n",
       "      <td>13.802497</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020</th>\n",
       "      <td>67</td>\n",
       "      <td>459</td>\n",
       "      <td>14.596950</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      DL_posts  Total  Percentage\n",
       "Year                             \n",
       "2014         8    562    1.423488\n",
       "2015        30   1167    2.570694\n",
       "2016       157   2146    7.315937\n",
       "2017       425   2957   14.372675\n",
       "2018       902   5475   16.474886\n",
       "2019      1216   8810   13.802497\n",
       "2020        67    459   14.596950"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_posts['Percentage'] = (df_posts.DL_posts/df_posts.Total) * 100\n",
    "df_posts.set_index('Year',inplace=True)\n",
    "df_posts"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The table clearly showcases an upward trend in the number of posts tagged as deep-learning. The year 2020 has seen less posts as it is the current year, hence is excluded from the analysis. The plot below visualizes the trend shown in the table for better understanding."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAl8AAAGNCAYAAAAvoRSkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOzdeVyN6f8/8FcbRUiRJVHiRNp3iZR1EGWnJI2l+ViGbNkGjUF2Glt2kmVK09hlH2RLiohp00Lata/n+v3hd8630znVydIJ7+fj4THTde5z3+/7Ote57ve5r+u+bynGGAMhhBBCCKkX0pIOgBBCCCHkR0LJFyGEEEJIPaLkixBCCCGkHlHyRQghhBBSjyj5IoQQQgipR5R8EUIIIYTUI0q+PpOdnR0mTZok6TDqxNPTE9ra2pIOo87OnDkDbW1tPHjwQNKh/BB+xPr+Efe5ofrRP4tJkybBzs5O0mGQr4SSL0Lq6PDhwzhz5oykwyCf6MGDB/Dx8UFubq6kQ/mm+fj44OrVq5IO45t25swZHD58WNJhEAmg5It8M0aMGIHIyEiYmZlJNI6jR48iKChIojGQT/fw4UP8+eefIpOvhtLGvgV//vknJV+fKSgoCEePHpV0GEQCKPki9a6iogJFRUV1fp+MjAwaN24MaWlqtuTroDZGvqT8/HxJh/DNKy4uRnl5uaTD+OKohxHTu3fv8Ouvv8LExATGxsZwd3dHYmJitcvfu3cPbm5uMDU1hZ6eHuzt7XHixAmRyz579gwzZ86EhYUFdHV1MWjQIOzevVuowfHmACQlJeGXX37hxzJz5kwkJSV99j6mpaVh5cqV6Nu3L3R1dWFtbY0VK1YgMzNTYLn3799j/fr1GDFiBMzMzKCnp4chQ4bA19cXFRUVAsvy5m3cu3cPO3fuRP/+/aGvr4+LFy8C+L85c7GxsZg+fTqMjIxgYmKCOXPmID09XeS6Ks8B4ZWFhobiwIED6N+/P78ORZ2dqqiowM6dO2Fra8v/XC5cuAAfHx9oa2sjOTm5xjrS1tZGSkoKHj58CG1tbf4/3vvu3LmDuXPnol+/ftDX14epqSnc3Nzw8OFDkeu7fPkyhg8fDj09PfTt2xd//vkn7t27B21tbaGhzeTkZMyePRvGxsYwNjbGL7/8gqSkpGrnHdalDZ4+fRqDBw+Grq4uBgwYgMOHD6OuTx67evUqHBwcoKenBxsbG2zbtg13794V2pea6vpz9+XJkyeYOnUqevXqBT09PfTu3RvTpk3D06dPAXyc7/jnn38CAPr168f//Hx8fABUP88oKysLq1evho2NDXR1dWFjY4PVq1cjOztbYLm6tsebN2/C2dkZFhYW0NfXR9++fTFr1izEx8fXVt38uoqKioKLiwuMjIxgbm6OxYsXC31n67IPJSUl8PHxwaBBg2BgYABTU1PY29vD29sbwMd2yJszGhQUJPA9qKy+29+jR48wZcoUmJiYQF9fH46Ojvjrr78Elpk7dy50dXWF9hkA4uLioK2tjT/++EOg/MKFC5gwYQKMjIxgYGCAMWPG4NKlS0Lv19bWhqenJ0JDQ/nL//LLL9XGa2dnh4cPHyIlJUWgDqu2vffv38PDwwNmZmYwMDDAzz//LLJ9lJaWYs+ePRg6dCj09PRgamoKd3d3vHjxosZ6Az5OpdDW1sbdu3dFrtfCwgIuLi4C5eIetyIjI+Hp6clvT0ZGRhg/fjxCQkKEtsWbj5yVlYUlS5bAysoKhoaGSE1NrXUfvjWykg7gW5CbmwsnJyekpqZi/Pjx0NLSwqNHj+Di4oLi4mKh5U+dOoWVK1fC0NAQ7u7uUFBQwL1797Bq1SokJiZi8eLF/GVv3ryJWbNmoVOnTnBzc0OLFi3w9OlT7NixAy9fvsSOHTsE1l1YWIhJkyZBX18fHh4eePPmDfz9/REREYGgoCC0bt36k/bx7du3GDduHMrKyjB69Gh07NgRb968wYkTJ/DgwQMEBgaiWbNmAIBXr17hypUrGDBgADp27IiysjL8+++/2Lx5M5KTk+Hl5SW0fm9vb5SXl2Ps2LFo2rQpNDU1+a+9f/8eLi4u6N+/PxYtWoTo6GicOnUK+fn5OHjwoFjxb926FcXFxRg3bhwaNWqEEydOwNPTEx07doSJiQl/OS8vL5w8eRIWFhZwc3PjH5DU1NTE2s6GDRuwbt06tGzZEu7u7vxyZWVlAB8PRh8+fICDgwPatm2L9+/f46+//oKrqyuOHj0KU1NT/nsuXLgADw8PdOzYEbNmzYKMjAz+/vtvXL9+XWi72dnZcHJyQmZmJsaPH4/OnTsjLCwMkydPRmFhodDydWmDhw8fxrp169CtWzd4eHigqKgIBw8ehIqKilh1AgAhISGYPXs21NTUMHPmTMjIyODMmTO4deuW2Ouojrj7EhcXBzc3N7Rq1QouLi5QUVFBZmYmwsLCEB0dDUNDQ4wbNw75+fkICQnBkiVL0LJlSwCo8QKUvLw8TJgwAW/evMGoUaOgo6ODly9f4sSJE7h//z7++usvKCoqCrxHnPb48OFD/PLLL+jatStmzJiBZs2aIS0tDaGhoUhMTBT4jlQnNTUVrq6uGDhwIAYNGoQXL14gMDAQz58/R0BAABQUFOq8D6tXr0ZgYCAcHBxgZGSEiooKJCQk8JMCZWVlbNiwAYsWLYKpqSnGjh37yZ8Z8GXa3/Xr1zFr1iy0atUKU6ZMgaKiIs6fP4/ly5cjOTkZ8+bNAwA4Ojri4sWLOH/+PJydnQXWERwczF+m8ue4Z88e9O7dG7/++iukpaUREhKCX3/9Fb/99hucnJwE1vH8+XNcvnwZY8eOFViPKEuXLsXmzZuRnZ2NJUuW8Mu1tLT4/19YWAhnZ2cYGBhg3rx5SE5OxtGjR/G///0P586dg4yMDACgrKwMP//8M8LDwzFixAg4OTkhPz8fp0+fxoQJE+Dn5wc9Pb1qYxkxYgQ2b96MwMBA9OrVS+C1kJAQ5OTkYMyYMfyyuhy3QkJCEBcXh8GDB0NNTQ05OTkICgrCrFmzsGnTJtjb2wvFM2XKFLRq1Qr/+9//UFhYiCZNmtRYl98kRmq1efNmxuFwWEBAgED5mjVrGIfDYc7Ozvyy9+/fM11dXebh4SG0nt9//51169aNJSYmMsYYKy4uZlZWVmzixImsrKxMYNlDhw4xDofD7t+/zy9zdnZmHA6HrVmzRmDZK1euMA6Hw1asWCHW/ixevJhxOByBMnd3d2ZpacnevXsnUB4ZGcm6d+/OduzYwS8rKipiXC5XaL0LFixg3bp1Y+/fv+eXBQYGMg6HwwYOHMgKCwuF3mNra8s4HA47f/68QPmqVasYh8NhsbGxQuuqXCe8shEjRrCSkhJ+eWpqKuvRowebN28ev+z169eMw+EwNzc3VlFRwS+Pjo5m3bp1YxwOhyUlJQlXmIiYK3/mlRUUFAiVpaenM3NzczZ16lR+WVlZGbO2tmY9e/ZkOTk5/PL8/HxmZ2fHOBwOCwwM5Jd7e3szDofDgoODBdbNK//UNvjhwwdmYGDAfvrpJ4HP5927d8zQ0FCovkUpLy9nNjY2zNzcnGVmZvLLc3NzWd++fYX2ZceOHdXWddW6rcu+HDlyhHE4HBYREVFjvDVtX1Qb27JlC+NwOMzPz09gWT8/P8bhcNjWrVuF3i9Oe1y7di3jcDgsIyOjxnirw/vuHDp0SKCc13fs3bv3k/bBzMxMoK1Wh8PhsMWLFwuVS6L99e3bl5mYmLDU1FR+eUlJCRs3bhzr1q0bi4+P5y/bq1cvNmrUKIF1cLlc1rdvXzZs2DB+2fPnzxmHw2GbN28W2uYvv/zCjIyMWF5enkB9cDgcdvfu3RrjrczZ2ZnZ2tpW+xqHw2G+vr4C5fv27WMcDofdvn2bX8b7zCuXMcZYXl4es7Gxqba/qszDw4Pp6uqy7OxsgXJXV1dmZmbGiouLGWN1P26J6hMLCwvZwIED2U8//SRQzjs2zZ8/v9Z4v3U07CiGq1evolWrVnBwcBAonzZtmtCyly9fRmlpKUaPHo2srCyBf3Z2duByubh37x4A4O7du8jIyMDIkSORm5srsGyfPn34y1Q1ffp0gb8HDBgATU1NXLt27ZP2Ly8vDzdv3oSdnR0aNWokEIeamho6duwoEIe8vDykpKQAfDwlnZOTg6ysLFhbW4PL5eL58+dC25gwYQL/V3hVqqqqGDJkiECZpaUlAODNmzdi7cPEiRPRqFEj/t9t2rSBpqYmEhIS+GU3btwAALi4uAjM6dHW1oa1tbVY26lN5V9oBQUFyM7OhrS0NAwMDBAZGcl/LSoqCmlpaXB0dESLFi345U2bNsX48eOF1nvjxg20bt0aw4YNEyj/+eefhZatSxu8c+cOioqK4OTkJPD5tG3bVuQvUlGioqLw7t07jBw5kn8GEACaNWsmcl/qoi77wjsze+3aNZSUlHzWdisLCQmBsrIyxo0bJ1A+btw4KCsri5x0Lk575MV7+fLlT57ToqioiIkTJwptW1FRUWBYpy77oKioiJiYGLx+/fqTYpJE+3v79i1GjRqFNm3a8MsbNWqEqVOngsvl8vtGGRkZ2Nvb49mzZ4iNjeUv++DBA7x9+1bgbNXZs2chJSUFBwcHkftRUFDAH87m6datG6ysrOpeadWQlpYWGu4T1Tf+888/6Ny5M3r06CEQZ2lpKaysrBAWFiZylKaysWPHorS0FGfPnuWXJScnIzQ0FPb29mjcuDGAuh+3KveJRUVFyM7ORlFRESwtLREbGytyXpyofu17Q8OOYkhKSoKenh7/FC+PqqoqmjdvLlDG+0K7urpWu76MjAyBZZcuXVrrsjzNmzcXObSopaWFq1ev8k/RVp0vJScnByUlJZHbiI+PB5fLRUBAAAICAkQuo66uzv//8vJy+Pr6Ijg4GG/evBGamyHqKrKahlAqr5uHF2tOTk617xNnHSkpKfy/eXOMOnfuLDK+27dvi7WtmiQmJmLr1q24c+eOUD3wEtbKsYiqF1FlycnJ0NfXF5oIrqKi8lltsKY6qTz8URPefMPPWUd16rIvQ4cOxT///IM9e/bg8OHDMDAwgLW1NYYOHSr2sLIoycnJ0NXVhaysYHcpKysLDQ0NkXNqxGmPTk5OuHbtGlavXo1NmzbBxMQEvXv3xrBhwwSS2Jqoq6sLJHnAx6RDXV1dYB5oXfZh6dKlWLRoEezt7aGurg4LCwvY2trCzs5OrAsR6rv98dbRpUsXode6du0KAAJ14eDggIMHDyI4OBgeHh4APg458hKzyvvBGMNPP/1U637waGhoiBWzuFRVVflJD4+ovjE2NhbFxcXo2bNntevKzs5Gu3btqn3dwsICGhoaCAgI4M+7PHPmDBhjAkOOdT1uZWZmYtu2bbh27ZrIuYi5ublCw/Zfuh4bIkq+vjBeIuLt7Q1VVVWRy/A6Zt6yixYtQvfu3UUuW906alP1TI65uTmOHTtWY8zDhw+vdp5C5Q5g/fr1OHbsGIYMGQJ3d3coKytDTk4OUVFR2LRpE7hcrtD75eXlq421alIrKrbaNISr0woKCuDk5ISioiJMnjwZHA4HTZs2hbS0NPbu3Yv79+/XSxx1aYOSUDkJrarqGaC67EujRo1w6NAhREZG4t9//8Xjx4+xY8cO/Pnnn9i8eTMGDBjwhfagduK0x5YtWyIgIACPHz/GvXv38OjRI6xbtw4+Pj7w9fWFkZFRPUQqrH///rh+/Tpu3bqFR48e4d69ewgICICpqSkOHToklOxV1dDbn7a2Nrp3746zZ89i3rx5KC4uxuXLl9GrVy+BH7aMMUhJSWHfvn3V9lFVE77qzu5/KnH7RsYYOByOwNyxqsRJ6MeOHYsNGzbg+fPn0NHRQVBQEHR1ddGtWzeh7Ypz3GKMwc3NDbGxsXBxcYGuri6aNWsGGRkZBAYG4ty5cyKPF1+6HhsiSr7EoK6ujjdv3qCiokLgy5CWliZ0doOXsbds2bLW08+8ZRUUFMQ+VZ2bm4v09HShs1+xsbFQUVHhn+I9dOiQwOtVz45U1rFjR0hJSaGsrEysOIKDg2FmZoatW7cKlIs7RCgpHTp0APBxYnbVzl+cq8tqExoairS0NKxduxajRo0SeG3btm0Cf/POxIjarqgyNTU1vHnzBlwuV+DAnpmZ+VltsHKdVP3VXHlYpia8uoyLixN6TdQ6eMOsHz584G8f+HiVXXp6Ojp16sQvq8u+8Ojr60NfXx/Ax6uUHRwcsG3bNn7yVVPyJ4q6ujri4+NRXl4ucOaovLwcCQkJn5VIyMjIwMLCAhYWFgCA6OhojBo1Crt374avr2+t709KSkJpaalAQlRaWoqkpCSBs0l13QclJSWMGDECI0aMAGMMmzZtwv79+3Ht2rUazwQB9d/+eOuIiYkReo1XVnX/HBwcsG7dOty/fx/p6ekoKCgQ+uGpoaGBf//9F+3bt//sM7hfW6dOnZCdnQ1LS8vP+iHq6OiIrVu3IiAgAP369cPbt2+FprnU5bj16tUrREdHY+bMmZgzZ47Aa1WvRP3RSP50wTegX79+yMjIwN9//y1Qvm/fPqFlf/rpJzRq1Ag+Pj4ix9jz8vJQWloK4OPZKRUVFezbt0/k8FpxcbHI8fCqnXJISAji4+PRv39/fpmVlZXAP11d3Wr3r2XLlrCxsUFISIjQHAbg46+XrKws/t/S0tJCZ6QKCwsb/J2abW1tAXy8SWrlX1uvXr3CnTt3xF5P06ZNRX5evMS8at3cuXMHERERAmW6urpo3bo1/+pInoKCApw8eVJk7Onp6Th37pxA+YEDB4SWrUsb7NWrF+Tl5XH8+HGBe6+lpqYKzP2oSY8ePdC2bVucOXNGoJ3k5+eL3Bde582b98Nz+PBhoV/BddmXytvmadu2LZSVlQXqmPcDpXJZTfr374+srCyhg8Xp06eRlZUl8L2rC1Hxdu7cGY0bNxY7tvz8fPj7+wuU+fv7Iz8/XyAucfehoqJC5HC5jo4OAAjVo6jvgSTaX/v27XHmzBmB6RZlZWU4cOAApKSk0K9fP4H32NvbQ1ZWFsHBwQgODkazZs2Elhk+fDgAYMuWLUK30AGEhxw/RdOmTfHhw4c631ajKgcHB6Snpwv96OYRN1ZlZWX0798f586dw/Hjx6GgoCA0964uxy1eIlh1/16/fi3yVhM/EjrzJYapU6fi3LlzWLFiBaKiotClSxc8fPgQT58+5V+qztO2bVusWrUKy5cvx5AhQzB8+HCoqakhKysLr1+/xtWrV3H+/Hl06NABTZo0gbe3N2bOnInBgwdj1KhR6NSpE3JzcxEXF4eQkBD8+eef/F/FwMdEKSQkBGlpaTA3N+ffaqJVq1aYNWvWJ+/jqlWrMHHiRDg7O2PEiBHQ0dEBl8tFUlISrl27BgcHB8yePRsAMGjQIJw6dQpz586FlZUVMjIyEBgYWO2csoaia9euGDduHE6dOgVXV1cMGDAAWVlZ8Pf3R/fu3REVFSXWWREDAwMEBARg27Zt0NLSgrS0NGxtbWFiYoLWrVvD29sbKSkpaNu2LV6+fIng4GBwOByBCcyysrJYvHgxFixYgDFjxmD06NGQkZFBUFAQlJSUkJycLBDLtGnTcO7cOSxduhSRkZH8W02Eh4d/Vhts0aIFfv31V3h7e2P8+PFwcHBAUVERTp48We18pqpkZGSwZMkSzJ07F2PGjMHYsWP5wwpKSkp4+/atwPJWVlbQ1NTEjh07kJOTgw4dOiAsLAwRERGftS+7d+/G3bt30bdvX3To0AGMMdy4cQNxcXGYOnWqwOcHgH+Ze+PGjdG1a1dwOByR+zd16lRcunQJXl5eePHiBbp3746XL18iICAAmpqaAuuuixUrViA1NRXW1tZo3749iouLcfHiRRQUFGDEiBFiraNjx47YuXMn/vvvP/To0QNRUVEIDAxE586dBe6XJu4+FBQUwNraGnZ2dtDR0YGysjKSk5Nx4sQJtGjRgv8DBgAMDQ0RGhoKX19ftG/fHlJSUhg6dKhE2t+KFSswa9YsjB49mn87m4sXL+Lp06dwd3cXmkOkoqKC3r174/LlyygpKcHo0aOF5lbp6+tj9uzZ8PHxgYODAwYNGoQ2bdogLS0NUVFRuH37tsiLi+rCwMAAN27cgJeXF4yMjCAjIwNLS8s63WYD+HgR0b1797Bhwwbcv38flpaWUFRUxNu3b3H//n00atSo2mknVY0bNw4XL17EjRs34OjoKDQfqy7HLS0tLXTt2hX79+9HcXExNDU1ER8fj1OnToHD4SAqKqpO+/k9oeRLDC1atMDx48exfv16/tkvc3NzHD16VOSk0lGjRkFDQwMHDx7EqVOnkJeXByUlJWhqauLXX38VGDLs3bs3AgIC4Ovri3/++QfZ2dlo3rw5OnbsCFdXV6H7DzVp0gRHjhzB2rVrsXnzZjDG0Lt3b3h6en7y/DAAaNeuHQIDA7Fv3z5cv34d//zzDxo3box27drB1tZWYKhhyZIlaNq0KS5duoRr166hXbt2GDduHPT09GqcZNsQrFy5EqqqqggICIC3tzc0NTWxcuVKPHv2DFFRUTXOTeOZN28ePnz4AH9/f+Tm5oIxhmvXrqFDhw7Yv38/Nm7cCD8/P5SXl0NXVxf79u1DQECA0NVjvF/fu3btwo4dO9CqVSuMHj0a2tramDVrlsDBQFlZGf7+/vD29kZgYCCkpKRgYWGBI0eOYPTo0UJx16UNurm5oUmTJjh06BA2b96Mdu3awc3NDc2aNatxUm1lgwcPxo4dO7Bz5074+PhARUUFjo6OMDMzg5ubm8CyMjIy2L17N9asWQM/Pz/IycmhV69e8PPzw4QJE4TWLe6+9O/fH+np6bh06RIyMjIgLy+PTp06Yc2aNRg9ejR/fSYmJliwYAFOnjyJFStWoLy8HLNmzao2+WrWrBlOnDiBHTt24Pr16zhz5gxUVFQwfvx4zJ49W+jgJK4RI0bgzJkzCAoKQlZWFhQVFdGlSxfs2LEDgwYNEmsdbdu2xbZt2+Dt7Y3z589DTk4O9vb2WLx4scBVZuLug7y8PCZPnozQ0FCEhoaioKAAqqqqsLOzw4wZMwSuJly5ciW8vLywZ88eFBQUAPh40QNQ/+3Pzs4Ohw8fxu7du3HgwAGUlZVBS0sLa9asEZgsXpmjoyP/Cujqkt1Zs2ZBV1cXx44dw9GjR1FYWAgVFRV07doVy5YtEyu2mri6uiIpKQmXL1/GyZMnweVycfTo0TonX3Jycti7dy/8/f0RHBzMv2mwqqoq9PT0ar3nWGWWlpbo1KkT3rx5I/C9qUzc45aMjAz27t0Lb29vBAUFoaioCF27doW3tzeio6N/6ORLin3u+U5SbyZNmoSUlBSRN+Ekn8fd3R33799HWFhYjZNc68PBgwfh7e2NU6dOwdDQsMZlefM8xo0bJ/LmtpL24MEDuLi4YN26dRg5cqSkw/mu2NnZQU1NTewzGoSIa+jQoaioqBB5J3/yZdCcL/JDETUHJTo6Grdv34alpWW9Jl6lpaVCc0kKCgpw/PhxKCkp8efZ8IiKnTf/r+pdqQkh5FOEhoYiJiZG5JMLyJdDw47khxIUFITg4GDY2NhAWVkZcXFxOH36NOTk5ISuxvnakpKSMG3aNAwdOhQdOnRAeno6goKCkJycjFWrVgld0j9t2jSoqanx5+Pdv38fN27cgJGR0SdP+iaEEOBj0pWUlIS9e/dCWVmZkq+vjJIv8kPp0aMHrl69imPHjuHDhw9o2rQpLCwsMGvWLKEzTV+bsrIyDA0NcfbsWWRmZkJWVhYcDgfz588XuuM/8PGKx7///hshISEoKSlBmzZt4Obmxn+WIiGEfKpdu3YhLCwMWlpa8Pb2/uS5jERM9fUcoy+pf//+As8tq8rf35+Zm5t/lW3PnTuXzZ49+6usu6GKiYlhHA6HvXr1StKhfHN27tzJBg0aJOkw2NKlS5mbm5ukw2CMNbzv0JgxY9j69eslHUaDd+vWLcbhcFh+fr6kQxHpa/b7nyo6OpqNGjWK6erqNoh+oK6o76+7d+/eMQ6Hw8LDw2tcrsY5X9ra2jX+8/T0rK8csU4cHR1x4cKFz1pHbGwstLW1P/n5Zt+qsWPHwtvbW6BMQ0MDd+7