lab_regression_PCA.ipynb

{
 "metadata": {
  "name": "",
  "signature": "sha256:88140ec775550b11bdb480c6f05c1af3ed4f0e3e08c97c826d91b680ecdf9085"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#Summary\n",
      "- 3 interfaces of scikit-learn\n",
      "    - estimator: building and fitting models\n",
      "    - predictor: making predictions   \n",
      "            clf = LogisticRegression()\n",
      "            clf.fit(X_train, y_train)\n",
      "            clf.predict(X_test)\n",
      "    - transformer: converting data\n",
      "            pca = RandomizedPCA(n_components=2)\n",
      "            train_x = pca.fit_transform(train_x)\n",
      "            test_x = pca.transform(test_x)\n",
      "    - meta-estimator: allows us to wrap CV and other methods to build simpler modesl\n",
      "            from sklearn.multiclass import OneVsOneClassifier\n",
      "            clf=OneVsOneClassifier(LogisticRegression())\n",
      "    - model selection: using 2 distinct meta-estimators (GridSearchCV and RandomizedSearchCV\n",
      "- All data has to be stored in 2-D array (numpy arrays) to be [n_samples, n_features]"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%matplotlib inline\n",
      "import pandas as pd\n",
      "import numpy as np\n",
      "import matplotlib.pyplot as plt\n",
      "from matplotlib import rcParams\n",
      "\n",
      "pd.set_option('display.width', 500)\n",
      "pd.set_option('display.max_columns', 100)\n",
      "rcParams['figure.figsize'] = (10, 6)\n",
      "plt.style.use('bmh')\n",
      "\n",
      "colors = [\"#348ABD\", \"#A60628\", \"#7A68A6\", \"#467821\", \"#CF4457\", \"#188487\", \"#E24A33\"]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 1
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "states_abbrev_dict = {\n",
      "        'AK': 'Alaska',\n",
      "        'AL': 'Alabama',\n",
      "        'AR': 'Arkansas',\n",
      "        'AS': 'American Samoa',\n",
      "        'AZ': 'Arizona',\n",
      "        'CA': 'California',\n",
      "        'CO': 'Colorado',\n",
      "        'CT': 'Connecticut',\n",
      "        'DC': 'District of Columbia',\n",
      "        'DE': 'Delaware',\n",
      "        'FL': 'Florida',\n",
      "        'GA': 'Georgia',\n",
      "        'GU': 'Guam',\n",
      "        'HI': 'Hawaii',\n",
      "        'IA': 'Iowa',\n",
      "        'ID': 'Idaho',\n",
      "        'IL': 'Illinois',\n",
      "        'IN': 'Indiana',\n",
      "        'KS': 'Kansas',\n",
      "        'KY': 'Kentucky',\n",
      "        'LA': 'Louisiana',\n",
      "        'MA': 'Massachusetts',\n",
      "        'MD': 'Maryland',\n",
      "        'ME': 'Maine',\n",
      "        'MI': 'Michigan',\n",
      "        'MN': 'Minnesota',\n",
      "        'MO': 'Missouri',\n",
      "        'MP': 'Northern Mariana Islands',\n",
      "        'MS': 'Mississippi',\n",
      "        'MT': 'Montana',\n",
      "        'NA': 'National',\n",
      "        'NC': 'North Carolina',\n",
      "        'ND': 'North Dakota',\n",
      "        'NE': 'Nebraska',\n",
      "        'NH': 'New Hampshire',\n",
      "        'NJ': 'New Jersey',\n",
      "        'NM': 'New Mexico',\n",
      "        'NV': 'Nevada',\n",
      "        'NY': 'New York',\n",
      "        'OH': 'Ohio',\n",
      "        'OK': 'Oklahoma',\n",
      "        'OR': 'Oregon',\n",
      "        'PA': 'Pennsylvania',\n",
      "        'PR': 'Puerto Rico',\n",
      "        'RI': 'Rhode Island',\n",
      "        'SC': 'South Carolina',\n",
      "        'SD': 'South Dakota',\n",
      "        'TN': 'Tennessee',\n",
      "        'TX': 'Texas',\n",
      "        'UT': 'Utah',\n",
      "        'VA': 'Virginia',\n",
      "        'VI': 'Virgin Islands',\n",
      "        'VT': 'Vermont',\n",
      "        'WA': 'Washington',\n",
      "        'WI': 'Wisconsin',\n",
      "        'WV': 'West Virginia',\n",
      "        'WY': 'Wyoming'\n",
      "}\n",
      "abbrev_states_dict = {v: k for k, v in states_abbrev_dict.items()}"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 2
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#Linear Regression with PCA"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "census_data = pd.read_csv('data/lab_scikit_learn/census_demographics.csv')\n",
      "census_data.head()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th></th>\n",
        "      <th>state</th>\n",
        "      <th>per_black</th>\n",
        "      <th>per_hisp</th>\n",
        "      <th>per_white</th>\n",
        "      <th>educ_hs</th>\n",
        "      <th>educ_coll</th>\n",
        "      <th>average_income</th>\n",
        "      <th>median_income</th>\n",
        "      <th>pop_density</th>\n",
        "      <th>vote_pop</th>\n",
        "      <th>older_pop</th>\n",
        "      <th>per_older</th>\n",
        "      <th>per_vote</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>0</th>\n",
        "      <td>    ALABAMA</td>\n",
        "      <td> 26.5</td>\n",
        "      <td>  4.0</td>\n",
        "      <td> 66.8</td>\n",
        "      <td> 81.4</td>\n",
        "      <td> 21.7</td>\n",
        "      <td> 22984</td>\n",
        "      <td> 42081</td>\n",
        "      <td>  94.4</td>\n",
        "      <td>  3001712.500</td>\n",
        "      <td>  672383.600</td>\n",
        "      <td> 0.140</td>\n",
        "      <td> 0.625</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>1</th>\n",
        "      <td>     ALASKA</td>\n",
        "      <td>  3.6</td>\n",
        "      <td>  5.8</td>\n",
        "      <td> 63.7</td>\n",
        "      <td> 90.7</td>\n",
        "      <td> 27.0</td>\n",
        "      <td> 30726</td>\n",
        "      <td> 66521</td>\n",
        "      <td>   1.2</td>\n",
        "      <td>   475548.444</td>\n",
        "      <td>   58540.158</td>\n",
        "      <td> 0.081</td>\n",
        "      <td> 0.658</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>2</th>\n",
        "      <td>    ARIZONA</td>\n",
        "      <td>  4.5</td>\n",
        "      <td> 30.1</td>\n",
        "      <td> 57.4</td>\n",
        "      <td> 85.0</td>\n",
        "      <td> 26.3</td>\n",
        "      <td> 25680</td>\n",
        "      <td> 50448</td>\n",
        "      <td>  56.3</td>\n",
        "      <td>  3934880.535</td>\n",
        "      <td>  920515.710</td>\n",
        "      <td> 0.142</td>\n",
        "      <td> 0.607</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>3</th>\n",
        "      <td>   ARKANSAS</td>\n",
        "      <td> 15.6</td>\n",
        "      <td>  6.6</td>\n",
        "      <td> 74.2</td>\n",
        "      <td> 81.9</td>\n",
        "      <td> 19.1</td>\n",
        "      <td> 21274</td>\n",
        "      <td> 39267</td>\n",
        "      <td>  56.0</td>\n",
        "      <td>  1798043.148</td>\n",
        "      <td>  428944.934</td>\n",
        "      <td> 0.146</td>\n",
        "      <td> 0.612</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>4</th>\n",
        "      <td> CALIFORNIA</td>\n",
        "      <td>  6.6</td>\n",
        "      <td> 38.1</td>\n",
        "      <td> 39.7</td>\n",
        "      <td> 80.7</td>\n",
        "      <td> 30.1</td>\n",
        "      <td> 29188</td>\n",
        "      <td> 60883</td>\n",
        "      <td> 239.1</td>\n",
        "      <td> 24009747.944</td>\n",
        "      <td> 4409953.704</td>\n",
        "      <td> 0.117</td>\n",
        "      <td> 0.637</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 3,
       "text": [
        "        state  per_black  per_hisp  per_white  educ_hs  educ_coll  average_income  median_income  pop_density      vote_pop    older_pop  per_older  per_vote\n",
        "0     ALABAMA       26.5       4.0       66.8     81.4       21.7           22984          42081         94.4   3001712.500   672383.600      0.140     0.625\n",
        "1      ALASKA        3.6       5.8       63.7     90.7       27.0           30726          66521          1.2    475548.444    58540.158      0.081     0.658\n",
        "2     ARIZONA        4.5      30.1       57.4     85.0       26.3           25680          50448         56.3   3934880.535   920515.710      0.142     0.607\n",
        "3    ARKANSAS       15.6       6.6       74.2     81.9       19.1           21274          39267         56.0   1798043.148   428944.934      0.146     0.612\n",
