Untitled.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Connect to POSTGRE and create tables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import psycopg2\n",
    "from psycopg2.extensions import ISOLATION_LEVEL_AUTOCOMMIT # <-- ADD THIS LINE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from configparser import ConfigParser\n",
    "def config(filename='database.ini', section='postgresql'):\n",
    "    # create a parser\n",
    "    parser = ConfigParser()\n",
    "    # read config file\n",
    "    parser.read(filename)\n",
    " \n",
    "    # get section, default to postgresql\n",
    "    db = {}\n",
    "    if parser.has_section(section):\n",
    "        params = parser.items(section)\n",
    "        for param in params:\n",
    "            db[param[0]] = param[1]\n",
    "    else:\n",
    "        raise Exception('Section {0} not found in the {1} file'.format(section, filename))\n",
    " \n",
    "    return db"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def connect():\n",
    "    \"\"\" Connect to the PostgreSQL database server returns connection \"\"\"\n",
    "    conn = None\n",
    "    try:\n",
    "        # read connection parameters\n",
    "        params = config()\n",
    " \n",
    "        # connect to the PostgreSQL server\n",
    "        print('Connecting to the PostgreSQL database...')\n",
    "        conn = psycopg2.connect(**params)\n",
    " \n",
    "        # create a cursor\n",
    "        cur = conn.cursor()\n",
    "        \n",
    " # execute a statement\n",
    "        print('PostgreSQL database version:')\n",
    "        cur.execute('SELECT version()')\n",
    " \n",
    "        # display the PostgreSQL database server version\n",
    "        db_version = cur.fetchone()\n",
    "        print(db_version)\n",
    "       \n",
    "     # close the communication with the PostgreSQL\n",
    "        cur.close()\n",
    "    except (Exception, psycopg2.DatabaseError) as error:\n",
    "        print(error)\n",
    "    finally:\n",
    "        return conn;"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def create_tables(conn):\n",
    "    \"\"\" create tables in the PostgreSQL database and load data from csv\"\"\"\n",
    "    commands = (\n",
    "        \"\"\"\n",
    "        CREATE TABLE adult (\n",
    "            age INTEGER,\n",
    "            workclass VARCHAR,\n",
    "            fnlwgt integer,\n",
    "            education varchar,\n",
    "            education_num integer,\n",
    "            marital_status varchar,\n",
    "            occupation varchar,\n",
    "            relationship varchar,\n",
    "            race varchar,\n",
    "            sex varchar,\n",
    "            capital_gain integer,\n",
    "            capital_loss integer,\n",
    "            hours_per_week integer,\n",
    "            native_country varchar,\n",
    "            data_class varchar\n",
    "        )\n",
    "        \"\"\",\n",
    "        \"\"\"\n",
    "        copy public.adult (\n",
    "            age,\n",
    "            workclass,\n",
    "            fnlwgt,\n",
    "            education,\n",
    "            education_num,\n",
    "            marital_status,\n",
    "            occupation,\n",
    "            relationship,\n",
    "            race,\n",
    "            sex,\n",
    "            capital_gain,\n",
    "            capital_loss,\n",
    "            hours_per_week,\n",
    "            native_country,\n",
    "            data_class\n",
    "        ) \n",
    "        FROM 'D:/UMass/Spring 18/645/SeeDB/adult_data.csv' \n",
    "        DELIMITER ',' CSV ENCODING 'UTF8';\n",
    "        \"\"\")\n",
    "    conn = None\n",
    "    try:\n",
    "        # read the connection parameters\n",
    "        params = config()\n",
    "        # connect to the PostgreSQL server\n",
    "#         conn = psycopg2.connect(**params)\n",
    "        cur = conn.cursor()\n",
    "        # create table one by one\n",
    "        for command in commands:\n",
    "            cur.execute(command)\n",
    "        # close communication with the PostgreSQL database server\n",
    "        cur.close()\n",
    "        # commit the changes\n",
    "        conn.commit()\n",
    "    except (Exception, psycopg2.DatabaseError) as error:\n",
    "        print(error)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def create_views(conn):\n",
    "    \"\"\" create views of married and unmarried, split into 10 subsets for phases\"\"\"\n",
    "    commands = (\n",
    "        \"\"\"\n",
    "        CREATE VIEW Married \n",
    "        AS SELECT * \n",
    "        FROM adult\n",
    "        WHERE marital_status IN (' Married-AF-spouse', ' Married-civ-spouse', ' Married-spouse-absent', ' Separated')\n",
    "        \"\"\",\n",
    "        \"\"\"\n",
    "        CREATE VIEW Unmarried \n",
    "        AS SELECT * \n",
    "        FROM adult\n",
    "        WHERE marital_status IN (' Never-married', ' Widowed', ' Divorced')\n",
    "        \"\"\"\n",
    "    )\n",
    "    conn = None\n",
    "    try:\n",
    "        # read the connection parameters\n",
    "        params = config()\n",
    "        # connect to the PostgreSQL server\n",
    "#         conn = psycopg2.connect(**params)\n",
    "        cur = conn.cursor()\n",
    "        # create table one by one\n",
    "        for command in commands:\n",
    "            cur.execute(command)\n",
    "        # close communication with the PostgreSQL database server\n",
    "        cur.close()\n",
    "        # commit the changes\n",
    "        conn.commit()\n",
    "    except (Exception, psycopg2.DatabaseError) as error:\n",
    "        print(error)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def close_connection(conn):\n",
    "    \"\"\" Close postgres connection\"\"\"\n",
    "    try:\n",
    "        if conn is not None:\n",
    "            conn.close()\n",
    "    except (Exception, psycopg2.DatabaseError) as error:\n",
    "        print(error)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "a = [\"workclass\", \"education\", \"occupation\", \"relationship\", \"race\", \"sex\", \"native_country\",\"data_class\"]\n",
    "m = [\"age\", \"fnlwgt\", \"hours_per_week\", \"capital_gain\", \"capital_loss\"]\n",
    "f = [\"avg\", \"sum\", \"min\", \"max\", \"count\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "tuples = [(x, m, f) for x in a]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Admin\\Anaconda2\\lib\\site-packages\\ipykernel_launcher.py:6: DeprecationWarning: You passed a bytestring as `filenames`. This will not work on Python 3. Use `cp.read_file()` or switch to using Unicode strings across the board.\n",
      "  \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Connecting to the PostgreSQL database...\n",
      "PostgreSQL database version:\n",