cEfkIkPpSUlICbW1t/tVJhHwpvr6+mDlzpqTDqLNNmzZ90kUMn/o+UndbtmyBkpISLl26JHQ/toamofb99eH27dvQ1tbmX7VbH2ocdqx848mbN29i+fLlAmXVXZZfVlYGOTm5LxRi3cnLy4t1ywAiHhkZGZHPkySkoapLH9TQ709HJKvqEwTq4s2bN3B0dPysZ4tKEvX9X0+NZ75at27N/9esWTORZbwzRBcvXoSzszP09PT498J69OgRJkyYAH19fdjY2OD3338XyCzHjh2LtWvXwtvbG+bm5ujVqxf/3lU8aWlpmDFjBvT19WFnZ4fg4OBad+rEiRMCNybl/dL7+++/YWdnB2NjY8yZM6fau0iXlJTw59zY29tDW1tb6Cnr+/fvh7W1NczNzbF8+XL+HZsBgMvlYvfu3bCzs4O+vj7s7e1rPRNXXl6ONWvWwNTUFBYWFvD29saSJUsEtivql8m8efMEJorXtm3GGLZt24a+fftCV1cX1tbW/PvVzJs3DxERETh48CD/7GZ6errIs4ChoaEYNWoUfx0bNmxAWVmZQKy1fbYXLlzAsGHDoK+vD3Nzc0yaNAnZ2dki68fOzg7Ax1tCaGtrY/DgwQA+nqF0d3eHlZUVjIyMMHLkSKG71b9//x7Tpk3jt6F//vkHAwYMEHhSQExMDMaPHw89PT0MGTIEd+/ehY6ODs6fP89fJiUlBXPmzIGpqSnMzc3h7u4u8MBe4OO8iZ49e8LIyAhLlixBSUmJyP2pbN26dRg4cCA/vq1btwq0p6SkJMyYMQNmZmYwNDTEkCFDcOXKlWrXJ6otVb1zvDhttLb95X2vjh8/jj59+sDAwKDG71V1xImltjrixXLq1CnY2dlBT08P5eXlYrXDqt+rXr16wdfXF0uWLIGRkRFsbGxw9OhRgXjEaS+VXbt2Dfr6+kJPrFi3bh3/XkrZ2dmYP38+LC0toaenhwEDBlR7tp4zzSQAACAASURBVOTEiRPYt28foqKi+N9V3raTkpLg7u4OIyMjfl/Hu/N7Te/z9fXFsGHDYGhoCBsbG6xcuVLkEzZqUlvdVXcGm3eft8rLnD59GtOnT4eBgQEGDx6Mx48fIzk5Ga6urjA0NMTIkSPx6tUroRguX76MgQMHQk9PD1OmTBG6ye+VK1fg4OAAPT099OvXDz4+PgJ9V69evbBnzx4sWrQIxsbG1d5rrKKiAtu3b0efPn2gq6uLESNG4NatWwL7EB8fjy1btkBbW7vax0VR319z38+ry5MnT+Lnn3/m9wFV+4gXL15g0qRJ0NfXh4WFBZYtWybQfnmvGxkZwcjICA4ODnj8+DFiY2Mxbdo0AICxsTG0tbXx22+/iVWnABAeHo4RI0ZAT08PI0eOFP/eZeKOY168eJFxOByhct6YcL9+/VhISAhLTExkqamp7NmzZ8zAwIAdOXKEJSQksCdPnrCRI0ey+fPn8987ZswYZmxszHbu3Mni4+NZcHAw69atG7ty5Qp/mUmTJrHhw4ez8PBw9uzZMzZ+/HhmYGBQpzlfGzduZIaGhmz27Nns1atX7NGjR8za2pr9/vvv1a7j0aNHjMPhsAcPHrC0tDSWk5PDGPs4X8XY2JitWrWKxcTEsJs3bzJDQ0N26NAh/nvXrVvHhgwZwv7991+WmJjIzpw5w/T09Njdu3er3Z6Pjw8zMzNjly9fZjExMWzFihXMyMhIYJ6OqLkpVefP1Lbt4OBgZmZmxm7fvs1SUlJYREQEO378OGOMsdzcXObo6MhWrlzJ0tLSWFpaGquoqBAa909KSmK6urrMy8uLxcTEsJCQEGZhYcG2bNkiEGtNn21ycjLr3r07O3bsGEtKSmLR0dHsxIkTLCsrS2T9vH37lnE4HPb333+ztLQ0lpmZyRhjLDIykp06dYq9evWKxcfHs+3btzNdXV2WmJjIf6+zszNzdHRkERER7Pnz58zJyUmgDZWVlbEBAwYwNzc39vLlS/bo0SM2cuRIpq2tzc6dO8cYYywvL4/Z2tqy5cuXs+joaBYTE8MWLlzI+vfvz0pKShhjjAUFBTFdXV12+vRpFhcXx7Zv386MjIxqnevh4+PDnjx5wpKSkti1a9dYr1692K5du/ivu7q6smnTprHo6GiWmJjIbt68+dltqbZ2Is7+8r5Xrq6u/HobOHAgmzNnTo37W9c2K04d8WKZNm0ae/HiBXv58iXjcrli9TFVv1dWVlbMwsKC+fv7s4SEBLZ//37G4XDY8+fPGWPitZeqSktLmbm5OQsMDOSXVVRUsF69erEjR44wxhhbtmwZGzlyJIuMjGRJSUns3r177PLlyyLXV1RUxLy8vNjw4cP539Xi4mJWXl7OhgwZwpycnNjz58/Z06dPmaOjIxs3blyN72OMsQMHDrD79++zpKQkFhoayn766Se2bNky/jbFmfNVW90VFxczDofDrl+/LvS+Y8eOCSxjY2PDLly4wOLi4tjs2bNZ7969maurK7tx4waLjY1lkydPZiNHjuSvw9/fn/Xo0YONGjWKf7wYO3YsGz16NH+Zq1evMhMTExYUFMQSExPZ3bt3Wb9+/QT6LisrK2ZiYsIOHTrE3rx5wxISEkTu6549e5iJiQk7f/48i42NZRs2bGA6OjosJiaGMcZYWloas7W1ZVu2bGFpaWmsoKBA5Hqo76+57+e1BwsLC4G+tVu3biw6Opox9rG/srS0ZHPmzGGvXr1i9+7dY3Z2dszDw4O/ngEDBjBPT08WGxvLEhIS2KVLl1hERAQrLy9nZ8+eZRwOhyUmJrK0tDSWl5cnVp3m5uYyMzMz5uHhwV6/fs1u3rzJBgwYINacry+WfPn5+QmUz5kzh61atUqgLDw8nHE4HP6OjRkzhjk7OwssM2HCBLZ69WrGGGMvX75kHA6HPXv2jP96fHw843A4n5R8VW78W7duZUOHDq12HdVNNJw7dy7r168fq6io4JctWLCATZ8+nTHG2IcPH1iPHj1YRESEwPtWrFjBZs6cWe32zM3N2YEDB/h/l5eXM1tb2zp9AcXZ9u7du9mwYcNYeXm5yDhEbaNqXfAaJJfL5S/j7+/P9PX1WWlpKX89NX22YWFhTFtbm6WlpVVbJ5VV12mLMnz4cLZ//37G2P+1oRcvXvBff/PmjUAbCgkJYT169GAZGRn8Ze7du8c4HA7/YOrn5yfUXkpLS5mhoSG7evUqY4wxBwcH5uXlJbDM+PHj6zzR9tChQwLbGjhwYI3tvara2pI47USc/d24cSPr0aMHe//+PX+Zu3fvMm1tbZaSklJtfHVts6JUraONGzcyXV1dlp2dLbBcbe2Qt0zV5MvT01PgPX369OG3KXHaiygrV65krq6u/L/v3r3LdHR0+OuZMmUK++2336p9f1UbN25kjo6OAmXXrl1jOjo6LDU1lV/G+/4+evSo2veJEhISwgwNDfl/i5t81VR3dUm+fHx8+K9HRkYKHWeqxuPv71/t8YK376NGjWL79u0T2Pa5c+eYmZmZQCziXBBibm7OfH19BcrGjBnDli5dyv+7tovDeOuhvr96vPYgqm/l1fXRo0eZubk5Kyoq4r9+69Ytpq2tzd6+fcu4XC7T1dVl58+fF7kNUW1bnDo9cuQIs7Cw4P+AYYyx06dPi5V8fbFbTVR9cHNUVBRSU1MFHkbN/v9px8TERP5l/VUfn6OqqorMzEwAH590Lycnhx49evBf19DQEHr+mzjU1dUFHrdReTt11bVrV4Enx6uqqiI+Ph7AxweGlpWVYfLkyQLvKSsrq3bSYkZGBnJycgTuZi4jIwNdXd06TQAUZ9tDhw7F8ePH0a9fP1hbW6N3796ws7Or0xy92NhYGBkZCTx70MTEBMXFxUhOToampiaAmj9bPT09mJqaYvDgwejVqxd69eqFgQMH1vmzzc/Ph4+PD27fvo309HRUVFSgpKQE7969A/CxDTVq1AjdunXjv6djx44C24mLi4OamprAIz14z//jiYqKQlxcHIyMjATKi4qK+ENxcXFxQo/SMTQ0rPUigXPnzsHPzw9JSUkoLCxEeXm5wOfh4uKCtWvX4vr16+jZsycGDhyI7t27i1yXOG1JnHYizv4CgJqamsBjrQwNDcEYQ1xcHBISEgQmsq9fv17osTnifl9qqyMA6NChg8j5WzW1w+rU1i/V1l5EGT58OJycnJCWlgZVVVWcPXsWVlZW/PVMnDgRHh4eiIiIgJWVFezs7GBqalrreiuLjY2FmpqawGOAtLS0oKSkhNjY2BrXd+fOHfj6+iI+Ph75+fngcrkoLi5GTk5OnebFfUp917aeVq1aVVuWlZWFpk2bAkC1x4vY2FiYmJjg5cuXeP36NXbu3MlfhrefHz58QIsWLQAIH8+qyszMRE5ODoyNjQXKTUxMEB4eLvY+Ut8vft9f9WkfhoaG/LqOjY2Fjo6OwFxvExMTMMYQGxuLdu3awdXVFQsXLsRff/0FS0tLDBo0SOiZn5WJU6e87VZ+FFxtTyXh+WLJl4KCgsDfXC4XTk5OmDhxotCy7dq14/9/1Q9eSkpKaH7KlyArK7irn7OdmmLm/Xffvn1CExU/9yIEaWlpoafDl5eX878I4mxbXV0dISEhuHv3LkJDQ7FmzRrs2bMHJ0+eFHqw7Keo/KWsqZ7k5ORw9OhRhIeH4+7duzhx4gQ2b96MEydOQEtLS+zt/fHHHwgLC8PChQvRsWNHyMvLY968eQLzgb4Exhj09fWF5l0A+KQfAzwPHjzAokWLMHfuXPTs2RPNmjXDpUuXBA4OTk5OsLW1xa1btxAaGooxY8Zgzpw5mD59+idtU5x28iX218jISODHF+9gWddYxKkjQLgPqroeHnG++6L6i6rfvboyNjaGmpoaLly4gAkTJuDKlStYtWoV//X+/fvjxo0b/M956tSpGDFiBFavXv1Z2+Wp6aHxCQkJcHd3h7OzMzw8PNCiRQtERERg8eLFAvN5xFFT3fFiENWP1bQe3vtElYnbjzPGwOVyMX/+fPTr10/o9cr3tar8Q70hoL7/8+OaP38+HB0dcfv2bdy5cwc+Pj5Yu3Ythg8fLvJ9X/NYDnzFxwv16NEDMTEx6NSpk9A/ca8c6dy5M8rKygSebP/mzZtqJ+Z9SbwY65qgcTgcyMrKIjU1VWi/27dvL/I9rVq14nd2PBUVFXj+/LnAci1btuRPnOXFVnnCqbjblpeXR79+/bB8+XKcPHkSL168QGRkJICPjarqI2+q0tLSQnh4uEBnEBYWBnl5+Tpd1SMtLQ0TExPMmTMHZ86cQfPmzXHx4kWRy8rIyEBKSkootrCwMIwaNQoDBgyAtrY2WrdujeTkZP7rnTt3RmlpqUA9JSUlCbShzp07IyUlBVlZWfwyXn3w6OjoICEhASoqKkJ127x5c/56Kn+GAIT+rurJkydQV1fH9OnToaenBw0NDaSkpAgt1759e0yYMAE7duyAu7s7Tp06JXJ94rQlcdqJOPsLfJyUX7lNRkREQEpKCp07d4aCgoLA+3hnJyoTJxZx66i+iNNeqmNvb49//vkH169fB5fLFXoygYqKCkaOHImNGzdi5cqVCAgIqLYPkpOTE3pNS0sLKSkpeP/+Pb8sNjYWOTk5/AObqPdFRkZCSkoKnp6eMDQ0hKamJlJTU8Xap7po1KgRFBUVBdrMu3fvkJOT80XWX93xQktLC9LS0ujevTsSEhJEHpfqcqNiFRUVKCkp4cmTJwLlYWFh6NKli9jrob6/9r6fR1TfyjsDpaWlhRcvXgg8gi0sLIzfF/F07twZrq6u2L9/P+zt7REQEADg/5Kpyt8LcepUS0sLL1++FPix//TpU7Hq4Kvd4X7GjBkYP3481qxZg1GjRkFBQQGxsbG4c+cOVq5cKdY6unXrxr+acNWqVZCVlcXatWvr5TYSrVu3hpycHG7fvo3WrVujcePGYt3xV0lJCS4uLvjjjz9QXl4OY2Nj5OfnIzw8HI0bN672KfEuLi7Ys2cP1NXVoampiaNHjwp1SJaWlti2bRtu3boFdXV1+Pn5CZzOF2fbp0+fhoyMDPT09KCgoIDg4GDIycmhY8eOAD4OI0VERCAlJQUKCgoiz3I4Ozvj+PHj+OOPPzBhwgTEx8dj+/btcHV1FfsXwePHjxEWFsYfdnn27BnS09Or7bhkZWXRtm1bhIaGwsDAAI0bN0bz5s2hoaGBy5cvw8bGBgCwbds2gV/RvDa0bNkyrFy5ErKysli/fr3AWZK+ffuiffv2WLx4MebPn4+CggJs3rwZUlJS/F9Njo6OOHz4MGbOnInZs2ejTZs2ePfuHa5cuYIpU6ZATU0NLi4uWLlyJbp37w5jY2OcP38e0dHRAsNyVfESiQsXLkBXVxc3b94UupLRy8sLdnZ26NSpE3Jzc3H37t0aO/ja2pI47USc/QU+dlqenp5YuHAh8vPz4eXlhYEDB1b7Q6MqcWIRp47qkzjtpTrDhw/Hrl27sHv3bgwYMECgHW7ZsgUGBgbo0qULSktLcfXqVWhqagpMcahMTU0NiYmJ/DamqKgIGxsbaGhoYMGCBfD09ERFRQVWrVoFY2NjmJiYVPs+DQ0NlJaWws/PDzY2NggLC/tq96WytLTEsWPHoKenBy6Xiy1btnzyrRyqkpOTw+rVq7Fs2TLIysri999/h66uLn+4debMmZgzZw5UVVUxaNAgSEtL49WrV4iOjoaHh0edtvXzzz9j79696NChA7p164bAwEBERUVh3bp1dVoP9f019/0858+fF+hbw8PD+VclOjo6YteuXfD09MT//vc/ZGVlwcvLC0OHDkW7du2Ql5eH7du3Y+DAgVBTU8P79+8RHh4OKysrAB+nLADAjRs3YG1tDXl5ebHq1MHBAT4+Pli2bBlmzJiBlJQU7N+/X6x6+GrJl66uLvz8/LB9+3b+0GOHDh34twgQ16ZNm7B8+XI4OzujVatWmDNnDtLS0r5GyALk5eWxZMkS+Pr6YuvWrbCyssKBAwfEeu/ChQuhqqqKPXv2IDk5Gc2aNYOOjg7/clZRZsyYgaysLCxevBjS0tIYOXIkBg0aJLCvEyZMQExMDBYuXAgZGRlMmjQJffr0Efi1Utu2mzVrhoMHD2Lt2rXgcrno0qULdu3axZ8jMn36dHh6emLIkCEoLi4Wum0D8PFz9PX1xcaNG3Hy5Em0aNECI0eOxKxZs8SqH14cDx8+xOHDh5Gfn4/27dtj3rx5NbaPJUuW8Leprq6OS5cuYcWKFVi6dCnGjRuHli1bws3NTegB1Js2bcKyZcvg5OSE1q1bY968eYiPj+efapeVlcWuXbuwfPlyjB49Gurq6li8eDHc3d35BwVFRUX4+/tj06ZNmDVrFgoKCqCqqsofBgM+dgApKSnYuHEjSkpKMHDgQDg7O9eYKAwePBgRERFYvXo1SktL0adPH8ycORMbN27kL1NeXo5Vq1bh/fv3UFRUhJWVFZYsWVLtOsVpS7W1E3H2FwA0NTVha2uLadOm4cOHD7CxsYGXl1e1sYlSWyzi1FF9Eqe9VEdTUxN6enp49uwZFi1aJLTeTZs24e3bt5CXl4eRkRF8fHyqXdeQIUNw/fp1ODs7Iy8vD1u2bMHQoUOxZ88e/PHHH3B2doaUlBSsra2xYsWKWt+3aNEi7NmzBxs3boSpqSkWLFiAhQsXfl5libB8+XIsXbqU/3309PQUecuIT9G0aVNMmTIFHh4eeP/+PUxMTLBp0yb+6/369eMnv76+vpCTk4Ompma1P4pr8vPPP6OoqAhr165FdnY2OnfujJ07d9Z56Iz6/tr7fgCYM2cOzp49i1WrVqF169bYuHEjfy6voqIiDhw4wL91i4KCAvr378/vJ2VlZZGZmYlFixYhIyMDLVu2hJ2dHRYsWADg45Csu7s71q9fj6ysLIwdOxZeXl611mnz5s2xe/dueHl5wcHBAV26dMHChQvFummzFPvciQzkq1m2bBlSU1PFTvqIeNLS0tCnTx/s2bMHffv2FblMREQExo4di3PnzqFr1671G+A3YtOmTbh37x7OnDkj6VAkjtoL+ZKo7/8/JSUl0NfXx549e2BrayvpcL4YerA2+e7duXMHpaWl6Nq1K9LT07F582a0bduWf8oZAC5evIgWLVqgY8eOSEpKwtq1a6Gvr08HUiIStRdCyOeg5It890pLS7F582YkJyejSZMmMDIygre3t8AQUX5+PrZs2YLU1FQoKSmhZ8+eDfbZpUTyqL0QQj4HDTsSQgghhNSjr3arCUIIIYQQIoySL0IIIYSQekTJFyGEEEJIPaLkixBCCCGkHlHyRQghhBBSjyj5IoQQQgipR5R8EUIIIYTUI0q+CCGEEELqESVfhBBCCCH1iJIvQgghhJB6RMkXIYQQQkg9ouSLEEIIIaQeUfJFCCGEEFKPKPkihBBCCKlHlHwRQgghhNQjSr4IIYQQQuoRJV+EEEIIIfWIki9CCCGEkHokK+kACCGEEEK+hJy8YkTGZCAyJgP6XVpBv0srKDWTl3RYQqQYY0zSQRBCCCGEfI6cvGKs2BuKhHe5/DKNds3x+4yeDS4Bo2FHQgghhHzzImMyBBIvAEh4l4tnsZkSiqh6lHwRQggh5JsXGZMhuvy/9HqOpHaUfBFCCCHkmxYVmwHN9i1EvqbftXU9R1M7Sr4IIYQQ8k0qr+Di4NkoeO66C/U2itBo11zgdY12zaGnpSKh6KpHE+4JIYQQ8s3JyCnChmOP8TIhC2qtm2KpqzmaN22EZ7GZiPwvHfpdW0NPS6XBTbYHKPkihBBCyDfmyas0bD4ehtyCUlgbtMfssYZoIi8n6bDERskXIYQQQr4JFVyGk1de4dTVV5CRlsLU4boY0ksTUlJSkg6tTugmq4QQQghp8LLzirH5eBgi/suAqnITLJ5kCk7HlpIO65NQ8kUIIYSQBu15bAY2+j1GVm4JzHXaYt4EIyg2aSTpsD4ZDTsSQgghpEHichnO3IzBsYsvAQCTh3SHY98u39wwY1V05osQQgghDU5eYSm2nniCRy/eQ7m5PBZNMkWPzg3vthGfgpIvQgghhDQorxOzsf7oI6RnF8Gwa2vMdzKBUrPGkg7ri6HkixBCCCENAmMM5+7E4+DZ56jgMkwcqI2xA7QhI/1tDzNWRckXIYQQQiSusLgMO04/xd2It2ih2AgLnExgyFGVdFhfBSVfhBBCCJGo+LcfsO7II7zLKICOpjIWTTKFSgsFSYf11VDyRQghhBCJYIwh5GEi9p6JRGk5F6Nsu2DST90hI/N9P3r6+947QgghhDRIxSXl2HYyHD6nn6KRnAxWuFnAdViPL5Z4LVmyBD179sSwYcP4Zd7e3hg8eDDs7e0xc+ZM5ObmAgCys7MxadIkGBkZwcvLS2A9paWlWLFiBQYNGoTBgwfj8uXLnx0b3eeLEEIIIfUq6X0e1h99hMTUPHRVV8JiFzO0UW7yRbfx6NEjNGnSBIsXL8a5c+cAAHfu3IGlpSVkZWWxceNGAMDChQtRWFiIFy9e4L///sN///2H3377jb+eHTt2oKKiAvPmzQOXy0VOTg6UlZU/KzYadiSEEEJIvbn5JBk7/3qK4tIKDLPWhJt9D8jJynzx7ZiZmSE5OVmgzNramv//hoaGuHTpEgCgSZMmMDU1RWJiotB6AgMDcfHiRQCAtLT0ZydeACVfhBBCCKkHpWUV2B/8HBdDE6DQWBaLJpmit6GaxOIJDAzETz/9VOMyvGHJ7du34+HDh1BXV8dvv/2GVq1afda2ac4XIYQQQr6q1MwCLPrzX1wMTYBGu+bYOs9GoonX7t27ISMjg+HDh9e4XHl5OVJTU2FkZISgoCAYGRnB29v7s7dPZ74IIYQQ8tWEPnuL7SfDUVBcjgHmHTFjpD4ay335YUZxnTlzBjdv3sThw4drfUZky5YtoaCggIEDBwIABg8ejICAgM+OgZIvQgghhHxx5RVcHDn/An/fikUjORnMHW+EfmYdJRrT7du3sX//fvj5+UFBofb7iElJScHW1hYPHjxAz549ERoaCi0trc+Og652JIQQQsgXlZ5dhA3HHiH6TTbUWitiyWQzdGrXvF5j8PDwwMOHD5GdnQ0VFRXMnj0bvr6+KC0thZKSEgDAwMCAf2sJOzs75Ofno6ysDM2aNcPBgwfRpUsXpKSkYNGiRcjNzYWysjLWrVuH9u3bf1ZslHwRQggh5IsJi36PzcefIK+wFH0M1TBzjAGayMtJOqwGhZIvQgghhHy2Ci7DicvROH3tNWSkpTHNQRc/9dSodV7Vj4jmfBFCCCHks2TnFmPT8TBExmSgjXITeLqYoYu6kqTDarAo+SKEEELIJ3sWk4GNfo+RnVcCix5tMXe8ERSbNJJ0WA0aDTsSQgghpM64XIaA6//h+KWXkJKSguswHYzoo0XDjGKgM1+EEEIIqZPcglJs8Q9DWHQaVFrIY9EkU+hoqkg6rG8GJV+EEEIIEVv0myx4H32MjJwiGGurwmOiMVooNpZ0WN8USr4IIYQQUivGGP75Nw6HzkaBMQbnwd0wph8H0tI0zFhXlHwRQgghpEYFRWXYfiococ/eQUmxMRY4m8Cga2tJh/XNouSLEEIIIdWKTc6B99HHeJdZgB6dVbBokimUm8tLOqxvGiVfhBBCCBHCGMPl+2/g+/czlJVzMaZfVzgN6gYZGWlJh/bNo1tNEEIIIURAUUk5dgVG4GZYMhQV5OAx0RhmOm0lHdZ3g858EUIIIYQvMTUX648+QtL7fGh3bIlFk0yhqtxE0mF9Vyj5IoQQQggA4EZYEnYGRKCktALDe3eG67AekJOlYcYvjYYdCSGEkB9caVkFfP9+hsv330ChsSx+HWeEXgbtJR3Wd4vOfBFCCCE/sLcZ+fA+8hhxbz9As31zeE42Q/tWipIO67tGyRchhBDyg7ob+RY7ToWjsLgcgyw7YZqDHhrLyUg6rO8eDTsSQgghP5iyci4On4vCP//GoXEjGfxvlAHsTNUlHdYPg858EUIIIT+QtOxCbDj6GK8Ss6HeRhGLXczQqW1zSYf1Q6HkixBCCPlBPH75Hlv8w5BXWIa+xh3wv9EGUGhMqUB9o2FHQggh5DtXUcHF8cvR+Ovaf5CTlcZ0Bz0MsuwEKSl6KLYkULpLCCGEfMeycoux0e8xnsdmoq1KEyx2MUOXDkqSDuuHRskXIYQQ8p2KjEnHRr8w5OSVoKdeO/w6zghNFeQkHdYPj4YdCSGEkO8Ml8vw17XX8L8cDSkpKUyx74HhvTvTMGMDQWe+CCGEkO/Ih/wSbPF/giev0tCqhTwWu5ihm4aypMMilVDyRQghhHwnXsZnYcOxR8j4UAyTbqqYN8EYLRQbSzosUgUlX4QQQsg3jjGG4NuxOHzuBRhjmPRTd4y26wppaRpmbIgo+SKEEEK+YflFZdh+8gnuP0+FUrPGWOhsAv0urSUdFqkBJV+EEELINyomKQfrjz7C+6xC6Gm1wkJnE7RsLi/psEgtKPkihBBCvjGMMVwKTYDv389RXsHF2P4cTByoDRkZaUmHRsRAyRchhBDyDSkqKcfOvyJwKzwZzZo0gsdEY5h2byPpsEgdUPJFCCGEfCPevMvF+qOPkJyWD+1OLbF4khlat1SQdFikjij5IoQQQr4B1x8nYmdAJErLKjCijxYmD9WBnCwNM36L6A73hBBCSANWUlaBvWciEfIwEU3kZTF3vBF66rWXdFjkM9CZL0IIIaSBepuej3VHHiHhXS46q7WAp4sZ2rVqKumwyGei5IsQQghpgO5EpGDHqacoKinH4J4amDZCF43kZCQdFvkCaNiREEIIaUDKyrk4ePY5zt2Jh3wjGcwcbYC+JuqSDot8QXTmixBCCGkg0rIK4X3sEV4n5kC9TTMsmWwG9TbNJB0W+cIo+SKEEEIagIcvUrHV/wnyi8pga9IB/xtlAPnGdJj+9aAWMgAAIABJREFUHtGwIyGEECJBFRVcHLv4EoE3YiAnK40ZjvoYaNERUlL0UOzvFaXUhBBCiIRkfijCRr8wRMVlol2rpvB0MUNntRaSDot8ZZR8EUIIIRLw9HUaNh0Pw4f8Uljpt8OcsUZoqiAn6bBIPaBhR0IIIaQeVXAZTl99jRNXoiEjLYUp9j1gb92Zhhl/IHTmixBCCKknH/JLsPl4GMJfp6N1SwUsnmQK7U7Kkg6L1DNKvgghhJB6EBWXiY1+j5H5oRim3dtg3gRjNG/aSNJhEQmg5IsQQgj5ihhjCLoZiyMXXgCMwWVId4yy7QppaRpm/FFR8kUIIYR8JfmFpdh2MhwPolKh3LwxFjibQk+rlaTDIhJGyRchhBDyFfyXlI31Rx8jLasQ+l1aYYGzCVo2k5d0WKQBoOSLEEII+YIYY7hwLwH7g5+jgsvFuAEcTBjYDTI0zEj+P0q+CCGEkC+ksLgMO/+KwO2nKWjetBHmTzSBcTdVSYdFGhhKvgghhJAvIOFdLtYfeYiU9AJ011DGokmmaKWkIOmwSANEyRchhBDyma4+TMTuM5EoLauAg40WJg/VgayMtKTDIg0UJV+EEEJIHeXkFSMyJgORMRngqLdEG5UmaNNSAS5DdWCp207S4ZEGjh4vRAghhNRBTl4xVuwNRcK7XH6ZRrvm+G2qJVrTMCMRA50TJYQQQuogMiZDIPECPs73ik7IklBE5FtDyRchhBBSB5ExGaLL/0uv50jIt4qSL0IIIURM5RVcdOmgJPI1/a6t6zka8q2i5IsQQggRA2MMuwIi0K5VU2i0ay7wmka75tDTUpFQZORbQ1c7EkIIIWI4fe01Qh4mIiu3GKun90RUXCYi/0uHftfW0NNSgRI9OoiIia52JIQQQmpx80kyNh8PQ+uWCtg0pw+Um1OiRT4dDTsSQgghNXgem4HtJ8PRRF4WK6daUuJFPhslX4QQQkg1ktPy8Mehh2CMYelkc3Rq27z2NxFSC0q+CCGEEBFy8kqwev995BeVYdYYAxhw6GpG8mVQ8kUIIYRUUVJWgTUHHyA1sxDjBnDQ37yTpEMi3xFKvgghhJBKuFyGLf5heJWYjb4mHeA0qJukQyLfGUq+CCGEkEoOn3+Be5HvoKulgjljDSElJSXpkMh3hpIvQggh5P87fzceQTdj0EFVEctczSEnKyPpkMh3iJIvQgghBMCjF6nwDYpEC8VGWDnVEopNGkk6JPKdouSLEELIDy8mOQcbjj2GrIw0VrhZoK1KU0mHRL5jlHwRQgj5oaVlF+L3A/dRUlaB+U4m0O6kLOmQyHeOki9CCCE/rIKiMnjtv4+s3BK42evCSr+9pEMiPwBKvgghhPyQyiu4WH/0Ed6k5mFYL02M6NNZ0iGRHwQlX4QQQn44jDHsCojA09fpMNdpi6kOenRLCVJvKPkihBDyw/nr2n8IeZgIrQ4tsMDZBDLSlHiR+kPJFyGEkB/KzSfJOHbxJVopKeC3ny2h0FhW0iGRHwwlX4QQQn4YUXGZ2H4yHE3kZbFqqiWUm8tLOiTyA6LkixBCyA8hOS0Pfxx6AMYYlkw2Q6d2zSUdEvlBUfJFCCHku/chvwSr999HXmEZZo0xgCFHVdIhkR8YJV+EEEK+ayVlFfj94AOkZhZi3AAO+pt3knRI5AdHyRf5f+3deXiU9b3+8TuZyTLZyUYSQGICIZCQhDUgVtyKKwUVwRatBbH2aMVedvlpPW5tbY/n9Kq91FOXggpt3RCq9hiVlqXaQgIYIJKEJUCiQHaykkwymXl+fwQiCMqWzDPL+/WPlzMh3OOV5Ln9zCffBwB8lstl6HevfqJdVU26dPxQzb8q0+xIAOULAOC7XnmvTBtKqpWVFqfF8/I4ywsegfIFAPBJBRv266/rKzQkIUIPLZisIKvF7EiAJMoXAMAHbSmv1QurShQdEazH7pyiyLBgsyMBfShfAACfsvdAs55cvllWS6AeXpivpLhwsyMBJ6B8AQB8Rn1Tp36xtFBdDqd+PH+CRg2PNTsScBLKFwDAJxzpdOjxJRt1uLVLC2dm6aKcFLMjAadE+QIAeL0ep0v/tXyzqmradN20CzXrknSzIwFfifIFAPBqhmHoD29t17bd9Zo0ZrDunJXNkRLwaJQvAIBXe2vtHv1902dKHxqtn946URYLlzZ4Nr5CAQBe65/FB7S8oFzxMTY9cscU2UKsZkcCTovyBQDwSqX7GvX717cqLNSqxxZNUWxUqNmRgDNC+QIAeJ0DdW164uUiGYahB2+fpOHJUWZHAs4Y5QsA4FVa2rv0+JJCtXU4dM+cXOVlJJodCTgrlC8AgNfocjj1q5eKVNPYoXlXZuib+cPNjgScNcoXAMAruFyGnnq1WDurmjR93FDNvzrT7EjAOaF8AQC8wrL3yvTvkkPKSovTfbfkcZYXvBblCwDg8Qo27Neq9RUakhChhxZMVpDVYnYk4JxRvgAAHm1Lea1eWFWi6IhgPXbnFEWGBZsdCTgvlC8AgMfae6BZTy7fLKslUP+5MF9JceFmRwLOG+ULAOCR6ps69YulRepyOHX//AnKHB5rdiSgX1C+AAAep8Pu0C+WFupwq10LZ2ZpWk6K2ZGAfkP5AgB4lB6nS/+1bLMqq1t13bQLNeuSdLMjAf2K8gUA8BiGYei5lSXaurtek8YM1p2zsjlSAj6H8gUA8Bhvrd2j1UVVShsSrZ/eOlEWC5cp+B6+qgEAHuGfxQe0vKBc8TE2PXJHvmwhVrMjAQOC8gUAMF3pvkb9/vWtCgu16tFFUxQXbTM7EjBgKF8AAFMdrG/XEy8XyTAMPXj7JKUmR5kdCRhQlC8AgGla2rv0+B8L1dbh0D1zcpWXkWh2JGDAUb4AAKbocjj1q5eKVN14RPOuzNA384ebHQlwC8oXAMDtXC5DT71arJ1VTZo+bqjmX51pdiTAbShfAAC3W/Zemf5dckhZaXG675Y8zvKCX6F8AQDc6v0N+7VqfYWGJETooQWTFWS1mB0JcCvKFwDAbbaU1+r5VSWKjgjWY3dOUWRYsNmRALejfAEA3GLvgWY9uXyzrJZA/efCfCXFhZsdCTAF5QsAMODqmzr1i6VF6nI4df/8CcocHmt2JMA0lC8AwIDqsDv0i6WFOtxq14LrszQtJ8XsSICpKF8AgAHT43TpyeVbVFndqmsvStXs6elmRwJMR/kCAAwIwzD03MoSFe+q08TRg/X92WM5UgIQ5QsAMEDeWrtHq4uqlDYkWj+7baIsFi45gET5AgAMgI+2HtDygnLFx9j0yB35soVYzY4EeAzKFwCgX5Xua9RTr21VWKhVjy6aorhom9mRAI9C+QIA9JuD9e164uUiGYahB747SanJUWZHAjwO5QsA0C9a2rv0+B8L1dbh0N1zcjVuVKLZkQCPRPkCAJy3LodTv3qpSNWNRzT3ygzNyB9udiTAY1G+AADnxeUy9NRrxdpZ1aTp44bq1qszzY4EeDTKFwDgvCwvKNO/tx9SVlqc7rslj7O8gNOgfAEAztn7G/Zr5boKDUmI0EMLJivIajE7EuDxKF8AgHOypbxWz68qUVR4sB5dNEWRYcFmRwK8AuULAHDW9h1s0X//abOslkA9vDBfyfHhZkcCvAblCwBwVhqaO/X4kkLZu526f/4EZabGmh0J8CqULwDAGeuwO/T4kkIdbrVrwfVZmpaTYnYkwOtQvgAAZ6TH6dKTy7eosrpV116UqtnT082OBHglyhcA4LQMw9BzK0tUvKtOE0cP1vdnj+VICeAcUb4AAKf11to9Wl1UpbQh0frZbRNlsXD5AM4V3z0AgK/18daDWl5QrvjoUD1yR75sIVazIwFejfIFAPhKpfsa9dTrxbKFWPXonVMVF20zOxLg9ShfAIBTOlTfrideLpLLZejB2ycpNTnK7EiAT6B8AQBO0tLepcf+WKi2DofunpOrcaMSzY4E+AzKFwDgBN0Op554eZOqG49o7pUZmpE/3OxIgE+hfAEA+rhchp56rVjllYd1ybghuvXqTLMjAT6H8gUA6LO8oEz/2n5IWWlx+tEt4zjLCxgAlC8AgCTp/Y2VWrmuQkMSwvXz701WkNVidiTAJ1G+AADaUl6r51eVKCo8WI8umqqo8GCzIwE+i/IFAH5u38EW/fefNssaGKCHF+YrOT7c7EiAT6N8AYAfa2ju1ONLCmXvdur++ROUmRprdiTA51G+AMBPddgdenxJoQ632vW967I0LSfF7EiAX6B8AYAf6nG69OTyLaqsbtU1F6XqhkvTzY4E+A3KFwD4GcMw9PyqEhXvqtPE0YN11+yxHCkBuBHlCwD8zMp1FfqwsEppQ6L1s9smymLhUgC4E99xAOBHPt56UMveK1N8dKgeuSNfthCr2ZEAv0P5AgA/Uba/UU+9XixbiFWP3jlVcdE2syMBfonyBQB+4FB9u3710iY5XYYeuH2SUpOjzI4E+C3KFwD4uJb2Lj22pFBtHd26+6ZcjR+VaHYkwK9RvgDAh3U7nHri5U2qbjiim68YqaumDDc7EuD3KF8A4KNcLkNPvVas8srDumTcEN169WizIwEQ5QsAfNbygjL9a/shZaXF6Ue3jFNgIGd5AZ6A8gUAPuiDjZVaua5CQxLC9fPvTVaQ1WJ2JABHUb4AwMd8srNWz60qUVR4sB5dNFVR4cFmRwJwHMoXAPiQ/Yda9OTyzbIEBujhhflKjg83OxKAL6F8AYCPaGju1ONLCtXZ5dSPvzNBmamxZkcCcAqULwDwAR12hx5fUqjGFrsWXJ+labkpZkcC8BUoXwDg5ZxOl5780xZVVrfqmotSdcOl6WZHAvA1KF8A4MUMw9Bzq0pUvLNOE0cP1l2zxyoggCMlAE9G+QIAL7ZqXYU+LKxSWkq0fnbbRFks/FgHPB3fpQDgpT7edlCvvFem+OhQPbIoX7YQq9mRAJwByhcAeKGy/Y166rVi2UKsemTRFMVF28yOBOAMUb4AwMscqm/Xr17aJKfL0AO3T9KFKdFmRwJwFihfAOBFWtq79NiSQrV1dOvum3I1flSi2ZEAnCXKFwB4iZ4el/70frmqG47o5itG6qopw82OBOAcBBiGYZgdAgDw1Zrb7CqpaND2PQ1KGxKtURcMUtqQaAUGcqQE4I341RgA8GDNbXY9/MJGVVa39j2WmhylX941VTGRoSYmA3CueNsRADzY9j0NJxQvSaqsbtWnextNSgTgfFG+AMADGYahHXsbVFLRcMrnS/bUuzkRgP7C244A4GFqD3fo2RXb1NnVo8smDDvlx+SMTHBzKgD9hckXAHgIp8vQux/v1Q//Z6227a5XhC1I+VlJSk2OOuHjUpOjNDY9zqSUAM4Xv+0IAB7gs5pWPfPmNu2salJkWJDunD1Wl44fqoCAADW32fXp3kaV7KlXzsgEjU2PY9ke8GKULwAwkaPHpVXr9uj1v+9Wj9Olb+QN0fdnj1VMZIjZ0QAMEHa+AMAkez5v0tNvbFNldatio0L1HzflaEp2stmxAAwwyhcAuJm9u0evfrhL7/yzQi5DumrKcH3v+ixF2ILMjgbADShfAOBGn1Y06JkV21TdcERJcWH64c15yuU3FwG/QvkCADc40unQK++V6YONlQoMkGZPT9f8qzMVGsyPYcDf8F0PAANsU2mN/rByuxpb7BqeFKnF88Yp44JBZscCYBJ+2xEABkhLe5defPtTfbT1oKyWAM29cpTmXD5SQVaOWAT8GZMvAOhnhmHon8UH9OLbO9TW0a1RFwzSvfPyNDwp6vR/GIDPY/IFAP2ovqlTf1i5XVvKaxUSbNFt14zW9RenyRIYYHY0AB6CyRcA9AOXy9AHhZV65f/K1NnVo7yRCbrn5lwlxYWbHQ2Ah2HyBQDn6WB9u555c5tK9zUq3BakRd/K0hWTLlBAANMuACdj8gUA58jpdOntf+7Vqx/uVHePS1PHJusHN+YoNor7LgL4aky+AOAc7DvYoqff3Kq9B1oUExGiH9yYo4tykpl2ATgtJl8AcBa6HU69/vddWrmuQi6XocsnDtOiWdmKDAs2OxoAL0H5AoAzVLa/UU+/sU0H69uVOMime+bkaXxmotmxAHgZyhcAnEaH3aE/FZTrvQ37JUkzv5Gm264ZLVsIP0IBnD1+cgDA1yjeWadn39qm+qZODU2M0L1z8zTmwjizYwHwYizcA8AptB7p1tJ3d2jtls9lCQzQTZeP1LwrMxQcZDE7GgAvx+QLAI5jGIb+XXJIL6z6VM3tXUofGq3Fc8cpbUi02dEA+AgmXwBwVGNLp55fVaLCHTUKtgZq/tWZmnVJuiwWboQNoP8w+QLg9wzD0N83faaX3t2hI/YeZaXFafHcPKUkRJgdDYAPYvIFwK/VNB7RM29uU0lFg2whVi2YmaWr8ocrkBthAxggTL4A+CWny9DfPt6nP39Qrq5upyaOHqx75uQqPsZmdjQAPo7JFwC/U1Xdqmfe3KZdnzUpMixY379hrKaPG8KtgQC4BZMvAH7D0ePSW2t26801u9XjNDR93FDdOTtb0REhZkcD4EcoXwD8wu7PmvT0G1tVVdOmuOhQ3T0nV5PHJJkdC4AfonwB8Gn2rh795cOdevejvXIZ0jVTU3X7dWMUbgsyOxoAP0X5AuCztu+p17MrtqmmsUPJ8eG6d26exqbHmx0LgJ+jfAHwOe2dDr38t1KtLqpSYIB002Uj9O2rMhXCrYEAeADKFwCfUrijWs+t3K7DrV1KTY7S4nl5GjlskNmxAKAPR00A8AlNbXa9+NdP9a/th2S1BOqWGRm66bKRsnJrIAAehskXAK9mGIbWfXJAS975VG0dDmUOH6TF88Zp2OBIs6MBwCkx+QLgteoOd+h/V25X8c46hQZb9N1rx+jaaRfKwq2BAHgwJl8AvI7LZej9Dfu1rKBMnV1O5WUk6Ic352lwbJjZ0QDgtJh8AfAqB+ra9Myb21S2/7AibEFaNCtbl08cxq2BAHgNJl8AvEKP06W/rq/Qa6t3ydHj0rScFN11w1gNigo1OxoAnBXKFwCPV3GgWc+8sU37DrVoUGSIfnBjji7KSTE7FgCcE8oXAI/V5XDq9dW7tGp9hVwuQ9+cfIEWzsxSRFiw2dEA4JxRvgB4pNJ9jXrmza06WH9EibFhuvfmXOVlJJodCwDOG+ULgEfpsDu07L0yFWyoVECA9K1L0nTb1aMVGsKPKwC+gZ9mADzGlvJa/e9b29XQ3KlhgyO1eG6eMlNjzY4FAP2KoyYAmK6lvUtL3t2h9Z8ckCUwQDdfkaG5V45UkJUbYQPwPUy+AJjGMAz9a9shvfB2iVrauzViWIwWz83ThSnRZkcDgAHD5AuAKRpbOvXcyhIVldYoOMiiW6/O1Le+kSYLN8IG4OOYfAFwK5fL0OqiKr38f6XqsPdobHq8fjg3VynxEWZHAwC3YPIFwG0ONbTr2Te369O9DQoLtWrhzCzNyB/OrYEA+BUmXwAGnNPp0rsf79OfP9ipbodTk8ck6e45OYqLtpkdDQDcjskXgAFVWd2qp9/Yqj2fNys6Ilh3zc7RxXkpTLsA+C0mXwAGhKPHqTf+sVtvrdkjp8vQpROGatG3shUdEWJ2NAAwFeULQL/bWXlYT7+5VZ/Xtis+xqZ75uRq4ujBZscCAI9A+QLQbzq7evTn98v1t3/tk2FI116UqtuvG6Ow0CCzowGAx6B8AegXW3fV6dm3tqvucIdS4sN179w8ZafHmx0LADwOC/cAzkt7R7de+lup/r7pMwUGBujGS0folhmjFBLErYEA4FSYfAE4ZxtKDun5VSVqautSWkq07p2XpxFDY8yOBQAejckXgLPW1GrX838t0YaSagVZA/XtGaN0w6UjZOXWQABwWky+AJwxwzC0ZvPnWvruDrV3OjQ6NVb3zs3TsMGRZkcDAK/B5AvAGak93KFnV2zTtt31soVYdPu1Y3TNRRcqMJDDUgHgbDD5AvC1nC5DBf/er+UFZbJ3OzU+M1H33JSrxNgws6MBgFdi8gWgT3ObXSUVDSqpaFDOiHiNTo3VC38tUVFprSLDgrRo1lhdNmEotwYCgPNA+QIgqbd4PfzCRlVWt/Y9lpocpUWzsvXBxkp9/4axGhQZal5AAPARvO0IQJJUUtFwQvGSem+KfbjFrv/33UkmpQIA38PvhQN+rqW9Szv2Nmj7noZTPl+2v9HNiQDAtzH5AvzQwfp2Fe2oVuGOGu2sOqyMCwbpsgnDTvmxOSMT3JwOAHwb5QvwA06Xod1VTSoq7S1cB+vbJUkBAdLo1FjlZyVp8pjB+mBj1Ek7X2PT48yKDQA+iYV7wEfZu3u0fXe9ikprtLmsVs3tXZKkkGCLxmUkKD8rWZPGDFZ0REjfn2lus+vTvY0q2VOvnJEJGpsepxiW7AGgX1G+AB/S3NalzWU1Kiqt0dbd9ep2OCVJMZEhmjwmSfnZScodmcBNrwHARLztCHi5z2vbtKm0t3DtrDqsY/87NWxwhPKzkpWfnaSMYYM4iR4APATlC/AyTpehnZWHjxauah2sPyJJCgyQxlwYp/ysJOVnJSklIcLkpACAU+FtR8AL2Lt6tHV3vTaV1mhzeY1a2rsl9e5vjR+VqPysJE0cfeL+FgDAM1G+AA/V1GbXptJabSqt0bbdderucUmSBkWGaPLR6VbuyAQFs78FAF6Ftx0BD2EYhg7UtatwR7U2ldZo12dNfftbFyRF9r2dOJL9LQDwapQvwETH9reOFa5DDV/sb2WlHdvfSlZyfLjJSQEA/YW3HQE3693fqlPhjt7zt9o6eve3QoMtGp95bH8rSVHhwSYnBQAMBMoX4AaHW+3aXFajwh012r6nXo6j+1uxUSGanJWs/Kwk5YyIZ38LAPwAbzsCA8AwDH127PytHb37W8cMT4pUfnZv4RoxNIb9LQDwM0y+gH7idLpUduz8rR01qm48ur8VGKCsC+OUn927MJ8Ux/4WAPgzyhdwHjq7elS8q673/K3j9rdsIRaNHzVY+dm9529FhrG/BQDoRfkCztLhVruKSmu0qfTL+1uhvb+dmN27vxVkZX8LAHAyv9z5euWVV7RixQoFBAQoIyNDv/nNbxQSwsngODXDMPRZTZsKS3uPg9j9WXPfc6nJUX2Fa8TQGAUEsL8FAPh6fjf5qq2t1be//W0VFBQoNDRU9913n6ZPn64bb7zR7GjwIE6nS2X7D/cVrprGDkm9+1vZab37W5PHsL8FADh7fjn5cjqdstvtslqtstvtSkxMNDsSPECH3aGtu+pVWFqtLWW1au90SJJsIVZdnJui/OxkTcxMVAT7WwCA8+B35Wvw4MFauHChLrvsMoWEhGjatGm6+OKLzY4FkzS2dGpTaY0KS2tUsqdBPc7e/a346FBdMm6I8rOSNXZEHPtbAIB+43flq6WlRWvWrNGaNWsUGRmp++67T++8845mzZpldjS4gWEYqqppU9GOahWW1qji8y/2ty5MiVJ+VrLys5OUPiSa/S0AwIDwu/K1YcMGDR06VLGxsZKkGTNmaOvWrZQvH9bjdKlsf6OKdtSoqLRGtYd797csgQHKHRnfW7iykpQYG2ZyUgCAP/C78pWSkqLt27ers7NToaGh2rhxo7Kzs82OhX7WYXeoeFedinbUaEv5F/tbYaFWfSNviPKzkjRh9GBF2IJMTgoA8Dd+V75yc3N11VVX6YYbbpDVatXo0aM1b948s2OhHzQ0d2pTWe/p8iUVx+1vxdg0ffxQ5WclKTs9XkHWQJOTAgD8md8dNQHfYRiGKqtbVbijRptKq1VxoKXvubQh0b3nb2UlKY39LQCAB6F8wav0OF0q3dvYd/5WXVOnpN79rbEj4pWflaTJWUlKHMT+FgDAM1G+4PE67A59Ul6nwtJqfVJeqyP2HklSeKhVE0YP1pSsZI3PTFQ4+1sAAC9A+YJHqm/q1KbS3uMgduxtUI+z98s0YZBN+VlJmpKVrDFpcexvAQC8jt8t3MNzNLfZVVLRoJKKBuWMiFfm8FhtKDmkdcUHtO/gF/tb6UOjlZ+VrCnZSUpNjmJ/CwDg1Zh8wW0cPS512B3q7OpRYGCAfrm0SJXVrX3PpyZHadGsbD32x40amx6v/OxkTR6TpIRBNhNTAwDQv5h84Ws5nS51dvWow95z4j+7HCc8dqxUddp71NF17HHHcc/39B39MGr4IF02YdgJxUuSKqtb1dhi12u/vFahIXxpAgB8E1c4H+RyGbJ3H1+YHKcsT31FyX7qxzvsPep2OM8pQ0BA7w2pbSFWRUcEKykuTGEhQbKFWjV+VOIJt/U5Xvn+Rl0+cdj5vHwAADwa5ctDGIahrm7nCVOjEwqT3dH33Bfl6MTJUmdXjzq7HOrsOrfCJEmhwRbZQqwKD7UqPsamsKMFKiz02D+D+v6977Gjper4jwsNtiow8Kt3sz4KPaAPi6pOejxnZMI5ZwcAwBv4Rfn68mJ3zoh4xUSGnvfnNQzj6B7TlyZJxxWmvnLUV5pOnCz1lqnej3Od4/ZdsDWwrxTFRIacVIrCTihPVtmOK1DHPi4s1KrQEKssX1