        "4  CALIFORNIA        6.6      38.1       39.7     80.7       30.1           29188          60883        239.1  24009747.944  4409953.704      0.117     0.637"
       ]
      }
     ],
     "prompt_number": 3
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def capitalize(s):\n",
      "    s = s.title()\n",
      "    s = s.replace(\"Of\", 'of')\n",
      "    return s\n",
      "\n",
      "census_data['State'] = census_data.state.map(capitalize)\n",
      "del census_data['state']\n",
      "census_data['State'] = census_data['State'].replace(abbrev_states_dict)\n",
      "census_data.set_index('State', inplace=True)\n",
      "census_data.head()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th></th>\n",
        "      <th>per_black</th>\n",
        "      <th>per_hisp</th>\n",
        "      <th>per_white</th>\n",
        "      <th>educ_hs</th>\n",
        "      <th>educ_coll</th>\n",
        "      <th>average_income</th>\n",
        "      <th>median_income</th>\n",
        "      <th>pop_density</th>\n",
        "      <th>vote_pop</th>\n",
        "      <th>older_pop</th>\n",
        "      <th>per_older</th>\n",
        "      <th>per_vote</th>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>State</th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>AL</th>\n",
        "      <td> 26.5</td>\n",
        "      <td>  4.0</td>\n",
        "      <td> 66.8</td>\n",
        "      <td> 81.4</td>\n",
        "      <td> 21.7</td>\n",
        "      <td> 22984</td>\n",
        "      <td> 42081</td>\n",
        "      <td>  94.4</td>\n",
        "      <td>  3001712.500</td>\n",
        "      <td>  672383.600</td>\n",
        "      <td> 0.140</td>\n",
        "      <td> 0.625</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>AK</th>\n",
        "      <td>  3.6</td>\n",
        "      <td>  5.8</td>\n",
        "      <td> 63.7</td>\n",
        "      <td> 90.7</td>\n",
        "      <td> 27.0</td>\n",
        "      <td> 30726</td>\n",
        "      <td> 66521</td>\n",
        "      <td>   1.2</td>\n",
        "      <td>   475548.444</td>\n",
        "      <td>   58540.158</td>\n",
        "      <td> 0.081</td>\n",
        "      <td> 0.658</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>AZ</th>\n",
        "      <td>  4.5</td>\n",
        "      <td> 30.1</td>\n",
        "      <td> 57.4</td>\n",
        "      <td> 85.0</td>\n",
        "      <td> 26.3</td>\n",
        "      <td> 25680</td>\n",
        "      <td> 50448</td>\n",
        "      <td>  56.3</td>\n",
        "      <td>  3934880.535</td>\n",
        "      <td>  920515.710</td>\n",
        "      <td> 0.142</td>\n",
        "      <td> 0.607</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>AR</th>\n",
        "      <td> 15.6</td>\n",
        "      <td>  6.6</td>\n",
        "      <td> 74.2</td>\n",
        "      <td> 81.9</td>\n",
        "      <td> 19.1</td>\n",
        "      <td> 21274</td>\n",
        "      <td> 39267</td>\n",
        "      <td>  56.0</td>\n",
        "      <td>  1798043.148</td>\n",
        "      <td>  428944.934</td>\n",
        "      <td> 0.146</td>\n",
        "      <td> 0.612</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>CA</th>\n",
        "      <td>  6.6</td>\n",
        "      <td> 38.1</td>\n",
        "      <td> 39.7</td>\n",
        "      <td> 80.7</td>\n",
        "      <td> 30.1</td>\n",
        "      <td> 29188</td>\n",
        "      <td> 60883</td>\n",
        "      <td> 239.1</td>\n",
        "      <td> 24009747.944</td>\n",
        "      <td> 4409953.704</td>\n",
        "      <td> 0.117</td>\n",
        "      <td> 0.637</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 4,
       "text": [
        "       per_black  per_hisp  per_white  educ_hs  educ_coll  average_income  median_income  pop_density      vote_pop    older_pop  per_older  per_vote\n",
        "State                                                                                                                                                \n",
        "AL          26.5       4.0       66.8     81.4       21.7           22984          42081         94.4   3001712.500   672383.600      0.140     0.625\n",
        "AK           3.6       5.8       63.7     90.7       27.0           30726          66521          1.2    475548.444    58540.158      0.081     0.658\n",
        "AZ           4.5      30.1       57.4     85.0       26.3           25680          50448         56.3   3934880.535   920515.710      0.142     0.607\n",
        "AR          15.6       6.6       74.2     81.9       19.1           21274          39267         56.0   1798043.148   428944.934      0.146     0.612\n",
        "CA           6.6      38.1       39.7     80.7       30.1           29188          60883        239.1  24009747.944  4409953.704      0.117     0.637"
       ]
      }
     ],
     "prompt_number": 4
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "df = census_data[['educ_coll', 'average_income', 'per_vote']]\n",
      "df.head()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th></th>\n",
        "      <th>educ_coll</th>\n",
        "      <th>average_income</th>\n",
        "      <th>per_vote</th>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>State</th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>AL</th>\n",
        "      <td> 21.7</td>\n",
        "      <td> 22984</td>\n",
        "      <td> 0.625</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>AK</th>\n",
        "      <td> 27.0</td>\n",
        "      <td> 30726</td>\n",
        "      <td> 0.658</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>AZ</th>\n",
        "      <td> 26.3</td>\n",
        "      <td> 25680</td>\n",
        "      <td> 0.607</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>AR</th>\n",
        "      <td> 19.1</td>\n",
        "      <td> 21274</td>\n",
        "      <td> 0.612</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>CA</th>\n",
        "      <td> 30.1</td>\n",
        "      <td> 29188</td>\n",
        "      <td> 0.637</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 5,
       "text": [
        "       educ_coll  average_income  per_vote\n",
        "State                                     \n",
        "AL          21.7           22984     0.625\n",
        "AK          27.0           30726     0.658\n",
        "AZ          26.3           25680     0.607\n",
        "AR          19.1           21274     0.612\n",
        "CA          30.1           29188     0.637"
       ]
      }
     ],
     "prompt_number": 5
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "print(plt.style.available)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "[u'dark_background', u'bmh', u'grayscale', u'ggplot', u'fivethirtyeight']\n"
       ]
      }
     ],
     "prompt_number": 6
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from pandas.tools.plotting import scatter_matrix\n",
      "with plt.style.context('ggplot'):\n",
      "    axeslist = scatter_matrix(df[['educ_coll', 'average_income', 'per_vote']], alpha=1, c='k', s=150, figsize=(10, 10), diagonal='kde')\n",
      "    for ax in axeslist.flatten():\n",
      "        ax.grid(False)\n",
      "    df.corr()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAm8AAAJzCAYAAABZIvRoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x107925290>"
       ]
      }
     ],
     "prompt_number": 7
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#Important note (scikit-learn format)\n",
      "- $n_{samples}$ X $n_{features}$\n",
      "- X needs to be a matrix (51,1)\n",
      "- y needs to be a vector (51, )"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#Linear Regression"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from sklearn.cross_validation import train_test_split\n",
      "from sklearn.linear_model import LinearRegression\n",
      "\n",
      "X_HD = df[['educ_coll', 'average_income']].values\n",
      "\n",
      "educ_coll_std_vec = X_HD[:,0]\n",
      "educ_coll_std = educ_coll_std_vec.reshape(-1,1)\n",
      "\n",
      "average_income_std_vec = X_HD[:,1]\n",
      "average_income_std = average_income_std_vec.reshape(-1,1)\n",
      "\n",
      "X = educ_coll_std\n",
      "y = average_income_std_vec\n",
      "\n",
      "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.5)\n",
      " \n",
      "clf = LinearRegression()\n",
