022-FinalProject.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# COGS108 - Final Project\n",
    "## import MostCommonPythonChallenges from StackOverflowAndPiazza as FinalProject\n",
    "#### Team Members: Nathaniel Qin, Edward Chen, Michael Tran, Austin Hermida\n",
    "\n",
    "## Project Overview:\n",
    "- Reseach question, hypothesis\n",
    "    - Why most common challenges in Python?\n",
    "    - What do we hope to achieve?\n",
    "- Datasets and Background & Prior Work\n",
    "    - Brief summary of our datasets\n",
    "    - Previous work that have been done using these datasets\n",
    "- Data Cleaning and Analysis\n",
    "    - Python StackOverflow data\n",
    "    - 10% of all StackOverflow data\n",
    "    - Piazza data from w2017 & w2018\n",
    "- Discussion and Moving Forward\n",
    "    - Conclusion and Implications\n",
    "    - What further questions we may have\n",
    "    - How to improve our study\n",
    "\n",
    "# NOTE: Please do not run any of the cells in this notebook- in particular, the kmeans cells. As they use random centers each time, our categorizing algorithm will not work correctly :(\n",
    "\n",
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Group Member Contributions:\n",
    "---\n",
    "Note: PIDs removed for anonymity\n",
    "\n",
    "####Team Contributions:\n",
    "- XXXXX \n",
    "\t- Creating TF-IDF models and kmeans analysis\n",
    "\t- Cleaned and sorted StackOverflow dataset\n",
    "\t- Used sorted data to create most of the graphs\n",
    "\t- Developed color schemes \n",
    "\t- Created piazza.py/piazzaparser.py/testingNLP.py\n",
    "- XXXXX \n",
    "\t- Contributed the most to writing up the final project\n",
    "\t- Created jsonparser.py and cleaned Piazza dataset\n",
    "\t- Determined topics from top30 key words and questions\n",
    "\t- Wrote script to automatize sorting topics \n",
    "- XXXXX \n",
    "\t- Emailing professor Voytek for Piazza datasets\n",
    "\t- Added to outside scripts (in repo) for testing purposes\n",
    "\t- Initalized FinalProject\n",
    "- XXXXX \n",
    "\t- Contributed to analysis and inline comments\n",
    "\t- Helped label graphs\n",
    "\t- Helped with outside scripts (in repo) for testing purposes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Research Question\n",
    "\n",
    "According to insights.dice.com, \"Python is one of the few languages with growth. . . [Python is] in the top seven [languages] with positive job openings in 2018 compared to 2017\" (https://insights.dice.com/2017/12/18/developer-languages-hot-2018/). With Python still continually growing in popularity, we would expect more students and other coders to also be interested in this learning this language. However, learning a new language is not easy, as it is with learning new things in general.\n",
    "\n",
    "From our own personal experiences within this class, we have frequently perused the unparseable plethora of StackOverflow posts to clear up simple confusions about Python syntax, or to discover the most efficient way to analyze a given set of data using Python libraries. And through our endeavors, we became more interested in how we could formulate a way to tailor Python curricula more effectively. In the end, we believe that if we could aggregate the most common questions about Python topics and analyze this set of data, we could not only expedite our own learning processes, but hopefully provide useful insight that could be utilized by instructors who wish to improve in their pedagogical approaches.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hypothesis\n",
    "\n",
    "We predict the most popular Python questions on StackOverflow will be errors related to syntax and understanding or debugging error statements. We also predict that questions regarding higher-level, intellectually conceptual, and more abstract concepts will be less popular on StackOverflow because we believe that StackOverflow is generally used more by beginners to programming, and for questions about specific pieces of code.\n",
    "\n",
    "As a general measure, popularity in this analysis will be determined based on the question's number of upvotes, comments, and follow-up responses."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Datasets\n",
    "    Dataset Name: Python Questions from Stack Overflow: Full text of Stack Overflow Q&A about the Python programming language\n",
    "    Link to the dataset: https://www.kaggle.com/stackoverflow/pythonquestions/data\n",
    "    Number of observations: 844,352\n",
    "    Actual number of observations used: 200,000\n",
    "\n",
    "This dataset contains a variety of different questions that have been asked on StackOverflow about the Python programming language, including questions about many different topics, such as syntax, errors, or algorithms. This dataset specifically contains the question id, owner user id, creation date, question's score, title of the question, and the body of the question itself. We plan to use this dataset to determine which questions are the most popular on StackOverflow regarding Python. Popularity will be largely based on each question's score, normalized with the frequency of questions within each topic.\n",
    "\n",
    "To determine the popularity of each question, which is necessary for our sorting algorithm, we will take into consideration both the number of cumulative upvotes all of the questions within a topic have, as well as the number itself of questions that were asked within each topic as well. These measures will be normalized in our final analysis.\n",
    "\n",
    "    Dataset Name: StackSample: 10% of Stack Overflow Q&A: Text from 10% of Stack Overflow questions and answers on programming topics\n",
    "    Link to the dataset: https://www.kaggle.com/stackoverflow/stacksample/data\n",
    "    Number of observations: 520,192\n",
    "    Actual number of observations used: 64,601\n",
    "\n",
    "This dataset contains an aggregation of 10% of all the questions and answers asked on StackOverflow. We will clean this data further so that we only work with questions with a python-related tag. Our approach to this dataset is similar to our approach for the previous one. First, we would need to sort the different questions based on their overall topics, taking into account the popularity that is associated with the specific question. Our analysis of this data will also involve sorting these questions into a topics, and then gauging the popularity of each topic.\n",
    "\n",
    "    Dataset Name: Piazza - COGS108\n",
    "    Link to the dataset: https://piazza.com/class/jbocgn8rmq2434\n",
    "                         (We have also included posts from last year's COGS 108 course with Dr. Voytek's permission)\n",
    "    Number of observations: roughly 600-900, (closer to 600 because we do not have access to private posts).\n",
    "\n",
    "This dataset is simply an accumulation of all of the Piazza posts that have been made on the Piazza webpage for COGS 108 this quarter. We will read this data using the Piazza-API package. This dataset needs further cleaning, compared to the previous datasets. As this data does come from Piazza, we do expect a lot of the posts to be geared more towards course logistics rather than python itself. We also know many of the questions on Piazza are directly related to homework, and will take this into consideration by attempting to organize homework questions into direct topics such as syntax or error messages based on key word searches.\n",
    "\n",
    "---\n",
    "\n",
    "After all of our analysis, we plan to connect our findings from this dataset back to out findings from the StackOverflow datasets and come up with a final conclusion. We will attempt to map out connections, and compare the struggles from COGS 108 to the struggles that programmers face on StackOverflow in general.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Background and Prior Work\n",
    "_____________________________________________________________________________\n",
    "\n",
    "At the start of this quarter, we wanted to minimize the learning curve for students who did not have prior programming experience. Personally, although our group members have all had programming experience, we also feel that learning the intricacies of Python syntax and its many libraries was fairly difficult, especially given its difference from other common programming languages like Java or C/C++. Our own personal experiences with the programming assignments led us to our hypothesis about which topics will be the most commonly asked about. We hope by analyzing the StackOverflow datasets and further comparing our analysis with the Piazza data, we can help data science instructors better understand the needs of their students.\n",
    "\n",
    "_____________________________________________________________________________\n",
    "\n",
    "References (include links):\n",
    "\n",
    "_____________________________________________________________________________\n",
    "\n",
    "- 1) https://www.kaggle.com/mithrillion/identifying-clusters-of-related-questions\n",
    "\n",
    "Identifying Clusters of Related Questions - Mithrillion\n",
    "\n",
    "This reference is a kernel on Kaggle.com that utilizes the same dataset, Python Questions from Stack Overflow. In this kernel, user Mithrillion utilizes sklearn and other language processing libraries to sort questions in the dataset into their respective clusters. This kernel demonstrated the starting steps to clustering and sorting the questions in the Python dataset semantically and topically.\n",
    "\n",
    "This information would definitely be of use to our project because aside from checking the commmon challenges Python users may encounter, we want to check the different challenges based on their respective conceptual topic. By accurately being able to sort each of the questions in this dataset into their respective clusters, we can aggregate a common score based on a question's overall topic, as well as the various popularities of different questions about the specific topic.\n",
    "\n",
    "_____________________________________________________________________________\n",
    "\n",
    "- 2) https://www.kaggle.com/jiashenliu/if-we-treat-stack-overflow-as-a-social-network\n",
    "\n",
    "If we treat Stack-Overflow as a Social Network - Jason Liu\n",
    "\n",
    "This reference is another kernel on Kaggle.com that utilizes the same dataset, Python Questions from Stack Overflow. In this kernel, user Jason Liu determines if similar questions and answers appear from specific users and, given a set of your personal questions and answers, if we could predict which questions you would be wanting to search for next. Jason Liu uses many Python libraries to accurately determine the meaning of each question and create predictive models to determine possible follow-up questions that the user would be interested in.\n",
    "\n",
    "This information again falls in line with needing to determine the meaning of each question, a process which we will need to implement ourselves when trying to sort our questions by topic. What is interesting is that the predictive models created by Jason tries to guess which questions a user would be interested in next based on the user's own question and answer data. We could potentially also implement this type of predictive modelling by synthesizing our findings on the most common issues with the inter-relatedness of certain tags on StackOverflow, to see what issues become increasingly common as time passes, or as the user's own programming skill develops."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Methods\n",
    "_____________________________________________________________________________\n",
    "*Data Cleaning*"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Although our datasets pulled from Kaggle are already fairly clean in that they have successfully followed the Safe Harbour Method and organized much of the data into their respective categories, we still need to clean the actual data values themselves. For example, with the Python and 10% StackOverflow data, we will need to process and remove any unwanted strings in the body of the questions and answer data columns. \n",
    "\n",
    "Furthermore, for the 10% StackOverflow dataset, we will have to sort all of the questions by their corresponding tags, and only keep the questions that are related to python. Since this is a list of 10% of all of the questions that have been asked on StackOverflow, we will take measures to ensure that our analysis is only conducted on questions regarding python. This is necessary in order to focus this dataset into the scope of our research question."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We also need to sort the questions and answer data into their respective categories and calculate the most common questions out of each particular category. In order to accomplish this, we will utilize Counter in order to count the frequency of words within a certain grouping of questions. Then, using the most common words in each grouping, we will devise a set of topics for each grouping to be labelled under."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In regards to our methods for the Piazza posts, we will employ many similar processes to clean the data. After reading in the raw data as a json using the Piazza-API, there will still be a lot of unnecessary format tags that we do not want to get in the way of our actual language processing. Removing these tags will be necessary as a part of data cleaning. \n",
    "\n",
    "Of course, ethics and privacy is one of our main concerns with this dataset, as we will be receiving this data in its raw, unfiltered state. We will need to clean this dataset using the Safe Harbour method, which is outlined further in the Ethics and Privacy section."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "_____________________________________________________________________________\n",
    "\n",
    "*Data Analysis*\n",
    "\n",
    "After processing the data initially, we plan to use the kmean scores of all of the questions and plot them all on a scatter plot. This will make it somewhat easier to see whether or not the questions actually do fall into natural groupings based on semantic meanings. For each topic, we will create a bar graph to first to depict the actual number of questions that fall within each category to gauge the frequency of this issue. We plan to then create a histogram representing the frequency of scores within each topic (upvotes for StackOverflow data, and number of followups for Piazza data). And finally, we will normalize these two measures together to form a final popularity score for each group, and plot this on yet another bar graph. In doing so, we essentially will have the most common questions for each category. \n",
    "\n",
    "For our Piazza data, we will apply similar data analytic strategies, by first plotting the kmean clusters on a scatter plot, and then by utilizing bar graphs to visualize the most common topics that are asked about in Piazza questions.\n",
    "\n",
    "Finally, as we have found it quite difficult to apply a statistical test to our findings, we will endeavor to synthesize our findings manually, by discerning which questions were the most popular in our cumulative data findings. This may be able to help instructors when developing their curricula. \n",
    "_____________________________________________________________________________"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*What to report*\n",
    "\n",
    "We aim to report the topics of the most popular questions about learning and coding in the Python language. We hope to utilize our analysis of the StackOverflow data in comparison with relevant Piazza posts in COGS 108 to perhaps draw a comparison of what Python programmers in general struggle the most with, and also what beginners to Python struggle the most with. And finally, we hope to synthesize our findings together and discover a general consensus as to major python struggles.\n",
    "\n",
    "_____________________________________________________________________________"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Beginning of Actual Data Analysis\n",
    "_____________________________________________________________________________\n",
    "\n",
    "We begin by importing relevant packages and libraries. The first four packages are fairly standard for data science.\n",
    "\n",
    "NLTK is a Python package that contains useful corpuses to parse out English stopwords. Since our project heavily relies on natural language processing, we will also be utilizing sklearn; in particular, TfidfVectorizer and MiniBatchKMeans, to sort our datapoints into relevant topic clusters. Finally, we will be using Counter from the Python collections library to determine each kmeans cluster topic by the most frequent words in each of the clusters.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import bs4 as bs #BeautifulSoup\n",
    "import pandas as pd #Pandas\n",
    "import numpy as np #Numpy\n",
    "import matplotlib.pyplot as plt #matplotlib.pyplot\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package punkt to\n",
      "[nltk_data]     /Users/NathanielQin/nltk_data...\n",
      "[nltk_data]   Package punkt is already up-to-date!\n",
      "[nltk_data] Downloading package stopwords to\n",
      "[nltk_data]     /Users/NathanielQin/nltk_data...\n",
      "[nltk_data]   Package stopwords is already up-to-date!\n"
     ]
    }
   ],
   "source": [
    "# Import NLTK package \n",
    "#   PennTreeBank word tokenizer \n",
    "#   English language stopwords\n",
    "import nltk\n",
    "from nltk.tokenize import word_tokenize\n",
    "from nltk.corpus import stopwords\n",
    "\n",
    "#Download punkt and stopwords from NLTK\n",
    "nltk.download('punkt')\n",
    "nltk.download('stopwords')\n",
    "stop_words = set(stopwords.words('english'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Import sklearn modules \n",
    "from sklearn.svm import SVC\n",
    "from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.metrics import classification_report, precision_recall_fscore_support\n",
    "from sklearn.cluster import MiniBatchKMeans"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Import collections for Counter\n",
    "from collections import OrderedDict, Counter"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1 - Python Questions from StackOverflow\n",
    "Our main analysis is regarding Python data. Therefore we need to first clean the Python data and sort each of the questions into a topic. \n",
    "\n",
    "Before we can proceed to finding the most popular question in each dataset, we need to first organize each of the question into some general topic. To do so, we will utilize TF-IDF to measure the relevance of each of the words in each question. We decided to use TF-IDF rather than a direct word count because TF-IDF has the benefit of removing words that appear frequently throughout the entire dataset, such as stop words (in English: \"the\", \"a\", \"my\", etc.); that's the inverse-document frequency part. Therefore, TF-IDF should leave only the most frequent and relevant words as markers. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 275,
   "metadata": {},
   "outputs": [],
   "source": [
    "#read in dataset\n",
    "#modulate number of rows want to read - for simplicity's sake we have it set \n",
    "#to only 200000 lines. When we attempted reading the whole 800000+ data points, we ran into constant memory errors.\n",
    "df_py = pd.read_csv('pythonquestions/Questions.csv', \n",
    "                           encoding= 'ISO-8859-1', nrows=200000)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   Score                                              Title  \\\n",
      "0     21  How can I find the full path to a font from it...   \n",
      "1     27            Get a preview JPEG of a PDF on Windows?   \n",
      "2     40  Continuous Integration System for a Python Cod...   \n",
      "3     25     cx_Oracle: How do I iterate over a result set?   \n",
      "4     28  Using 'in' to match an attribute of Python obj...   \n",
      "\n",
      "                                                Body  \n",
      "0  <p>I am using the Photoshop's javascript API t...  \n",
      "1  <p>I have a cross-platform (Python) applicatio...  \n",
      "2  <p>I'm starting work on a hobby project with a...  \n",
      "3  <p>There are several ways to iterate over a re...  \n",
      "4  <p>I don't remember whether I was dreaming or ...  \n"
