kfold-cv-lda-logreg-roc-py35.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# LDA & Logit K-Fold Cross Validation\n",
    "### Compatible with Python 2.7, 3.5, 3.6\n",
    "#### Plot ROC Curve with Optimal Cutoffs and Accuracy Rates\n",
    "John Bonfardeci\n",
    "2017-04-09\n",
    "Modified: 2017-04-17\n",
    "\n",
    "#### What this notebook does:\n",
    "* Create randomized K cross-validation (CV) groups\n",
    "* For-each CV:\n",
    "    * Split randomized data into (k-1)/k training data and 1/k testing\n",
    "    * Run LDA K times\n",
    "        * Find best cutoff with test cutoff set\n",
    "        * Output Confusion Matrix, misclassification, and accuracy rates\n",
    "        * Plot ROC combined with K curves (does not plot cutoff points - these are printed out though)\n",
    "        * Optional - output images and data to Excel\n",
    "    \n",
    "    * Run Logistic Regression K times\n",
    "        * Find best cutoff with test cutoff set\n",
    "        * Output Confusion Matrix, misclassification, and accuracy rates\n",
    "        * Plot ROC combined with K curves (does not plot cutoff points - these are printed out though)\n",
    "        * Optional - output images and data to Excel\n",
    "    \n",
    "Directions in Jupyter Notebook: Select `Cell -> Run All`. \n",
    "Look for printed output and ROC plots at bottom of notebook."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Change These Parameters\n",
    "\n",
    "# relative file path; can place in same working directory as notebook\n",
    "file_path = 'hof_data.csv'\n",
    "\n",
    "# the column names of the continuous predictors\n",
    "feature_names = ['H','HR','RBI','AVG','SLG','OBP']\n",
    "\n",
    "# the column names of the categorical predictors\n",
    "categorical_features = ['POS'] # DOES NOT HANDLE CATEGORICAL PREDICTORS AT THIS TIME\n",
    "# ...here's why: \n",
    "# http://stats.stackexchange.com/questions/158772/can-we-use-categorical-independent-variable-in-discriminant-analysis#158781)\n",
    "\n",
    "# name of target column\n",
    "target_name = 'HOF'\n",
    "\n",
    "# what value represents the positive value in a binary response: e.g. 1, 'Y', 'T', ...\n",
    "target_value = 'Y' \n",
    "\n",
    "# test thresholds to try\n",
    "test_cutoffs = [0.10, 0.25, 0.50, 0.75, 0.90] \n",
    "\n",
    "# LDA Solver to use. \n",
    "#    Options: \n",
    "#        'lsqr' (Least squares solution) - can use with shrikage\n",
    "#        'svd' (Singular value decomposition) - use with large num features\n",
    "#        'eigen' (Eigenvalue decomposition) - can use with shrikage\n",
    "# http://scikit-learn.org/stable/modules/generated/sklearn.discriminant_analysis.LinearDiscriminantAnalysis.html#sklearn.discriminant_analysis.LinearDiscriminantAnalysis\n",
    "lda_solver = 'lsqr'\n",
    "\n",
    "# change to None to run LDA without shrinkage\n",
    "# Options: None, 'auto', float between 0 and 1\n",
    "# see http://scikit-learn.org/stable/auto_examples/classification/plot_lda.html#sphx-glr-auto-examples-classification-plot-lda-py\n",
    "lda_shrinkgage = 'auto'\n",
    "\n",
    "# number of cross-validation groups to run\n",
    "num_cv = 5\n",
    "\n",
    "# Use Dimensionality reduction with Principal Component Analysis\n",
    "use_pca = True\n",
    "pca_n_components = 2 # number of eigenvalues to keep; should explain 90% or more of the variance\n",
    "\n",
    "# Turn on feature (variable) selection and remove features with low variance.\n",
    "# More at http://scikit-learn.org/stable/modules/feature_selection.html\n",
    "#feature_selection = True\n",
    "\n",
    "# Feature selection removes all features whose variance doesn’t meet this threshold.\n",
    "#feature_selection_threshold = 0.8\n",
    "\n",
    "output_data_to_excel = False\n",
    "output_dir = 'lda-logit\\\\'\n",
    "save_figs = False"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "All of the links below proved helpful in the researh for writing this notebook.\n",
    "* http://sebastianraschka.com/Articles/2014_python_lda.html#introduction\n",
    "* http://scikit-learn.org/stable/modules/generated/sklearn.discriminant_analysis.LinearDiscriminantAnalysis.html#sklearn.discriminant_analysis.LinearDiscriminantAnalysis\n",
    "* https://en.wikipedia.org/wiki/Sensitivity_and_specificity\n",
    "* https://en.wikipedia.org/wiki/Youden%27s_J_statistic\n",
    "* http://stats.stackexchange.com/questions/29719/how-to-determine-best-cutoff-point-and-its-confidence-interval-using-roc-curve-i\n",
    "* https://sasshowcase.wordpress.com/category/sas-macro/\n",
    "* http://stackoverflow.com/questions/28719067/roc-curve-and-cut-off-point-python\n",
    "* https://www.medcalc.org/manual/roc-curves.php\n",
    "* http://www.ncss.com/wp-content/themes/ncss/pdf/Procedures/NCSS/One_ROC_Curve_and_Cutoff_Analysis.pdf\n",
    "* http://www.umich.edu/~ners580/ners-bioe_481/lectures/pdfs/1978-10-semNucMed_Metz-basicROC.pdf\n",
    "* http://scikit-learn.org/stable/modules/cross_validation.html\n",
    "* http://stats.stackexchange.com/questions/95797/how-to-split-the-dataset-for-cross-validation-learning-curve-and-final-evaluat\n",
    "* http://scikit-learn.org/stable/modules/generated/sklearn.model_selection.StratifiedKFold.html\n",
    "* http://scikit-learn.org/stable/auto_examples/model_selection/plot_roc.html\n",
    "* http://gim.unmc.edu/dxtests/roc2.htm\n",
    "* https://codeyarns.com/2014/10/27/how-to-change-size-of-matplotlib-plot/\n",
    "* https://en.wikipedia.org/wiki/Receiver_operating_characteristic\n",
    "* http://scikit-learn.org/stable/modules/model_evaluation.html#confusion-matrix\n",
    "* http://scikit-learn.org/stable/modules/feature_selection.html"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# DON'T change below this line unless you know what your're doing.\n",
    "\n",
    "import math\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from scipy import interp\n",
    "import matplotlib.pyplot as plt\n",
    "from itertools import cycle\n",
    "from copy import copy\n",
    "\n",
    "from sklearn.metrics import roc_curve, auc, confusion_matrix\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis\n",
    "#from sklearn.feature_selection import VarianceThreshold\n",
    "from sklearn.decomposition import PCA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1 0 0 0 0]\n",
      "[[3771  755 2297  305  555  374]\n",
      " [1022   57  366  254  347  292]\n",
      " [1832  336 1122  277  485  337]\n",
      " [ 999  130  433  255  412  320]\n",
      " [ 260   37  109  236  393  304]]\n"
     ]
    }
   ],
   "source": [
    "# load the data into Pandas Dataframe object\n",
    "usecols = copy(feature_names).append(target_name)\n",
    "\n",
    "df = pd.read_csv(file_path, usecols=usecols) # returns new Pandas DataFrame \n",
    "\n",
    "# randomize rows (http://quabr.com/29576430/shuffle-dataframe-rows)\n",
    "#df = _df.iloc[np.random.permutation(len(df))]\n",
    "\n",
    "# create new series with target value converted to 1/0 for the binary responses\n",
    "y = pd.Series( map(lambda x: 1 if x == target_value else 0, df[target_name]) ).as_matrix() # <class 'pandas.core.series.Series'>\n",
    "\n",
    "# get our predictor variables as a multidimensional array (matrix)\n",
    "X = df[feature_names].as_matrix() # <class 'pandas.core.frame.DataFrame'>\n",
    "\n",
    "cat = df[categorical_features].as_matrix() # <class 'pandas.core.frame.DataFrame'>\n",
    "\n",
    "# check the X and y data\n",
    "print(y[0:5])\n",
    "print(X[0:5])\n",
    "\n",
    "# good to go..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def get_balanced_accuracy(tpr, fpr):\n",
    "    \"\"\"\n",
    "    Balanced Accuracy\n",
    "    Dr. Alan Dabney\n",
    "    \n",
    "    @param tpr: float (True Positive Rate - the Sensitivity)\n",
    "    @param fpr: float (False Positive Rate - the 1-Specificity)\n",
    "    @returns float\n",
    "    \"\"\"\n",
    "    return (tpr + (1-fpr)) / 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Confusion Matrix\n",
    "def get_confusion_matrix(cutoff, actual, prob):\n",
    "    \"\"\"\n",
    "    Putting a threshold/cutoff on the output of a regression problem, \n",
    "    to determine whether the value is greater or lower than a fixed threshold, \n",
    "    is a binary classification problem.\n",
    "    \n",
    "    TN | FP\n",
    "    -------\n",
    "    FN | TP\n",
    "    \n",
    "    For example: \n",
    "    \n",
    "       n=165   | Predicted NO | Predicted YES\n",
    "    ------------------------------------------\n",
    "    Actual NO  |       50     |      10       |\n",
    "    ------------------------------------------\n",
    "    Actual YES |       5      |      100      |\n",
    "    ------------------------------------------\n",
    "    \n",
    "    The diagonal elements represent the number of points for which the predicted label is equal to the true label,\n",
    "    while off-diagonal elements are those that are mislabeled by the classifier.\n",
    "    The higher the diagonal values of the confusion matrix the better, indicating many correct predictions.\n",
    "    \"\"\"\n",
    "    pred = []\n",
    "    for (x, y) in prob:\n",