      "('PostgreSQL 10.2, compiled by Visual C++ build 1800, 64-bit',)\n",
      "('Number of rows in table ', 29933L)\n",
      "('Number of phases ', 10)\n",
      "('Size of each phase ', 2994L)\n"
     ]
    }
   ],
   "source": [
    "conn = connect()\n",
    "cur = conn.cursor()\n",
    "\n",
    "#find total rows in db then split into 10 phase - size of each phase\n",
    "cur.execute(\"SELECT COUNT(age) FROM adult;\")\n",
    "num_of_rows = cur.fetchall()[0][0];\n",
    "print(\"Number of rows in table \", num_of_rows)\n",
    "phase = 10\n",
    "size = num_of_rows/phase+1;\n",
    "print(\"Number of phases \", phase)\n",
    "print(\"Size of each phase \", size)\n",
    "\n",
    "#query from adult\n",
    "cur.execute(\"SELECT * FROM adult;\")\n",
    "cur2 = conn.cursor()\n",
    "# for i in ['a_'+`i` for i in range(10)]:\n",
    "for i in range(10):\n",
    "    create_tbl = \"CREATE TEMP TABLE a_\" + str(i) + \" (age INTEGER,workclass VARCHAR,fnlwgt integer,education varchar, education_num integer, marital_status varchar, occupation varchar, relationship varchar, race varchar, sex varchar, capital_gain integer, capital_loss integer, hours_per_week integer,native_country varchar,data_class varchar);\"\n",
    "    cur2.execute(create_tbl)\n",
    "    commands = \"INSERT INTO a_\"+ str(i) +\" VALUES(%s, %s, %s,%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s);\"\n",
    "    results = cur.fetchmany(size) #fetching every size-rows from adult\n",
    "    for result in results:\n",
    "        #add to view i \n",
    "#         print(result)\n",
    "        cur2.execute(commands,result)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#closing connection or cur2 will result in dropping temp tables!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Pruning Optimization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import normalize\n",
    "import scipy\n",
    "import numpy as np\n",
    "view_distances = {}\n",
    "num_of_phases = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def normalization(row):\n",
    "    \"\"\"\n",
    "    Function to normalize the f(m) column\n",
    "    :param rows:\n",
    "    :return:\n",
    "    \"\"\"\n",
    "    new_row = []\n",
    "    for val in row:\n",
    "        if(val is not None):\n",
    "            new_row.append(val)\n",
    "    new_row = np.array(new_row)\n",
    "    return normalize(new_row[:,np.newaxis], axis=0).ravel()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def compute_KL_divergence(target,reference):\n",
    "    if(len(target)==len(reference)):\n",
    "        return scipy.stats.entropy(target,reference)\n",
    "    else:\n",
    "        return 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def execute_query(conn,query):  \n",
    "    cur = conn.cursor()\n",
    "    cur.execute(query)\n",
    "    rows = cur.fetchall()\n",
    "    cur.close()    \n",
    "    return np.array(rows).T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import math"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def get_mean(view_distances):\n",
    "    mean_dict = {}\n",
    "    my_values = []\n",
    "    for key, value in view_distances.items():\n",
    "        for v in value:\n",
    "            if not math.isinf(v):\n",
    "                my_values.append(v)\n",
    "        mean_dict[key] = (np.mean(my_values),np.std(my_values))\n",
    "    return mean_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def get_confidence_interval(m, mu):\n",
    "    delta = 0.05\n",
    "    n = phase\n",
    "    term1 = 1 - (m-1)/n\n",
    "    #print m\n",
    "    #print math.log(m)\n",
    "    term2 = 2 * math.log(math.log(m))\n",
    "    term3 = math.log((3.14*3.14)/(3*delta))\n",
    "    numerator = term1 * (term2 + term3)\n",
    "    denominator = 2 * m\n",
    "    epsilon = math.sqrt(numerator/denominator)\n",
    "    confidence_interval = (mu-epsilon, mu+epsilon)\n",
    "    return confidence_interval\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def prune(view_distances, m, phase):\n",
    "    views_CI_dict = {}\n",
    "    ci_dict = {}\n",
    "    prunedViews = []\n",
    "    mean_dict = get_mean(view_distances)\n",
    "    for key, value in mean_dict.items():\n",
    "        #print mean_dict[key]\n",
    "        ci_dict[key] = get_confidence_interval(m, value[0])\n",
    "        #print type(get_confidence_interval(m, mean_dict[key][0]))\n",
    "        #print \"aaaaaa\"\n",
    "        #ci_dict[key] = scipy.stats.norm.interval(0.95, loc=value[0], scale=value[1])\n",
    "        #print type(ci_dict[key])\n",
    "    ci_dict_copy = ci_dict.copy()\n",
    "    viewItems = list(ci_dict_copy)\n",
    "    viewItems.sort(key = lambda x:x[1][1])\n",
    "    topFiveViews = viewItems[-5:]\n",
    "    topFiveViews.sort(key = lambda x:x[1][0])\n",
    "    lowestLowerBound =topFiveViews[:1][0][1][0]\n",
    "    key_topFiveViews =[k for k in topFiveViews]\n",
    "    for key, value in ci_dict_copy.items():\n",
    "        if(value[0] < lowestLowerBound and key not in key_topFiveViews):\n",
    "            ci_dict_copy.pop(key)\n",
    "            view_distances.pop(key)\n",
    "    return view_distances"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def share_opt(conn):\n",
    "    view_distances = {}\n",
    "    # Phase 0\n",
    "    for a,measures,functions in tuples:\n",
    "        #print a\n",
    "         for m in measures:\n",
    "            for f in functions:\n",
    "                #view_distances[(a,m,f)] = []\n",
    "                target_command =\"select {},{}({}) from a_{} where marital_status IN (' Married-AF-spouse', ' Married-civ-spouse', ' Married-spouse-absent', ' Separated') group  by {}\".format(a,f,m,0,a)\n",
    "                reference_command = \"select {},{}({}) from a_{} WHERE marital_status IN (' Never-married', ' Widowed', ' Divorced') group  by {}\".format(a,f,m,0,a)\n",
    "                target_array = execute_query(conn,target_command)\n",
    "                reference_array = execute_query(conn,reference_command)\n",
    "                view_distances[(a,m,f)]= [compute_KL_divergence(normalization(target_array[1]), normalization(reference_array[1]))]\n",
    "    #view_distances = prune(view_distances, 1, phase)\n",
    "    for i in range(1, phase):\n",
    "        for key, value in view_distances.items():\n",
    "                #print key\n",
    "                #print value\n",
    "                a = key[0]\n",
    "                m = key[1]\n",
    "                f = key[2]\n",
    "                target_command =\"select {},{}({}) from a_{} where marital_status IN (' Married-AF-spouse', ' Married-civ-spouse', ' Married-spouse-absent', ' Separated') group  by {}\".format(a,f,m,i,a)\n",
    "                reference_command = \"select {},{}({}) from a_{} WHERE marital_status IN (' Never-married', ' Widowed', ' Divorced') group  by {}\".format(a,f,m,i,a)\n",