     ]
    }
   ],
   "source": [
    "#check data\n",
    "#drop rows we are not going to use in our analysis\n",
    "df_py = df_py.drop(['Id', 'OwnerUserId', 'CreationDate'], axis=1)\n",
    "print(df_py.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#for every question, parse HTML to actual text\n",
    "#place in new column\n",
    "\n",
    "#create a function to convert HTML text\n",
    "def convert_body(body):\n",
    "    return bs.BeautifulSoup(body, 'lxml').get_text()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#create new column after parsing HTML\n",
    "df_py[\"pBody\"] = pd.Series(df_py[\"Body\"].apply(convert_body), index = df_py.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   Score                                              Title  \\\n",
      "0     21  How can I find the full path to a font from it...   \n",
      "1     27            Get a preview JPEG of a PDF on Windows?   \n",
      "2     40  Continuous Integration System for a Python Cod...   \n",
      "3     25     cx_Oracle: How do I iterate over a result set?   \n",
      "4     28  Using 'in' to match an attribute of Python obj...   \n",
      "\n",
      "                                                Body  \\\n",
      "0  <p>I am using the Photoshop's javascript API t...   \n",
      "1  <p>I have a cross-platform (Python) applicatio...   \n",
      "2  <p>I'm starting work on a hobby project with a...   \n",
      "3  <p>There are several ways to iterate over a re...   \n",
      "4  <p>I don't remember whether I was dreaming or ...   \n",
      "\n",
      "                                               pBody  \n",
      "0  I am using the Photoshop's javascript API to f...  \n",
      "1  I have a cross-platform (Python) application w...  \n",
      "2  I'm starting work on a hobby project with a py...  \n",
      "3  There are several ways to iterate over a resul...  \n",
      "4  I don't remember whether I was dreaming or not...  \n"
     ]
    }
   ],
   "source": [
    "#check data\n",
    "print(df_py.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Create a color map for each label returned from kmeans\n",
    "LABEL_COLOR_MAP1= {0 : '#0c2a50ff',\n",
    "                   1 : '#253582ff',\n",
    "                   2 : '#593d9cff',\n",
    "                   3 : '#7e4e90ff',\n",
    "                   4 : '#a65c85ff',\n",
    "                   5 : '#cc6a70ff',\n",
    "                   6 : '#eb8055ff',\n",
    "                   7 : '#f9a242ff',\n",
    "                   8 : '#f7cb44ff',\n",
    "                   9 : '#e8fa5bff',\n",
    "                  }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#in order to sort the question into topic-clusters\n",
    "#create a TF-IDF object and transform data\n",
    "tfidf = TfidfVectorizer(sublinear_tf=True, analyzer='word', \n",
    "                        max_features=500, tokenizer=word_tokenize, max_df=0.8)\n",
    "\n",
    "#Transform data\n",
    "tf_pBody = tfidf.fit_transform(df_py[\"pBody\"])\n",
    "tf_pBody = tf_pBody.toarray() #transformed vectors into a matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(200000, 500)\n"
     ]
    }
   ],
   "source": [
    "#shape should equal 200000, 500\n",
    "print(tf_pBody.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Plot data on a 2 dimensional grid for better visualization \n",
    "pca = PCA(n_components=2).fit(tf_pBody)\n",
    "data2D = pca.transform(tf_pBody)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#use MiniBatchKMeans to make clusters to fit the questions\n",
    "#(minibatch cause over 10000 clusters)\n",
    "\n",
    "kmeans = MiniBatchKMeans(n_clusters=10).fit(tf_pBody) #default threshold\n",
    "centers2D = pca.transform(kmeans.cluster_centers_)\n",
    "clusters= kmeans.cluster_centers_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a37f57860>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot clusters based on their kmeans labels\n",
    "#mark cluster centers with red X's\n",
    "label_color_py = [LABEL_COLOR_MAP1[l] for l in kmeans.labels_]\n",
    "plt.scatter(data2D[:, 0], data2D[:, 1], c=label_color_py)\n",
    "plt.scatter(centers2D[:,0], centers2D[:,1], marker='x', s=200, linewidths=3, c='r')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Each of the points on the graph represents a question from the dataset and each of the color clusters represents how kmeans has grouped these questions together based on their TF-IDF values. Unfortunately, since kmeans uses random centers each time, our question-grouping method only works once :)\n",
    "\n",
    "Finally, we need to sort the questions into their respective categories. To do so we will relabel our original dataframe with the labels created by kmeans. Then we will print out the top 30 words from each of the categories and the 10 top most popular questions from each category (these cells have been deleted as they were extremely long), as this will assist us when we devise topics to label each cluster with. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#set the label to their corresponding questions in df_questions\n",
    "df_py[\"label\"] = pd.Series(kmeans.labels_, index = df_py.index)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   Score                                              Title  \\\n",
      "0     21  How can I find the full path to a font from it...   \n",
      "1     27            Get a preview JPEG of a PDF on Windows?   \n",
      "2     40  Continuous Integration System for a Python Cod...   \n",
      "3     25     cx_Oracle: How do I iterate over a result set?   \n",
      "4     28  Using 'in' to match an attribute of Python obj...   \n",
      "\n",
      "                                                Body  \\\n",
      "0  <p>I am using the Photoshop's javascript API t...   \n",
      "1  <p>I have a cross-platform (Python) applicatio...   \n",
      "2  <p>I'm starting work on a hobby project with a...   \n",
      "3  <p>There are several ways to iterate over a re...   \n",
      "4  <p>I don't remember whether I was dreaming or ...   \n",
      "\n",
      "                                               pBody  label  \n",
      "0  I am using the Photoshop's javascript API to f...      5  \n",
      "1  I have a cross-platform (Python) application w...      5  \n",
      "2  I'm starting work on a hobby project with a py...      6  \n",
      "3  There are several ways to iterate over a resul...      5  \n",
      "4  I don't remember whether I was dreaming or not...      5  \n"
     ]
    }
   ],
   "source": [
    "#print head to verify that the labels column exists\n",
    "print(df_py.head())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#create a function that converts an entire column into 1 string\n",
    "def getStrings(input_df):\n",
    "    output = \"\"\n",
    "    for string in input_df:\n",
    "        output = output + \" \" +string\n",
    "    return output.lower()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#create a list of each of the topic clusters based on their labels from kmeans\n",
    "qpy_dict = [df_py[df_py['label'] == i] for i in range(10)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#sort each question by descending order based on the question's score\n",
    "topic_questions = []\n",
    "for i in range(10):\n",
    "    qpy_dict[i] = qpy_dict[i].sort_values(by=['Score'], ascending=False)\n",
    "    topic_questions.append(qpy_dict[i]['pBody'][:1000])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[(\"''\", 5010), ('``', 3386), ('row', 3339), ('data', 1180), ('file', 1142), ('--', 862), ('csv', 802), ('like', 793), ('column', 754), ('rows', 744), ('import', 739), ('print', 736), ('python', 671), (\"'m\", 629), ('code', 605), ('using', 564), ('...', 540), ('def', 530), (\"n't\", 506), ('would', 498), ('return', 497), ('open', 492), ('want', 476), ('get', 463), (\"'s\", 454), ('list', 449), ('way', 447), ('values', 445), ('none', 434), ('array', 431)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[(\"''\", 1086), ('list', 990), ('``', 756), ('like', 608), ('--', 605), ('python', 540), ('way', 478), ('array', 429), ('print', 413), (\"'s\", 392), (\"n't\", 384), ('want', 384), ('would', 361), ('return', 355), (\"'m\", 339), ('...', 332), ('import', 310), ('code', 298), ('def', 295), ('example', 294), ('values', 290), ('data', 287), ('function', 283), ('using', 282), ('get', 281), ('value', 263), ('range', 248), ('following', 246), ('one', 230), ('numpy', 223)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[('--', 3328), (\"''\", 2854), ('``', 2037), ('table', 1576), ('column', 1462), ('id', 766), ('data', 765), ('query', 743), ('like', 694), ('using', 692), ('database', 648), (\"'m\", 605), ('integer', 588), ('class', 586), (\"n't\", 552), ('select', 538), ('import', 536), ('sqlalchemy', 512), (\"'s\", 495), ('python', 487), ('name', 476), ('want', 451), ('...', 449), ('get', 436), ('create', 430), ('way', 424), ('string', 423), ('would', 408), ('use', 392), ('sql', 389)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[(\"''\", 6684), ('``', 6000), ('file', 4693), ('line', 4294), ('--', 2910), ('module', 1852), ('import', 1735), ('python', 1445), ('error', 1258), ('call', 1227), ('traceback', 1187), ('last', 1150), ('recent', 1121), ('...', 736), (\"n't\", 698), ('get', 672), ('print', 645), ('return', 604), (\"'m\", 561), ('code', 560), ('using', 479), ('following', 441), ('object', 437), ('class', 433), ('run', 417), (\"'s\", 414), ('def', 409), ('stdin', 379), ('install', 369), ('trying', 363)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[(\"''\", 6234), ('``', 1721), ('python', 813), ('td', 617), ('/td', 592), ('import', 578), ('like', 561), ('print', 490), ('...', 488), ('string', 481), ('get', 443), (\"n't\", 440), (\"'m\", 435), ('using', 419), ('value', 396), ('code', 392), ('want', 382), (\"'s\", 377), ('way', 367), ('return', 347), ('file', 323), ('http', 319), ('def', 309), ('html', 304), ('class', 301), ('would', 298), ('use', 271), ('function', 267), ('error', 261), ('form', 254)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[('python', 882), ('way', 354), ('like', 267), ('want', 259), ('using', 239), ('would', 204), (\"'s\", 200), ('use', 196), ('string', 185), ('file', 181), (\"n't\", 167), ('function', 166), (\"''\", 161), ('get', 159), (\"'m\", 154), ('list', 150), ('need', 150), ('``', 145), ('know', 139), ('code', 135), ('one', 118), ('object', 115), ('something', 111), ('value', 93), ('module', 91), ('find', 88), ('example', 87), ('without', 77), ('possible', 75), ('import', 72)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[('python', 1450), (\"n't\", 945), (\"''\", 894), (\"'m\", 843), (\"'s\", 806), ('like', 779), ('``', 738), ('using', 631), ('would', 609), ('use', 590), ('file', 500), ('way', 485), ('code', 470), ('want', 431), ('--', 397), ('get', 378), ('need', 363), ('...', 358), (\"'ve\", 355), ('one', 352), ('something', 343), ('import', 298), ('run', 297), ('time', 296), ('know', 283), ('error', 283), ('module', 263), ('work', 262), ('function', 256), ('data', 249)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[(\"''\", 3378), ('self', 3330), ('def', 3230), ('class', 2756), ('``', 2374), ('return', 1487), ('--', 1058), ('print', 1037), ('__init__', 1000), ('object', 981), ('python', 805), ('import', 781), (\"n't\", 708), ('code', 668), ('like', 653), ('method', 614), (\"'s\", 613), ('...', 591), (\"'m\", 527), ('name', 523), ('way', 507), ('function', 498), ('==', 446), ('want', 439), ('value', 436), ('use', 432), ('would', 430), ('using', 390), ('none', 377), ('example', 371)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[(\"''\", 5522), ('``', 4185), ('print', 1118), ('python', 970), ('--', 967), ('import', 860), ('file', 652), ('string', 582), ('code', 578), ('def', 502), ('...', 497), ('return', 496), (\"'s\", 480), (\"n't\", 479), ('using', 468), (\"'m\", 450), ('like', 450), ('get', 431), ('error', 391), ('function', 335), ('line', 334), ('way', 322), ('value', 312), ('want', 304), ('use', 297), ('would', 285), ('type', 285), ('data', 284), ('open', 265), ('following', 265)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[(\"''\", 2884), ('class', 2561), ('``', 1875), ('model', 1419), ('models.charfield', 1103), ('user', 1100), ('models.model', 1094), ('return', 909), ('django', 859), ('def', 826), ('self', 790), ('name', 752), ('like', 680), ('form', 652), ('field', 652), ('models.foreignkey', 638), (\"'s\", 625), ('import', 616), (\"n't\", 575), ('...', 568), (\"'m\", 559), ('want', 490), ('models', 477), ('--', 467), ('file', 451), ('get', 442), ('way', 440), ('blank=true', 438), ('would', 432), ('fields', 432)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "#Create a list called top30py_dict to hold onto the top 30 most frequently used words in each\n",
    "#question cluster\n",
    "top30py_dict = []\n",
    "\n",
    "#filter out stopwords and single characters and count top 30 most frequently used words\n",
    "for i in range(10):\n",
    "    word_tokens = word_tokenize(getStrings(topic_questions[i]))\n",
    "    filtered_sentence = [w for w in word_tokens if not w in stop_words]\n",
    "    filtered_punct = [w for w in filtered_sentence if len(w) > 1]\n",
    "    #create a Counter object to count words\n",
    "    cnt = Counter(filtered_punct)\n",
    "    top30py_dict.append(cnt.most_common(30))\n",
    "    #save top 30 words from each label\n",
    "    print(cnt.most_common(30))\n",
    "    print()\n",
    "    print(\"-------------------------------------------\")\n",
    "    print()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#create a topic map from cluster\n",
    "LABEL_TOPIC_MAP1 = {0 : 'managing data structures',                       # if 'data' maps to 1180\n",
    "                    1 : 'lists',                                          # if 'list' maps to 990\n",
    "                    2 : 'using and managing databases',                   # if 'database' maps to 648\n",
    "                    3 : 'debugging errors',                               # if 'traceback' maps to 1187\n",
    "                    4 : 'advanced python concepts',                       # if 'class' maps to 301\n",
    "                    5 : 'basic python concepts',                          # if 'python' maps to 882\n",
    "                    6 : 'python concepts w/ minimal coding background',   # if 'code' maps to 470\n",
    "                    7 : 'python concepts w/ extensive coding background', # if 'method' maps to 614\n",
    "                    8 : 'dictionaries and strings',                       # if 'string' maps to 582\n",
    "                    9 : 'modelling data'                                  # if 'model' maps to 1419\n",
    "                   }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['managing data structures', 'lists', 'using and managing databases', 'debugging errors', 'advanced python concepts', 'basic python concepts', 'python concepts w/ minimal coding background', 'python concepts w/ extensive coding background', 'dictionaries and strings', 'modelling data']\n"
     ]
    }
   ],
   "source": [
    "#assign a topic to each cluster\n",
    "assign_topics1 = ['none']*10\n",
    "for i in range(10):\n",
    "    for j in range(10):\n",
    "        should_break = False;\n",
    "        for k,v in top30py_dict[j]:\n",
    "            if k == 'data' and v == 1180 and assign_topics1[j] == 'none':\n",
    "                assign_topics1[j] = LABEL_TOPIC_MAP1[0]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'list' and v == 990 and assign_topics1[j] == 'none':\n",
    "                assign_topics1[j] = LABEL_TOPIC_MAP1[1]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'database' and v == 648 and assign_topics1[j] == 'none':\n",
    "                assign_topics1[j] = LABEL_TOPIC_MAP1[2]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'traceback' and v == 1187 and assign_topics1[j] == 'none':\n",
    "                assign_topics1[j] = LABEL_TOPIC_MAP1[3]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'class' and v == 301 and assign_topics1[j] == 'none':\n",
    "                assign_topics1[j] = LABEL_TOPIC_MAP1[4]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'python' and v == 882 and assign_topics1[j] == 'none':\n",
    "                assign_topics1[j] = LABEL_TOPIC_MAP1[5]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'code' and v == 470 and assign_topics1[j] == 'none':\n",
    "                assign_topics1[j] = LABEL_TOPIC_MAP1[6]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'method' and v == 614 and assign_topics1[j] == 'none':\n",
    "                assign_topics1[j] = LABEL_TOPIC_MAP1[7]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'string' and v == 582 and assign_topics1[j] == 'none':\n",
    "                assign_topics1[j] = LABEL_TOPIC_MAP1[8]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'model' and v == 1419 and assign_topics1[j] == 'none':\n",
    "                assign_topics1[j] = LABEL_TOPIC_MAP1[9]\n",
    "                should_break = True\n",
    "                break\n",
    "        if should_break:\n",
    "            break\n",
    "print(assign_topics1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After we have determined each of the cluster's topics, we will now plot histograms and bar graphs to visualize which question topic contains the highest score. Each bar graph represents a different measure: cumulative question score, frequency, and a normalized measure of the first two.\n",
    "\n",
    "Each histogram represents the frequency of questions (y-axis) within each score grouping (x-axis)\n",
    "\n",