      "clf.fit(X_train, y_train)\n",
      "print \"The intercept is: %.5f\" % clf.intercept_\n",
      "print \"The coefficent is: %.5f\" % clf.coef_\n",
      " \n",
      "predicted_train = clf.predict(X_train)\n",
      "predicted_test = clf.predict(X_test)\n",
      "\n",
      "plt.scatter(X, y, s=50, marker=\"o\", c='k', lw=1.5, label='Data')\n",
      "plt.plot(X_train, predicted_train, c=colors[3], alpha=.8, ls=\"-\", lw=3, label='Training')\n",
      "plt.plot(X_test, predicted_test, c=colors[0], alpha=.8, ls=\"-\", lw=3, label='Test')\n",
      "plt.xlabel(\"educ_coll\")\n",
      "plt.ylabel(\"average_income\")\n",
      "plt.legend(loc='best')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "The intercept is: 4422.38214\n",
        "The coefficent is: 833.63860\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 8,
       "text": [
        "<matplotlib.legend.Legend at 0x10a361f90>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAnYAAAGCCAYAAABtv45rAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x10347da10>"
       ]
      }
     ],
     "prompt_number": 8
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plt.scatter(predicted_train, predicted_train - y_train, c=colors[3], alpha=.8, s=40, label='Train')\n",
      "plt.scatter(predicted_test, predicted_test - y_test, c=colors[0], alpha=.8, s=40, label='Test')\n",
      "plt.axhline(y=0, lw=4, color='k')\n",
      "plt.xlabel(\"Fitted Value\")\n",
      "plt.ylabel(\"Residual\")\n",
      "plt.legend(loc='best')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 9,
       "text": [
        "<matplotlib.legend.Legend at 0x10a571c90>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAoEAAAGCCAYAAACfLxmLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nREQkgf4fbmYMzs72JH4AAAAASUVORK5CYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0x107eac3d0>"
       ]
      }
     ],
     "prompt_number": 9
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#Using Standardized Variables"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from sklearn.cross_validation import train_test_split\n",
      "from sklearn.linear_model import LinearRegression\n",
      "\n",
      "X_HD = df[['educ_coll', 'average_income']].values\n",
      "X_HDn = (X_HD - X_HD.mean(axis=0)) / X_HD.std(axis=0)\n",
      "\n",
      "educ_coll_std_vec = X_HDn[:,0]\n",
      "educ_coll_std = educ_coll_std_vec.reshape(-1,1)\n",
      "\n",
      "average_income_std_vec = X_HDn[:,1]\n",
      "average_income_std = average_income_std_vec.reshape(-1,1)\n",
      "\n",
      "X = educ_coll_std\n",
      "y = average_income_std_vec\n",
      "\n",
      "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.5)\n",
      " \n",
      "clf = LinearRegression()\n",
      "clf.fit(X_train, y_train)\n",
      "print \"The intercept is: %.5f\" % clf.intercept_\n",
      "print \"The coefficent is: %.5f\" % clf.coef_\n",
      " \n",
      "predicted_train = clf.predict(X_train)\n",
      "predicted_test = clf.predict(X_test)\n",
      "\n",
      "plt.scatter(X, y, s=50, marker=\"o\", c='k', label='Data')\n",
      "plt.plot(X_train, predicted_train, c=colors[0], alpha=.8, ls=\"-\", lw=3, label='Training')\n",
      "plt.plot(X_test, predicted_test, c=colors[3], alpha=.8, ls=\"-\", lw=3, label='Test')\n",
      "plt.xlabel(\"educ_coll_std\")\n",
      "plt.ylabel(\"average_income_std\")\n",
      "plt.legend(loc='best')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "The intercept is: 0.06816\n",
        "The coefficent is: 0.84219\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 10,
       "text": [
        "<matplotlib.legend.Legend at 0x10a752390>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAmEAAAGCCAYAAACozRT6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x1090dcc90>"
       ]
      }
     ],
     "prompt_number": 10
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "#Residual Plot\n",
      "plt.scatter(predicted_train, predicted_train - y_train, c=colors[0], alpha=.8, s=40, label='Train')\n",
      "plt.scatter(predicted_test, predicted_test - y_test, c=colors[3], alpha=.8, s=40, label='Test')\n",
      "plt.axhline(y=0, lw=4, color='k')\n",
      "plt.xlabel(\"Fitted Value\")\n",
      "plt.ylabel(\"Residual\")\n",
      "plt.legend(loc='best')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 11,
       "text": [
        "<matplotlib.legend.Legend at 0x10a906b50>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAmsAAAGCCAYAAAC/74QzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\no1gDAAAIMYo1AACAEPv/FjLiCVK8C6UAAAAASUVORK5CYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0x10a5c8590>"
       ]
      }
     ],
     "prompt_number": 11
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "print \"Training score:\", clf.score(X_train, y_train)\n",
      "print \"Testing score:\", clf.score(X_test, y_test)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Training score: 0.729866245512\n",
        "Testing score: 0.846952137308\n"
       ]
      }
     ],
     "prompt_number": 12
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#Doing a PCA on the data"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from sklearn.decomposition import PCA\n",
      "\n",
      "pca = PCA(n_components=2)\n",
      "X = pca.fit_transform(X_HDn)\n",
      "\n",
      "print \"Variance Explained:\", pca.explained_variance_ratio_\n",
      "print \"Variance Explained by 1st component: \", pca.explained_variance_ratio_[0]\n",
      "print \"Variance Explained by 2nd component: \", pca.explained_variance_ratio_[1]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Variance Explained: [ 0.94703283  0.05296717]\n",
        "Variance Explained by 1st component:  0.947032827533\n",
        "Variance Explained by 2nd component:  0.0529671724671\n"
       ]
      }
     ],
     "prompt_number": 13
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "###Take the X which is transformed and plot into PCA space"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plt.scatter(X[:, 0], X[:, 1], s=50, marker=\"o\", c='k', label='PCA')\n",
      "plt.legend(loc='best')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 14,
       "text": [
        "<matplotlib.legend.Legend at 0x10a920b90>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAlkAAAFwCAYAAACLs24kAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x10a92e350>"
       ]
      }
     ],
     "prompt_number": 14
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "###Since 1st component explain most of the variance, drop the 2nd component"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "pca1 = PCA(n_components=1) # keep only the 1st component\n",
      "X_E = pca1.fit_transform(X_HDn)\n",
      "X_reconstructed = pca1.inverse_transform(X_E)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 15
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plt.scatter(X_reconstructed[:,0], X_reconstructed[:,1], c=colors[1], alpha=.8, s=40, label='Reconstructed (PCA)')\n",
      "plt.scatter(educ_coll_std_vec, average_income_std_vec, c='k', alpha=.8, s=40, label='Data')\n",
      "plt.plot(X_train, predicted_train, c=colors[0], alpha=.8, ls=\"-\", lw=3, label='Training')\n",
      "plt.plot(X_test, predicted_test, c=colors[3], alpha=.8, ls=\"-\", lw=3, label='Test')\n",
      "plt.legend(loc='best')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 16,
       "text": [
        "<matplotlib.legend.Legend at 0x10acbccd0>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAk8AAAFwCAYAAAChA5+LAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJk9EREREEWDyRERERBQBJk9EREREEfh/Pz+Nn5bIRzoAAAAASUVORK5CYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0x109080e90>"
       ]
      }
     ],
     "prompt_number": 16
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "**The red line (reconstructed from PCA) is steeper because, the principle behind this is that regression minimizes the y-residuals while the PCA minimizes the norms (trying to get the greatest variance of the set)**"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from IPython.display import Image as Im\n",
      "from IPython.display import display\n",