OY+lPOiHilJkedtPM1Nj3OLX8/AABm8fmF++Y2ux5+YeNJF/nH7pwiR8+J+0x95ehL06SOY1Onvrfmjk6i7D1ynmNjsloCZDtlOQo6rhRZj06Ugk4oT2FfKk9Wi3cet9DcZtenextVsqdeOSMTNDY9rl9KMQAAnszny9dHWw/of/78yUmP/8eNOVq75XPt+qzpjD9XYIBk+1I5+nIR6i1TX5SqL8rViW/XBVkt/fkyAQCAl/D5tx1LKhpO+fi+gy26YvIFGnlBzMm7TMeVqONLVUiQhTOmAADAefH58pUzIl4fFp682J2bkaBv5A0xIREAAPBn3rksdBaOLXYfj8VuAABgFp/f+ZJY7AYAAJ7DL8oXAACAp/D5tx0BAAA8CeULAADAjShfAAAAbkT5AgBF16+CAAADkUlEQVQAcCPKFwAAgBtRvgAAANyI8gUAAOBGlC8AAAA3onwBAAC4EeULAADAjShfAAAAbkT5AgAAcCPKFwAAgBtRvgAAANyI8gUAAOBGlC8AAAA3onwBAAC4kVeXr+rqat1222269tprdd1112nZsmWSpObmZi1YsEAzZszQggUL1NLSIknau3ev5s2bp+zsbC1duvSkz+d0OjV79mzdddddbn0dAADgq/Xn9X7ZsmW6/vrrdd111+mVV15x90uR5OXly2Kx6IEHHlBBQYHeeOMNvfrqq6qoqNCLL76oqVOnavXq1Zo6dapefPFFSVJMTIweeugh3XHHHaf8fMuXL1d6ero7XwIAADiN/rre7969WytWrNCKFSv0zjvvaP369aqqqnL76/Hq8pWYmKisrCxJUkREhNLS0lRbW6s1a9Zo9uzZkqTZs2frH//4hyQpLi5OOTk5slqtJ32umpoarV+/XnPmzHHfCwAAAKfVX9f7vXv3KicnRzabTVarVZMmTdLq1avd+2Lk5eXreAcOHFB5eblyc3PV2NioxMRESVJCQoIaGxtP++d//etf66c//akCA33mPwkAAD7nfK73GRkZ+uSTT9TU1KTOzk599NFHqqmpcUfsE5w8AvJCR44c0eLFi/Xzn/9cERERJzwXEBCggICAr/3z69atU2xsrLKzs1VUVDSQUQEAwDk63+t9enq6Fi1apDvuuEM2m02ZmZmmDF28fszjcDi0ePFizZw5UzNmzJDUO26sq6uTJNXV1Sk2NvZrP0dxcbHWrl2ryy+/XPfff78KCwv1k5/8ZMCzAwCAM9Mf13tJuvnmm7Vq1Sr95S9/UXR0tFJTUwcy9il5dfkyDEMPPfSQ0tLStGDBgr7HL7/8cr399tuSpLfffltXXHHF136eH//4x/roo4+0du1a/e53v9OUKVP029/+dkCzAwCAM9Nf13tJfW9NHjp0SKtXr9bMmTMHJvTXCDAMw3D739pPtmzZovnz5ysjI6NvbHj//fcrJydHP/rRj1RdXa2UlBT9/ve/V0xMjOrr63XTTTepvb1dgYGBCgsLU0FBwQmjy6KiIr300kt64YUXzHpZAADgOP15vf/Od76j5uZmWa1WPfjgg5o6darbX49Xly8AAABv49VvOwIAAHgbyhcAAIAbUb4AAADciPIFAADgRpQvAAAAN6J8AQAAuBHlCwAAwI0oXwAAAG70/wElRG1W21RPvAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.style.use('seaborn')\n",
    "sns.set_style('white')\n",
    "\n",
    "plt.figure(figsize=(10,6))\n",
    "sns.lineplot(x=df_posts.index[:-1],y=df_posts.iloc[:-1].DL_posts, markers=\"o\",style=True, legend=False)\n",
    "plt.xlabel(\"\")\n",
    "plt.ylabel(\"\")\n",
    "plt.title(\"deep-learning tagged questions posted over the year\",fontsize=18,loc='center',y=1.06)\n",
    "plt.suptitle(\"Trend in the questions tagged as deep-learning vs total number of questions posted\",y=0.93,fontsize=14)\n",
    "plt.tick_params(left=False,bottom=False)\n",
    "plt.yticks([])\n",
    "plt.xticks([2014,2019])\n",
    "plt.text(max(df_posts.index[:-1]),max(df_posts.DL_posts)-90,max(df_posts.DL_posts))\n",
    "plt.text(min(df_posts.index[:-1]),min(df_posts.DL_posts)+50,min(df_posts.DL_posts))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "The visualization confirms the trend shown in the table. The following conclusions are drawn from the plot :-\n",
    "* Every year since 2014 there has been an upward trend (increase) in the number of questions tagged as deep-learning.\n",
    "* Over the years (between 2014 and 2019) there has been a 12% increase in the posts (questions) tagged as deep-learning."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the previous analysis, it was found that Deep Learning was associated with other tags such as :-\n",
    "\n",
    "* keras\n",
    "* machine-learning\n",
    "* neural-networks\n",
    "* tensorflow\n",
    "\n",
    "From these four tags, Neural networks is the underlying technology of Deep Learning, similarly, Keras and Tensorflow are frameworks built to support and implement deep learning. These three tags are highly associated to Deep Learning. Even though Machine Learning and Deep Learning intersect, Machine Learning consists of topics out of scope for Deep Learning.<br>\n",
    "\n",
    "By this logic, the assumption is made that any post containing either of these tags,\n",
    "\n",
    "* deep-learning\n",
    "* keras\n",
    "* neural-network\n",
    "* tensorflow\n",
    "\n",
    "is in general labeled as relevant to the topic and hence the post is categorized as a Deep Learning question.<br>\n",
    "A new column is created, *is_relevant*, to indicate the above. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "relevant_tags = ['deep-learning','keras','tensorflow','neural-network']\n",
    "\n",
    "df['is_relevant'] = df.Tags.apply( \n",
    "    lambda x: 1 if list(np.intersect1d(x,relevant_tags)) else 0 \n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Similar to the previous logic, the grouping will happen based on years and aggregate on relevance of the post indicated by *is_relevant*."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Year</th>\n",
       "      <th>DL_posts</th>\n",
       "      <th>Total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2014</td>\n",
       "      <td>30</td>\n",
       "      <td>562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015</td>\n",
       "      <td>114</td>\n",
       "      <td>1167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2016</td>\n",
       "      <td>417</td>\n",
       "      <td>2146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2017</td>\n",
       "      <td>883</td>\n",
       "      <td>2957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2018</td>\n",
       "      <td>1876</td>\n",
       "      <td>5475</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2019</td>\n",
       "      <td>2695</td>\n",
       "      <td>8810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2020</td>\n",
       "      <td>130</td>\n",
       "      <td>459</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Year  DL_posts  Total\n",
       "0  2014        30    562\n",
       "1  2015       114   1167\n",
       "2  2016       417   2146\n",
       "3  2017       883   2957\n",