    "Each of the bar graphs represents the number of questions and scores of each topic"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 249,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAakAAAEWCAYAAADcsGj7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzt3XmcHFW99/HPl4SdQFgCQhJI1CCLC2LYFK8IiAGV4IIGUcKiQQ2Kj3pl0QcUWZ97FfBeBLmyBEUCLkj04mUPXFSWQCKrSFgTE8lAQsKugd/zxzmdKTo9PZ2Z6e6ame/79erXdJ2q6vpV9en6VZ2qOaWIwMzMrIxWa3cAZmZmXXGSMjOz0nKSMjOz0nKSMjOz0nKSMjOz0nKSMjOz0hqQSUrSxZJObtOyJekiSUsk3dGOGFaFpPsl7dHC5R0s6dpWLa87ksZICklD2x1Lb0j6qKR5kp6X9M52x9Mqkn4vaXKLlhWS3tyKZVmnliQpSY9LekrSuoWyz0ma2Yrlt9juwAeAURGxc60JJI2SdKmkZyS9IOkOSfs1O7BayTsito+Imc1edmF5l0bEPq1aXl/L392vJD0taamkeyUd2u64gH8HjoqI9SJidruDaZWI2DciprU7jqJmHvg0+wBc0kxJn2vW5/dEK8+khgJHt3B5fULSkFWcZSvg8Yh4oYvP2wi4FfgHsD2wCXAmMF3SAb2J1Vrip8A80ve8MXAI8FRfLqCHO7etgPt7uLxVreNWUu1uEWjK8iOi6S/gceBYYDEwPJd9DpiZ348BAhhamGcm8Ln8/lDgD6Sd+bPAo8C7c/k8YBEwuTDvxcB5wHXAc8DNwFaF8dvkcYuBh4BPVs17LnA18AKwd4312QKYkeefC3w+lx8BvAy8CjwPfLfGvN8D7gNWqyo/Jq+Xutseefhw4EFgCXBNZf3y/GfmbbIUuAd4KzAF+CcpOT4P/Lbw3eyd368JnAUsyK+zgDXzuD2A+cDX82cvBA4rxLMf8EDe3n8DvtFFXTgUuLUwHMAXgIfzupwDqIt5VyPVo0eAZ4ArgI0K438B/D2v9y3A9oVxawPfB57I42/NZZVtPRl4Enga+Faduvw8sEOd8bsDfyTV03nAobl8A+ASoCPH8O1KHeD19XsxcHK977hqeWvmmIJUXx/J5dvmOvMsKXntv4p1fCZwcl6X54HfkpLypcAy4E5gTGH6s/P6LgPuAt5bGPed/F1dkuvH/cD4wvjKd/pcrkMfLYwbkr+3p4HHgKMo/DZYeT9xK+msckmeft/CZ43N9eI54HpSXftZne/yX0n1fEH+LgJ4cx73IWB2Xt95wHcK8z2Zp30+v3YD3gTcSKq3T+ftOLyL5fbkN3xMnu4V0gnBilgL3/nJheGJwJwc/yPABOAU0r7r5fz5/8mq7Zu7rbtdrVvd/FFvZF+98kbcG/h1YSVWNUktBw4jVdqTc0U4h/Qj3SdXvPUKX8hzwL/k8WeTd4zAurlSHZa/zB1zpdm+MO9S4D2kneJaNdbnZuBHwFrADqQdz161dsI15r2N2slrbN4G4xrYHgeQkuO2eR2+Dfwxj/sgaScxPFeIbYHNa1XU4neT35+U49sUGEHaQX0vj9sjfwcnAauTktKLwIZ5/ELyjgnYENixi/V/3fbJ6/m7HO+WeVtO6GLer+b4RuXv9cfAZYXxhwPD6Ey2cwrjzsnbcCSpDr07T1fZ1v9FSlrvIP3Qt+0ihutJP8pJwJZV47Yk1buD8jbamJzQSDvoq3J8Y4C/AkdU1e8v5+9z7XrfcRdxFXegq+d5jwfWAPbMcb1lFer4zPwZbyIl2AdyzHvneC4BLipM/5m8vkNJBzJ/r3wuKUm9nOvMEOA04LbCvAeSDvxWAz5FSpyVOvuFvOxRpHp1PfWT1D+Bz+flfJGUYJTH/4mUwNYgHUwso4skRdppP0VKDusCP6/axnsAb8sxvz1Pe0Cd/dmbSZcB1iT9tm4Bzupi2T35Dc8BRgNrV9eH6vmAnfP3/4Ec/0hgm+rtuYr75obqbr1167JuN5JkevuiM0m9NW+cEax6knq4MO5tefrNCmXP0LlDuBiYXhi3HukIYTTpR/C/VfH9GDixMO8lddZldP6sYYWy04CLC7HWS1JzgS/UKF8rr9O7G9gevyfv4PLwaqSEsRVph/RXYFdWPltbUVGrv5v8/hFgv6ofy+OFH+VLVTEtAnbN758EjgTW76YuvG775PXcvTB8BXBsF/M+SD4YyMObk3ZKQ2tMOzx/9gZ5+7wEvKPGdJVtPapQdgcwqYsYNgROJ50NvEraOeyUxx0HXFljniGkxLddoexIOuv/ocCTVfN0+R13EVdxB/peUpJYrTD+MvLRPt3U8UJ9+1Zh+PvA7wvDH6FwEFBj/iWV7U1KUtcXxm0HvFRn3jnAxPz+RuDIwri9qZ+k5hamXSdP+wbSAcRyYJ3C+J/RdZK6EDi9MLw1VTv+qunPAs6sqlMr1cvC9AcAs7sY15Pf8OFd1Yfq+Uj7uzPrfO+rmqQarrv11q2rV0vv7ouI+0hHzcf2YPZiu/9L+fOqy9YrDM8rLPd50qnoFqQNtYukZysv4GBSRV5p3hq2ABZHxHOFsidIRyONeJq0c61WKeto4DO2As4uxL+YdFQyMiJuJJ2mnwM8Jel8Ses3GNsWpHWpeCKXVTwTEcsLwy/Suc0/TjpSfkLSzZJ2a3CZkHaotT6z2lbAlYX1fpCUKDaTNETS6ZIekbSM9MOFdM1vE9JBwCO9jSEilkTEsRGxPbAZaYf6G0kiHcDUWsYmpKP36m1brDPVda7L77jOOlRsAcyLiNdWYXm1VP++uvy9Sfq6pAfzzSTPkg4ONilMX71916pcv5B0iKQ5hXV9a2HeLapi7S7uFcuJiBfz2/Xo/N2+WJi2u995cXzxu0PSLpJuktQhaSnpjK+4vlRNv6mk6ZL+luvnz7qavoe/4Ua+z4qu6mlPNVx3e7Ju7bgF/UTS6XjxB1O5yWCdQlkxafTE6MobSesBG5FO/ecBN0fE8MJrvYj4YmHeqPO5C4CNJA0rlG1Jug7TiOuBj0uq3vafJF3zeYTut8c80tFlcR3Wjog/AkTEDyPiXaQbM7Ymta13t16Vdduqar0WNLJSEXFnREwkNRX+hnRG1Nfmka4xFNd7rYj4G/BpUjv73qQd5Jg8j0gHBi+Tmq76TEQ8TWo+2oJUv+Z1sYynSWd81du2WGeqv5u633E3FgCjq+pYd8vrMUnvJV0T+SSp+Xc4qcVEDcy7Famp9Shg4zzvfYV5F5Ka+ipG0zMLSb/b4m+q3mctrBq/ZdX4n5OuS4+OiA1I18ArMdfatqfl8rdHxPqk5tEut08PfsPV5S9Sf//R1W+h+nMa2TevUt2ts241tTxJRcRc4HLgK4WyDtIP6DP5iPhwer9D2U/S7pLWIN2scHtEzCOdyW0t6bOSVs+vnSRt22D880jXak6TtJakt5NumLi0wbjOBNYHLpD0hvwZBwH/l9Tk+FoD2+M84DhJ2wNI2kDSgfn9Tvkob3VSBavcyAHpSPiNdWK7DPi2pBGSNgFOIB3x1SVpjfz/TxtExD9Jbf2vdjdfD5wHnJJ3bOQ4J+Zxw0hNas+QflCnVmbKZxQXAj+QtEXeprtJWnNVA5B0hqS3ShqaD1S+SGpieoZUB/aW9Mk8fmNJO0TEq6SkfYqkYTn+r1F/23b5HTfgdtJ3/81cv/cgNc9NX9X1bdAwUlNaBzBU0gmkOt6IdUk7uQ4ASYeRzqQqrgCOljRS0nBSMlxlEfEEMAv4Tq6vu5G2SVeuAA6VtF1ObCdWjR9GOjN7WdLOpIOkig7gNV7/WxtGuhnhWUkjqbNj7uVvuGIO8Olc1ycA7yuMuwA4TNJeklbL23abWp/fw31zT/dPNbXrn3lPIlXOos+TvrhnSBm2kSPGen5OqliLgXeRmvTIzXT7kC58LyA1D5xBuqDZqINIR+oLgCtJyeW6RmbMO7PdSc1PD5Aq7iXA1Ii4sDBpl9sjIq7MMU/PTQf3Afvm0euTjkyXkJooniEd7UOqnNvl0/Df1AjvZNIP+R7gXuDuXNaIzwKP53i+QDpS7Gtnk45er5X0HOkmil3yuEtI6/s30na9rWreb5DW6U5SnTiDntX/dUjfeeUu062A/QEi4klSk+fX8zLmkG7EgHRh+YU8z62k+ln8vl+nm++4roj4R45pX9JZ3I+AQyLiL6uwnqviGtJ1iL+SvoOXabD5KSIeIF3v+hNpB/k20o0pFf8FXEuqk7NJdyQup2cHQQeT7rR7hlSvLycd2NSK6/ek60w3kq4j31g1yZeAk3I9PIFCy0FuUjwF+EP+re0KfJd0k9ZS4L9JN5F1pTe/4YqjSUm4cjljxbQRcQfpxrEzczw303mWfzbwCaXOCH6Yy1Zp39yL/VNNlbterE1ye+wfSBfcT2h3PGZlJmlf4LyI2Krbibv/rMuBv0RE9VmSlciA7BapP4mIZaSj71cl9fY6nNmAImltSfvl5tORpNaRK3v4WTtJelNu4ppAuoZZ72zESsBnUmZWWvl60M2kf8B/idRUdnQ+uFvVz/oIqelzY9JNSqdFxEV9GK41gZOUmZmVlpv7zMystPr14wm6sskmm8SYMWPaHYaZWb9y1113PR0RI9odR9GATFJjxoxh1qxZ7Q7DzKxfkfRE91O1lpv7zMystJykzMystJykzMystJykzMystJykzMystJykzMystJqWpCRdKGmRpPsKZf8m6S+S7pF0Ze56vzLuOElzJT0k6YOF8gm5bK6knjws0czM+qlmnkldDEyoKrsOeGtEvJ3Urf9xAJK2Iz06Y/s8z4/ys0uGkJ7guC/pkdMH5WnNzGwQaFqSiohbSM/UKZZdW3j8+G10PnFzIjA9Il6JiMdIz2/ZOb/mRsSj+Rk50/O0ZmY2CLSzx4nDSQ8dg/Qo+eJD6ubT+Xj5eVXlu9Bkp57T+QDT46dOavbizMysC225cULSt0hP16w8cl01Jos65bU+c4qkWZJmdXR09E2gZmbWVi1PUpImAx8GDo7O54TMB0YXJhtFejR7V+UriYjzI2J8RIwfMaJU/SOamVkPtTRJ5adhHgPsHxEvFkbNACZJWlPSWGAccAdwJzBO0lhJa5BurpjRypjNzKx9mnZNStJlwB7AJpLmkx77fBywJnCdJIDbIuILEXG/pCuAB0jNgFMj4tX8OUcB1wBDgAsj4v5mxWxmZuXStCQVEQfVKL6gzvSnAKfUKL8auLoPQzMzs37CPU6YmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpNS1JSbpQ0iJJ9xXKNpJ0naSH898Nc7kk/VDSXEn3SNqxMM/kPP3DkiY3K14zMyufZp5JXQxMqCo7FrghIsYBN+RhgH2Bcfk1BTgXUlIDTgR2AXYGTqwkNjMzG/ialqQi4hZgcVXxRGBafj8NOKBQfkkktwHDJW0OfBC4LiIWR8QS4DpWTnxmZjZAtfqa1GYRsRAg/900l48E5hWmm5/LuipfiaQpkmZJmtXR0dHngZuZWeuV5cYJ1SiLOuUrF0acHxHjI2L8iBEj+jQ4MzNrj1YnqadyMx7576JcPh8YXZhuFLCgTrmZmQ0CrU5SM4DKHXqTgasK5Yfku/x2BZbm5sBrgH0kbZhvmNgnl5mZ2SAwtFkfLOkyYA9gE0nzSXfpnQ5cIekI4EngwDz51cB+wFzgReAwgIhYLOl7wJ15upMiovpmDDMzG6CalqQi4qAuRu1VY9oApnbxORcCF/ZhaGZm1k+U5cYJMzOzlThJmZlZaTlJmZlZaTlJmZlZaTlJmZlZaTlJmZlZaTlJmZlZaXWbpCQdLWn93BvEBZLulrRPK4IzM7PBrZEzqcMjYhmpS6IRpN4gTm9qVGZmZjSWpCo9ke8HXBQRf6Z27+RmZmZ9qpEkdZeka0lJ6hpJw4DXmhuWmZlZY333HQHsADwaES9K2pjcAayZmVkzdZukIuI1SU8B20lqWoe0ZmZm1bpNOpLOAD4FPAC8mosDuKWJcZmZmTXU3HcA8JaIeKXZwZiZmRU1cuPEo8DqzQ7EzMysWiNnUi8CcyTdAKw4m4qIrzQtKjMzMxpLUjPya1A69ZzpK94fP3VSGyMxMxt8Grm7b5qkNYCtc9FDEfHP5oZlZmbW2N19ewDTgMdJPU2MljQ5Inx3n5mZNVUjzX3fB/aJiIcAJG0NXAa8q5mBmZmZNXJ33+qVBAUQEX/Fd/uZmVkLNHImNUvSBcBP8/DBwF3NC8nMzCxpJEl9EZgKfIV0TeoW4EfNDMrMzAwau7vvFeAH+WVmZtYyXV6TknRF/nuvpHuqX71ZqKT/I+l+SfdJukzSWpLGSrpd0sOSLs+3vSNpzTw8N48f05tlm5lZ/1HvTOro/PfDfblASSNJTYfbRcRLORlOIj2v6syImC7pPNIjQs7Nf5dExJslTQIqHd6amdkA1+WZVEQszG+/FBFPFF/Al3q53KHA2vnRH+sAC4E9gV/m8dNIHdsCTMzD5PF7SfKTgc3MBoFGbkH/QI2yfXu6wIj4G/DvwJOk5LSUdLfgsxGxPE82HxiZ348E5uV5l+fpN67+XElTJM2SNKujo6On4ZmZWYnUuyb1RUn3AttUXY96DOjxNSlJG5LOjsYCWwDrUjvpRWWWOuM6CyLOj4jxETF+xIgRPQ3PzMxKpN41qZ8DvwdOA44tlD8XEYt7scy9gcciogNA0q+BdwPDJQ3NZ0ujgAV5+vnAaGB+bh7cAOjN8s3MrJ+od01qaUQ8Dnwb+Hu+FjUW+Iyk4b1Y5pPArpLWydeW9iI99fcm4BN5msnAVfn9jDxMHn9jRKx0JmVmZgNPI9ekfgW8KunNwAWkRPXzni4wIm4n3QBxN3BvjuF84Bjga5Lmkq45XZBnuQDYOJd/jdef1ZmZ2QDWSI8Tr0XEckkfA86KiP+QNLs3C42IE4ETq4ofBXauMe3LwIG9WZ6ZmfVPjZxJ/VPSQcAhwO9ymTuYNTOzpmskSR0G7AacEhGPSRoL/Ky5YZmZmTXWd98Dko4BtszDjwGnNzswMzOzbs+kJH0EmAP8Tx7eQdKMZgdmZmbWSHPfd0g3NDwLEBFzSHf4mZmZNVUjSWp5RCytKvP/KZmZWdM1cgv6fZI+DQyRNI7Ug/kfmxuWmZlZY2dSXwa2B14BLgOWAV9tZlBmZmbQ2N19LwLfyi8zM7OW6TZJSbqJ2r2O79mUiMzMzLJGrkl9o/B+LeDjwPIupjUzM+szjTT33VVV9AdJNzcpHjMzsxUaae7bqDC4GvAu4A1Ni8jMzCxrpLnvLtI1KZGa+R4DjmhmUGZmZtBYc597lzAzs7ZopLnvY/XGR8Sv+y4cMzOzTo009x0BvBu4MQ+/H5gJLCU1AzpJmZlZUzSSpALYLiIWAkjaHDgnIg5ramRmZjboNdIt0phKgsqeArZuUjxmZmYrNHImNVPSNaR++wKYBNzU1KjMzMxo7O6+oyR9FPiXXHR+RFzZ3LDMzMwaO5MiJyUnJjMza6lGrkmZmZm1hZOUmZmVVpdJStIN+e8ZrQvHzMysU71rUptLeh+wv6TppL77VoiIu5samZmZDXr1ktQJwLHAKOAHVeMC6PFDDyUNB34CvDV/1uHAQ8DlwBjgceCTEbFEkoCzgf2AF4FDnSDNzAaHLpv7IuKXEbEv8P8i4v1Vr94+lfds4H8iYhvgHcCDpIR4Q0SMA27IwwD7AuPyawpwbi+XbWZm/UQj/yf1PUn70/l/UjMj4nc9XaCk9fNnHZo//x/APyRNBPbIk00j9Q94DDARuCQiArhN0nBJm1f1gmFmZgNQt3f3SToNOBp4IL+OzmU99UagA7hI0mxJP5G0LrBZJfHkv5vm6UcC8wrzz89l1XFOkTRL0qyOjo5ehGdmZmXRyC3oHwI+EBEXRsSFwIRc1lNDgR2BcyPincALdDbt1aIaZbFSQcT5ETE+IsaPGDGiF+GZmVlZNPp/UsML7zfo5TLnA/Mj4vY8/EtS0noq97Be6Wl9UWH60YX5RwELehmDmZn1A40kqdOA2ZIuljSN9Dj5U3u6wIj4OzBP0lty0V6kZsQZwORcNhm4Kr+fARyiZFdgqa9HmZkNDo3cOHGZpJnATqSmt2NyoumNLwOXSloDeBQ4jJQwr5B0BPAkcGCe9mrS7edzSbeg+zlWZmaDRKMdzC4kndH0iYiYA4yvMWqvGtMGMLWvlm1mZv2H++4zM7PScpIyM7PSqpukJK0m6b5WBWNmZlZUN0lFxGvAnyVt2aJ4zMzMVmjkxonNgfsl3UH6x1sAImL/pkVlZmZGY0nqu02PwszMrIZG/k/qZklbAeMi4npJ6wBDmh+amZkNdt0mKUmfJz0iYyPgTaTOXc+jxv80DSannjN9xfvjp05qYyRmZgNXI7egTwXeAywDiIiH6eyh3MzMrGkaSVKv5Gc+ASBpKDV6ITczM+trjSSpmyUdD6wt6QPAL4DfNjcsMzOzxpLUsaSHFN4LHEnq8PXbzQzKzMwMGru777X8iI7bSc18D+VOX83MzJqqkbv7PkS6m+8R0qM6xko6MiJ+3+zgzMxscGvkn3m/D7w/IuYCSHoT8N+Ak5SZmTVVI9ekFlUSVPYonY92NzMza5ouz6QkfSy/vV/S1cAVpGtSBwJ3tiA2MzMb5Oo1932k8P4p4H35fQewYdMiMjMzy7pMUhFxWCsDMTMzq9bI3X1jgS8DY4rT+1EdZmbWbI3c3fcb4AJSLxOvNTccMzOzTo0kqZcj4odNj8TMzKxKI0nqbEknAtcCr1QKI+LupkVlZmZGY0nqbcBngT3pbO6LPGxmZtY0jSSpjwJvLD6uw8zMrBUa6XHiz8Dwvl6wpCGSZkv6XR4eK+l2SQ9LulzSGrl8zTw8N48f09exmJlZOTWSpDYD/iLpGkkzKq8+WPbRwIOF4TOAMyNiHLAEOCKXHwEsiYg3A2fm6czMbBBopLnvxL5eqKRRwIeAU4CvSRLpGten8yTTgO8A5wIT83uAXwL/KUl+XIiZ2cDXyPOkbm7Ccs8CvgkMy8MbA89GxPI8PB8Ymd+PBOblWJZLWpqnf7r4gZKmAFMAttxyyyaEbGZmrdZtc5+k5yQty6+XJb0qaVlPFyjpw6Se1e8qFteYNBoY11kQcX5EjI+I8SNGjOhpeGZmViKNnEkNKw5LOgDYuRfLfA+wv6T9gLWA9UlnVsMlDc1nU6OABXn6+cBoYL6kocAGwOJeLN/MzPqJRm6ceJ2I+A29+B+piDguIkZFxBhgEnBjRBwM3AR8Ik82Gbgqv5+Rh8njb/T1KDOzwaGRDmY/VhhcDRhPjea2PnAMMF3SycBsUn+B5L8/lTSXdAY1qQnLNjOzEmrk7r7ic6WWA4+T7rjrtYiYCczM7x+lRjNiRLxMetCimZkNMo1ck/JzpczMrC3qPT7+hDrzRUR8rwnxmJmZrVDvTOqFGmXrknqA2BhwkjIzs6aq9/j471feSxpG6sboMGA68P2u5jMzM+srda9JSdoI+BpwMKmroh0jYkkrAjMzM6t3TerfgI8B5wNvi4jnWxaVmZkZ9f+Z9+vAFsC3gQWFrpGe6023SGZmZo2qd01qlXujMDMz60tORGZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlpOUmZmVlqNPD7eVsGp50xf8f74qZPaGImZWf/nMykzMystJykzMystJykzMystJykzMystJykzMystJykzMyutlicpSaMl3STpQUn3Szo6l28k6TpJD+e/G+ZySfqhpLmS7pG0Y6tjNjOz9mjHmdRy4OsRsS2wKzBV0nbAscANETEOuCEPA+wLjMuvKcC5rQ/ZzMzaoeVJKiIWRsTd+f1zwIPASGAiMC1PNg04IL+fCFwSyW3AcEmbtzhsMzNrg7Zek5I0BngncDuwWUQshJTIgE3zZCOBeYXZ5uey6s+aImmWpFkdHR3NDNvMzFqkbUlK0nrAr4CvRsSyepPWKIuVCiLOj4jxETF+xIgRfRWmmZm1UVuSlKTVSQnq0oj4dS5+qtKMl/8uyuXzgdGF2UcBC1oVq5mZtU877u4TcAHwYET8oDBqBjA5v58MXFUoPyTf5bcrsLTSLGhmZgNbO3pBfw/wWeBeSXNy2fHA6cAVko4AngQOzOOuBvYD5gIvAoe1NlwzM2uXliepiLiV2teZAPaqMX0AU5salJmZlZJ7nDAzs9JykjIzs9Lyk3lbxE/sNTNbdT6TMjOz0nKSMjOz0nKSMjOz0nKSMjOz0nKSMjOz0nKSMjOz0nKSMjOz0nKSMjOz0nKSMjOz0nKSMjOz0nKSMjOz0nKSMjOz0nIHs23mjmfNzLrmMykzMystJykzMystJykzMystJykzMystJykzMyst391XUsW7/uD1d/51dUeg7xQ0s4HGZ1JmZlZaTlJmZlZabu4bBNw8aGb9lZOUrcTJy8zKot8kKUkTgLOBIcBPIuL0Noc06Dh5mVmr9YskJWkIcA7wAWA+cKekGRHxQHsjs4q+alJc1c/py8TpJGxWPv0iSQE7A3Mj4lEASdOBiYCTVD/VzoTQrsTZk3Xuq+1Uxpj66vPrTd/s76gsBzZliaMZFBHtjqFbkj4BTIiIz+XhzwK7RMRRhWmmAFPy4FuAh3qxyE2Ap3sxf3802NZ5sK0veJ0Hi96s81YRMaIvg+mt/nImpRplr8uuEXE+cH6fLEyaFRHj++Kz+ovBts6DbX3B6zxYDLR17i//JzUfGF0YHgUsaFMsZmbWIv0lSd0JjJM0VtIawCRgRptjMjOzJusXzX0RsVzSUcA1pFvQL4yI+5u4yD5pNuxnBts6D7b1Ba/zYDGg1rlf3DhhZmaDU39p7jMzs0HIScrMzErLSapA0gRJD0maK+nYdsfTDJIulLTMAfG1AAAEnElEQVRI0n2Fso0kXSfp4fx3w3bG2NckjZZ0k6QHJd0v6ehcPmDXW9Jaku6Q9Oe8zt/N5WMl3Z7X+fJ8I9KAIWmIpNmSfpeHB/T6Akh6XNK9kuZImpXLBkzddpLKCl0v7QtsBxwkabv2RtUUFwMTqsqOBW6IiHHADXl4IFkOfD0itgV2Babm73Ygr/crwJ4R8Q5gB2CCpF2BM4Az8zovAY5oY4zNcDTwYGF4oK9vxfsjYofC/0cNmLrtJNVpRddLEfEPoNL10oASEbcAi6uKJwLT8vtpwAEtDarJImJhRNyd3z9H2omNZACvdyTP58HV8yuAPYFf5vIBtc6SRgEfAn6Sh8UAXt9uDJi67STVaSQwrzA8P5cNBptFxEJIO3Rg0zbH0zSSxgDvBG5ngK93bvqaAywCrgMeAZ6NiOV5koFWx88Cvgm8loc3ZmCvb0UA10q6K3cPBwOobveL/5NqkW67XrL+TdJ6wK+Ar0bEsnSgPXBFxKvADpKGA1cC29aarLVRNYekDwOLIuIuSXtUimtMOiDWt8p7ImKBpE2B6yT9pd0B9SWfSXUazF0vPSVpc4D8d1Gb4+lzklYnJahLI+LXuXjArzdARDwLzCRdjxsuqXJwOpDq+HuA/SU9Tmqq35N0ZjVQ13eFiFiQ/y4iHYzszACq205SnQZz10szgMn5/WTgqjbG0ufytYkLgAcj4geFUQN2vSWNyGdQSFob2Jt0Le4m4BN5sgGzzhFxXESMiogxpN/ujRFxMAN0fSskrStpWOU9sA9wHwOobrvHiQJJ+5GOvipdL53S5pD6nKTLgD1I3fk/BZwI/Aa4AtgSeBI4MCKqb67otyTtDvwvcC+d1yuOJ12XGpDrLentpAvmQ0gHo1dExEmS3kg609gImA18JiJeaV+kfS83930jIj480Nc3r9+VeXAo8POIOEXSxgyQuu0kZWZmpeXmPjMzKy0nKTMzKy0nKTMzKy0nKTMzKy0nKTMzKy0nKbNekPSt3Mv4PbkX6l3aHZPZQOJukcx6SNJuwIeBHSPiFUmbAD1+FISkoYV+5swMn0mZ9cbmwNOVfw6NiKdzH2o7SfpjfpbTHZKG5ec7XZSf+zNb0vsBJB0q6ReSfgtcm8v+VdKd+ezsu+1bPbP285mUWc9dC5wg6a/A9cDlwJ/y309FxJ2S1gdeIj3niIh4m6RtSL1Wb50/Zzfg7RGxWNI+wDhS/2sCZkj6l/yIFbNBx2dSZj2Un9f0LmAK0EFKTkcCCyPizjzNstyEtzvw01z2F+AJoJKkrit0WbNPfs0G7ga2ISUts0HJZ1JmvZAfhzETmCnpXmAqtR8HUe+5IC9UTXdaRPy4z4I068d8JmXWQ5LeIql4lrMDqafxLSTtlKcZlh8VcQtwcC7bmtTx50M1PvYa4PD87CskjczPCTIblHwmZdZz6wH/kR+JsRyYS2r6uyiXr026HrU38CPgvHy2tRw4NN8R+LoPjIhrJW0L/CmPex74DP34eUBmveFe0M3MrLTc3GdmZqXlJGVmZqXlJGVmZqXlJGVmZqXlJGVmZqXlJGVmZqXlJGVmZqX1/wFC0rVNLCrEdwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a22e314a8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a2aaa94e0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAbkAAAEWCAYAAAD7HukTAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzt3Xm8HFWZ//HPl4QtEAiEsCWBoAYERBEjizIOAiJBJIyCAwMSFkURHRx1JLixyQC/UVFGRXGIBFSQiQKRQSECgVHZwiI7EhBITCCBBAigSOD5/XFOk0rT3bfuvd335tb9vl+vfnXVqaqup9anz6nqakUEZmZmVbRKfwdgZmbWKU5yZmZWWU5yZmZWWU5yZmZWWU5yZmZWWU5yZmZWWStdkpN0vqSv99O8JenHkpZIuqU/YugOSfdK2q0P53eIpKv7an5dkTROUkga2t+x9Iakf5I0V9Lzkt7ej3H8WtLk/pp/WZIelbRnf8fRG5J+IOmrfTSvjq0vSbMkfawTn90uXSa5vIKelLRWoexjkmZ1NLL+sSvwPmBMROzYaARJYyT9VNLTkl6QdIukfTodWKPkHxHbRsSsTs+7ML+fRsRefTW/dsvb7heSnpL0rKS7JR3e33EB3wA+HRFrR8Qd/RVEREyMiGn9Nf/BJCI+GRGn9ncc9fKXxjf1dxztVLYmNxQ4rpOBdIKkId2cZHPg0Yh4ocnnrQ/8Dvg7sC2wAXAWcLGk/XsTq/WJC4G5pO08EjgMeLKdM+hhrXJz4N4ezq+7+7jZ4BIRLV/Ao8AUYDEwIpd9DJiVu8cBAQwtTDML+FjuPhz4PSkZPAM8Arwrl88FFgKTC9OeD/wAmAksBa4HNi8Mf3Methh4EPhI3bTnAFcCLwB7NlieTYEZefo5wMdz+VHA34BXgOeBkxtMeypwD7BKXfnxebnU1frI/UcC9wNLgKtqy5enPyuvk2eBu4C3AEcDL5OS6/PArwrbZs/cvTrwbWB+fn0bWD0P2w2YB3w+f/YC4IhCPPsA9+X1/RfgC032hcOB3xX6A/gk8FBelu8BajLtKqT96GHgaeASYP3C8P8BnsjLfQOwbWHYmsA3gcfy8N/lstq6ngw8DjwFfLnFvvw8sH2L4bsCfyDtp3OBw3P5usAFwKIcw1dq+wAr7t+Lga+32sZ181s9xxSk/fXhXL513meeISW//bq5jz9aLAdOAn6Su9cAfpK3wTPArcBGTY7b35FqmUuAPwMTC5+5Rd5OS4Hf5m3/kybrdT3girz+luTuMXXHx6l5PS4FrgY2KAz/aF7vTwNfrl++unmdD3wf+HVet78HNiYdD0uAB4C3F8av7ZNLScfAP9Xv7y3WwRF5Gy8lHf+fqIvli6RjbT7pnBnAmwpx1vaV3Wh9fI4EfgU8l7fX1ykchw3WQdP1BewI3Ji3/QLgu8BqedgNLN8Xnwf+ueS2Ox24hXRsXk7547rpeQfYF7gzx/kH4K1159u/5OkeBPZoti4ionSS2xP4ZWGjdDfJLcs7xJC8gR4nHRSrA3vlYNcubPylwHvy8O/UNiiwFunkcwSpdrkD6cS2bWHaZ4F3k06qazRYnutJB8EawPZ54+3R6CTeYNqbaJz8tsjrYHyJ9bE/KblunZfhK8Af8rD3A7cBI0gJb2tgk/qDotHJDDglx7chMCrvGKcWDqJleZxVSTvXi8B6efgC4B8KJ6Qdmiz/CusnL+cVOd7N8rrcu8m0n83xjcnb9YfARYXhRwLDWZ6s7ywM+15eh6NJ+9C78ni1df0jUtJ7G/ASsHWTGH5LOukdBGxWN2wz0n53cF5HI8kJkZTgLs/xjQP+BBxVt39/Jm/PNVtt4yZxFU9+q+ZpvwSsBuye49qqG/v4a/tF7j+J5UnuE6QT5rC8Lt8BrNPkuH0Z+Hge7xjSyVp5+I2kk/9qpC8Hz9E8yY0EPpznOZx04rus7vh4GNgyr79ZwBl52DakE27tfPCtvL5bJbmn8nKtAVxLSk6Hsfz8c11h/ANJX3xXIZ3UX2D5MdfVOvgA8EbSsfqPpGNqhzxsb9LJfdu83BfSOsm1Oj4vzq9heX3Mpcl5qqv1ldfLzqT9chwpSX+20b7YjW33F9KX8bWAXxT3A1of1w3PO6Tz+kJgp7zeJ5P26dWBrfLyb1rIP29smcNaDSweMHkhniWdQLub5B4qDNsuj79Roexplp9QzgcuLgxbm1S7GkvaCf+vLr4fAicWpr2gxbKMzZ81vFB2OnB+o5N4g+nnAJ9sUL5GXqZ3lVgfvyafIHP/KqQdenPSCe1PpJ2wvrZ4Pq2T3MPAPoVh7yc1vUI6iP5aF9NCYOfc/Tjp5LdOF/vCCusnL+euhf5LgClNpr2fwjcuYBPSCWRog3FH5M9eN6+fvwJvazBebV0Xv1neAhzUJIb1gDNItaNXSN8U35mHnQBc2mCaIaTEuU2h7BMs3/8PBx6vm6bpNm4SV/Hk9w+kk+MqheEXASeV2cfr94vcfxLLk9yR1H0zbnHczikMG5bj3Jj0hWAZMKww/Cc0SXIN5rM9sKRuvl8p9H8K+E3u/horng/WIrVotEpyPyr0fwa4v9C/HfBMi9juBCZ1tQ6aTHsZcFzungqcXhj2JlonuYbHZ97/XiZ/ycnDmtbkerC+Pkthv6cuyZXcdmcU+rfJ8xvSYNrXjuvc3/C8Q2qpOLWu7EHSF4k35XWzJ7Bqmf2t9N2VEXEP6Vv7lLLTFBSve/w1f1592dqF/rmF+T5PagbalJQIdpL0TO0FHEI68F43bQObAosjYmmh7DFSDaGMp0gn53q1skUlPmNz4DuF+BeTvgmOjohrSc0H3wOelHSupHVKxrYpaVlqHstlNU9HxLJC/4ssX+cfJn17fEzS9ZJ2KTlPSCfkRp9Zb3Pg0sJy309KNBtJGiLpDEkPS3qOdJKGdM1zA9KXiId7G0NELImIKRGxLbAR6YR2mSSRvgA1mscGpNpK/bot7jP1+1zTbdxiGWo2BeZGxKvdmF93XEhqPr1Y0nxJ/0/Sqk3GfW29RsSLuXNtlh9HLxbGbRqTpGGSfijpsbx9bwBG1F1PbLYNN2XF88ELpC/FrdSfW5qeayQdJunOwrZ6C2mbvy6uunWApImSbpK0OE+7T2HaFeKm623W7PgcRap1lf2slutL0paSrpD0RN4W/8GKy7uCktuuGM9jpNroBl0c19D8vLM58Pm68/xYUu1tDikxnwQslHSxpOJ57nW6+xOCE0lV9+IBV7tJY1ihrJh0emJsrUPS2sD6pGaCucD1ETGi8Fo7Io4pTBstPnc+sL6k4YWyzUjV7TJ+C3xYUv16+wipTf1hul4fc0lt98VlWDMi/gAQEWdHxDtIzRxbAv9eYrlqy7Z53XLNL7NQEXFrREwiNXVeRqqRtdtc0vWM4nKvERF/Af4FmET6drYuqYYGKTE8RbpW+sZ2BhMRT5Ga2zYl7V9zm8zjKdI36fp1W9xn6rdNy23chfnA2Lp9rKv51XuBJvtfRLwcESdHxDaklod9SU153bGAdBwV5zG22cika01bATtFxDqkpjRI27fMvIrng2GkJrRek7Q5qan708DIiBhBuubeZVySVic1zX2D1Co1gnSdtDbtAlLTfE2r9dPKIlKtuexndbW+ziFdlxyft8WXaL28ZbZdMZ7NSMfLU7Q+rludd+YCp9UdP8Mi4qI83c8iYlfSMRnAmS3i716Sy1n058C/FsoWkQ7AQ3PmPpLen5D2kbSrpNVIF6Rvjoi5pJrklpI+KmnV/HqnpK1Lxj+X1FRzuqQ1JL2VdMPJT0vGdRawDnCepI3zZxwMfJXUZPpqifXxA+AESdsCSFpX0oG5+52SdsrfrF9g+Y0wkL6NvqFFbBcBX5E0StIGpGaLn3S1QJJWU/r927oR8TLp2sorXU3XAz8ATssnFnKck/Kw4aQmwadJJ+f/qE2UazRTgW9J2jSv013ySaZbJJ0p6S2ShuYvOseQmqOeJu0De0r6SB4+UtL2EfEK6eA7TdLwHP/naL1um27jEm4mbfsv5v17N+CDpGsyZd0JHJSnnwAcUFgH75W0Xf4m/hzphNSt7R0RjwGzgZPy/rNLjrGZ4aQa1DNKdyif2I3ZTQf2LZwPTqF9v+9di3SSXAQg6QhSTa6M1UjXiBYByyRNJN1fUHMJcISkrXOi+VpPAsz73y9J63qYpDfT+ktJV+trOGm7P58/65i66evPM2W23aGStsnLeQowPcfd9Lju4rzzI+CT+VwoSWtJ+kA+/raStHs+/v+WY2u5//ZkZzmFtHMUfZxU43iaVAMp8421lZ+RVuZi0oXSQwByM+NepBsH5pOaEs4k7WxlHUz6RjEfuJSUnGaWmTCfDHclNZ/dR7rAewFwbERMLYzadH1ExKU55otzFf4eYGIevA5pAy9h+d1R38jDzgO2ydX3yxqE93XSiecu4G7g9lxWxkeBR3M8nwQOLTldd3yHdFfr1ZKWkm5C2SkPu4C0vH8hrdeb6qb9AmmZbiXtE2fSs313GGmb1+7y3RzYDyAiHic1nXw+z+NO0o0skK7rvJCn+R1p/yxu7xV0sY1bioi/55gmkr4Nfx84LCIe6MZyfpX0xWoJcHKOt2Zj0onwOVKT8fWU+DLUwCHALqR99OukL78vNRn326QbSp4ibdvflJ1JRNwLHEtahgWkZZrXg3gbffZ9pLt2bySd3Lcj3ZhUZtqlpC/7l+SY/oW0f9eG/xo4G7iOdC3/xjyo2Tpq5dOkmtATpObmi5p9Ton19YUc61LSuebndR9xEjAtn2c+QrltdyHpGuMTpHNjrRLU1XHd8LwTEbNJ59Dv5vjnkK6PQjrXn5HjeYJUC/xSo3VRU7tLyHpA6XrZ70kXbnv0Tc2sCiT9HHggIrpTSxs0cmvTPaSf9SzravwuPutM0s0vk9sSXMWtdI/1Gkgi4jnSt/9XJPX2OqTZgJGb1t8oaRVJe5OuvTRqYRi0lB7Xtpqk9Ug1+1/1JMFJerOkt+amux1Jl1gubXe8VTWgn/m3MsjX+U7u7zjM+tjGpGtFI0nNYcdEPz6SbCX1CVIz3iukZuFP9fBzhpOaKDcl3T7/TdLvNq0EN1eamVllubnSzMwqq5LNlRtssEGMGzeuv8MwMxtQbrvttqciYlR/x9FOlUxy48aNY/bs2f0dhpnZgCLpsa7HGljcXGlmZpXlJGdmZpXlJGdmZpXlJGdmZpXlJGdmZpXlJGdmZpXlJGdmZpXlJGdmZpXlJGdmZpVVySee9Nb0qbe/1n3AkTv0YyRmZtYbrsmZmVllOcmZmVllOcmZmVllOcmZmVllOcmZmVllOcmZmVllOcmZmVllOcmZmVllOcmZmVllOcmZmVllOcmZmVlldSzJSZoqaaGkexoM+4KkkLRB7peksyXNkXSXpB0K406W9FB+Te5UvGZmVj2drMmdD+xdXyhpLPA+4PFC8URgfH4dDZyTx10fOBHYCdgROFHSeh2M2czMKqRjSS4ibgAWNxh0FvBFIAplk4ALIrkJGCFpE+D9wMyIWBwRS4CZNEicZmZmjfTpNTlJ+wF/iYg/1g0aDcwt9M/LZc3KG3320ZJmS5q9aNGiNkZtZmYDVZ8lOUnDgC8DX2s0uEFZtCh/fWHEuRExISImjBo1queBmplZZfRlTe6NwBbAHyU9CowBbpe0MamGNrYw7hhgfotyMzOzLvVZkouIuyNiw4gYFxHjSAlsh4h4ApgBHJbvstwZeDYiFgBXAXtJWi/fcLJXLjMzM+tSJ39CcBFwI7CVpHmSjmox+pXAI8Ac4EfApwAiYjFwKnBrfp2Sy8zMzLo0tFMfHBEHdzF8XKE7gGObjDcVmNrW4MzMbFDwE0/MzKyynOTMzKyynOTMzKyynOTMzKyynOTMzKyynOTMzKyynOTMzKyynOTMzKyynOTMzKyynOTMzKyynOTMzKyynOTMzKyynOTMzKyynOTMzKyynOTMzKyynOTMzKyynOTMzKyynOTMzKyyOpbkJE2VtFDSPYWy/5T0gKS7JF0qaURh2AmS5kh6UNL7C+V757I5kqZ0Kl4zM6ueTtbkzgf2riubCbwlIt4K/Ak4AUDSNsBBwLZ5mu9LGiJpCPA9YCKwDXBwHtfMzKxLHUtyEXEDsLiu7OqIWJZ7bwLG5O5JwMUR8VJE/BmYA+yYX3Mi4pGI+DtwcR7XzMysS/15Te5I4Ne5ezQwtzBsXi5rVv46ko6WNFvS7EWLFnUgXDMzG2j6JclJ+jKwDPhprajBaNGi/PWFEedGxISImDBq1Kj2BGpmZgPa0L6eoaTJwL7AHhFRS1jzgLGF0cYA83N3s3IzM7OW+rQmJ2lv4Hhgv4h4sTBoBnCQpNUlbQGMB24BbgXGS9pC0mqkm1Nm9GXMZmY2cHWsJifpImA3YANJ84ATSXdTrg7MlARwU0R8MiLulXQJcB+pGfPYiHglf86ngauAIcDUiLi3UzGbmVm1dCzJRcTBDYrPazH+acBpDcqvBK5sY2hmZjZIdNlcKek4SesoOU/S7ZL26ovgzMzMeqPMNbkjI+I5YC9gFHAEcEZHozIzM2uDMs2Vtdv49wF+HBF/VL6gNhhMn3r7a90HHLlDP0ZiZmbdVaYmd5ukq0lJ7ipJw4FXOxuWmZlZ75WpyR0FbA88EhEvShpJarI0MzNbqXWZ5CLiVUlPAttI6vMfj5uZmfVUl0lL0pnAP5N+w/ZKLg7ghg7GZWZm1mtlamb7A1tFxEudDsbMzKydytx48giwaqcDMTMza7cyNbkXgTslXQO8VpuLiH/tWFRmZmZtUCbJzcAPRTYzswGozN2V0/I/AGyZix6MiJc7G5aZmVnvlbm7cjdgGvAo6eknYyVNjgjfXWlmZiu1Ms2V3wT2iogHASRtCVwEvKOTgZmZmfVWmbsrV60lOICI+BO+29LMzAaAMjW52ZLOAy7M/YcAt3UuJDMzs/Yok+SOAY4F/pV0Te4G4PudDMrMzKwdytxd+RLwrfwyMzMbMJpek5N0SX6/W9Jd9a+uPljSVEkLJd1TKFtf0kxJD+X39XK5JJ0taU7+/B0K00zO4z8kaXLvFtfMzAaTVjW54/L7vj387POB7wIXFMqmANdExBmSpuT+44GJwPj82gk4B9hJ0vrAicAE0kOhb5M0IyKW9DAmMzMbRJrW5CJiQe78VEQ8VnwBn+rqg/Pv6BbXFU8i/eaO/L5/ofyCSG4CRkjaBHg/MDMiFufENhPYu+zCmZnZ4FbmJwTva1A2sYfz26iWPPP7hrl8NDC3MN68XNas3MzMrEtNmyslHUOqsb2x7hrccOD3bY5DDcqiRfnrP0A6GjgaYLPNNmtfZGZmNmC1qsn9DPggcHl+r73eERGH9nB+T+ZmSPL7wlw+DxhbGG8MML9F+etExLkRMSEiJowaNaqH4ZmZWZW0uib3bEQ8CnwFeCJfi9sCOFTSiB7ObwZQu0NyMimB1soPy3dZ7gw8m5szrwL2krRevhNzr1xmZmbWpTLX5H4BvCLpTcB5pET3s64mknQRcCOwlaR5ko4CzgDeJ+kh0rW+M/LoV5L+nHUO8CPyjS0RsRg4Fbg1v07JZWZmZl0q88STVyNimaQPAd+OiP+SdEdXE0XEwU0G7dFg3CA9VaXR50wFppaI08zMbAVlanIvSzoYOAy4Ipf5Ac1mZrbSK5PkjgB2AU6LiD9L2gL4SWfDMjMz670yz668T9LxwGa5/88sv5ZmZma20uqyJifpg8CdwG9y//aSZnQ6MDMzs94q01x5ErAj8AxARNxJusPSzMxspVYmyS2LiGfryho+dcTMzGxlUuYnBPdI+hdgiKTxpD9P/UNnwzIzM+u9MjW5zwDbAi8BFwHPAZ/tZFBmZmbtUObuyheBL+eXmZnZgNFlkpN0HQ2uwUXE7h2JyMzMrE3KXJP7QqF7DeDDwLLOhGNmZtY+ZZorb6sr+r2k6zsUj5mZWduUaa5cv9C7CvAOYOOORWRmZtYmZZorb2P5v3QvA/4MHNXJoMzMzNqhTHOln25iZmYDUpnmyg+1Gh4Rv2xfOGZmZu1TprnyKOBdwLW5/73ALOBZUjOmk5yZma2UyiS5ALaJiAUAkjYBvhcRR3Q0MjMzs14q81ivcbUElz0JbNmheMzMzNqmTJKbJekqSYdLmgz8L3Bdb2Yq6d8k3SvpHkkXSVpD0haSbpb0kKSfS1otj7t67p+Th4/rzbzNzGzw6DLJRcSngR8AbwO2B86NiM/0dIaSRpP+yWBCRLwFGAIcBJwJnBUR44ElLP+ZwlHAkoh4E3BWHs/MzKxLZWpyRMSlEfFv+XVpG+Y7FFhT0lBgGLAA2B2YnodPA/bP3ZNyP3n4HpLUhhjMzKziSiW5doqIvwDfAB4nJbdnST84fyYias/EnAeMzt2jgbl52mV5/JH1nyvpaEmzJc1etGhRZxfCzMwGhD5PcpLWI9XOtgA2BdYCJjYYtfbPB41qbY3+FeHciJgQERNGjRrVrnDNzGwAa5rkJF2T39t9DWxP4M8RsSgiXib9zu5dwIjcfAkwBpifu+cBY3MsQ4F1gcVtjsnMzCqoVU1uE0n/COwn6e2Sdii+ejHPx4GdJQ3L19b2AO4j3bF5QB5nMnB57p6R+8nDr42I19XkzMzM6rX6MfjXgCmkWtW36oYF6UaRbouImyVNB24nPfD5DuBc0k8TLpb09Vx2Xp7kPOBCSXNINbiDejJfMzMbfJomuYiYDkyX9NWIOLWdM42IE4ET64ofAXZsMO7fgAPbOX8zMxscyvwLwamS9gPek4tmRcQVnQ3LzMys97q8u1LS6cBxpOtm9wHH5TIzM7OVWpkHNH8A2D4iXgWQNI10zeyETgZmZmbWW2V/Jzei0L1uJwIxMzNrtzI1udOBOyRdR/ph9ntwLc7MzAaAMjeeXCRpFvBOUpI7PiKe6HRgZmZmvVWmJkf+P7kZHY7FzMysrfr82ZVmZmZ9xUnOzMwqq2VzpaRVgLvyn5tawfSpt7/WfcCRvXmUp5mZdUrLmlz+bdwfJW3WR/GYmZm1TZkbTzYB7pV0C/BCrTAi9utYVGZmZm1QJsmd3PEozMzMOqDM7+Sul7Q5MD4ifitpGDCk86GZmZn1TpkHNH8cmA78MBeNBi7rZFBmZmbtUOYnBMcC7waeA4iIh4ANOxmUmZlZO5RJci9FxN9rPZKGkv4Z3MzMbKVWJsldL+lLwJqS3gf8D/CrzoZlZmbWe2WS3BRgEXA38AngSuArvZmppBGSpkt6QNL9knaRtL6kmZIeyu/r5XEl6WxJcyTdJcm/vDYzs1LK3F35av6j1JtJzZQPRkRvmyu/A/wmIg6QtBowDPgScE1EnCFpCim5Hg9MBMbn107AOfndzMyspTJ3V34AeBg4G/guMEfSxJ7OUNI6pP+kOw8gIv4eEc8Ak4BpebRpwP65exJwQSQ3ASMkbdLT+ZuZ2eBR5sfg3wTeGxFzACS9Efhf4Nc9nOcbSM2fP5b0NuA24Dhgo/yXPkTEAkm1OzhHA3ML08/LZQt6OH8zMxskylyTW1hLcNkjwMJezHMosANwTkS8nfSosCktxleDstc1l0o6WtJsSbMXLVrUi/DMzKwqmtbkJH0od94r6UrgElJyORC4tRfznAfMi4ibc/90UpJ7UtImuRa3CcsT6TxgbGH6McD8+g+NiHOBcwEmTJjgnziYmVnLmtwH82sN4EngH4HdSE2N6/V0hhHxBDBX0la5aA/gPtI/j0/OZZOBy3P3DOCwfJflzsCztWZNMzOzVprW5CLiiA7O9zPAT/OdlY8AR5AS7iWSjgIeJ9UYIf1kYR9gDvBiHtfMzKxLXd54ImkLUlIaVxy/N3+1ExF3AhMaDNqjwbhBerSYmZlZt5S5u/Iy0u3+vwJe7Ww4ZmZm7VMmyf0tIs7ueCRmZmZtVibJfUfSicDVwEu1woi4vWNRmZmZtUGZJLcd8FFgd5Y3V0buNzMzW2mVSXL/BLyh+Hc7ZmZmA0GZJ578ERjR6UDMzMzarUxNbiPgAUm3suI1uR7/hMDMzKwvlElyJ3Y8CjMzsw4o839y1/dFIGZmZu1W5oknS1n+1P/VgFWBFyJinU4GZmZm1ltlanLDi/2S9gd27FhEZmZmbVLm7soVRMRl+DdyZmY2AJRprvxQoXcV0oOV/X9tZma20itzd+UHC93LgEeBSR2JxszMrI3KXJPz/7eZmdmA1DTJSfpai+kiIk7tQDxmZmZt06om90KDsrWAo4CRgJOcmZmt1JomuYj4Zq1b0nDgOOAI4GLgm82mMzMzW1m0vCYnaX3gc8AhwDRgh4hY0heBmZmZ9VbT38lJ+k/gVmApsF1EnNTOBCdpiKQ7JF2R+7eQdLOkhyT9XNJquXz13D8nDx/XrhjMzKzaWv0Y/PPApsBXgPmSnsuvpZKea8O8jwPuL/SfCZwVEeOBJaRrf+T3JRHxJuCsPJ6ZmVmXmia5iFglItaMiOERsU7hNby3z62UNAb4APDfuV+kp6hMz6NMA/bP3ZNyP3n4Hnl8MzOzlrr9WK82+TbwReDV3D8SeCYiluX+ecDo3D0amAuQhz+bx1+BpKMlzZY0e9GiRZ2M3czMBog+T3KS9gUWRsRtxeIGo0aJYcsLIs6NiAkRMWHUqFFtiNTMzAa6Mo/1ard3A/tJ2gdYA1iHVLMbIWlorq2NAebn8ecBY4F5koYC6wKL+z7scqZPvf217gOO3KEfIzEzsz6vyUXECRExJiLGAQcB10bEIcB1wAF5tMnA5bl7Ru4nD782IvyAaDMz61J/XZNr5Hjgc5LmkK65nZfLzwNG5vLPAVP6KT4zMxtg+qO58jURMQuYlbsfocGfsUbE34AD+zQwMzOrhJWpJmdmZtZWTnJmZlZZTnJmZlZZTnJmZlZZTnJmZlZZTnJmZlZZTnJmZlZZTnJmZlZZTnJmZlZZTnJmZlZZTnJmZlZZTnJmZlZZTnJmZlZZTnJmZlZZTnJmZlZZTnJmZlZZ/fqnqYPJ9Km3v9Z9wJE79GMkZmaDh2tyZmZWWX2e5CSNlXSdpPsl3SvpuFy+vqSZkh7K7+vlckk6W9IcSXdJcjXIzMxK6Y+a3DLg8xGxNbAzcKykbYApwDURMR64JvfoTfGrAAAHqElEQVQDTATG59fRwDl9H7KZmQ1EfZ7kImJBRNyeu5cC9wOjgUnAtDzaNGD/3D0JuCCSm4ARkjbp47DNzGwA6tdrcpLGAW8HbgY2iogFkBIhsGEebTQwtzDZvFxW/1lHS5otafaiRYs6GbaZmQ0Q/ZbkJK0N/AL4bEQ812rUBmXxuoKIcyNiQkRMGDVqVLvCNDOzAaxfkpykVUkJ7qcR8ctc/GStGTK/L8zl84CxhcnHAPP7KlYzMxu4+uPuSgHnAfdHxLcKg2YAk3P3ZODyQvlh+S7LnYFna82aZmZmrfTHj8HfDXwUuFvSnbnsS8AZwCWSjgIeBw7Mw64E9gHmAC8CR/RtuGZmNlD1eZKLiN/R+DobwB4Nxg/g2I4GZWZmleQnnpiZWWX52ZX9zM+0NDPrHNfkzMysspzkzMysspzkzMysspzkzMysspzkzMyssnx35UrMd16amfWOa3JmZlZZrskNQK7hmZmV45qcmZlVlpOcmZlVlpOcmZlVlpOcmZlVlm88qRDfkGJmtiInuUHAyc/MBis3V5qZWWW5JjeIuYZnZlXnJGev0yz5OSma2UAzYJKcpL2B7wBDgP+OiDP6OSQr6G4CdCI1s74wIJKcpCHA94D3AfOAWyXNiIj7+jcy60qnk1aZZNnbeTvxmg1cAyLJATsCcyLiEQBJFwOTACe5QaKdiaZdtchOfE6n5t2u8p7oyy869fOowheUKixDf1JE9HcMXZJ0ALB3RHws938U2CkiPl0Y52jg6Ny7FfBgL2a5AfBUL6YfiAbbMg+25QUv82DRm2XePCJGtTOY/jZQanJqULZCdo6Ic4Fz2zIzaXZETGjHZw0Ug22ZB9vygpd5sBiMy9zKQPmd3DxgbKF/DDC/n2IxM7MBYqAkuVuB8ZK2kLQacBAwo59jMjOzldyAaK6MiGWSPg1cRfoJwdSIuLeDs2xLs+cAM9iWebAtL3iZB4vBuMxNDYgbT8zMzHpioDRXmpmZdZuTnJmZVZaTXIGkvSU9KGmOpCn9HU8nSJoqaaGkewpl60uaKemh/L5ef8bYbpLGSrpO0v2S7pV0XC6v7HJLWkPSLZL+mJf55Fy+haSb8zL/PN/IVRmShki6Q9IVub/Sywsg6VFJd0u6U9LsXFbZfbu7nOSywqPDJgLbAAdL2qZ/o+qI84G968qmANdExHjgmtxfJcuAz0fE1sDOwLF521Z5uV8Cdo+ItwHbA3tL2hk4EzgrL/MS4Kh+jLETjgPuL/RXfXlr3hsR2xd+H1flfbtbnOSWe+3RYRHxd6D26LBKiYgbgMV1xZOAabl7GrB/nwbVYRGxICJuz91LSSfB0VR4uSN5Pveuml8B7A5Mz+WVWmZJY4APAP+d+0WFl7cLld23u8tJbrnRwNxC/7xcNhhsFBELICUEYMN+jqdjJI0D3g7cTMWXOzfd3QksBGYCDwPPRMSyPErV9vFvA18EXs39I6n28tYEcLWk2/LjDaHi+3Z3DIjfyfWRLh8dZgObpLWBXwCfjYjn0hf96oqIV4DtJY0ALgW2bjRa30bVGZL2BRZGxG2SdqsVNxi1Estb590RMV/ShsBMSQ/0d0ArE9fklhvMjw57UtImAPl9YT/H03aSViUluJ9GxC9zceWXGyAingFmka5HjpBU+3JbpX383cB+kh4lXWrYnVSzq+ryviYi5uf3haQvMzsySPbtMpzklhvMjw6bAUzO3ZOBy/sxlrbL12bOA+6PiG8VBlV2uSWNyjU4JK0J7Em6FnkdcEAerTLLHBEnRMSYiBhHOnavjYhDqOjy1khaS9LwWjewF3APFd63u8tPPCmQtA/p21/t0WGn9XNIbSfpImA30t9xPAmcCFwGXAJsBjwOHBgR9TenDFiSdgX+D7ib5ddrvkS6LlfJ5Zb0VtINB0NIX2YviYhTJL2BVNNZH7gDODQiXuq/SNsvN1d+ISL2rfry5uW7NPcOBX4WEadJGklF9+3ucpIzM7PKcnOlmZlVlpOcmZlVlpOcmZlVlpOcmZlVlpOcmZlVlpOcWS9J+nJ+0v9d+UnwO/V3TGaW+LFeZr0gaRdgX2CHiHhJ0gZAj//ORdLQwrMWzayXXJMz651NgKdqPzCOiKfycwTfKekP+f/cbpE0PP/H24/zf3/dIem9AJIOl/Q/kn4FXJ3L/l3Srbl2eHL/LZ7ZwOaanFnvXA18TdKfgN8CPwduzO//HBG3SloH+Cvpv86IiO0kvZn05Pgt8+fsArw1IhZL2gsYT3oGoYAZkt6T/ybJzLrBNTmzXsj/2fYO4GhgESm5fQJYEBG35nGey02QuwIX5rIHgMeAWpKbWXjs0l75dQdwO/BmUtIzs25yTc6sl/Jf2swCZkm6GziWxn/p0uq/fV6oG+/0iPhh24I0G6RckzPrBUlbSSrWsrYnPe1/U0nvzOMMz3/3cgNwSC7bkvTw3AcbfOxVwJH5/++QNDr/V5iZdZNrcma9szbwX/lvbZYBc0hNlz/O5WuSrsftCXwf+EGu7S0DDs93ZK7wgRFxtaStgRvzsOeBQxnE/wlm1lP+FwIzM6ssN1eamVllOcmZmVllOcmZmVllOcmZmVllOcmZmVllOcmZmVllOcmZmVll/X8cKZF/jgoCSAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a42cb1fd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a30653940>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAasAAAEWCAYAAADYRbjGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzt3Xm4HFWd//H3hyQsSiAEAkISEpSggChiWFQGETECIiAzKAxgwCguqOEZN3AcEYEBHkdRZ0BFQMARMCJLdPQHEQi4AgkgOxI2ExNJIAESQDTw/f1xTudWOt19K8nt7rr3fl7P08+tOrV9a+n+1qk6t0oRgZmZWZWt0+0AzMzMeuNkZWZmledkZWZmledkZWZmledkZWZmledkZWZmlTdgkpWkiySd1qVlS9IPJC2RdGs3Ylgdku6VtHcHl3ekpOs6tbzeSBovKSQN7XYsa0PS+yTNlbRM0pu6sPyQtG2nl7smJO0taV6HlvUVSf/biWUNJm1LVpIek/SEpFcWyj4saWa7ltlFewLvAsZExG6NRpA0RtKPJD0l6TlJt0o6oN2BNUriEbFjRMxs97ILy/tRREzq1PL6Wt53P5X0pKRnJN0t6ZhuxwX8F/DJiNgwIu7odjCDUSeTYH/WF9up3TWrocDUNi+jz0kaspqTjAMei4jnmsxvJPAb4O/AjsBmwNnA5ZIOWZtYrSN+CMwl7edNgQ8CT/TlAtawljcOuHcNl7e6x7hZd0VEWz7AY8CJwGJgRC77MDAzd48HAhhamGYm8OHcfQzwW9KP+tPAI8Bbc/lcYCEwuTDtRcB3gRnAUuAmYFxh+OvysMXAg8D766b9DvAL4Dlg3wbrsxUwPU8/B/hILp8C/A14CVgGnNJg2lOBe4B16sq/kNdLvW2P3P8h4H5gCXBtbf3y9GfnbfIMcBfweuA44B+kJLkM+Flh3+ybu9cDvgnMz59vAuvlYXsD84DP5HkvAI4txHMAcF/e3n8BPtvkWDgG+E2hP4CPAQ/ldTkHUJNp1yEdRw8DTwHTgJGF4T8B/prX+2Zgx8KwDYCvA4/n4b/JZbVtPRn4M/Ak8O8tjuVlwM4thu8J/I50nM4FjsnlGwOXAItyDF+qHQOsfHwvBk5rtY/rlrdejilIx+vDuXz7fMw8TUpiB63mMX5sXvZS0nH50brhn8vHwPwcZwDbAnvkfTCkMO77gLty927A73NcC4D/AdYtezwAHynEdR+wS+E7+dO8fR8FPl237y/K87svxz6vxT4M4NN5vZ8EvkY69tbL+2enwribAy+QThZeAF7O+2NZjukrpOP0khzzvcDEwvS97adzgP/L094CvKYNx95vSDXzJXnb7V+Y50jgB3k/LwGuLgw7ELgzL+93wBvqfvNPytt7SZ7H+sArm2yn3YBZwLOkk79vtMwprQauzScHvi9wJT1fxNVNVstJX6AhwGmkH5Zz8gE0Ke/MDQs7eSmwVx7+LfIPZN5Yc/O8hgK7kA7IHQvTPgO8jXSArt9gfW4Czs0bf+d8ELyz0Y9xg2n/QOMktk3eBhNKbI9DSEly+7wOXwJ+l4e9G5gNjCAlru2BLQvrdlqjfZO7v5rj2xwYlQ/AU/OwvfM++CowjJScngc2ycMXAP+Uuzch/4g0WM+Vtk9ez5/neLfO23K/JtOekOMbk/fr94DLCsM/BAynJ+neWRh2Tt6Go0nH0FvzeLVt/X3Sj9obgReB7ZvE8CtSYjkc2Lpu2Nak4+6IvI02JSc20o/FNTm+8cCfgCl1x/en8v7coNU+bvEDu23uHpan/SKwLrBPjuu1q3GMvwd4TT6G3p73dS0x7Ef6QXk96ft0ad3yHwbeVZjXT4ATc/ebSQltaN4O9wMnlDkegMNIJ0K75ri2JSWJdUjH/Jfz+r6alGjenac7E/g16Yd3LOlksbdkdWMef+u8r2rfvXOBswrjTqXnxG/v+vmSktXfSN+XIcAZwB9WYz8tJv2QDwV+BFzeJOa1Ofb+QToJGAJ8nJSYlIf/H/Bj0nd6GPD2XL4L6aR19zzdZNJvSe3k9rG8ncfm7fhben77G22n3wNH5+4NgT1a5pTeks6afuhJVq8nfUlGsfrJ6qHCsJ3y+FsUyp4q7JyLijs1r/xLecN9APh1XXzfA04uTHtJi3UZm+c1vFB2BnBRox/jBtPPAT7WoHz9vE5vLbE9flk72HL/OqQfk3GkA/5PpB+E+trbRbROVg8DBxSGvZt0SbN2gL1QF9PC2kFFOnn4KLBRL8fCStsnr+eehf5p5B+2BtPeTz4pyP1bkr5oQxuMOyLPe+O8fV4A3thgvNq2HlMouxU4vEkMm5B+/O7Nx8GdwK552EnAVQ2mGUJKgDsUyj5Kz/F/DPDnumma7uMmcRWTxT+RajfrFIZfBnylzDHeZP5XA1Nz94XAmYVh29Ut/zTgwtw9nFR7axb3CcVt1up4INUupzaYx+4Ntt9JwA9y9yMUToBIVxl6S1bF8T8BXF9Y1lx6aiazyFdmaJ6sflXo3wF4YTX20/mFYQcADzSJeW2OvTmFYa/I6/8q0vfrZfIJad18v0M+kS2UPUhPMnuMwu9cjr1W62+0nW4GTgE2K3M8tr01YETcQzprOnENJi/eF3ghz6++bMNC/9zCcpeRzlC2Iv2g7y7p6doHOJK0c1aZtoGtgMURsbRQ9jjpjL2MJ0kHQb1a2aIS8xgHfKsQ/2LSmeboiLiBdGnlHOAJSedJ2qhkbFuR1qXm8VxW81RELC/0P0/PNv9n0gH5uKSbJL2l5DIhfWEbzbPeOOCqwnrfT0oYW0gaIulMSQ9Lepb0ZYF0T3Az0snAw2sbQ0QsiYgTI2JHYAtSsrpakkgnMo2WsRnpzLl+2xaPmfpjruk+brEONVsBcyPi5dVY3kok7S/pD5IW5+UfkNdjxfzr5l10KXCopPWAQ4HbI+LxPN/tJP1c0l/zfvrPwnxrmu2LZtt3HLBV3Xf6i6T9UybeRurH3wogIm4hJd+3S3odqXY3vZd51a/P+vm+ZJn9VPa7sTbH3oplRMTzuXPDPM/FEbGkwXzHAZ+p2+ZjWfn3ouE2bGIK6aTnAUm3STqwxbgda7p+MqnKWdxYtcYIryiUFZPHmhhb65C0IakqOp+0AW+KiBGFz4YR8fHCtNFivvOBkZKGF8q2Jl2eKONXwD9Lqt/e7yfdE3qY3rfHXNI9hOI6bBARvwOIiG9HxJtJDTi2I12j7229aus2rm695pdZqYi4LSIOJl1CvJp0RtzX5pKupxfXe/2I+Avwr8DBpBr8xqQaE6Qf+CdJl2Je05fBRMSTpGv9W5GOr7lNlvEkqQZYv22Lx0z9vmm5j3sxHxhbd4z1trwVcpL5KWndtoiIEaT7W8qjLKDw/crz7plxxH2kH6f9Sfvl0sLg7wAPABMiYiNSUhHlNNu+c4FH67bV8IiotbBtGW8T9eMXvwcXA0cBRwNXRMTfcnlv3696ZfZTWWtz7LWa50hJI5oMO71um78iIi4rjNNsG66ynSLioYg4gvT7cRZwRbH1eL2OJKuImEO6BvrpQtki0sY7Kp8hf4i1/2E5QNKektYlNWq4JSLmkmp220k6WtKw/NlV0vYl459LupdzhqT1Jb2BdFbwo5JxnQ1sBFwg6VV5HkcA/0G6FPlyie3xXeAkSTsCSNpY0mG5e1dJu0saRkp6tQYfkGqnr24R22XAlySNkrQZ6R5Ar/8jImnd/P9TG0fEP0g3SV/qbbo18F3gdEnj8nJHSTo4DxtOutzxFCnJ/2dtonzmeiHwDUlb5W36lvyjvFoknSXp9ZKG5hOWj5MuozxFOgb2lfT+PHxTSTtHxEuk5H26pOE5/n+j9bZtuo9LqJ39fz4f33sD7wUuLzn9uqT7eYuA5ZL2J90XrpkGHCNpB0mvIJ2A1ruU9B3fi3TPqmY46fhYlmsmH28wbTPnA5+V9GYl2+ZteSvwrKQvSNog79/XS9q1EO9JkjaRNIZ0b7A3n8vjjyXdl/pxYdgPSY1GjiLdD6p5AthU0sYl12dt91NRXx57AETEAtLl6HPzthgmaa88+PvAx/JvjSS9UtJ76k7ij1f6V4+RpJOS2jZcZTtJOkrSqPxdfToXN/0N6eQ/BX+VdGO26COkGsBTpBpBmTPIVi4lfYkWk27qHgmQL99NIt0gn0+qAp9F+nKWdQTpzH0+cBUpycwoM2H+UduTdFnqPlJrmEuA4yPiwsKoTbdHRFyVY748X0q5h3QWCykRfp/UAufxPP1/5WEXADvkavvVDcI7jXQN/i7gbuD2XFbG0cBjOZ6Pkb7Ife1bpEsu10laSmpssXsedglpff9C2q5/qJv2s6R1uo10TJzFmh3zryDt81qr1HHAQQAR8WfS5bLP5GXcSWqwAekH8rk8zW9Ix2dxf6+kl33cUkT8Pce0P+nM+lzggxHxQMnpl5ISzTTScfSvFC51RcQvSQ1YbiDdg72hwWwuI92buCHXQGs+m+e3lHSc/njVSZvG9RPgdNK2W0qqwY/MP8jvJTV2epS0zueTatiQ7oU8noddR0o2vbmG1GjjTlIjgwsKccwjfTeC1HCjVv5AXu9H8nes1WWvtd5PdfPqs2OvztGkmtkDpHvUJ+TlzSL9Rv0P6RiZQ7r/VXQpaXs/kj+n5Wkbbaf9gHslLSN9zw8v1FhXUWv9YR2kdD/pt6Sbo1/udjxmg52kIF2mnNNinAuB+RHxpc5F1n9IeozUIOxX7Zj/gHncUn8SEc+SzohekrS29+nMrM0kjSc1HLmg9ZjWLk5WXRIRcyPilIj4a+9jm1m3SKr9U//XIuLRbsczWPkyoJmZVZ5rVmZmVnn9+hUJzWy22WYxfvz4bodhZtavzJ49+8mIGNXtOBoZkMlq/PjxzJo1q9thmJn1K5LKPOmjK3wZ0MzMKs/JyszMKs/JyszMKs/JyszMKs/JyszMKs/JyszMKs/JyszMKs/JyszMKs/JyszMKm9APsFibd195c0runc6dK8WY5qZWSe4ZmVmZpXnZGVmZpXnZGVmZpXnZGVmZpXnZGVmZpXnZGVmZpXnZGVmZpXnZGVmZpXnZGVmZpXnZGVmZpXnZGVmZpXnZGVmZpXnZGVmZpXnZGVmZpXnZGVmZpXnZGVmZpXnZGVmZpXnZGVmZpXnZGVmZpXnZGVmZpXnZGVmZpXX9mQlaYikOyT9PPdvI+kWSQ9J+rGkdXP5erl/Th4+vjCPk3L5g5Le3e6YzcysWjpRs5oK3F/oPws4OyImAEuAKbl8CrAkIrYFzs7jIWkH4HBgR2A/4FxJQzoQt5mZVURbk5WkMcB7gPNzv4B9gCvyKBcDh+Tug3M/efg78/gHA5dHxIsR8SgwB9itnXGbmVm1tLtm9U3g88DLuX9T4OmIWJ775wGjc/doYC5AHv5MHn9FeYNpVpB0nKRZkmYtWrSor9fDzMy6qG3JStKBwMKImF0sbjBq9DKs1TQ9BRHnRcTEiJg4atSo1Y7XzMyqa2gb5/024CBJBwDrAxuRalojJA3NtacxwPw8/jxgLDBP0lBgY2BxobymOI2ZmQ0CbatZRcRJETEmIsaTGkjcEBFHAjcC/5JHmwxck7un537y8BsiInL54bm14DbABODWdsVtZmbV086aVTNfAC6XdBpwB3BBLr8A+KGkOaQa1eEAEXGvpGnAfcBy4PiIeKnzYZuZWbd0JFlFxExgZu5+hAat+SLib8BhTaY/HTi9fRGamVmV+QkWZmZWeU5WZmZWeU5WZmZWeU5WZmZWeU5WZmZWeU5WZmZWeU5WZmZWeU5WZmZWeU5WZmZWeU5WZmZWeU5WZmZWeU5WZmZWeU5WZmZWeU5WZmZWeU5WZmZWeU5WZmZWed14U3C/cveVN6/o3unQvboYiZnZ4NVrzUrSVEkbKblA0u2SJnUiODMzMyh3GfBDEfEsMAkYBRwLnNnWqMzMzArKJCvlvwcAP4iIPxbKzMzM2q5Mspot6TpSsrpW0nDg5faGZWZm1qNMA4spwM7AIxHxvKRNSZcCzczMOqLXZBURL0t6AthBklsPmplZx/WafCSdBXwAuA94KRcHcHPTiczMzPpQmZrSIcBrI+LFdgdjZmbWSJkGFo8Aw9odiJmZWTNlalbPA3dKuh5YUbuKiE+3LSozM7OCMslqev6YmZl1RZnWgBdLWhfYLhc9GBH/aG9YZmZmPcq0BtwbuBh4jPTkirGSJkeEWwOamVlHlLkM+HVgUkQ8CCBpO+Ay4M3tDMzMzKymTGvAYbVEBRARf8KtA83MrIPK1KxmSboA+GHuPxKY3b6QzMzMVlYmWX0cOB74NOme1c3Aue0MyszMrKhMa8AXgW/kj5mZWcc1TVaSpkXE+yXdTXoW4Eoi4g1tjczMzCxrVbOamv8euCYzlrQ+6ZLhenk5V0TEyZK2AS4HRgK3A0dHxN8lrQdcQmpl+BTwgYh4LM/rJNKrSl4CPh0R165JTGZm1j81bQ0YEQty5yci4vHiB/hEiXm/COwTEW8kvQ9rP0l7AGcBZ0fEBGAJKQmR/y6JiG2Bs/N4SNoBOBzYEdgPOFfSkNVdUTMz67/KNF1/V4Oy/XubKJJluXdY/gSwD3BFLr+Y9FR3gINzP3n4OyUpl18eES9GxKPAHGC3EnGbmdkA0TRZSfp4vl/1Okl3FT6PAneVmbmkIZLuBBYCM4CHgacjYnkeZR4wOnePBuYC5OHPAJsWyxtMU1zWcZJmSZq1aNGiMuGZmVk/0eqe1aXAL4EzgBML5UsjYnGZmUfES8DOkkYAVwHbNxot/1WTYc3K65d1HnAewMSJE1cZbmZm/Vere1bP5AYOXwL+mu9VbQMclZNPaRHxNDAT2AMYIamWJMcA83P3PGAsQB6+MbC4WN5gGjMzGwTK3LP6KfCSpG2BC0gJ69LeJpI0qpbUJG0A7AvcD9wI/EsebTJwTe6envvJw2+IiMjlh0taL7cknADcWiJuMzMbIMo8weLliFgu6VDgmxHx35LuKDHdlsDFueXeOsC0iPi5pPuAyyWdBtxBSoDkvz+UNIdUozocICLulTQNuA9YDhyfLy+amdkgUSZZ/UPSEcAHgffmsl4fZBsRdwFvalD+CA1a80XE34DDmszrdOD0ErGamdkAVOYy4LHAW4DTI+LRfCnuf9sblpmZWY8yzwa8T9IXgK1z/6PAme0OzMzMrKbXmpWk9wJ3Av8v9+8saXq7AzMzM6spcxnwK6R7TE8DRMSdpBaBZmZmHVEmWS2PiGfqyvxPt2Zm1jFlWgPeI+lfgSGSJpBewvi79oZlZmbWo0zN6lOkJ56/CFwGPAuc0M6gzMzMisq0Bnwe+Pf8MTMz67hek5WkG2n84Nh92hKRmZlZnTL3rD5b6F4f+GfSY4/MzMw6osxlwNl1Rb+VdFOb4jEzM1tFmcuAIwu96wBvBl7VtojMzMzqlLkMOJuelyAuBx4FprQzKDMzs6IylwH9tAozM+uqMpcBD201PCKu7LtwzMzMVlXmMuAU4K3ADbn/HaRX1D9DujzoZGVmZm1VJlkFsENELACQtCVwTkQc29bIzMzMsjKPWxpfS1TZE8B2bYrHzMxsFWVqVjMlXUt6LmAAhwM3tjUqMzOzgjKtAT8p6X3AXrnovIi4qr1hmZmZ9ShTsyInJycoMzPrijL3rMzMzLrKycrMzCqvabKSdH3+e1bnwjEzM1tVq3tWW0p6O3CQpMtJzwZcISJub2tkZmZmWatk9WXgRGAM8I26YQH45YtmZtYRTZNVRFwBXCHpPyLi1A7GZGZmtpIy/2d1qqSD6Pk/q5kR8fP2hmVmZtaj19aAks4ApgL35c/UXGZmZtYRZf4p+D3AzhHxMoCki4E7gJPaGZiZmVlN2f+zGlHo3rgdgZiZmTVTpmZ1BnCHpBtJzdf3wrUqMzProDINLC6TNBPYlZSsvhARf213YGZmZjVlH2S7AJje5ljMzMwa8rMBzcys8pyszMys8lomK0nrSLqnU8GYmZk10jJZ5f+t+qOkrVd3xpLGSrpR0v2S7pU0NZePlDRD0kP57ya5XJK+LWmOpLsk7VKY1+Q8/kOSJq9uLGZm1r+VaWCxJXCvpFuB52qFEXFQL9MtBz4TEbdLGg7MljQDOAa4PiLOlHQi6WG5XwD2Bybkz+7Ad4DdJY0ETgYmkh6gO1vS9IhYshrraWZm/ViZZHXKmsw4tyBckLuXSrofGA0cDOydR7sYmElKVgcDl0REAH+QNELSlnncGRGxGCAnvP2Ay9YkLjMz63/K/J/VTZLGARMi4leSXgEMWZ2FSBoPvAm4BdgiJzIiYoGkzfNoo4G5hcnm5bJm5fXLOA44DmDrrVf7quVqu/vKm1d073ToXi3GNDOztVXmQbYfAa4AvpeLRgNXl12ApA2BnwInRMSzrUZtUBYtylcuiDgvIiZGxMRRo0aVDc/MzPqBMk3XjwfeBjwLEBEPAZu3nCKTNIyUqH4UEVfm4ify5T3y34W5fB4wtjD5GGB+i3IzMxskyiSrFyPi77UeSUNpULOpJ0nABcD9EVF80/B0oNaibzJwTaH8g7lV4B7AM/ly4bXAJEmb5JaDk3KZmZkNEmUaWNwk6YvABpLeBXwC+FmJ6d4GHA3cLenOXPZF4ExgmqQpwJ+Bw/KwXwAHAHOA54FjASJisaRTgdvyeF+tNbYwM7PBoUyyOhGYAtwNfJSUVM7vbaKI+A2N7zcBvLPB+EG65NhoXhcCF5aI1czMBqAyrQFfzi9cvIV0+e/BnFjMzMw6otdkJek9wHeBh0k1pW0kfTQiftnu4MzMzKDcZcCvA++IiDkAkl4D/B/gZGVmZh1RpjXgwlqiyh6hp7m5mZlZ2zWtWUk6NHfeK+kXwDTSPavD6GmZZ2Zm1natLgO+t9D9BPD23L0I2KRtEZmZmdVpmqwi4thOBmJmZtZMmdaA2wCfAsYXxy/xihAzM7M+UaY14NWkxyb9DHi5veGYmZmtqkyy+ltEfLvtkZiZmTVRJll9S9LJwHXAi7XCiLi9bVGZmZkVlElWO5EeSLsPPZcBI/ebmZm1XZlk9T7g1cXXhJiZmXVSmSdY/BEY0e5AzMzMmilTs9oCeEDSbax8z8pN183MrCPKJKuT2x6FmZlZC2XeZ3VTJwIxMzNrpswTLJaSWv8BrAsMA56LiI3aGZiZmVlNmZrV8GK/pEOA3doWkZmZWZ0yrQFXEhFX4/+xMjOzDipzGfDQQu86wER6LguamZm1XZnWgMX3Wi0HHgMObks0ZmZmDZS5Z+X3WpmZWVe1eq39l1tMFxFxahviMTMzW0WrmtVzDcpeCUwBNgWcrMzMrCNavdb+67VuScOBqcCxwOXA15tNZ2Zm1tda3rOSNBL4N+BI4GJgl4hY0onAzMzMalrds/oacChwHrBTRCzrWFRmZmYFrf4p+DPAVsCXgPmSns2fpZKe7Ux4ZmZmre9ZrfbTLczMzNrBCcnMzCrPycrMzCrPycrMzCqvzLMBbTXcfeXNK7p3OnSvLkZiZjZwuGZlZmaV52RlZmaV17ZkJelCSQsl3VMoGylphqSH8t9NcrkkfVvSHEl3SdqlMM3kPP5Dkia3K14zM6uudtasLgL2qys7Ebg+IiYA1+d+gP2BCflzHPAdWPG4p5OB3YHdgJNrCc7MzAaPtiWriLgZWFxXfDDpGYPkv4cUyi+J5A/ACElbAu8GZkTE4vxMwhmsmgDNzGyA6/Q9qy0iYgFA/rt5Lh8NzC2MNy+XNStfhaTjJM2SNGvRokV9HriZmXVPVRpYqEFZtChftTDivIiYGBETR40a1afBmZlZd3U6WT2RL++R/y7M5fOAsYXxxgDzW5Sbmdkg0ulkNR2oteibDFxTKP9gbhW4B/BMvkx4LTBJ0ia5YcWkXGZmZoNI255gIekyYG9gM0nzSK36zgSmSZoC/Bk4LI/+C+AAYA7wPOmNxETEYkmnArfl8b4aEfWNNszMbIBrW7KKiCOaDHpng3EDOL7JfC4ELuzD0MzMrJ+pSgMLMzOzppyszMys8pyszMys8pyszMys8pyszMys8pyszMys8pyszMys8vxa+w7x6+7NzNaca1ZmZlZ5TlZmZlZ5TlZmZlZ5TlZmZlZ5TlZmZlZ5TlZmZlZ5TlZmZlZ5TlZmZlZ5TlZmZlZ5TlZmZlZ5TlZmZlZ5fjZgl/mZgWZmvXPNyszMKs/JyszMKs/JyszMKs/JyszMKs/JyszMKs+tASvKrQTNzHq4ZmVmZpXnZGVmZpXnZGVmZpXne1b9kO9nmdlg45qVmZlVnpOVmZlVni8DDiC+PGhmA5VrVmZmVnmuWQ0CrnGZWX/nZDWIOYmZWX/hZGWrcBIzs6rpN8lK0n7At4AhwPkRcWaXQxp0miUxJzcza7d+kawkDQHOAd4FzANukzQ9Iu7rbmTWm9VNcE6IZtZIv0hWwG7AnIh4BEDS5cDBgJPVILc2SW9NpulW+drOq5m+PDnoTyca7Y6piuvc3/WXZDUamFvonwfsXhxB0nHAcbl3maQH12J5mwFPrsX0/c1gW1/wOg8WXufVM64vA+lL/SVZqUFZrNQTcR5wXp8sTJoVERP7Yl79wWBbX/A6DxZe54Gjv/xT8DxgbKF/DDC/S7GYmVmH9ZdkdRswQdI2ktYFDgemdzkmMzPrkH5xGTAilkv6JHAtqen6hRFxbxsX2SeXE/uRwba+4HUeLLzOA4QiovexzMzMuqi/XAY0M7NBzMnKzMwqz8mqQNJ+kh6UNEfSid2Opx0kXShpoaR7CmUjJc2Q9FD+u0k3Y+xrksZKulHS/ZLulTQ1lw/Y9Za0vqRbJf0xr/MpuXwbSbfkdf5xbrA0YEgaIukOST/P/QN9fR+TdLekOyXNymUD8rh2ssoKj3TaH9gBOELSDt2Nqi0uAvarKzsRuD4iJgDX5/6BZDnwmYjYHtgDOD7v24G83i8C+0TEG4Gdgf0k7QGcBZyd13kJMKWLMbbDVOD+Qv9AX1+Ad0TEzoX/rRqQx7WTVY8Vj3SKiL8DtUc6DSgRcTOwuK74YODi3H0xcEhHg2qziFgQEbfn7qWkH7PRDOD1jmRZ7h2WPwHsA1yRywfUOksaA7wHOD/3iwG8vi0MyOPayapHo0c6je5SLJ2KysJhAAADjUlEQVS2RUQsgPTDDmze5XjaRtJ44E3ALQzw9c6XxO4EFgIzgIeBpyNieR5loB3j3wQ+D7yc+zdlYK8vpBOQ6yTNzo+cgwF6XPeL/7PqkF4f6WT9m6QNgZ8CJ0TEs+nEe+CKiJeAnSWNAK4Ctm80Wmejag9JBwILI2K2pL1rxQ1GHRDrW/C2iJgvaXNghqQHuh1Qu7hm1WMwP9LpCUlbAuS/C7scT5+TNIyUqH4UEVfm4gG/3gAR8TQwk3S/boSk2knqQDrG3wYcJOkx0iX8fUg1rYG6vgBExPz8dyHphGQ3Buhx7WTVYzA/0mk6MDl3Twau6WIsfS7fu7gAuD8ivlEYNGDXW9KoXKNC0gbAvqR7dTcC/5JHGzDrHBEnRcSYiBhP+u7eEBFHMkDXF0DSKyUNr3UDk4B7GKDHtZ9gUSDpANLZWO2RTqd3OaQ+J+kyYG/SawSeAE4GrgamAVsDfwYOi4j6Rhj9lqQ9gV8Dd9NzP+OLpPtWA3K9Jb2BdHN9COmkdFpEfFXSq0k1j5HAHcBREfFi9yLte/ky4Gcj4sCBvL553a7KvUOBSyPidEmbMgCPaycrMzOrPF8GNDOzynOyMjOzynOyMjOzynOyMjOzynOyMjOzynOyMlsLkv49P9X8rvzk6927HZPZQOTHLZmtIUlvAQ4EdomIFyVtBqzxKygkDS08x87MClyzMltzWwJP1v7JNCKezM9p21XS7/K7pG6VNDy/X+oH+d1Dd0h6B4CkYyT9RNLPgOty2eck3ZZra6d0b/XMqsM1K7M1dx3wZUl/An4F/Bj4ff77gYi4TdJGwAuk9ywRETtJeh3pSdnb5fm8BXhDRCyWNAmYQHrGm4DpkvbKr3YxG7RcszJbQ/l9UW8GjgMWkZLUR4EFEXFbHufZfGlvT+CHuewB4HGglqxmFB6HMyl/7gBuB15HSl5mg5prVmZrIb+GYyYwU9LdwPE0fg1Fq/eRPFc33hkR8b0+C9JsAHDNymwNSXqtpGKtZ2fSk823krRrHmd4fkXFzcCRuWw70kNGH2ww22uBD+V3byFpdH5Xkdmg5pqV2ZrbEPjv/CqO5cAc0iXBH+TyDUj3q/YFzgW+m2tfy4FjcgvClWYYEddJ2h74fR62DDiKAfJOIrM15aeum5lZ5fkyoJmZVZ6TlZmZVZ6TlZmZVZ6TlZmZVZ6TlZmZVZ6TlZmZVZ6TlZmZVd7/B/krQfHdtadcAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a33640ac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZ4AAAEWCAYAAABWn/G6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzt3Xm4HFW57/Hvj4SZQBgCkgESlRkVIUyKigwxgAIiKBzQgFEcUHHgKChHUOAA9x5F9KjAFWQQGQ4IREEGgeB0ICSAMomEMSGRBMKsoIH3/rFWJ5Wmd+/aQ1fv3fv3eZ797KpV07uqV9fbq6q6WhGBmZlZVZZrdwBmZja0OPGYmVmlnHjMzKxSTjxmZlYpJx4zM6uUE4+ZmVVqyCceSedKOrFN25akn0p6RtKMdsTQE5LulbRzhds7WNL1VW2vO5LGSwpJw9sdS19I+qCkOZJelPT2fljfdEmf6I/YCut8l6QH+nOdvYwjJL253XF0mgGXeCQ9KulJSasWyj4haXobw2qVnYDdgbERsV2jGSSNlXShpKclvSRphqQ9Wx1Yo4QcEVtExPRWb7uwvQsjYlJV2+tv+bW7XNJTkp6TdLekQ9sdF/BfwOciYrWIuLPdwTQSEb+LiE2q3GYrEmgn6o/9NOASTzYcOLLdQfSUpGE9XGRD4NGIeKmL9a0F/B74J7AFsA5wGnCxpH37EqtV4gJgDul1Xhv4GPBkf26gl72vDYF7e7m9nrZxs9eLiAH1BzwKHA0sAkbmsk8A0/PweCCA4YVlpgOfyMOHAn8gHaCfBR4G3pHL5wALgCmFZc8FzgBuAF4AbgE2LEzfNE9bBDwAfLhu2R8D1wAvAbs1qM9oYFpefjbwyVw+FXgZeBV4EfhWg2VPAO4Blqsr/1qul7rbH3n848D9wDPAdbX65eVPy/vkOeDPwJbA4cC/SAnvReCXhddmtzy8IvA9YF7++x6wYp62MzAX+Epe93zgsEI8ewL35f39BHBUF23hUOD3hfEAPg08mOvyQ0BdLLscqR09BDwNXAqsVZj+P8Dfcr1/C2xRmLYy8B3gsTz997mstq+nAI8DTwHfaNKWXwS2ajJ9J+CPpHY6Bzg0l68BnA8szDEcW2sDLNu+FwEnNnuN67a3Yo4pSO31oVy+WW4zz5IS0t49bOPTgZOBGXl/XdWDfd2wLdTaUGG+ccAv8j55GvjvLvbp8cBlwCV5nXcAb8vT/h24vG7+H5Da7kmk9+LLeR/9d3dtjtTGjs2v0YL8mq1Rd5wq21Yatrk8be/8ujyb9/VmdcfLo0jv3edyvVcqTN8HuAt4nvRemFxoY2eT3ptPACcCw+ra2A/yOv8C7JqnvW4/0cVxpOlxvkwyqPIv78jdciOrval6mngWA4cBw/IOfTw3mBWBSblBrlZ4Y70AvDtPP518sANWJR0QDiP1wrbODWiLwrLPAe/MjXClBvW5BfgRsBKwFemNs2sh1t832Re30jghTcj7YKMS+2NfUsLbLNfhWOCPedr7gFnAyNx4NgPWL9TtxEavTR7+do5vXWAU6QB6QuGgsTjPszzp4PJ3YM08fT7wrjy8JrB1F/VfZv/kev4qx7tB3peTu1j2izm+sfl1PRO4qDD948AIlibQuwrTfpj34RhSG3pHnq+2r/8f6UDxNuAVCgeCuhh+Q3oDHwhsUDdtA1K7Oyjvo7XJSYp0ALsqxzce+Cswta59fz6/nis3e427iCuAN+fh5fOyXwdWAHbJcW3SgzY+nXTw2pL0nrkc+FnJfd2wLVBIPPk1+BPp4LYq6b20Uxd1O570oWn/XLejgEfy8Pqk5Fn7QDucdLDcpv59U6bN5XrNBt4IrEY6Zl1Qd5wq21a6anMb55h3z3X4at7mCoX35AzSB9y1SB8+Pp2nbZdfu93zazcG2DRPu5L0nliV9B6eAXyqro19KW/zI3k9azXaTzQ5jnTZBvsjWfTnH0sTz5a5sqPoeeJ5sDDtLXn+9QplT7P0TX4ucHFh2mqkjD4u7/Df1cV3JnBcYdnzm9RlXF7XiELZycC5hVibJZ7ZtUZUV75SrtM7SuyPX5MPWnl8OVIS2JB0kPkrsAOv71WdS/PE8xCwZ13je7Rw0PhHXUwLgB3y8OPAp4DVu2kLy+yfXM+dCuOXAkd3sez95ASfx9cnHZCGN5h3ZF73Gnn//IP8Kbluvtq+HlsomwEc2EUMawKnkD6tvkr65LltnnYMcEWDZYaRDlCbF8o+xdL2fyjweN0yXb7GXcRVTDzvIvVGlitMvwg4vkwbL7S3Uwrjm5N6y8Oa7etmbYFlE8+OpAP+6167Bus/Hri1bl8Uk9uvWXrW4f3AfY3eN2XaHHAj8NnCtE1qbawnbaWbNvcfwKV18z4B7Fx4Tx5SmP5/gDPy8JnAaQ3WuV5uYysXyg4Cbi60sXkUzibk2D/aaD/R5DjS1d9AvcZDRNxD+qRxdC8WL55H/0deX33ZaoXxOYXtvkg6hTGadHDeXtKztT/gYOANjZZtYDSwKCJeKJQ9RvrkUcZTpANmvVrZwhLr2BA4vRD/ItKnkjERcROpq/xD4ElJZ0lavWRso0l1qXksl9U8HRGLC+N/Z+k+/xCpF/SYpFsk7Vhym5AOko3WWW9D4IpCve8nHfzXkzRM0imSHpL0POnNC+ka2jqkxP5QX2OIiGci4uiI2IL0Zr8LuFKSSB9KGm1jHVLPo37fFttMfZvr8jVuUoea0cCciHitB9trpDjPY6RPyut0s6+hXFsYBzxW155KxZLrNZelbfM84JA8fAjpOlx3unq9G70HhpNe6+6WLWrW5pbZRq7PHJZ9fbraRldtbEPS6zO/0GbOJPV8ap6InFWy+vf3Er05jgzYxJMdB3ySZXdy7UL8KoWyYiLojXG1AUmrkbqs80gv8C0RMbLwt1pEfKawbNC1ecBakkYUyjYgfWIp4zfAhyTVv04fJr2ZHqL7/TGH1IUu1mHliPgjQER8PyK2Id28sDHpPHh39arVbcO6es0rU6mIuD0i9iE19CtJnyL72xxgj7p6rxQRTwD/Rjr3vRuplzM+LyNSsn8ZeFN/BhMRT5HuJqudEpnTxTaeIn1qrt+3xTZT/9o0fY27MQ8YV9fGutteI+MKwxuQ6vAUzfd12bYwB9igBzdSFN/Py5FOt9ba5pXAWyVtSerxXFhYrkw9ixq9BxbT8xtImrW5ZbZR+NBS5hjSVRubQ+rxrFNoL6vnD0g1Y/K2aorv79ftpybHkYYGdOKJiNmki2VfKJQtJO30Q/KnqY/T94PEnpJ2krQC6YL+bRExh9Tj2ljSRyUtn/+2lbRZyfjnkK59nCxpJUlvJd1UcGHzJZc4DVgdOFvSG/I6DiJ1v4+LiNdK7I8zgGMkbQEgaQ1JB+ThbSVtL2l5UgKr3ewA6c3zxiaxXQQcK2mUpHWAbwI/665CklbI389ZIyL+Rbro+Wp3y/XCGcBJkjbM2x0laZ88bQTpjfc0KWH/Z22h/InyHOC7kkbnfbqjpBV7GoCkUyVtKWl4/vDxGWB2RDxNagO7Sfpwnr62pK0i4lXSwfckSSNy/F+m+b7t8jUu4TbSa//V3L53Bj4AXNzD6h4iaXNJq5Cu7V2W69Llvu5BW5hBOl12iqRV8/vgnU1i2UbSfjlRfTFv/1aAiHiZdPPBz4EZEfF4Ybnu2ny9i4AvSZqQP7D+J3BJD3pm5JiatblLgb0k7Zrfp1/J9SnzoeJs4LC87HKSxkjaNCLmA9cD35G0ep72JknvKSy7LvCF3CYOIF23uSZPW2Y/dXMcaWhAJ57s26QLYEWfJGXUp0kZtsyL0MzPSb2rRcA2pNNp5FNkk0gXh+eRurSnki76lXUQ6VPePOAKUsK4ocyC+QC1E6kbfh/pLpLzgSMi4pzCrF3uj4i4Isd8cT7VcQ+wR568Ouni5zOkrvTTpE/lkBrt5rkrfmWD8E4EZpLuYLmbdPdQ2S/ifhR4NMfzaZae+uhPp5PuJrxe0gukA8/2edr5pPo+Qdqvt9YtexSpTreT2sSp9O69sgrpNa/dXbkh6Q4l8gFvT9KBZBHpNNzb8nKfJ72BHybd3fRz0oGpoW5e46Yi4p85pj1In7x/BHwsIv7Sg3pCOmV1Luk9shJLPyx2t6+7bQs5gX0AeDPpmtBc0vXXrlyVpz+T179fTmw155Gu/dafZjsd2F/pC93fb7L+mnPyOn5LuoHhZdJr1xsN21xEPEDaJz8gvT4fAD6QX7emImIG6cao00jXy29hae/pY6RTuveR9tNlLHta/zbSzUtPke5k2z8fj+D1+6nZcaSh2m2BNgjk86Z/IF2U/ma74zEbaCQdT7pxossPM5I2IN0i/IaIeL6q2AYLpS85fyIidmrVNgZDj8ey/CbZE3hVUl+va5kNOfmaz5dJd7I66bTJoH7m1FCUrxt9q91xmA02So/hepJ0Omhym8MZ0nyqzczMKuVTbWZmVqmOPNW2zjrrxPjx49sdhpnZoDJr1qynImJUq7fTkYln/PjxzJw5s91hmJkNKpIe636uvvOpNjMzq5QTj5mZVcqJx8zMKuXEY2ZmlXLiMTOzSjnxmJlZpZx4zMysUk48ZmZWKSceMzOrVEc+uaCv5l197ZLh0Xv5IbZmZv3JPR4zM6uUE4+ZmVXKicfMzCrlxGNmZpVy4jEzs0o58ZiZWaWceMzMrFJOPGZmViknHjMzq5QTj5mZVcqJx8zMKuXEY2ZmlXLiMTOzSjnxmJlZpZx4zMysUk48ZmZWKSceMzOrlBOPmZlVyonHzMwq1dLEI+lLku6VdI+kiyStJGmCpNskPSjpEkkr5HlXzOOz8/TxhfUck8sfkPS+VsZsZmat1bLEI2kM8AVgYkRsCQwDDgROBU6LiI2AZ4CpeZGpwDMR8WbgtDwfkjbPy20BTAZ+JGlYq+I2M7PWavWptuHAypKGA6sA84FdgMvy9POAffPwPnmcPH1XScrlF0fEKxHxCDAb2K7FcZuZWYu0LPFExBPAfwGPkxLOc8As4NmIWJxnmwuMycNjgDl52cV5/rWL5Q2WWULS4ZJmSpq5cOHC/q+QmZn1i1aealuT1FuZAIwGVgX2aDBr1BbpYlpX5csWRJwVERMjYuKoUaN6F7SZmbVcK0+17QY8EhELI+JfwC+AdwAj86k3gLHAvDw8FxgHkKevASwqljdYxszMBplWJp7HgR0krZKv1ewK3AfcDOyf55kCXJWHp+Vx8vSbIiJy+YH5rrcJwEbAjBbGbWZmLTS8+1l6JyJuk3QZcAewGLgTOAu4GrhY0om57Oy8yNnABZJmk3o6B+b13CvpUlLSWgwcERGvtipuMzNrrZYlHoCIOA44rq74YRrclRYRLwMHdLGek4CT+j1AMzOrnJ9cYGZmlXLiMTOzSjnxmJlZpZx4zMysUk48ZmZWKSceMzOrlBOPmZlVqqXf4+kE866+dsnw6L0mtzESM7PO4B6PmZlVyonHzMwq5cRjZmaVcuIxM7NKOfGYmVmlnHjMzKxSTjxmZlYpJx4zM6uUE4+ZmVXKicfMzCrlxGNmZpXqNvFIOlLS6krOlnSHpElVBGdmZp2nTI/n4xHxPDAJGAUcBpzS0qjMzKxjlUk8yv/3BH4aEX8qlJmZmfVImcQzS9L1pMRznaQRwGutDcvMzDpVmd/jmQpsBTwcEX+XtDbpdJuZmVmPdZt4IuI1SU8Cm0vyD8eZmVmfdJtIJJ0KfAS4D3g1Fwfw2xbGZWZmHapMD2ZfYJOIeKXVwZiZWecrc3PBw8DyrQ7EzMyGhjI9nr8Dd0m6EVjS64mIL7QsKjMz61hlEs+0/GdmZtZnZe5qO0/SCsDGueiBiPhXa8MyM7NOVeautp2B84BHSU8sGCdpSkT4rjYzM+uxMqfavgNMiogHACRtDFwEbNPKwMzMrDOVuatt+VrSAYiIv+K73MzMrJfK9HhmSjobuCCPHwzMal1IZmbWycokns8ARwBfIF3j+S3wo1YGZWZmnavMXW2vAN/Nf2ZmZn3S5TUeSZfm/3dL+nP9X5mVSxop6TJJf5F0v6QdJa0l6QZJD+b/a+Z5Jen7kmbnbWxdWM+UPP+Dkqb0tdJmZtY+zXo8R+b/7+/D+k8Hro2I/fN3gVYBvg7cGBGnSDoaOBr4GrAHsFH+2x74MbC9pLWA44CJpIeTzpI0LSKe6UNcZmbWJl32eCJifh78bEQ8VvwDPtvdiiWtDrwbODuv758R8SywD+l7QeT/++bhfYDzI7kVGClpfeB9wA0RsSgnmxuAyT2uqZmZDQhlbqfevUHZHiWWeyOwEPippDsl/UTSqsB6taSW/6+b5x8DzCksPzeXdVW+DEmHS5opaebChQtLhGdmZu3Q7BrPZyTdDWxad33nEaDMNZ7hwNbAjyPi7cBLpNNqXW6yQVk0KV+2IOKsiJgYERNHjRpVIjwzM2uHZtd4fg78GjiZZRPGCxGxqMS65wJzI+K2PH5ZXs+TktaPiPn5VNqCwvzjCsuPBebl8p3ryqeX2L6ZmQ1Aza7xPBcRjwLHAn/L13YmAIdIGtndiiPib8AcSZvkol1Jv2I6DajdmTYFuCoPTwM+lu9u2wF4Lp+Kuw6YJGnNfAfcpFxmZmaDUJkvkF4OTJT0ZtKNAtNIvaE9Syz7eeDCfEfbw8BhpGR3qaSpwOPAAXnea/I6Z5N+A+gwgIhYJOkE4PY837dL9rjMzGwAKpN4XouIxZL2A74XET+QdGeZlUfEXaTboOvt2mDeID0hodF6zgHOKbNNMzMb2Mrc1fYvSQcBHwN+lcv8kFAzM+uVMonnMGBH4KSIeETSBOBnrQ3LzMw6VZlntd0n6WvABnn8EeCUVgdmZmadqdsej6QPAHcB1+bxrSRNa3VgZmbWmcqcajse2A54FpbcMDChhTGZmVkHK5N4FkfEc3Vlr3tygJmZWRllbqe+R9K/AcMkbUT6Qbg/tjYsMzPrVGV6PJ8HtgBeAS4Cnge+2MqgzMysc5W5q+3vwDfyn5mZWZ90m3gk3Uzjp0Hv0pKIzMyso5W5xnNUYXgl4EPA4taEY2Zmna7MqbZZdUV/kHRLi+IxM7MOV+ZU21qF0eWAbYA3tCwiMzPraGVOtc1i6S+BLgYeAaa2MigzM+tcZU61+SkFZmbWb8qcatuv2fSI+EX/hWNmZp2uzKm2qcA7gJvy+HuB6cBzpFNwQzLxzLv62iXDo/ea3MZIzMwGlzKJJ4DNI2I+gKT1gR9GxGEtjczMzDpSmUfmjK8lnexJYOMWxWNmZh2uTI9nuqTrSM9pC+BA4OaWRmVmZh2rzF1tn5P0QeDdueisiLiitWGZmVmnKtPjIScaJxszM+uzMtd4zMzM+o0Tj5mZVarLxCPpxvz/1OrCMTOzTtfsGs/6kt4D7C3pYtKz2paIiDtaGpmZmXWkZonnm8DRwFjgu3XTAvAPwZmZWY91mXgi4jLgMkn/EREnVBiTmZl1sDLf4zlB0t4s/R7P9Ij4VWvDMjOzTtXtXW2STgaOBO7Lf0fmMjMzsx4r8wXSvYCtIuI1AEnnAXcCx7QyMDMz60xlv8czsjC8RisCMTOzoaFMj+dk4E5JN5NuqX437u2YmVkvlbm54CJJ04FtSYnnaxHxt1YHZmZmnansQ0LnA9NaHIuZmQ0BflabmZlVyonHzMwq1TTxSFpO0j192YCkYZLulPSrPD5B0m2SHpR0iaQVcvmKeXx2nj6+sI5jcvkDkt7Xl3jMzKy9miae/N2dP0naoA/bOBK4vzB+KnBaRGwEPANMzeVTgWci4s3AaXk+JG1O+rntLYDJwI8kDetDPGZm1kZlTrWtD9wr6UZJ02p/ZVYuaSzpC6g/yeMiPVz0sjzLecC+eXifPE6evmuefx/g4oh4JSIeAWYD25XZvpmZDTxl7mr7Vh/W/z3gq8CIPL428GxELM7jc4ExeXgMMAcgIhZLei7PPwa4tbDO4jJLSDocOBxggw360kEzM7NW6rbHExG3AI8Cy+fh24Fuf4tH0vuBBRExq1jcaBPdTGu2TDHOsyJiYkRMHDVqVHfhmZlZm5R5SOgnSae+zsxFY4ArS6z7naQfkXsUuJh0iu17wEhJtZ7WWGBeHp4LjMvbHE56NM+iYnmDZczMbJApc43nCFISeR4gIh4E1u1uoYg4JiLGRsR40s0BN0XEwcDNwP55tinAVXl4Wh4nT78pIiKXH5jvepsAbATMKBG3mZkNQGWu8bwSEf9M1/mX9EZed6qrB74GXCzpRNJTrs/O5WcDF0iaTerpHAgQEfdKupT0kwyLgSMi4tU+bN/MzNqoTOK5RdLXgZUl7Q58FvhlTzYSEdOB6Xn4YRrclRYRLwMHdLH8ScBJPdmmmZkNTGVOtR0NLATuBj4FXAMc28qgzMysc5V5OvVr+cffbiOdYnsgX3sxMzPrsW4Tj6S9gDOAh0i3Nk+Q9KmI+HWrgzMzs85T5hrPd4D3RsRsAElvAq4GnHjMzKzHylzjWVBLOtnDwIIWxWNmZh2uyx6PpP3y4L2SrgEuJV3jOYD09AIzM7Mea3aq7QOF4SeB9+ThhcCaLYvIzMw6WpeJJyIOqzIQMzMbGsrc1TYB+Dwwvjh/ROzdurDMzKxTlbmr7UrS42x+CbzW2nDMzKzTlUk8L0fE91seiZmZDQllEs/pko4DrgdeqRVGRLe/yTMUzbv62iXDo/ea3MZIzMwGpjKJ5y3AR0m/p1M71RZ53MzMrEfKJJ4PAm+MiH+2OhgzM+t8ZZ5c8CdgZKsDMTOzoaFMj2c94C+SbmfZazy+ndrMzHqsTOI5ruVRmJnZkFHm93huqSIQMzMbGso8ueAF0l1sACsAywMvRcTqrQzMzMw6U5kez4jiuKR9ge1aFpGZmXW0Mne1LSMirsTf4TEzs14qc6ptv8LocsBElp56MzMz65Eyd7UVf5dnMfAosE9LojEzs45X5hqPf5fHzMz6TbOfvv5mk+UiIk5oQTxmZtbhmvV4XmpQtiowFVgbcOIxM7Mea/bT19+pDUsaARwJHAZcDHynq+XMzMyaaXqNR9JawJeBg4HzgK0j4pkqAjMzs87U7BrP/wX2A84C3hIRL1YWlZmZdaxmXyD9CjAaOBaYJ+n5/PeCpOerCc/MzDpNs2s8PX6qgZmZWXecXMzMrFJOPGZmViknHjMzq5QTj5mZVcqJx8zMKlXm6dTWD+Zdfe2S4dF7TW5jJGZm7dWyHo+kcZJulnS/pHslHZnL15J0g6QH8/81c7kkfV/SbEl/lrR1YV1T8vwPSprSqpjNzKz1WnmqbTHwlYjYDNgBOELS5sDRwI0RsRFwYx4H2APYKP8dDvwYljy25zhge9JPbh9XS1ZmZjb4tCzxRMT8iLgjD78A3A+MIf2I3Hl5tvOAffPwPsD5kdwKjJS0PvA+4IaIWJSfE3cD4HNVZmaDVCU3F0gaD7wduA1YLyLmQ0pOwLp5tjHAnMJic3NZV+X12zhc0kxJMxcuXNjfVTAzs37S8sQjaTXgcuCLEdHsGW9qUBZNypctiDgrIiZGxMRRo0b1LlgzM2u5liYeScuTks6FEfGLXPxkPoVG/r8gl88FxhUWHwvMa1JuZmaDUCvvahNwNnB/RHy3MGkaULszbQpwVaH8Y/nuth2A5/KpuOuASZLWzDcVTMplZmY2CLXyezzvBD4K3C3prlz2deAU4FJJU4HHgQPytGuAPYHZwN9Jv3ZKRCySdAJwe57v2xGxqIVxm5lZC7Us8UTE72l8fQZg1wbzB3BEF+s6Bzin/6IzM7N28SNzzMysUk48ZmZWKSceMzOrlBOPmZlVyonHzMwq5Z9FaDP/XIKZDTXu8ZiZWaWceMzMrFJOPGZmViknHjMzq5QTj5mZVcqJx8zMKuXEY2ZmlXLiMTOzSjnxmJlZpZx4zMysUk48ZmZWKT+rbYDyM9zMrFO5x2NmZpVy4jEzs0o58ZiZWaV8jWcQ8vUfMxvM3OMxM7NKOfGYmVmlnHjMzKxSvsbTQXztx8wGA/d4zMysUk48ZmZWKZ9qGwJ8Cs7MBhL3eMzMrFLu8Qxh7gmZWTs48djrOCGZWSv5VJuZmVXKPR4rzT0hM+sPTjzWZ05IZtYTTjzWUk5KZlbPicfaoquE1F/lZjZwDZrEI2kycDowDPhJRJzS5pBsACuTqJpNK5PEnPTMemdQJB5Jw4AfArsDc4HbJU2LiPvaG5kNBf2VkPpSXnaZnsZk1g6DIvEA2wGzI+JhAEkXA/sATjxmBe1Kkv25ripOww60mHrzIaN+mUbzDFSDJfGMAeYUxucC2xdnkHQ4cHgefVHSA33Y3jrAU31YfrAZavUF13mocJ17ZsP+DKQrgyXxqEFZLDMScRZwVr9sTJoZERP7Y12DwVCrL7jOQ4XrPDANlicXzAXGFcbHAvPaFIuZmfXBYEk8twMbSZogaQXgQGBam2MyM7NeGBSn2iJisaTPAdeRbqc+JyLubeEm++WU3SAy1OoLrvNQ4ToPQIqI7ucyMzPrJ4PlVJuZmXUIJx4zM6uUE0+BpMmSHpA0W9LR7Y6nFSSdI2mBpHsKZWtJukHSg/n/mu2Msb9JGifpZkn3S7pX0pG5vGPrLWklSTMk/SnX+Vu5fIKk23KdL8k363QMScMk3SnpV3m80+v7qKS7Jd0laWYuG/Dt2oknKzyWZw9gc+AgSZu3N6qWOBeo/2rz0cCNEbERcGMe7ySLga9ExGbADsAR+bXt5Hq/AuwSEW8DtgImS9oBOBU4Ldf5GWBqG2NshSOB+wvjnV5fgPdGxFaF7+4M+HbtxLPUksfyRMQ/gdpjeTpKRPwWWFRXvA9wXh4+D9i30qBaLCLmR8QdefgF0oFpDB1c70hezKPL578AdgEuy+UdVWdJY4G9gJ/kcdHB9W1iwLdrJ56lGj2WZ0ybYqnaehExH9JBGli3zfG0jKTxwNuB2+jweufTTncBC4AbgIeAZyNicZ6l09r494CvAq/l8bXp7PpC+jBxvaRZ+bFhMAja9aD4Hk9Fun0sjw1uklYDLge+GBHPpw/EnSsiXgW2kjQSuALYrNFs1UbVGpLeDyyIiFmSdq4VN5i1I+pb8M6ImCdpXeAGSX9pd0BluMez1FB+LM+TktYHyP8XtDmefiefQ0nzAAAC50lEQVRpeVLSuTAifpGLO77eABHxLDCddH1rpKTaB85OauPvBPaW9CjpNPkupB5Qp9YXgIiYl/8vIH242I5B0K6deJYayo/lmQZMycNTgKvaGEu/y+f6zwbuj4jvFiZ1bL0ljco9HSStDOxGurZ1M7B/nq1j6hwRx0TE2IgYT3rv3hQRB9Oh9QWQtKqkEbVhYBJwD4OgXfvJBQWS9iR9Sqo9luekNofU7yRdBOxMenT6k8BxwJXApcAGwOPAARFRfwPCoCVpJ+B3wN0sPf//ddJ1no6st6S3ki4sDyN9wLw0Ir4t6Y2kHsFawJ3AIRHxSvsi7X/5VNtREfH+Tq5vrtsVeXQ48POIOEnS2gzwdu3EY2ZmlfKpNjMzq5QTj5mZVcqJx8zMKuXEY2ZmlXLiMTOzSjnxmPWBpG/kpz//OT8hePt2x2Q20PmROWa9JGlH4P3A1hHxiqR1gF4/dl/S8MJzxcw6lns8Zr23PvBU7QuJEfFUfm7WtpL+mH8LZ4akEfn3cX6afzvlTknvBZB0qKT/kfRL4Ppc9u+Sbs+9qG+1r3pmreEej1nvXQ98U9Jfgd8AlwD/m/9/JCJul7Q68A/S78QQEW+RtCnpicIb5/XsCLw1IhZJmgRsRHrmloBpkt6df87CrCO4x2PWS/n3brYBDgcWkhLOp4D5EXF7nuf5fPpsJ+CCXPYX4DGglnhuKDzSZFL+uxO4A9iUlIjMOoZ7PGZ9kH96YDowXdLdwBE0fvR+s99geKluvpMj4sx+C9JsgHGPx6yXJG0iqdgb2Yr0BOjRkrbN84zIj+X/LXBwLtuY9ADHBxqs9jrg4/m3g5A0Jv/WilnHcI/HrPdWA36Qf35gMTCbdNrtp7l8ZdL1nd2AHwFn5F7RYuDQfCfcMiuMiOslbQb8b572InAIA/A3Vcx6y0+nNjOzSvlUm5mZVcqJx8zMKuXEY2ZmlXLiMTOzSjnxmJlZpZx4zMysUk48ZmZWqf8P4kU/wW4zXUAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11bdb3668>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a34d06470>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a22330b38>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a22de3208>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a3184c208>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Print out questions in each topic cluster and organize them by popularity\n",
    "bins = np.linspace(0, 50, 100)\n",
    "for i in range(10):\n",
    "    fig, ax = plt.subplots()\n",
    "    ax.hist(qpy_dict[i]['Score'], bins, alpha=0.5, color=LABEL_COLOR_MAP1[i])\n",
    "    ax.set_ylabel('Number of questions')\n",
    "    ax.set_xlabel('Score')\n",
    "    title = 'Number of Questions in each Score for '+assign_topics1[i]\n",
    "    ax.set_title(title)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#want to find which topic has highest popularity and which question in each\n",
    "#topic is the most popular\n",
    "\n",
    "#aggregate sum of all scores and save to list called scores\n",
    "scores_py = np.zeros(10)\n",
    "for i in range(10):\n",
    "    scores_py[i] = qpy_dict[i]['Score'].sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 251,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a34db97b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "color1 = [LABEL_COLOR_MAP1[l] for l in range(10)]\n",
    "#total ratings\n",
    "fig, ax = plt.subplots()\n",
    "ind = np.arange(len(scores_py))\n",
    "ax.barh(ind, scores_py, alpha=0.5, color=color1)\n",
    "ax.set_yticks(ind)\n",
    "ax.set_yticklabels(assign_topics1)\n",
    "ax.set_xlabel('Sum of Total Scores')\n",
    "ax.set_title('Summation of Scores in Each Topic')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 252,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#number of questions in each dataframe\n",
    "num_qpy = np.zeros(10)\n",
    "for i in range(10):\n",
    "    num_qpy[i] = len(qpy_dict[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 253,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a30d70898>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#total ratings\n",
    "fig, ax = plt.subplots()\n",
    "ind = np.arange(len(num_qpy))\n",
    "ax.barh(ind, num_qpy, alpha=0.5, color=color1)\n",
    "ax.set_yticks(ind)\n",
    "ax.set_yticklabels(assign_topics1)\n",
    "ax.set_xlabel('Total Number of Questions')\n",
    "ax.set_title('Number of Questions in Each Topic')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#normalize data based on its total number of questions\n",
    "norm_scores_py = np.zeros(10)\n",
    "for i in range(10):\n",
    "    norm_scores_py[i] = scores_py[i]/num_qpy[i]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 255,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a34cda908>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#normalized ratings ratings\n",
    "fig, ax = plt.subplots()\n",
    "ind = np.arange(len(norm_scores_py))\n",
    "ax.barh(ind, norm_scores_py, alpha=0.5, color=color1)\n",
    "ax.set_yticks(ind)\n",
    "ax.set_yticklabels(assign_topics1)\n",
    "ax.set_xlabel('Normalized Score')\n",
    "ax.set_title('Normalized Score Based on Question Population')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "Based on our analysis, we see that basic python concepts are the most common type of question asked on StackOverflow. Following this are questions regarding lists and questions formed by programmers with a minimal coding background. This is closely aligned with our initial hypothesis that many questions on StackOverflow would be formed by beginners, or people with little python experience. Although, it should be noted that the labelling of \"basic\" versus \"advanced\" python concepts may be somewhat arbitrary, as it was completely based on the subjectivity of one person (Nathaniel). Nonetheless, it's safe to say that these questions are those regarding python concepts in some form.\n",
    "\n",
    "As an example, some of the top 10 questions pulled from this basic python concepts category are:\n",
    "- What are metaclasses? What do you use them for?\n",
    "- What's the difference between the list methods append() and extend()?\n",
    "- If I create a global variable in one function, how can I use that variable in another function?\n",
    "\n",
    "These are all short, one-line questions that are specifically targeted toward a certain concept. These are crucial for beginners to understand if they want to truly become fluent in the language.\n",
    "\n",
    "It should also be noted that the topic of dictionary and strings also have a high count of questions, but their questions are not scored highly. Our conjecture is that these questions are often very specific to special cases that each programmer might be facing, and as a result, may not be questions that a lot of other people have, since they would also need to be in this specific case. One example of such a specific question is a question from a user who, coming from a C++ background, was trying to translate one of his complicated functions for reading from stdin to python. This is a very niche question, and yet was included in the list of top 10 questions- it's likely that other questions in this category are similarly specific."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 2 - 10% of Questions on StackOverflow\n",
    "Since we are synthesizing the two different StackOverflow datasets, we would also need to be careful to ensure that our primary focus is still on Python-related questions. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "64601\n"
     ]
    }
   ],
   "source": [
    "#following similar steps for the python data:\n",
    "df_stack = pd.read_csv('stacksample/Questions.csv', encoding= 'ISO-8859-1')  \n",
    "df_tag = pd.read_csv('stacksample/Tags.csv', encoding= 'ISO-8859-1')\n",
    "\n",
    "#separate only questions that contain the Python tag, get list of question ids\n",
    "df_tag = df_tag[df_tag[\"Tag\"] == \"python\"]\n",
    "ids = df_tag.iloc[:,0]\n",
    "\n",
    "#sort out all questions with only id in the ids list\n",
    "df_stack = df_stack[df_stack['Id'].isin(ids)]\n",
    "\n",
    "#check len of remaining dataframe == 64601\n",
    "print(len(df_stack))\n",
    "\n",
    "#drop unused columns\n",
    "df_stack = df_stack.drop(['Id', 'OwnerUserId', 'CreationDate'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "113    This is a difficult and open-ended question I ...\n",
      "184    I'm creating an ZIP file with ZipFile in Pytho...\n",
      "204    I have a bunch of files (TV episodes, although...\n",
      "331    I've been trying to wrap my head around how th...\n",
      "370    A reliable coder friend told me that Python's ...\n",
      "Name: pBody, dtype: object\n"
     ]
    }
   ],
   "source": [
    "#convert question from HTML format to String\n",
    "df_stack[\"pBody\"] = pd.Series(df_stack[\"Body\"].apply(convert_body), index = df_stack.index)\n",
    "print(df_stack[\"pBody\"].head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Create a second color map specific for 10% StackOverflow data\n",
    "LABEL_COLOR_MAP2 = {0 : '#461969ff',\n",
    "                    1 : '#45337dff',\n",
    "                    2 : '#3c4f8aff',\n",
    "                    3 : '#33638dff',\n",
    "                    4 : '#29788eff',\n",
    "                    5 : '#25a186ff',\n",
    "                    6 : '#35b479ff',\n",
    "                    7 : '#5cc863ff',\n",
    "                    8 : '#c3df2eff',\n",
    "                    9 : '#fde725ff'\n",
    "                   }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Use TF-IDF to transfrom questions\n",
    "tfidf = TfidfVectorizer(sublinear_tf=True, analyzer='word', \n",
    "                        max_features=500, tokenizer=word_tokenize, max_df=0.8)\n",
    "\n",
    "tf_pBody = tfidf.fit_transform(df_stack[\"pBody\"].values.astype('U'))\n",
    "tf_pBody = tf_pBody.toarray() #transformed vectors into a matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#plot on 2D\n",
    "pca = PCA(n_components=2).fit(tf_pBody)\n",
    "data2D = pca.transform(tf_pBody)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Use kmeans to organize and cluster each question\n",
    "kmeans = MiniBatchKMeans(n_clusters=10).fit(tf_pBody) #default threshold\n",
    "centers2D = pca.transform(kmeans.cluster_centers_)\n",
    "clusters = kmeans.cluster_centers_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11707a898>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "label_color_st = [LABEL_COLOR_MAP2[l] for l in kmeans.labels_]\n",
    "plt.scatter(data2D[:, 0], data2D[:, 1], c=label_color_st)\n",
    "plt.scatter(centers2D[:,0], centers2D[:,1], \n",
    "            marker='x', s=200, linewidths=3, c='r')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since both datasets are from the same originating source (StackOverflow), we expect to find a similar distribution as the Python dataset. \n",
    "\n",
    "We will also use the same procedure for finding the most common words to determine the topics of each cluster.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df_stack[\"label\"] = pd.Series(kmeans.labels_, index = df_stack.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "top30st_dict = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[(\"''\", 105698), ('``', 39664), ('--', 6731), ('import', 5405), ('return', 4825), ('python', 4461), ('url', 4343), ('code', 4310), ('class=', 4280), ('file', 4274), ('data', 4036), ('get', 3912), ('print', 3867), ('def', 3595), ('like', 3494), ('using', 3479), ('http', 3362), (\"'m\", 3356), ('...', 3279), ('request', 3254), ('name', 3190), ('error', 2963), (\"n't\", 2938), ('form', 2849), ('class', 2753), ('html', 2602), ('function', 2550), ('name=', 2515), ('div', 2494), (\"'s\", 2479)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[('python', 14272), (\"''\", 10595), (\"n't\", 8206), ('--', 8141), ('``', 8050), ('file', 7973), (\"'m\", 7878), ('using', 7231), ('error', 5948), (\"'s\", 5605), ('like', 5329), ('code', 5317), ('use', 5312), ('import', 5293), ('get', 4828), ('would', 4767), ('run', 4480), ('script', 4290), ('want', 3796), (\"'ve\", 3364), ('...', 3304), ('way', 3286), ('need', 3164), ('trying', 3125), ('install', 3084), ('data', 3039), ('one', 3028), ('running', 3019), ('problem', 2992), ('files', 2982)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[('python', 6168), ('using', 3163), ('way', 2740), ('like', 2347), ('want', 2084), ('use', 2070), ('file', 2034), ('would', 1788), (\"'m\", 1584), ('need', 1339), ('data', 1232), ('know', 1198), ('code', 1197), ('--', 1166), (\"n't\", 1154), ('one', 1146), (\"''\", 1134), (\"'s\", 1123), ('get', 1121), ('thanks', 1105), ('function', 954), ('possible', 902), ('image', 888), ('``', 858), ('files', 831), ('example', 808), ('script', 786), ('find', 779), ('django', 742), ('...', 702)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[(\"''\", 10095), ('``', 5568), ('import', 3446), ('def', 2990), ('--', 2864), ('print', 2603), ('return', 2474), ('code', 2273), ('range', 1994), ('data', 1934), ('function', 1908), ('==', 1683), ('like', 1532), ('list', 1490), ('python', 1473), ('using', 1427), (\"'m\", 1400), (\"n't\", 1358), ('array', 1301), ('want', 1222), ('would', 1192), ('line', 1160), ('numpy', 1151), ('plot', 1145), ('len', 1130), ('get', 1119), ('self', 1115), (\"'s\", 1109), ('values', 1101), ('...', 1079)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[(\"''\", 32043), ('self', 21879), ('``', 20789), ('def', 19645), ('class', 14636), ('return', 8979), ('--', 8796), ('import', 8672), ('print', 5693), ('__init__', 4707), ('code', 4248), ('object', 4087), ('name', 3351), (\"n't\", 3233), ('==', 3093), ('file', 3089), (\"'m\", 3076), ('get', 2909), ('none', 2898), ('user', 2786), ('like', 2773), ('...', 2772), ('data', 2668), ('python', 2637), (\"'s\", 2625), ('true', 2386), ('method', 2330), ('using', 2269), ('error', 2236), ('value', 2151)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[(\"''\", 40965), ('``', 34258), ('file', 26604), ('line', 24322), ('--', 19187), ('import', 8770), ('error', 7398), ('module', 7362), ('python', 5448), ('call', 5159), ('traceback', 5041), ('last', 4837), ('recent', 4647), ('return', 4012), ('...', 3517), ('print', 3171), ('code', 3153), ('get', 2960), ('using', 2608), (\"'m\", 2418), ('run', 2368), (\"n't\", 2305), ('def', 2198), ('self', 2098), ('trying', 2051), ('following', 2006), ('name', 1950), ('data', 1631), (\"'s\", 1542), ('use', 1540)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[(\"''\", 12256), ('--', 7036), ('``', 6688), ('list', 3456), ('print', 3425), ('...', 2784), ('like', 2727), ('nan', 2533), ('data', 2513), ('10', 2414), ('array', 2288), ('==', 2087), ('want', 2013), ('range', 1977), ('return', 1965), ('import', 1933), ('code', 1900), ('def', 1801), ('row', 1781), ('python', 1726), ('would', 1719), ('values', 1615), (\"'m\", 1610), ('get', 1590), ('value', 1589), ('column', 1485), (\"n't\", 1449), ('len', 1386), ('way', 1381), ('example', 1341)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[(\"''\", 27188), ('``', 15943), ('import', 6618), ('print', 6405), ('--', 6008), ('list', 5954), ('file', 5875), ('code', 4995), ('data', 4961), ('like', 3827), ('python', 3629), ('line', 3513), ('get', 3340), (\"'m\", 3298), ('want', 3051), ('using', 2980), ('...', 2948), ('def', 2762), ('open', 2755), (\"n't\", 2747), ('row', 2735), ('return', 2538), ('error', 2532), (\"'s\", 2290), ('trying', 2243), ('would', 2187), ('value', 2120), ('output', 2091), ('name', 2087), (\"'a\", 2052)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[(\"''\", 16421), ('``', 10239), ('--', 8298), ('like', 6799), ('would', 5208), ('want', 5034), ('list', 4575), ('data', 4366), (\"n't\", 4332), ('python', 4301), ('...', 4235), (\"'m\", 4202), ('using', 4078), ('way', 4007), ('code', 3784), ('string', 3677), ('one', 3673), ('function', 3499), (\"'s\", 3393), ('get', 3331), ('use', 3188), ('file', 3187), ('value', 3115), ('example', 2848), ('need', 2839), ('values', 2561), ('return', 2426), ('def', 2353), ('class', 2339), ('print', 2289)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[(\"''\", 55284), ('``', 48293), ('print', 20488), ('--', 12466), ('==', 9234), ('import', 8181), ('def', 7693), ('code', 6420), ('return', 5095), ('file', 5014), ('input', 4825), ('python', 4696), ('else', 4120), ('elif', 3747), ('line', 3696), ('true', 3389), (\"'m\", 3375), (\"n't\", 3367), ('str', 3323), ('number', 3306), (\"'s\", 2886), ('error', 2755), ('get', 2745), ('data', 2710), ('int', 2663), ('function', 2549), ('time', 2487), ('program', 2468), ('name', 2455), ('enter', 2402)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for i in range(10):\n",