      "Im('./data/lab_scikit_learn/pcavsfit.png')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAeAAAAHgCAIAAADytinCAAAC7mlDQ1BJQ0MgUHJvZmlsZQAAeAGF\nVM9rE0EU/jZuqdAiCFprDrJ4kCJJWatoRdQ2/RFiawzbH7ZFkGQzSdZuNuvuJrWliOTi0SreRe2h\nB/+AHnrwZC9KhVpFKN6rKGKhFy3xzW5MtqXqwM5+8943731vdt8ADXLSNPWABOQNx1KiEWlsfEJq\n/IgAjqIJQTQlVdvsTiQGQYNz+Xvn2HoPgVtWw3v7d7J3rZrStpoHhP1A4Eea2Sqw7xdxClkSAog8\n36Epx3QI3+PY8uyPOU55eMG1Dys9xFkifEA1Lc5/TbhTzSXTQINIOJT1cVI+nNeLlNcdB2luZsbI\nEL1PkKa7zO6rYqGcTvYOkL2d9H5Os94+wiHCCxmtP0a4jZ71jNU/4mHhpObEhj0cGDX0+GAVtxqp\n+DXCFF8QTSeiVHHZLg3xmK79VvJKgnCQOMpkYYBzWkhP10xu+LqHBX0m1xOv4ndWUeF5jxNn3tTd\n70XaAq8wDh0MGgyaDUhQEEUEYZiwUECGPBoxNLJyPyOrBhuTezJ1JGq7dGJEsUF7Ntw9t1Gk3Tz+\nKCJxlEO1CJL8Qf4qr8lP5Xn5y1yw2Fb3lK2bmrry4DvF5Zm5Gh7X08jjc01efJXUdpNXR5aseXq8\nmuwaP+xXlzHmgjWPxHOw+/EtX5XMlymMFMXjVfPqS4R1WjE3359sfzs94i7PLrXWc62JizdWm5dn\n/WpI++6qvJPmVflPXvXx/GfNxGPiKTEmdornIYmXxS7xkthLqwviYG3HCJ2VhinSbZH6JNVgYJq8\n9S9dP1t4vUZ/DPVRlBnM0lSJ93/CKmQ0nbkOb/qP28f8F+T3iuefKAIvbODImbptU3HvEKFlpW5z\nrgIXv9F98LZua6N+OPwEWDyrFq1SNZ8gvAEcdod6HugpmNOWls05Uocsn5O66cpiUsxQ20NSUtcl\n12VLFrOZVWLpdtiZ0x1uHKE5QvfEp0plk/qv8RGw/bBS+fmsUtl+ThrWgZf6b8C8/UXAeIuJAABU\ng0lEQVR42u2dB3gURf/HsYFiF3lFBcWuWLG9WP5SfAEFpSNSQkKLIE1EpApEuoQWKdIxdJASQEGk\nSgmBkJAAKSSkkN7LJZdctsz/tzdwHMnlcknuLnt338/Dw3Nlc7szO/PZ787tzdZiAAAAVEktVAEA\nAEDQAAAAIGgAAICgAQAAQNAAAABBAwAAgKABAABA0AAAAEEDAACAoAEAAIIGAAAAQQMAAICgAQAA\nggYAAABBAwAABA0AAACCBgAACBoAAAAEDQAAAIIGAAAIGgAAAAQNAAAQNAAAAAjaMoKDg/2MOHjw\n4KVLlyRJUuGm7t27l7bQnmsMCgqiNSYmJlplMTuXxbAT7V9v1q3qkydP0ls5OTn2bDapqalr1649\nfPiwKIqW9J2//vrrxIkTGo2m1DJpaWm0AQsWLNixY0d6enrZD/H396c/p4+ycG+eP3+elk9KSnJc\nS1ahv7iuoAcMGFCrDK1atcrKylLbpt599920bfY8eLi5udEat2zZYpXF7FaW3NzcUaNGLV68mD+t\nXbs2rUvl7dBMHf7f//0fvXXmzBm7NRtvb+/bbrvtrrvuoj987LHHMjMzTS7Wr1+/Uh2nUaNGR44c\nMSywfPnye++91/DuQw89tHHjRuNPKCoqohfprWeffVaWZZNrodfXr1/frVs3/rRPnz60/LZt2xxX\n0FXoL64u6C+++GLhwoV0nJ8yZQq1SHpl4sSJattU2rYJEyaU145twfbt22mNoaGhVlnMbmUZM2YM\n7UHamw4kaDN1aGdBC4JQv359Eitl9v3799Pfzpgxw4ygv/zyy/nz50+bNu3dd9+lp8888wx/d9eu\nXfSU1v7TTz/9/vvvnTt3pqf0scbhd8OGDQZ908mrybWcO3eO3v3ggw8gaNcV9KxZswyvUPLiIZo/\n1Wq1v/7665AhQ3755RfjtlVQUDBv3rxhw4Zt3br1ypUrs2fPPnr0KL1OTYceX7169ccffyRHlJSU\n0Iv79u377rvvqAfSaaDhE+iMb86cOd988w2tnf7W0IXKe51yDX2yQWq0iqVLl3777bfUPQwnj3xL\njh8/TmeCtLqxY8eGhISULfXZs2dpsYsXL9KGjRgxgj6hsLDw2rVrXl5etNkBAQF8MXqXFgsLC6PH\ne/bsocd05ksNi0pNW5iXl1d2MV9fX3pMb/n4+AwfPnz37t30IpWCik8VSCe8pcpy6tSp2bdC58J8\nmcjIyKlTp9K66FzbcKJND6hXjxw58ocffvjjjz/KnuhQ0T755BN+0CUvGARNuqENoE/buXOnYeHy\ndi4nOzubtscgeoL+ll65fPlyqSVJrPQ61RvtkXHjxtGuqez2G9chQbuGdh8daS5cuGAQNFUXLUNV\nZxg9oKdU4fwp7cE///yT1j569OhVq1ZR0coKurymVYqnnnqK/mTv3r3ff/89PTCs0aSg6QP5U41G\nc88999ArVAraVArF9JgKa1i+devWb7/9tvF4S8uWLfmeov+7du1adhX5+flDhw6ld2mTeLvigqZO\nx9sh9UHqiYblTda5MWaacXl9SqfT8Xo+efIk7Tg6aFnYyHlrpNrjH0h9s6ygK2zPEPQtgqZzrq++\n+ope+fnnn3n7aNKkCT3lLa9evXrcd9T6+evU+elk8K233qLH1J3orc8++4wev/jii/x/eoW38jv1\n0MK0q/iJ2yuvvHL77bfTMg888AAtMHPmTDOvl+pppOC6desa0sejjz7KrcpjC2WZOnXq3HffffSY\nHhhahgFq1vTWG2+8Qaex/GM//vhjOnXgn0mF5VHOuCX16tWLHn/00Uf0gfzk97///W/ZBtesWTP+\nyffffz8Vlh5TdLrjjjv4Wt555x1+gDGUhZJXqdPkTp060QKHDx82bAz9T87l/Y3iGD1t0KABl8h7\n771Hia9s0Tj0rkHQL7zwAq9PYsmSJWZ2rvHJNbdMUFAQHwb9z3/+Q2VJTk4uVZ+rV6827HSCPqey\n229ch8eOHeP1Q/9TbdOGcUHTBvAXDUOZxnuhd+/e9JR2Om0hPSD3lWo2ZppWKXiw5XuZ7Fl2ZNmk\noBMSEqiF0yvR0dH0mG+qmeROi1ELefzxx0mXd+lJSUkptQyFBuO2ERcXxwVNe5M+nO/Z5s2b84XL\nq3NjzDTj8voUFZ+e0nbyDkXn1hY2chI3PaXX+R8++OCDVORS+7rC9gxBD+AWo7qm/cqrmwRHLYPe\nnTRpEj2lCiVxb968mbdXep1SFa9N6qsRERGNGjUqJeimTZv6+/ufPn2aEhB9Ji1AH0itlvYEneVl\nZGTQCR3/ZGrBZAoKNXzMtLzXjXsabQzvYPQWNW6KyfT4pZdeore4oKmdBQcH01M6D+CLmRQ0tQna\nkvj4eN4i6ZhEjYOfivLYaFLQ1F7pT3gf4GGhrKD79+9PfYOyA69b2hhaEdUwPeXfjRjKQjmFAill\nLjpu8S5NmZoWoOLQUzrxp8L27NmTHlMcpuUffvhhsiT/xozSyuTJk0t92UKbN3jwYFp+ypQpPATx\nbswDOwUrg7zK27nGUJ3Q6xRLeUSix23bti3birigSU8UDyl4VmH7jeuQTufpMUUq+lu+wRUKmsK+\nu7v7oEGDqNrpMR/55WI1VLWZpmVMcXExfQ5fKTUzM19OckFTU6djAx1WyWj09LXXXjPUFR3ezHS9\n8ePH0zKU9+kx/bnJAwYVh84tuPWoougpFzS1QwpJJHTuPmpd5dV5eYIu1YzN9CkuaIJyLkmcerGF\njZyaHJ000FroQ3i3WrNmjfG+tqQ9Q9CKoClZUITk0YOOaYZ3+ckyReB169ZR5dJuoKZvSCuGUz86\nhSklaDrH5G/RGTQ9pbPUdXo+/PBDekqnoqQkHjceeeQROrmjfkhNhJYv73XjnnbkyBG+zYYexZsp\nnd9xQdPq+Fu0v+kppVSTgqYOxp/yZsoP7zNmzKDH1EBNCprO6fifvPzyy/SUDk4mBU2ngUx//QA9\nprNa/if8POP8+fNlB0bpYMaPjnxskToMNxEVnyqNOgY99fDwoLdIBzzc0f6aPn06RSoLx6B57KWD\nAT2mLTGzc42h3kWRk4uGn2vzr7lWrlzZ5gbUabmgDe6uwvYb6pCOIrQl9JhHBFIAJS9LEjQfGKHj\nXJcuXXiR+eHTUNVmmpYx/KDeoUMHvivphHLq1KnUos6dO2f+S0KqKPJRbGws0195wlNnef2OykXv\n0jKHDh0ir1H98/Hrsl9LnD17tuwY9KJFi/hTfg5E54hm6rysoMs2YzN9yiBow2CIhY2cjymtX7+e\nEgMvLNmgVH+xpD1D0NeHOP7555879NBBmL9L9c6DwCtGFBQUfP3118YW5h40FjQpmL9Fn8xPl4w/\ngQ+z/v333y1atOB9hsd2/iflvW7oabRr+ZczhlLw7EDZkwuauih/nXY5PfXy8jIpaG5hgg7jhsDF\nTw7KEzQlTf4ndH5HT8PDw00KOjAwkNcnPaYK4X/CD078LWNB04Ghfv369JQShGEokA8fGVcaxTp+\nzkte4F/98+TCt6FCQfOeT/3NkPLK27mlPo2kQ4vR2S71McpHhYWFhlVwtm7dygVt0EEVtt9QhyRN\nfl5ssCd3kEHQ1Jn56/SKQdC07/h47nPPPUcHEl6f/ATCuKrLa1oGyHT8Q6i66JD2xBNP8ILQ9lCu\nNCnoCRMmUKikDGjsVnIuV7bxoCrpaeLEiYavNGqZgrbQEkFTnfOnb775Jm+HZuq8rKDLNmMzfYoL\n2vjgbWEjp75PNUDBvHv37nzv8LE14/5iSXuGoG+OQVMD4tcD8fzy448/Gk5MqMPQ2RYPCBSg6HV+\n6Q9lAX4YNBY05QLj76A///xz/vTYsWMXLlwoKSmhz6fTYXpMJ5v04Omnn6bFaOHyXjfuabQNfLSR\n9+GYmBjepamTcEFT6rFE0PzM3SBorh4uaCq4SUEbDl3mBc0HbXnbbdeuHf8TfuZeStCZmZl86HbY\nsGHGW9i4cWMyEe/eZNV///2XSkf2pKRD1UJ/SP2WnzbyajeGn5/OnTu37FUcxoIub+eWYtOmTfyk\n2Picg3Rw9AbkQS7ob775psrbb1yHXIt8qIe0yIfd+FUc/FyHfwFIkjIImmIaPSYR8KFzOms2nPgb\nqtpM0zJAK+Wjq/wpz5UEab3CMehSNG/enN4dPnw4f5qSksJP/3msoYROj0lnPW7Aj5dlvyqkNEqv\nv//++yav4jBuhybr3KSgyzZjM32KC5r6SClBm2/k9Fe0MXT2w3M3P8kuJWgL2zMEfVPQpE6+z/gh\nkXzKr+6kM0eemnknpMjAuz21Gzov44+NBX348GHDF02PPfYYZRY65/L29qYHtHxubu6JEyf4uTOd\nkS1fvpwiD30IfWx5r5eKQjzWvfDCC6NGjeJnT3wsz+EETQc5PnpLJfpUD7cMH0+g022qBDoPJUkd\nPHiQtpCKTI+nTZtGXdQ4lRjDS01bMnv2bDOCLm/nloJ6LB9kIIwv8i07Bk1nsoZXKrv9Zb84ouRF\ne+HVV181FjS3GO0L2k0NGzY0CJraG/+WcufOnf379+dbyy8mMVS1maZlDE+O7du3p2hMUdqQbTt2\n7KjT6SwX9MWLF/nhhOIL6Zhvxuuvv05HF8rmtMdLfd3KR5/o9VJfFfJQT7uA+hctb0bQJuvcQkGb\n6VNc0PXq1auUoMkkDRo0oG347bff6NSBf/3IPWPY1xa2Zwj6lsvsgoOD+de7vLcsW7aM+4s6DInP\ncPUSZWTqLfRWz549+aUIfPC6lKCJkJAQMgI/INP+M1zEtnTpUmq71GforSeffNJweVl5rxsLmqzh\n7u7OxUHdgPoS7zwOJ2h+DZkx1Kx5r6CK5W6lQ5phvIL+vG3btvxLMHqXDntlrxO4dOnSI488Yrgm\ntzxBm9m5peDfOj711FPlXbhdVtCV3X7jOqTN6Nu3L/VbagM99RgEffLkSf5dHB31+Vm5YQzaw8OD\nf2ybNm26dOliOJE3bjblNS1jqP3zM0K+hXRm8+eff1LcNmRhCwVNUPP45JNP+IgKrZQOvfxaRj7u\nR9tZanneeEp9VUh1zhsJdR+qBDOCLq/OLRR0eX2qaoKmx7///ju/NoNCAP9Gmo89Gu9rS9qzSwva\nQhITE40vf6GGQmdq1Fv4U177dNw28wk5esq+np2dTZGhbM8v73Vj+HmrmR/jOjoUQ0z+qJdep3NS\nfo25Sahr8W/Pq7Bzy0K65IOtdtt+gk6NTTYYahLlfdFPh1hLLqS1pGnRSR4FcEPToo0xv7VmKC4u\npo8y+YWk5VDSL+/waWGdW4h1+xRVsiUbY0l7gKArATU4/nXit99+S+fFlL8oJth0dB/UCD4+Pp9/\n/jn/3TMZHxUCHBoXms2Ozo5feOEFOiuhs6E333zzjz/+wO53PujoS3amE9VSk0gAAEE7APacGQPU\nCOqc2hAACBoAACBoAAAAEDQAAEDQAAAAIGgAAAAQNAAAQNAAAAAgaAAAgKABAABA0AAAACBoAACA\noAEAAEDQAAAAQQMAAICgAQAAggYAAABBAwAAgKABAACCBgAAAEEztn379s0AAKAODh8+DEFf548/\n/vjggw9+AwDYne++29u8+aWlS1ETt9CkSZPg4GAIWmHLli3Lli3DiQ8AdiYri3XowKKjUROl+f77\n78+fPw9BQ9AA1BgTJrC1a1ENEDQEDYDK+Ocf1q8fE0XUBAQNQQOgJrKzWceOLCoKNQFBQ9AAqIyJ\nE9nq1agGCBqCBkBlHD7MPDyYIKAmIGgIGgA1kZPDOndmERGoCQgaggZAZUyezFauRDVA0BA0sBKR\nkZFnzpxJSUlBVVSTo0eZuzsrKUFNQNAQNLAGc+fO7dKly48//vjoo4/u3LkTFVJlcnOVwY3wcNQE\nBA1BA2uwffv2nj17ivqLddPT09966y3b/QDX6fHyYr/9hmqAoCFoYCU8PT0vXbpED9KScgRZmbZl\n/vz5qJYqcPw4c3PD4AYEDUED6zFixIjAwEBF0OejAvPYht37li5dimqpLHl5yuDG5cuoCQgaggbW\nzH3H27Vrl5mZKZ0IPhkU8sNvm4JS8ylKg0oxbRrDcQ2ChqCB9dm1a9fzzz8funxjt27dQkIvxmjZ\nuTyWhVN1izlxgvXuzXQ61AQEDUED20AJ+uYJu8DO57MrhUyQUDEVkJ/PunRh+mF8AEFD0MD2glae\nyiwWUdoCpk9nixejGiBoCBrYUdDX46HAgvJZZCErQZQ2xenTrFcvDG5A0BA0qAlB8ygdp4/SmYjS\nt6LRsG7dWGgoagKChqBBDQnaOEpHFCBK32TWLObjg2qAoCFoUNOCvh6li9jZPJaBM3rGzpxhX3/N\niotRExA0BA1UIOjr5/UCC3b5KF1QoAxuXLiAVgNBQ9BATYLmUTretaP0nDls4UI0GQgaggbqE/T1\nKC0qUTq8gOlcLEoHBLAePVhREZoMBA1BA7UKmpBldk0fpdNdJkoXFrJefRnm+4OgS5Oeni5U9R5n\nEDSwhaA5BSK7kM/CXCNKz53L1p9iqTe+Gzxx4oSfn18oLrVzQUG7ubmF66f+joiIaNeuXZ06derV\nqzd06NCSyk9oCEED2wnaOEqnOXWUDgxkX33FcgpYqIZd0rBh34/28PBYsGBBo0aN5s2bhybkWoJu\n2rQpnwfS3d2dCpOVlRUZGdm+fXsvLy8IGqhK0K4QpbVaxc76HqkckDYeODp71+EUfZQuLi5u3br1\nkSNH0IpcUdANGjTIz8/nLwYHB7ds2RKCBioUNDdXgj5Kp1YUpbVa7cqVK+fOnbtjxw6HqB+KyN7e\nN58OGDAgJDySonRmbBo93b1798yZM9GKXEvQPj4+ycnJnTp1io2N5S/6+vpSwSBoYJku5T///HPj\nxo1nzpyxj6A5hSIL0bDLBay4nCgtiiKdC06dOpU27+OPP/76669VXpNkEorPhYU3XxkxYgRFZjog\nac4os/RTV12wYAGanAsJ2tvbm86b6tevX7t2bd6Cp0yZ8uijj4aEhEDQwBL69+8/dOjQFStWPPPM\nMxWOjFlR0DxKJ5YfpWfPnr1o0SLD04EDB6o5R2u1ynV1587d8mJoaGiLFi3Cw8Op3ig+16pVKzc3\nF03OhQRtIDExMSwsjB74+/trNBrzC69fv75LGV555ZUOHTqgTbgUv/7666RJk/jjkpKSL7744tCh\nQ3YTtHGUvqQpHaU9PT0jIyPpQWxUak4Joxw9d+5c1dYkJeNffjHxenBwMMX/oMXrBg0aFB8fjybn\nooKu5NFem12GwYMHd+/eHW3CpfDw8IiIiKAHCbHpJEpKeXPmzLGzoK9H6WIlSqcYTVvx448/0lkd\nPSjwvxSYx7w37ly1Zq06qzE4WPlVt/Hghn3qDYJ2TkGb5LvvvlP/MB+wLsOGDeOROedsOPnx1807\nlyxZYn9BX88N4vVL04qk6yeFzZo