       "4  2018      1876   5475\n",
       "5  2019      2695   8810\n",
       "6  2020       130    459"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grouped_1 = df.groupby(by=['Year'])\n",
    "grouped_2 = df[df.is_relevant == 1].groupby(by=['Year'])\n",
    "\n",
    "temp_1 = pd.DataFrame(grouped_2.size(),columns=['DL_posts']).reset_index()\n",
    "temp_2 = pd.DataFrame(grouped_1.size(),columns=['Total']).reset_index()\n",
    "\n",
    "df_posts = pd.merge(temp_1,temp_2,left_on=['Year'],right_on=['Year'],how='inner')\n",
    "df_posts"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Dividing the columns DL_posts and Total gives the proportion of posts relevant to the analysis as compared to total posts for that year. These proportions can be used to visualize the trend in the number of posts relevant through the years. \n",
    "\n",
    "NOTE : A relevant post here means that the tag on the post is atleast one of the above 4 concluded tags. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>DL_posts</th>\n",
       "      <th>Total</th>\n",
       "      <th>Percentage</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Year</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014</th>\n",
       "      <td>30</td>\n",
       "      <td>562</td>\n",
       "      <td>5.338078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015</th>\n",
       "      <td>114</td>\n",
       "      <td>1167</td>\n",
       "      <td>9.768638</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016</th>\n",
       "      <td>417</td>\n",
       "      <td>2146</td>\n",
       "      <td>19.431500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017</th>\n",
       "      <td>883</td>\n",
       "      <td>2957</td>\n",
       "      <td>29.861346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018</th>\n",
       "      <td>1876</td>\n",
       "      <td>5475</td>\n",
       "      <td>34.264840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019</th>\n",
       "      <td>2695</td>\n",
       "      <td>8810</td>\n",
       "      <td>30.590238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020</th>\n",
       "      <td>130</td>\n",
       "      <td>459</td>\n",
       "      <td>28.322440</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      DL_posts  Total  Percentage\n",
       "Year                             \n",
       "2014        30    562    5.338078\n",
       "2015       114   1167    9.768638\n",
       "2016       417   2146   19.431500\n",
       "2017       883   2957   29.861346\n",
       "2018      1876   5475   34.264840\n",
       "2019      2695   8810   30.590238\n",
       "2020       130    459   28.322440"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_posts['Percentage'] = (df_posts.DL_posts/df_posts.Total) * 100\n",
    "df_posts.set_index('Year',inplace=True)\n",
    "df_posts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.style.use('seaborn')\n",
    "sns.set_style('white')\n",
    "\n",
    "plt.figure(figsize=(10,6))\n",
    "sns.lineplot(x=df_posts.index[:-1],y=df_posts.iloc[:-1].DL_posts, markers=\"o\",style=True, legend=False)\n",
    "plt.xlabel(\"\")\n",
    "plt.ylabel(\"\")\n",
    "plt.title(\"Deep Learning questions posted over the year\",fontsize=18,loc='center',y=1.06)\n",
    "plt.suptitle(\"Trend in the questions from the Deep Learning category vs total number of questions posted\",y=0.93,fontsize=14)\n",
    "plt.tick_params(left=False,bottom=False)\n",
    "plt.yticks([])\n",
    "plt.xticks([2014,2019])\n",
    "plt.text(max(df_posts.index[:-1]),max(df_posts.DL_posts)-170,max(df_posts.DL_posts))\n",
    "plt.text(min(df_posts.index[:-1]),min(df_posts.DL_posts)+70,min(df_posts.DL_posts))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following inferences are made from the plot :-\n",
    "\n",
    "* Similar to previous plot, there exists an upward trend in number of questions posted on Deep Learning category.\n",
    "* From the year 2014 to 2019, posts on Deep Learning category have seen a 25% increase."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The conclusion, gathered from the project are :-\n",
    "\n",
    "* The following topics (tags) are most popular in the Data science community on Stack Exchange,\n",
    "\n",
    "        * classification      * clustering       * cnn            * dataset\n",
    "        * deep-learning       * keras            * lstm           * machine-learning\n",
    "        * neural-network      * nlp              * pandas         * python\n",
    "        * regression          * scikit-learn     * tensorflow     * time-series\n",
    "        \n",
    "* Among the most popular topics, there exists association between certain topics such as,\n",
    "        \n",
    "        * deep-learning : keras : neural-network\n",
    "        * deep-learning : machine-learning : neural-network\n",
    "        * machine-learning : python : keras\n",
    "\n",
    "* The Deep Learning category consists of the following tags associated with the deep-learning tag,\n",
    "\n",
    "        * keras       \n",
    "        * tensorflow\n",
    "        * neural-networks\n",
    "\n",
    "* Deep Learning category has seen an upward trend in terms of number of posts since 2014.\n",
    "\n",
    "* Currently Deep Learning seems to be the most sought after category in the Data science community on Stack Exchange.\n",
    "\n",
    "* Creating posts related to Deep Learning would widely benefit the community and be popular amongst it."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}