    "    word_tokens = word_tokenize(getStrings(df_stack[df_stack[\"label\"] == i][\"pBody\"]))\n",
    "    filtered_sentence = [w for w in word_tokens if not w in stop_words]\n",
    "    filtered_punct = [w for w in filtered_sentence if len(w) > 1]\n",
    "    #create a Counter object to count words\n",
    "    cnt = Counter(filtered_punct)\n",
    "    #save top 30 words from each label\n",
    "    top30st_dict.append(cnt.most_common(30))\n",
    "    print(top30st_dict[i])\n",
    "    print()\n",
    "    print(\"-------------------------------------------\")\n",
    "    print()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "qstack_dict = [df_stack[df_stack['label'] == i] for i in range(10)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 256,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Topic map for 10% StackOverflow Dataset\n",
    "LABEL_TOPIC_MAP2 = {0 : 'web development',                                # if 'url' maps to 4343\n",
    "                    1 : 'learning python from another language',          # if 'python' maps to 14272\n",
    "                    2 : 'basic python concepts',                          # if 'python' maps to 6168\n",
    "                    3 : 'functions and packages',                         # if 'def' maps to 2990\n",
    "                    4 : 'classes',                                        # if 'class' maps to 14636\n",
    "                    5 : 'debugging errors',                               # if 'error' maps to 7398\n",
    "                    6 : 'lists and dictionaries',                         # if 'list' maps to 3456\n",
    "                    7 : 'file-reading and strings',                       # if 'file' maps to 5875\n",
    "                    8 : 'data output',                                    # if 'data' maps to 4366\n",
    "                    9 : 'scripting'                                       # if 'print' maps to 20488\n",
    "                   }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 257,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['web development', 'learning python from another language', 'basic python concepts', 'functions and packages', 'classes', 'debugging errors', 'lists and dictionaries', 'file-reading and strings', 'data output', 'scripting']\n"
     ]
    }
   ],
   "source": [
    "assign_topics2 = ['none']*10\n",
    "for i in range(10):\n",
    "    for j in range(10):\n",
    "        should_break = False;\n",
    "        for k,v in top30st_dict[j]:\n",
    "            if k == 'url' and v == 4343 and assign_topics2[j] == 'none':\n",
    "                assign_topics2[j] = LABEL_TOPIC_MAP2[0]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'python' and v == 14272 and assign_topics2[j] == 'none':\n",
    "                assign_topics2[j] = LABEL_TOPIC_MAP2[1]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'python' and v == 6168 and assign_topics2[j] == 'none':\n",
    "                assign_topics2[j] = LABEL_TOPIC_MAP2[2]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'def' and v == 2990 and assign_topics2[j] == 'none':\n",
    "                assign_topics2[j] = LABEL_TOPIC_MAP2[3]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'class' and v == 14636 and assign_topics2[j] == 'none':\n",
    "                assign_topics2[j] = LABEL_TOPIC_MAP2[4]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'error' and v == 7398 and assign_topics2[j] == 'none':\n",
    "                assign_topics2[j] = LABEL_TOPIC_MAP2[5]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'list' and v == 3456 and assign_topics2[j] == 'none':\n",
    "                assign_topics2[j] = LABEL_TOPIC_MAP2[6]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'file' and v == 5875 and assign_topics2[j] == 'none':\n",
    "                assign_topics2[j] = LABEL_TOPIC_MAP2[7]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'data' and v == 4366 and assign_topics2[j] == 'none':\n",
    "                assign_topics2[j] = LABEL_TOPIC_MAP2[8]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'print' and v == 20488 and assign_topics2[j] == 'none':\n",
    "                assign_topics2[j] = LABEL_TOPIC_MAP2[9]\n",
    "                should_break = True\n",
    "                break\n",
    "        if should_break:\n",
    "            break\n",
    "print(assign_topics2)\n",
    "           "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, we will plot similar graphs for the 10% StackOverflow dataset like we did for the Python dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 258,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a34d06550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeoAAAEWCAYAAACpPdRYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzt3Xm4HFWd//H3hxAWJRACFwayEMSggEvECCioiBgWl+CCggoB0aDCiCPOCI4jIGaQ+amgM8jigAQUYkTRyKAQlsigsiQSgRCRsCYkQwKBsGk08P39cU6TSqe7b98lt+re+3k9Tz+36tR2TvXp+tapOrdKEYGZmZlV0wZlZ8DMzMyac6A2MzOrMAdqMzOzCnOgNjMzqzAHajMzswpzoDYzM6uwygVqSRdL+npJ25akH0h6UtJtZeShKyTNl7RvH27vY5Ku7avtdUbSWEkhacOy89ITkt4vaZGkZyW9oRfWN1vSJ3sjb93Ydp/Wye6S9JCk/ftgO/3qmNITkvaVtHg9rv8oSTevr/VXWaeBOlfoxyS9vJD2SUmz12vOyrEP8C5gVETs0WgGSaMk/UjSE5Kek3SbpIPXd8YancBExG4RMXt9b7uwvR9FxMS+2l5vy9/dTyU9LmmlpLskHVV2voBvAsdHxGYRcUfZmemJvq6T7Sjz5J82jin9VT5JfmXZ+RgM2m1RbwicsD4zsj5IGtLFRXYAHoqI55qsbwRwM/A3YDdga+AsYLqkQ3qSV+sTlwKLSN/zVsCRwGO9uYFutu53AOZ3c3tdrePd1t+vXJSks2OK92kD3i91IqLlB3gIOAlYAQzPaZ8EZufhsUAAGxaWmQ18Mg8fBfyWFNCeAh4A3pLTFwHLgMmFZS8GzgNmAc8AvwF2KEx/dZ62ArgX+HDdsucCVwPPAfs3KM/2wMy8/ELgUzn9GOCvwAvAs8BpDZY9Hbgb2KAu/Uu5XOpsf+TxTwALgCeBa2rly8uflffJSuBO4DXAFODvpBOEZ4FfFr6b/fPwxsDZwJL8ORvYOE/bF1gMnJjXvRQ4upCfg4F78v5+FPhik7pwFHBzYTyATwP35bKcA6jJshuQ6tH9wBPADGBEYfpPgP/L5b4J2K0wbVPgW8DDefrNOa22rycDjwCPA//aoi4/C4xvMX0f4HekeroIOCqnbwFcAizPefhKrQ6wdv1eAXy91Xdct72Nc56CVF/vz+m75DrzFCmAv6+LdXw2bdS3PO07uaxPA3OBtxamnQpcAfwwT/9kTpuR98czOX8T6o4X+xeWbzXv7sAdedpPgB/X9l+Tuvdb4D9zHfgT8M487VBgbt38JwI/p/Vv54uk39jKvO1NCst/inR8WEE6Xmzf1XpPg2MKa36LXyLV90vb3N5n8/aeIR2HdgJ+n7+XGcBGTfbbTsANpN/c48CPyMfxnuwH0m+0Vm+fBT5C58eZjUlXjx4hnSCfB2xad4xaa790cvzprO52q+7Vb6fwHbwyD787L/t03v6pdfMeSTpOPAH8G2v/JloeB5semzqdIW8E+FmhIF0N1KuBo4EhwNfzF3VO/uIm5p21WeFA9Azwtjz9O7WdBrw875ijSa383UmVb7fCsiuBvfMO2aRBeX4DfA/YBBhPOvi+s5DXm1vsi1toHMB3zPtgXBv74xBSxd8ll+ErwO/ytANIFW44KWjvAmxXKNvXG303efhrOX/bAB2kgHN64UewOs8zlBSYnwe2zNOXkis5sCWwe4uDZX2gvirnd0zelwc2WfbzOX+j8vd6PnB5XTAZxpoTjnmFaefkfTiSVIfekuer7evvkwL364FVwC5N8nAd6WB/GDCmbtoYUr07PO+jrchBnfRj/0XO31jgz8AxdfX7H/P3uWmr77hJvooHgaF52S8DGwH75Xy9qgt1fDZt1Lc8/eO5rBuSDrD/V1sn6WD397yODXLZTiUFn4Pzd3EGcEuTOtl03ly2h0lX6oYCHyAF01aBejXwT3n+j+T9MCLXhRXF7510IP1gJ7+d20gn7iNIJzKfztP2Ix1Xds/r/k/gpm7W+6NY+zezby7HmXndm7a5vZnA5qQreauA64FXkE4i76HQ2Knb/itJl943Jh0XbgLO7sX98MoGZWt2nDk7l2ME6bf0S+CMZvuljX3ZWd3tVt2r306D3+i+wGtJv4nXkU46DsnTdiWduOyTt/NN0m+o9ptoeRxseozodIY1gfo1pB9GB10P1PcVpr02z79tIe0J1hwULwamF6ZtRjojHU36cf5vXf7OB04pLHtJi7KMzusaVkg7A7i42RdUt/xCciWuS98kl+ktbeyPX5EP8nl8A1Jl3oH0w/gzsBfrttovpnWgvh84uDDtANIlt1rF+ktdnpYBe+XhR4Bjgc07qQtr7Z9czn0K4zOAk5osu4B8QpTHtyNV4A0bzDs8r3uLvH/+Ary+wXy1fT2qkHYbcFiTPGwJfIN0dv0CMA94U552MnBlg2WGkA6MuxbSjmVN/T8KeKRumabfcZN8FQ8CbyUdcDYoTL+cfNZOJ3W8K/WtybJP1vY16WB3U930U4HrCuO7An9pUiebzks6EX+UQkuUdKWkVaBeUjf/bcARefhcYGoe3i2Xo3ZF6eL69eZ8frww/h/AeXn4QuA/CtM2I9XVsd2o90exbqD+G2u3WtvZ3t6F6XOBLxXGv0Uh+HZSNw4B7ujF/VAfqBseZ0gNj+eAnQrT3gw82Gy/dLYv26i73ap7jbZTX9a6aWcDZ+Xhr7J2A+RluVy130Tbx8Hip+1e3xFxN+ks8qR2lyko3gf8S15ffdpmhfFFhe0+Szpb3p4UzPaU9FTtA3wM+IdGyzawPbAiIp4ppD1Maqm143HSjq1XS1vexjp2AL5TyP8KUiUeGRE3AP9FakE+JukCSZu3mbftSWWpeTin1TwREasL48+zZp9/kHTm+bCk30h6c5vbhBRUGq2z3g7AlYVyLyAFy20lDZH0DUn3S3qadPCA1Adga9KJ0P09zUNEPBkRJ0XEbsC2pED9c0kincQ12sbWrDkDr6mvM/V1rul33KIMNdsDiyLixS5sr5WWeZF0oqQFuXPdU6STo6072Vb9/t6kxT3FZvNuDzwa+WjVZrnq5y/W8WnAR/N3eQQwIyJWdbK+ZvVmrd9SPgY9wdrfQbv1vpHlEfHXwng726s/XrY6fr5E0jaSpkt6NP+2fsja3y/0bD/Ua3ac6SAFrbmFuvjrnF5Tv19aaqPu9mbdK253T0k3SlouaSXpNkhtu9uzdvx6nrTPapoeB1tts6v/nnUK6Z5F8YuqdZJ4WSGtGDi7Y3RtQNJmpEslS0g74DcRMbzw2SwiPlNYNmhuCTBC0rBC2hjS2VU7rgM+KKl+v32YdH/lfjrfH4uAY+vKsGlE/A4gIr4bEW8ktQp2Bv65jXLVyrZDXbmWtFOoiLg9IiaRLpv/nNRC6G2LgIPqyr1JRDwKfBSYRLpyswWppQwpoDxOuoS1U29mJiIeJ12Wql3yW9RkG4+Tznjr922xztR/Ny2/404sAUbX1bHOttdK07xIeivpnuCHSZcnh5Oumqmb2+qKpcDIHFhrRjebOauf/6U6HhG3kFoubyXVp0sL83W1DGv9lvJ/vGxF+8eJztTnZ31u74y8vddFxOaky8Vqvch6ydfjpBOK3Qr1cIuIKJ5gtP09tVl3m+ms7j1H4fgtqT6eXUa6hD86IrYg3WuvrWsp6bJ2bdlNSfusptVxsKkuBeqIWEi66f65Qtpy0hf38dwy+gQ9P6geLGkfSRuROk7cGhGLSC36nSUdIWlo/rxJ0i5t5n8R6d7tGZI2kfQ6UoePH7WZr7NI94kulPQPeR2HkzoMnBIRL7axP84DTpa0G4CkLSQdmofflM/WhpIqS60jCqQz6Fe0yNvlwFckdUjamnQJ5oedFUjSRkr/H71FRPyd1EHihc6W64bzgKmSdsjb7ZA0KU8bRrq8/ATpB/LvtYVyy/Ii4NuSts/79M2SNu5qBiSdKek1kjbMJ2ufARZGxBOkOrC/pA/n6VtJGh8RL5BOXKZKGpbz/wVa79um33EbbiV99/+S6/e+wHuB6V0tbxt5GUa6L7gc2FDSV0n1uy/8nlTPjs/7exLQ2b8vbQN8Lu+XQ0n33a8uTL+EdEVqdUQU/9+2s99OvcuAoyWNz/Xs30nHoIe6sI6uWJ/bG0a6Z/qUpJGsOfHvjXy1vV/z7/j7wFmStgGQNFLSAV3IT1FP6m5nde+PwG653JuQLqPXb3tFRPxV0h6kE8OaK4D3SnpLjl+nsfbJQ6vjYFPdeeDJ10iduoo+RaoAT5Bagu20HFq5jNR6XwG8kXR5m3zJeiKpM9AS0qWNWueDdh1OarEtAa4kBdhZ7SyYD+j7kC7F3kP6AVwCHBcRFxVmbbo/IuLKnOfp+VLU3cBBefLmpMr8JGt6DX4zT7sQ2DVfMvl5g+x9HZhD6r15F/CHnNaOI4CHcn4+TTrr7m3fIZ2FXivpGVKHij3ztEtI5X2UtF9vqVv2i6Qy3U6qE2fSvbr7MtJ3Xvvvgx2A9wFExCOky/8n5m3MI3VOg9RR7Lm8zM2k+ln8vtfSyXfcUkT8LefpIFIr5HvAkRHxpy6Us928XEO6h/1n0v7/K127rN5tuZwfIJ0oP0Wqc1eRTtiauZXUYfNxYCrwofybrLmU1Jfm0rrlOvvt1OftetLJ909JLaSdSMec9WI9b+80UmewlcD/kDoF91a+TgWm5f364TZW+SVSP59bcl28DnhVu/mp0+2621ndi4g/k+LcdaSe9vUPWfks8LV8HPsqhSuQETGfdLyYTtpnz5Du09fqdavjYFNa+zK9dYXS/ePfkjohfbXs/Jj1Z5JuJXVk+kGDaUeROsjt02L5TUkHxd0j4r71llEbcFrVvR6udzPSycC4iHiwu+up3CNE+5OIeJrUCnuhwX0MM2tB0tvzLaQNJU0m/avLr3uwys8AtztIW2fWQ90rrvu9kl6W7+l/k3Q18KGerNNPf+mhfN/7tLLzYdYPvYp02XAzUkfMD0XE0u6sSNJDpHuBfkKgtaPX6l4Dk0i3X0S6HXlY9PDStS99m5mZVZgvfZuZmVWYL31309Zbbx1jx44tOxtmZv3G3LlzH4+Ijs7ntCIH6m4aO3Ysc+bMKTsbZmb9hqSHO5/L6vnSt5mZWYU5UJuZmVWYA7WZmVmFOVCbmZlVmAO1mZlZhTlQm5mZVZgDtZmZWYU5UJuZmVWYA7WZmVmF+clkJbjk/BteGj7y2P1KzImZmVWdW9RmZmYV5kBtZmZWYQ7UZmZmFeZAbWZmVmEO1GZmZhXmQG1mZlZhDtRmZmYV5kBtZmZWYf0+UEsaIukOSVfl8R0l3SrpPkk/lrRRTt84jy/M08cW1nFyTr9X0gHllMTMzGxd/T5QAycACwrjZwJnRcQ44EngmJx+DPBkRLwSOCvPh6RdgcOA3YADge9JGtJHeTczM2upXwdqSaOAdwP/nccF7AdckWeZBhyShyflcfL0d+b5JwHTI2JVRDwILAT26JsSmJmZtdavAzVwNvAvwIt5fCvgqYhYnccXAyPz8EhgEUCevjLP/1J6g2XWImmKpDmS5ixfvrw3y2FmZtZQvw3Ukt4DLIuIucXkBrNGJ9NaLbN2YsQFETEhIiZ0dHR0Kb9mZmbd0Z/fnrU38D5JBwObAJuTWtjDJW2YW82jgCV5/sXAaGCxpA2BLYAVhfSa4jJmZmal6rct6og4OSJGRcRYUmewGyLiY8CNwIfybJOBX+ThmXmcPP2GiIicfljuFb4jMA64rY+KYWZm1lJ/blE38yVguqSvA3cAF+b0C4FLJS0ktaQPA4iI+ZJmAPcAq4HjIuKFvs+2mZnZugZEoI6I2cDsPPwADXptR8RfgUObLD8VmLr+cmhmZtY9/fbSt5mZ2WDgQG1mZlZhDtRmZmYV5kBtZmZWYQ7UZmZmFeZAbWZmVmEO1GZmZhXmQG1mZlZhDtRmZmYV5kBtZmZWYQ7UZmZmFeZAbWZmVmEO1GZmZhXmQG1mZlZhDtRmZmYV1m8DtaRNJN0m6Y+S5ks6LadfLOlBSfPyZ3xOl6TvSloo6U5JuxfWNVnSffkzuawymZmZ1duw7Az0wCpgv4h4VtJQ4GZJv8rT/jkirqib/yBgXP7sCZwL7ClpBHAKMAEIYK6kmRHxZJ+UwszMrIV+26KO5Nk8OjR/osUik4BL8nK3AMMlbQccAMyKiBU5OM8CDlyfeTczM2tXvw3UAJKGSJoHLCMF21vzpKn58vZZkjbOaSOBRYXFF+e0ZumNtjdF0hxJc5YvX96rZTEzM2ukXwfqiHghIsYDo4A9JL0GOBl4NfAmYATwpTy7Gq2iRXqj7V0QERMiYkJHR0eP829mZtaZfh2oayLiKWA2cGBELM2Xt1cBPwD2yLMtBkYXFhsFLGmRbmZmVrp+G6gldUganoc3BfYH/pTvOyNJwCHA3XmRmcCRuff3XsDKiFgKXANMlLSlpC2BiTnNzMysdP251/d2wDRJQ0gnHDMi4ipJN0jqIF3Sngd8Os9/NXAwsBB4HjgaICJWSDoduD3P97WIWNGH5TAzM2uq3wbqiLgTeEOD9P2azB/AcU2mXQRc1KsZNDMz6wX99tK3mZnZYOBAbWZmVmEO1GZmZhXmQG1mZlZhDtRmZmYV5kBtZmZWYaUHakknSNo8P4jkQkl/kDSx7HyZmZlVQemBGvhERDxNeiJYB+lBJN8oN0tmZmbVUIVAXXspxsHADyLijzR+UYaZmdmgU4VAPVfStaRAfY2kYcCLJefJzMysEqrwCNFjgPHAAxHxvKStyM/hNjMzG+xKD9QR8aKkx4BdJZWeHzMzsyopPTBKOhP4CHAP8EJODuCm0jJlZmZWEaUHatI7o18VEavKzoiZmVnVVKEz2QPA0LIzYWZmVkVVCNTPA/MknS/pu7VPZwtJ2kTSbZL+KGm+pNNy+o6SbpV0n6QfS9oop2+cxxfm6WML6zo5p98r6YD1VE4zM7Muq8Kl75n501WrgP0i4llJQ4GbJf0K+AJwVkRMl3QeqVf5ufnvkxHxSkmHAWcCH5G0K3AYsBuwPXCdpJ0j4oVGGzUzM+tLpbeoI2IacDkwN38uy2mdLRcR8WweHZo/AewHXJHTp5HugQNMyuPk6e+UpJw+PSJWRcSDwEJgjx4XzMzMrBeUHqgl7QvcB5wDfA/4s6S3tbnsEEnzgGXALOB+4KmIWJ1nWQyMzMMjgUUAefpKYKtieoNlzMzMSlWFS9/fAiZGxL0AknYmtbDf2NmC+fL0eEnDgSuBXRrNlv82eixptEhfh6QpwBSAMWPGdJY9MzOzHiu9RQ0MrQVpgIj4M13sBR4RTwGzgb2A4YUHp4wCluThxcBogDx9C2BFMb3BMvXbuSAiJkTEhI6Ojq5k0czMrFuqEKjn5Ndb7ps/3yfdq25JUkduSSNpU2B/YAFwI/ChPNtk4Bd5eGYeJ0+/ISIipx+We4XvCIwDbuulspmZmfVIFS59fwY4Dvgc6TL0TaR71Z3ZDpgmaQjphGNGRFwl6R5guqSvA3cAF+b5LwQulbSQ1JI+DCAi5kuaQXoy2mrgOPf4NjOzqig9UOcnkn07f7qy3J3AGxqkP0CDXtsR8Vfg0CbrmgpM7cr2zczM+kJpgVrSjIj4sKS7aNB5KyJeV0K2zMzMKqXMFvUJ+e97SsyDmZlZpZXWmSwilubBz0bEw8UP8Nmy8mVmZlYlVej1/a4GaQf1eS7MzMwqqMx71J8htZx3knRnYdIw4Lfl5MrMzKxayrxHfRnwK+AM4KRC+jMRsaKcLJmZmVVLmfeoV0bEQ8BXgP/L96Z3BD5ee5CJmZnZYFeFe9Q/BV6Q9ErSQ0l2JLW2zczMBr0qBOoX89usPgCcHRH/RHrqmJmZ2aBXhUD9d0mHA0cCV+W0Lr2Uw8zMbKCqQqA+GngzMDUiHswvxvhhyXkyMzOrhCo86/seSV8CxuTxB4FvlJsrMzOzaii9RS3pvcA84Nd5fLykmeXmyszMrBpKD9TAqaS3XT0FEBHzSD2/zczMBr0qBOrVEbGyLm2dt2mZmZkNRqXfowbulvRRYIikccDngN+VnCczM7NKqEKL+h+B3YBVwOXA08DnO1tI0mhJN0paIGm+pBNy+qmSHpU0L38OLixzsqSFku6VdEAh/cCctlDSSY22Z2ZmVobSW9QR8Tzwr/nTFauBEyPiD5KGAXMlzcrTzoqIbxZnlrQrcBjppGB74DpJO+fJ55De4rUYuF3SzIi4p3slMjMz6z2lB2pJN9LgnnRE7Ndqufw+66V5+BlJC4CRLRaZBEyPiFXAg5IWkjqxASyMiAdyfqbneR2ozcysdKUHauCLheFNgA+SWsttkzQWeANwK7A3cLykI4E5pFb3k6QgfkthscWsCeyL6tL3bLKdKcAUgDFjxnQli2ZmZt1S+j3qiJhb+Pw2Ir5Ak0DZiKTNSC/2+HxEPA2cC+wEjCe1uL9Vm7XR5lukN8rrBRExISImdHR0tJtFMzOzbiu9RS1pRGF0A+CNwD+0uexQUpD+UUT8DCAiHitM/z5rnh++GBhdWHwUsCQPN0s3MzMrVemBGpjLmpbtauBB4JjOFpIk0msxF0TEtwvp2+X71wDvB+7OwzOByyR9m9SZbBxwW97uuPyM8UdJHc4+2gvlMjMz67HSA3VEdPcpZHsDRwB3SZqX074MHC5pPCn4PwQcm7czX9IMUiex1cBxEfECgKTjgWuAIcBFETG/m3nqskvOv+Gl4SOPbdl/zszMBqHSA7WkD7SaXruk3SD9ZhrfX766xbqmAlMbpF/dajkzM7OylB6oSZe53wLUmpbvAGYDK0mt4oaB2szMbDCoQqAOYNfafWVJ2wHnRMTR5WbLzMysfKX/exYwttD5C+AxYOdmM5uZmQ0mVWhRz5Z0Dek530HqdX1juVkyMzOrhtIDdUQcL+n9wNty0gURcWWZeTIzM6uK0gM1QA7MDs5mZmZ1qnCP2szMzJpwoDYzM6uw0gK1pOvz3zPLyoOZmVnVlXmPejtJbwfel98BvdZTxiLiD+Vky8zMrDrKDNRfBU4iva3q23XTAvCDr83MbNArLVBHxBXAFZL+LSJOLysfZmZmVVb6v2dFxOmS3sea/6OeHRFXtVrGzMxssCi917ekM4ATSK+fvAc4IaeZmZkNeqW3qIF3A+Mj4kUASdOAO4CTS82VmZlZBZTeos6GF4a3aGcBSaMl3ShpgaT5kk7I6SMkzZJ0X/67ZU6XpO9KWijpTkm7F9Y1Oc9/n6TJvVoyMzOzHqhCoD4DuEPSxbk1PRf49zaWWw2cGBG7AHsBx0naldST/PqIGAdcn8cBDgLG5c8U4FxIgR04BdgT2AM4pRbczczMylZ6oI6Iy0mB9mf58+aImN7Gcktr/2sdEc8AC4CRwCRgWp5tGnBIHp4EXBLJLcDw/O7rA4BZEbEiIp4EZgEH9loBzczMeqAK96jJ76Oe2d3lJY0F3gDcCmxbe791RCyVtE2ebSSwqLDY4pzWLL3RdqaQWuOMGTOmu9k1MzNrW+kt6p6StBnwU+DzEfF0q1kbpEWL9HUTIy6IiAkRMaGjo6PrmTUzM+uifh2oJQ0lBekfRcTPcvJj+ZI2+e+ynL4YGF1YfBSwpEW6mZlZ6UoN1JI2kHR3N5cVcCGwICKKjyCdCdR6bk8GflFIPzL3/t4LWJkvkV8DTJS0Ze5ENjGnmZmZla7Ue9QR8aKkP0oaExGPdHHxvYEjgLskzctpXwa+AcyQdAzwCHBonnY1cDCwEHgeODrnYYWk04Hb83xfi4gV3S6UmZlZL6pCZ7LtgPmSbgOeqyVGxPtaLRQRN9P4/jLAOxvMH8BxTdZ1EXBRuxk2MzPrK1UI1KeVnQEzM7OqKj1QR8RvJO0AjIuI6yS9DBhSdr7MzMyqoPRe35I+BVwBnJ+TRgI/Ly9HZmZm1VF6oCbdN94beBogIu4Dtmm5hJmZ2SBRhUC9KiL+VhuRtCFNHjhiZmY22FQhUP9G0peBTSW9C/gJ8MuS82RmZlYJVQjUJwHLgbuAY0n/7/yVUnNkZmZWEVXo9f1ifr3lraRL3vfm/3k2MzMb9EoP1JLeDZwH3E96gMmOko6NiF+VmzMzM7PylR6ogW8B74iIhQCSdgL+B3CgNjOzQa8K96iX1YJ09gBr3nhlZmY2qJXWopb0gTw4X9LVwAzSPepDWfOCDDMzs0GtzEvf7y0MPwa8PQ8vB7bs++yYmZlVT2mBOiKOLmvbZmZm/UXpnckk7Qj8IzCWQn46e82lmZnZYFB6oCa9gONC0tPIXiw5L2ZmZpVShV7ff42I70bEjRHxm9qnnQUlXSRpmaS7C2mnSnpU0rz8Obgw7WRJCyXdK+mAQvqBOW2hpJN6t3hmZmbdV4UW9XcknQJcC6yqJUbEH9pY9mLgv4BL6tLPiohvFhMk7QocBuwGbA9cJ2nnPPkc4F3AYuB2STMj4p5ulMXMzKxXVSFQvxY4AtiPNZe+I4+3FBE3SRrb5nYmAdMjYhXwoKSFwB552sKIeABA0vQ8rwO1mZmVrgqB+v3AK4qvuuwFx0s6EpgDnBgRTwIjgVsK8yzOaQCL6tL3bLRSSVOAKQBjxozpxeyamZk1VoV71H8Ehvfi+s4FdgLGA0tJjyiF9BzxetEifd3EiAsiYkJETOjo6OiNvJqZmbVUhRb1tsCfJN3O2veou/XvWRHxWG1Y0veBq/LoYmB0YdZRwJI83CzdzMysVFUI1Kf05sokbRcRS/Po+4Faj/CZwGWSvk3qTDYOuI3Uoh6X/5/7UVKHs4/2Zp7MzMy6q/RA3e6/YjUi6XJgX2BrSYtJQX9fSeNJl68fAo7N25kvaQapk9hq4LiIeCGv53jgGmAIcFFEzO92gczMzHpR6YFa0jOsuSe8ETAUeC4iNu9s2Yg4vEHyhS3mnwpMbZB+NXB1Wxk2MzPrQ6UH6ogYVhyXdAhr/m3KzMxsUKtCr++1RMTPaeN/qM3MzAaD0lvUhfdSQzpxmECTf48yMzMbbEoP1Kz9XurVpA5gk8rJipmZWbWUHqj9XmozM7PmSgvUkr7aYnJExOl9lhkzM7OKKrNF/VyDtJcDxwBbAYM6UF9UAlaZAAAKKElEQVRy/g0vDR95rPvWmZkNVqUF6oioPYMbScOAE4CjgemseT63mZnZoFbqPWpJI4AvAB8DpgG75zddmZmZGeXeo/5/wAeAC4DXRsSzZeXFzMysqsp84MmJpJdjfAVYIunp/HlG0tMl5svMzKwyyrxHXbmnopmZmVWNg6WZmVmFOVCbmZlVmAO1mZlZhfXrQC3pIknLJN1dSBshaZak+/LfLXO6JH1X0kJJd0ravbDM5Dz/fZIml1EWMzOzRvp1oAYuBg6sSzsJuD4ixgHX53GAg4Bx+TMFOBde+l/uU4A9Se/BPqUW3M3MzMrWrwN1RNwErKhLnkR6eAr57yGF9EsiuQUYLmk74ABgVkSsyA9bmcW6wd/MzKwU/TpQN7FtRCwFyH+3yekjgUWF+RbntGbp65A0RdIcSXOWL1/e6xk3MzOrNxADdTNqkBYt0tdNjLggIiZExISOjo5ezZyZmVkjAzFQP5YvaZP/Lsvpi4HRhflGAUtapJuZmZVuIAbqmUCt5/Zk4BeF9CNz7++9gJX50vg1wERJW+ZOZBNzmpmZWelKfXtWT0m6HNgX2FrSYlLv7W8AMyQdAzwCHJpnvxo4GFgIPE96pSYRsULS6cDteb6vRUR9BzUzM7NS9OtAHRGHN5n0zgbzBnBck/VcBFzUi1kzMzPrFQPx0reZmdmA4UBtZmZWYQ7UZmZmFeZAbWZmVmEO1GZmZhXmQG1mZlZhDtRmZmYV5kBtZmZWYf36gSeD0SXn3/DS8JHH7ldiTszMrC+4RW1mZlZhDtRmZmYV5kBtZmZWYQ7UZmZmFeZAbWZmVmEO1GZmZhU2YAO1pIck3SVpnqQ5OW2EpFmS7st/t8zpkvRdSQsl3Slp93Jzb2ZmlgzYQJ29IyLGR8SEPH4ScH1EjAOuz+MABwHj8mcKcG6f59TMzKyBgR6o600CpuXhacAhhfRLIrkFGC5puzIyaGZmVjSQA3UA10qaK2lKTts2IpYC5L/b5PSRwKLCsotz2lokTZE0R9Kc5cuXr8esm5mZJQP5EaJ7R8QSSdsAsyT9qcW8apAW6yREXABcADBhwoR1ppuZmfW2Aduijogl+e8y4EpgD+Cx2iXt/HdZnn0xMLqw+ChgSd/l1szMrLEBGaglvVzSsNowMBG4G5gJTM6zTQZ+kYdnAkfm3t97AStrl8jNzMzKNFAvfW8LXCkJUhkvi4hfS7odmCHpGOAR4NA8/9XAwcBC4Hng6L7PspmZ2boGZKCOiAeA1zdIfwJ4Z4P0AI7rg6yZmZl1yYAM1IOR31NtZjYwDch71GZmZgOFA7WZmVmFOVCbmZlVmAO1mZlZhTlQm5mZVZh7fQ8C7hFuZtZ/uUVtZmZWYQ7UZmZmFeZAbWZmVmG+Rz2I+d61mVn1uUVtZmZWYQ7UZmZmFeZL37YOXxI3M6sOt6jNzMwqzC1qa1uzlrZb4GZm648DdSbpQOA7wBDgvyPiGyVnqd9rJ7DXTzMzs7U5UAOShgDnAO8CFgO3S5oZEfeUm7PBp6utdrfyzWygc6BO9gAWRsQDAJKmA5MAB+oBpicBv37a+thGT09Cqnbi4qsnZj2niCg7D6WT9CHgwIj4ZB4/AtgzIo6vm28KMCWPvgq4t5ub3Bp4vJvL9lcu8+Aw2Mo82MoLPSvzDhHR0ZuZGQzcok7UIG2dM5iIuAC4oMcbk+ZExISerqc/cZkHh8FW5sFWXhicZS6b/z0rWQyMLoyPApaUlBczM7OXOFAntwPjJO0oaSPgMGBmyXkyMzPzpW+AiFgt6XjgGtK/Z10UEfPX4yZ7fPm8H3KZB4fBVubBVl4YnGUulTuTmZmZVZgvfZuZmVWYA7WZmVmFOVD3IUkHSrpX0kJJJ5Wdn/VF0kWSlkm6u5A2QtIsSfflv1uWmcfeJGm0pBslLZA0X9IJOX0gl3kTSbdJ+mMu82k5fUdJt+Yy/zh3zhxQJA2RdIekq/L4gC6zpIck3SVpnqQ5OW3A1u0qcqDuI4XHlB4E7AocLmnXcnO13lwMHFiXdhJwfUSMA67P4wPFauDEiNgF2As4Ln+3A7nMq4D9IuL1wHjgQEl7AWcCZ+UyPwkcU2Ie15cTgAWF8cFQ5ndExPjC/08P5LpdOQ7Ufeelx5RGxN+A2mNKB5yIuAlYUZc8CZiWh6cBh/RpptajiFgaEX/Iw8+QDuIjGdhljoh4No8OzZ8A9gOuyOkDqswAkkYB7wb+O4+LAV7mJgZs3a4iB+q+MxJYVBhfnNMGi20jYimkwAZsU3J+1gtJY4E3ALcywMucLwHPA5YBs4D7gaciYnWeZSDW8bOBfwFezONbMfDLHMC1kubmxyjDAK/bVeP/o+47bT2m1PovSZsBPwU+HxFPp8bWwBURLwDjJQ0HrgR2aTRb3+Zq/ZH0HmBZRMyVtG8tucGsA6bM2d4RsUTSNsAsSX8qO0ODjVvUfWewP6b0MUnbAeS/y0rOT6+SNJQUpH8UET/LyQO6zDUR8RQwm3R/frikWgNgoNXxvYH3SXqIdOtqP1ILeyCXmYhYkv8uI52Q7cEgqdtV4UDddwb7Y0pnApPz8GTgFyXmpVfl+5QXAgsi4tuFSQO5zB25JY2kTYH9SffmbwQ+lGcbUGWOiJMjYlREjCX9fm+IiI8xgMss6eWShtWGgYnA3Qzgul1FfjJZH5J0MOkMvPaY0qklZ2m9kHQ5sC/pdXiPAacAPwdmAGOAR4BDI6K+w1m/JGkf4H+Bu1hz7/LLpPvUA7XMryN1IhpCOuGfERFfk/QKUmtzBHAH8PGIWFVeTtePfOn7ixHxnoFc5ly2K/PohsBlETFV0lYM0LpdRQ7UZmZmFeZL32ZmZhXmQG1mZlZhDtRmZmYV5kBtZmZWYQ7UZmZmFeZAbVZxkv41v6HqzvwGoz3LzpOZ9R0/QtSswiS9GXgPsHtErJK0NdDt1yhK2rDwXGoz6wfcojartu2Ax2sP0IiIx/Nzl98k6Xf5fdC3SRqW3xH9g/zu4DskvQNA0lGSfiLpl8C1Oe2fJd2eW+mnlVc8M+uMW9Rm1XYt8FVJfwauA34M/D7//UhE3C5pc+AvpPckExGvlfRq0huPds7reTPwuohYIWkiMI70zGYBMyW9Lb+e1Mwqxi1qswrL73x+IzAFWE4K0McCSyPi9jzP0/ly9j7ApTntT8DDQC1Qzyo84nFi/twB/AF4NSlwm1kFuUVtVnH5dZKzgdmS7gKOo/GrFFu9V/O5uvnOiIjzey2TZrbeuEVtVmGSXiWp2NodT3pL1faS3pTnGZZfs3gT8LGctjPphQn3NljtNcAn8vuzkTQyv2vYzCrILWqzatsM+M/8SsnVwELSZfAf5PRNSfen9we+B5yXW92rgaNyT/G1VhgR10raBfh9nvYs8HH8TmGzSvLbs8zMzCrMl77NzMwqzIHazMyswhyozczMKsyB2szMrMIcqM3MzCrMgdrMzKzCHKjNzMwq7P8DMl2bxBLOp+UAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a287f60f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a462fffd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a27aaa588>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a311e5668>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a26249710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a359b1588>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a1fd6f978>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a37dfa470>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY4AAAEWCAYAAABxMXBSAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzt3XmcHFW99/HPNxAEIRCWgCEEiBpkcYkQNuFqVIiACooLoCKbBhUEHuFKUK5scsXXI6g8F8EoEfAqMaJA5KIQkOBFL5AEwhIWGSCQkJgEwhLAGw35PX+c00llmOmpSqane2a+79erX1N1qqrrd7pr+ld1Ti2KCMzMzMoa0OwAzMysd3HiMDOzSpw4zMysEicOMzOrxInDzMwqceIwM7NKnDh6AUlXSPp2k9YtST+T9Lyku5sRQxWSZksa04Pr+6ykm3tqfV2RtL2kkLRus2NZG5I+LmmupJclvbvZ8dSszfYladtcn3W6Oawe58SxBiTNkbRQ0oaFsi9ImtbEsBplX2B/YJuI2KOjGSRtI+kXkp6T9IqkuyUd1OjAOkqoEbFLRExr9LoL6/tFRIztqfV1t/zd/UbSs5JelPSApKObHRfwPeDEiNgoIu5tdjA1Vbav/DuxX2HZp3N9XmtYgD3EiWPNrQuc3OwgqlqDvZ3tgDkR8Uon77cZcAfwD2AXYAvg+8AkSR9bm1itR/wcmEv6njcHPg8s7M4VrOHRz3bA7DVcX7fv0ff2I7huFxF+VXwBc4DxwBJgcC77AjAtD28PBLBuYZlpwBfy8NHAn0k/sC8ATwDvyeVzgUXAUYVlrwAuA6YCS4Hbge0K03fM05YAjwKfbrfspcCNwCvAfh3UZ2tgSl6+DfhiLj8O+F/gNeBl4JwOlj0PeBAY0K789FwvdfV55PFjgYeB54GbavXLy38/fyYvAvcDbwfGAf8kJayXgd8Vvpv98vAbgB8A8/PrB8Ab8rQxwDzg1PzeC4BjCvEcBDyUP+9ngNM62RaOBu4ojAfwJeCxXJdLAHWy7ADSdvQ48BwwGdisMP3XwN9yvf8E7FKYtgFwIfBUnn5HLqt91kcBTwPPAt+ssy2/DIyqM31f4C+k7XQucHQu3wS4ClicYziztg2w+va9BPh2ve+43frekGMK0vb6eC7fKW8zL5ASysEVt/GjSdvjUuBJ4LOFaV/McS3N3/muhW3pdNI2t4y0sziHVdvX2cA1wK/ysvcA78rTfg6sAP6e6/N12v0f5Pqclz+rpcDNwBaFuD6fP9vngH8rrrvZr6YH0BtftS8Q+G3hn6Jq4lgOHAOsA3yb9E9+Sf7HGZs3pI0K/xhLgffm6T8k/1gBG5L+oY/JG/aupB+LXQrLvgjsQ/qhWr+D+twO/AhYHxhF+jH4YCHWO+p8FnfScUIZkT+DkSU+j4+REtZOuQ5nAn/J0z4EzAQGk5LITsDQQt2+3dF3k4fPzfFtCQwh/QCel6eNyd/BucBAUqJ4Fdg0T18A/Ese3pT8Y9JBPVf7fHI9b8jxbps/ywM6WfaUHN82+Xv9MXB1YfqxwCBWJcBZhWmX5M9wGGkbek+er/ZZ/4SUSN5F+tHbqZMYbiH9cB0ObNtu2rak7e6I/BltTk4ypKRxfY5ve+CvwHHttu+v5u9zg3rfcSdxBfDWPDwwL/sNYD3gAzmut5XZxkn/Iy8V5h/Kqv+PT5F2DHYnbV9vZdVOyxxgFjAc2KCD7ets0s7LJ3OMp5GS0sD283b0u5C/v8eBHfJnNA24IE/bmZRw9s11/l5elxNHb32xKnG8PW+wQ6ieOB4rTHtHnn+rQtlzhX/SK4BJhWkbkY4ChgOHAf/dLr4fA2cVlr2qTl2G5/caVCj7DnBFIdZ6iaMN+FIH5evnOr2nxOfxe/KPTh4fQPoR3470I/FXYC9ef1RzBfUTx+PAQYVpHyI1u0FKHH9vF9MiYK88/DRwPLBxF9vCap9Prue+hfHJwPhOln2YnKDz+ND847BuB/MOzu+9Sf58/k7eu203X+2z3qZQdjdweCcxbApcQNqLf430Q7l7nnYGcG0Hy6xDSkY7F8qOZ9X2fzTwdLtlOv2OO4mrmDj+hXTkNaAw/Wrg7JLb+IakI5VPkBNAYdpNwMl1/s+PrbN9nQ3c2a5OxR2OlfO2+26KiePMwvSvAH/Iw99i9Z2IN5KOrlsicbiPYy1ExIOkvcvxa7B4sR357/n92pdtVBifW1jvy6QmgK1JP657Snqh9gI+C7ypo2U7sDWwJCKWFsqeIu3JlvEs6QevvVrZ4hLvsR3ww0L8S0h7f8Mi4o/Af5D2sBdKmiBp45KxbU2qS81TuazmuYhYXhh/lVWf+SdIRyFPSbpd0t4l1wnpR66j92xvO+DaQr0fJv14byVpHUkXSHpc0kukHyFIfUhbkBLz42sbQ0Q8HxHjI2IXYCtS4rhOkkg7FR2tYwvSXnD7z7a4zbTf5jr9juvUoWZrYG5ErKiwvpUi9c8dRmpCXCDpvyTtmCd3Vscu37f99BzfPFbfxrrS2fe0dbv3fpW0M9kSnDjW3lmkNtLiRlzrSH5joaz4Q74mhtcGJG0EbEZqt58L3B4RgwuvjSLiy4Vlo877zgc2kzSoULYt6fC9jFuAT0hqvy19mvRP9Dhdfx5zgePb1WGDiPgLQERcHBG7kTrfdwD+tUS9anXbrl295pepVERMj4hDSM1c15GOHLrbXODAdvVePyKeAT4DHEI6st2EtLcK6cf2WVLf01u6M5iIeJbUJLI1afua28k6niUdGbX/bIvbTPvvpu533IX5wPB221hX61tNRNwUEfuTdmgeITXl1eKq9zl2tY0V/y8HkJoda9tYV8vWsyC/V+29NyA1FbYEJ461FBFtpM6xkwpli0kb9efynuOxrP0/+UGS9pW0HqlD7a6ImEs64tlB0pGSBubX7pJ2Khn/XFLb/3ckrS/pnaRO8V+UjOv7wMbA5ZLelN/jCFJn3lkRsaLE53EZcIakXQAkbSLpU3l4d0l7ShpISkC1znpIR21vrhPb1cCZkoZI2oJ0+P+fXVVI0nr5+oxNIuKfpPbxRpxCeRlwvqTt8nqHSDokTxtEag56jpRw/722UN6znQhcJGnr/JnuLekNVQOQ9F1Jb5e0bt55+DLQFhHPkbaB/SR9Ok/fXNKoSKeTTs6xD8rxf436n22n33EJd5G++6/n7XsM8FFgUsk6biXp4Hz6/DJS30Ht+/wpcJqk3ZS8tfZ9lLSbpEPzWVen5Pe/M0/ravus5xrgo5Lek//nzyHtNLQEJ47ucS6pHbXoi6Q94+dIe8pl9qzq+SXp6GYJsBupOYrcxDSW1Lk5n3To+11SR2lZR5D2aOcD15J+8KeWWTD/wOxLajp5iPRPeRVwQkRMLMza6ecREdfmmCflZpkHgQPz5I1Je4fPs+oMk+/laZcDO+fmj+s6CO/bwAzSWTEPkM56KXsh5ZHAnBzPl4DPlVyuih+Szma7WdJS0g/OnnnaVaT6PkP6XO9st+xppDpNJ20T32XN/p/fSPrOa2f3bQccDOm6A1Jz3al5HbNIne2QOr5fycvcQdo+i9/3arr4juuKiH/kmA4kHe38CPh8RDxSso4Dch3m53q8j9SfQET8Gjg/x7+UdHS5Wcn3hXSCwGGk7fNI4NC8swGpr/DMvH2eVuE9iYjZpM94EunoYympD25ZlfdpFOWOF7Nukfsf/kzqVP1Ws+MxaxRJZ5M68BuxU9F+XRuRkvvIiHiy0evrio84rFtFxEukvdTXJK1tv45ZvyXpo5LemJvYvkc6wpzT3KiShiWO3NZ9t6T7lO7vck4uHyHpLkmPSfpVbr9D0hvyeFuevn3hvc7I5Y9K+lCjYrbuERFzI+KciPhb13ObWScOYdXFqyNJp1S3RBNRw5qq8ul8G0bEy7lj8w7SLTq+Bvw2IiZJugy4LyIulfQV4J0R8SVJhwMfj4jDJO1M6uTcg3S2xy3ADtEH7vdiZtYbNeyII5KX8+jA/ArSBV3X5PIrSVeUQsquV+bha4AP5uRzCOnit2W5ba+NlETMzKwJGnrjLqWbjc0kXcZ/Cemc/hcKF13NY9X1D8PIF7xExHJJL5LOWx7G6meUFJcprmsc6f5FbLjhhrvtuOOO7WcxM7M6Zs6c+WxEDOlqvoYmjtycNErSYNIpfx1dW1BrK+voHOWoU95+XROACQCjR4+OGTNmrFHMZmb9laSnup6rh86qiogXSPdl2QsYrFW3KC5eZTmPfBVmnr4J6ZzrleUdLGNmZj2skWdVDclHGrXL5fcj3YvnNtLdJCHd+vn6PDwlj5On/zGfQTAFODyfdTWCdHZByz+Jzsysr2pkU9VQ4MrczzEAmBwRN0h6iHT16LeBe0lX/5L//lxSG+lI43BIV1BKmky6enY56Ypkn1FlZtYkffLKcfdxmJlVJ2lmRIzuaj5fOW5mZpU4cZiZWSVOHGZmVokTh5mZVeLEYWZmlTT0yvHeasXSi1cODxh0Up05zcz6Hx9xmJlZJU4cZmZWiROHmZlV4sRhZmaVOHGYmVklThxmZlaJE4eZmVXixGFmZpU4cZiZWSVOHGZmVokTh5mZVeLEYWZmlThxmJlZJU4cZmZWiROHmZlV4sRhZmaVOHGYmVklThxmZlaJE4eZmVXixGFmZpU4cZiZWSUNSxyShku6TdLDkmZLOjmXny3pGUmz8uugwjJnSGqT9KikDxXKD8hlbZLGNypmMzPr2roNfO/lwKkRcY+kQcBMSVPztO9HxPeKM0vaGTgc2AXYGrhF0g558iXA/sA8YLqkKRHxUANjNzOzTjQscUTEAmBBHl4q6WFgWJ1FDgEmRcQy4ElJbcAeeVpbRDwBIGlSnteJw8ysCXqkj0PS9sC7gbty0YmS7pc0UdKmuWwYMLew2Lxc1ll5+3WMkzRD0ozFixd3cw3MzKym4YlD0kbAb4BTIuIl4FLgLcAo0hHJhbVZO1g86pSvXhAxISJGR8ToIUOGdEvsZmb2eo3s40DSQFLS+EVE/BYgIhYWpv8EuCGPzgOGFxbfBpifhzsrNzOzHtbIs6oEXA48HBEXFcqHFmb7OPBgHp4CHC7pDZJGACOBu4HpwEhJIyStR+pAn9KouM3MrL5GHnHsAxwJPCBpVi77BnCEpFGk5qY5wPEAETFb0mRSp/dy4ISIeA1A0onATcA6wMSImN3AuM3MrI5GnlV1Bx33T9xYZ5nzgfM7KL+x3nJmZtZzfOW4mZlV4sRhZmaVOHGYmVklThxmZlaJE4eZmVXixGFmZpU4cZiZWSVOHGZmVokTh5mZVeLEYWZmlThxmJlZJU4cZmZWiROHmZlV4sRhZmaVOHGYmVklThxmZlaJE4eZmVXSZeKQdLKkjZVcLukeSWN7IjgzM2s9ZY44jo2Il4CxwBDgGOCChkZlZmYtq0ziqD03/CDgZxFxHx0/S9zMzPqBMoljpqSbSYnjJkmDgBWNDcvMzFrVuiXmOQ4YBTwREa9K2pzUXGVmZv1Ql4kjIlZIWgjsLKlMojEzsz6sy0Qg6bvAYcBDwGu5OIA/NTAuMzNrUWWOID4GvC0iljU6GDMza31lOsefAAY2OhAzM+sdyhxxvArMknQrsPKoIyJOalhUZmbWssoccUwBzgP+AswsvOqSNFzSbZIeljRb0sm5fDNJUyU9lv9umssl6WJJbZLul7Rr4b2OyvM/JumoNamomZl1jzJnVV0paT1gh1z0aET8s8R7LwdOjYh78rUfMyVNBY4Gbo2ICySNB8YDpwMHAiPza0/gUmBPSZsBZwGjSZ3yMyVNiYjnq1TUzMy6R5l7VY0BHgMuAX4E/FXSe7taLiIWRMQ9eXgp8DAwDDgEuDLPdiWp851cflUkdwKDJQ0FPgRMjYglOVlMBQ4oX0UzM+tOZfo4LgTGRsSjAJJ2AK4Gdiu7EknbA+8G7gK2iogFkJKLpC3zbMOAuYXF5uWyzsrNzKwJyvRxDKwlDYCI+CsVzrKStBHwG+CUfLPETmftoCzqlLdfzzhJMyTNWLx4cdnwzMysojKJY0a+nfqY/PoJJTrHASQNJCWNX0TEb3PxwtwERf67KJfPA4YXFt8GmF+nfDURMSEiRkfE6CFDhpQJz8zM1kCZxPFlYDZwEnAy6QryL3W1kCQBlwMPR8RFhUlTgNqZUUcB1xfKP5/PrtoLeDE3ad0EjJW0aT4Da2wuMzOzJihzVtUy4KL8qmIf4EjgAUmzctk3SM/ymCzpOOBp4FN52o2kO/C2ka4dOSavf4mk84Dpeb5zI2JJxVjMzKybdJo4JE2OiE9LeoAO+hQi4p313jgi7qDz53Z8sIP5Azihk/eaCEystz4zM+sZ9Y44Ts5/P9ITgZiZWe/QaR9H7ZRZ4CsR8VTxBXylZ8IzM7NWU6ZzfP8Oyg7s7kDMzKx3qNfH8WXSkcVbJN1fmDQI+HOjAzMzs9ZUr4/jl8Dvge+Q7idVs9RnNZmZ9V/1+jhejIg5wJnA33Lfxgjgc5IG91B8ZmbWYsr0cfwGeE3SW0kX9I0gHY2YmVk/VCZxrIiI5cChwA8i4v8AQxsblpmZtaoyieOfko4APg/ckMv8KFkzs36qTOI4BtgbOD8inpQ0AvjPxoZlZmatqsy9qh6SdDqwbR5/knS/KTMz64fKPAHwo8As4A95fJSkKY0OzMzMWlOZpqqzgT2AFwAiYhbpzCozM+uHyiSO5RHxYruy190t18zM+ocyzxx/UNJngHUkjSQ90OkvjQ3LzMxaVZkjjq8CuwDLgKuBl4BTGhmUmZm1rjJnVb0KfDO/zMysn+sycUi6jY6fAPiBhkRkZmYtrUwfx2mF4fWBTwDLGxOOmZm1ujJNVTPbFf1Z0u0NisfMzFpcmaaqzQqjA4DdgDc1LCIzM2tpZZqqZpL6OERqonoSOK6RQZmZWesq01Tlq8TNzGylMk1Vh9abHhG/7b5wzMys1ZVpqjoOeA/wxzz+fmAa8CKpCcuJw8ysHymTOALYOSIWAEgaClwSEcc0NDIzM2tJZW45sn0taWQLgR0aFI+ZmbW4Mkcc0yTdRLpPVQCHA7c1NCozM2tZXR5xRMSJwGXAu4BRwISI+GpXy0maKGmRpAcLZWdLekbSrPw6qDDtDEltkh6V9KFC+QG5rE3S+KoVNDOz7lXmiIOIuBa4tuJ7XwH8B3BVu/LvR8T3igWSdiYdyewCbA3cIqnWHHYJsD8wD5guaUpEPFQxFjMz6yalEseaiIg/Sdq+5OyHAJMiYhnwpKQ20lMHAdoi4gkASZPyvE4cZmZNUqZzvLudKOn+3JS1aS4bBswtzDMvl3VW/jqSxkmaIWnG4sWLGxG3mZlRJ3FIujX//W43ru9S4C2kvpIFwIW11XUwb9Qpf31hxISIGB0Ro4cMGdIdsZqZWQfqNVUNlfQ+4ODcRLTaj3hE3FN1ZRGxsDYs6SfADXl0HjC8MOs2wPw83Fm5mZk1Qb3E8S1gPOnH+qJ20wKo/CAnSUML14R8HKidcTUF+KWki0id4yOBu0nJaqSkEcAzpA70z1Rdr5mZdZ9OE0dEXANcI+nfIuK8qm8s6WpgDLCFpHnAWcAYSaNIiWcOcHxe12xJk0md3suBEyLitfw+JwI3AesAEyNidtVYzMys+yiiwy6D1WeSDgbem0enRcQN9eZvttGjR8eMGTPWePkVSy9eOTxg0EndEZKZWcuTNDMiRnc1X5dnVUn6DnAy6WjgIeDkXGZmZv1Qmes4PgyMiogVAJKuBO4FzmhkYGZm1prKXscxuDC8SSMCMTOz3qHMEcd3gHsl3UY6y+m9+GjDzKzfKvPo2KslTQN2JyWO0yPib40OrFW4o9zMbHVlb3K4gHSthZmZ9XPNuFeVmZn1Yk4cZmZWSd3EIWlA8UFMZmZmdRNHvnbjPknb9lA8ZmbW4sp0jg8FZku6G3ilVhgRBzcsKjMza1llEsc5DY/CzMx6jTLXcdwuaTtgZETcIumNpDvVmplZP1TmJodfBK4BfpyLhgHXNTIoMzNrXWVOxz0B2Ad4CSAiHgO2bGRQZmbWusokjmUR8Y/aiKR16eS532Zm1veVSRy3S/oGsIGk/YFfA79rbFhmZtaqyiSO8cBi4AHSo15vBM5sZFBmZta6ypxVtSI/vOkuUhPVo1HmebNmZtYndZk4JH0YuAx4nHRb9RGSjo+I3zc6ODMzaz1lLgC8EHh/RLQBSHoL8F+AE4eZWT9Upo9jUS1pZE8AixoUj5mZtbhOjzgkHZoHZ0u6EZhM6uP4FDC9B2IzM7MWVK+p6qOF4YXA+/LwYmDThkVkZmYtrdPEERHH9GQgZmbWO5Q5q2oE8FVg++L8vq26mVn/VOasquuAy0lXi69obDhmZtbqypxV9b8RcXFE3BYRt9deXS0kaaKkRcVHz0raTNJUSY/lv5vmckm6WFKbpPsl7VpY5qg8/2OSjlqjWpqZWbcpkzh+KOksSXtL2rX2KrHcFcAB7crGA7dGxEjg1jwOcCAwMr/GAZdCSjTAWcCewB7AWbVkY2ZmzVGmqeodwJHAB1jVVBV5vFMR8SdJ27crPgQYk4evBKYBp+fyq/KtTO6UNFjS0Dzv1IhYAiBpKikZXV0ibjMza4AyiePjwJuLt1ZfC1tFxAKAiFggqfZcj2HA3MJ883JZZ+WvI2kc6WiFbbfdthtCNTOzjpRpqroPGNzgONRBWdQpf31hxISIGB0Ro4cMGdKtwZmZ2Spljji2Ah6RNB1YVitcw9NxF0oamo82hrLq1iXzgOGF+bYB5ufyMe3Kp63Bes3MrJuUSRxndeP6pgBHARfkv9cXyk+UNInUEf5iTi43Af9e6BAfC5zRjfGYmVlFZZ7H0eWptx2RdDXpaGELSfNICegCYLKk44CnSfe9gvRwqIOANuBV4Ji87iWSzmPVvbHOrXWUm5lZc5S5cnwpq/oV1gMGAq9ExMb1louIIzqZ9MEO5g3ghE7eZyIwsas4zcysZ5Q54hhUHJf0MdI1FWZm1g+VOatqNRFxHV1cw2FmZn1XmaaqQwujA4DRdHJKrJmZ9X1lzqoqPpdjOTCHdKW3mZn1Q2X6OPxcDjMzW6neo2O/VWe5iIjzGhCPmZm1uHpHHK90ULYhcBywOeDEYWbWD9V7dOyFtWFJg4CTSRfmTQIu7Gw5MzPr2+r2ceTnYXwN+CzpNui7RsTzPRGYmZm1pnp9HP8XOBSYALwjIl7usajMzKxl1bsA8FRga+BMYL6kl/JrqaSXeiY8MzNrNfX6OCpfVW5mZn2fk4OZmVXixGFmZpU4cZiZWSVOHGZmVokTh5mZVeLEYWZmlThxmJlZJU4cZmZWiROHmZlV4sRhZmaVlHl0rHVgxdKLVw4PGHRSEyMxM+tZPuIwM7NKnDjMzKwSJw4zM6vEicPMzCppSuKQNEfSA5JmSZqRyzaTNFXSY/nvprlcki6W1Cbpfkm7NiNmMzNLmnnE8f6IGBURo/P4eODWiBgJ3JrHAQ4ERubXOODSHo/UzMxWaqWmqkOAK/PwlcDHCuVXRXInMFjS0GYEaGZmzUscAdwsaaakcblsq4hYAJD/bpnLhwFzC8vOy2WrkTRO0gxJMxYvXtzA0M3M+rdmXQC4T0TMl7QlMFXSI3XmVQdl8bqCiAnABIDRo0e/brqZmXWPphxxRMT8/HcRcC2wB7Cw1gSV/y7Ks88DhhcW3waY33PRmplZUY8nDkkbShpUGwbGAg8CU4Cj8mxHAdfn4SnA5/PZVXsBL9aatMzMrOc1o6lqK+BaSbX1/zIi/iBpOjBZ0nHA08Cn8vw3AgcBbcCrwDE9H7KZmdX0eOKIiCeAd3VQ/hzwwQ7KAzihB0IzM7MSWul0XDMz6wWcOMzMrBInDjMzq8SJw8zMKnHiMDOzSpw4zMysEicOMzOrxInDzMwqceIwM7NKmnV33D5rxdKLVw4PGHRSEyMxM2sMH3GYmVklThxmZlaJE4eZmVXixGFmZpU4cZiZWSVOHGZmVokTh5mZVeLEYWZmlThxmJlZJU4cZmZWiROHmZlV4ntV9RDfw8rM+gofcZiZWSVOHGZmVokTh5mZVeI+jiZz34eZ9TY+4jAzs0qcOMzMrJJe01Ql6QDgh8A6wE8j4oImh9RQxSYsWL0Zy81bZtZMvSJxSFoHuATYH5gHTJc0JSIeam5kraWzhOJEY2bdqVckDmAPoC0ingCQNAk4BHDiWAtVE029BNRdSau7kpyTpVnjKCKaHUOXJH0SOCAivpDHjwT2jIgTC/OMA8bl0bcBj67FKrcAnl2L5Xuj/lbn/lZfcJ37i7Wp83YRMaSrmXrLEYc6KFst40XEBGBCt6xMmhERo7vjvXqL/lbn/lZfcJ37i56oc285q2oeMLwwvg0wv0mxmJn1a70lcUwHRkoaIWk94HBgSpNjMjPrl3pFU1VELJd0InAT6XTciRExu4Gr7JYmr16mv9W5v9UXXOf+ouF17hWd42Zm1jp6S1OVmZm1CCcOMzOrxImjQNIBkh6V1CZpfLPjaQRJEyUtkvRgoWwzSVMlPZb/btrMGLubpOGSbpP0sKTZkk7O5X223pLWl3S3pPtync/J5SMk3ZXr/Kt8skmfIWkdSfdKuiGP9+n6AkiaI+kBSbMkzchlDd22nTiywm1NDgR2Bo6QtHNzo2qIK4AD2pWNB26NiJHArXm8L1kOnBoROwF7ASfk77Yv13sZ8IGIeBcwCjhA0l7Ad4Hv5zo/DxzXxBgb4WTg4cJ4X69vzfsjYlTh+o2GbttOHKusvK1JRPwDqN3WpE+JiD8BS9oVHwJcmYevBD7Wo0E1WEQsiIh78vBS0g/LMPpwvSN5OY8OzK8APgBck8v7VJ0lbQN8GPhpHhd9uL5daOi27cSxyjBgbmF8Xi7rD7aKiAWQfmSBLZscT8NI2h54N3AXfbzeudlmFrAImAo8DrwQEcvzLH1tG/8B8HVgRR7fnL5d35oAbpY0M996CRq8bfeK6zh6SJe3NbHeTdJGwG+AUyLipbRD2ndFxGvAKEmDgWuBnTqarWejagxJHwEWRcRMSWNqxR3M2ifq284+ETFf0pbAVEmPNHqFPuJYpT/f1mShpKEA+e+iJsfT7SQNJCWNX0TEb3Nxn683QES8AEwj9e/94/L1AAACz0lEQVQMllTbYexL2/g+wMGS5pCamT9AOgLpq/VdKSLm57+LSDsIe9DgbduJY5X+fFuTKcBRefgo4PomxtLtclv35cDDEXFRYVKfrbekIflIA0kbAPuR+nZuAz6ZZ+szdY6IMyJim4jYnvS/+8eI+Cx9tL41kjaUNKg2DIwFHqTB27avHC+QdBBpL6V2W5PzmxxSt5N0NTCGdOvlhcBZwHXAZGBb4GngUxHRvgO915K0L/DfwAOsav/+Bqmfo0/WW9I7SZ2i65B2ECdHxLmS3kzaI98MuBf4XEQsa16k3S83VZ0WER/p6/XN9bs2j64L/DIizpe0OQ3ctp04zMysEjdVmZlZJU4cZmZWiROHmZlV4sRhZmaVOHGYmVklThxma0HSN/PdZ+/Pdyfds9kxmTWabzlitoYk7Q18BNg1IpZJ2gJY49t2S1q3cF8ls5blIw6zNTcUeLZ2QVlEPJvvGbS7pL/kZ2HcLWlQfj7Gz/JzE+6V9H4ASUdL+rWk3wE357J/lTQ9H8Wc07zqmXXMRxxma+5m4FuS/grcAvwK+J/897CImC5pY+DvpOdEEBHvkLQj6W6mO+T32Rt4Z0QskTQWGEm635CAKZLem2+Hb9YSfMRhtoby8y52A8YBi0kJ43hgQURMz/O8lJuf9gV+nsseAZ4CaoljauF2EGPz617gHmBHUiIxaxk+4jBbC/nW5dOAaZIeAE6g41t317uH+yvt5vtORPy424I062Y+4jBbQ5LeJql4NDCKdAfarSXtnucZlG/r/Sfgs7lsB9LN5x7t4G1vAo7Nzw5B0rD8nAWzluEjDrM1txHw//Lty5cDbaRmq5/l8g1I/Rv7AT8CLstHJcuBo/OZWKu9YUTcLGkn4H/ytJeBz9FHnxVivZPvjmtmZpW4qcrMzCpx4jAzs0qcOMzMrBInDjMzq8SJw8zMKnHiMDOzSpw4zMyskv8PyVMqzOk2BLUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a2cc96748>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "bins = np.linspace(0, 50, 100)\n",
    "for i in range(10):\n",
    "    fig, ax = plt.subplots()\n",
    "    ax.hist(qstack_dict[i]['Score'], bins, alpha=0.5, color=LABEL_COLOR_MAP2[i])\n",
    "    ax.set_ylabel('Number of questions')\n",
    "    ax.set_xlabel('Score')\n",
    "    title = 'Number of Questions in each Score for '+assign_topics2[i]\n",
    "    ax.set_title(title)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 259,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "for i in range(10):\n",
    "    qstack_dict[i] = qstack_dict[i].sort_values(by=['Score'], ascending=False)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 260,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a33f41128>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "scores_st = np.zeros(10)\n",
    "for i in range(10):\n",
    "    scores_st[i] = qstack_dict[i]['Score'].sum()\n",
    "\n",
    "color2 = [LABEL_COLOR_MAP2[l] for l in range(10)]\n",
    "fig, ax = plt.subplots()\n",
    "ind = np.arange(len(scores_st))\n",
    "ax.barh(ind, scores_st, alpha=0.5, color=color2)\n",
    "ax.set_yticks(ind)\n",
    "ax.set_yticklabels(assign_topics2)\n",
    "ax.set_xlabel('Sum of Total Scores')\n",
    "ax.set_title('Summation of Scores in Each Topic')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 262,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a2a7935f8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#number of questions in each dataframe\n",
    "num_qstack = np.zeros(10)\n",
    "for i in range(10):\n",
    "    num_qstack[i] = len(qstack_dict[i])\n",
    "    \n",
    "#total ratings\n",
    "fig, ax = plt.subplots()\n",
    "ind = np.arange(len(num_qstack))\n",
    "ax.barh(ind, num_qstack, alpha=0.5, color=color2)\n",
    "ax.set_yticks(ind)\n",
    "ax.set_yticklabels(assign_topics2)\n",
    "ax.set_xlabel('Total Number of Questions')\n",
    "ax.set_title('Number of Questions in Each Topic')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 263,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a359b1ba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#normalize data based on its total number of questions\n",
    "norm_scores_st = np.zeros(10)\n",
    "for i in range(10):\n",
    "    norm_scores_st[i] = scores_st[i]/num_qstack[i]\n",
    "    \n",
    "fig, ax = plt.subplots()\n",
    "ind = np.arange(len(norm_scores_st))\n",
    "ax.barh(ind, norm_scores_st, alpha=0.5, color=color2)\n",
    "ax.set_yticks(ind)\n",
    "ax.set_yticklabels(assign_topics2)\n",
    "ax.set_xlabel('Normalized Score')\n",
    "ax.set_title('Normalized Score Based on Question Population')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the Normalized Score bar chart, we see that the most common topic is still basic Python concepts. Following behind this are questions that have been asked by users who are learning python coming from another language. These are highly related categories, and thus, these findings still support our hypothesis that many of the questions that are asked on StackOverflow are by beginners to the language who are struggling to attain genuine understanding.\n",
    "\n",
    "Here are a few examples of the most popular questions within the basic python concepts category, as well as from the learning python from another language category:\n",
    "- A reliable coder friend told me that python's current multi-threading implementation is seriously buggy - enough to avoid using altogether. What can said about this rumor?\n",
    "- Is there a way using python's standard library to easily determine (i.e. one function call) the last day of a given month?\n",
    "- Is there any performance difference between tuples and lists when it comes to instantiation and retrieval of elements?\n",
    "\n",
    "Once again, many of these are simply, yet thought-provoking questions that are truly indicative of a desire to understand the python language more deeply than simply being able to code using it.\n",
    "\n",
    "Many of us in this project group would also fall into the category of learning python from another language. And it's easy to see why these questions, along with basic python concepts, would be popular- in order to truly grasp any programming language, we need to understand the core concepts that fuel it. Simply practicing programming would not help. Sure, it's great to get immediate help whenever we run into confusing error statements, but if we want to better ourselves in the long run, it's the act of trying to understand the concepts behind what we are trying to code that will ultimately steer us towards more consistent success.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3 - Piazza data from COGS108s2017 & COGS108w2018\n",
    "\n",
    "For the Piazza data, we will employ many similar processes to clean the data. Using piazza.py, we were able to parse the questions and determine their topic, and group this information together to see which questions are the most frequently asked. We mainly looked at the number of follow-ups each question had to determine the question's popularity. Of course, we also need to clean this dataset using the Safe Harbour method, which is outlined further in the Ethics and Privacy section."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from piazza_api import Piazza\n",
    "import json"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#create dataframe of piazza posts\n",
    "df_piazza = pd.read_json(\"posts.json\", orient='values')\n",
    "df_piazza2 = pd.read_json(\"lastyearposts.json\", orient='values')\n",
    "df_piazza = df_piazza.append(df_piazza2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 276,
   "metadata": {},
   "outputs": [],
   "source": [
    "#aggregated list of names/words we need to remove from the dataframe (including times, locations, and course-specific materials)\n",
    "names = ['thomas', 'david', 'harshita', 'ilkay', 'wan', 'tianyu', 'tom', 'novak', 'washington', 'wills', 'kevin', 'gael', 'oscar', 'mengting', 'josh', 'altintas', 'shuai', 'megan', 'martin', 'mande', 'screen_shot_20180128_at_8.32.21_pm.png', 'jim', 'monday', 'tuesday', 'wednesday', 'thursday', 'friday', 'saturday', 'sunday', 'mon', 'tues', 'wed', 'thurs', 'fri', 'sat', 'sun', 'tue', 'thur', '3:00-4:00pm', 'january', 'february', 'march', 'april', 'may', 'june', 'july', 'august', 'september', 'october', 'november', 'december', 'jan', 'feb', 'mar', 'apr', 'jun', 'jul', 'aug', 'sept', 'sep', 'oct', 'nov', 'dec', '14th', 'martin', 'luther', 'king', 'jr.', 'day', 'san', 'diego', 'bradley', '//github.com/cogs108/overview/blob/master/wi18.md', 'adam', '-zachariah', 'nye', 'zjgutier', '3:30-4:30pm', 'nick.jones25', '//github.com/cogs108/tutorials/blob/master/11-testingdistributions.ipynb', 'gmail.com', '//goo.gl/forms/crdl0qjsqxdzacmi2', '//github.com/cogs108/assignments', '27th', 'bill', 'reasonscree_shot_20180128_at_4.35.36_pm.png', '//gihub.com/cogs108/assignments/tree/master/a1', '2016', '2015', '2014', '12:30-1:30pm', 'feb.12', 'dylan', \"''screen_shot_20180213_at_10.00.39_pm.png\", 'mark', 'bob', 'john', 'voytek', 'screen_shot_20180128_at_9.19.33_pm.png', 'centr', 'content']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#method used to clean each post in df_piazza\n",
    "def convert_body(body):\n",
    "    body = str(body).lower()\n",
    "    #body = body[11:len(body) - 1]\n",
    "    body = body.replace('\\\\n', ' ')\n",
    "    body = body.replace('\\\\xa0', ' ')\n",
    "    for name in names:\n",
    "        body = body.replace(name, ' ')    \n",
    "    return bs.BeautifulSoup(body, 'lxml').get_text()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df_piazza.columns = ['followups','pBody']\n",
    "df_piazza['pBody'] = df_piazza['pBody'].apply(convert_body)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "tfidf = TfidfVectorizer(sublinear_tf=True, analyzer='word', \n",
    "                        max_features=500, tokenizer=word_tokenize, max_df=0.8)\n",
    "\n",
    "#Transform data\n",
    "tf_pBody = tfidf.fit_transform(df_piazza[\"pBody\"])\n",
    "tf_pBody = tf_pBody.toarray() #transformed vectors into a matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "pca = PCA(n_components=2).fit(tf_pBody)\n",
    "data2D = pca.transform(tf_pBody)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "kmeans = MiniBatchKMeans(n_clusters=10).fit(tf_pBody) #default threshold\n",
    "centers2D = pca.transform(kmeans.cluster_centers_)\n",
    "clusters = kmeans.cluster_centers_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "LABEL_COLOR_MAP3= {0 : '#ff7575',\n",
    "                   1 : '#ff8775',\n",
    "                   2 : '#ffba79',\n",
    "                   3 : '#ffcb71',\n",
    "                   4 : '#fff671',\n",
    "                   5 : '#dbe0e8',\n",
    "                   6 : '#dbedf4',\n",
    "                   7 : '#b3e5eb',\n",
    "                   8 : '#7ec5dd',\n",
    "                   9 : '#2d76c4'\n",
    "                  }\n",
    "            "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a458d9b00>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "label_color_pi = [LABEL_COLOR_MAP3[l] for l in kmeans.labels_]\n",
    "plt.scatter(data2D[:, 0], data2D[:, 1], c=label_color_pi)\n",
    "plt.scatter(centers2D[:,0], centers2D[:,1], \n",