    "                reference_array = execute_query(conn,reference_command)\n",
    "                #print compute_KL_divergence(normalization(target_array[1]), normalization(reference_array[1]))\n",
    "                view_distances[(a,m,f)].append(compute_KL_divergence(normalization(target_array[1]), normalization(reference_array[1])))        \n",
    "        if(i!=phase-1):\n",
    "            view_distances = prune(view_distances, i+1, phase)\n",
    "    return view_distances\n",
    "        \n",
    "        \n",
    "        \n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Admin\\Anaconda2\\lib\\site-packages\\scipy\\stats\\_distn_infrastructure.py:2505: RuntimeWarning: invalid value encountered in true_divide\n",
      "  pk = 1.0*pk / np.sum(pk, axis=0)\n",
      "C:\\Users\\Admin\\Anaconda2\\lib\\site-packages\\scipy\\stats\\_distn_infrastructure.py:2512: RuntimeWarning: invalid value encountered in true_divide\n",
      "  qk = 1.0*qk / np.sum(qk, axis=0)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('race', 'hours_per_week', 'avg')\n",
      "[0.0017189504469477747, 0, 0, 0, 0, 0, 0, 0, 0, 0]\n",
      "('data_class', 'hours_per_week', 'max')\n",
      "[0.0085249160277568806, 0.00055039133792055342, 0.0041892080024039519, 0.00054232321500322783, 0.0041892080024039519, 0.00054232321500322783, 0.00054232321500322783, 0.0041892080024039519, 0.009671128861736239, 0.009671128861736239]\n",
      "('race', 'hours_per_week', 'count')\n",
      "[0.018818956934680629, 0, 0, 0, 0, 0, 0, 0, 0, 0]\n",
      "('relationship', 'hours_per_week', 'avg')\n",
      "[0, 0, 0, 0, 0, 0, 0, 0, 0, 0]\n",
      "('education', 'hours_per_week', 'min')\n",
      "[0.69494322461374769, 0, 0, 0, 0, 0, 0, 0, 0, 0]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Admin\\Anaconda2\\lib\\site-packages\\ipykernel_launcher.py:21: DeprecationWarning: unorderable dtypes; returning scalar but in the future this will be an error\n"
     ]
    }
   ],
   "source": [
    "top5_views = share_opt(conn)\n",
    "for key , value in top5_views.items():\n",
    "    print key\n",
    "    print value\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import operator\n",
    "A = {'a':(10,11), 'b':(843,999), 'c': (39,99), 'd':(40,63), 'e':(55,79), 'f':(9,10), 'g': (22,55), 'x' : (10, 99999)}\n",
    "copyA = {'a':(10,11), 'b':(843,999), 'c': (39,99), 'd':(40,63), 'e':(55,79), 'f':(9,10), 'g': (22,55), 'x' : (10, 99999)}\n",
    "newA = dict(sorted(A.iteritems(), key=operator.itemgetter(1), reverse=True)[:5])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[('f', (9, 10)), ('a', (10, 11)), ('g', (22, 55)), ('d', (40, 63)), ('e', (55, 79)), ('c', (39, 99)), ('b', (843, 999)), ('x', (10, 99999))]\n",
      "[('x', (10, 99999)), ('c', (39, 99)), ('d', (40, 63)), ('e', (55, 79)), ('b', (843, 999))]\n"
     ]
    }
   ],
   "source": [
    "b = copyA.items()\n",
    "b.sort(key = lambda x:x[1][1])\n",
    "print b\n",
    "top5 = b[-5:]\n",
    "top5.sort(key = lambda x:x[1][0])\n",
    "print(top5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[('x', (10, 99999)), ('c', (39, 99)), ('d', (40, 63)), ('e', (55, 79)), ('b', (843, 999))]\n",
      "[('x', (10, 99999))]\n",
      "10\n",
      "10\n",
      "22\n",
      "9\n",
      "{'a': (10, 11), 'c': (39, 99), 'b': (843, 999), 'e': (55, 79), 'd': (40, 63), 'g': (22, 55), 'f': (9, 10), 'x': (10, 99999)}\n",
      "{'a': (10, 11), 'g': (22, 55)}\n"
     ]
    }
   ],
   "source": [
    "prunedViews = []\n",
    "print(top5)\n",
    "print(top5[:1])\n",
    "lowestLowerBound = top5[:1][0][1][0]\n",
    "print lowestLowerBound\n",
    "key_top5 ={k[0] for k in top5}\n",
    "for key in key_top5:\n",
    "    copyA.pop(key, None)\n",
    "for key, value in copyA.items():\n",
    "    print value[0]\n",
    "    if(value[0] < lowestLowerBound):\n",
    "        copyA.pop(key)\n",
    "print(A)\n",
    "print(copyA)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualize Top 5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "#import matplotlib.pyplot as plt2\n",
    "def draw_viz(conn, target, reference ,title, y_axis, x_axis):\n",
    "#     print(\"Conn \", conn)\n",
    "#     print(\"Target \", target)\n",
    "#     print(\"ref \", reference)\n",
    "#     print(\"title \", title)\n",
    "#     print(\"y axis \", y_axis)\n",
    "#     print(\"x_axis \", x_axis)\n",
    "    # Setting the positions and width for the bars\n",
    "    width = 0.25 \n",
    "\n",
    "    target_values = list(target[1])\n",
    "    ref_values = list(reference[1])\n",
    "    if(len(target[1]) < len(reference[1])):\n",
    "        diff = len(reference[1]) - len(target[1])  \n",
    "        for i in range(diff):\n",
    "            target_values.append(0)\n",
    "    else:\n",
    "        diff = len(target[1]) - len(reference[1])\n",
    "        for i in range(diff):\n",
    "            ref_values.append(0)\n",
    "    \n",
    "    pos = list(range(len(target_values)))\n",
    "    # Plotting the bars\n",
    "    fig, ax = plt.subplots()\n",
    "\n",
    "    # Create a bar with pre_score data,\n",
    "    # in position pos,\n",
    "    plt.bar(pos, \n",
    "            #using df['pre_score'] data,\n",
    "            target_values, \n",
    "            # of width\n",
    "            width, \n",
    "            # with alpha 0.5\n",
    "            alpha=0.5, \n",
    "            # with color\n",
    "            color='blue') \n",
    "\n",
    "    # Create a bar with mid_score data,\n",
    "    # in position pos + some width buffer,\n",
    "    plt.bar([p + width for p in pos], \n",
    "            #using df['mid_score'] data,\n",
    "            ref_values,\n",
    "            # of width\n",
    "            width, \n",
    "            # with alpha 0.5\n",
    "            alpha=0.5, \n",
    "            # with color\n",
    "            color='lightblue') \n",
    "\n",
    "\n",
    "    # Set the y axis label\n",
    "    ax.set_ylabel(y_axis)\n",
    "\n",
    "    # Set the chart's title\n",
    "    ax.set_title(title)\n",
    "\n",
    "    # Set the position of the x ticks\n",
    "    ax.set_xticks([p + 1.5 * width for p in pos])\n",
    "\n",
    "    # Set the labels for the x ticks\n",
    "    ax.set_xticklabels(target[0], rotation =\"vertical\")\n",
    "\n",
    "    # Setting the x-axis and y-axis limits\n",
    "    plt.xlim(min(pos)-width, max(pos)+width*4)\n",
    "#     plt.ylim([0,max(max(target_values), max(ref_values))] )\n",
    "\n",
    "    # Adding the legend and showing the plot\n",
    "    plt.legend(['Married', 'Unmarried'], loc='upper left')\n",
    "    plt.grid()\n",
    "    plt.savefig(\"plot_\"+x_axis+\"_\"+y_axis+\".png\", bbox_inches='tight')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYMAAAFmCAYAAACP7fYjAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3XmclXXd//HXG0RGYEBR4kYxcVdUcEFxyQLJJXfNLDPD\n8r6tX4uZ0WJp4tKdlXdlq2EW4kaZZqZWkoJYWurIIoqGmhpKqeAyyC6f3x/XNeNhnGHmbNfFnPN+\nPh7nMefazvl8j3g+57tc368iAjMzq2898g7AzMzy52RgZmZOBmZm5mRgZmY4GZiZGU4GZmaGk4F1\nM5LeKWmppJ4lXDtMUkjaqBqxdWfpZ7pdideGpB0qHZNly8nAyibpGUmrJG3RZv+s9ItiWKXeKyKe\ni4h+EfFmpV6zkiRNlnRJm30bfBJKP9On847D8uNkYJXyT+CUlg1JewB9Sn2x9r44N+Qv0+7An6mt\nj5OBVco1wEcLtscDUwpPkHRUWlt4XdK/JE0sONby6/kMSc8Bd3eyb6P0ugGSrpK0SNLzki5paUKS\n1FPSZZJelvQ0cFRHwUv6sqTftNl3uaQfpM9Pl/S0pGZJ/5R0aqkfVFqTmiBprqTXJP1KUkN6bIyk\nhZK+JOnFtFzHSzpS0j8kLZH01YLX2k/S/ZJeTc/9kaSNC46HpE9LWgAs6GTfDunz3unn9pyk/0i6\nQtImBa/5xfS9XpD08VI/B9vARIQffpT1AJ4B3gs8AewK9AQWAtsAAQxLzxsD7EHyI2QE8B/g+PTY\nsPTcKUBfYJNO9m2UXvdb4Gfp8XcADwCfSI99Engc2BoYCEwvvLZNGbYBlgGN6XZPYBGwf/rarwM7\np8eGALt18FlMBi5ps69tzM+kcW6ZxjUf+GTBZ7QG+DrQC/gf4CXgeqAR2A1YDmybnr9PGuNG6fvM\nB84ueO8ApqXvs0kn+3ZIn38PuDU93gj8HvhmeuyI9L/b7unncn3htX5030fuAfjR/R+8lQzOA76Z\nfmFMS7+gWpNBO9d9H/he+rzlC3O7guPr27cRMBhY2fKFlh4/BZiePr+75Us23T6so2SQHv8L8NH0\n+aHAU+nzvsCrwPsL36uD1+hqMvhIwfFvA1ekz8ekX/Y90+3G9NrRBec3kSbRdt7/bOC3BdsBHNLm\nnI727QAIeAPYvuDYAcA/0+e/AC4tOLaTk0FtPNxeaJV0DTAT2JY2TUQAkkYDl5L8qtwY6A3c2Oa0\nf7Xzuu3tg+TXfC9gkaSWfT0Kzt+yzbXPdhL/9STJZArw4XSbiHhD0geBCcBVkv4KfCEiHm/nNdak\nMRXqBaxNHy3+XfB8WRpri8XxVgf58vTvfwqOLwf6AUjaCfguMIqkj2YjkmRRqJjPdFD6Ok0Fn6lI\nakqkcRa+fmefqXUT7jOwiomIZ0k6ko8Ebm7nlOtJmh+2jogBwBUkXzTrvEx7L93BW/6LpGawRURs\nmj76R8Ru6fFFJE1ELd7ZSRFuBMZIGgqckMabBBDxp4g4lKSJ6HHgyg5e4zmSmkChbYF/RcTat59e\ntp+m8ewYEf2Br1LeZ/oySbLZreAzHRAR/dLjxX6m1k04GVilnUHSBPFGO8cagSURsULSfiS/vksW\nEYuAO4H/k9RfUg9J20t6T3rKr4GzJA2VtBnwlU5e7yVgBvBLkmaR+QCSBks6TlJfkuSzlHV/5Re6\nCThK0mFpB/aWJM1nU8sp63o0kvRnLJW0C/D/ynmxNGFdCXxP0jsAJG0l6fD0lF8Dp0saLqkPcEE5\n72cbDicDq6iIeCoiHurg8KeAiyQ1k3SQ/roCb/lRkianx4BXgN+Q/HqH5EvtT8Ac4GHar620dT1J\n/8f1Bft6AOcALwBLgPfQwZduRDxK0tT0zfTc+4G/AxcWUaZiTCBJqs0k5f1VBV7zy8CTwN8kvQ78\nGdgZICL+QNLXc3d6zt0VeD/bACjCi9uYmdU71wzMzMzJwMzMnAzMzAwnAzMzg+5z09kWW2wRw4YN\nyzuMDr3xxhv07ds37zByUc9lB5ff5d+wy9/U1PRyRAzq7LxukwyGDRvGQw91NGIxfzNmzGDMmDF5\nh5GLei47uPwu/4ZdfkldukvczURmZuZkYGZmTgZmZkY36jNoz+rVq1m4cCErVqzIOxQGDBjA/Pnz\n8w6johoaGhg6dCi9erWdhNPMak23TgYLFy6ksbGRYcOGUTDdbi6am5tpbGzMNYZKiggWL17MwoUL\n2XbbbfMOx8yqrFs3E61YsYLNN98890RQiySx+eabbxC1LjOrvm6dDAAngiryZ2tWP7p9MjAzs/J1\n6z6DtiZOzP71JHHqqafy05/+FIA1a9YwZMgQRo8ezW233VbW+x944IHcd999XT5/4sSJ9OvXjwkT\nJpT1vmZWf2oqGeShb9++zJs3j+XLl9PY2Mi0adPYaqutinqNNWvWsNFGG71tu5hE8MIL0NwMa9cm\nz4u15Zadn2NmtcvNRBVw5JFH8qc//QmAG264gVNOOaX12AMPPMABBxzAXnvtxYEHHsgTTzwBwOTJ\nkzn22GM55JBDGDduHDNmzODggw/m2GOPZfjw4QD069ev9XW+853vsO+++zJixAguuOCtlQa/8Y1v\nsNNOO3H88e/iqaeeyKK4ZlaDnAwq4EMf+hA33XQTK1asYO7cuYwePbr12C677MK9997LrFmzuOii\ni/jqV7/aeuzhhx/mN7/5Dffcc0/r9uWXX84//vGPdV7/zjvvZMGCBTzwwAPMnj2bpqYmZs6cSVNT\nE1OnTmX27Nlcc80dzJnzYDYFNrOa42aiChgxYgTPPfccN9xwA0ceeeQ6x1577TXGjx/PggULkMTq\n1atbjx166KEMHDiwdXu//fZrd0z/nXfeyZ133slee+0FwNKlS1mwYAHNzc2ccMIJ9OnTh8ZGOPTQ\nY6tUQjOrda4ZVMj73vc+JkyYsE4TEcD555/P2LFjmTdvHr///e/XGbffdtrbjqbBjQjOPfdcZs+e\nzezZs3nyySc544wzKl8IM6tbrhlUyGmnncbgwYPZY489mDFjRuv+1157rbVDefLkySW99uGHH875\n55/PqaeeSr9+/Xj++efp1asX7373uzn99NM599xzWbp0DdOm/Z7TTvtEBUpj1n2UO4qw0qMQu6ua\nSgZ5/kfdaqutOOuss962/0tf+hLjx4/nkksu4aijjirptQ877DDmz5/PAQccACQdy9deey177703\nH/zgBxk5ciSbbvoO9txz37LKYGb1SxGRdwxdMmrUqGi7uM38+fPZddddc4poXXnPTVTKcNJCHQ0t\n7cpnvKEv7lFtLn++5c+7ZpB3+TsjqSkiRnV2Xs3UDKr1ZWhmVg9qJhmY1au8fxlbbfBoIjMzc83A\nrN499nJz3iHYBsA1AzMzczIwM7Maayb618rSq7uvvvz2fcO36Hyo6DPPPMPRRx/N/fff37ov76mk\n58x5iN/8ZgoXX/yDLl8zZswYLrvsMkaN6nQEmpnVoJpKBvWovemvR44cxciR/lI3s65zMqiiMWPG\nMHr0aKZPn86rr77KVVddxcEHH8zkyZO55ZZbeOONN1iwYAETJkxg1apVXHPNNfTu3Zs77riDgQMH\ncuWVVzJp0iRWrVrFDjvswDXXXEOfPn04/fTTaWhoYNasWRx00EH079+fRx55iueee5qttnonH/nI\nJ7jiisuYMuU2li17g/PO+yxPPDGP1atX84UvTOTww49j+fLlnHPOx3jssTnssMMuLF++PO+Py8xy\nlEmfgaSekmZJui3dHihpmqQF6d/NsogjD2vWrOGBBx7g+9//PhdeeGHr/nnz5nHzzTfz4IMP8rWv\nfY0+ffowa9YsDjjgAKZMmQLAiSeeyIMPPsicOXPYddddueqqq1qvX7hwIffddx/f/e53AViw4DGm\nTv0zP/nJDeu8/+WXf4ODDjqE229/gBtvnM7FF3+RZcveYMqUn7LJJn245575fOELF9LU1JTBp2Fm\nG6qsOpA/B8wv2P4KcFdE7AjclW53Sx0tGt+y/8QTTwRgn3324Zlnnmk9PnbsWBobGxk0aBADBgzg\nmGOOAWCPPfZoPW/evHkcfPDB7LHHHlx33XU8+uijrdd/4AMfoGfPnq3bhx12LJtsssnb4pg5805+\n/ONLOfTQPTnppDGsXLmC559/jr//fSYnnvgRAIYPH8GIESNK/xDMrNurejORpKHAUcA3gHPS3ccB\nY9LnVwMzgC9XO5Zq2HzzzXnllVfW2bdkyZLWdQl69+4NQM+ePVmzZk3rOS37AXr06NG63aNHj9bz\nTj/9dG655RZGjhzJ5MmT15kNte101336dDz99aRJN7HDDjuXWEIzqwdZ9Bl8H/gSUDg0Z3BELEqf\n/xsY3N6Fks4EzgQYPHjwOl+GAAMGDKC5ORlB1Ls3bLRmZclBrmrn0uYuDk4aPHgw06dPZ+zYsSxZ\nsoQ77riDM844gzfffJM33niD5uZmli5dSkTQ3NzMihUrWLVqVWvsEcHSpUvp3bv3Osdef/11Ghsb\nWbJkCVOmTGHIkCE0NzezevVqli9f3nr9ypUraWjoRe/eyXavXsvo0WMNvXs3M27cWKZM+T8uvfQy\nJDF37hxGjBjJQQeN5tZbr2bcuH2ZP/8x5s6d2xproRUrVrztc29r6dKlnZ5Ty/Iu/85l5vnnZq8t\n6/o1K5Z16/KXG3re//0rparJQNLRwIsR0SRpTHvnRERIanfq1IiYBEyCZNbStjMDzp8/v3Wm0Bde\ngCEblT5raDkT1V177bV88pOf5Otf/zoAF154ISNHjqRnz5707duXxsZGVq5ciSQaGxtpaGhg4403\nbo1dEv369XvbsUsuuYRx48YxaNAgRo8e3Tozaq9evdhkk01ar+/duzdvvtmblSuT7dWr+7B27Uas\nXNnIZz5zMRdccDYHH3wQa9euZeutt2XKlNv48IfP5pxzPsb+++/Hjjvuyj777NMaa6GGhobWFdY6\nsqHP2lhteZe/3LmFTj60vDuQX5zX1K3L32Y9qqLl/d+/Uqo6hbWkbwKnAWuABqA/cDOwLzAmIhZJ\nGgLMiIj15vfOprDOe9ZST2E9prwAurG8y192MvhMfScDT2GdqGoHckScGxFDI2IY8CHg7oj4CHAr\nMD49bTzwu2rGYWZm65fXfQaXAr+WdAbwLHByTnFYDcj7l6FZLcgsGUTEDJJRQ0TEYmBchV63w+Gd\nVp7usgqemZWvW09U19DQwOLFi/2lVQURweLFi2loaMg7FDPLQLeejmLo0KEsXLiQl156iVdfLe+1\nXnutvOtXrFiR6xdnNcrf0NDA0KFDy3thsw2c13NIdOtk0KtXr9abu/JuN54xY0anQzCrKe/ym1n3\n1q2biczMrDKcDMzMzMnAzMycDMzMjG7egWyV4xEVZvXNNQMzM3MyMDMzNxO1cjNJ/fJ/ezPXDMzM\nDCcDMzPDycDMzHAyMDMznAzMzAwnAzMzw8nAzMxwMjAzM5wMzMwMJwMzM6OEZCCpr6Se1QjGzMzy\n0WkykNRD0ocl3S7pReBxYJGkxyR9R9IO1Q/TzMyqqSs1g+nA9sC5wH9FxNYR8Q7gXcDfgG9J+kgV\nYzQzsyrryqyl742I1W13RsQS4CbgJkm9Kh6ZmZllptOaQUsikHRG22OSLi08x8zMuqdi1jN4v6QV\nEXEdgKQfAw3VCcvMzLJUVDIAbpW0FjgCeDUi3lZbMDOz7qfTZCBpYMHmfwO3AH8FLpQ0MO07MDOz\nbqwrNYMmIAAV/D0qfQSwXdWiMzOzTHSaDCJi2ywCMTOz/HT5DmRJfSSdJ2lSur2jpKOrF5qZmWWl\nmOkofgmsAg5Mt58HLql4RGZmlrliksH2EfFtYDVARCwj6T8wM7NurphksErSJiSdxkjaHlhZlajM\nzCxTxdxncAHwR2BrSdcBBwGnVyMoMzPLVpeTQURMk/QwsD9J89DnIuLlqkVmZmaZKWY0kYD3AftE\nxG1AH0n7dXJNg6QHJM2R9KikC9P9AyVNk7Qg/btZWaUwM7OyFNNn8BPgAOCUdLsZ+HEn16wEDomI\nkcCewBGS9ge+AtwVETsCd6XbZmaWk2KSweiI+DSwAiAiXgE2Xt8FkViabvZKHwEcB1yd7r8aOL6Y\noM3MrLKKSQar0+UuW0YTDQLWdnaRpJ6SZgMvAtMi4u/A4IhYlJ7yb2BwcWGbmVklKSK6dqJ0KvBB\nYG+SX/MnAedFxI1dvH5T4LfAZ4G/RMSmBcdeiYi39RtIOhM4E2Dw4MH7TJ06tcPXX7Sow0Ndstmg\nTvPaeq1ZsYx+/fqVF0QZ8ix/PZcdXH6XP9/yd2bs2LFNETGqs/OKGU10naQmYBzJaKLjI2J+Ede/\nKmk6yfTX/5E0JCIWSRpCUmto75pJwCSAUaNGxZgxYzp8/YkTuxpJ+04+tLms61+c18T64qu2PMtf\nz2UHl9/lz7f8lVLMaKKLga2ByRHxo64kAkmD0hoB6Q1rhwKPA7cC49PTxgO/KzZwMzOrnGJuOnua\nZCTRDyQ1A/cCMyNifV/kQ4Cr076GHsCvI+I2SfcDv06X0nwWOLm08M3MrBKKaSb6JfBLSf9F8uU9\ngaQ9v3E918wF9mpn/2KS5iYzM9sAdDkZSPo5MBz4D0mt4CTg4SrFZWZmGSpmaOnmQE/gVWAJ8HJE\nrKlKVGZmlqlimolOAJC0K3A4MF1Sz4gYWq3gzMwsG8U0Ex0NHAy8G9gUuJukucjMzLq5YkYTHUHy\n5X95RLxQpXjMzCwHxTQTfWZ9xyXdHxEHlB+SmZllrZgO5M40VPC1zMwsQ5VMBl2b5MjMzDY4lUwG\nZmbWTVUyGaiCr2VmZhnqUjJI1ySY3slpp1UgHjMzy0GXkkFEvAmslTRgPefMq1hUZmaWqWLuM1gK\nPCJpGvBGy86IOKviUZmZWaaKSQY3pw8zM6sxxdx0dnW6QM07I+KJKsZkZmYZK2als2OA2cAf0+09\nJd1arcDMzCw7xQwtnQjsRzKFNRExG9iuCjGZmVnGikkGqyPitTb71lYyGDMzy0cxHciPSvow0FPS\njsBZwH3VCcvMzLJUTM3gs8BuwErgBuB14OxqBGVmZtkqZjTRMuBrkr6VbEZz9cIyM7MsFTOaaF9J\njwBzSW4+myNpn+qFZmZmWSmmz+Aq4FMRcS+ApHcBvwRGVCMwMzPLTjF9Bm+2JAKAiPgLsKbyIZmZ\nWdaKqRncI+lnJJ3HAXwQmCFpb4CIeLgK8ZmZWQaKSQYj078XtNm/F0lyOKQiEZmZWeaKGU00dn3H\nJY2PiKvLD8nMzLJWyZXOPlfB1zIzswx52UszM6toMogKvpaZmWXINQMzM6toMvhrBV/LzMwyVMx0\nFJ+T1F+JqyQ9LOmwluMR8ZnqhGhmZtVWTM3g4xHxOnAYsBlwGnBpVaIyM7NMFZMMWvoEjgSuiYhH\ncT+BmVlNKCYZNEm6kyQZ/ElSI17pzMysJnTpDmRJAr4ODAKejohlkjYHPlbN4MzMLBtdSgYREZLu\niIg9CvYtBhZXLTIzM8tMMc1ED0vat5gXl7S1pOmSHpP0qKTPpfsHSpomaUH6d7OiojYzs4oqJhmM\nBu6X9JSkuZIekTS3k2vWAF+IiOHA/sCnJQ0HvgLcFRE7Anel22ZmlpNiprA+vNgXj4hFwKL0ebOk\n+cBWwHHAmPS0q4EZwJeLfX0zM6sMRXRtSiFJ72xvf0Q818XrhwEzgd2B5yJi03S/gFdatttccyZw\nJsDgwYP3mTp1aoevv2hRV6Lo2GaDyhsYtWbFMvr161deEGXIs/z1XHZw+V3+fMvfmbFjxzZFxKjO\nziumZnA7yWR0AhqAbYEngN06u1BSP+Am4OyIeD35/k+kndPtZqSImARMAhg1alSMGTOmw/eYOLGL\npejAyYc2l3X9i/OaWF981ZZn+eu57ODyu/z5lr9SilncZo/C7XS5y091dp2kXiSJ4LqIuDnd/R9J\nQyJikaQhwItFxGxmZhVW8kR16ZrHo9d3TtoEdBUwPyK+W3DoVmB8+nw88LtS4zAzs/J1uWYg6ZyC\nzR7A3sALnVx2EMkcRo9Imp3u+yrJnEa/lnQG8CxwcpcjNjOziiumz6Cx4Pkakj6Em9Z3QUT8hY7n\nLxpXxHubmVkVFdNncCG0dgYTEUurFZSZmWWrmPUMdpc0C3gUeFRSk6TdqxeamZllpZgO5EnAORGx\nTURsA3wh3WdmZt1cMcmgb0RMb9mIiBlA34pHZGZmmSumA/lpSecD16TbHwGernxIZmaWtaKWvSRZ\nz+Dm9DEo3WdmZt1cMaOJXgHOqmIsZmaWk2JuOtsJmAAMK7wuIg6pfFhmZpalYvoMbgSuAH4OvFmd\ncMzMLA/FJIM1EfHTqkViZma56TQZSBqYPv29pE8BvwVWthyPiCVVis3MzDLSlZpBE2+tYwDwxYJj\nAWxX6aDMzCxbnSaDiNg2i0DMzCw/nd5nIOldnRzv7zmKzMy6t640E71f0reBP5I0Gb1EsuzlDsBY\noGWeIjMz66a60kz0+bQT+f3AB4AhwHJgPvCzdM0CMzPrxro0tDQdMXRl+jAzsxpT6rKXLV4DmiJi\ndjvHzMysmyhmorpRwCeBrdLHJ4AjgCslfakKsZmZWUaKuQN5KLB3y3KXki4gWQf53SQdy9+ufHhm\nZpaFYmoG76DgzmNgNTA4Ipa32W9mZt1MMTWD64C/S/pdun0McL2kvsBjFY/MzMwyU8x6BhdL+gNw\nULrrkxHxUPr81IpHZmZmmSlmNNEPgKkRcXkV4zEzsxwU02fQBJwn6SlJl0kaVa2gzMwsW11OBhFx\ndUQcCewLPAF8S9KCqkVmZmaZKaZm0GIHYBeSOYker2w4ZmaWhy4nA0nfTmsCFwGPAKMi4piqRWZm\nZpkpZmjpU8CBJIvZ9AZGSCIiZlYlMjMzy0wxyWAtcDfJncizgf2B+4FDqhCXmZllqJg+g7NIOo+f\njYixwF7Aq1WJyszMMlVMMlgRESsAJPWOiMeBnasTlpmZZamYZqKFkjYFbgGmSXoFeLY6YZmZWZaK\nmY7ihPTpREnTgQEkS2GamVk3V0zNoFVE3FPpQMzMLD+l3HRmZmY1xsnAzMycDMzMrMrJQNIvJL0o\naV7BvoGSpklakP7drJoxmJlZ56pdM5gMHNFm31eAuyJiR+CudNvMzHJU1WSQzlu0pM3u44Cr0+dX\nA8dXMwYzM+ucIqK6byANA26LiN3T7VcjYtP0uYBXWrbbufZM4EyAwYMH7zN16tQO32fRovLi3GzQ\n2rKuX7NiGf369SsviDLkWf56Lju4/C5/vuXvzNixY5siotPFyEq6z6BSIiIkdZiNImISMAlg1KhR\nMWbMmA5fa+LE8mI5+dDmsq5/cV4T64uv2vIsfz2XHVx+lz/f8ldKHqOJ/iNpCED698UcYjAzswJ5\nJINbgfHp8/HA73KIwczMClR7aOkNJGse7CxpoaQzgEuBQ9NV096bbpuZWY6q2mcQEad0cGhcNd/X\nzMyK4zuQzczMycDMzJwMzMwMJwMzM8PJwMzMcDIwMzOcDMzMDCcDMzPDycDMzHAyMDMznAzMzAwn\nAzMzw8nAzMxwMjAzM5wMzMwMJwMzM8PJwMzMcDIwMzOcDMzMDCcDMzPDycDMzHAyMDMznAzMzAwn\nAzMzw8nAzMxwMjAzM5wMzMwMJwMzM8PJwMzMcDIwMzOcDMzMDCcDMzPDycDMzHAyMDMznAzMzAwn\nAzMzw8nAzMxwMjAzM3JMBpKOkPSEpCclfSWvOMzMLKdkIKkn8GPgfcBw4BRJw/OIxczM8qsZ7Ac8\nGRFPR8QqYCpwXE6xmJnVPUVE9m8qnQQcERH/nW6fBoyOiM+0Oe9M4Mx0c2fgiUwDLc4WwMt5B5GT\nei47uPwu/4Zd/m0iYlBnJ22URSSliohJwKS84+gKSQ9FxKi848hDPZcdXH6XvzbKn1cz0fPA1gXb\nQ9N9ZmaWg7ySwYPAjpK2lbQx8CHg1pxiMTOre7k0E0XEGkmfAf4E9AR+ERGP5hFLBXWL5qwqqeey\ng8vv8teAXDqQzcxsw+I7kM3MzMnAzMycDKxIknpKuizvOPIkSe3s651HLFmT1EPSyXnHYZXnZFAG\nSYMlHZ0+3pF3PFmIiDeBd+UdR86uKtyQ1A+4I6dYMhURa4Ev5R1HniT1kXS+pCvT7R0lHZ13XOVy\nMihR+uvoAeADwMnA39M7q+vBLEm3SjpN0oktj7yDytBCST8BkLQZcCdwbb4hZerPkiZI2lrSwJZH\n3kFl6JfASuCAdPt54JL8wqkMjyYqkaQ5wKER8WK6PQj4c0SMzDey6pP0y3Z2R0R8PPNgciLp20B/\nYB/g0oi4KeeQMiPpn+3sjojYLvNgctByx7GkWRGxV7pvTnf/f3+Dno5iA9ejJRGkFlMnNa2I+Fje\nMeShTe3n78D5JLXDkHRiRNycT2TZioht844hZ6skbQIEgKTtSWoK3ZqTQen+KOlPwA3p9geBP+QY\nT2Yk7QT8FBgcEbtLGgEcGxHdvqrciWPabM8CeqX7A6iLZCCpD3AO8M6IOFPSjsDOEXFbzqFlZSLw\nR2BrSdcBBwHd/geSm4nKIOn9JP8QAO6NiN/mGU9WJN0DfBH4WUE1eV5E7J5vZJYFSb8CmoCPpj8G\n+gD3RcSeOYeWGUmbA/sDAv4WERvyrKVdUhfNGtWSthNPJOk8uqeOOtH6RMQDbfatySWSHEi6WtKm\nBdubSfpFnjFlbPuI+DawGiAilpF8KdYFSXdFxOKIuD0ibouIlyXdlXdc5XIzUYkkfQK4EFgBrCX5\nnyGAeuhEezltJ21pMz0JWJRvSJkaERGvtmxExCuS9sozoIzVZJt5ZyQ1AH2ALdJRZC0JsD+wVW6B\nVYiTQekmALvXQvWwBJ8mmZxrF0nPA/8EPpJvSJnqIWmziHgFIK0R1tP/Sxfw9jbz03ONKBufAM4G\ntgQeLtj/OvCjXCKqIPcZlEjSH4ET0ypyXZLUl2RUVXPesWRJ0keBrwI3kvw6PAn4RkRck2tgGarF\nNvOukvTZiPhh3nFUmpNBidJmgV+SDDFsrSJHxFm5BVVlks5Z3/GI+G5WseRN0m7A2HTz7oh4LM94\nsiZpK2Cv7VeEAAAMq0lEQVQbCmpEETEzv4iqT9IhEXF3RzdYdvehxfVUta20nwF3A4+Q9BnUg8b0\n787Avry1INExJOPt60ZEPCrpJaABQNI7I+K5nMPKhKRvkQylfpS3/u0HUNPJAHgPyf/zbYcYQw0M\nLXbNoESFdx/WG0kzgaNamockNQK3R8S7840sG5KOBf6PpO34RZJfyPMjYrdcA8uIpCdIOtFrvtO4\nnrhmULo/SDoT+D3rNhMtyS+kzAwGVhVsr0r31YuLSdrL/xwRe0kaS311oD9NcrNdXSaDdIba9wPD\nWLeZ7KK8YqoEJ4PSnZL+PbdgX70MLZ0CPCCp5Sa744Grc4wna6sjYnE6nXOPiJgu6ft5B1Vtkn5I\n8m98GTA7HVtfF/1lbfwOeI3kxruaSYhuJrKSSNqHt6aynhkRs/KMJ0uS/kySAL8JbEHSVLRvRByY\na2BVJmn8eg5HREzJLJgc1erd9k4GRar1EQVdJaknSdNQYTW5XjpQ+5LcbCjgVGAAcF1ELM41sIxI\n+lxEXN7ZvlolaRLww4h4JO9YKsnJoEiSLoyIC+p5GmdJnyW58eg/wJukd19HxIhcA7NMSHo4IvZu\ns6/mB1RIeoSkmWwjYEeSvpOV1Mi/fyeDEknq3XY0haSB9dCBLOlJYHS9/BJuIamZdAqGtodIvgz6\nZxxSpiSdAnyYpHnw3oJDjcDaiBiXS2AZkbTN+o5HxLNZxVIN7kAu3c2SjouINQCS/gu4nWSxk1r3\nL5IOtLoSEY2dn1XT7iOZg2oLkqG1LZqBublElK3Pk3wGf42I5/MOptJcMyiRpP8BjiSZimBrkhuw\nJkTEnbkGlgFJV5HceHY7644mqZs7kKF1Xv/hwDN1Nh3DdiTlBngsIp7OM56sSPoMcGD6gCQx3Af8\nFZiTrg/dbTkZlEHSp4EjSMYbfyIi7ss3omxIuqC9/RFxYdaxZCm92ewHwBLgPODHJP0mw4AvR0RN\nD6+V1B/4OUntd066e0+SIZZnRMTrecWWNUlb8lZiOBZ4R3dvJnQyKFKb+XkEfJSkijwL6u/XcT1J\n173+AMnooekkd+E+LekdwF0RsUeuAVaZpMnAM8BFLb+CJYlk+c8dIuKj+UWXjbS8e5AkgYNIakgv\nAfd39x9D7jMoXtt245s72F+zJA0CvgTsRjo3D0BEHJJbUNlYGxH/gGRR+JbmkYh4UVI9LO5zUESc\nXrgjkl+TF0lakE9I2ZE0jWTtgtnA34D/jYj5+UZVOU4GRWov+0vqAfSro2rydcCvgKOBTwLjSX4d\n1boe6aImPYC1bRY4qfdVA+thpbOngREkw0oXkyzy9FKt9Be5mahEkq4n+SJ8E3iQ5BfD5RHxnVwD\ny4CkpojYR9LclrHVkh6MiH3zjq2aJD3DW6vatRURUdNTkUi6GngKuDgKvjgknQ/sFBGn5RZchtK+\nk/1Jmor2BwYB8yJifXdob/BcMyjd8Ih4XdKpwB+Ar5B0pNV8MiBd+xZYJOko4AWg5td/johheceQ\ns88CVwFPSpqd7tuTpL/sjNyiyt5KkvmZlqfPhwIb5xpRBTgZlK6XpF4kc9T8KCJWS6qXatYlkgYA\nXwB+SFIr+ny+IVm1pc2gH0jXPC4cWvpUjmFlRtL3SGoDO5IkwPuBK4DxhWtid1dOBqX7GcnIijnA\nzPTuxLroM4iI29Knr/HWal9WJ9Iv/7pIAG38E7gWmB0Rb+YdTKW5z6BC0iFnPVvuSK5FBVMYt6uO\npjA2qzmuGRRJ0vcj4uz0eetMjRERkn4OnJ5nfFX2UN4BmFl1uGZQpMIZG9vO3tjebI61rg6H1ZrV\npHofG10KdfC8bki6XlL/dF7/ecBjkr6Yd1xmVjong+L1kLSZpM0Lng+UNBDomXdwGRme1gSOJxlW\nuy1QF2PMzWqV+wyKN4DkfoKWWsHDBcfqpc2tnofVmtUkJ4Mi+cYjoI6H1ZrVKncgV4CkiRExMe84\n8iRpo1oeVmtW65wMKqBeRhG1mb77bTx9t1n35WaiyqiXUUV1M023Wb1xzaACJPXo7kvemVl9czIo\nUbrAy/+QLHnYWsOKiI/nFVMe6qWJzKzWuZmodL8D7gX+TLKmQb2qlyYys5rmZFC6PhHx5byD2ADc\nnncAZlY+34FcutskHZl3EHmQtK2kBoCIOE/SJpKG5RuVmZXDfQYlktQM9CVZ6Wg1SXNJRET/XAPL\ngKSHgAMjYlW6vTHw11pf9tKslrmZqEQRUc/DLDdqSQQAEbEqTQhm1k05GZRB0mYkS+A1tOyLiJn5\nRZSZlyQdGxG3Akg6Dng555jMrAxuJiqRpP8GPkeyGPZsYH/g/og4JNfAMpCugXsdsCVJ89i/gI9G\nxJO5BmZmJXMyKJGkR4B9gb9FxJ6SdgH+NyJOzDm0zEjqBxARS/OOxczK42ai0q2IiBWSkNQ7Ih6X\ntHPeQWVF0lHAbkBDsvwzRMRFuQZlZiVzMijdQkmbArcA0yS9Ajybc0yZkHQF0AcYC/wcOAl4INeg\nzKwsbiaqAEnvIVn05o+Fo2xqlaS5ETGi4G8/4A8RcXDesZlZaVwzKJKk/hHxerrMZYtH0r/9gCU5\nhJW