127FjR8HB0zNmzfYY73U2V8pW33SmRUXKjEgBAaym6g2CVq+g\nvb29Z5li165dEDSokLCwsI8++ujs2bMlxwLXbdw0wMv7fI5AUbqmRENROskoSufn51OODl+9dfbs\n2QUFBbkCO5/HogqZqu5Xu2iRMqeoeSBoFxU0JaBatWq5u7t/eyurVq2CoIElXLlypX379heW+o4a\nNSo9PT1Vp/gxoUhxZU2JhkfpizeitPFKRZnx+9WqJEqHhCiDGxV96QNBu/AQB+l48ODBGOIA1cHY\nIDpJ+SHJhXzlRyU1JRpDlE4uNrHSm/errdEoXVzMevZk/v6Vq17gWoLWaDS9evWq8MoNCBpUyiBp\n+ih97dYobWfRUJSmHB16JV1b5lBhiNI1eL/aX39lM2ZUsXqBqwjaWkDQEHQpykbpGhFNUnAMHSoo\nUJcddcnT32TrSk3crzY0lHXtyizMRRA0BA1BAysLmpNuFKVrRDS00iJJubojVMPKRmlJZrF2v1+t\nTqfcqPvUKStUL4CgIWhQLYNQlA4vUKbez9f/ZLmmti1FPyptMkrz+9VG2itKL1nCpk+3WvUCCBqC\nBtU1SIaOBcTkxhcpobWmts18lI6zS5S+fJl17sxuTE0GQUPQEDRQgaCJ4lOhEfoorRFqctt4lE40\nFaVtfb9anY716cNOnLBJ9QIIGoIGVTcIXyxDPyodZ68obXLbivVROkTDCk1G6SIlStvifrXLlrHK\nz7UOQUPQEDSwl6AJyqeRhcqwb75Qk9tm5mc1PEpb9361YWHK4EZeHgQNQUPQQMWC5mSWKEE1Vmvb\nKG1+2yhKXy4oN0rHW+9+tSUlzM2NHTtmw+oFEDQEDawmaPtEaUu2La38KF0gWidKL1/OpkyxbfUC\nCBqCBtYUNCfLllHawm0z8wt1w/1qqxylIyKUwY0qz7YPQUPQEDSoMUETgqT8ou98vvLrvhrZNuMo\nfa2cKF21+9UKAuvblx09aqciAAgaggZWFrRxlI7RKnNl1Iig2Y2f1ZiP0mmVidKrVrGffrJrEQAE\nDUED6wtayZuyPkrnWS1KV81u/BfqJn9Wo9xkK9/c/WqNiYxknTqxnBwIGoKGoIHjC5qTXcIC89hV\na0TpKtvN8At1jakonVD+/WqNBzc8PNjhwzVWBAgagoaggfUFzaN0VKGi6VzBHttWHhnmo7TGXJRe\nvZpNnFhjJwEAgoagga0Eba0oXX27lUisvF+oK/er5VG6uPRbUVHK4EZWFgQNQUPQwEkFzfST60fr\no3SOYMNtszBKx5m6FtBwv1pDlBZF1r8/O3jQrtULIGgIGthb0ByyMzmaTF3ZKG1Fu5n5WU2p+9Wu\nXcvGj7d39QIIGoIGNSNoHqWvavVRusT622Y5Zn6hzqP0qUTWoYN1BjcgaAgaggaOIWhO7o0obeEt\nX21htxL9z2pMRmlBYD8tYAcO1ED1AggaggY1LGjjKJ1dYr+VlsXkL9R9fdlv+60zvxIEDUFD0MDx\nBG2I0ufzlEvxzEdpm9qt1C/UY2JYx44sMcPKU5VC0BA0BA0cTNA8Ssfo71NlJkrbwW78WsDoAjZo\nENu/X3ml+vMrQdAQNAQNHFvQHEqvlGGvlBOl7WM3WvX2U2zClFterPL8ShA0BA1BAycRtHGUziqp\nGbvFxipXbqSnl369avMrQdAQNAQNnEfQhigdxKO0ZFe7SRLz9GT79pW7QKXmV4KgIWgIGjihoJn+\nPlWx+iidWWK/lW7cyEaPrmAZC+dXgqAhaAgaOK2gOfn6KB1ZqFywbOuVxscrgxtpaRYtXOH8ShA0\nBA1BAycXNI/ScfoonR54xYZrkdiQIWzPnkr8iZn5lSBoCBqCBi4haI6GonR0VkSBEqVtwZYtbNSo\nqvxh2fmVIGgIGoIGriVoQjgRHKdPrBk6K39yQgL78kuWmlrFPy81vxIEDUFD0MDlBM1XSlHaur/u\nI71++y3bvbu6n2OI0kUSBA1BQ9DAJQXN9KPS8db7dd+2bWzkSKu5nkfp5GIIGoKGoIFLCpqjEa0Q\npRP1E4qmpFhzO3mUvkhRWoSgIWgIGrikoJk1JsoYOpTt3GmTrU2+EaVlGYKGoCFo4HqC5lR5oow/\n/mDDh9twgylBX9RHaa0IQUPQEDRwSUGzKk2UkZSkDG7Q/7aGR+kkfZSGoCFoCBq4nKCNo7SFv+6j\n7EwJ2j4UScrVHaEaVuB/CW0MgoaggSsKmlk8UcbOvWzYMHtvf0oxC4jJTbp1VBpA0BA0cBVBc8xP\nlJGSwr75kQWmWHovRCuiPX2RR2mtiLYGQUPQwCUFzYwnyigTpUeOVK59Lm/WaTsUIVU/Kk2bhygN\nQUPQwBUFfT2x6qO08UQZu3crvxvkZjQ567R9ikDbQwGftq0QURqChqCBawr6epS+MVFGaqoy50ZC\ngtFn3ph12j5RulQRKN3ThiUgSkPQEDRwTUEbonSohs38TZm1rizGs07buQjXo3S+cgkKJy8vLyEh\nQavVojVC0BA0cAlBE3v2sLFeyrzPpj/8xqzTmSU1UIQ0fZS+VkTHj61t2rRxc3N7+umnAwIC0CAh\naAgaOL+g09KUn6XEx1ewGI/Stpt12kwRdBI7HJU0fMGq5Ow8ehoSEvLUU09FR0ejTULQEDRwckF/\n/z3btMmytVCUts2s0xUWgTR04mIErTo/Q3H0li1bfHx80CYhaAgaOLOg9+1T7tUtVSYUW33WaUuK\nMHLkyODgYFpj1rlIerpz584FCxagTULQEDRwWkGnpyuDG7GxlV/djVmnrRilzRdhz549b7/9dklJ\nCS2WlJRUq1atixcvok1C0BA0cFpBjxnDNmyo+kqtMuu05UWYN2/es88+G7l2e4cOHQ4ePIgGCUFD\n0MBpBf3XX8rghli9H4NUf9bpShUhMTEx78DJjIwMtEYIGoIGTitoUlzHjiwmxjprr/Ks045SvRC0\nPRAEITs7G4IGEPTYsczX15obUIVZpyFoCFpBp9ONHz++YcOGt912W61aterWrfvqq6+uWbMGggau\nKegDB9iAAdUd3DAfpYslp6peCNqGeHp6dujQISQkRKPRyLKcm5t7+vTp999/f+nSpRA0cDVBZ2Up\nV25cvWqrLbFw1mkIGoK+TuPGjVPK3JrY39+/TZs2EDRwNUGPH8/WrrX59pifdRqChqBv8sUXX2wq\n80upiRMn9u7dG4IGLiXogwfZLxtZZrE9NsnMrNMQNAR9k6CgoJdffvm9994jt3p6evbo0eONN95o\n0qRJXFwcBA1cR9BZWaxTJ3b5KgvMY1GFdrpnStlZpyFoCLo0RUVFR44cWbNmzcyZM1euXHns2DG5\nollmN2zY0K0Mzz//fPPmzdEmbEFgYGCvXr0+//xzOoJW82IbCNokEycy/tW4KLOrWkXT2XaZ6Nl4\n1mkIGoK2DoWFhZll+Oabb7p37442YXXi4+M//PDDc+fOabVaPz+///3vf7m5ueXtl6VLl86aNWv7\n9u0QtOUrPXyYeXgwQbj5Sq7Azufp75lirygdqo/SRRIEDUHbBgxx2IghQ4bQaY3iX60yYLlw4UKT\n19gIgtC2bdsZM2bs37+fhG7nfeG4gs7JUQY3IiNLv05ROsa+UTpJH6WTiyFoCPoG3t7es0yxa9cu\nCFoluLm5XblyhR7EXrwWXch27No9Z86csotNmzZt2bJlhqcDBgyowk50QUH/9BNbtarcd/MoStvx\n9oMUpS9qlH9FIgQNQTM2bNiwWrVqubu7f3srq8y0WQjavixevNjT01PJyCeCI/OFn3x3HToTWHax\ngQMHRkVFKR6/kkqhb8+ePfPnz4egza/0yBHm7n7L4IaJP7fv7QdNRmkI2nWHOEjHgwcPxhCHahEE\noU2bNu3bt0/Y+mfnzp0XrVpn8kqD0aNHb9HfOK/A/xItMNt3+1rf9RC0mZXm5rLOnVlEhEUfYrfb\nD3KKbkRprQhBu7agNRpNr1696H8IWs3s378/Zcff586dY+VcaZCYmEgnQ8uXL9f8fcpr2rSBk2ac\nzZXtM37qoIKeOpUtX16Zz7HL7QeNoRAdkKcEahGCdllBWwsI2s6i4VcaGEfp/Pz8sWPHhq/e6u3t\nrdVq8+x4KYLDCfrYMebmxkoqr1p7R2lJuboj5EqGVlRp9ULQEDQw3f1MXmlgvBhfwA7jp44l6Lw8\nZXAjLKyqH6i//eA529x+0HSUDr7Ko7T5nyhA0BA0BK0iQV/Xza1XGpjwke0vRXAsQf/8M6v8bGCl\n4bcftN2dvEsVoVgfpUNvjEpD0BA0BO0Ygma3XmlgcjFbX4rgQII+cYL16cN01gi/trj9oPkipOov\n8EgsMh2lrVi9MlFmGjUIGoIGVbcbHx6NCEsuL9PZbvzUUQSdn68Mbly+bM3Pt+7tByssAkXpywXK\nJB6Fok2qV46IFCdNEZ55SfxmtJycDEFD0MBqEqRMF3PxmpkrDWx0KYKjCHr6dLZkiQ1WcSNKp+vs\ntOtTdcrqEm6N0tWpXjk9XfJZLLz3oVCrjnDHPWLbL6QdDn/zWQgaglaXoPliFSbl61nbeuOnDiHo\nU6dYr17WGdwwCb9nii2itMl6ux6l829G6SpUr6zVSpu3iu07CnfWVdTc9H1p/iI+uOEEI9oQNASt\nRkEzC640sO6lCOoXtEbDunZloaG2XVf1bz9Y2XpL00fpa/oobXn1ypIkHT4iegwU7q+nePmp58Vx\nE+XLYTW+syBoCNolBM2p8EoDa12KoH5Bz5zJfv3VTmu0yp28La83Wguti9aYf6biwXU59KI4/ifh\nyWcULz9YXxzwjXTU9AzDEDQEDWwuwQqvNLDKpQgqF7S/P+vZkxUX22+lhtsPptnrBCVdxwJicq+V\nc4GHnJQkeS8Q3nyXvCzO8BG/7Cxt+0MuKlLbzoKgIWjXEvT1pFzRlQbVvBRBzYLO+zukWzcWElID\nu6wKtx+sTr0VnQqlPUj7seDGqLSs0Ui/rxdbtxNuv1uJzM3+T1q8VDpyVrU7C4KGoF1R0MyCKw0M\n46dVuBRBzYKeOThp4cIa22tVvpN3lauXdt/ZXDnu7MWSPh5C3YcULz/3ijjlZ1k/2SFzpamXIGgI\n2mEEzanwSoOqjZ9asc8fPnx4x44dFy5csEqlBQSwHp9mmz2Vt1+UtuT2g9WsXjnwvPjdaO0Lr10e\nPyto+ZbcSdPk0/4OdDSFoCFolxY0s+BKgypcimCtbRs+fPiAAQN8fHyeeuqpuXPnVleLhaxbN3Z+\nTYQadl/Z2w+eOHFiz549oRZcWVLxPo2Lk2bOEeevVPJynfvFbl9Lu/eka4VzecppkyRD0BA0BO0g\ngrYwKVcqSltl25YvX/7DDz/wx8XFxW3atDl69Gh1auyXX9j8+eoSjeH2g8O//8HDw2PhwoUNGzac\nN29e1epNzs2VVq4Wmv9PuK2O8q/PQHoq5+QYFiiRlEt0gvOVy3U4mZmZBf/4WzK9MAQNQYMaEzSz\n4EoDyy9FsMq29evXLzw8nB4kxmUUSWz3btM397KQc+dYjx5Mq1WdaKhWNx44OnvXYR6l6VD06aef\n8ttOWlhvckmJ5LeXYjKFZSUyv/y6NGM2hejy/jyzRLngPU7LfNdv+Oyzz8JWb33yySePHz8OQUPQ\nQL2Cvj4OUNGVBpZcimCVbRsxYgT3VPbZCDoq/LbNz8fHp8qDG199xXjfVKFoBg4cGBIeSVE6MzaN\nnu7Zs2fmzJmW1Jt82l/8doRQ73HFy481Ekd+LwdaJCCK0gcvx45a4puaV0CfFhkZ+eyzz0aWvVcu\nBA1BA1UJmllwpUGFC1hl20JDQz/55JPLly+LJ4K37d7rOdMnKFtnyUT1ZfH2ZoZhAxWKZuTIkUeO\nHKFa1eh/WkLHoYVmLzSRDp4Wp/wsPPeK4uW6D4m9+kp/HZDFylUN2er0xQiK0rkBymnKjh07Fi1a\nBEFD0EDtgjZOymauNDCzgLW2LSQkpEWLFkGL1w0ZMiQxMTFZ/61ahRPVlyIwUInPWq16RcOLGRER\nQdvm5+dXq1at3NxcE8fFzExpyTLhg0/EaYuE2+8W//e59Pt6uap3p+O2oiidk5oDQUPQwMEEzUxd\naWDhAtbdNuPFiiyYqN4Y8jLZOTBQ7aIJCgr6+OOP6VDk6ekZHx9/SyUXFUnbd4gdugh33atE5jfe\nkbwXyklJ1Vzj/v37mzRpUlhYSI/j4uLoqBBh9o65EDQEDdQl6OuOu3GlQZFk6QK2EzQnxeIoPX++\ncvGGo4jGeNuUOfKPHRcHDhYerK94+clnxDHj5NCLVlzdsmXLnnnmGTo76dChg/mvJSFoCBqoVNA8\nKSdVFKWNF7C1oC2M0sHByoXP+ozoSIKWL4eJ4ycJTz2vePn+eqLHQOnQYVmyyfz/aWlpFJxNjqhA\n0BA0cAxBVzZKF56+aJ8iXL/nk6koXVSkXFcXEOAwopFTU6XN+4S3/6t4+c66YrsO0uatsmHs3EHa\nGwQNQYOaEbRxUk42G6UDYnKTi+1UBH771LL3fFq4kM2e7QCikQsLpQ2bxM++VG5cMvA74b0PpUWL\n5bQ0B21vEDQEDWpM0IakfFGj/CtveKHQ/5L5BaxehFL3fLpwQRncKChQr2iUOfIP/iO69RPue0SJ\nzI1fFCdOliMiHb29QdAQNKhhQXPMXPTGV1rhVXHWLYLh9qk5WkZN78wZlYpGDr4gjv5RePxpxcsP\nPyZ6fiv/e0Ku1JWDEDQEDUFD0BVSJCkxuew3dYaVmv8qzxZFSNOxTQHMzA/xako0ckKCNHuu8FpT\nxcu17xM7dZN27JLtedcACBqChqBdStCcshe9lVppeVfF2aIIoaGsV19m5tcbdhaNnJcnrVkn9vS4\nPkf+Ry2kZcvlrCxnbW8QNAQN1CXoskm57EpNRmmrF0GnU27Uffp0zVeILAjSvr/EHr2Fex5Ubis1\nb4X08wz56lWnb28QNAQNVCdo46ScWMzEclaacutVcdYqQlZW1owZM8aMGfP11yd+/lmq2QqRA86K\nw0cJ9Z9U8vKjT4jDvpPPBDiB3SBoCBo4tqCZ4aK3KxmFotkF9FfFWaUIWq32448/XrZs2ZZtl557\n6dz7738qCIL9K0SOiaGMLLz4quLlux8Qv+ol7f1TvrElEDQEDUFD0GrpV8lBV5WkXFTu9Rv8qrhr\nIXGWXLxgvgg//fSTr6+vTsfc+7MDMWy6z7L169fbrULk7GzptxXCRy34HPliyzbS6rVyXp7z2Q2C\nhqCBkwiaVmq46M1clI5INbOAhUX45ptvIiMjly5lc6dlaQS2Nyxh7aHTJZJtK0TW6aSdu8XO3YVO\nvZTI/Opb0uy58rVrTmw3CBqCBs4jaOOknFBOlKbF0swuYEkRvLy8Jk3a1LkzyzkQQk+7dOu2/fgZ\n+swMnfUrRJnA6MRJ8ZuhwsOPKV5+/Glx3Q45+II69wIEDUFD0BB0BSvVScqdDC/kK3c1NLmYmQUs\nKUJaWvadd27y8FiTve/YyJEjW7ZsSRbViMqd90zerbxqFSJHRIqTpgjPvKR4+d6HxT4e0t8HZVF0\nnTuuQtAQNHBCQV/XKB90vjUpGy9mcgFLirB8OZs0Sfjll1+ifXeuXr1avHFjEUlW7mBNn5muq3qF\nyOnpks9i4b0PFS/fcY/Y9gtp/UbZ6CfkEDQEDUEDhxe0yaRcajFaINxUlDZThPBw1rkz45NlmlyM\n34zcOEpbUiGyVitt3iq27yjcWVdRc9P3pfmL5JQUx9oLEDQEDUFD0JVbabpRUja5GF+Awq8kV/Bp\nJSWsb19mmGi+vMVoRdeMbkZuZtuUCYwOHxH7DVJmYf7BS3jqeXHcRPlymIPuBQgagoagIehKr5Qn\n5eB8lncmzPwCGtHcp61cySZPtnSlPEpThC86HWpCzRcviT+OFxo+q+TlB+uL/T2lvccsmcAIgoag\nIWjgVILmZFBSjskxTsomFtBHacHUp0VEKIMbOTmVWCm/GXlATG7ajVFpOSlJ8l4gvPmu4uW77hW/\n7Cxt3S4XFTG73BQGgoagIWgIOli121Z8KjSCJ+VyfvpXIjFaICg6q9QCgsA8PNjhw1VZqebM5ZDs\nkksBlwq7fH19AqP/fiwtXipnZNiueiFoCBqChqAdTNB8MUrK5/JYXPlROv18VKkFVq1ikyZVeqXK\nVXH7D4grNpfc90h8V/eAHYeT12ySo6LsUL0QNAQNQUPQDilonpQjC1lQPssXTC9mvAAZtWNHlp1d\niZXK54PE70YLjzVS8nK3vuKQ4fKp04Wi8kPHSxrlN40QNAQNQQMI2tximSVKlI7Vlo7ShsVogbO5\nbMQ49