    "            marker='x', s=200, linewidths=3, c='r')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df_piazza[\"label\"] = pd.Series(kmeans.labels_, index = df_piazza.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "top30pi_dict = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[('office', 18), ('hours', 15), ('pm', 13), (\"'s\", 8), ('professor', 5), ('csb', 5), ('ta', 5), ('week', 5), ('class', 4), ('know', 4), ('hour', 4), ('go', 3), ('section', 3), (\"'hi\", 3), ('time', 3), ('thanks', 3), ('tas', 2), ('1-2', 2), ('2-3', 2), ('instead', 2), ('sections', 2), ('project', 2), (\"'hey\", 2), ('moved', 2), ('every', 2), ('still', 2), ('115', 2), ('post', 2), ('``', 2), (\"''\", 2)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[(\"''\", 81), ('``', 55), (\"'m\", 9), ('data', 8), ('column', 7), ('columns', 5), ('dataframe', 5), (\"'i\", 5), (\"'for\", 5), ('mean', 5), ('function', 4), ('label', 4), ('new', 4), ('using', 4), ('use', 4), ('notimplementederror', 4), ('import', 4), ('review', 3), ('say', 3), ('supposed', 3), (\"'s\", 3), ('getting', 3), ('values', 3), ('wrong', 3), ('labels', 3), ('string', 3), ('part', 3), (\"'when\", 3), ('steps', 3), ('income', 3)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[('thanks', 24), (\"'how\", 23), ('string', 21), (\"'s\", 20), (\"n't\", 20), (\"'is\", 20), ('would', 18), ('assignment', 15), ('values', 14), ('says', 14), ('know', 13), (\"'what\", 12), ('data', 12), ('lecture', 12), (\"'when\", 11), (\"'i\", 11), ('major', 11), ('people', 11), ('error', 10), ('first', 10), ('correct', 10), ('sure', 9), (\"'can\", 9), ('``', 9), (\"''\", 9), ('class', 9), ('names', 8), ('dataframe', 8), ('us', 8), (\"'m\", 8)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[(\"n't\", 106), ('error', 92), ('code', 66), (\"'m\", 65), (\"'s\", 62), ('get', 56), ('wrong', 51), ('know', 45), ('tried', 43), ('help', 35), ('run', 35), (\"'i\", 34), (\"''\", 29), ('``', 28), ('getting', 28), ('work', 27), ('got', 26), ('still', 26), ('test', 26), ('anyone', 25), ('would', 25), ('import', 24), ('one', 22), ('something', 22), ('first', 22), ('cell', 21), ('assert', 21), ('print', 20), ('using', 20), ('however', 20)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[('--', 512), ('df', 125), ('code', 74), ('==', 72), ('``', 69), (\"''\", 68), ('error', 61), ('zip_code', 55), (\"'zip\", 54), (\"'m\", 50), (\"n't\", 48), (\"'income\", 47), ('value', 46), ('self', 46), ('...', 46), ('get', 45), (\"'i\", 45), ('zip_dict', 38), ('use', 36), ('list', 34), ('like', 31), ('df_zip', 31), ('population', 31), ('print', 30), ('key', 30), ('last', 29), (\"'age\", 29), ('index', 28), ('getting', 27), ('tried', 27)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[('project', 29), ('final', 17), ('group', 14), ('data', 12), ('github', 11), ('code', 11), ('repo', 10), ('would', 9), ('one', 9), ('file', 8), ('like', 7), (\"n't\", 7), ('work', 7), ('check', 6), ('proposal', 6), ('notebook', 6), ('show', 6), ('``', 6), (\"'m\", 5), ('wondering', 5), ('assignments', 5), ('us', 5), ('know', 5), ('need', 5), (\"''\", 5), (\"'ve\", 4), (\"'s\", 4), ('grade', 4), ('ted', 4), ('plot', 4)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[(\"'i\", 98), (\"n't\", 76), (\"'m\", 62), ('file', 39), ('could', 32), ('get', 31), ('sure', 31), ('error', 30), ('github', 29), ('anaconda', 29), (\"'ve\", 27), ('tried', 27), ('ca', 27), ('find', 26), ('jupyter', 25), ('know', 23), ('please', 22), ('using', 21), ('thank', 21), ('still', 20), ('push', 19), ('notebook', 19), ('python', 19), ('thanks', 18), ('open', 18), ('figure', 17), ('anyone', 16), ('git', 15), (\"'hi\", 15), ('getting', 15)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[(\"'m\", 61), ('getting', 44), ('error', 39), (\"'i\", 29), ('sure', 25), (\"n't\", 24), ('know', 23), ('anyone', 16), ('keep', 13), ('help', 10), ('wrong', 10), ('code', 10), ('like', 7), ('confused', 6), ('get', 6), (\"'when\", 6), (\"'hi\", 6), ('using', 6), ('take', 5), ('right', 5), ('thank', 5), ('would', 5), ('dataframe', 5), ('really', 5), (\"'s\", 5), (\"'ve\", 5), ('appreciated', 5), ('``', 4), ('function', 4), ('assert', 4)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[('data', 100), ('use', 100), (\"n't\", 95), ('code', 80), ('get', 77), ('part', 68), (\"''\", 68), ('``', 66), ('like', 60), (\"'s\", 56), ('one', 54), ('assignment', 52), ('make', 50), ('supposed', 50), (\"'for\", 46), (\"'m\", 44), ('know', 44), ('value', 43), ('would', 42), ('thanks', 40), ('values', 39), ('also', 37), ('income', 37), ('question', 36), ('using', 35), ('rows', 34), ('variable', 34), ('dataframe', 34), ('zip', 34), ('add', 33)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n",
      "[('grade', 4), ('assignment', 3), (\"'do\", 3), ('need', 3), ('class', 3), ('drop', 2), ('way', 2), ('website', 2), (\"'hey\", 1), ('wondering', 1), ('check', 1), ('see', 1), ('breakdown', 1), ('many', 1), ('points', 1), ('got', 1), ('question', 1), ('create', 1), ('countvectorizer', 1), ('rows', 1), (\"'year\", 1), ('invalid', 1), (\"'we\", 1), ('turn', 1), ('one', 1), ('week', 1), ('late', 1), ('25', 1), ('deduction', 1), ('right', 1)]\n",
      "\n",
      "-------------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for i in range(10):\n",
    "    word_tokens = word_tokenize(getStrings(df_piazza[df_piazza[\"label\"] == i][\"pBody\"]))\n",
    "    filtered_sentence = [w for w in word_tokens if not w in stop_words]\n",
    "    filtered_punct = [w for w in filtered_sentence if len(w) > 1]\n",
    "    #create a Counter object to count words\n",
    "    cnt = Counter(filtered_punct)\n",
    "    #save top 30 words from each label\n",
    "    top30pi_dict.append(cnt.most_common(30))\n",
    "    print(top30pi_dict[i])\n",
    "    print()\n",
    "    print(\"-------------------------------------------\")\n",
    "    print()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 212,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "qpiazza_dict = [df_piazza[df_piazza['label'] == i] for i in range(10)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "LABEL_TOPIC_MAP3 = {0 : 'office hours',                                   # if 'office' maps to 18\n",
    "                    1 : 'dataframes',                                     # if 'dataframe' maps to 5\n",
    "                    2 : 'assignment clarification',                       # if ''how' maps to 23\n",
    "                    3 : 'stuck on errors',                                # if 'error' maps to 92\n",
    "                    4 : 'decoding error statements',                      # if 'error' maps to 61\n",
    "                    5 : 'final project',                                  # if 'project maps to 29\n",
    "                    6 : 'external packages and github',                   # if 'github' maps to 29\n",
    "                    7 : 'debugging errors',                               # if 'error' maps to 39\n",
    "                    8 : 'data analysis',                                  # if 'data' maps to 100\n",
    "                    9 : 'grades'                                          # if 'grade' maps to 4\n",
    "                   }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 264,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['office hours', 'dataframes', 'assignment clarification', 'stuck on errors', 'decoding error statements', 'final project', 'external packages and github', 'debugging errors', 'data analysis', 'grades']\n"
     ]
    }
   ],
   "source": [
    "assign_topics3 = ['none']*10\n",
    "for i in range(10):\n",
    "    for j in range(10):\n",
    "        should_break = False;\n",
    "        for k,v in top30pi_dict[j]:\n",
    "            if k == 'office' and v == 18 and assign_topics3[j] == 'none':\n",
    "                assign_topics3[j] = LABEL_TOPIC_MAP3[0]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'dataframe' and v == 5 and assign_topics3[j] == 'none':\n",
    "                assign_topics3[j] = LABEL_TOPIC_MAP3[1]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == '\\'how' and v == 23 and assign_topics3[j] == 'none':\n",
    "                assign_topics3[j] = LABEL_TOPIC_MAP3[2]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'error' and v == 92 and assign_topics3[j] == 'none':\n",
    "                assign_topics3[j] = LABEL_TOPIC_MAP3[3]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'error' and v == 61 and assign_topics3[j] == 'none':\n",
    "                assign_topics3[j] = LABEL_TOPIC_MAP3[4]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'project' and v == 29 and assign_topics3[j] == 'none':\n",
    "                assign_topics3[j] = LABEL_TOPIC_MAP3[5]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'github' and v == 29 and assign_topics3[j] == 'none':\n",
    "                assign_topics3[j] = LABEL_TOPIC_MAP3[6]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'error' and v == 39 and assign_topics3[j] == 'none':\n",
    "                assign_topics3[j] = LABEL_TOPIC_MAP3[7]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'data' and v == 100 and assign_topics3[j] == 'none':\n",
    "                assign_topics3[j] = LABEL_TOPIC_MAP3[8]\n",
    "                should_break = True\n",
    "                break\n",
    "            elif k == 'grade' and v == 4 and assign_topics3[j] == 'none':\n",
    "                assign_topics3[j] = LABEL_TOPIC_MAP3[9]\n",
    "                should_break = True\n",
    "                break\n",
    "        if should_break:\n",
    "            break\n",
    "print(assign_topics3)\n",
    "           "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 265,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#sort each question inside each group by ascending order\n",
    "questions_pi_dict = [df_piazza[df_piazza['label'] == i] for i in range(10)]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 266,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "followup_pi = np.zeros(10)\n",
    "for i in range(10):\n",
    "    followup_pi[i] = questions_pi_dict[i]['followups'].sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 267,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a384f2668>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "color3 = [LABEL_COLOR_MAP3[l] for l in range(10)]\n",
    "fig, ax = plt.subplots()\n",
    "ind = np.arange(len(followup_pi))\n",
    "ax.barh(ind, followup_pi, alpha=0.5, color=color3)\n",
    "ax.set_yticks(ind)\n",
    "ax.set_yticklabels(assign_topics3)\n",
    "ax.set_xlabel('Total Number of Follow-ups')\n",
    "ax.set_title('Summation of Follow-ups in Each Topic')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 268,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#normalize data based on its total number of questions\n",
    "num_questions_pi = np.zeros(10)\n",
    "for i in range(10):\n",
    "    num_questions_pi[i] = len(questions_pi_dict[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 269,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a31059d68>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#total ratings\n",
    "fig, ax = plt.subplots()\n",
    "ind = np.arange(len(num_questions_pi))\n",
    "ax.barh(ind, num_questions_pi, alpha=0.5, color=color3)\n",
    "ax.set_yticks(ind)\n",
    "ax.set_yticklabels(assign_topics3)\n",
    "ax.set_xlabel('Total Number of Questions')\n",
    "ax.set_title('Number of Questions in Each Topic')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 270,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "norm_followups = np.zeros(10)\n",
    "for i in range(10):\n",
    "    norm_followups[i] = followup_pi[i]/num_questions_pi[i]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 271,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a2362a550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "ind = np.arange(len(norm_followups))\n",
    "ax.barh(ind, norm_followups, alpha=0.5, color=color3)\n",
    "ax.set_yticks(ind)\n",
    "ax.set_yticklabels(assign_topics3)\n",
    "ax.set_xlabel('Normalized Follow-ups')\n",
    "ax.set_title('Normalized Number of Follow-ups Based on Question Population')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 4 - Comparing All Datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 272,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a2736cd68>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#normalized ratings ratings\n",
    "fig, ax = plt.subplots()\n",
    "ind = np.arange(len(norm_scores_py))\n",
    "ax.barh(ind, norm_scores_py, alpha=0.5, color=color1)\n",
    "ax.set_yticks(ind)\n",
    "ax.set_yticklabels(assign_topics1)\n",
    "ax.set_xlabel('Normalized Score')\n",
    "ax.set_title('Normalized Score Based on Question Population')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 273,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a348af4a8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "ind = np.arange(len(norm_scores_st))\n",
    "ax.barh(ind, norm_scores_st, alpha=0.5, color=color2)\n",
    "ax.set_yticks(ind)\n",
    "ax.set_yticklabels(assign_topics2)\n",
    "ax.set_xlabel('Normalized Score')\n",
    "ax.set_title('Normalized Score Based on Question Population')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 274,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a345d94a8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "ind = np.arange(len(norm_followups))\n",
    "ax.barh(ind, norm_followups, alpha=0.5, color=color3)\n",
    "ax.set_yticks(ind)\n",
    "ax.set_yticklabels(assign_topics3)\n",
    "ax.set_xlabel('Normalized Follow-ups')\n",
    "ax.set_title('Normalized Number of Follow-ups Based on Question Population')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Ethics and Privacy\n",
    "\n",
    "For the StackOverflow data, the datasets are already publically available on Kaggle.com and seem to have been anonymized following the Safe Harbour Method. StackOverflow probably has little to no biases because of the large number of active users. It also contains questions from people from all different backgrounds of programming experience, so we expected to see a large range of questions that were asked.\n",
    "\n",
    "As for the Piazza data, we needed to take a few more precautionary steps. First, we had to remove any identifying properties from all of the questions, such as student names or identifying labels in the question bodies themselves. This was a painstaking process that I'm sure could have been implemented more efficiently, but involved manually going through all of the unique words within the dataset and removing words that were reminiscient of names, locations, dates, etc. \n",
    "\n",
    "This Piazza dataset has demonstrated itself to also contain biases towards data science university students in UCSD who are generally active on Piazza; thus, our results from comparing this dataset with the StackOverflow dataset cannot be used to generalize all university students. Many of the questions were in fact directed towards course-specific logistics, including programming assignments and projects. For these questions, there may have been a potential academic integrity violation if the post contains answers or solutions to any course-related work. Therefore our results do not display the direct questions themselves if they contain specific information regarding course material, and we have ensured that they have been removed. In general, we hope to ensure that our data analysis does not violate privacy conditions of both the student and of the course."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Discussion\n",
    "---\n",
    "With all of our findings, it has become clear that a general consensus for the most common struggles for python programmers are generally those related to more beginner concepts. As stated earlier, this may probably be attributed to the general userbase of StackOverflow, which tends to contain more beginner programmers than highly experienced ones. However, this is likely not the only driving factor. Many of the questions that we reviewed were those with specific coding questions, from users who were asking others to either debug their code or provide a simpler implementation for code that they had already written. These specific, tailored, questions are likely only asked by beginner programmers, as more advanced programmers would be able to more clearly discern simply issues and instead, face more conceptual errors that do not need exact examples in order to be relevant.\n",
    "\n",
    "That said, we did also notice that there were many conceptual questions that seemed to be asked by beginners as well, and this demonstrates a genuine interest to learn and better oneself in the development of their own coding expertise. Without understanding the basic concepts underneath, programmers will lack a solid foundation on which they can build their knowledge, and as a result, anything more advanced or more abstract that they learn later on will crumble. Thus, StackOverflow has proved to be a good resource for those who truly want to learn the science behind programming in Python, and not just simple pointers for coding.\n",
    "\n",
    "In regards to Piazza, we saw a heavy skew towards course-related questions, including questions that asked about things such as grades, office hours, or clarifications on assignments and the final project. However, we did also see other topics that were more Python-based, including questions about dataframes. This is probably expected because as this is a data science class, we have worked a lot this quarter with dataframes in particular. By and large, the biggest grouping of questions were those related to errors, and in general, we saw 3 different groupings: questions about understanding error statements, questions about initial debugging steps, and questions by students who had already debugged errors but were unable to find a solution.\n",
    "\n",
    "In the end, using all 3 of our data sets and data analyses, it seems that the most common difficulties that programmers face are basic concepts about the very language that they are using. In order to remedy this, an immediate solution may be to provide programmers with more concept-based curriculum so that they are truly able to understand the different data structures and libraries they are using before jumping straight in based on pure, surface-understanding examples. From Piazza, and for COGS 108 specifically, we saw that most of the questions were based on errors. It may be useful, then, to accumulate a list of most common errors when working with external libraries such as pandas or numpy, and provide a handy FAQ for students to refer to as they work on their assignments. This would greatly reduce the time spent on debugging specific errors, which may be able to help students better understand the overall concept of what they are doing, rather than the specific coding requirements. This is simply a suggestion; feel free to use our data as you please :)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Moving Forward\n",
    "---\n",
    "Having completed this assignment, we have learned a lot. Initially, we planned on using Levenshtein distances in order to complete our analyses, but after discussing with the TAs, we found much simpler, and more useful mechanisms for our specific data analysis. However, looking back, there are other things that we definitely could have done with more efficiency. For example, rather than using kmeans to discover clusters, we could have used LDA in order to glean a more useful and potentially more accurate clustering of questions. Although we may never know without trying, it seems that LDA may have worked better for us. Furthermore, we could have also benefitted from a better method for giving labels to the clusterings that were generated by kmeans. As it stands, they were quite subjective, but we are not sure if a better, more objective method exists.\n",
    "\n",
    "As for our project itself, we hope that the data we have analyzed will be useful for not only Prof. Voytek as he continues to develop this curriculum, but perhaps our analysis strategies could also be applied to Piazza datasets from other classes. As data science emerges as a new field of study at UCSD, our algorithms could potentially be applied to other DSC classes at UCSD, which may be especially useful since many of the data science professors are also newly developing their curricula.\n",
    "\n",
    "Finally, from this assignment, we have learned a lot about the practice of data science. We have experienced firsthand just how difficult it might be to clean data and effectively analyze it in ways that make sense and are useful for the real world. Not only this, but we've learned better how to work as a team, as we managed to overcome obstacles and disagreements to reach this point. Thanks for a great quarter Professor Voytek and all the TAs!"
   ]
  }
 ],
 "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}