    "        pred.append(1 if y > cutoff else 0)\n",
    "\n",
    "    return confusion_matrix(actual, pred)\n",
    "\n",
    "\n",
    "def get_tpr_fpr(cm):\n",
    "    \"\"\"\n",
    "    Sensitivity: TruePos / (True Pos + False Neg) \n",
    "    Specificity: True Neg / (False Pos + True Neg)\n",
    "    TN | FP\n",
    "    -------\n",
    "    FN | TP\n",
    "    \"\"\"\n",
    "    \n",
    "    tn = float(cm[0][0])\n",
    "    fp = float(cm[0][1])\n",
    "    fn = float(cm[1][0])\n",
    "    tp = float(cm[1][1])\n",
    "    \n",
    "    tpr = tp / (tp + fn)\n",
    "    fpr = 1-(tn / (fp + tn))\n",
    "    \n",
    "    return [tpr, fpr]   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def get_best_cutoff(actual, prob):  \n",
    "    \"\"\"\n",
    "    Get the best cutoff according to Balanced Accuracy\n",
    "    'Brute-force' technique - try all cutoffs from 0.01 to 0.99 in increments of 0.01\n",
    "    \n",
    "    @param actual list<float>\n",
    "    @param prob list<tuple<float, float>>\n",
    "    @returns list<float>\n",
    "    \"\"\"\n",
    "    best_tpr = 0.0; best_fpr = 0.0; best_cutoff = 0.0; best_ba = 0.0; \n",
    "    cutoff = 0.0\n",
    "    while cutoff < 1.0:\n",
    "        cm = get_confusion_matrix(cutoff=cutoff, actual=actual, prob=prob)\n",
    "        _tpr, _fpr = get_tpr_fpr(cm)\n",
    "        ba = get_balanced_accuracy(tpr=_tpr, fpr=_fpr)\n",
    "        \n",
    "        if(ba > best_ba and _tpr < 1.0):\n",
    "            best_ba = ba\n",
    "            best_cutoff = cutoff\n",
    "            best_tpr = _tpr\n",
    "            best_fpr = _fpr\n",
    "        \n",
    "        cutoff += 0.01\n",
    "        \n",
    "    return [best_tpr, best_fpr, best_cutoff, best_ba]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def get_accuracy(C):\n",
    "    \"\"\"\n",
    "    Get the accuracy rate of a confusion matrix\n",
    "    TN | FP\n",
    "    -------\n",
    "    FN | TP\n",
    "    \"\"\"\n",
    "    assert C.shape == (2,2), \"Confusion matrix should be from binary classification only.\"\n",
    "    # true negative, false positive, etc...\n",
    "    tn = C[0,0]; fp = C[0,1]; fn = C[1,0]; tp = C[1,1];\n",
    "\n",
    "    NP = fn+tp # Num positive examples\n",
    "    NN = tn+fp # Num negative examples\n",
    "    N  = NP+NN\n",
    "    accuracy = (tp+tn+0.)/N\n",
    "    return accuracy\n",
    "\n",
    "\n",
    "def show_confusion_matrix(C, class_labels=['0','1'], figsize=(6,6), fontsize=12, filename='roc-curve'):\n",
    "    \"\"\"\n",
    "    C: ndarray, shape (2,2) as given by scikit-learn confusion_matrix function\n",
    "    class_labels: list of strings, default simply labels 0 and 1.\n",
    "\n",
    "    Draws confusion matrix with associated metrics.\n",
    "    https://notmatthancock.github.io/2015/10/28/confusion-matrix.html\n",
    "    \"\"\"\n",
    "    assert C.shape == (2,2), \"Confusion matrix should be from binary classification only.\"\n",
    "    \n",
    "    # true negative, false positive, etc...\n",
    "    tn = C[0,0]; fp = C[0,1]; fn = C[1,0]; tp = C[1,1];\n",
    "\n",
    "    NP = fn+tp # Num positive examples\n",
    "    NN = tn+fp # Num negative examples\n",
    "    N  = NP+NN\n",
    "\n",
    "    fig = plt.figure(figsize=figsize)\n",
    "    ax  = fig.add_subplot(111)\n",
    "    ax.imshow(C, interpolation='nearest', cmap=plt.cm.gray)\n",
    "\n",
    "    # Draw the grid boxes\n",
    "    ax.set_xlim(-0.5,2.5)\n",
    "    ax.set_ylim(2.5,-0.5)\n",
    "    ax.plot([-0.5,2.5],[0.5,0.5], '-k', lw=2)\n",
    "    ax.plot([-0.5,2.5],[1.5,1.5], '-k', lw=2)\n",
    "    ax.plot([0.5,0.5],[-0.5,2.5], '-k', lw=2)\n",
    "    ax.plot([1.5,1.5],[-0.5,2.5], '-k', lw=2)\n",
    "\n",
    "    # Set xlabels\n",
    "    ax.set_xlabel('Predicted', fontsize=fontsize)\n",
    "    ax.set_xticks([0,1,2])\n",
    "    ax.set_xticklabels(class_labels + [''])\n",
    "    ax.xaxis.set_label_position('top')\n",
    "    ax.xaxis.tick_top()\n",
    "    # These coordinate might require some tinkering. Ditto for y, below.\n",
    "    ax.xaxis.set_label_coords(0.34,1.06)\n",
    "\n",
    "    # Set ylabels\n",
    "    ax.set_ylabel('Actual', fontsize=fontsize, rotation=90)\n",
    "    ax.set_yticklabels(class_labels + [''],rotation=90)\n",
    "    ax.set_yticks([0,1,2])\n",
    "    ax.yaxis.set_label_coords(-0.09,0.65)\n",
    "\n",
    "\n",
    "    # Fill in initial metrics: tp, tn, etc...\n",
    "    ax.text(0,0,\n",
    "            'True Neg: %d\\n(Num Neg: %d)'%(tn,NN),\n",
    "            va='center',\n",
    "            ha='center',\n",
    "            bbox=dict(fc='w',boxstyle='round,pad=1'))\n",
    "\n",
    "    ax.text(0,1,\n",
    "            'False Neg: %d'%fn,\n",
    "            va='center',\n",
    "            ha='center',\n",
    "            bbox=dict(fc='w',boxstyle='round,pad=1'))\n",
    "\n",
    "    ax.text(1,0,\n",
    "            'False Pos: %d'%fp,\n",
    "            va='center',\n",
    "            ha='center',\n",
    "            bbox=dict(fc='w',boxstyle='round,pad=1'))\n",
    "\n",
    "\n",
    "    ax.text(1,1,\n",
    "            'True Pos: %d\\n(Num Pos: %d)'%(tp,NP),\n",
    "            va='center',\n",
    "            ha='center',\n",
    "            bbox=dict(fc='w',boxstyle='round,pad=1'))\n",
    "\n",
    "    # Fill in secondary metrics: accuracy, true pos rate, etc...\n",
    "    ax.text(2,0,\n",
    "            'False Pos Rate: %.2f'%(fp / (fp+tn+0.)),\n",
    "            va='center',\n",
    "            ha='center',\n",
    "            bbox=dict(fc='w',boxstyle='round,pad=1'))\n",
    "\n",
    "    ax.text(2,1,\n",
    "            'True Pos Rate: %.2f'%(tp / (tp+fn+0.)),\n",
    "            va='center',\n",
    "            ha='center',\n",
    "            bbox=dict(fc='w',boxstyle='round,pad=1'))\n",
    "\n",
    "    ax.text(2,2,\n",
    "            'Accuracy: %.2f'%((tp+tn+0.)/N),\n",
    "            va='center',\n",
    "            ha='center',\n",
    "            bbox=dict(fc='w',boxstyle='round,pad=1'))\n",
    "\n",
    "    ax.text(0,2,\n",
    "            'Neg Pre Val: %.2f'%(1-fn/(fn+tn+0.)),\n",
    "            va='center',\n",
    "            ha='center',\n",
    "            bbox=dict(fc='w',boxstyle='round,pad=1'))\n",
    "\n",
    "    ax.text(1,2,\n",
    "            'Pos Pred Val: %.2f'%(tp/(tp+fp+0.)),\n",
    "            va='center',\n",
    "            ha='center',\n",
    "            bbox=dict(fc='w',boxstyle='round,pad=1'))\n",
    "\n",
    "\n",
    "    if save_figs:\n",
    "        if(use_pca):\n",
    "            filename = filename+'-pca'\n",
    "        \n",
    "        filename = '%s%s.png' % (output_dir, filename)\n",
    "        plt.savefig(filename)\n",
    "        print('Confusion matrix image was saved to: %s' % (filename))\n",
    "        \n",
    "    plt.tight_layout()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def show_roc(kfolds, mean_tpr, mean_fpr, title, lw=2, filename='roc'):\n",
    "    \"\"\"\n",
    "    Display and save ROC curve\n",
    "    \"\"\"\n",
    "    plt.figure(figsize=(12,12))\n",
    "    colors = cycle(['cyan', 'red', 'seagreen', 'darkorange', 'blue'])\n",
    "    \n",
    "    # Plot the ROC Curve for this CV group\n",
    "    i=0\n",
    "    for (k, color) in zip(kfolds, colors):\n",
    "        tpr, fpr = k[0], k[1]\n",
    "        roc_auc = auc(fpr, tpr)\n",
    "        plt.plot(fpr, tpr, lw=lw, color=color, label='ROC fold %d (area = %0.2f)' % (i+1, roc_auc))\n",
    "        i += 1\n",
    "    \n",
    "    # Plot the ROC Curve for logistic regression\n",
    "    plt.plot([0, 1], [0, 1], linestyle='--', lw=lw, color='k', label='Luck')\n",
    "\n",
    "    mean_tpr /= len(kfolds)\n",
    "    mean_tpr[-1] = 1.0\n",
    "    mean_auc = auc(mean_fpr, mean_tpr)\n",
    "    \n",
    "    plt.plot(mean_fpr, mean_tpr, color='g', linestyle='--', label='Mean ROC (area = %0.2f)' % mean_auc, lw=lw)\n",
    "    plt.xlim([-0.05, 1.05])\n",
    "    plt.ylim([-0.05, 1.05])\n",
    "    plt.xlabel('False Positive (1-Specificity)')\n",
    "    plt.ylabel('True Positive (Sensitivity)')\n",
    "    plt.title(title)\n",
    "    plt.legend(loc=\"lower right\")\n",
    "    \n",
    "    if save_figs:\n",
    "        if(use_pca):\n",
    "            filename = filename+'-pca'\n",
    "        \n",
    "        filename = '%s%s.png' % (output_dir, filename)\n",
    "        plt.savefig(filename)\n",
    "        print('ROC image was saved to: %s' % (filename))\n",
    "        \n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Output to Excel\n",
    "def output_excel(filename, prob, x, y, predicted, sheetname='Sheet1'):\n",
    "    \"\"\"\n",
    "    Output data k-fold data to Excel\n",
    "    \"\"\"\n",
    "    prob_cols = [target_name, 'Prob[0]', 'Prob[1]', 'Predicted']\n",
    "    col_names = []\n",
    "    \n",
    "    if(use_pca and len(feature_names) > len(x[0])):\n",
    "        for i in range(len(x[0])):\n",
    "            col_names.append('Prin%d'%(i))\n",
    "            \n",
    "    else:\n",
    "        col_names = np.asarray(feature_names)    \n",
    "    \n",
    "    col_names = np.append( np.asarray(col_names), np.asarray(prob_cols))\n",
    "    \n",
    "    data = []\n",
    "    \n",