15enfZZK2BBYDQ3KMx8zK5GRQvOuBo0mWvgzWnZsngO3yCCpjt6VNZN8hWfYzgCvzDcnMyuFm\nIiuLpN5AQ0S8lncsZlY6J4MiSVrvdM0R8XBWsWRN0o7AZcD2JE1jEyLi+XyjMrNKcDIokqTp6dMG\nYBTJovACRgAPRcQBecVWbZLuBaYAM4FjgQPq6b4Ks1rmWUuLFBFjI2IssAjYOyJGRcQ+wF5Arf9K\nboyIKyPiiYj4DsnCPmZWA9yBXLqdI6JlFBERMU/SrnkGlIEGSXvxVqf5JoXbtdxEZlbr3ExUIkk3\nAG8A16a7TgX6RcQp+UVVXQVNZO2JepiXyaxWORmUKF3c5f8B7053zQR+GhEr8ovKzKw0TgZWFkmT\nIuLMvOMws/K4z6BEkg4CJgLbUPA5RkQ93HRWaFTeAZhZ+ZwMSncV8HmSO5HfzDmWPL2YdwBmVj43\nE5VI0t8jYnTecZiZVYKTQYkkXQr0BG4GVrbsr4fhlZJ2Ar7I25vIPJrIrJtyMihRB8Ms62J4paQ5\nwBW0aSKLiKbcgjKzsjgZWNEkNaV3XZtZjXAyKJKkc9Z3PCK+m1UseZE0kaTj+Les20RWD2s5mNUk\njyYqXmPeAWwAxqd/v1iwr17WcjCrSa4ZmJmZawaVIOnhiFjvOge1RtLuwHCSqbwBiIgp+UVkZuVw\nMqgMdX5K7ZB0ATCGJBncAbwP+AvJWgdm1g15PYPKuD3vADJ2EjAO+HdEfAwYCQzINyQzK4eTQQkk\n9Sy8zyAizssznhwsj4i1wBpJ/UlGFm2dc0xmVgY3E5UgIt6UtFbSgDpdCP4hSZsCV5LceLYUuD/f\nkMysHB5NVCJJvyNZ6nIaySI3AETEWbkFlQNJw4D+ETE351DMrAxOBiWSNL69/RFxddaxZEXSLhHx\nuKR2R07Vw7xMZrXKyaAMkjYB3hkRT+QdSxZaFrKp53mZzGqVk0GJJB0DXAZsHBHbStoTuCgijs05\nNDOzonk0UekmAvsBrwJExGzqZDoGSR+Q1Jg+P0/SzZL2yjsuMyudk0HpVrczkmhtLpFk7/yIaJb0\nLuC9JKu+XZFzTGZWBieD0j0q6cNAT0k7SvohcF/eQWWkZQ2Do4BJEXE7sHGO8ZhZmZwMSvdZYDeS\nKZxvAF4Hzs41ouw8L+lnwAeBOyT1xv+WzLo1dyBb0ST1AY4AHomIBZKGALtHxLScQzOzEjkZlEjS\nKOCrwDDWXQd4RF4x5UFSX+BE4EMRcVTe8ZhZaTwdRemuI1nc5RHqp+MYAEkbk/QXfBg4HLgJdyCb\ndWtOBqV7KSJuzTuILEk6DDgFOAyYTjJl9b7pzKVm1o25mahEksaRfDHexbrrAN+cW1BVJmktcC9w\nekT8M933dETUxf0VZrXMNYPSfQzYBejFW81EAdRsMgD2Bj4E/FnS08BUoGe+IZlZJbhmUCJJT0TE\nznnHkRdJB5LUjN4PzAF+GxGT8o3KzErlseGlu0/S8LyDyEtE3BcRnwWGAt8D9s85JDMrg2sGJZI0\nH9ge+CdJn4FIZu6sq6GlZlYbnAxKJGmb9vZHxLNZx2JmVi4ngwpIb7w6ATjFN16ZWXfkPoMSSdpY\n0gmSbgQWAePwjVdm1k25ZlCkdm68+hXww4gYlmdcZmblcDIokm+8MrNa5JvOiucbr8ys5rhmUAbf\neGVmtcLJoAIk9SBZ/vFDEfHxvOMxMyuWk4GZmXloqZmZORmYmRlOBmZmhpOBmZnhZGBmZjgZmJkZ\nTgZmZgb8f+LC7EAIixHxAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xbcd7d30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYkAAAElCAYAAAARAx4oAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xbcd7d68>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZAAAAFmCAYAAABDdMzIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3XucVXW9//HXG1DGkcEbxkFRIUMUCTNRvB+QvPzS0NRI\nTyZeTmqZVufYRcufeDvHX1ons5OFmYg3MjM1b3md9Bw1FEFFyEC8BFEopg53Bj6/P9Ya3IwDzN57\n9loze7+fj8d+zF7fdfusPTCf/b2s71JEYGZmVqxueQdgZmZdkxOImZmVxAnEzMxK4gRiZmYlcQIx\nM7OSOIGYmVlJnECsqknaUdJiSd1L2HeApJDUoxKxdWXpZ/rREvcNSR/r6Jgse04gljlJr0taKalP\nq/Jp6R+XAR11roh4MyJ6RcTqjjpmR5I0UdJlrco6feJKP9O5ecdh+XICsby8BpzYsiDp40B9qQdr\n649tZ/4D3BX4M7WNcQKxvNwEnFywPA6YVLiBpCPTWsn7kv4iaXzBupZv6adLehN4bCNlPdL9tpB0\nvaQFkuZLuqyleUtSd0lXSXpb0lzgyPUFL+nbku5oVXa1pB+n70+RNFdSk6TXJH2h1A8qrbGdJ+lF\nSe9J+pWkunTdSEnzJH1L0sL0uo6R9GlJf5b0jqQLCo61j6SnJb2bbvsTSZsWrA9JZ0uaDczeSNnH\n0vc908/tTUl/l/QzSZsVHPOb6bn+Kum0Uj8H63ycQCwvzwC9Je2W/gE/Abi51TZLSJLMliR/zL8s\n6ZhW2/wzsBtw+EbKWkwEmoGPAXsChwH/mq77EnBUWj4cOH4D8U8GPi2pAZLkA4wFbpW0OfBj4P9E\nRAOwPzB9A8dqj7HAEcBAYBhwSsG6fwLqgO2B/wtcB5wE7AUcBFwoaWC67WrgG0AfYD9gNPCVVuc6\nBhgBDNlIWYsrgF2AT5B8ri1xIOkI4DzgUGAQ8KliLto6uYjwy69MX8DrJH9Ivgf8J8kfxoeBHkAA\nA9az34+A/0rfD0i3/WjB+g2V9QD6AiuAzQrWnwg8nr5/DDirYN1hLfuuJ57/AU5O3x8KvJq+3xx4\nFziu8FzrOcZE4LJWZQMKz5t+XicVrP8+8LP0/UhgGdA9XW5I9x1RsP1U4Jj1nP/rwG8LlgM4pNU2\n6yv7GCCSRL9zwbr9gNfS978ErihYt0vLvnn/O/Sr/JfbMy1PNwFPkHyrntR6paQRJN9uhwKbAj2B\nX7fa7C9tHLetMoCdgE2ABZJayroVbL9dq33f2Ej8t5IkoEnAv6TLRMQSSZ8n+eZ9vaT/Bf49Iv7U\nxjGa05gKbQKsSV8t/lbwfmkaa4tF8cEggWXpz78XrF8G9AKQtAvwQ5IaVj1JYp3a6vzFfKbbpseZ\nWvCZCmgZ9bZdq+Nv7DO1LsRNWJabiHiDpDP908CdbWxyK3APsENEbAH8jOSP0zqHaevQ6znlX0hq\nIH0iYsv01Tsidk/XLwB2KNh+x41cwq+BkZL6A59N400CiPh9RBwK9AP+RNKs1JY3SWochQYCf4mI\nNR/evGzXpvEMiojewAWU95m+TZKgdi/4TLeIiF7p+mI/U+tCnEAsb6eTNI8saWNdA/BORCyXtA/J\nt/ySRcQC4CHgB5J6S+omaWdJ/5xucjtwrqT+krYCvrOR470FNAI3kDTZzAKQ1FfS0WlfyApgMevW\nJgr9BjhS0mFpJ/52JE17k8u51g1oAN4HFkvaFfhyOQdLk9x1wH9J+giApO0ltfQ/3Q6cImmIpHrg\nonLOZ52LE4jlKiJejYjn1rP6K8AlkppIOmVv74BTnkzSHDYT+AdwB0ktAZI/hL8HXgCep+1aUWu3\nkvTn3FpQ1g34N+CvwDsknfpt/qGOiJdJmsH+M932aeCPwMVFXFMxziNJxE0k1/urDjjmt4E5wDOS\n3gceAQYDRMQDJH1Xj6XbPNYB57NOQhF+oJSZmRXPNRAzMyuJE4iZmZXECcTMzEriBGJmZiWp6hsJ\n+/TpEwMGDMg7jDYtWbKEzTffPO8wcuPr9/X7+jvn9U+dOvXtiNi2PdtWdQIZMGAAzz23vhGi+Wps\nbGTkyJF5h5EbX7+v39c/Mu8w2iSp3bMFuAnLzMxK4gRiZmYlcQIxM7OSVHUfSFtWrVrFvHnzWL58\nea5xbLHFFsyaNSvXGCqhrq6O/v37s8kmrSeYNbNqU3MJZN68eTQ0NDBgwAAKpp/OXFNTEw0NDbmd\nvxIigkWLFjFv3jwGDhy48R3MrEuruSas5cuXs8022+SaPKqVJLbZZpvca3dmlo2aSyCAk0cF+bM1\nqx01mUDMzKx8NdcH0tr48dkfTxJjx47lV79KHsXQ3NxMv379GDFiBPfee29Z599///156qmn2r39\n+PHj6dWrF+edd15Z5zWz2lPzCSQPm2++ObNmzWLZsmVsttlmPPzww2y//fZFHaO5uZkePXp8aLmY\n5GFWq8r94tjRXzy7Kjdh5eSwww7jvvvuA+C2227jxBNPXLtuypQp7Lfffuy5557sv//+vPLKKwBM\nnDiRMWPGcMghhzB69GgaGxs56KCDGDNmDEOGDAGgV69ea49z5ZVXsvfeezNs2DAuuuiDJ4lefvnl\n7LLLLhx44IFrj21mViwnkJwcd9xxTJ48meXLl/Piiy8yYsSItet23XVXnnzySaZNm8Yll1zCBRdc\nsHbd888/zx133MEf/vCHtctXX301f/7zn9c5/kMPPcTs2bOZMmUK06dPZ+rUqTzxxBNMnTqVyZMn\nM336dO6//36effbZbC7YzKqOm7ByMnToUF5//XVuu+02Pv3pT6+z7r333mPcuHHMnj0bSaxatWrt\nukMPPZStt9567fI+++zT5j0XDz30EA899BB77rknAIsXL2b27Nk0NTXx2c9+lvr6egDGjBlTicsz\nsxrgGkiOxowZw3nnnbdO8xXAhRdeyKhRo5gxYwa/+93v1rmvovUU0OubEjoiOP/885k+fTrTp09n\nzpw5nH766R1/EWZWs5xAcnTaaadx0UUX8fGPf3yd8vfee29tp/rEiRNLOvbhhx/OL3/5SxYvXgzA\n/PnzWbhwIQcffDB33XUXy5Yto6mpid/97ndlXYOZ1a6ab8LKczRF//79Offccz9U/q1vfYtx48Zx\n2WWXceSRR5Z07MMOO4xZs2ax3377AUnn+s0338wnP/lJPv/5z7PHHnvwkY98hL333rusazCz2qWI\nyDuGihk+fHi0fqDUrFmz2G233XKK6APVOBdWi/Z8xp35gTpZ8PXne/15D+PN+/o3RNLUiBjenm3d\nhGVmZiVxAjEzs5I4gZiZWUmcQMzMrCROIGZmVpKaH8ZrZlasmW835R1Cp1DRBCLpl8BRwMKIGJqW\nXQl8BlgJvAqcGhHvpuvOB04HVgPnRsTv0/K9gInAZsD9wNeig8Yfd/Q/hCF9Njw09/XXX+eoo47i\n6aefXluW95Tqzz33HJMmTeLHP/5xu/cZOXIkV111FcOHt2u0n5lVoUo3YU0EjmhV9jAwNCKGAX8G\nzgeQNAQ4Adg93eenkrqn+1wLfAkYlL5aH9Paqbm5+UPLw4cPLyp5mJlBhWsgEfGEpAGtyh4qWHwG\nOD59fzQwOSJWAK9JmgPsI+l1oHdEPAMgaRJwDPBAJWPPw8iRIxkxYgSPP/447777Ltdffz0HHXQQ\nEydO5K677mLJkiXMnj2b8847j5UrV3LTTTfRs2dP7r//frbeemuuu+46JkyYwMqVK/nYxz7GTTfd\nRH19Paeccgp1dXVMmzaNAw44gN69e/Pqq68yd+5cdtxxR84880yuuuoq7r33XpYsWcI555zDjBkz\nWLVqFePHj+foo49m2bJlnHrqqbzwwgvsuuuuLFu2LO+Py8xylncn+ml8kAi2B/5SsG5eWrZ9+r51\neVVqbm5mypQp/OhHP+Liiy9eWz5jxgzuvPNOnn32Wb773e9SX1/PtGnT2G+//Zg0aRIAxx57LM8+\n+ywvvPACu+22G9dff/3a/efNm8dTTz3FD3/4QwBmzpzJI488wm233bbO+S+//HIOOeQQpkyZwuOP\nP843v/lNlixZwrXXXkt9fT2zZs3i4osvZurUqRl8GmbWmeXWiS7pu0AzcEsHH/cM4AyAvn370tjY\nuM76LbbYgqamD/o9Vq5Y0ZGnp2kjXSpLlixhzZo1rF69em0cK1asYNNNN2X16tUcccQRNDU1MXjw\nYObOnUtTUxPLly/nwAMPBKCuro7evXszatQompqaGDRoEDNmzKCpqYkpU6Zw6aWX8t5777FkyRJG\njx5NU1MTq1at4qijjmLp0qVrz3f44YfT3NxMU1MTS5cuXfv+wQcf5K677uL73/8+AMuWLWPWrFk8\n9thjnHXWWTQ1NTFw4ECGDh3KkiVL1vksWyxfvvxDn3trixcv3ug21czXn+/1Dx5c3v5vTl9T1v7N\ny5dWxe8/lwQi6RSSzvXRBZ3h84EdCjbrn5bNT9+3Lm9TREwAJkAyF1br+WZmzZq1zhxUm3Zs/tjo\n/FY77bQT7733Ht27d1+77ZIlS9h1113p3r07W221FQ0NDaxYsYI1a9bQ0NBAXV0dvXr1Wrt99+7d\n2WabbWhoaKC+vp5u3brR0NDAV77yFe666y722GMPJk6cSGNjIw0NDWyyySb06dNn7f49e/Zc53j1\n9fX06NGDhoYGJPHb3/6Wwa3+h/Xo0YP6+vq1+3Tr1o3NN9+8zeutq6tb+xyS9enMcwFlwdfftefC\nGntoeYNvFs6YWhW//8ybsCQdAXwLGBMRSwtW3QOcIKmnpIEkneVTImIB8L6kfSUJOBm4O+u4