s8/Fq1Ujo+XZs4RXnlD8XKd+8WuPaRdfrLu5nXRsqxMrWfybuUQNAQNQQMI+hZMRmnjxdat\nZzMWVvBpcm6utHK10Px/fAIj4f9aSStWycbfJ96KybuVQ9AQNAQNIGgTZN0apQ2LRUezDh1YeXcj\nkf4Nkvz2it2+prCsROaXX5emz5Lj4izZEsPdyg1RGoKGoCFoAEGbRpDYlUJ2Pp/lCdcXE0U2YAD7\n+29TevU/I347Qpy9VPHyY43Ekd/LgVXph8ZRGoKGoCFoAEGbg0fpq5cSRJn9/jsbN+5WL0dFiVN+\nFp5voni57kPikO+kvw7IZmfrtzxKJwXHQNAQNAQNIGhzCDK7cjnpeDIbO5/F669UljMzpSXLhA8+\nUbx8+93ip59J63xlM7eJrVqUvpJ+sZybkUPQEDQEDSDo65QcCx7jxfYd1J0LS4xetKr4vkcUNb/x\njvTLPDkpyXZFSC7nZuQQNAQNQQMImo85yL+PDhv9+t/Cg/WL7n0kctKcwH8v5IRF2acIJm9GDkFD\n0BA0cHVBy2Hh4oSfop5o/uWLUcn3vii6D5AOHZYlKbuEBeaxq1omynYqQkr5URqChqAhaOBCgpZT\nU6UFPsLb/xVq1dHdce/ARgf2jjkua7XGywgyiy5UNJ0r2KkI5UVpCBqChqCB8wtaLiyUNm4WP/tS\nuOMeZYj53Q+kRYs3LM394Ydy/zzHllHaZBF4lE40itIQNAQNQQOnFbQyR/7Bf8S+/QX+1V/jF8WJ\nk+WISHo3Pl75WUpamrlPEG9E6RzBTkUo1kdpw83IIWgIGoIGTiho+UKIuG6H8PjTipcf+o/o+a38\n7wnDHPmSxAYPZnv3WrQ6sjM5mkxtxShtvgj8buWJRUyEoCHomhJ0enq6UNXL/iFoCNq0lxMSpDne\nwmtNFS/3HCh26ibt2CUXl56saPNm6k6VWCOp+apW0XR2iZ2ql6L05QJ2ISqjUFTpXoCgnU3Qbm5u\n4eHhTLldRUS7du3q1KlTr169oUOHlpRUutVD0BD0LV7Oz5fWrBNbtb0+R/6HzaVly+VyptW4dk0Z\n3EhNrfR6c/VROqpQ+RbRPtWbEhRNUTqhiJm/PRYEDUFbgaZNmwYGBtIDd3d3KkxWVlZkZGT79u29\nvLzM/NWmTZt6lOHFF19s0aIFNOrigpYFQdr3l9ijt3DPg4qXX2gieU2Xr141N/QhsyFDmJ9fFVfN\no/S5akdpy6u37N3KIWgI2raCbtCgQX5+Pn8xODi4ZcuWZv6qoKAgrQyenp7du3eHRl1W0HLAWXH4\nKKH+k4qXH31CHPadfCbAkr/dupVOv6q7AXkCO5+nTLdU5Shd2epNM7pbOQQNQdtK0D4+PsnJyZ06\ndYqNjeUv+vr6fl+p4UAMcbiwoOWYGGnaTHHeCsXLdz8gftVL2vunbPEQWUKCMriRkmKFLaEoHaMf\nlc4qsVP1monSEDQEbQW8vb1bt25dv3792rVrc71OmTLl0UcfDQkJgaAhaHNezs6WflshfNySz5Ev\n9vWUVq+V8/Iq9yEyGzqU7dplzQ1TonS+PkpLdqredFNRGoKGoK1JYmJiWFgYPfD399dUaZ4wCNoV\nBC3rdNLO3WLn7kLt+5TI/Opb0qxf5GvXqvZp27ezESNsUFGyMvf/uUpG6epUL0XpcH63cpHFxMSM\nGDHi0opNY8eOLS5zsQoEDUHXDBC0EwtaJk6cFL8ZKjz8mOLlx58Wvx8jB1VrRYmJyuBGcrKtqitf\nUO6kFVmo3FXLPtWboWOnMnStew/YvdtPe+jM6tWr27Rpo7HqtKgQNAQNQUPQRmqOvCJOmiI885Li\n5XsfFvt4SH8flEWx+p88bBjbscPGNSazOH2UziyxU/WOnTBxm38IReniU8og4axZs5YuXQpBQ9AQ\nNARtzX4lp6dLPouF9z8Se3kKd9wjtv1CWr9RLiiwVllIzSRo+8CjdERBBVHaKtU7cODA8PBwitKp\nkcq81fv27Zs9ezYEDUFD0BC0FfqVrNVKW7aJ7TsKd9ZVInPT96XN+2SrXGNhRHKyMriRmGjHqqMo\nXaREaVKnTat3/vz5EyZMMDx96623du/eDUG7iqAXLFgwderUmJgYFdYFBO24glYmMDp8ROw3SHjg\nUcXLjZ4Tx02UL122US8dOVL5etD+aATle7zyorRVqreoqOiNN95wc3Pbvn173759Bw0a5PTtDYK+\nyZUrV0aPHv3EE080b9587dq1qvr+AYJ2REHLFy+JP44XGj6rePnB+mJ/T+noMdnoejGr99Jdu5RL\n62S5hupQZvFFylVxZaO0taqXDnabNm1asmTJX3/95QrtDYIujSiKBw4c6NevX8OGDd3d3f39/SFo\nCLpSi8nJyZL3AuHNdxUv33Wv+GVnaet2uajI1r00JUUZ3EhIqOGa1IhKlA4vUC6Ps8MJCgTtWoJO\nT0/38fH56KOPGjduTFvw3//+t0ePHhA0OkyFi8kajeS7QWzdTrk39s8Lhf9+LC1eKmdk2K2XjhrF\ntmxRRWVShL+mj9LpOggagraSoP/999927drdf//9PXv2PHjwIJ1M8VOqBg0aGH6oDUFD0KVlJIrS\n/gNib3eh7kNKZH72ZXGyl3TwtJ17qZ+fMilSTQ1umKRAVH6ozaM0BA1BV1fQq1atWrZsWW5ubqnX\n9+/fX4UJQiFopxe0fD5I/G608FgjxcuPNBAHD5NPna6RwJiaqgxuxMerrlYNUTo1KBqChqCtMMSh\nTiBo9Qhajo+XZs4RXnlD8XKd+8WuPaRdfrJOZ90Bk0p2FWVKftVCUTo4KjPs1lFpCBqChqCB1TqM\nnJsrrVoj/rqOT2Ak/F8racUqOSfHugMmVWDfPvbNN0ySVF294ongBH2UTtNB0BA0BA2s1GHkkhLJ\nb6/Yvadw9wNKZP7KXZo+S46Ls/qASdVIT1cGNyraHLVUb6Go3B/2coFygysIGoKGoEHVO4zsf0b8\ndoRQ73HFy/9pKI4YJZ8LtPqASTU3e9ZytuWAI1WvLDMepVN1EDQEDUGDyhokOlqaOk14voni5boP\niT3dpD/3y5W82699BP3nn2yKDzuaxUI1FQ/vqspHPEpf0pSO0hA0BA1BA1Md5ug5acky4YNPFC/f\nfrf46WfSOl/5xo3KbNT9qtNLMzKUwY2YGJZUxI5ns39zKhjeVd0JiswSi5UonXJjVucTJ06k7T4U\nGhoKQUPQEDTQa6K4WNq+Q+zYVZzho6j5jXekX+bJ1Z5qyA6C/vFH5ut7/XGR/sd7FKVDNOUO76rT\nR1pRif8UpcdOmuzu7h6/eW+jRo3mz58PQUPQELQLe5k4/q84cLDwYH3Fy08+I06cKoeE2rn7VbmX\n7t/PBg5kpSaOTr4RpVPUd6cRc7/AlNmG/UfWnLpIm02LFRUVtWrV6vjx4xA0BA1Bu56aw8LFCT8J\nT7+gePn+eqL7AOmfQ7K1L1KzqaAzM1nHjuzqVRNvFUn6KJ3NQvLVFaUrnOg5JDySonRmYCRTfhXp\nN3PmTAgagoagXcbLqanSAh/hnWaKl++sK37+pbRpi1xYWLPdr2q9dNw4tm6duQWSi/VROpulFDmG\nj0aMGHH06FGK0gUaJfwv0gNBQ9AQtLN7ubBQ2rhZ/OxL4Y57FDW/+4G08Fc5LU0l3a8KvfTgQTZg\nAKvwrlgUny/oo/QFdURp8yW9cOFC8+bNw8PDeXyuVatW2RkXIGgIGoJ2Fi9LkvTPIbFvf+G+R8Qx\nPwuNXxQnTpbDI9TW/SrbS7OylMGNqChLl08pVnI0penkmo7SFZY0KCjo448/btWq1aBBg+JVOKsI\nBA1BQ9BWUPOFEPGHscITjZW8/NB/xEFDpH23zJHv0IKeMIGtXVu5LaH4HKKP0sF5ygg1fARBQ9AQ\ntN29nJAgzfEWXmuqeLn2fWKnbtIfO+Xi4prqCbYQ9KFDzMODVe2W34YonVRDURqChqAhaNfzcn6+\ntPZ3sVVb4fa7FTV/2FxatlzOyqrxnmB1QWdns06d2JUrVd8kQ5QOylOum4aPIGgIGoK2jZcFQfpz\nvzhwiHDPg4qXX2gieU2XTV535iyCnjSJrVplhQ1L0ykXSvMobc+BHwgagoagXUDNZ8+JI0YJ9Z8k\nL4uzl4rDvpPPBKiwJ1hX0IcPK4MblZwOpFx0EruouR6ltaK6KgSChqAhaAf0ckyMNG2m8NJrSl6+\n+wHxq17Snn3Sv0Gq7QlWFHRurjK4EWHtK1DSdddHpRPtEqUhaAgagnY6L2dnS8tXCh+35HPkiy1a\nS6vXynl5NhpGUKegJ09mK1bYZCMNUfq87aM0BA1BQ9DO4mWdTtq5W+zylVD7PiUyv/qWNOsX+do1\nmw4jqFPQx46xvn2ZTe+ISVH6RI7NozQEDUFD0I6v5pOnxMHDhHbdFC83eEr8fowcFGyHlKpOQefm\nss6dmf7ndbaFovQlG0dpCBqChqAd1suRV8SfpgrPvKR4+d6HxRWbpb8PyhVd8ev0gvbyYsuW2W+b\nM/RR+lg2S9BaP0pD0BA0BO1oXk5Pl35dIrz/keLlO+4R27SX1m+UCwqsa14HFfS//zI3N6az7xz8\nJRK7rI/SgXnKTU/gIwgagnYJQWu12pUrV86dO9fPz0/WaqUt28QvOgl31lXU3PR9ad5COSXFRuZ1\nREHn5SmDG5cv18zOyiy5HqXjrRelIWgIGoJWKYIgtG/f3svL6+T0mTvrPZZ/l/6rv0bPieMmypcu\n29q8jijoadPY0qU1ucuUKF2gj9K5rEC0X4VA0BA0BG1vSM1LlixRfvs3+AfhgUePPP/yqZmzzExg\n5OKCPnmS9e5t78GN8qL0SX2Ujqt2lIagIWgIWqX069cvJiZGLi6ODYjIKSj28/Mzf9M5Vxa0RsO6\ndGEXL6pl31GUDtNH6bPVi9IQNAQNQat343ft2kUPCkXl26fZv29b57segja52IwZ7NdfVbcHs6od\npSFoCBqCVimxsbHNmjXbs2dPdHT09FmzBk2ZfS5XyhEg6NKLnT7NevZUxeBGWQSZhd+I0hrRVhUC\nQUPQEHQNkJOTQ0Xo37//nDlzCgsLyc4UpaMLmShD0DcHN7p2ZaGhqt6P2TeidKyWSbL1KwSChqAh\naFVAar6qVTSdUwJBK8yaxRziRqkUpSP0UTqAorRg5QqBoCFoCFpF5JqK0i4o6DNnGO1e/R1gHAOK\n0qdyFE3HWBalIWgIGoJ27CidXeKigi4oYN26sQsXHGzHUZSOtDhKQ9AQNATt2FH6fB6LKlS6vasJ\n+pdf2IIFjrrjckrY6VxlVDqm0FyUhqAhaAja4aN0jD5KZ56LdB1Bnz3LevRgWq1j7zgepc/ksnwB\ngoagIWjnJY+idHT2FX2UdnpBa/650L07Cwpyhh2XIyhRmjR91VSUhqAhaAjaSSg5eYGi9Lk85fcR\nzi3oOd8mzpvnVOdAUYXXo3SeAEFD0BC087Zd6uFB+UyJ0pJzCjowkHVvme3Qgxsmyb0RpfnFOTqd\nbsOGDde27Pvrr7/gZQgagnaetktnyrHlR2mHFjR5+auv2NlVEU65Bw1R2j9HHjD8u3HjxqXuPPjO\nO+8MHDgQaoagIWinarv5+igdWahM3OM0gp43j82d6+Qn/nQOtCUkdvuVDJK1qC9pnz59/Pz8YGcI\nGoJ2qrZLUTpOH6UzS5xB0EFBrHt3Vljo/COznt8M9o9LO5bNoi8m0NN9+/bNmTMHdoagIWgnbLs8\nSkcUXI/SDiporVa5ru7sWSfppeYZPXr0zp07KUqfy5OvFLKxEyYus+dtFiFoCBqCtmfbVaJ0ETub\np9zh1EEFvXAhM4RIpxd0XFxcs2bN9u/fH3ft2up9h6Zu3pdaWAI7Q9AQtDO3XY3AgvNZWHiKTnIw\nQQcHK7/qLihwFUETmZmZgwcPdnNz8/LySsrJN/l1AgQNQUPQTtV2lSgdGs+jtKMIurhYGdwICHCq\nXlrpmjH1dQIEDUFD0E7Vdv/999+03YcuRERRlA4vYGaitHoE7eOjzCnqZL20apT6OgGChqAhaOdp\nu8OHD+/fv3/85r0NGjRYsnTpNf2odLpO1YIODWXDJymz8kPQN6N0kRKlM3QuWXwIGoJ2yra7ZMmS\n0aNH88UKCgpatWp16tSpApFdyFdub1o2SqtB0MXFrFcv9k8ku6RhxRIEfRP+dYILRmkIWkWCFgQh\nOzsbgrZK23Vzc4uMVKa7S7wQS7Lbvn373Llz6aksMx6l03SqE/Svvyp3g6UtTCpWtjClmO3evbt1\n69Yhv2147bXXIiIimAtDUTr+xpU5EDQEbT9B63S68ePHN2zY8LbbbqtVq1bdunVfffXVNWvWQNDV\nabtDhgw5efIkPcjOL6ZevWzzTsrUhncNUdoQVGtc0Bcvsi5dbg5uaEV2JDZ9+oZdCanptFhISEjL\nli0TExNdPUqLrMKvEyBoCNqaeHp6dujQgXqgRqORZTk3N/f06dPvv//+0qVLzfzV5s2be5bhxRdf\npG4MQRPUJlq0aHHlyhWq0u179g35ZWlwtq7IqFdTUE3QJ7JUXc0LWqdjvXsz/QHlJsOHDz8RdpW2\nsMD/Ej1dt26d+SbhIhjOgdJ1rt7IIWh7CLpx48YpKSmlXvT392/Tpo25KKHRpJRh0KBB3bt3h6A5\nZ8+e/eCDD1q1akWHwITERMO4gTGFIgvRsMsFTHv6Yg0WgcQ7bVrpd2lvXrp0iaJ0VmIWPd2xY8cC\nx72lirUx83UCBA1BW5Mvvvhi06ZNpV6cOHFib8pUGOKwatsl2YVqlK/gSkXpxCIWEJObqquZIly+\nzDp3Zvn5pd+lVvHpp59KkrKtmZmZtWrVsl1Dd+gonaZDI4egbUZQUNDLL7/83nvvkVsp6/Xo0eON\nN95o0qRJXFwcBG31tmv4Ci751ihd4H8pRFP68gk7FEGnY25u7N9/TS8wduzY119/fcKECZ06dfrz\nzz8hZVeL0hB0zQuaKCoqOnLkyJo1a2bOnLly5cpjx47JslyFz4GgLY/SFzXKP3pg+DQlSpsaBrFp\nEX77jXl5mVsmPDz85MmTSUlJcLGZg26Ck0ZpCFoVgrYWEHSlSNbrmAK1bHSPcJPDIDbi0vqwzp1Z\nbi4cawUMXycUS2jkEDQE7RRtlyzMo3Sh/koJ88Mg1qWkhPVpm3XsGNRq/SidqkMjh6AhaGdpuyTi\ngJhcHqXLDoMUiTbZ/hUr2E8eKbCqjaK0nb9OgKAhaAjahm1Xe/oidelQo1Fpm0bpiAjWqRPL3h8C\nn9ooStv56wQIGoKGoG3Ydvmnpeh7deKtUbqozDeK1UQQmLs7O3IEk2zYFnt+nQBBQ9AQtM0FTdB5\nMXXpkDI6VoZBbnyjWE1WrWI//eQk3U/9UTrJkaM0BA1BQ9AmPi1Vp4/SRbdGab27Q6sXpa9c0Q9u\nZEPQiNIQNAQNQVf10yhKXy5QonThrTpOqUaUFkXm4cEOHXKe7udwUTq52HUbOQQNQTte292zZ0/r\n1q0vLFv/5ptvRkVFlXqXR+mEW6N0cVWj9Jo1bOJEp+p+Dhelrft1AgQNQUPQNmy7wcHBn332WUpK\nCn0aNaNPP/207G/2SMdhBcrviQtu7dWpxSaGQcxA8u/YkWVlQdA1jPEPlCBoCBqCVm/bHTJkSGBg\nID3IORtO/69evbq8KT3T9FH6WpkobXIYxOTgxoAB7OBBZ+t+DopVvk6AoCFoCNq2bbd///7h4Yqa\nryXmUI/dtsvPzJSeuvKitKlhkFKsW8fGjXPC7ufQpKg+SkPQELRLC9rX17dt27Z8aqqIjPxJ6/1O\nhUWb/5N0M1E633SUvnpVGdzIzISgEaUhaAgagq4MI0aMaNq06aRJk7788ku//X9bcj0WRenwAuXe\nSxqx4mEQPrhx4IBzdj9nitKJ6ovSEDQE7eqCJkJDQ48fP37t2jVWmeuxMvQ6ji9S7mdqZhjE15f9\n+KPTdj+nwfADpULRORs5BA1BO0/b5T9tqHCmpJtRWjA9DHI1hnXowDIyIGjHwJKvEyBoCBqCVkXb\nTb4Rpc13Vx6l48pE6csa5rWY/fWXM3c/p4zSFl6ZA0FD0BB0DbddC2dKKpFYRAELymf5RlF6wwY2\neaaTdz9nJU0dURqChqAhaEujdIXXY2WWsHMUpbVKlI6LUwY30tMhaEelvKsqIWgIGoJWXdu18Hos\nitKRhex8Hvt2ONu3z/m7n4tE6Ws1FKUhaAgagq4EFv60YetfbPIcl+h+LhKlw2soSkPQEDQEXTkq\nnCkpPl4Z3EhLg6CdinRTV1VC0BA0BK3GtlveTEn0dMgQ5ufnKt3PBaN02R8oQdAQNASturZr8nqs\nLVvYqFEu1P1ckAw7RmkIGoKGoKsXpY2ux0pIUAY3UlJcqPu5JvyqyrI/UIKgIWgIWnVtl0fpC/ls\n2Ai2e7drdT8Xj9LnyvxACYKGoCFoNbbdHYfZpFku1/0QpSMLS/9ACYKGoCFodbXdxERlcCM5GYJ2\nRfgPlGK11o/SEDQEDUFbgWHD2M6drtj9gE2jNAQNQUPQ1eWPP9jw4S7a/YAxWdaO0hA0BA1BV4vk\nZGVwIzERggYKgsSuFLLz+SxPcJ5GDkFD0I7adkeMYNu3u273AybJLmGBeSxGy0TZGRo5BA1BO2Tb\n3blTGX2u8jQ6ELQzR2mZRemnzcoVHLuRQ9AQtEO23ZSUqg9uQNAuFaWvVjVKQ9AQNARd5dpmW7e6\n+gkssDBKB1YpSkPQEDQEXRX8/JRJkWSXH2EEFpJTpSgNQUPQEHSlSU1lvfox/U3AIWhgKaTmaH2U\nzhFcqIVA0BC0vRk1iu0MtcJdNiBoV4zSguJoMrUlURqChqAh6Mqxdy8bPJgVC1a4YR0E7bJR+qpW\nH6VLIGgIGoK2HmlpypUb8fHXn6ZX74Z1ELQrk6uP0lGFyreIEDQEDUFbgdGj2caNt7xSnbtsQNCI\n0jxKZ5dA0BA0BF099u1jnp5Mkky8VbUb1kHQgEfp8+VEaQgagoagLSI9XRnciI0td4GbUdqVvqMH\n1orSMVploqVSURqChqAhaIsYM4atX1/xYvyGdRbeZQOCBsbkCcosS1f0Ubq4uHjVqlVxm/b88ccf\nEDQEDUGbY/9+NmgQEy0bZTbcsK7CqYEhaFC6ScjKbKVnc+XPuveaPHlyht/h5s2bd+vWTZZlBy0R\nBA1B25bMTGVwIyamkn+lnxo4zuzUwBA0MInPqrXL/jkTSVFa30IGDRq0bds2CBqChqBNMG4c+/33\nqvxhhXfZgKCBSYYMGRIWHk5ROjX4Kj09cODA7NmzIWgIGoIuzd9/swEDLB3cMImZu2xA0KCcTDBu\nw4YNhqcjRoxYtmwZBA1BQ9C3ujWLdezIoqOr+znl3WUDggYmSUpKatas2bZt26KioubMmfPkk0+W\nlJQ4aFkgaAjaVowfz9autZ7u9VGa32WjuLh43bp18Zv3EvARKItGoxkzZoynp+esWbPoseMWBIKG\noG3CP/+wfv2qNbhhIkrLSpQOzJPchwyfPHly2u5/3n333YEDB8JHwFmBoCFo65OdrQxuREXZ5MMX\nrv598/krytTA+iK4ubn5+fmhJwMIGoKGoC1i4kS2erWtPrx///5RV2OiCln6eeUIsHfvXm9vb/Rk\nAEFD0BB0xRw+zDw8mCDYastHjRq1c+dOw9OhQ4euWrUKPRlA0BA0BF0BOTmsc2cWEWHDLY+NjW3W\nrNm+ffvowezZs5966inBdkcDACBoCNppBD15Mlu50uYbn5WVNWzYMA8PjxkzZjj0d/QAuJag09PT\nq5ynIOhqcvQoc3dnDnvJKQAQtLVxc3MLDw+nBxEREe3atatTp069evWGDh1ahUvTIejqkJvLOnWy\n7eAGABC0gwm6adOmgYGB9MDd3Z0KQye/kZGR7du39/LygqDtKWiq799+Q4cCAII2JegGDRrk5+fz\nF4ODg1u2bAlB203Qx4/TqQwGNwCAoMsI2sfHJzk5uVOnTrE37tjh6+tLBTPzV1u3bu1ThpdffrkK\nWoeg8/KUKzcuX0ZvAgCCvhVvb+/WrVvXr1+/du3aPP9OmTLl0UcfDQkJMeuUvMQyLFy4cM6cORB0\nZfn5Z7Z0KboSABB0+ZBhw8LC6IG/v3/VLr3asmWL405LWFOCPnGC9e7NdDp0JQAg6IoYNGgQpeOq\n/S0EXVny81mXLuzSJfQjACBoC6hTp05aWhoEbR9BT5/OFi9GJwIAgoagVSbo06dZr14Y3AAAgrYY\nT09PDHHYQdAaDevWjYWGogcBAEHbBQjacmbNYj4+6D4AQNAQtMoE7e/PevZkxcXoPgBA0BC0mgRd\nUMC6dmVmrzIHAEDQEHRNCHrOHLZwIToOABA0BK0yQQcEsB49WFEROg4AEDQErSZBFxYqV24EB6PX\nAABBQ9AqE/Qvv7D589FlAICgIWiVCTowkH31FdNq0WUAgKAhaDUJmrxMdtZPuw0AgKAhaDUJet48\n5u2NzgIABA1Bq0zQ1Dy6d8fgBgAQNAStMkGTl3v0YOfOoacAAEFD0CoT9IIFysUbAAAIGoJWl6CD\ng5ULnwsL0U0AgKAhaDUJuqhIGdwICEAfAQCChqBVJuhFi5Q5RQEAEDQErS5Bh4QogxtVuvUuAACC\nhqBtJujiYmW6Z39/9A4AIGgIWmWC/vVXNnMmugYAEDQErTJBh4Yq8/FjcAMACBqCVpegdTrlRt2n\nTqFfAABBQ9AqE/SSJWz6dHQKACBoCFplgr58mXXuzPLz0SkAgKAhaDUJWqdjffqwEyfQIwCAoCFo\nlQmayu3lhe4AAAQNQatM0HxwIy8P3QEACBqCVpOgS0qYmxs7fhx9AQAIGoJWmaCXL2dTpqAjAABB\nQ9AqE3REhDK4kZuLjgAABA1Bq0nQJSXM3Z0dPYpeAAAEDUGrTNArV7LJk9EFAICgIWiVEb4xrFMn\nlpODLgAABA1BqwZRFLdt29m9RcKCBSHoAABA0BC0WpAkqWfPnm3abBnRJfqFF16YMGEC+gAAEDQE\nrQrmzp07evSy4ZNZ5l+hFKW7dOly8OBBdAMAIGgIuubp3btvt255fwWzjMAr9HTnzp2kbHQDACBo\nCLrmadXK9+uvY+iBVifS/+PHj1+xYgW6AQAQNARdw0RHk6Dz33uv7alTp7KyshYvXvzAAw/odDp0\nAwAgaAi6JhFFNmAA+/tvFhcX17179y5dulB8zsF1dgBA0BB0jePry8aORYMHAIKGoFVGTAzr2JFl\nZKDBAwBBQ9BqQhTZoEFs/360dgAgaAhaZaxfz8aMQVMHAIKGoFVGbCzr0IGlp6OpAwBBQ9BqQpKY\npyfbtw/tHAAIGoJWGRs3stGj0cgBgKAhaJURH68MbqSloZEDAEFD0GpCktiQIWzPHrRwACBopxB0\ncHDw7bff/oiDU6dOHfr/gQc+ue++3g5dBCfYCygCilBNateunZiYCEFfp2XLlo5+yG3RogWKgCKg\nCCgCBI1GiSKgCCgCBA1Bo1GiCCgCigBBY4+iCCgCigBBQ9BolCgCioAiQNAQNIqAIqAIKAIEjUaJ\nIqAIKAIEbROSk5MdfY+iCCgCioAiOKegAQDARYCgAQAAggYAAABBAwAABA0AAACCBgAACBoAAAAE\nDQAAwFkELQiCLMsOXeNOUAQA0AvsUwRHEnRqamqjRo2io6P503PnzjUywnY3NbBdEbKzs7t37/7k\nk0++9tprgYGBDrQvpk6daqj5L7/80rG61tGjRz/88MMGDRp07dpVo9E4oh0cuv6doBfYzUUOI+gN\nGza8+uqrd911l6FSli1b1rNnz7gbiKLocEXo0aPH+PHj6VC8d+/eJ554oqioyFF2R6tWrdavX89r\nPiUlxYHskJmZ+fjjj58+fVqr1Q4ePNjT09MRBe249e8EvcCeLnIMQet0uk8//TQ8PLx+/fqGSqHe\nRfUSHx9/7do1By3C/fffn5GRwR+//fbbBw8edJRuVq9ePdry0NDQ/Px8x1LbgQMHDPNt0Y546KGH\nHFHQDlr/TtAL7OwiBxuDNq6UZs2avfTSS40bN65VqxYdgR2uCDk5ObVr1za83q5dOzoyO0QREhIS\nKD40adLkP//5z4MPPnj06FEHakIrVqzo3bs3f1xYWEiNh7qcY/UCh65/p+kF9nGRAwt6wIABhw4d\nogdnzpy5++676dTVsYoQFRX18MMPG17v3r37b7/95hBFoPjQp08fqnA6LZ00aVLbtm0dqAnNnDnT\nMKwhyzL1qKysLMfqBQ5d/07TC+zjIvUK+vHHH39Aj/Epj3GlGEMnrStXrnSsImRnZ995552G16mb\n+fn5qXZ3fPvtt7wsQ4YMMX49LS2NEpCjHB2J5cuX9+zZkz/Oz8+vU6eOQ19O4HD179C9oDxB285F\n6hX0yZMnj+uhXVi2UoqKiubNm2c4Of3kk0+2bdvmWEUg6tata5hM9pVXXgkKClLt7oiMjORloQcX\nL17cvn07fz0lJeXee+/VarWOooZ9+/Z99NFH/PGlS5eee+45h5OyQ9e/Q/cCk0WwqYsceIiDutm6\ndet4N7vnnnvy8vIcrgh9+/YdN26cJEl///33888/X1xc7BBFiI+Pf/DBB9PT0yl7Tps2zbEu86Lu\nRLvA399fEITBgwd///33Didoh65/p+kF9nGRAwv61KlTVC+vv/76Qw89tHbtWkcswtWrVykyNG7c\nmHLckSNHHGhHUGSgLafu9Oqrr4aFhTlWK9q0adP999//zDPP0Nmow12F4gT17xy9wD4ucvifemdl\nZdGx16GLQFHIETeb4pvh6iiHg+Kz8bgT6h+9QJ0uwlwcAACgUiBoAACAoAEAAEDQAAAAQQMAAICg\nAQAAggYAAABBAwAAgKABAACCBgAAAEEDAAAEDQAAAIIGAAAAQQMAAAQNAAAAggYAAAgaAAAABA0A\nABA0AAAACBoAAAAEDQAAEDQA6ufq1auvv/56dHQ0PV67dm3Xrl1lWUa1AAgaAFUwcuTIDh06pKSk\n1KtXLyAgABUCIGgA1EJ+fn7Dhg2bNm06evRo1AaAoAFQF7Nnz77tttuSkpJQFQCCBkBF5ObmUoL+\n8MMP+/Xrh9oAEDQAKmLQoEFubm6ZmZn169c/ePAgKgRA0ACogiNHjtSrVy89PZ0e+/r6Nm7cuKCg\nANUCIGgAAAAQNAAAQNAAAAAgaAAAgKABAABA0AAAACBoAACAoAEAAEDQAAAAQQMAAICgAQAAQNAA\nAABBAwAAgKABAACCBgAAAEEDAAAEDQAAAIIGAABgnv8HJQLUmJVK0sUAAAAASUVORK5CYII=\n",
       "prompt_number": 17,
       "text": [
        "<IPython.core.display.Image at 0x1092a3f90>"
       ]
      }
     ],
     "prompt_number": 17
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "X_HD = df[['educ_coll', 'average_income']].values\n",
      "X_HDn = (X_HD - X_HD.mean(axis=0)) / X_HD.std(axis=0)\n",
      "\n",
      "educ_coll_std_vec = X_HDn[:,0]\n",
      "educ_coll_std = educ_coll_std_vec.reshape(-1,1)\n",
      "\n",
      "average_income_std_vec = X_HDn[:,1]\n",
      "average_income_std = average_income_std_vec.reshape(-1,1)\n",
      "\n",
      "X = average_income_std\n",
      "y = educ_coll_std_vec\n",
      "\n",
      "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.5)\n",
      " \n",
      "clf = LinearRegression()\n",
      "clf.fit(X_train, y_train)\n",
      "print \"The intercept is: %.5f\" % clf.intercept_\n",
      "print \"The coefficent is: %.5f\" % clf.coef_\n",
      " \n",
      "predicted_train = clf.predict(X_train)\n",
      "predicted_test = clf.predict(X_test)\n",
      "\n",
      "plt.scatter(X, y, s=50, marker=\"o\", c='k', label='Data')\n",
      "plt.plot(X_train, predicted_train, c=colors[0], alpha=.8, ls=\"-\", lw=3, label='Training')\n",
      "plt.plot(X_test, predicted_test, c=colors[3], alpha=.8, ls=\"-\", lw=3, label='Test')\n",
      "plt.xlabel(\"average_income_std\")\n",
      "plt.ylabel(\"educ_coll_std\")\n",
      "plt.legend(loc='best')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "The intercept is: 0.06075\n",
        "The coefficent is: 0.80501\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 18,
       "text": [
        "<matplotlib.legend.Legend at 0x10afa0810>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAmEAAAGCCAYAAACozRT6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x1092f2d10>"
       ]
      }
     ],
     "prompt_number": 18
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#Logistic Regression"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "!head \"data/lab_scikit_learn/chall.txt\""
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "66   0\r\n",
        "70   1\r\n",
        "69   0\r\n",
        "68   0\r\n",
        "67   0\r\n",
        "72   0\r\n",
        "73   0\r\n",
        "70   0\r\n",
        "57   1\r\n",
        "63   1\r\n"
       ]
      }
     ],
     "prompt_number": 19
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "data = np.array([[float(j) for j in datae.strip().split()] for datae in open('./data/lab_scikit_learn/chall.txt')])\n",
      "data"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 20,
       "text": [
        "array([[ 66.,   0.],\n",
        "       [ 70.,   1.],\n",
        "       [ 69.,   0.],\n",
        "       [ 68.,   0.],\n",
        "       [ 67.,   0.],\n",
        "       [ 72.,   0.],\n",
        "       [ 73.,   0.],\n",
        "       [ 70.,   0.],\n",
        "       [ 57.,   1.],\n",
        "       [ 63.,   1.],\n",
        "       [ 70.,   1.],\n",
        "       [ 78.,   0.],\n",
        "       [ 67.,   0.],\n",
        "       [ 53.,   1.],\n",
        "       [ 67.,   0.],\n",
        "       [ 75.,   0.],\n",
        "       [ 70.,   0.],\n",
        "       [ 81.,   0.],\n",
        "       [ 76.,   0.],\n",
        "       [ 79.,   0.],\n",
        "       [ 75.,   1.],\n",
        "       [ 76.,   0.],\n",
        "       [ 58.,   1.]])"