    "    i = 0\n",
    "    for xrow, yrow, prob, pred in zip(x, y, prob, predicted):\n",
    "        a = []\n",
    "        for col in xrow:\n",
    "            a.append(col)\n",
    "\n",
    "        a.append(yrow)\n",
    "        a.append(prob[0])\n",
    "        a.append(prob[1])\n",
    "        a.append(pred)\n",
    "        data.append(a)\n",
    "        i+=1\n",
    "\n",
    "    df = pd.DataFrame(data=data, columns=col_names)\n",
    "    if(use_pca):\n",
    "        filename = filename+'-pca'\n",
    "        \n",
    "    filename = '%s%s.csv'%(output_dir, filename)\n",
    "    #writer = pd.ExcelWriter(filename, engine='xlsxwriter')\n",
    "    #df.to_excel(writer, sheet_name=sheetname)\n",
    "    #writer.save()\n",
    "    df.to_csv(filename, sep=',', encoding='utf-8')\n",
    "    print('Excel file was saved to: %s' % (filename))\n",
    "    \n",
    "\n",
    "def output_roc(filename, sheetname, tpr, fpr, cutoffs):\n",
    "    \"\"\"\n",
    "    Output ROC data to Excel\n",
    "    \"\"\"\n",
    "    data = []\n",
    "    for (t, f, c) in zip(tpr, fpr, cutoffs):\n",
    "        data.append([t, f, c])\n",
    "        \n",
    "    df = pd.DataFrame(data=data, columns=['Sensitivity', '1-Specificity', 'Cutoff'])\n",
    "    if(use_pca):\n",
    "        filename = filename+'-pca'\n",
    "       \n",
    "    filename = '%s%s.csv'%(output_dir, filename) \n",
    "    #writer = pd.ExcelWriter(filename, engine='xlsxwriter')\n",
    "    #df.to_excel(writer, sheet_name=sheetname)\n",
    "    #writer.save()\n",
    "    df.to_csv(filename, sep=',', encoding='utf-8')\n",
    "    print('ROC image was saved to %s'%(filename))\n",
    "\n",
    "            \n",
    "def output_cms(filename, data, sheetname='Sheet1'):\n",
    "    \"\"\"\n",
    "    Output Cutofs, TPR, FPR, Confusion, ... to Excel \n",
    "    \"\"\"\n",
    "    \n",
    "    cols = ['Method', 'FoldNum', 'Cutoff', 'Sensitivity', '1-Specificity', \n",
    "            'TrueNeg', 'FalsPos', 'FalseNeg', 'TruePos', 'Accuracy', 'AUC', 'Balanced Accuracy']\n",
    "\n",
    "    df = pd.DataFrame(data=data, columns=cols)\n",
    "    df.sort_values(by=['FoldNum', 'Cutoff'])\n",
    "    \n",
    "    if(use_pca):\n",
    "        filename = filename+'-pca'\n",
    "        \n",
    "    filename = '%s%s.csv'%(output_dir, filename)\n",
    "    #writer = pd.ExcelWriter(filename, engine='xlsxwriter')\n",
    "    #df.to_excel(writer, sheet_name=sheetname)\n",
    "    #writer.save()\n",
    "    df.to_csv(filename, sep=',', encoding='utf-8')\n",
    "    print('Excel file was saved to: %s' % (filename))\n",
    "    \n",
    "    \n",
    "def get_cm_row(method, foldnum, cutoff, actual, prob):\n",
    "    \"\"\"\n",
    "    Create new row for output: Cutofs, TPR, FPR, Confusion, ... to Excel \n",
    "\n",
    "    TN | FP\n",
    "    -------\n",
    "    FN | TP\n",
    "    \"\"\"\n",
    "    cm = get_confusion_matrix(cutoff, actual, prob)\n",
    "    tn = cm[0][0]; fp = cm[0][1]; fn = cm[1][0]; tp = cm[1][1]\n",
    "    tpr, fpr = get_tpr_fpr(cm)\n",
    "    accuracy = get_accuracy(cm)\n",
    "    _fpr, _tpr, cutoffs = roc_curve(actual, prob[:, 1])\n",
    "    _auc = auc(_fpr, _tpr)\n",
    "    ba = get_balanced_accuracy(tpr, fpr)\n",
    "    \n",
    "    return(method, foldnum, cutoff, tpr, fpr, tn, fp, fn, tp, accuracy, _auc, ba)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.95461248  0.04136611]\n"
     ]
    }
   ],
   "source": [
    "# Dimension Reduction with PCA\n",
    "def get_pca(X, n_components, svd_solver='full'):\n",
    "    pca = PCA(n_components=(n_components if len(X[0]) > 1 else 1), svd_solver=svd_solver)\n",
    "    pca.fit(X)\n",
    "    eigenvalues = pca.explained_variance_ratio_\n",
    "    print(eigenvalues)\n",
    "    return pca.transform(X)\n",
    "\n",
    "\n",
    "if use_pca:\n",
    "    X = get_pca(X, pca_n_components)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fold 1, TPR: 0.90000, FPR: 0.03608, Cutoff: 0.16000, BA: 0.93196\n",
      "Fold 2, TPR: 0.90000, FPR: 0.02577, Cutoff: 0.41000, BA: 0.93711\n",
      "Fold 3, TPR: 0.90000, FPR: 0.03109, Cutoff: 0.22000, BA: 0.93446\n",
      "Fold 4, TPR: 0.90000, FPR: 0.06218, Cutoff: 0.10000, BA: 0.91891\n",
      "Fold 5, TPR: 0.88889, FPR: 0.01554, Cutoff: 0.27000, BA: 0.93667\n",
      "Means:  TPR: 0.89778, FPR: 0.03413, Cutoff: 0.23200, BA: 0.93182\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtQAAALJCAYAAACUZbS1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x158a341c5c0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAbQAAAGmCAYAAAAUKnHWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n0IT/A1gIA4BVqWMqAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x158a399fba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Run LDA with cross-validation and plot ROC curves\n",
    "rand = np.random.RandomState(0)\n",
    "cv = StratifiedKFold(n_splits=num_cv, shuffle=True, random_state=rand)\n",
    "mean_tpr = 0.0\n",
    "mean_fpr = np.linspace(0, 1, 100) # `linspace` returns evenly spaced numbers over a specified interval.\n",
    "roc_kfolds = []\n",
    "cm_folds = []\n",
    "lda_actual = []\n",
    "lda_prob = []\n",
    "best_mean_cutoff = 0.0\n",
    "best_mean_tpr = 0.0\n",
    "best_mean_fpr = 0.0\n",
    "best_mean_ba = 0.0\n",
    "\n",
    "# For each cross-validation batch, fit model 80:20 split, plot ROC, and get cutoff\n",
    "# TODO - add feature selection per CV\n",
    "# train, test are of type <type 'numpy.ndarray'>\n",
    "i = 0\n",
    "for (train, test) in cv.split(X, y):\n",
    "\n",
    "    test_actual = y[test]\n",
    "    test_X = X[test]\n",
    "    train_X = X[train]\n",
    "    train_y = y[train]\n",
    "    \n",
    "    # train LDA on training dataset\n",
    "    lda = LinearDiscriminantAnalysis(solver=lda_solver,  \n",
    "                                     shrinkage=lda_shrinkgage).fit(train_X, train_y)\n",
    "       \n",
    "    # test LDA on test dataset\n",
    "    # predict probability returns <type 'numpy.ndarray'> (n_samples, n_classes)\n",
    "    prob = lda.predict_proba(test_X)\n",
    "    \n",
    "    # get predicted values: 1/0; default threshold/cutoff is 0.5\n",
    "    predicted = lda.predict(test_X)\n",
    "    train_predicted = lda.predict(train_X)\n",
    "    \n",
    "    # Compute ROC curve\n",
    "    # fpr = false positive rate (1-specificity)\n",
    "    # tpr = true positive rate (sensitivity)\n",
    "    fpr, tpr, thresholds = roc_curve(test_actual, prob[:, 1]) # returns <type 'numpy.ndarray'> x 3\n",
    "    \n",
    "    # add mean Sensitivity\n",
    "    mean_tpr += interp(mean_fpr, fpr, tpr)\n",
    "    mean_tpr[0] = 0.0\n",
    "  \n",
    "    if output_data_to_excel:\n",
    "        output_excel('lda-fold-%d-test' % (i+1), prob, test_X, test_actual, predicted, 'TestSet%d' % (i+1))\n",
    "        output_excel('lda-fold-%d-train' % (i+1), \n",
    "                     lda.predict_proba(train_X), train_X, train_y, train_predicted, 'TrainSet%d' % (i+1))\n",
    "    # /if...\n",
    "    \n",
    "    # get_cm_row(method, foldnum, cutoff, actual, prob)\n",
    "    # append confusion matrix rows for output to Excel\n",
    "    for cutoff in test_cutoffs:\n",
    "        cm_folds.append( get_cm_row('LDA', i+1, cutoff, test_actual, prob) )\n",
    "    \n",
    "    \n",
    "    lda_actual.extend(test_actual)\n",
    "    lda_prob.extend(prob)\n",
    "    \n",
    "    roc_kfolds.append([tpr, fpr])\n",
    "    \n",
    "    #output_cutoffs(i, tpr, fpr, thresholds, prob)\n",
    "    \n",
    "    best_tpr, best_fpr, best_cutoff, best_ba = get_best_cutoff(actual=test_actual, prob=prob)\n",
    "    print('Fold %d, TPR: %0.5f, FPR: %0.5f, Cutoff: %0.5f, BA: %0.5f' % (i+1, best_tpr, best_fpr, best_cutoff, best_ba))\n",
    "    best_mean_cutoff += best_cutoff\n",
    "    best_mean_tpr += best_tpr\n",
    "    best_mean_fpr += best_fpr\n",
    "    best_mean_ba += best_ba\n",
    "    \n",
    "    i += 1\n",
    "#/for (train, test)...\n",
    "\n",
    "best_mean_cutoff /= float(num_cv)\n",
    "print('Means:  TPR: %0.5f, FPR: %0.5f, Cutoff: %0.5f, BA: %0.5f'%(best_mean_tpr/float(num_cv),\n",
    "                                                    best_mean_fpr/float(num_cv), \n",
    "                                                    best_mean_cutoff, \n",
    "                                                    best_mean_ba/float(num_cv)))\n",
    "\n",
    "show_roc(roc_kfolds, mean_tpr, mean_fpr, 'ROC Curves for LDA with K-fold Cross-validation', 2, 'lda-roc')\n",
    "\n",
    "if output_data_to_excel:\n",
    "    output_cms('lda-cms', cm_folds, 'LDA_Test_Folds') # output to excel\n",
    " \n",
    "\n",
    "#Determine the best cutoff points and output confusion matrix\n",
    "best_tp = get_confusion_matrix(best_mean_cutoff, logit_actual, logit_prob)\n",
    "#best_accuracy =  get_confusion_matrix(0.5, logit_actual, logit_prob)\n",
    "show_confusion_matrix(best_tp, ['0','1'], (6,6), 12, 'lda-best-tp-cm')\n",
    "#show_confusion_matrix(best_accuracy, ['0','1'], (6,6), 12, 'lda-best-accuracy-cm')\n",
    " "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fold 1, TPR: 0.80000, FPR: 0.02577, Cutoff: 0.30000, BA: 0.88711\n",