O0qv\nXr3o16/f2icKvvPOOzz44INraxjlaGpqol+/fqxatYpbbimtYnf44YdzzTXX0JLXp02bBsDBBx/M\nrbfeCiTNaS+++GLZ8ZpZ11bpYby3ASOBPpLmAReRjLrqCTyc5AOeiYizIuJlSbcDM0mats6OiNXp\nob7CB8N4H6ADO9A3Nuy2EiZNmsRZZ53F9773PQAuuugidt5557KPe+mllzJixAi23XZbRowY0Wbz\n0sZceOGFfP3rX2fYsGGsWbOGgQMHcu+99/LlL3+ZU089ld12243ddtuNvfbaq+x4zaxr83TuOfF0\n7m7C8fWPzO38ZTdhfbV6m7A8nbuZmVWcE4iZmZWkJhNINTfb5c2frVntqLkEUldXx6JFi/yHrgIi\ngkWLFlFXV5d3KGaWgZqbjbd///7MmzePt956K9c4li9fXpV/aOvq6ujfv//GNzSzLq/mEsgmm2zC\nwIED8w6DxsbGjd5sZ2bWmdVcE5aZmXUMJxAzMyuJE4iZmZXECcTMzEriBGJmZiVxAjEzs5I4gZiZ\nWUmcQMzMrCROIGZmVhInEDMzK4kTiJmZlcQJxMzMSuIEYmZmJXECMTOzkrRrOndJ/YETgIOA7YBl\nwAzgPuCBiFhTsQjNzKxT2mgCkXQDsD1wL/D/gIVAHbALcATwXUnfiYgnKhmomZl1Lu2pgfwgIma0\nUT4DuFPSpsCOHRuWmZl1dhvtA2lJHpL2ar1O0lERsTIi5rS1r6RfSlooaUZB2daSHpY0O/25VcG6\n8yXNkfSKpMMLyveS9FK67seSVOyFmplZxyqmE/06SUNbFiSdCFy4kX0mkjRzFfoO8GhEDAIeTZeR\nNISkn2X3dJ+fSuqe7nMt8CVgUPpqfUwzM8tYMQnkeGCSpF0lfQn4CnDYhnZI+0XeaVV8NHBj+v5G\n4JiC8skRsSIiXgPmAPtI6gf0johnIiKASQX7mJlZTto1CgsgIuZKOgG4C3gTOCwilpVwzr4RsSB9\n/zegb/p+e+CZgu3mpWWr0vety9sk6QzgDIC+ffvS2NhYQoiVt3jx4k4bWxZ8/b7+PK9/8ODy9n9z\nenkDT5uXL62K3397RmG9BERB0dZAd+CPkoiIYaWePCJCUmx8y6KOOQGYADB8+PAYOXJkRx6+wzQ2\nNtJZY8uCr9/Xn+f1jx9f3v5jD20qa/+FM6ZWxe+/PTWQozr4nH+X1C8iFqTNUwvT8vnADgXb9U/L\n5qfvW5ebmVmO2jMK642WF8kf+EPS90vbs38b7gHGpe/HAXcXlJ8gqaekgSSd5VPS5q73Je2bjr46\nuWAfMzPLSbv7QCRdBAwHBgM3AJsANwMHbGCf24CRQB9J84CLgCuA2yWdDrwBjAWIiJcl3Q7MBJqB\nsyNidXqor5CM6NoMeCB9mZlZjtqdQIDPAnsCzwNExF8lNWxoh4g4cT2rRq9n+8uBy9sofw4Y+uE9\nzMwsL8U0Qa1Mh9EGgKTNKxOSmZl1BcUkkNsl/RzYMr0P5BHgusqEZWZmnV0x94FcJelQ4H2SfpD/\nGxEPVywyMzPr1IrpAwH4M8ntG49IqpfUEBHlDYg2M7Muqd1NWGmz1R3Az9Oi7UnuSjczsxpUTB/I\n2SRDdt8HiIjZwEcqEZSZmXV+xSSQFRGxsmVBUg/WneLEzMxqSDEJ5A+SLgA2SzvTfw38rjJhmZlZ\nZ1dMAvkO8BbwEnAmcD/wvUoEZWZmnV8xo7BGATdHhO/9MDOzomogJwMvSHpG0pWSPlP4OFozM6st\nxdxIOA5A0nYkTyf8b2C7Yo5hZmbVo5jZeE8CDgI+DrwN/AR4skJxmZlZJ1dM7eFHwKvAz4DHI+L1\nikRkZmZdQrv7QCKiD3AaUAdcLmmKpJsqFpmZmXVqxUxl0hvYEdgJGABsAZT3ZHkzM+uyimnC+p+C\n108iYl5lQjIzs66gmFFYwza0XtI1EXFO+SGZmVlXUMx9IBuz3mejm5lZ9enIBGJmZjXECcTMzErS\nkQlEHXgsMzPr5NqVQCR1l3TVRja7upgTS/qGpJclzZB0m6Q6SVtLeljS7PTnVgXbny9pjqRXJB1e\nzLnMzKzjtSuBRMRq4MCNbDOxvSeVtD1wLjA8IoYC3YETSKaMfzQiBgGPpstIGpKu3x04AvippO7t\nPZ+ZmXW8Ypqwpkm6R9IXJR3b8irj3D1IHk7VA6gH/gocDdyYrr8ROCZ9fzQwOSJWRMRrwBxgnzLO\nbWZmZSrmRsI6YBFwSEFZAHcWe9KImJ82ib0JLAMeioiHJPWNiAXpZn8D+qbvtweeKTjEvLTMzMxy\nUsyNhKd21EnTvo2jgYHAu8Cv09l+C88Xkop+5rqkM4AzAPr27UtjY2P5AVfA4sWLO21sWfD1+/rz\nvP7Bg8vb/83p5c3i1Lx8aVX8/ouZzn0X4Fqgb0QMlTQMGBMRl5Vw3k8Br0XEW+mx7wT2B/4uqV9E\nLJDUD1iYbj8f2KFg//5p2YdExARgAsDw4cNj5MiRJYRXeY2NjXTW2LLg6/f153n948eXt//YQ5vK\n2n/hjKlV8fsvpg/kOuB8YBVARLxI0rFdijeBfSXVSxIwGpgF3AOMS7cZB9ydvr8HOEFST0kDgUHA\nlBLPbWZmHaCYPpD6iJiS/L1fq7mUk0bEHyXdATyfHmMaSa2hF3C7pNOBN4Cx6fYvS7odmJluf3Y6\nMszMzHJSTAJ5W9LOJB3nSDoeWLDhXdYvIi4CLmpVvIKkNtLW9pcDl5d6PjMz61jFJJCzSWoJu0qa\nD7wGfKEiUZmZWadXzCisucCnJG0OdIuI8nqRzMysSyvmiYTbSPox8CTQKOlqSdtULjQzM+vMihmF\nNRl4CzgOOD59/6tKBGVmZp1fMX0g/SLi0oLlyyR9vqMDMjOzrqGYGshDkk6Q1C19jQV+X6nAzMys\ncysmgXwJuJVkqO0KkiatMyU1SXq/EsGZmVnnVcworIYNrZe0e0S8XH5IZmbWFXTkEwlv6sBjmZlZ\nJ+dH2pqZWUk6MoEUPfW6mZl1XR2ZQMzMrIZ0ZAJZ2YHHMjOzTq6YqUwOSOfBQtJJkn4oaaeW9RGx\nbyUCNDOzzqmYGsi1wFJJewD/DrwKTKpIVGZm1ukVk0CaIyJInmX+k4j4b2CD94aYmVn1KmYurCZJ\n5wMnAQdL6gZsUpmwzMyssyumBvJ5kilMTo+IvwH9gSsrEpWZmXV67aqBSOoO3BYRo1rKIuJN3Adi\nZlaz2lUDiYjVwBpJW1Q4HjMz6yKK6QNZDLwk6WFgSUthRJzb4VGZmVmnV0wCuTN9mZmZFTWd+42V\nDMTMzLqWYu5Ef03S3NavUk8saUtJd0j6k6RZkvaTtLWkhyXNTn9uVbD9+ZLmSHpF0uGlntfMzDpG\nMU1Ywwve1wGfA7Yu49xXAw9GxPGSNgXqgQuARyPiCknfAb4DfFvSEOAEYHdgO+ARSbuknftmZpaD\ndtdAImJRwWt+RPwIOLKUk6ajuQ4Grk+PvTIi3iW5y72lqexG4Jj0/dHA5IhYERGvAXOAfUo5t5mZ\ndYx210AkfbJgsRtJjaSYGkyhgcBbwA3p3FpTga8BfSNiQbrN34C+6fvtgWcK9p+XlrUV5xnAGQB9\n+/alsbGxxBAra/HixZ02tiz4+n39eV7/4MHl7f/m9DVl7d+8fGlV/P6LSQA/KHjfDLwOjC3jvJ8E\nzomIP0q6mqS5aq2ICElFP6QqIiYAEwCGDx8eI0eOLDHEympsbKSzxpYFX7+vP8/rHz++vP3HHtpU\n1v4LZ0ytit9/MaOwRm18q3abB8yLiD+my3eQJJC/S+oXEQsk9QMWpuvnAzsU7N8/LTMzs5wUMwpr\ni/QZIM+lrx+Uemd6OpfWXyS1VCRHAzOBe4Bxadk44O70/T3ACZJ6ShoIDAKmlHJuMzPrGMU0Yf0S\nmMEHzVZfBG4Aji3x3OcAt6QjsOYCp5IktNslnQ680XKuiHhZ0u0kSaYZONsjsMzM8lVMAtk5Io4r\nWL5Y0vRSTxwR01l3aHCL0evZ/nLg8lLPZ2ZmHauY6dyXSTqwZUHSAcCyjg/JzMy6gmJqIGcBk9J+\nDwHvAKdUIigzM+v8ihmF9QKwh6Te6fL7FYvKzMw6vWJuJOwJHAcMAHpIAiAiLqlIZGZm1qkV04R1\nN/AeyV3jKyoTjpmZdRXFJJD+EXFExSIxM7MupZhRWE9J+njFIjEzsy5lozUQSS8BkW57avoMkBUk\nI7EiIoZVNkQzM+uM2tOEdVTFozAzsy6nPQlkUUQs3tAGknptbBszM6su7ekDuTudOPFgSZu3FEr6\nqKTTJf0ecOe6mVmN2WgNJCJGS/o0cCZwQPqc8mbgFeB+YFw6u66ZmdWQdg3jjYj7SZKFmZkZUNzz\nQB5tT5mZmdWG9gzjrQPqgT5p85XSVb1Zz3PJzcys+rWnCetM4OvAdiTTmLQkkPeBn1QoLjMz6+Ta\n04l+NXC1pHMi4poMYjIzsy6gmOncr5G0P+lsvAXlkyoQl5mZdXLFTOd+E7AzMB1oeR55AE4gZmY1\nqJjZeIcDQyIiKhWMmZl1HcXMxjsD+KdKBWJmZl1LMTWQPsBMSVMoeKBURIzp8KjMzKzTKyaBjK9U\nEGZm1vUUMwrrDx19ckndgeeA+RFxlKStgV+RjPR6HRgbEf9Itz0fOJ2kA//ciPh9R8djZmbtV8xU\nJk2S3k9fyyWtlvR+mef/GjCrYPk7wKMRMQh4NF1G0hDgBGB3kpl/f5omHzMzy0m7E0hENERE74jo\nDWwGHAf8tNQTS+oPHAn8oqD4aODG9P2NwDEF5ZMjYkVEvAbMAfYp9dxmZlY+lTMqV9K0iNizxH3v\nAP4TaADOS5uw3o2ILdP1Av4REVtK+gnwTETcnK67HnggIu5o47hnAGcA9O3bd6/JkyeXdG2Vtnjx\nYnr16pV3GLnx9fv687z+BQvK23+rbdeUtX/z8qWd9vc/atSoqRExvD3bFnMj4bEFi91I7gtZXmRs\nLcc6ClgYEVMljWxrm4gISUVnt4iYAEwAGD58eIwc2ebhc9fY2EhnjS0Lvn5ff57XP358efuPPbSp\nrP0XzphaFb//YkZhfabgfTNJJ/fRJZ73AGBM+qCqOqC3pJuBv0vqFxELJPUDFqbbzwd2KNi/f1pm\nZmY5KWYU1qkdddKIOB84HyCtgZwXESdJuhIYB1yR/rw73eUe4FZJPySZFXgQMKWj4jEzs+IVMwqr\nv6TfSlqYvn6TdoR3pCuAQyXNBj6VLhMRLwO3AzOBB4GzI2L1eo9iZmYVV0wT1g3ArcDn0uWT0rJD\nywkgIhqBxvT9ImD0era7HLi8nHOZmVnHKWYurG0j4oaIaE5fE4FtKxSXmZl1csUkkEWSTpLUPX2d\nBCyqVGBmZta5FZNATgPGAn8DFgDHA6dUICYzM+sCiukDuQQYVzA31dbAVSSJxczMakwxNZBhLckD\nICLeAUq6C93MzLq+YhJIN0lbtSykNZBiajBmZlZFikkAPwCelvTrdPlzeFitmVnNKuZO9EmSngMO\nSYuOjYiZlQnLzMw6u6KaoNKE4aRhZmZF9YGYmZmt5QRiZmYlcQIxM7OSOIGYmVlJnEDMzKwkTiBm\nZlYSJxAzMyuJE4iZmZXECcTMzEriBGJmZiVxAjEzs5I4gZiZWUmcQMzMrCS5JBBJO0h6XNJMSS9L\n+lpavrWkhyXNTn8WPsDqfElzJL0i6fA84jYzsw/kVQNpBv49IoYA+wJnSxoCfAd4NCIGAY+my6Tr\nTgB2B44Afiqpey6Rm5kZkFMCiYgFEfF8+r4JmAVsDxwN3JhudiNwTPr+aGByRKyIiNeAOcA+2UZt\nZmaFFBH5BiANAJ4AhgJvRsSWabmAf0TElpJ+AjwTETen664HHoiIO9o43hnAGQB9+/bda/LkyZlc\nR7EWL15Mr1698g4jN75+X3+e179gQXn7b7XtmrL2b16+tNP+/keNGjU1Ioa3Z9uinkjY0ST1An4D\nfD0i3k9yRiIiQlLR2S0iJgATAIYPHx4jR47soGg7VmNjI501tix09euf+XZTeQeYMbVLX3+58v79\njx9f3v5jDy3v97+wSn7/uY3CkrQJSfK4JSLuTIv/Lqlfur4fsDAtnw/sULB7/7TMzMxyktcoLAHX\nA7Mi4ocFq+4BxqXvxwF3F5SfIKmnpIHAIGBKVvGamdmH5dWEdQDwReAlSdPTsguAK4DbJZ0OvAGM\nBYiIlyXdDswkGcF1dkSszj5sMzNrkUsCiYj/AbSe1aPXs8/lwOUVC8rMzIriO9HNzKwkTiBmZlYS\nJxAzMytyTk32AAARXElEQVSJE4iZmZXECcTMzEriBGJmZiVxAjEzs5LkOheWWV7Kngvpqx0ShlmX\n5hqImZmVxAnEzMxK4gRiZmYlcQIxM7OSOIGYmVlJnEDMzKwkTiBmZlYSJxAzMyuJE4iZmZXECcTM\nzEriqUxqVLlTeZS7v5l1fa6BmJlZSVwDsZLMfLsp7xDMLGeugZiZWUmcQMzMrCRdqglL0hHA1UB3\n4BcRcUWe8bgZx8xqWZdJIJK6A/8NHArMA56VdE9EzCz1mH6okJlZ6bpMAgH2AeZExFwASZOBo4GS\nE4hZrSr/y5Nr3waKiLxjaBdJxwNHRMS/pstfBEZExFdbbXcGcEa6OBh4JdNA268P8HbeQeTI1+/r\n9/V3TjtFxLbt2bAr1UDaJSImABPyjmNjJD0XEcPzjiMvvn5fv6+/619/VxqFNR/YoWC5f1pmZmY5\n6EoJ5FlgkKSBkjYFTgDuyTkmM7Oa1WWasCKiWdJXgd+TDOP9ZUS8nHNY5ej0zWwV5uuvbb7+KtBl\nOtHNzKxz6UpNWGZm1ok4gZiZWUmcQKziJHWXdFXeceRJktoo65lHLFmT1E3S2LzjsI7nBJIhSX0l\nHZW+PpJ3PFmJiNXAgXnHkbPrCxck9QLuzymWTEXEGuBbeceRJ0n1ki6UdF26PEjSUXnHVS4nkIyk\n38CmAJ8DxgJ/TO+urxXTJN0j6YuSjm155R1UhuZJ+imApK2Ah4Cb8w0pU49IOk/SDpK2bnnlHVSG\nbgBWAPuly/OBy/ILp2N4FFZGJL0AHBoRC9PlbYFHImKPfCPLhqQb2iiOiDgt82ByIun7QG9gL+CK\niPhNziFlRtJrbRRHRHw082By0HLnuaRpEbFnWvZCV///32XuA6kC3VqSR2oRNVQDjIhT844hD61q\nWX8ELiSpiYakYyPiznwiy1ZEDMw7hpytlLQZEACSdiapkXRpTiDZeVDS74Hb0uXPAw/kGE+mJO0C\nXAv0jYihkoYBYyKiy1fjN+IzrZanAZuk5QHURAKRVA/8G7BjRJwhaRAwOCLuzTm0rIwHHgR2kHQL\ncADQ5b9UuQkrQ5KOI/mHA/BkRPw2z3iyJOkPwDeBnxdU4WdExNB8I7MsSPoVMBU4Of0CUQ88FRGf\nyDm0zEjaBtgXEPBMRHTW2XjbrWaaUDqDtM17PEnn2R9qrBOxPiKmtCprziWSHEi6UdKWBctbSfpl\nnjFlbOeI+D6wCiAilpL8Ia0Jkh6NiEURcV9E3BsRb0t6NO+4yuUmrIxIOhO4GFgOrCH5zxNATXQi\nAm+n7b4tbcDHAwvyDSlTwyLi3ZaFiPiHpD3zDChjVdkHsDGS6oB6oE86+q4lafYGts8tsA7iBJKd\n84Ch1VBtLdHZJBPI7SppPvAacFK+IWWqm6StIuIfAGnts5b+/13Eh/sATsk1omycCXwd2A54vqD8\nfeAnuUTUgdwHkhFJDwLHplX3miVpc5IRaTX1TFRJJwMXAL8m+RZ6PHB5RNyUa2AZqsY+gPaSdE5E\nXJN3HB3NCSQjaXPFDSRDOddW3SPi3NyCyoCkf9vQ+oj4YVax5E3S7sCodPGxiJiZZzxZk7Q9sBMF\nNa+IeCK/iCpP0iER8dj6bprt6sO4a6kKnbefA48BL5H0gdSKhvTnYGBvPngI2GdI7oeoGRHxsqS3\ngDoASTtGxJs5h5UJSf+PZOj6y3zw7z+Aqk4gwD+T/L9vPZwbqmAYt2sgGSm8A7UWSXoCOLKl6UpS\nA3BfRBycb2TZkDQG+AFJW/hCkm/isyJi91wDy4ikV0gGElR9x3ktcQ0kOw9IOgP4Hes2Yb2TX0iZ\n6gusLFhemZbViktJ2v8fiYg9JY2itgYRzCW5gbImE0g68/JxwADWbcK7JK+YOoITSHZOTH+eX1BW\nS8N4JwFTJLXcPHkMcGOO8WRtVUQsSqc27xYRj0v6Ud5BVZqka0j+nS8Fpqf3PtRMH2CBu4H3SG6m\nrJok6iYsy4ykvfhgWvcnImJanvFkSdIjJEnzP4E+JM1Ye0fE/rkGVmGSxm1gdUTEpMyCyVG1zrrg\nBFJh1T4KoxiSupM0WxVW4WulE3lzkptIBXwB2AK4JSIW5RpYRiR9LSKu3lhZtZI0AbgmIl7KO5aO\n5ARSYZIujoiLan06c0nnkNxM9ndgNemd+BExLNfALBOSno+IT7Yqq/qBJZJeImnC6wEMIukLWkGV\n/Pt3AsmIpJ6tR6BI2rpWOtElzQFG1Mo37haSmkin72i9iuQPSO+MQ8qUpBOBfyFpunyyYFUDsCYi\nRucSWEYk7bSh9RHxRlaxVII70bNzp6SjI6IZQNI/AfeRPFyoFvyFpBOxpkREw8a3qmpPkcx51odk\nGHOLJuDFXCLK1jdIPoP/jYj5eQfT0VwDyYikLwGfJpnCYgeSG+rOi4iHcg0sI5KuJ7mZ8D7WHYVT\nM3eiw9rnYgwBXq+xqTw+SnLdADMjYm6e8WRF0leB/dMXJMnkKeB/gRfS58V3WU4gGZJ0NnAEyVjw\nMyPiqXwjyo6ki9oqj4iLs44lS+kNhD8G3gG+B/w3ST/QAODbEVHVQ5kl9QZ+QVLTfiEt/gTJcNbT\nI+L9vGLLmqTt+CCZjAE+0tWbMJ1AKqzVXFACTiapuk+D2vsGXmskvQB8jmTU1eMkd2PPlfQR4NGI\n+HiuAVaYpInA68AlLd+2JYnk0b4fi4iT84suG+n1fpwkcRxAUhN7C3i6q3+Bch9I5bVuA79zPeVV\nTdK2wLeA3UnnggKIiENyCyobayLizwCSXmtpuomIhZJq4YFaB0TEKYUFkXxrvUTS7HxCyo6kh0me\n/TEdeAb4j4iYlW9UHccJpMLa+oYhqRvQq5aq78AtwK+Ao4CzgHEk38KqXbf0QULdgDWtHipU608E\nrYUnEs4FhpEM4V1E8mC1t6ql/8tNWBmRdCvJH87VwLMk30qujogrcw0sI5KmRsRekl5sGfsu6dmI\n2Dvv2CpJ0ut88ATK1iIiqnoqG0k3Aq8Cl0bBHxtJFwK7RMQXcwsuQ2lf0L4kzVj7AtsCMyJiQ3fq\nd3qugWRnSES8L+kLwAPAd0g6EmsigZA+CxtYIOlI4K9A1T8TPiIG5B1Dzs4BrgfmSJqeln2CpA/w\n9Nyiyt4KkvnAlqXv+wOb5hpRB3ACyc4mkjYhmQ/pJxGxSlItVf8uk7QF8O/ANSQ1sG/kG5JVWtpM\n+7n0GeiFw3hfzTGszEj6L5JaxyCSpPk08DNgXES8m2dsHcEJJDs/JxmN8gLwRHqHas30gUTEvenb\n9/jgqXxWI9KEURNJo5XXgJuB6RGxOu9gOpr7QHKSDu3r3nJnerUqmM67TTU0nbdZ1XENpMIk/Sgi\nvp6+Xzv7aESEpF8Ap+QZXwaeyzsAM6sM10AqrHAW0tYzkrY1Q2ktqNFhzGZVp9bHoWdB63lfUyTd\nKql3+lyMGcBMSd/MOy4zK50TSOV1k7SVpG0K3m8taWuge97BZWhIWuM4hmQY80CgJu4BMKtW7gOp\nvC1I7vdoqX08X7CultoPa30Ys1nVcQKpMN9ItlZND2M2q0buRM+BpPERMT7vOPImqUe1D2M2q2ZO\nIDmopdFXraaz/xBPZ2/WdbkJKx+1NBqrpqatN6slroHkQFK3rv4oSzMzJ5CMpA9U+hLJo0zX1vwi\n4rS8YspLLTXhmVUzN2Fl527gSeARkmeC1LJaasIzq1pOINmpj4hv5x1EJ3Ff3gGYWfl8J3p27pX0\n6byDyIukgZLqACLie5I2kzQg36jMrBzuA8mIpCZgc5Knka0iacaJiOida2AZkfQcsH9ErEyXNwX+\nt9ofaWtWzdyElZGIqPXhrD1akgdARKxMk4iZdVFOIBmStBXJoy3rWsoi4on8IsrUW5LGRMQ9AJKO\nBt7OOSYzK4ObsDIi6V+BrwH9genAvsDTEXFIroFlJH0m9i3AdiTNd38BTo6IObkGZmYlcwLJiKSX\ngL2BZyLiE5J2Bf4jIo7NObRMSeoFEBGL847FzMrjJqzsLI+I5ZKQ1DMi/iRpcN5BZUnSkcDuQF3y\nSHiIiEtyDcrMSuYEkp15krYE7gIelvQP4I2cY8qMpJ8B9cAo4BfA8cCUXIMys7K4CSsHkv6Z5EFT\nDxaOTKpmkl6MiGEFP3sBD0TEQXnHZmalcQ2kwiT1joj300fYtngp/dkLeCeHsPKwLP25VNJ2wCKg\nX47xmFmZnEAq71bgKJLH2gbrzgMVwEfzCCoH96ZNeFeSPNY3gOvyDcnMyuEmLMucpJ5AXUS8l3cs\nZlY6J5AKk7TBacsj4vmsYsmDpEHAVcDOJE1350XE/HyjMrOO4ARSYZIeT9/WAcOBF0iasYYBz0XE\nfnnFlgVJTwKTgCeAMcB+tXbvi1m18my8FRYRoyJiFLAA+GREDI+IvYA9gVr4Jt4QEddFxCsRcSXJ\nA7XMrAq4Ez07gyOiZfQVETFD0m55BpSROkl78sHggc0Kl6u9Cc+smrkJKyOSbgOWADenRV8AekXE\niflFVXkFTXhtiVqZC8ysGjmBZCR9mNKXgYPToieAayNieX5RmZmVzgnEMidpQkSckXccZlYe94Fk\nRNIBwHhgJwo+94iolRsJCw3POwAzK58TSHauB75Bckf66pxjydvCvAMws/K5CSsjkv4YESPyjsPM\nrKM4gWRE0hVAd+BOYEVLea0MY5W0C/BNPtyE51FYZl2UE0hG1jOctWaGsUp6AfgZrZrwImJqbkGZ\nWVmcQCwTkqamd+CbWZVwAqkwSf+2ofUR8cOsYsmTpPEknee/Zd0mvFp5HopZ1fEorMpryDuATmJc\n+vObBWW19DwUs6rjGoiZmZXENZAcSHo+Ijb4nJBqJGkoMIRkansAImJSfhGZWTmcQPKhjW9SXSRd\nBIwkSSD3A/8H+B+SZ4WYWRfk54Hk4768A8jB8cBo4G8RcSqwB7BFviGZWTmcQDIgqXvhfSAR8b08\n48nJsohYAzRL6k0yImuHnGMyszK4CSsDEbFa0hpJW0TEe3nHk5PnJG0JXEdyM+Fi4Ol8QzKzcngU\nVkYk3U3yGNuHSR4sBUBEnJtbUDmRNADoHREv5hyKmZXBCSQjksa1VR4RN2YdS5Yk7RoRf5LU5qiz\nWpkLzKwaOYFkSNJmwI4R8UresWSl5eFRtT4XmFk1cgLJiKTPAFcBm0bEQEmfAC6JiDE5h2ZmVhKP\nwsrOeGAf4F2AiJhODU3jIelzkhrS99+TdKekPfOOy8xK5wSSnVVtjMBak0sk+bgwIpokHQh8iuQJ\njT/LOSYzK4MTSHZelvQvQHdJgyRdAzyVd1AZankGyJHAhIi4D9g0x3jMrExOINk5B9idZCrz24D3\nga/nGlG25kv6OfB54H5JPfG/P7MuzZ3olglJ9cARwEsRMVtSP2BoRDycc2hmViInkIxIGg5cAAxg\n3WeCD8srprxI2hw4FjghIo7MOx4zK42nMsnOLSQPU3qJ2uo8B0DSpiT9H/8CHA78Bneim3VpTiDZ\neSsi7sk7iKxJOgw4ETgMeJxk+va90xl5zawLcxNWRiSNJvlD+ijrPhP8ztyCyoCkNcCTwCkR8Vpa\nNjciauYeGLNq5RpIdk4FdgU24YMmrACqOoEAnwROAB6RNBeYDHTPNyQz6wiugWRE0isRMTjvOPIk\naX+SWthxwAvAbyNiQr5RmVmpPA4/O09JGpJ3EHmKiKci4hygP/BfwL45h2RmZXANJCOSZgE7A6+R\n9IGIZDbamhvGa2bVwQkkI5J2aqs8It7IOhYzs47gBJKD9Ea6zwIn+kY6M+uq3AeSEUmbSvqspF8D\nC4DR+EY6M+vCXAOpsDZupPsVcE1EDMgzLjOzcjmBVJhvpDOzauUbCSvPN9KZWVVyDSRDvpHOzKqJ\nE0gOJHUjeazrCRFxWt7xmJmVwgnEzMxK4mG8ZmZWEicQMzMriROImZmVxAnEzMxK4gRiZmYlcQIx\nM7OSOIGYmVlJ/j9l7l3mgcwqBwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xd86a588>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYMAAAFHCAYAAABDHSCwAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xd86a080>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYMAAAFFCAYAAAAO1YG7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nzuGFgXPOObwwcM45hxcGzjnngP8PiDPPCTMXTacAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xdaf0be0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for key , value in top5_views.items():\n",
    "    a = key[0]\n",
    "    m = key[1]\n",
    "    f = key[2]\n",
    "#     print key , value\n",
    "    target_command =\"select {},{}({}) from a_{} where marital_status IN (' Married-AF-spouse', ' Married-civ-spouse', ' Married-spouse-absent', ' Separated') group  by {}\".format(a,f,m,0,a)\n",
    "    reference_command = \"select {},{}({}) from a_{} WHERE marital_status IN (' Never-married', ' Widowed', ' Divorced') group  by {}\".format(a,f,m,0,a)\n",
    "    target_array = execute_query(conn,target_command)\n",
    "    reference_array = execute_query(conn,reference_command)\n",
    "#     print target_array\n",
    "#     print reference_array\n",
    "    draw_viz(conn, target_array, reference_array ,\"Married vs Unmarried\", f + \"(\"+ m +\")\", a)\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}