       ]
      }
     ],
     "prompt_number": 20
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "temps, pfail = data[:,0], data[:,1]\n",
      "df = pd.DataFrame(dict(temps=temps, pfail=pfail))\n",
      "df.head()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th></th>\n",
        "      <th>pfail</th>\n",
        "      <th>temps</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>0</th>\n",
        "      <td> 0</td>\n",
        "      <td> 66</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>1</th>\n",
        "      <td> 1</td>\n",
        "      <td> 70</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>2</th>\n",
        "      <td> 0</td>\n",
        "      <td> 69</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>3</th>\n",
        "      <td> 0</td>\n",
        "      <td> 68</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>4</th>\n",
        "      <td> 0</td>\n",
        "      <td> 67</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 21,
       "text": [
        "   pfail  temps\n",
        "0      0     66\n",
        "1      1     70\n",
        "2      0     69\n",
        "3      0     68\n",
        "4      0     67"
       ]
      }
     ],
     "prompt_number": 21
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plt.scatter(df.temps, df.pfail, c='k', s=50, label='Data')\n",
      "plt.legend(loc='best')\n",
      "plt.xlabel('temps')\n",
      "plt.ylabel('pfail')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 22,
       "text": [
        "<matplotlib.text.Text at 0x10ac5b910>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAm4AAAGCCAYAAABZxk93AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSTC4AQAAlASDGwAAQEn8fwahY2nCdWJIAAAAAElFTkSuQmCC\n",
       "text": [
        "<matplotlib.figure.Figure at 0x107ace1d0>"
       ]
      }
     ],
     "prompt_number": 22
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      ">The default C in sklearn is 1.The meaning of C is: the larger the C, the lesser the regularization. The smaller the C the higher the regularization."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from sklearn.linear_model import LogisticRegression\n",
      "\n",
      "reg=1000 #No regularization. If we need regularization, reduce the number\n",
      "clf = LogisticRegression(C=reg)\n",
      "clf.fit(temps.reshape(-1,1), pfail)\n",
      "\n",
      "tempsnew = np.linspace(20, 90, 15)\n",
      "probs = clf.predict_proba(tempsnew.reshape(-1,1))[:, 1]\n",
      "predicts = clf.predict(tempsnew.reshape(-1,1))\n",
      "\n",
      "plt.scatter(df.temps, df.pfail, c='k', s=50, label='Actual')\n",
      "plt.scatter(tempsnew, predicts, c=colors[4], s=50, label='Predicted')\n",
      "plt.plot(tempsnew, probs, c=colors[4], ls='-', lw=5, marker='s', label='Logistic')\n",
      "plt.legend(loc='best')\n",
      "plt.xlabel('temps')\n",
      "plt.ylabel('pfail')\n",
      "\n",
      "pd.crosstab(pfail, clf.predict(temps.reshape(-1,1)), rownames=[\"Actual\"], colnames=[\"Predicted\"])"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th>Predicted</th>\n",
        "      <th>0.0</th>\n",
        "      <th>1.0</th>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>Actual</th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>0</th>\n",
        "      <td> 16</td>\n",
        "      <td> 0</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>1</th>\n",
        "      <td>  3</td>\n",
        "      <td> 4</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 23,
       "text": [
        "Predicted   0  1\n",
        "Actual          \n",
        "0          16  0\n",
        "1           3  4"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAnEAAAGCCAYAAACPY5USAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
       "text": [
        "<matplotlib.figure.Figure at 0x10796f8d0>"
       ]
      }
     ],
     "prompt_number": 23
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "reg=1\n",
      "clf = LogisticRegression(C=reg)\n",
      "clf.fit(temps.reshape(-1,1), pfail)\n",
      "\n",
      "tempsnew = np.linspace(20, 90, 15)\n",
      "probs = clf.predict_proba(tempsnew.reshape(-1,1))[:, 1]\n",
      "predicts = clf.predict(tempsnew.reshape(-1,1))\n",
      "\n",
      "plt.scatter(df.temps, df.pfail, c='k', s=50, label='Actual')\n",
      "plt.scatter(tempsnew, predicts, c=colors[4], s=50, label='Predicted')\n",
      "plt.plot(tempsnew, probs, c=colors[4], ls='-', lw=5, marker='s', label='Logistic')\n",
      "plt.legend(loc='best')\n",
      "plt.xlabel('temps')\n",
      "plt.ylabel('pfail')\n",
      "\n",
      "pd.crosstab(pfail, clf.predict(temps.reshape(-1,1)), rownames=[\"Actual\"], colnames=[\"Predicted\"])"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th>Predicted</th>\n",
        "      <th>0.0</th>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>Actual</th>\n",
        "      <th></th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>0</th>\n",
        "      <td> 16</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>1</th>\n",
        "      <td>  7</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 24,
       "text": [
        "Predicted   0\n",
        "Actual       \n",
        "0          16\n",
        "1           7"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAnEAAAGCCAYAAACPY5USAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nooGSxEVEREQ0UJK4iIiIiAb6G1ohzg9HnkmPAAAAAElFTkSuQmCC\n",
       "text": [
        "<matplotlib.figure.Figure at 0x10af2fa50>"
       ]
      }
     ],
     "prompt_number": 24
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#Logistic Regression with CV"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from sklearn.linear_model import LogisticRegression\n",
      "from sklearn.grid_search import GridSearchCV\n",
      "\n",
      "def prepare_features(train_df, X_train, y_train):\n",
      "    X_HD = train_df[[X_train, y_train]].values\n",
      "    X_std_vec = X_HD[:,0]\n",
      "    X_std = X_std_vec.reshape(-1,1)\n",
      "    y_std_vec = X_HD[:,1]\n",
      "    y_std = y_std_vec.reshape(-1,1)\n",
      "    X = X_std\n",
      "    y = y_std_vec\n",
      "    return y, X\n",
      "\n",
      "def fit_logistic(train_df, X_train, y_train, reg=0.0001, penalty=\"l2\"):\n",
      "    y, X = prepare_features(train_df, X_train, y_train)\n",
      "    clf = LogisticRegression(C=reg, penalty=penalty)\n",
      "    clf.fit(X, y)\n",
      "    return clf\n",
      "\n",
      "def cv_optimize(train_df, X_train, y_train, penalty=\"l2\", n_folds=10, num_p=100):\n",
      "    y, X = prepare_features(train_df, X_train, y_train)\n",
      "    clf = LogisticRegression(penalty=penalty)\n",
      "    parameters = {\"C\": np.logspace(-4, 3, num=num_p)}\n",
      "    gs = GridSearchCV(clf, param_grid=parameters, cv=n_folds)\n",
      "    gs.fit(X, y)\n",
      "    return gs.best_params_, gs.best_score_\n",
      "\n",
      "def cv_and_fit(train_df, X_train, y_train, penalty=\"l2\", n_folds=5):\n",
      "    y, X = prepare_features(train_df, X_train, y_train)\n",
      "    bp, bs = cv_optimize(train_df, X_train, y_train, penalty=penalty, n_folds=n_folds)\n",
      "    print \"Best Params ['C']: \", bp['C'] \n",
      "    print \"Best Score: \", bs\n",
      "    clf = fit_logistic(train_df, X_train, y_train, penalty=penalty, reg=bp['C'])\n",
      "    return y, X, clf\n",
      "\n",
      "y, X, clf = cv_and_fit(challenger_df, 'temps', 'pfail')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "ename": "NameError",
       "evalue": "name 'challenger_df' is not defined",
       "output_type": "pyerr",
       "traceback": [
        "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m\n\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
        "\u001b[0;32m<ipython-input-25-4001be82ab68>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     34\u001b[0m     \u001b[0;32mreturn\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mclf\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     35\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 36\u001b[0;31m \u001b[0my\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mclf\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcv_and_fit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mchallenger_df\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'temps'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'pfail'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
        "\u001b[0;31mNameError\u001b[0m: name 'challenger_df' is not defined"
       ]
      }
     ],
     "prompt_number": 25
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "train_probs = clf.predict_proba(X)[:,1]\n",
      "train_predicts = clf.predict(X)\n",
      " \n",
      "X_new = np.linspace(20., 90., 15)\n",
      "probs = clf.predict_proba(X_new.reshape(-1,1))[:, 1]\n",
      "predicts = clf.predict(X_new.reshape(-1,1))\n",
      " \n",
      "plt.scatter(X, y, c='k', s=50, label='Actual')\n",
      "plt.scatter(X, train_probs, c=colors[0], s=50, label='train_probs')\n",
      "plt.scatter(X, train_predicts, c=colors[0], s=50, label='train_predicts')\n",
      "plt.scatter(X_new, predicts, c=colors[2], s=50, label='Predicted')\n",
      " \n",
      "plt.plot(X_new, probs, c=colors[4], ls='-', lw=5, marker='s', label='Logistic')\n",
      "plt.legend(loc='best')\n",
      "plt.xlabel('temps')\n",
      "plt.ylabel('pfail')\n",
      " \n",
      "print \"Crosstab of Predictions:\"\n",
      "print pd.crosstab(y, clf.predict(X), rownames=[\"Actual\"], colnames=[\"Predicted\"])\n",
      "print \"Check the failures in the prediction:\"\n",
      "zip(X, y, clf.predict(X))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#Complete Version"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "data = np.array([[float(j) for j in datae.strip().split()] for datae in open('./data/lab_scikit_learn/chall.txt')])\n",
      "temps, pfail = data[:,0], data[:,1]\n",
      "\n",
      "final_df = pd.DataFrame(dict(temps=temps, pfail=pfail))\n",
      "final_df = challenger_df.sort(['temps'])"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from sklearn.linear_model import LogisticRegression\n",
      "from sklearn.grid_search import GridSearchCV\n",
      " \n",
      "def prepare_features(train_df, X_train, y_train):\n",
      "    X_HD = train_df[[X_train, y_train]].values\n",
      "    X_std_vec = X_HD[:,0]\n",
      "    X_std = X_std_vec.reshape(-1,1)\n",
      "    y_std_vec = X_HD[:,1]\n",
      "    y_std = y_std_vec.reshape(-1,1)\n",
      "    X = X_std\n",
      "    y = y_std_vec\n",
      "    return y, X\n",
      " \n",
      "def fit_logistic(train_df, X_train, y_train, reg=0.0001, penalty=\"l2\"):\n",
      "    y, X = prepare_features(train_df, X_train, y_train)\n",
      "    clf = LogisticRegression(C=reg, penalty=penalty)\n",
      "    clf.fit(X, y)\n",
      "    return clf\n",
      " \n",
      "def cv_optimize(train_df, X_train, y_train, penalty=\"l2\", n_folds=10, num_p=100):\n",
      "    y, X = prepare_features(train_df, X_train, y_train)\n",
      "    clf = LogisticRegression(penalty=penalty)\n",
      "    parameters = {\"C\": np.logspace(-4, 3, num=num_p)}\n",
      "    gs = GridSearchCV(clf, param_grid=parameters, cv=n_folds)\n",
      "    gs.fit(X, y)\n",
      "    return gs.best_params_, gs.best_score_\n",
      " \n",
      "def cv_and_fit(train_df, X_train, y_train, penalty=\"l2\", n_folds=5):\n",
      "    y, X = prepare_features(train_df, X_train, y_train)\n",
      "    bp, bs = cv_optimize(train_df, X_train, y_train, penalty=penalty, n_folds=n_folds)\n",
      "    print \"Best Params ['C']: \", bp['C'] \n",
      "    print \"Best Score: \", bs\n",
      "    clf = fit_logistic(train_df, X_train, y_train, penalty=penalty, reg=bp['C'])\n",
      "    return y, X, clf\n",
      " \n",
      "y, X, clf = cv_and_fit(final_df, 'temps', 'pfail') #Change here"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "train_probs = clf.predict_proba(X)[:,1]\n",
      "train_predicts = clf.predict(X)\n",
      " \n",
      "X_new = np.linspace(20., 90., 15)\n",
      "probs = clf.predict_proba(X_new.reshape(-1,1))[:, 1]\n",
      "predicts = clf.predict(X_new.reshape(-1,1))\n",
      " \n",
      "plt.scatter(X, y, c='k', s=50, label='Actual')\n",
      "plt.scatter(X, train_probs, c=colors[0], s=50, label='train_probs')\n",
      "plt.scatter(X, train_predicts, c=colors[0], s=50, label='train_predicts')\n",
      "plt.scatter(X_new, predicts, c=colors[2], s=50, label='Predicted')\n",
      " \n",
      "plt.plot(X_new, probs, c=colors[4], ls='-', lw=5, marker='s', label='Logistic')\n",
      "plt.legend(loc='best')\n",
      "plt.xlabel('temps')\n",
      "plt.ylabel('pfail')\n",
      " \n",
      "print \"Crosstab of Predictions:\"\n",
      "print pd.crosstab(y, clf.predict(X), rownames=[\"Actual\"], colnames=[\"Predicted\"])\n",
      "print \"Check the failures in the prediction:\"\n",
      "zip(X, y, clf.predict(X))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#Principal Component Analysis"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from PIL import Image\n",
      "STANDARD_SIZE = (322, 137)\n",
      "def img_to_matrix(filename, verbose=False):\n",
      "    img = Image.open(filename)\n",
      "    if verbose==True:\n",
      "        print \"changing size from %s to %s\" % (str(img.size), str(STANDARD_SIZE))\n",
      "    img = img.resize(STANDARD_SIZE)\n",
      "    img = list(img.getdata())\n",
      "    img = map(list, img)\n",
      "    img = np.array(img)\n",
      "    return img\n",
      "\n",
      "def flatten_image(img):\n",
      "    s = img.shape[0] * img.shape[1]\n",
      "    img_wide = img.reshape(1, s)\n",
      "    return img_wide[0]                         "
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import os\n",
      "checks_dir = 'data/lab_scikit_learn/images/images/checks/'\n",
      "dollars_dir = 'data/lab_scikit_learn/images/images/dollars/'\n",
      "def images(img_dir):\n",
      "    return [img_dir+f for f in os.listdir(img_dir)]\n",
      "checks=images(checks_dir)\n",
      "dollars=images(dollars_dir)\n",
      "images=checks+dollars\n",
      "labels = [\"check\" for i in range(len(checks))] + [\"dollar\" for i in range(len(dollars))]\n",
      "len(labels), len(images)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from IPython.display import Image as Im\n",
      "from IPython.display import display\n",
      "for i in range(3):\n",
      "    display(Im(checks[i]))\n",
      "for i in range(3):\n",
      "    display(Im(dollars[i]))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "i0=images[20]\n",
      "display(Im(i0))\n",
      "i0m=img_to_matrix(i0)\n",
      "print i0m.shape\n",
      "plt.imshow(i0m[:,1].reshape(137,322))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "data = []\n",
      "for image in images:\n",
      "    img = img_to_matrix(image)\n",
      "    img = flatten_image(img)\n",
      "    data.append(img)\n",
      "\n",
      "data = np.array(data)\n",
      "data.shape"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "y = np.where(np.array(labels)==\"check\", 1, 0)\n",
      "y.shape"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "data = pd.DataFrame(data)\n",
      "data.head()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from sklearn.decomposition import PCA\n",
      "def do_pca(d,n):\n",
      "    pca = PCA(n_components=n)\n",
      "    X = pca.fit_transform(d)\n",
      "    print pca.explained_variance_ratio_\n",
      "    print \"Percentage of variance captured:\", np.sum(pca.explained_variance_ratio_)\n",
      "    return X, pca\n",
      "\n",
      "X2, pca2 = do_pca(data, 2)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "#Plotting 2-D PCA\n",
      "df = pd.DataFrame({\"x\": X2[:, 0], \"y\": X2[:, 1], \"label\":np.where(y==1, \"check\", \"dollar\")})\n",
      "colors = [\"red\", \"blue\"]\n",
      "for label, color in zip(df['label'].unique(), colors):\n",
      "    mask = df['label']==label\n",
      "    plt.scatter(df[mask]['x'], df[mask]['y'], c=color, label=label)\n",
      "plt.legend()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#Classification Boundary"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from matplotlib.colors import ListedColormap\n",
      "def points_plot(Xtr, Xte, ytr, yte, clf):\n",
      "    X=np.concatenate((Xtr, Xte))\n",
      "    h = .02\n",
      "    x_min, x_max = X[:, 0].min() - .5, X[:, 0].max() + .5\n",
      "    y_min, y_max = X[:, 1].min() - .5, X[:, 1].max() + .5\n",
      "    xx, yy = np.meshgrid(np.linspace(x_min, x_max, 50),\n",
      "                         np.linspace(y_min, y_max, 50))\n",
      "\n",
      "    # just plot the dataset first\n",
      "    cm = plt.cm.RdBu\n",
      "    cm_bright = ListedColormap(['#FF0000', '#0000FF'])\n",
      "    f,ax = plt.subplots()\n",
      "    # Plot the training points\n",
      "    ax.scatter(Xtr[:, 0], Xtr[:, 1], c=ytr, cmap=cm_bright)\n",
      "    # and testing points\n",
      "    ax.scatter(Xte[:, 0], Xte[:, 1], c=yte, cmap=cm_bright, marker=\"s\", s=50, alpha=0.9)\n",
      "    ax.set_xlim(xx.min(), xx.max())\n",
      "    ax.set_ylim(yy.min(), yy.max())\n",
      "    ax.set_xticks(())\n",
      "    ax.set_yticks(())\n",
      "    Z = clf.predict_proba(np.c_[xx.ravel(), yy.ravel()])[:, 1]\n",
      "    Z = Z.reshape(xx.shape)\n",
      "    ax.contourf(xx, yy, Z, cmap=cm, alpha=.4)\n",
      "    cs2 = ax.contour(xx, yy, Z, cmap=cm, alpha=.4)\n",
      "    plt.clabel(cs2, fmt = '%2.1f', colors = 'k', fontsize=14)\n",
      "    return ax"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "is_train = np.random.uniform(0, 1, len(data)) <= 0.7\n",
      "train_x, train_y = data[is_train], y[is_train]\n",
      "test_x, test_y = data[is_train==False], y[is_train==False]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "pca = PCA(n_components=2)\n",
      "train_x = pca.fit_transform(train_x)\n",
      "test_x = pca.transform(test_x)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def plot_decision_boundary(clf, X):\n",
      "    w = clf.coef_.ravel()\n",
      "    a = -w[0] / w[1]\n",
      "    xx = np.linspace(np.min(X[:, 0]), np.max(X[:, 0]))\n",
      "    yy = a * xx - clf.intercept_ / w[1]\n",
      "    plt.plot(xx, yy)\n",
      "    plt.xticks(())\n",
      "    plt.yticks(())"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plot_decision_boundary(clf, X)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}