      "Fold 2, TPR: 0.90000, FPR: 0.06186, Cutoff: 0.16000, BA: 0.91907\n",
      "Fold 3, TPR: 0.90000, FPR: 0.01036, Cutoff: 0.25000, BA: 0.94482\n",
      "Fold 4, TPR: 0.80000, FPR: 0.02073, Cutoff: 0.21000, BA: 0.88964\n",
      "Fold 5, TPR: 0.88889, FPR: 0.03109, Cutoff: 0.28000, BA: 0.92890\n",
      "Means:  TPR: 0.85778, FPR: 0.02996, Cutoff: 0.24000, BA: 0.91391\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtQAAALJCAYAAACUZbS1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzs3Xd8VFX+xvHPNyEQCAjSBESKq66CIiIugtJBKSLiSrEg\nFmyLvYsNLKvuomtFZC2IUlaQdQFZQAig2AugxB+4goKIgLQAgfTz++NeNMYkBMLMySTP21deZObe\nufPMZJAnZ849Y845RERERETkwMT5DiAiIiIiEstUqEVERERESkCFWkRERESkBFSoRURERERKQIVa\nRERERKQEVKhFREREREpAhVpEyhQzu8bMNprZLjOr5TtPQczsQjObG4HjDjezF4vYfomZLT7Y95vn\n+PXNbLGZ7TSzx/axbzcz+76I7a+b2YiDnbE0MrMKZubMrEl4+UUzG16cfQ/gvoaY2X8PNKuIFEyF\nWqSUM7PvzWxPWBA3mNk4M6uab592ZpYcFplUM5thZs3y7XOImT1pZmvDY60KL9cu5H7NzK43s+Vm\nlmZm68xsipmdEMnHWxJmlgA8AZzhnKvqnNtyEI75vZl1K3m6XznnJjjnzshzH87MjjoIx/2rc25o\neMwm4XErHOjx8j92MxtkZtvMrGMhN7kaWA8c4py740Dvt5jZDjezV8K/EzvM7P/M7H4zqxzJ+40G\n59xQ59xfS3ocMzvKzH7zYRPOuVedcz1LemwR+S0VapHY0Mc5VxVoCZwE3LV3g5m1BeYC/wEaAE2B\nZcD7ZnZkuE9FYD7QHOgBHAK0BTYDfyrkPp8CbgCuB2oCxwBvAb33N3xJSt1+OgxIBFL294bhLxD6\nf2IhzGwI8BzQ2zm3qJDdGgNfuwh/Ylj4S+CHQAWgjXPuEILXdW3gyAL2j9brT0TKKf3jIRJDnHMb\ngDkExXqvvwHjnXNPOed2Oue2OufuAT4CRoT7XAw0Avo55752zuU65zY55x5yzs3Kfz9mdjQwDDjf\nOZfsnMtwzu0OR1YfDfdZaGZD89zmN9MJwtHRYWb2P+B/Zva8mY3Kdz//MbObw+8bmNmbZvazmX1n\nZtfn2e9PZvZZOBK50cyeKCDzMcDK8OJ2M0sOr29nZp+GI/efmlm7PLdZaGYPm9n7wG4KKGNFMbMr\nzOxbM9tqZtPNrEGebWeY2crwfkeb2aK9z1fe58rM3g1vsix852BgAfezxsxODr+/MHxum4eXLzez\nt8LvR5jZ6+HN9h53e3jctnmONyocaf7OzPY5WmlmVwGPA2c65z4oZJ/XgAuB4eH9dTKzRDN72sx+\nMrMfzeyJ8Je7gm5/spktteBdlklApSIi3QpsAS52zq0BcM6tcc5d55xLsV+nRfzFzL4FVoT3cXr4\nOko1s0/MrE2e+788HJHfaWarzWxQeP0xZvZueJvNZjaxkPynhY8xLs91/c3si/D7tmb2kZltD5+P\npy14R6WgY/1muouZ3WnBSPyPwJB8+54dPm87LHj36d48m98N99kVfp1iZkPNbGGe2xf1nCw2s5Fm\n9kH4vMw2s5pF/FxEyi0VapEYYmYNgZ7At+HlKkA7YEoBu78BdA+/7wbMds7tKuZddQXWOec+KVli\nzgHaAM2AScBAMzMAMzsUOAOYHJaQGQQj64eH93+jmZ0ZHucp4KlwJPIP4WP7DefcNwQj8AA1nHNd\nwn/83waeBmoRTAd52347t3owcCVQDVhT3AdmZl2AR4ABQP3wtpPDbbWBqQTvJNQiKPrtCjqOc65D\n+O2J4TSVfxWw2yKgU/h9R2A10CHP5YJGjPdurxEe98PwcpswT22CX8Ze2vszKcQ1wANAV+fcZ4Xt\n5JwbDPwL+Gt4fwuB+4DWQAuCd1ZOI8+7K3uZWSWCd1heJng35D8Er53CdAOmFWMk/GzgFOCE8Gfy\nNsEvBrWAZ4BZZnaomR1C8Nro7pyrFub8MjzGw+HtDgUaEozSF+QDIIvg57HXBcDeAp5N8I5P7fD4\nPYCr9pEfMzsrvF0XgneJzsy3yy6CX2RqAH2AG8LbQPgaCH8eVZ1zn+Y7dqHPSb7HMITg3Z8k4OZ9\nZRYpj1SoRWLDW2a2E/gB2ATcH15fk+Dv8U8F3OYngn+8IfjHsqB9CrO/+xfmkXDEfA/wHuCA9uG2\n84APnXPrCUpPHefcA865TOfcauCfwKBw3yzgKDOr7Zzb5Zz7qJj33xv4n3PuNedctnNuEsFoZZ88\n+4xzzqWE27P247FdCLzsnPvCOZdBUBTbWnCyWC8gxTk3zTmXTVDoN+zHsfNbxK9FrT1Bkd97ubBC\nXZg1zrl/OudygFcJfhk4rIj9uxO82/HVfiUOXAiMcM797JzbRFDMBxew32kEr41nnHNZzrnJwJIi\njlvc1+dfnXPbwtdfH4KfyaTwZ/0awS8me6cwOeB4M0t0zv3knPs6vD4LaALUd86lO+feL+iOwnI/\nGTgfwMxqEJTfyeH2T51zH4f3vRoYy2/Ld2EGAC+F7yyl8eu7TnvvNzl8/eY655aF91ec48K+nxPC\n+/6fc243wS/uLQs6kEh5p0ItEhvOCUfOOgHH8mtR3gbkEpSi/OoTzJGG4O3xgvYpzP7uX5gf9n6T\nv3AQjHxNCL9vDDQI3w7fbmbbgeH8WvQuJxidW2HBtI29I3D70oDfjzqvIRgF/13G/fSbY4ej/1vC\nYzfg94993QHeDwSFub2Z1QfiCUboTwvLe3Vg6X4c65diH5YkgKqF7AvBCPUxwIt5R7ItWIli71SC\n2wu5bf7nP/9zn3e/dflGnIt6t6C4r8+8P9tCXwvOuR0Er8thwAYzm2nBFCKAW4AE4DMz+8qCueSY\n2b15Hv+z4b4TgT+HUzn+DHzsnFsX7n+smb0dTt3YQfDLRYEnBOfzm9dS/scQTiVZaMFUqVRgaDGP\nu/fY+/r7kfcXwd0U/VoRKbdUqEViSHgy2DhgVHg5jeDkrP4F7D6A4EREgHnAmWaWVMy7mg80NLPW\nReyTBlTJc7leQZHzXZ4EnGdmjQmmHrwZXv8D8J1zrkaer2rOuV4A4QjZ+UBd4DFgajEfy3qCsp5X\nI+DHIjIW12+OHeapFR77J4LpAXu3Wd7L+8s59y1BmbkOeDcsgBsIpqosds7lFnSzA72/fDYSTMFp\nD4zOk2lonqkEfyvktvmf//zP/V6/eb7y7FuYeUC/fUxVgd8+B0W+Fpxz/3XOdSMo6t8CL4TX/xQ+\n1voEhXusmTV1zj2Y5/FfG+77JcHP5Ux+O92D8HjLgaPCqUv3AfvKD8Fzc0S+zHlNJvh7dIRzrjrw\nYp7j7us1UJy/HyJSDCrUIrHnSaC7mZ0YXr4TGGLBEnfVwjmhDxGs4jEy3Oc1gtL6ZjhSFmdmtSxY\nt7hX/jtwzv2PoDxNsuDksooWnGA2yMzuDHdbCpxrZlUsWPLt8n0Fd84tIRg1fxGY45zbHm76BNhp\nZneYWWUzizez483sFAAzu8jM6oTFce9tCiqR+c0CjjGzCyw4UW0gwXzumcW4bV4J4ePf+1WB4JeD\nS82sZTgH+K8EI5LfE8xLPcHMzgn3HUbBv3DstZF9nxC5CLiWX6d3LMx3Ob+fCZ6j/TrRsiDhtJyu\nQA8z+8d+3HQScJ+Z1TazOsC9wOsF7LcYiDOza8Of0wCgVRHHHUUwCvuKmTWC4PwCM3vKwpM1CzAT\naG5mA8P7uAA4imBOfX0z6xOek5BJ8MtibnjcAWa2d8R2O0FJzSki20TgJoK/f1PzXF8NSAXSzOw4\nijF/OvQGcFn49zaJX6d75T3uVudcupmdyq/TpCCYHuYsXO2nAIU+J8XMJiIhFWqRGOOc+xkYTzDC\nhXNuMcGI2LkEo1lrCE4AOz0sxoRzfLsRzB9+B9hBUGJrAx8XclfXA88SnIS1HVgF9CM4eRDgHwTl\nYyPBXNwJBRyjIBPDLL+M3oXzec8imJ/5Hb+W7urhLj2AFDPbRXCC4qBwXmyRXLAO9VkEb9tvAW4H\nznLObS7yhr83C9iT52uEc24eQUF8k+B5/wNhmQmP35/gpL8tBCX+MyCjkOOPAF4Np7sMKGSfRQTl\n6d1CLv9GOJ3jYYLlE7eHZeuAOefWEpwYd56ZPVLMm40kONF0OcFJfh8TzP/Of+wMgtfWFQTTmPoR\nLNFYWJbNBIUV4NPw/IJ3CF43qwu5zc8EJyneQfAzuYngtbCNYBrNbQQ/xy0EJ5AOC2/aJryPNGAa\nMCx8LgozkeB5eic89l63EJzct5NgtLqgk08Lyj2D4O/gIuCb8HHmdQ3wSPgcDCfPCbvOuZ0Ez/fH\n4WvgN+847eM5EZH9YPs+SVpERErCglVM1gEXOucW+M4jIiIHl0aoRUQiwMzONLMa4XSQ4QTzWou7\nOomIiMQQFWoRkchoSzBNZjPB8mTnFGeaioiIxB5N+RARERERKQGNUIuIiIiIlEAF3wH2V+3atV2T\nJk18xxARERGRMu7zzz/f7Jyrs6/9Yq5QN2nShM8++8x3DBEREREp48ysqE9t/YWmfIiIiIiIlIAK\ntYiIiIhICahQi4iIiIiUgAq1iIiIiEgJqFCLiIiIiJSACrWIiIiISAmoUIuIiIiIlIAKtYiIiIhI\nCahQi4iIiIiUgAq1iIiIiEgJqFCLiIiIiJSACrWIiIiISAmoUIuIiIiIlIAKtYiIiIhICahQi4iI\niIiUgAq1iIiIiEgJqFCLiIiIiJSACrWIiIiISAmoUIuIiIiIlIAKtYiIiIhICahQi4iIiIiUgAq1\niIiIiEgJqFCLiIiIiJSACrWIiIiISAmoUIuIiIiIlEDECrWZvWxmm8xseSHbzcyeNrNvzexLM2sV\nqSwiIiIiIpESyRHqcUCPIrb3BI4Ov64Eno9gFhERERGRiKgQqQM75941syZF7NIXGO+cc8BHZlbD\nzOo7536KVKaomtYbvpu1z916vziTWSt6RyFQKVR1Aww6B5I2/X7bMyshNwE63wctXtd2bdd2bdd2\nbdf2crl9N8x5Cvd/A3+/bykSsUJdDIcDP+S5vC687neF2syuJBjFplGjRlEJV2LFKNNA+S3TAGl1\n4OPr4JxLIT6r4H0qb4FDvyv8GNqu7dqu7dqu7dpetrdXfBTnBmBmhe/vmQUDxBE6eDBCPdM5d3wB\n22YCjzrnFoeX5wN3OOc+K+qYrVu3dp99VuQupcPj4Q/9FseJowYBsOzWyb/bbe9rwznY+zKJ3E/E\ns7wPNpSZk8m6Het+t2vTGk0xMzbv3syOjB3aru3aru3aru3aXs62/7DpBy6++GKuuOAKrr/6+t/t\nGw1m9rlzrvU+9/NYqF8AFjrnJoWXVwKd9jXlQ4U6huV5sDf89wYaVGvAdW2uo0pCFb+5REREpFRw\nzvHaa68xYMAAEhMTycrKIiEhwVue4hZqn8vmTQcuDlf7OBVILTPzp6VIKZtSeOaTZ7h3wb1sSitg\n/rSIiIiUO8457rrrLoYMGcKAAQNwznkt0/sjYnOozWwS0AmobWbrgPuBBADn3BhgFtAL+BbYDVwa\nqSxSuty38D4cjitPvpImNZr4jiMiIiKeOee4+eabefLJJ6lQoQKDBw+mNM+Zzi+Sq3ycv4/tDhgW\nqfuX0unz+jDt/6aRWCGRu9vf7TuOiIiIeJabm8t1113H6NGjSUhIYMqUKfTt29d3rP3ic5UPKYfu\n7RL8ee0p11K/Wn2/YURERMS71atXM378eCpVqsS0adPo1auX70j7TYVaomrYJ7Cra3vuOP0O31FE\nRETEI+ccZsZRRx3FrFmzSE9Pp3v37r5jHRAVaomq3v+D3pe+6zuGiIiIeJSdnc3FF1/Mn/70J268\n8Ubat2/vO1KJ+FzlQ8qR5O+SuasrbK3sO4mIiIj4lJmZyaBBg5g0aRL33XcfGzdu9B2pxDRCHcOc\nc7gCVq02DDMrNdudc9w1/y4+aQ+HpsPtB/h4RUREJLZlZGTQv39/ZsyYQfXq1ZkzZw6HHXaY71gl\npkIdw8YvG88l/7nkd9cvvnQxpzU6rdRtr7srmEMtIiIi5U92djbnnHMOs2fPpmbNmsydO5eTTz7Z\nd6yDQlM+YpwV8F9p3F4pvhJ/eweSsg7igxcREZGYUaFCBdq1a0ft2rVJTk4uM2UaIvzR45Ggjx4P\npnoEt42dBc+B3z5YERERKRd27drF999/z/HHHw/Apk2bqFu3rudUxRMLHz0uB2jllpU0eKIBN/z3\nBt9RRERERAqVmprKmWeeSYcOHVi2bBlAzJTp/aFCHYPmr57Phl0b2LR7k+8oIiIiIgXatm0b3bt3\n54MPPqBq1aokJSX5jhQxOikxBiV/nwxAlyZdPCcRERER+b0tW7bQvXt3lixZQtOmTUlOTqZJkya+\nY0WMRqhjTK7LZcF3CwDoemRXz2lEREREfu+ee+5hyZIlHHXUUSxatKhMl2nQCHXMWbphKdvSt9G4\nemOa1mjqO46IiIjI7/z9738nPT2dhx9+mAYNGviOE3EaoY4xleIrcUnLS7jghAtib5UPERERKbPW\nrVvHkCFD2LVrF1WrVuWVV14pF2UaNEIdc5rXbc4rfV/xHUNERETkF2vWrKFLly6sXr2apKQkRo8e\n7TtSVGmEOoZk5WSxYvMKYm3tcBERESm7Vq9eTYcOHVi9ejWtW7fmoYce8h0p6lSoY8gnP37Ccc8d\nR5fxWt1DRERE/Pvmm2/o0KEDa9eupW3btsybN4+aNWv6jhV1KtQxZP538wFoXqe55yQiIiIisGfP\nHnbv3k379u2ZM2cO1atX9x3JC82hjiHJ34XrTzfVCLWIiIj4s2HDBurVq8eJJ57Ie++9R5MmTcr0\nB7fsi0aoY8TurN18uO5DDKNTk06+44iIiEg59cUXX9C8eXMefPBBAJo3b16uyzSoUMeM99e+T2ZO\nJifVP4malcvf3CQRERHx75NPPqFr165s3bqVTz75hJycHN+RSgVN+YgRJ9U/iXF9x5FYIdF3FBER\nESmHPvjgA3r06MHOnTvp168fkydPJj4+3nesUkGFOkbUrlKbIS2H+I4hIiIi5dCGDRs488wz2bVr\nFwMHDuS1114jISHBd6xSQ1M+YkBqeirPf/o832z5xncUERERKYfq1avHQw89xODBg3n99ddVpvPR\nCHUMWLRmEX+Z9RdOO+I0Fl+22HccERERKSdmz55NUlIS7du354YbbsA5h5n5jlXqqFDHgPmrg/Wn\nuzbt6jmJiIiIlBczZszgvPPOo2LFiixZsoSjjjpKZboQmvIRA5K/1/rTIiIiEj1vvvkm5557LpmZ\nmVx22WX84Q9/8B2pVFOhLuU27trI8k3LqVyhMqc2PNV3HBERESnjJk+ezMCBA8nOzua2227jySef\n1Mj0PqhQl3IfrfsIgNMbnU6lCpU8pxEREZGyzDnHxIkTycnJ4Z577uGxxx5TmS4GzaEu5foe25c1\nN65hR8YO31FERESkDMvKyiIhIYE33niDqVOnctFFF/mOFDM0Qh0DGlVvxPF1j/cdQ0RERMqo0aNH\nc+qpp7Jt2zYSExNVpveTCnUptjZ1LQOmDGD8svG+o4iIiEgZ9eSTTzJs2DC++OIL/vvf//qOE5NU\nqEux+avnM+XrKbz5f2/6jiIiIiJl0N/+9jduuukmAJ577jkuuOACz4lik+ZQl2J7l8vT+tMiIiJy\nsD311FPccccdmBljx45l6NChviPFLI1Ql1LOOZK/0/rTIiIiEhk9evTg8MMP55VXXlGZLiGNUJdS\nK7esZP3O9dRNqkvzOs19xxEREZEywDnH9OnTOfvss/njH//IypUrSUpK8h0r5mmEupTasGsDRx56\nJF2adtH6jyIiIlJizjluuukmzjnnHO666y4AlemDRCPUpVSnJp1Ydf0q0rPTfUcRERGRGJebm8u1\n117L888/T8WKFWnXrp3vSGWKCnUp5JzD4YizOBIrJPqOIyIiIjEsJyeHq666ipdeeolKlSrx1ltv\n0aNHD9+xyhRN+SiFlmxYQr1R9bhx9o2+o4iIiEiM++CDD3j55ZepXLkyM2fOVJmOAI1Ql0LJ3yXz\n8+6f2Zmx03cUERERiXHt27fn5ZdfpmnTpnTs2NF3nDJJI9SlkJbLExERkZLIzMzkggsuYPbs2QBc\ncsklKtMRpBHq0iYni3fXvAtA56adPYcRERGRWJOenk7//v2ZOXMmCxcuZNWqVVSuXNl3rDJNhbq0\nWf8paVlpHFf7OBpUa+A7jYiIiMSQPXv20K9fP+bMmUPNmjWZOXOmynQUqFCXNpUOYehJQ2lco7Hv\nJCIiIhJDdu/eTZ8+fUhOTqZOnTrMmzePFi1a+I5VLqhQlzZ1j+efZ//TdwoRERGJMRUrVqRWrVoc\ndthhJCcn06xZM9+Ryg0V6tIkOx22/A9X93h9OqKIiIgUS2pqKnv27KFevXpMmDCB9evX07ix3umO\nJq3yUZqsXQxjWnDG62f4TiIiIiIxYNu2bXTv3p3OnTuzceNGEhISVKY9UKEuTcLl8loe1tJzEBER\nESntNm/eTJcuXfj000/JyMggPT3dd6RyS4W6NPluPgBdj+zqOYiIiIiUZhs3bqRz584sXbqUo48+\nmnfffVcj0x6pUJcSqempsP4ziKvA6Y1O9x1HRERESrHLLruM5cuXc+yxx7Jo0SIaNmzoO1K5pkJd\nSry75l1wudDwVKpWrOo7joiIiJRizz33HL1792bhwoXUr1/fd5xyT6t8lNC1bz7Ge98t+d31y8JF\nOk4cNahYx2l3RDvo9zrsq0z37g2zZu1vTBEREYlx33//PWPHjuWhhx6iSZMmzJw503ckCalQl1BB\nZbog7ZueVOT2WlVqQYsL932gWC/TvXr5TiAiIhJzVq1aRZcuXVi7di3Vq1fnjjvu8B1J8lChPkiW\n3Tr5t1c8/q+Cry9IpR0898lr0LgDHHZC8e7Quf1MKCIiIrFo5cqVdOnShfXr19OuXTuuvvpq35Ek\nH82hLg2O+IBr/3stzLzSdxIREREpRVJSUujYsSPr16+nQ4cOzJ49m+rVq/uOJfmoUJcGjRYHfx6h\n1T1ERETkV6tWrfplvelZs2ZRrVo135GkAJryURoc8X7wp5bLExEREWDnzp1Uq1aNs88+m3feeYdT\nTz2VypUr+44lhdAItW/xmdDw4+D7I9r5zSIiIiLeffzxxxx55JFMmzYNgM6dO6tMl3Iq1B707g1m\nwRd1v4KEPRxb+1hIquM7moiIiHi0ePFiunfvzubNm3njjTd8x5FiUqH24Dcr3/10Ml2W/Mhr/V7z\nlkdERET8W7hwIT169GDnzp0MGjSI115TN4gVmkPt0a8r3zUIv0RERKQ8SklJoVevXuzZs4fBgwfz\nyiuvEB8f7zuWFJMKtUfOOS6bfhktD2vJX075C8Qn+I4kIiIiHhx33HGcf/75mBkvvPCCynSMUaH2\n6Nut3zJu6TgOSzqM69tc7zuOiIiIRNmsWbNo1qwZTZo0YezYsZgZcXGakRtr9BPzaPHaYP3p0xqd\nhpl5TiMiIiLRNHXqVPr27UuXLl3YunUr8fHxKtMxSj81j/YW6tP1gS4iIiLlysSJExk0aBDZ2dn0\n79+fQw891HckKQEVao8W//DrCLWIiIiUD6+++ioXXXQROTk53HvvvTz66KN6pzrGaQ61Lwm7cc5R\nJaEKJ9U7yXcaERERiYLMzExGjRqFc44HH3yQe+65x3ckOQhUqH3JqsI3131DanoqCVrdQ0REpMxz\nzlGxYkXmzp3LzJkzueKKK3xHkoNEUz48q55Y3XcEERERibAnnniCAQMGkJ2dTf369VWmyxgVal8u\nOItzJp/Dmu1rfCcRERGRCHr00Ue55ZZbmDp1KsnJyb7jSASoUPuQsBv+MIcZ38zg0Mo6q1dERKQs\ncs7xwAMPcNddd2FmvPjii5xxxhm+Y0kEaA61Dw0+hfhsWhzWkkMqHeI7jYiIiETAAw88wIgRI4iL\ni2PcuHEMHjzYdySJEI1Q+9BI60+LiIiUdW3btiUpKYmJEyeqTJdxGqH2odH7AJzeSIVaRESkLHHO\n8eGHH9KuXTvOOOMMvvvuO+rUqeM7lkSYRqh92NQcNjXTB7qIiIiUIbm5uVxzzTWcfvrpjB8/HkBl\nupxQofbhnb/D6BQaHtLQdxIRERE5CHJychg6dCgvvPAClSpVom7dur4jSRRpykeUpaanglUDp99l\nREREyoLs7GwuueQSJkyYQOXKlZkxYwZdu3b1HUuiSK0uyv4y6y9wey04ZqbvKCIiInIQvPHGG0yY\nMIGqVasye/ZslelySCPUUfbhDx9C5e2Q2sh3FBERETkIzj//fFJSUujduzft2rXzHUc80Ah1lG1M\n2xh8s62p3yAiIiJywNLT07n00ktZsWIFZsbDDz+sMl2OaYQ6inJdLruzdgcXsqr4DSMiIiIHZPfu\n3fTr14+5c+fy+eefs3TpUuLiNEZZnqlQR9EvZTqzCrh4v2FERERkv6WlpdGnTx8WLFhA3bp1mTBh\ngsq0aMpHtF198tWwbIjvGCIiIrKfdu7cSc+ePVmwYAH169dn4cKFnHDCCb5jSSmgEeooqlqxKs+f\n9Txj+vhOIiIiIvsrJyeHtLQ0GjZsSHJyMkcffbTvSFJKqFBHUa7LxTkHaLqHiIhIrNi2bRuJiYnU\nqFGDuXPnsmPHDpo21eIC8itN+YiiD3/4kAoPVoAhnX1HERERkWLYvHkznTt35txzzyUjI4NatWqp\nTMvvaIQ6inZl7gq+yU3wG0RERET2aePGjXTt2pWUlBT27NnDtm3bqFevnu9YUgpphDqKfinUmVX9\nBhEREZEirV+/nk6dOpGSkkKzZs1YtGiRyrQUSoU6itKy0oJvVKhFRERKLeccffr0YcWKFZxwwgks\nWLBAZVqKpEIdRRqhFhERKf3MjKeeeooOHTr8st60SFFUqKPouNrHcWWrK2Ht6b6jiIiISD7ffvst\nY8eOBeD0009n4cKF1KpVy3MqiQU6KTGKOjftTOemnRl7tu8kIiIikteKFSvo2rUr69ev59BDD6V/\n//6Yme9YEiM0Qh1FmTmZ5Lpc3zFEREQkj+XLl9OpUyfWr19Px44d6dmzp+9IEmNUqKPo2lnXEv9A\nPLR60XfHx42NAAAgAElEQVQUERERAZYtW0bnzp3ZuHEj3bp1Y9asWVStqnOdZP+oUEfRLyclZlX2\nG0REREQAmDdvHps3b6Znz57MmDGDKlWq+I4kMUhzqKPo10Kd5DeIiIhIOZeZmUnFihW55ZZbqF+/\nPn/+85+pVKmS71gSozRCHUVah1pERMS/xYsXc8wxx7B06VIALrjgApVpKREV6ijSOtQiIiJ+LVy4\nkB49erBmzRr++c9/+o4jZYSmfETRuceeS4u6LXhxZwPfUURERMqdd955h759+7Jnzx6GDBnC008/\n7TuSlBEq1FF0x+l3APBiqucgIiIi5czixYvp06cPGRkZXHHFFYwZM4a4OL1RLweHCnUU7cnaQ2KF\nREALxYuIiERTy5Ytad26NS1btuTpp59WmZaDSoU6ig597FAyczKhwm7ITvQdR0REpMx75513OPXU\nU6lWrRpz586lcuXK+gREOej061mUZOVkkZGTQZzFQbbOJBYREYm0iRMn0qNHD8466ywyMzOpUqWK\nyrREhAp1lOxdMi+pYhKa8iEiIhJZ48aN46KLLiI3N5fOnTuTkJDgO5KUYSrUUbJ3ybyqFbVknoiI\nSCSNHTuWSy+9FOccDz/8MCNGjNDItESU5lBHSVpmMEKtQi0iIhI527ZtY/jw4QD8/e9/59Zbb/Wc\nSMoDFeooqVqxKtf96TpqVa7FCN9hREREyqhDDz2UuXPn8vHHH3PNNdf4jiPlhAp1lBx+yOE83TNY\nQH6E3ygiIiJlziOPPEJmZib3338/rVq1olWrVr4jSTkS0TnUZtbDzFaa2bdmdmcB26ub2QwzW2Zm\nKWZ2aSTz+JSVk0VGdgbOOd9RREREygznHCNHjmT48OGMHDmSL7/80nckKYciVqjNLB54DugJNAPO\nN7Nm+XYbBnztnDsR6AQ8bmYVI5XJp2n/N43EhxMZ9OYg31FERETKBOccd999NyNGjCAuLo7x48fT\nokUL37GkHIrkCPWfgG+dc6udc5nAZKBvvn0cUM2CU2+rAluB7Ahm8mbvKh9JCUmek4iIiJQNt99+\nO4888gjx8fFMmjSJiy66yHckKaciWagPB37Ic3ldeF1ezwLHAeuBr4AbnHO5EczkRW9gaFioX8mz\nyofl+yr8AL3BLPgSERERAI444ggSEhKYMmUKAwYM8B1HyjHf61CfCSwFGgAtgWfN7JD8O5nZlWb2\nmZl99vPPP0c7Y4nNAggLNfsYoe5V4AFm5dupwL1ERETKvNzcXFasWAHA9ddfz4oVK+jXr5/nVFLe\nRbJQ/wgckedyw/C6vC4FprnAt8B3wLH5D+ScG+uca+2ca12nTp2IBY6o8JMSH8ozQu0K+Hq7qGM4\nF3y9XeReIiIiZVJOTg6XX345p5xyCh9++CEARx55pOdUIpEt1J8CR5tZ0/BEw0HA9Hz7rAW6ApjZ\nYcAfgdURzORPw1MZdsowTjn8FN9JREREYk52djYXX3wx48aNIzc3lz179viOJPKLiK1D7ZzLNrNr\ngTlAPPCycy7FzK4Ot48BHgTGmdlXBNOI73DObY5UJq/+eDbP/vFs3ylERERiTlZWFhdeeCFTpkyh\natWqzJo1i/bt2/uOJfKLiH6wi3NuFuEU4jzXjcnz/XrgjEhmKDUy08iKr0hCfILvJCIiIjHl8ccf\nZ8qUKRxyyCHMnj2btm3b+o4k8hu+T0osP6YOoOJDFXn7G81/FhER2R833ngjAwcOZP78+SrTUiqp\nUEdLuMpH1TwnJYqIiEjBdu/ezQ033MDWrVtJTExk8uTJtG7d2ncskQJFdMqH5KFCLSIiUiy7du2i\nT58+LFy4kFWrVjFz5kzfkUSKpEIdLXs/KbGiPilRRESkMDt27KBXr168//771K9fn1GjRvmOJLJP\nKtTRohFqERGRIm3fvp0ePXrw8ccf07BhQ5KTkzn66KN9xxLZJxXqaDlpKNfs/pkaiTV8JxERESmV\nNm/ezJo1a2jSpAnJyck0bdrUdySRYlGhjpbOIxntO4OIiEgptGPHDqpVq8ZRRx1FcnIySUlJNGrU\nyHcskWLTKh/R0DsXElOx+GzMfIcREREpPTZs2EDbtm259dZbcc5x3HHHqUxLzFGhjoaFW+GuGnDr\nYb9c1auXxzwiIiKlwI8//kjHjh35+uuvmT17Njt37vQdSeSAaMpHNFQMTkhsVL8qa5znLCIiIqXA\n2rVr6dKlC6tWreLEE0/knXfe4ZBDDvEdS+SAaIQ6GipqhQ8REZG90tPT6dy5M6tWreLkk08mOTmZ\nOnXq+I4lcsBUqKNBhVpEROQXiYmJ3H333bRt25Z58+ZRs2ZN35FESkSFOhrCQp2UoA91ERGR8mvF\nihXMnTsXgMsuu4z33nuPGjW0nKzEPs2hjoYdDeGDW/hzT62nKSIi5dPy5cvp2rUrqampLFq0iDZt\n2hAfH+87lshBoUIdDZuPhbmjGDbHdxAREZHoW7JkCd27d2fLli10796dE044wXckkYNKUz6iocIe\nqLiLXJfrO4mIiEhUffrpp3Tp0oUtW7bQu3dvpk+fTpUqVXzHEjmoVKij4ZTRMLwat8691XcSERGR\nqBo7dizbt2/nnHPOYdq0aSQmJvqOJHLQacpHNFRMA7TKh4iIlB+5ubnExcUxevRomjdvzrBhw0hI\nSPAdSyQiNEIdDVo2T0REypH58+fzpz/9iY0bN5KQkMCNN96oMi1lmgp1NKhQi4hIOTFnzhzOOuss\nPv/8c8aMGeM7jkhUqFBHg9ahFhGRcmDmzJmcffbZpKenc9VVV3Hvvff6jiQSFZpDHQ3/6wm7a3H8\nFcf7TiIiIhIRs2bN4txzzyUrK4vrrruOp556CjPzHUskKlSoo2H5+bD8fE5u4DuIiIhIZDRv3pz6\n9eszYMAA/va3v6lMS7miQh0NidsgpyLOVdH/YEREpExZvHgx7dq1o3HjxnzxxRfUrFlT/9ZJuaM5\n1NFwaUe4uyrLNi7znUREROSgefnll+nQoQNXX301zjlq1aqlMi3lkgp1NCRoHWoRESlbxowZw+WX\nX45zjiOPPFJFWso1TfmIBi2bJyIiZcjTTz/NDTfcAMATTzzBTTfd5DmRiF8q1NGgQi0iImXEqlWr\nuOWWWwB49tlnGTZsmOdEIv6pUEdYTm4OVNwNQJWEKp7TiIiIlMwf/vAHJk6cyPbt27niiit8xxEp\nFVSoIyzX5cK7w6FCBnGmKesiIhJ7nHOMHDmS448/nvPOO4/+/fv7jiRSqqhQR1hCfAIkP+w7hoiI\nyAFxznHXXXfx2GOPkZiYyGmnnUb9+vV9xxIpVVSoIywrJwuqpEJmVSDRdxwREZFic85x88038+ST\nT1KhQgXGjx+vMi1SABXqCEv5OQVuPwk2tAC0DrWIiMSG3NxcrrvuOkaPHk1CQgJTpkyhb9++vmOJ\nlEoq1BG2KzNY4SMYoRYREYkdGRkZVKpUiWnTptGrVy/fcURKLZ0lF2FpmcGHuqhQi4hILMjJyWHD\nhg3ExcXxwgsv8PHHH6tMi+yDCnWEaYRaRERiRXZ2NoMHD6Zt27b88MMPxMfHc+KJJ/qOJVLqqVBH\nmAq1iIjEgszMTAYNGsSkSZPYsmUL69at8x1JJGZoDnWENa/bHN67Ezac5DuKiIhIgTIyMhgwYADT\np0+nevXqzJkzhzZt2viOJRIzVKgjrHWD1jC/te8YIiIihbrtttuYPn06NWvWZO7cuZx88sm+I4nE\nFE35iLDU9FSovBXisnxHERERKdDw4cM57bTTSE5OVpkWOQAq1BE2YuEIuKMWtHnGdxQREZFf7Nq1\niwcffJDs7Gzq1avHe++9pxMQRQ6QpnxE2K8nJSb5DSIiIhJKTU2lV69efPDBB2zevJmnnnoKM/Md\nSyRmqVBHWFqW1qEWEZHSY9u2bfTo0YNPPvmERo0acf311/uOJBLzVKgjTMvmiYhIabFlyxa6d+/O\nkiVLaNq0KcnJyTRp0sR3LJGYpznUEaZCLSIipUVKSgpff/01Rx11FIsWLVKZFjlINEIdYReecCEL\nJrSGbUf6jiIiIuVURkYGlSpVokOHDsycOZNmzZrRoEED37FEygyNUEfY5a0uh3f+Btub+o4iIiLl\n0Lp16zjxxBN59dVXAejWrZvKtMhBpkIdYT/t/AkStwPOdxQRESln1qxZQ8eOHVm5ciXPPPMM2dnZ\nviOJlEkq1BF2/PPHw52HBh/uIiIiEiWrV6+mQ4cOrF69mtatWzN37lwqVNBMT5FI0N+sCPvlpMQs\nrUMtIiLRsWnTJjp06MCPP/5I27Zt+e9//0v16tV9xxIps1SoIyjLQWZOJuTGQ3Yl33FERKScqFOn\nDv379+ezzz5j1qxZVKtWzXckkTJNhTqC0vZ+k1kV0CdQiYhIZH311VckJCRw7LHH8sQTT5CRkUFi\nYqLvWCJlnuZQR9Cuvechag1qERGJsCVLltC5c2e6dOnC999/j5mpTItEiQr1wWL22y+gisGIBcDH\n1xW8T3G/REREivDJJ5/QpUsXtmzZQqtWrahXr57vSCLligp1BNU0uH8R8P4dJT9Yr14lP4aIiJQ5\nH3zwAd26dWP79u3069ePadOmaWRaJMo0h/pgcfnWmX7c2ONg+471UCstWOUj/z4iIiIl4Jzj7rvv\nZufOnQwYMIDXX3+dhIQE37FEyh2NUEfQ/Bxo8EQD6D/AdxQRESmDzIw333yT++67jwkTJqhMi3ii\nQh1BOilRREQiYfbs2Zx77rlkZGRQs2ZNRo4cqQ9tEfFIhTqCdu39Rh/qIiIiB8mMGTPo27cv//73\nvxk3bpzvOCKCCnVEpWmEWkREDqI333yTc889l8zMTK6//nquvPJK35FEBBXqiPplhFqFWkRESuhf\n//oXAwcOJDs7m9tuu40nn3wS09KqIqWCCnUEnR4P93a4F77r4juKiIjEuIYNG5KYmMjdd9/NY489\npjItUoroDIYI6hgPHTs/wIOrfScREZFYtXz5co4//nhOO+00UlJSaNy4se9IIpKPRqgj6GcHG3dt\nhPhM31FERCQGjR49mhYtWvDss88CqEyLlFIq1BE0LAPqPV4Pjn3LdxQREYkxTz75JMOGDcM5R1ZW\nlu84IlIEFeoI0jrUIiJyIB577DFuuukmAJ599tlfvheR0klzqCPo10KtdahFRKR4Fi9ezJ133omZ\n8cILL3DFFVf4jiQi+6BCHUFaNk9ERPbX6aefzogRI2jSpAlDhgzxHUdEikGFOoL0wS4iIlIczjlG\njhxJ3759Oemkk7j//vt9RxKR/aA51BF0fQLc3f5uSKvrO4qIiJRSzjluuukmRo4cSa9evUhLS/Md\nSUT2k0aoI2hYRaDLQzyc7juJiIiURrm5uVx77bU8//zzJCQkMGbMGJKSdN6NSKxRoY4Q5+B7B1XT\nfgbq+I4jIiKlTE5ODldddRUvvfQSlSpV4t///jc9e/b0HUtEDoAKdYSkA0fuhkr/OCK8JCIi8quM\njAxWrlxJ5cqVmT59Ot26dfMdSUQOkOZQR0if1ycAkLFTJySKiMivsrKy2LVrF1WqVGHWrFksWLBA\nZVokxqlQR8j8NW2Db8IVPnr18hhGRERKhczMTAYOHPjLyYfVqlWjTZs2vmOJSAlpykekVAxWoW52\ndBIpbh/7iohImZeRkcF5553HzJkzqV69OqtWraJFixa+Y4nIQaAR6kipGCx7VLWipnyIiJR3e/bs\noW/fvsycOZOaNWuSnJysMi1ShmiEOlJ2NoD5D3HFk4f5TiIiIp4NGTKEOXPmUKdOHebNm6cyLVLG\nqFBHSmojeO9uhrbyHURERHy78847Wb58OVOnTqVZs2a+44jIQaYpH5FSKRWqr2FHxg7fSURExIPU\n1FReeuklAFq1asVXX32lMi1SRqlQR8rxk+GmJtw29zbfSUREJMq2bt1Kt27dGDp0KM8++ywA8fHx\nnlOJSKRoykekhKt86KREEZHyZfPmzXTv3p2lS5fStGlT+vTp4zuSiESYRqgjRYVaRKTc2bhxI507\nd2bp0qUcffTRvPvuuzRu3Nh3LBGJMBXqSAkLdVLFJM9BREQkWubMmcPy5cs57rjjWLRoEQ0bNvQd\nSUSiQFM+IkUj1CIi5UZubi5xcXFcfPHF5Obm0qtXL+rWres7lohEiUaoI2VlX0h+gDaH6yNlRUTK\nsu+//56WLVvy/vvvA3DJJZeoTIuUMyrUkfJtD3j3Xk5ucLLvJCIiEiGrVq2iY8eOfPXVV4wYMQLn\nnO9IIuKBpnxESvW14IzMnMOoGF/RdxoRETnIVq5cSZcuXVi/fj1t27Zl6tSpmJnvWCLigUaoI+W8\nQXBzIz798VPfSURE5CBbvXo1nTp1Yv369XTo0IE5c+ZQvXp137FExBONUEeKTkoUESmzGjRoQIsW\nLWjWrBnTp08nKUkrOomUZyrUkaJCLSJS5nz55Zc0bNiQmjVr8tZbbwFQuXJlz6lExDdN+YgUFWoR\nkTLl448/pkOHDpxxxhmkpqZSuXJllWkRAVSoI0cf7CIiUma8//77dO/endTUVJo0aaIiLSK/oSkf\nkZL8EFTcSZX7qvhOIiIiJbBw4ULOOuss0tLSGDRoEK+99hoVKuifTxH5lf6PECkf3gxAnFZQEhGJ\nWVlZWQwdOpS0tDQGDx7MK6+8Qnx8vO9YIlLKaMpHBOzJ2gN1vobKW3xHERGREkhISGDGjBnccsst\nKtMiUigV6ghYvmk5DGsOg8/wHUVERA7Af/7zH+644w6ccxx33HGMGjVKZVpECqUpHxGwKW1T8E1a\nXb9BRERkv02dOpXzzz+f7Oxs2rdvz1lnneU7koiUchqhjgAVahGR2DRx4kQGDRpEdnY2t99+O717\n9/YdSURigAp1BPy8++fgm911/AYREZFie/XVVxk8eDA5OTnce++9PProo5jpzHIR2TdN+YgAjVCL\niMSe3NxccnNzefDBB7nnnnt8xxGRGKJCHQF9/9iXx0ceBt938h1FRET24YcffuCII47g0ksvpVWr\nVpx44om+I4lIjNGUjwho37g9fHAbrD/FdxQRESnCP/7xD4455hjmzZsHoDItIgdEhToCvtz4JRy6\nGuKyfUcREZFCPProo9x8882kp6ezatUq33FEJIapUEdA74m94YY/QLUffUcREZF8nHM88MAD3HXX\nXZgZL774IldddZXvWCISwzSH+iBzzv16UqJW+RARKXWmTJnC/fffT1xcHOPGjWPw4MG+I4lIjFOh\nPsh2ZOwgMycTMpMgq4rvOCIikk+/fv0YOHAg55xzDoMGDfIdR0TKAE35OMh+WYNaS+aJiJQazjke\neeQRNmzYQEJCApMmTVKZFpGDRoX6INMa1CIipUtubi7XXHMNw4cPp3fv3uTm5uoDW0TkoFKhPsia\n1GjCmN5j4KMbfUcRESn3cnJyGDp0KC+88AKJiYk8/PDDxMXpnz4RObgi+n8VM+thZivN7Fszu7OQ\nfTqZ2VIzSzGzRZHMEw0NqjXgqtZXwXK9lSgi4lN2djZDhgzhlVdeoXLlysycOZMePXr4jiUiZVCR\nJyWaWX1gINAeaADsAZYDbwNznXOuiNvGA88B3YF1wKdmNt0593WefWoAo4Eezrm1Zhbz8yRWb1tN\nanoqJDaF9Bq+44iIlFubN2/m/fffJykpiVmzZtGhQwffkUSkjCq0UJvZi0BTgvL8FLAJSASOAc4B\n7jez251ziws5xJ+Ab51zq8PjTQb6Al/n2ecCYJpzbi2Ac25TyR5OBPXuDbNmAXDtZe1577gGADzD\nu3Swn+DxfwHwVAY8nQW0fAI+uslXWhGRciszM5MKFSpQr149kpOT2bBhA23btvUdS0TKsKJGqJ9x\nzi0r4PqlwBtmlgg0KuL2hwM/5Lm8DmiTb59jgAQzWwhUA55yzo3PfyAzuxK4EqBRo6LuMoLCMg38\nUqaBoEznsWnvmL1OShQRibr09HTOO+886tevzwsvvEDTpk1p2rSp71giUsYVOod6b5k2s55WwOnQ\nzrl059w3Jbz/CsDJQG/gTOBeMzumgPsa65xr7ZxrXaeO5w9LyTPLZdmtk3+9/hYHtzg2HdEluKwP\ndRERiao9e/bQt29f3n77bf7973/zww8/7PtGIiIHQXFOShwC/M/M/mpmR+/HsX8EjshzuWF4XV7r\ngDnOuTTn3GbgXeDE/biPUufnNK1DLSISbWlpafTu3Zu5c+dSp04dFixYQOPGjX3HEpFyYp+F2jk3\nCGhNUIYnmtl7ZnaZmSXt46afAkebWVMzqwgMAqbn2+c/wOlmVsHMqhBMCfm//X4UpYjWoRYRiS7n\nHGeffTYLFiygXr16LFy4kBNOOMF3LBEpR4q1bJ5zbjswEXiVYN70+cAyM/tLEbfJBq4F5hCU5Dec\ncylmdrWZXR3u83/AbOBL4BPgRefc8hI8Hu+e7vk0T/V4CtI05UNEJBrMjGuuuYbGjRuzaNEimjVr\n5juSiJQzVsTKd8EOZr2AS4FmwOvAOOfcT+EI9dfOuai+p9a6dWv32WefRfMuA3unkTvHiaOCNaaX\n3ToZHg+vv8UVtruIiETA1q1b+fTTTznzzDOB4ITExMREz6lEpCwxs8+dc633tV+R61CHLgSed84l\n573SOZdmZlccaMCyKDU9laUblnL4IYcDR/mOIyJSZm3evJlu3bqRkpLC9OnT6dmzp8q0iHhTnCkf\na/KXaTP7K4Bzbm5EUsWopRuW0unVTlzy1iW+o4iIlFkbN26kU6dOLFu2jCOPPJIWLVr4jiQi5Vxx\nCnVBn9Pa+2AHKQv2npBYN0knJIqIRML69evp1KkTKSkpNGvWjIULF3L44Yf7jiUi5VxRn5R4FXA1\n8Ecz+yLPpmrA55EOFot+3h0smadCLSISGWPGjGHFihWccMIJzJs3j7p19f9bEfGvqDnUbwDzgUeA\nO/Ncv7NUf0S4RxqhFhGJDOccZsaIESNISEjgL3/5C7Vq1fIdS0QEKHrKR6Zz7lvgcuDnPF/pZnZI\nNMLFGhVqEZGD79tvv6VTp06sXbuWuLg47r33XpVpESlVihqhngr0BFIAB1i+PxtFPF2MGdpqKG0O\nb0Obhm18RxERKRNWrFhB165dWb9+PcOHD+f111/3HUlE5HcKLdTOuZ7hn0cUto/8Vqv6rWhVv5Xv\nGCIiZUJKSgpdu3Zl48aNdOzYkTFjxviOJCJSoH2u8mFm08ysv5lVjkagWDZ/9Xw+WvcRmTmZvqOI\niMS05cuX06lTJzZu3Ei3bt2YNWsWVatW9R1LRKRAxVk27zmgO7DSzCab2TlmVjHCuWLSgKkDaPtS\nW1LTU31HERGJabVr16ZWrVr07NmTGTNmUKVKFd+RREQKtc9PSnTOzQfmm1kFgmJ9OTAOqBHZaLEl\nOzebrXu2Yhg1K9f0HUdEJCZ9/fXXHH300dSrV49FixZRo0YNKlWq5DuWiEiRijNCjZlVAs4GLgFa\nA5MimCkmbd69GYDaVWoTHxfvOY2ISOxZvHgxbdq0YfDgwWRnZ3PYYYepTItITNjnCLWZTQROA94B\nXgQucM7lRDpYrNGSeSIiB27BggWcddZZ7N69m7i4Yo31iIiUGvss1MAEYIhzLivSYWLZz2nBpyTW\nSarjOYmISGyZO3cuffv2JT09nSFDhvDSSy8RH693+kQkdhT10eMdnXOLgHigp5n9ZrtzbnqEs8WU\nEw47gWkDplG1os5CFxEpru3btzNgwADS09O54oorGDNmjEaoRSTmFDVC3R1YBPQvYJsDVKjzqJtU\nl37H9fMdQ0QkptSoUYNJkyYxd+5cHn/8cZVpEYlJRX2wyz3ht3c759bm3WZm+pTEfJZuWMrGXRtp\ncVgL6ler7zuOiEipNmXKFNLS0rjkkkvo2bMnPXv29B1JROSAFWco4K1iXleujf50ND0m9OA/K//j\nO4qISKk2YcIEBg0axGWXXcaSJUt8xxERKbGi5lAfAxwHVDezs/NsOgRIjHSwWKNVPkRE9m3cuHFc\ndtllOOe4//77admype9IIiIlVtQc6ubAuQQf4JJ3HvVO4KpIhopFKtQiIkUbO3YsV10V/PPx8MMP\nM3z4cM+JREQOjqLmUP8b+LeZne6cWxzFTDHp593BsnntW9WBLZ7DiIiUQitXrgRg1KhR3HLLLZ7T\niIgcPEVN+bjFOfc48GczOzf/dufczRFNFmP2jlCTlmeEupefLCIipcnWrVupWbMmo0aNok+fPnTq\n1Ml3JBGRg6qokxJXhX8uB1IK+JI83hzwJkydCOk1cI5gYcG3facSEfHrr3/9K82bN+ebb77BzFSm\nRaRMKmrKx1vhny/tvc6CT3ep4pxLi0K2mNLtyG7Brx4iIoJzjpEjRzJy5EjMjM8++4xjjjnGdywR\nkYjY57J5ZjbezA4xsyrAV8C3ZqbpHnlsyoUpKVOg/ue+o4iIeOec4+6772bkyJHExcUxfvx4Lrjg\nAt+xREQipjjrULdwzu0AzgHeARoDl0QyVKz5IhcGTB0A3e7yHUVExLtnnnmGRx55hPj4eCZNmsRF\nF13kO5KISEQVp1AnmFkFoC/wH+dcJpAb2VixZZMLv0nTknkiIhdddBGnnHIKU6ZMYcCAAb7jiIhE\nXHEK9YvAWuBQYFH4seO7Ipoqxvz8S6Gu4zWHiIgvubm5PPvss6Snp1OzZk0++ugj+vXr5zuWiEhU\n7LNQO+f+4Zxr4Jw7wznngB+ALpGPFjs0Qi0i5VlOTg6XX3451133/+zdd3RVVd7G8e9OoYQSqU4E\nFKQZICRIEIRBIIChCAhIsQECFqQMyDiIfcQCzisi0kZpiiCOKFIMQbpDJ5EWerFAACkSSoDU/f4R\nyIC03JCbk5s8n7WyTO49Ofchk1l5srPP7/SnW7duAHh5ZWS9RkQkd7jRnRIBMMbkI23/dPk/Hf+u\nmzJ5HBVqEcmrkpOT6d69OzNmzMDPz4/nnnvO6UgiItkuI0sIs4EupJXplMve5KKh+eD7x76H/c2d\njlwzi0MAACAASURBVCIikm2SkpJ47LHHmDFjBoULFyYyMpKwMP0BU0TynpuuUAN3WWtruD2JB6vi\nBVUqt4JTTicREck+u3btIiIigqJFixIZGcn999/vdCQREUdkpFCvNcZUs9Zud3saDzUzCQrtmgfe\nD0JKfqfjiIi4VUpKCt7e3tSoUYPvv/+eQoUKERoa6nQsERHHZKRQ1wU2GmP2AgmAAay19l63JvMg\nPRPg/My24H1GhVpEcrVz587Rvn17HnroIfr370+jRo2cjiQi4riMFOqH3Z7Cg8VbOA8U8CnAhcRC\nTscREXGbs2fP0qZNG5YvX87mzZt58sknue2225yOJSLiuIyMzdsHlAIaXHw/DkhydzBPcWnCR+lC\npUlbvBcRyX1Onz5NixYtWL58OQEBASxfvlxlWkTkopsWamPMq8AbwKsXHyoAzHBnKE9yZaEWEcl9\nEhISePDBB1m1ahXlypXjxx9/5J577nE6lohIjpGRLR+PALWAnwCstbHGmKJuTeVBLt0lsZSf7pIo\nIrlT/vz5CQ8P5/fff2fp0qVUqFDB6UgiIjlKRuZQJ1y8Q6IFMMb4uTeSZ/mrN/y3ILzV5C2no4iI\nZKljx46xZcsWAN588002btyoMi0icg0ZKdTfGmPGAv7GmKeAH4DJ7o3lOW4zaaU69A6NjBKR3OPI\nkSM0btyYJk2aEBMTgzFGe6ZFRK7jpls+rLUjjDEtgUSgJvCOtXaB25N5iBUpcCAVGpz8GdDKjYh4\nvtjYWMLCwti9ezfVqlWjZMmSTkcSEcnRMrKHGmvtAmPMWuCvwCH3RvIsE5Pgi2SY+uuPqFCLiKf7\n7bffCAsLY9++fQQHB7No0SJKldI1IiIiN3LdLR/GmO+MMTUuvv8XYBvwPDDTGNM/m/LleJryISK5\nycsvv8y+ffuoXbs2S5cuVZkWEcmAG61QV7bWxlx8/ylgibX2yYsTPlYCH7s9nQdQoRaR3GT8+PH4\n+/vzzjvvaM+0iEgG3eiixMtv3tIUiACw1p4GUt0ZypOkj80rpFUcEfFMO3fupHPnzsTHx1OkSBHG\njh2rMi0i4oIbrVDHGmP6AAeB2kAnAGNMASBfNmTL8ay16SvUmkMtIp4oJiaGpk2bcvToUe6++26G\nDx/udCQREY9zo0LdC3gbeAh4zFp78uLj9YHP3B3MU0QVTFulLuhb0OkoIiIu2bhxI82bN+fEiRM0\nb96c119/3elIIiIe6bqF2lp7BOh9jceXAkvdGcpTGGOo6e10ChER123YsIEHH3yQuLg4WrduzaxZ\nsyhQoIDTsUREPNKNpnyMN8bcc53nChpjuhljHnVftJzvyNkjfJIE85OdTiIi4hofn7T1lIcffphv\nv/1WZVpE5BbcaMvHROBdY0wVYCtwDCgAVAZKAlOBse4OmJPtPrGbZxPgr15p+2JERHK6X3/9lbvu\nuotatWqxZs0aKlasiK+vr9OxREQ82nVXqK210dbaDqTtmZ4EbCDttuP9rLVB1toPrLUXsilnjpSU\nkjYIxdc4HEREJAOWLFlCtWrVePfddwG45557VKZFRLJARm49fhpYnA1ZPE5Salqh1sgTEcnpFi5c\nyMMPP8yFCxfYu3cv1lqM0WqAiEhWuNEcarmJ9BVqh3OIiNzIvHnzaNu2LRcuXODZZ59l4sSJKtMi\nIlnopivUkibmjWLUKHISPjBsvvhz6NvvvgJUqEUk5/r555/p2LEjSUlJ9O/fn48++khlWkQki2W4\nUBtj8ltrE9wZJierUeTkVY818Yb1BcG/3AMOJBIRubkKFSowfPhwDh8+zPvvv68yLSLiBjct1MaY\n+0i7KNEfuNMYEwz0ttb2d3e4HGmwTX+3GFDHuSQiItc1ffp0ypcvT4MGDXjhhRecjiMikqtlZA/1\naNKmwp0AsNZuBpq4M5Sn2HFsB59Ef8LK31Y6HUVEJN3kyZN58sknadmyJbGxsU7HERHJ9TJSqL2s\ntb/+6bEUd4TxNCt+XcGz859l2uZpTkcREQFg/Pjx9OrVC2stL7/8MmXKlHE6kohIrpeRPdQHLm77\nsMYYb6A/sNu9sTxD+pQPb12WKCLO++ijjxg4cCAAI0eOZNCgQQ4nEhHJGzJSqPuQtu3jTuB30mZS\n93FnKE9xaQ61r5cKtYg4KzU1lcjISADGjBlD3759HU4kIpJ3ZKRQJ1tru7o9iQfSCrWI5ATnz5+n\nYMGCfPvttyxevJg2bdo4HUlEJE/JyB7qDcaYCGNMd2NMEbcn8iBaoRYRJ1lref3112nYsCFxcXEU\nLFhQZVpExAEZufV4RWNMfaAr8E9jzCZgprV2ptvT5XC9avUivGI4txe+3ekoIpLHWGsZOnQoI0aM\nwNvbm9WrV9OqVSunY4mI5EkZuvW4tXa1tXYAcC9wGpju1lQeIqBIAHXK1OFO/zudjiIieYi1lhde\neIERI0bg4+PDzJkzVaZFRBx000JtjClsjHncGDMPWA8cA+q7PZkHWH1gNZ9Gf0rM0Rino4hIHvL6\n668zatQofH19mTVrFo888ojTkURE8rSMrFDHAPWA9621lay1g62169ycyyPMjJnJM/OfYcn+JU5H\nEZE85LHHHuPOO+/ku+++o127dk7HERHJ8zIy5eNua22q25N4IE35EJHskpKSwn/+8x+6du1KYGAg\nu3fvJn/+/E7HEhERblCojTEfWGsHA98YY+yfn7fWdnBrMg+gKR8ikh2Sk5Pp1q0bX375JTt27OCt\nt95SmRYRyUFutEL91cX/jsmOIJ4ovVBrhVpE3CQxMZHHHnuMb775hiJFihAeHu50JBER+ZPrFmpr\n7fqL7wZaa68o1caYfkCe3zicvuVDK9Qi4gYJCQl07tyZuXPn4u/vz8KFC6lbt67TsURE5E8yclFi\nz2s81iurg3iid8LeYV3vdYRX0oqRiGS95cuXM2/ePIoXL86SJUtUpkVEcqgb7aHuQtrNXCoYY769\n7KkiQJy7g3mCCsUqUKFYBadjiEguY63FGEN4eDhTpkwhJCSE4OBgp2OJiMh13GgP9XrgBFAWGHvZ\n42eAje4M5Sm+3/09h88epkWlFpQtWtbpOCKSC5w5c4ZOnToxePBgmjdvTvfu3Z2OJCIiN3GjPdQ/\nAz8Di7MvjmcZtW4Ui/cvZuETC1WoReSWnTp1ilatWrF69Wr27dvH9u3b8fXVNRoiIjndjbZ8rLDW\nNjLGnAQuH5tnAGutLe72dDmcLkoUkaxy8uRJwsPD2bBhA+XKlWPBggUq0yIiHuJGWz6aXPxvyewI\n4okujc3L553P4SQi4sni4uJo2rQpGzdupHz58ixbtozy5cs7HUtERDLoulM+Lrs7YjnA21qbAtwP\nPAsUyoZsOV5iSiKgOdQicmsKFSrE3XffTcWKFfnxxx9VpkVEPExGxuZ9B1hjTEVgClAZmOHWVB5C\nWz5E5FYcPnyYI0eO4Ovry4wZM1i1ahXlypVzOpaIiLjoRls+Lkm11iYZYzoAH1trRxtjNOUD+LLj\nl5xJPEOVElWcjiIiHiY2NpawsDB8fHxYvnw5pUqV4vbbb3c6loiIZEJGCnWyMaYT8CTw8MXHtCQL\nBJYKdDqCiHigX3/9lbCwMPbv309ISAjGGKcjiYjILcjonRKbAO9ba/cbYyoAX7o3lmf4YssXTPxp\nImcSzjgdRUQ8xP79+3nggQfYv38/oaGhLFmyhJIlde23iIgnu+kKtbU2xhgzAKhkjLkH2Gutfcf9\n0XK+IYuHcOjMIVpUakGR/EWcjiMiOZy1lu7du/Pbb79Rr149IiMj8ff3dzqWiIjcopuuUBtjGgJ7\ngUnAZGC3MaaBu4N5Al2UKCKuMMbw+eef07VrV3744QeVaRGRXCIjWz4+BFpZaxtYa+sDrYGP3BvL\nM1yaQ62xeSJyI1u3buXFF18kNTWVChUq8OWXX1KkiP6qJSKSW2TkosR81trtlz6w1u4wxuhOJmiF\nWkRubuPGjTRv3pwTJ05Qvnx5+vbt63QkERHJYhkp1D8ZYyYAX1z8+HFAY/PQCrWI3Nj69esJDw8n\nLi6O1q1b06tXL6cjiYiIG2Rky8dzwH7gHxff9pN2t8Q8q3VrMAYSx66FT9dR0Dc/mnolIpdbvXo1\nzZo1Iy4ujvbt2/Ptt99SoEABp2OJiIgb3HCF2hgTBFQEZltr38+eSDlfRMTFd47Uuuq5Vq2yN4uI\n5ExHjhzh3LlzdO7cmS+++AJfX/0lS0Qkt7puoTbGvAz0An4C6hhj3rLWTs62ZDmdVxLj1k6koG9B\neoT0cDqNiOQQJ06coESJEnTo0IEVK1ZQt25dfHwysrtOREQ81Y22fDwO1LTWdgLqAH2yJ5KH8D3H\n8xHPM2DBAKeTiEgOERkZSYUKFfjuu+8AaNCggcq0iEgecKNCnWCtjQew1h67ybF5j7cuSBSR/5k3\nbx7t2rXjzJkzLF261Ok4IiKSjW60dHK3Mebbi+8boOJlH2Ot7eDWZDmdl0bmiUiab775hq5du5Kc\nnMyAAQMYNWqU05FERCQb3ahQd/zTx2PcGcTjaIVaRIB169bRpUsXUlJS+Pvf/87777+P0dgfEZE8\n5bqF2lq7JDuDeBytUIsIUKdOHZ544gnKli3LsGHDVKZFRPIgXS2TWafLEv1MNN7G2+kkIuKAGTNm\n0KBBA+666y4mT56MMUZlWkQkj1KhzqyU/NwbcK/TKUTEAePGjaNv377cfffdbN68mcKFCzsdSURE\nHJThyR3GmPzuDOJxCh1l/IbxfLfzO6eTiEg2GjVqFH379gWgX79+KtMiInLzQm2Muc8YsxXYc/Hj\nYGPMx25PltMV28fzEc8zfOVwp5OISDYZMWIEgwYNAmDMmDHp74uISN6WkRXq0cBDwAkAa+1moIk7\nQ3kETfkQyVPi4+OZOnUqxhg++eST9FVqERGRjOyh9rLW/vqni21S3JTHc2jKh0ieYK0lNTWVQoUK\nsWTJElatWkWnTp2cjiUiIjlIRlaoDxhj7gOsMcbbGDMQ2O3mXDmfVqhFcj1rLUOGDOHxxx8nOTmZ\nO+64Q2VaRESukpFC3Qd4AbgT+B2od/GxvE0r1CK5mrWWQYMG8a9//YtvvvmGqKgopyOJiEgOddMt\nH9bao0DXbMjiWX59gOhnoimcT1f4i+Q2qamp9O3blwkTJuDr68vXX39NvXr1nI4lIiI51E0LtTHm\nU8D++XFr7TNuSeQpEvw1h1okl+rfvz8TJkwgf/78zJ49m5YtWzodSUREcrCMbPlYDCy5+LYKKA0k\nuDOURyi5kwlRE1jxywqnk4hIFmvTpg233XYb8+fPV5kWEZGbysiWj68u/9gYMw1Y6bZEnqLcavp8\n34ceIT1oVL6R02lE5BYlJSWxcuVKmjRpQosWLfjll1/w9/d3OpaIiHiADN8p8TIVgNuzOojH0UWJ\nIrlGYmIiXbp0oWnTpsyYMQNAZVpERDIsI3uoT/K/PdRewB/AS+4M5REujs3L553P4SAicisSEhJ4\n5JFHmD9/Pv7+/lSsWNHpSCIi4mFuWKhN2t1cgoHYiw+lWmuvukAxT9IKtYjHO3/+PO3bt2fhwoUU\nL16cRYsWce+9uthYRERcc8MtHxfLc4S1NuXim8r0Jbqxi4jHmzZtGgsXLqRUqVIsW7ZMZVpERDIl\nI7ce32SMqWWt3ejqyY0xLYCPAG9gorV2+HWOqwOsAbpaa2e5+jqO2NSdqP80pXSh0k4nEZFMevrp\npzl48CBdu3alWrVqTscREREPdd0VamPMpbJdC9hgjNlljPnJGLPRGPPTzU5sjPEGxgItgWrAo8aY\nq35iXTxuBPBDZv4Bjom/ndp31Kacfzmnk4iIC06dOkWHDh3YvXs3xhjeeustlWkREbklN1qhXg/c\nC7TN5LnvA/Zaa/cDGGNmAu2A7X86rj/wDVAnk6+Trcyld8quwURthrL14C8hTkYSkQw6efIkDz74\nIFFRUfz++++sXLmStEtFREREMu9Ge6gNgLV237XeMnDuMsCByz4+ePGx/72AMWWA9sD4G53IGPOM\nMSbKGBN17NixDLx0Nqj2DXzfB/Ytuu4hrbIxjojc2PHjxwkLCyMqKooKFSowY8YMlWkREckSN1qh\nLmWMeeF6T1prR2bB648ChlhrU2/0g81a+wnwCUBoaKijF0ZaLv6mcXHKx4fevgx0MpCI3NTRo0dp\n2rQpMTExVK5cmaVLl1K2bFmnY4mISC5xo0LtDRTmsl0OLooFLt9gXJb/jd+7JBSYebFMlwRaGWOS\nrbXfZfI1s4+3xuaJeApfX198fHwIDAxkyZIlBAQEOB1JRERykRsV6sPW2rdu4dwbgMrGmAqkFemu\nwGOXH2CtrXDpfWPMVGC+R5Rp+N8cao3NE8mxDh06RPHixSlWrBiLFi0iNTWV0qU1mUdERLLWTfdQ\nZ5a1NhnoBywEdgD/sdZuM8Y8Z4x57lbOnSNohVokR/vll1/461//yiOPPEJiYiIlS5ZUmRYREbe4\n0Qp101s9ubU2Aoj402MTrnNsj1t9vWy17C2ixvbnTv87nU4iIn+yb98+wsLC+O233yhZsiTnzp0j\nX758TscSEZFc6rqF2lr7R3YG8Tin7qT2HSrTIjnNrl27CAsL49ChQ9SvX5+IiAj8/f2djiUiIrnY\nDW89LjdQOYJ/R/2bX+N+dTqJiFyUnJxMmzZtOHToEA888ACRkZEq0yIi4nYq1Jl13xie+/45Yo7G\nOJ1ERC7y8fFh4sSJtG3bloiICIoUKeJ0JBERyQNUqDNLUz5EcoyffvqJCRPSLs944IEHmDNnDoUK\nFXI4lYiI5BU3uihRbkRTPkRyhHXr1hEeHs6pU6coV64crVu3djqSiIjkMVqhziytUIs4btWqVTRv\n3pxTp07RsWNHmjdv7nQkERHJg1SoM0sr1CKOWr58OeHh4Zw5c4auXbsyc+ZMjcYTERFHaMtHZn37\nBRu2nCawZKDTSUTypNWrVxMfH0+3bt2YPHky3t7eTkcSEZE8SoU6s05UIfQOp0OI5D3nzp3Dz8+P\noUOHEhgYSLt27fDy0h/bRETEOfoplFk1v+CT6E84deGU00lE8ow5c+ZQsWJFNm/ejDGG9u3bq0yL\niIjj9JMos5q8xrPzn+WP87qhpEh2mDVrFo888ghHjhzh22+/dTqOiIhIOhXqzPLWlA+R7DJjxgy6\ndu1KcnIy//jHP3jzzTedjiQiIpJOhTqzvBMBTfkQcbeIiAiefPJJUlJSeO211xg+fDjGGKdjiYiI\npNNFiZmlOdQi2aJhw4bUrVuXli1b8tprrzkdR0RE5Coq1JmlOdQibvXNN98QHh5OkSJFWL58uWZM\ni4hIjqUtH5k1cQ0bnt6An6+f00lEcp0PP/yQRx55hLZt25KSkqIyLSIiOZpWqDPraJDmUIu4wfDh\nwxk6dCgAnTp10g1bREQkx9MKdWaYVKgzlok/TXQ6iUiu8tZbbzF06FCMMUycOJE+ffo4HUlEROSm\nVKgzwzsBWvejX0Q/p5OI5BqHDx9m1KhReHl58dlnn9GrVy+nI4mIiGSItnxkhmZQi2QZay3GGAIC\nAvjhhx/Yv38/nTt3djqWiIhIhqlQZ4aXJnyIZAVrLQMHDqRUqVK8+uqrhIaGEhoa6nQsERERl6hQ\nZ4ZWqEVuWWpqKn379mXChAnky5ePRx99lIoVKzodS0RExGUq1JmhFWqRW5KSksIzzzzD5MmTKVCg\nALNnz1aZFhERj6VCnRnxt8P4TURs1u2PRVxlreWpp55i2rRpFCxYkHnz5tG0aVOnY4mIiGSapnxk\nRko++D2YmrfXdDqJiMcxxlCrVi0KFy5MZGSkyrSIiHg8FerM8DsOdcYya/ssp5OIeIzExES2bt0K\nwKBBg9i1axcPPPCAw6lERERunQp1Zvj/Cq378e5/33U6iYhHuHDhAh06dKBBgwasXbsWgDvu0K1G\nRUQkd1ChzgxN+RDJsPPnz9OuXTu+//578uXLR/78+Z2OJCIikqV0UWJmaMqHSIbEx8fTpk0bli1b\nRunSpVm8eDFBQUFOxxIREclSKtSZoRVqkQx55513WLZsGQEBASxZsoTAwECnI4mIiGQ5FerM0Aq1\nSIa89tprxMbG8uqrr1K5cmWn44iIiLiF9lBnxoH6MGEjY1qNcTqJSI7zxx9/0KtXL+Li4ihYsCCf\nffaZyrSIiORqWqHOoNYT5xOxszX8HaAIHAmhSgmnU4nkLMePH6d58+Zs2rSJ+Ph4Zs6c6XQkERER\nt9MKdQZF7Gz9vw9K7KZaj7Es2b/EuUAiOczvv/9O48aN2bRpE1WqVOH//u//nI4kIiKSLVSoXWQt\nfL5kHdvL92PKpilOxxHJEQ4dOkTjxo3Ztm0b1apVY/ny5ZQtW9bpWCIiItlChToTklI15UPkcmfO\nnOHkyZMEBQWlT/UQERHJK7SHOhOSUjTlQwTS9kyXKFGCqlWrps+aLlFCFxeIiEjeohXqTEhfoVah\nljxs79693HvvvQwZMgRrLYGBgSrTIiKSJ6lQZ0L6CrW2fEgetXPnTho1asSBAwdYtWoVCQkJTkcS\nERFxjLZ8ZMKTwU/SpEITShTUapzkPTExMTRr1ozff/+dRo0aMX/+fAoUKOB0LBEREceoUGdCSb+S\nlPQr6XQMkWx3+vRpmjZtytGjR2nWrBlz5szBz8/P6VgiIiKO0paPTFh7cC3jNowj+lC001FEslXR\nokV55513aN26NfPmzVOZFhERQYU6U+bumkvfiL5E7o10OopItli7di2LFi0CoHfv3sybN0/bPERE\nRC7Slo9M0EWJkpesXLmSli1bkpyczJo1awgJCcEY43QsERGRHEMr1JlwaWxePu98DicRca9ly5YR\nHh7O2bNnad++PTVq1HA6koiISI6jQp0JurGL5AU//PADrVq14ty5c3Tv3p1p06bh46M/aomIiPyZ\nCnUm6NbjkhdMmzaNCxcu8PTTTzN58mS8vb2djiQiIpIjGWut0xlcEhoaaqOiorL9dS9tGbUWDp4+\nyLH4Y5QtWpZShUplexYRd0pOTsbHx4ekpCSmTp1Kr1698PLS794iIpL3GGOirbWhNztOPyUzoWzR\nstQKqKUyLbnO119/Te3atTl69Ci+vr48/fTTKtMiIiI3oZ+UmRCxJ4JxG8ax/+R+p6OIZJnp06fT\ntWtXtmzZwsyZM52OIyIi4jFUqDPhk+hP6BvRl81HNjsdRSRLTJ06lSeffJLU1FTeeOMN+vfv73Qk\nERERj6FL9jNBFyVKbjJ9+nSeeuopAN555x1efvllhxOJiIh4FhXqTNDYPMlN6tevT7ly5fjb3/7G\n4MGDnY4jIiLicVSoM0Er1JIbLFy4kObNm1OhQgViYmIoWrSo05FEREQ8kvZQZ4JWqMXTvfvuu7Ro\n0YJ+/foBqEyLiIjcAq1QZ8JnD3/G6YTTVC5R2ekoIi6x1vLPf/6Tf/7znxhjqFOnjtORREREPJ4K\ndSZULF7R6QgiLrPW8sorr/Dee+/h5eXFZ599xhNPPOF0LBEREY+nQp0J07dM53TCabrW6EqxgsWc\njiOSIVu3buX999/H29ubGTNm0LlzZ6cjiYiI5Aoq1JnwzxX/ZM8fewirEKZCLR6jZs2afPHFF+TP\nn5/27ds7HUdERCTXUKHOBE35EE+RmppK//79adasGe3bt6dr165ORxIREcl1NOUjEzTlQzxBSkoK\nvXr1Yty4cfTo0YOTJ086HUlERCRX0gp1JmiFWnK65ORkunfvzowZM/Dz82P27NkUK6btSSIiIu6g\nQp0JWqGWnCwpKYnHH3+cr7/+msKFCxMREUHDhg2djiUiIpJrqVBnwsqeK0lITsC/gL/TUUSu4u3t\nTaFChfD39ycyMpJ69eo5HUlERCRXM9ZapzO4JDQ01EZFRWX76xqT9l8P+3JJHnLhwgVOnDhBmTJl\nSElJYf/+/VSurJsPiYiIZJYxJtpaG3qz43RRYiZ8vO5jxm8Yj6f9MiK517lz52jTpg0NGzbk4MGD\neHt7q0yLiIhkExVqF6WkpjAgcgD9FvTDXFq2FnHQ2bNnad26NYsXLyY+Pp5Tp045HUlERCRP0R5q\nF6VP+NAFiZIDnD59mlatWrFq1SoCAgJYunQp99xzj9OxRERE8hStULsoMSUR0Mg8yRn69evHqlWr\nKFu2LCtWrFCZFhERcYBWqF2kkXmSkwwfPpwjR47w73//mwoVKjgdR0REJE/SCrWLdFMXcdqxY8d4\n+eWXSU5O5o477uCHH35QmRYREXGQVqhdVKJgCbb22aoJH+KII0eO0LRpU7Zv3461lvfee8/pSCIi\nInmeCrWLfL19qVG6htMxJA+KjY0lLCyM3bt3U716dQYOHOh0JBEREUFbPlz2x/k/GL1uNDNjZjod\nRfKQ3377jUaNGrF7926Cg4NZtmwZt99+u9OxREREBBVql8WejuVvkX/j7R/fdjqK5CG7d+/mwIED\n1K5dm6VLl1KqVCmnI4mIiMhF2vLhIl2UKNkpPj6eQoUK0axZMyIjI6lVqxa33Xab07FERETkMlqh\ndpHG5kl22bFjB1WrVuWLL74AoEmTJirTIiIiOZAKtYu0Qi3ZISYmhsaNGxMbG8tnn32mqTIiIiI5\nmAq1i7RCLe62ceNGGjduzNGjR3nwwQeZM2cOxhinY4mIiMh1aA+1i+qUqcOW57ZQwKeA01EkF/r1\n118JCwsjLi6O1q1bM2vWLAoU0PeaiIhITqZC7aLC+QoTdHuQ0zEkl7rzzjt54oknOHjwIF999RX5\n8uVzOpKIiIjchAq1i/ac2EPk3kiqlKhCeKVwp+NILrFy5UpKly5NlSpV+Oijj0hJScHXV9uKRERE\nPIH2ULvop8M/MSByAJM2TnI6iuQSS5YsITw8nLCwMA4dOoSXl5fKtIiIiAdRoXaRpnxIVlq4cCEP\nPfQQ586do3nz5rr7oYiIiAdSoXaRpnxIVpk3bx5t27blwoULPPvss0yaNAlvb2+nY4mIiIiLTssQ\nmAAAIABJREFUVKhdlL5CrUIttyA1NZV33nmHxMRE+vfvz/jx4/Hy0v8dRUREPJEuSnRR+gq1tnxI\nJllr8fLyYv78+Xz++ecMGjRIc6ZFREQ8mPG0O7CFhobaqKiobH/dS33nePwJYs/EUqxAMcr5l8v2\nHOLZpk2bxqxZs/j66681Ek9ERCSHM8ZEW2tDb3acVqhdVMKvBCX8SjgdQzzQ5MmT6d27N9ZaZs+e\nTZcuXZyOJCIiIllAmzZdtO7gOj5e9zHrY9c7HUU8yIQJE+jVqxfWWt59912VaRERkVxEhdpFEXsi\nGBA5gIg9EU5HEQ8xbtw4+vTpA8AHH3zA0KFDHU4kIiIiWUmF2kWa8iGuCgoKolChQnz88ce88MIL\nTscRERGRLKY91C7SlA/JqA0bNlCnTh0aNmzI3r17+ctf/uJ0JBEREXEDrVC7SCvUcjPWWl5//XXu\nu+8+xo8fD6AyLSIikotphdpFiSmJgFao5dqstQwdOpQRI0bg5eWFv7+/05FERETEzVSoXfTaA6/x\nXOhzBBQOcDqK5DDWWl544QVGjRqFj48PM2bMoFOnTk7HEhERETdToXZRQJEAAoqoTMvVIiMjGTVq\nFL6+vnz99de0a9fO6UgiIiKSDVSoXbRgzwL2ndxHeMVwKpeo7HQcyUFatGjBG2+8wX333UerVq2c\njiMiIiLZRBclumjKpin0X9Cfnw7/5HQUyQFSUlIYMGAAmzdvxhjDm2++qTItIiKSx6hQuyh9yocu\nSszzkpOTeeKJJ/j44495+OGHSUxMdDqSiIiIOEBbPlx0acpHPu98DicRJyUmJvLYY4/xzTffUKRI\nEaZNm0a+fPqeEBERyYtUqF2UfmMXzaHOsxISEujUqRPz5s3D39+fhQsXUrduXadjiYiIiENUqF2k\nLR+SlJTEiRMnKFasGIsWLaJ27dpORxIREREHqVC7aEq7KZxOOM3dxe52Oopks3PnzpGamkrhwoVZ\nsGABBw4coHr16k7HEhEREYepULuo/G3lnY4gDjh79iwPPfQQxhi+//57ihYtqjItIiIigKZ8uGz6\nlumMWT+G4+eOOx1FssmpU6cIDw9nxYoV7Nq1iyNHjjgdSURERHIQrVC76L2V77Ht2DYa3dWIkn4l\nnY4jbnby5ElatGjB+vXrKVeuHEuXLuXuu7XdR0RERP5HhdpFuigxb+ncuTPr16+nfPnyLFu2jPLl\nyzsdSURERHIYFWoXaWxe3vLee+8RFxfHt99+S7ly5ZyOIyIiIjmQW/dQG2NaGGN2GWP2GmNeusbz\njxtjthhjthpjVhtjgt2ZJytohTr3O3z4MBMmTAAgNDQ0fbuHiIiIyLW4bYXaGOMNjAWaAweBDcaY\nudba7Zcd9jPQyFp70hjTEvgEyNF3yLh0p0StUOdOBw8eJCwsjD179uDj40Pv3r0xxjgdS0RERHIw\nd275uA/Ya63dD2CMmQm0A9ILtbV29WXHrwXKujFPlvixx49cSL6gCxJzoV9//ZWwsDD2799PcHAw\n7dq1czqSiIiIeAB3FuoywIHLPj7IjVefewEL3JgnS1QtWdXpCOIG+/fvp0mTJvz222+EhoaycOFC\nihcv7nQsERER8QA54qJEY0wT0gr1X6/z/DPAMwB33nlnNia72uh1o/EyXvQJ7YO3l7ejWSTrLF68\nmN9++4169eoRGRmJv7+/05FERETEQ7izUMcCl1/JVfbiY1cwxtQEJgItrbUnrnUia+0npO2vJjQ0\n1GZ91IwbGDkQi6VPaB8nY0gWSU5OxsfHh2eeeYZChQrRtm1bihQp4nQsERER8SDunPKxAahsjKlg\njMkHdAXmXn6AMeZO4FvgSWvtbjdmyRomBYvFYLQ6nQts2bKFe+65h7Vr1wLw+OOPq0yLiIiIy9xW\nqK21yUA/YCGwA/iPtXabMeY5Y8xzFw97HSgBjDPGbDLGRLkrT5bw1si83OKnn36iSZMm7Nu3j5Ej\nRzodR0RERDyYW/dQW2sjgIg/PTbhsvd7A73dmSFLeemmLrnB+vXrCQ8PJy4ujtatW/P55587HUlE\nREQ8mFtv7JLreKfNoM7nnc/hIJJZW7ZsoVmzZsTFxdG+fXu+/fZbChQo4HQsERER8WA5YsqHx0jw\nZ9vz27DW0esi5RZUrlyZunXrUqJECaZNm4avr/7aICIiIrdGhdoVqT5UK1XN6RSSCStXrqR69eoU\nK1aMuXPn4uvri4+Pvv1FRETk1mnLhysKxDFq7Si+2PKF00nEBZGRkTRr1ozw8HDOnj1LwYIFVaZF\nREQky6hQu6LQ7wxaOIhhPw5zOolk0Lx582jXrh0JCQnUrl0bPz8/pyOJiIhILqNC7QpvTfnwJN98\n8w0dOnQgMTGRAQMGMG7cOLy89C0vIiIiWUvtwhUXp3xoDnXOd/78ef72t7+RnJzM3//+d0aNGoUx\nxulYIiIikgtpI6krNIfaYxQsWJCFCxcyd+5cXnrpJZVpERERcRsValdc3PKhOdQ516RJk9i9ezfD\nhw+nevXqVK9e3elIIiIiksupULvi8L1se36bCnUONW7cOPr27QtA69ateeCBBxxOJCIiInmBCrUr\nkvw0hzqHGjVqFIMGDQJg5MiRKtMiIiKSbXRRoiuK7WPU2lHM3z3f6SRymX/961/pZXrMmDHp74uI\niIhkBxVqV9y+hUELBzHxp4lOJ5HLFC9eHC8vLz755JP0LR8iIiIi2UVbPlxxaQ61xuY5zlrLvn37\nqFSpEr169aJhw4ZUqVLF6VgiIiKSB2mF2hVemvKRE1hrGTJkCEFBQSxduhRAZVpEREQcoxVqV+hO\niY6z1jJo0CA++ugjfHx8OHnypNORREREJI9ToXbFpTslqlA7IjU1lX79+jF+/Hh8fX2ZNWsWbdu2\ndTqWiIiI5HEq1K7Y1pmYiAb4F/B3OkmeNGnSJMaPH0/+/PmZPXs2LVu2dDqSiIiIiAq1Sy7cRvXS\ntzmdIs/q0aMHy5Yto2fPnjRr1szpOCIiIiKALkp0TZn1fLT2I1YfWO10kjwjKSmJl156iaNHj+Lr\n68uMGTNUpkVERCRHUaF2xd2LGLhwoG7skk0SExPp2rUrI0aMoEOHDlhrnY4kIiIichVt+XCFpnxk\nm4SEBDp16sS8efPw9/fngw8+wBjjdCwRERGRq6hQu8JLN3bJDufPn6d9+/YsXLiQ4sWLs2jRIu69\n916nY4mIiIhckwq1K7RCnS1OnDjBjh07KFWqFIsXL6ZmzZpORxIRERG5LhVqV1ycQ607JbrHuXPn\nKFCgAGXLlmXp0qUkJCRQrVo1p2OJiIiI3JAKtSv+O5SYz57m9sK3O50k1zl16hQtW7akZs2ajB8/\nnooVKzodSURERCRDVKhdEX871UurTGe1kydPEh4ezoYNG4iNjeXo0aPcfru+ziIiIuIZNDbPFZUi\n+WjtR+w8vtPpJLnG8ePHCQsLY8OGDVSoUIEff/xRZVpEREQ8igq1K2pOY+DCgUQdinI6Sa6QkpLC\ngw8+yKZNm6hcuTI//vgjd911l9OxRERERFyiQu0KTfnIUt7e3gwZMoQaNWqwYsUKypYt63QkERER\nEZdpD7UrNIc6Sxw8eJAtW7bQqlUrunTpQocOHfD11ddURCSvS0pK4uDBg1y4cMHpKJLHXJoyltk+\nokLtCq1Q37JffvmFsLAwDh48yIIFC2jatKnKtIiIAGkLLkWKFKF8+fK6O65kG2stJ06c4ODBg1So\nUCFT59CWD1doDvUt2bdvH40aNeLnn38mJCREdz8UEZErXLhwgRIlSqhMS7YyxlCiRIlb+suIVqhd\nMWcSW/ec4i5/XTjnql27dhEWFsahQ4eoX78+ERER+Pv7Ox1LRERyGJVpccKtft+pULvidDlqlC7n\ndAqP9Omnn3Lo0CEeeOAB5s+fT5EiRZyOJCIiIpIltOXDFUHT+WjtR/x+9nenk3gMay0AI0aM4P33\n3yciIkJlWkREcixvb29CQkKoUaMGbdq0IS4uLv25bdu2ERYWRtWqValcuTLDhg1L/zkHsGDBAkJD\nQ6lWrRq1atVi8ODBV50/ISGBZs2aERISwldffXXdHI0bNyYq6uoxvVOnTqVfv35XPb5z507uv/9+\n8ufPz//93/9d97zWWsLCwjh9+vR1j3FadHQ0QUFBVKpUiQEDBlzxNb4kMTGRp556iqCgIIKDg1m+\nfHn6c19++SVBQUHUrFmTFi1acPz4cQDGjBnD5MmT3ZJZhdoV9T9g4MKBHDx90OkkHuGnn37ivvvu\n48CBA3h7e/Piiy9SqFAhp2OJiIhcV8GCBdm0aRMxMTEUL16csWPHAnD+/Hnatm3LSy+9xK5du9i8\neTOrV69m3LhxAMTExNCvXz+++OILtm/fTlRUFJUqVbrq/Bs3bgRg06ZNdOnSJctyFy9enNGjR/P3\nv//9hsdFREQQHBxM0aJFM3zulJSUW43nkj59+vDpp5+yZ88e9uzZQ2Rk5FXHfPrppwBs3bqVRYsW\nMXjwYFJTU0lOTuZvf/sby5YtY8uWLdSsWZMxY8YA0LNnTz7++GO3ZFahdoXG5mXYunXrCAsLIyoq\nivfff9/pOCIi4mGMm95ccf/99xMbGwvAjBkzaNCgAQ8++CAAfn5+jBkzhuHDhwPw/vvv88orr3DP\nPfcAaSvdffr0ueJ8R48e5YknnmDDhg2EhISwb98+lixZQq1atQgKCqJnz54kJCRclWPKlClUqVKF\n++67j1WrVl0za+nSpalTp85NJ2dNnz6ddu3apX/88MMPU7t2bapXr84nn3yS/njhwoUZPHgwwcHB\nrFmzhujoaBo1akTt2rUJDw/n8OHDQFqxrVOnDsHBwXTs2JFz587d8PVv5vDhw5w+fZp69ephjKFb\nt2589913Vx23fft2wsLC0v/tt912G1FRUVhrsdYSHx+PtZbTp09zxx13AGn/m5UvX57169ffUsZr\nUaF2haZ8ZMjKlStp3rw5p06domPHjnzwwQdORxIREXFJSkoKS5YsoW3btkDado/atWtfcUzFihU5\ne/Ysp0+fJiYm5qrn/6x06dJMnDiRhg0bsmnTJsqUKUOPHj346quv2Lp1K8nJyYwfP/6Kzzl8+DBv\nvPEGq1atYuXKlWzfvv2W/l2rVq26IufkyZOJjo4mKiqK0aNHc+LECQDi4+OpW7cumzdvpm7duvTv\n359Zs2YRHR1Nz549eeWVVwDo0KEDGzZsYPPmzQQGBjJp0qSrXnPZsmWEhIRc9Va/fv2rjo2Njb3i\nRm9ly5ZN/6XmcsHBwcydO5fk5GR+/vlnoqOjOXDgAL6+vowfP56goCDuuOMOtm/fTq9evdI/LzQ0\nlP/+97+Z/wJehy5KdIXmUN/UqlWraNGiBfHx8XTt2pVp06bh46NvMxERcc3Vu2azx/nz5wkJCSE2\nNpbAwECaN2/uttfatWsXFSpUoEqVKgB0796dsWPHMnDgwPRj1q1bR+PGjSlVqhQAXbp0Yffu3Zl+\nzT/++OOKa5lGjx7N7NmzAThw4AB79uyhRIkSeHt707Fjx/ScMTEx6V+LlJQUAgICgLStLq+++ipx\ncXGcPXuW8PDwq16zSZMmbNq0KdOZr6Vnz57s2LGD0NBQ7rrrLurXr4+3tzdJSUmMHz+ejRs3cvfd\nd9O/f3/ee+89Xn31VSDtl5qdO3dmaRZQoXbNxS0fWqG+vrvvvps77riDevXqMWXKFLy9vZ2OJCIi\nkmGX9lCfO3eO8PBwxo4dy4ABA6hWrRo//vjjFcfu37+fwoULU7RoUapXr050dDTBwcEOJc8YHx8f\nUlNT8fLyYvny5SxevJg1a9bg5+dH48aN02cxFyhQIP1nuLWW6tWrs2bNmqvO16NHD7777juCg4OZ\nOnXqFRcHXrJs2TIGDRp01eN+fn6sXr36isfKlCnDwYP/u1bt4MGDlClT5pr/jg8//DD94/r161Ol\nSpX04l6xYkUAOnfunL4tB9JmnRcsWPC6X5/M0pYPV3y2jK19tvKXwn9xOkmOExUVRXJyMgEBAaxe\nvVplWkREPJqfnx+jR4/mgw8+IDk5mccff5yVK1eyePFiIG0le8CAAfzjH/8A4MUXX+Tdd99NXz1O\nTU1lwoQJN3yNqlWr8ssvv7B3714Apk2bRqNGja44pm7duqxYsYITJ06QlJTE119/fUv/rqpVq7J/\n/34ATp06RbFixfDz82Pnzp2sXbv2up9z7Nix9EKdlJTEtm3bADhz5gwBAQEkJSUxffr0a37+pRXq\nP7/9uUwDBAQEULRoUdauXYu1ls8///yKPd+XnDt3jvj4eAAWLVqEj48P1apVo0yZMmzfvp1jx46l\nPxcYGJj+ebt376ZGjRoZ/XJlmAq1K/6oRI3SNXRR4p/MmTOH+vXr0717d1JSUihZsqTKtIiIeLxa\ntWpRs2ZNvvzySwoWLMicOXN4++23qVq1KkFBQdSpUyd9hF3NmjUZNWoUjz76KIGBgdSoUSO9uF5P\ngQIFmDJlCp06dSIoKAgvLy+ee+65K44JCAjgzTff5P7776dBgwZXlMPLHTlyhLJlyzJy5Ejefvtt\nypYte83ReK1bt05fRW7RogXJyckEBgby0ksvUa9evWueO1++fMyaNYshQ4YQHBxMSEhIehkeNmwY\ndevWpUGDBukXZN6qcePG0bt3bypVqkTFihVp2bIlAHPnzuX1118H0i7wvPfeewkMDGTEiBFMmzYN\ngDvuuIM33niDBx54gJo1a7Jp0yZefvnl9HOvWrXKLdt4zLVm++VkoaGh9lpzGd3NGKDuaEaNsvSp\n00fbPi6aNWsWjz76KMnJyQwcOJCRI0fqLlciIpIpO3bsuG5hlKxx+PBhunXrxqJFi5yOku02btzI\nyJEj08v3n13r+88YE22tDb3ZubVC7Yrm/2DgwoGkpGbvPMacasaMGXTt2pXk5GSGDBmiMi0iIpLD\nBQQE8PTTT+foG7u4y/Hjxxk2bJhbzq2LEl2hOdTpfv/9d3r37k1KSgqvvfYa//znP1WmRUREPEDn\nzp2djuAId05sUaHOKJMKXqkAeBvtD7799tv5z3/+w5YtW67YmyQiIiKS16hQZ5TX/2ZQ5+WV2HHj\nxlG4cGG6devGQw89xEMPPeR0JBERERFHqVBn0OnhBSgan7dnUH/44Ye88MILeHt7c//991O5cmWn\nI4mIiIg4ToU6g/yAnX6Q0iPr7//uCYYPH87QoUMBGDNmjMq0iIiIyEWa8pFB3gaqekG1UtWcjpLt\n3nrrLYYOHYoxhkmTJl01I1NERCS38Pb2JiQkhBo1atCmTRvi4uLSn9u2bRthYWFUrVqVypUrM2zY\nMC4fP7xgwQJCQ0OpVq0atWrVYvDgwVedPyEhgWbNmhESEsJXX3113RyNGzfmWmOCp06dmj77+nLT\np0+nZs2aBAUFUb9+fTZv3nzN81prCQsLy9FTPqKjowkKCqJSpUoMGDCAa414TkxM5KmnniIoKIjg\n4OD02dpnzpwhJCQk/a1kyZLpt3IfM2YMkydPdktmFeoMOm3hg0SYsnGK01GylbWWEydO4OXlxWef\nfUbPnj2djiQiIuI2l249HhMTQ/HixRk7diyQdmfEtm3b8tJLL7Fr1y42b97M6tWrGTduHAAxMTH0\n69ePL774gu3btxMVFUWlSpWuOv/GjRsB2LRpE126dMmy3BUqVGDFihVs3bqV1157jWeeeeaax0VE\nRBAcHEzRokUzfO6UlOwdF9ynTx8+/fRT9uzZw549e4iMjLzqmE8//RSArVu3smjRIgYPHkxqaipF\nihS54m6Md911Fx06dACgZ8+efPzxx27JrEKdQcct/D0R3vrxLaejZAtrLb///jvGGEaNGsXatWt5\n8sknnY4lIiJ5hTHueXPB/fffT2xsLJB274UGDRrw4IMPAmm3Jh8zZgzDhw8H4P333+eVV15Jv1ug\nt7c3ffr0ueJ8R48e5YknnmDDhg2EhISwb98+lixZQq1atQgKCqJnz54kJCRclWPKlClUqVKF++67\nj1WrVl0za/369SlWrBgA9erV4+DBg9c8bvr06Vfcyvvhhx+mdu3aVK9enU8++ST98cKFCzN48GCC\ng4NZs2YN0dHRNGrUiNq1axMeHs7hw4eBtGJbp04dgoOD6dixI+fOnbv5F/YGDh8+zOnTp6lXrx7G\nGLp168Z333131XHbt28nLCwMgNKlS3PbbbddtaK/e/dujh49SsOGDYG0/83Kly/P+vVZv31XhTqD\nki7+19cr98+gttYycOBA7r33Xvbu3Ysxhjp16jgdS0REJNukpKSwZMkS2rZtC6Rt96hdu/YVx1Ss\nWJGzZ89y+vRpYmJirnr+z0qXLs3EiRNp2LAhmzZtokyZMvTo0YOvvvqKrVu3kpyczPjx46/4nMOH\nD/PGG2+watUqVq5cyfbt22+afdKkSem36/6zVatWXZFz8uTJREdHExUVxejRozlx4gQA8fHx1K1b\nl82bN1O3bl369+/PrFmziI6OpmfPnrzyyisAdOjQgQ0bNrB582YCAwOZNGnSVa+5bNmyK7ZhXHqr\nX7/+VcfGxsZStmzZ9I/Lli2b/kvN5YKDg5k7dy7Jycn8/PPPREdHc+DAgSuOmTlzJl26dLliOlto\naCj//e9/b/TlyxRdlJhBSRe37+T2m7qkpqby/PPP8+9//5t8+fKxd+/ea/7JSkRExK2usW82O5w/\nf56QkBBiY2MJDAx0681Adu3aRYUKFahSpQoA3bt3Z+zYsel7fgHWrVtH48aNKVWqFABdunRh9+7d\n1z3nsmXLmDRpEitXrrzm83/88QdFihRJ/3j06NHMnj0bgAMHDrBnzx5KlCiBt7c3HTt2TM8ZExOT\n/rVISUkhICAASNvq8uqrrxIXF8fZs2cJDw+/6jWbNGnCpk2bMvx1yYiePXuyY8cOQkNDueuuu6hf\nvz7e3lfeJ2TmzJlX3Wa8dOnS7Ny5M0uzgAp1hiU9+RN8cm+uXqFOSUnh6aefZsqUKRQoUIDZs2fT\nokULp2OJiIhkm0t7qM+dO0d4eDhjx45lwIABVKtWjR9//PGKY/fv30/hwoUpWrQo1atXJzo6muDg\nYIeSw5YtW+jduzcLFiygRIkS1zzGx8eH1NRUvLy8WL58OYsXL2bNmjX4+fnRuHFjLly4AECBAgXS\nC6q1lurVq7NmzZqrztejRw++++47goODmTp1avrFgZdbtmwZgwYNuupxPz8/Vq9efcVjZcqUuWK7\nysGDBylTpsw1/x0ffvhh+sf169dP/8UEYPPmzSQnJ1/1V4MLFy5QsGDBa31pbom2fGRQYkoikLvn\nUL/99ttMmTKFggULMn/+fJVpERHJs/z8/Bg9ejQffPABycnJPP7446xcuZLFixcDaSvZAwYM4B//\n+AcAL774Iu+++2766nFqaioTJky44WtUrVqVX375hb179wIwbdo0GjVqdMUxdevWZcWKFZw4cYKk\npCS+/vrra57rt99+o0OHDkybNu2KYnmt19y/fz8Ap06dolixYvj5+bFz507Wrl173c85duxYeqFO\nSkpi27ZtQNpUjYCAAJKSkpg+ffo1P//SCvWf3/5cpgECAgIoWrQoa9euxVrL559/fsWe70vOnTtH\nfHw8AIsWLcLHx4dq1f43ie3LL7/k0Ucfverzdu/eTY0aNa779cksFeoMqnl7TXb03cGXHb90Oorb\n9OvXj/r16xMZGUnTpk2djiMiIuKoWrVqUbNmTb788ksKFizInDlzePvtt6latSpBQUHUqVMnfYRd\nzZo1GTVqFI8++iiBgYHUqFEjvbheT4ECBZgyZQqdOnUiKCgILy+vq0bTBgQE8Oabb3L//ffToEED\nAgMDr3mut956ixMnTvD8888TEhJCaGjoNY9r3bp1+ipyixYtSE5OJjAwkJdeeol69epd83Py5cvH\nrFmzGDJkCMHBwYSEhKSX4WHDhlG3bl0aNGiQfkHmrRo3bhy9e/emUqVKVKxYMX0/+Ny5c3n99deB\ntAs87733XgIDAxkxYsRVWzv+85//XLNQr1q1yi3beMy1ZvvlZKGhofZacxklcxITExk5ciSDBg0i\nf/78WGvz9K3VRUTEOTt27LhuYZSscfjwYbp168aiRYucjpLtNm7cyMiRI68q35dc6/vPGBNtrb32\nbyeX0Qp1Bv188mdGrhnJnJ1znI6SZS5cuEDHjh0ZOnRo+rxKlWkREZHcKyAggKeffjpH39jFXY4f\nP86wYcPccm5dlJhB245tY/APg2lVuRXt7rl6L4+nOX/+PA8//DA//PADJUqUuOKKYhEREcm9Onfu\n7HQER7hzYosKdQYlpaRNos4NUz7i4+Np06YNy5Yto3Tp0ixevJigoCCnY4mIiIh4JBXqDEpKvVio\nc8Ec6p07d7Ju3Tr+8pe/sHTpUu1XExEREbkFKtQZlBtWqJOSkvD19aV27dpEREQQEBBww9E6IiIi\nInJzuigxgzx9DvUff/xBgwYNGDduHACNGjVSmRYRERHJAirUGdQhsAM7+u7gnbB3nI7isuPHj9O0\naVM2bNjAhx9+yPnz552OJCIikiN5e3sTEhJCjRo1aNOmDXFxcenPbdu2jbCwMKpWrUrlypUZNmwY\nl48fXrBgAaGhoVSrVo1atWoxePDgq86fkJBAs2bNCAkJ4auvvrpujsaNG3OtMcFTp05Nn319uTlz\n5lCzZs30GdTXu/W4tZawsLAcPeUjOjqaoKAgKlWqxIABA7jWiOfExESeeuopgoKCCA4OvuIOjV99\n9RU1a9akevXqDBkyJP3xMf/f3p3HVVXtj/9/LSBFRNNUugSlJoaM56goCdck1Cz7iJZTfi01KtMy\nrdT0ZqX32qSVenH8aalpTp/0OtQ1Cwg1UVNIUAYRVAoIU3EKkfTA/v1xDvvDmEcPk/l+Ph7n8Wjv\nvYb32RvvfbNYe60FC1i+fHmNxCwJtZXudLyTDi074Na04vaX9dlvv/1GSEgICQkJPPDYKJWMAAAg\nAElEQVTAA+zcubNGttwUQggh/gpKth5PSkrirrvuYuHChYB5daywsDCmTp1KWloaiYmJ7N27V//L\nb1JSEuPGjeOLL74gJSWFuLg4PDw8KrR/6NAhABISEhg6dGi1xd2zZ08SExNJSEhg+fLlPP/885WW\n2759OwaDgaZNm1rddlFRUXWFaZWxY8eybNky0tPTSU9PZ8eOHRXKLFu2DIAjR44QGRnJxIkTKS4u\nJi8vj8mTJxMdHU1ycjKnTp0iOjoagPDwcObPn18jMcscaivF/RrH7p93E3BPAA+1fqiuw7HK77//\nTkhICEePHsXb25vo6Gj+9re/1XVYQgghxHUZPn6qRtpNnLTe6rLdunXj8OHDAKxdu5bg4GAeeeQR\nwLw1+YIFCwgJCeHll19m9uzZTJs2Td8t0N7enrFjx5Zp7/Tp0zz99NOcOXMGo9HIpk2byMzMZNKk\nSZhMJrp06cLixYtp2LBhmXorVqzggw8+oFmzZhgMhgrXAZydnfX/vnz5cpX7SqxZs0bfewJgwIAB\nZGVlUVhYyIQJE/Rrzs7OvPjii0RFRbFw4UIaNWrE66+/Tn5+Pi1btmTlypW4urqybNkyli5dytWr\nV/Hw8GD16tU4OTlZfY/Ly83N5dKlS/qujSNGjGDLli36boklUlJSCA0NBcDFxYVmzZoRFxeHUor2\n7dvTqlUrAHr16sWmTZvo2bMnTk5OtGnThgMHDtC1a9ebjrEyMkJtpZiTMUz8biLb0rbVdShWa9Kk\nCQMHDsTPz4+YmBhJpoUQQggrFRUVER0dTVhYGGCe7tG5c+cyZdq1a0d+fj6XLl0iKSmpwvXyXFxc\n+PTTT+nevTsJCQm4ubkxatQoNmzYwJEjRzCZTCxevLhMndzcXKZPn05sbCx79uwhJSWlyvY3b95M\nhw4dePzxx6uc2hAbG1smzuXLlxMfH09cXBwRERHk5eUB5qQ8MDCQxMREAgMDeeWVV9i4cSPx8fGE\nh4czbdo0AJ588kkOHjxIYmIiXl5efPbZZxX6jImJwWg0VvgEBQVVKJuTk4O7u7t+7O7uTk5OToVy\nBoOBbdu2YTKZOHnyJPHx8WRlZeHh4UFaWhqZmZmYTCa2bNlCVlaWXi8gIIAffvihynt4s2SE2kr6\nsnm3wCofmZmZXLx4EYPBwMyZM/nHP/5B48aN6zosIYQQwmo3MpJcna5cuYLRaCQnJwcvL68a3Qwk\nLS2Ntm3b6osEjBw5koULF5bZbO3HH38kJCREH3EdOnQox44dq7S9J554gieeeILdu3fz9ttvExUV\nVaHMuXPnaNKkiX4cERHB5s2bAcjKyiI9PZ0WLVpgb2/PwIED9TiTkpL0e1FUVISrqytgnury1ltv\nceHCBfLz8+nTp0+FPh9++GESEhJu+P78mfDwcFJTUwkICKB169YEBQVhb29P8+bNWbx4MUOHDsXO\nzo6goCCOHz+u13NxceHo0aPVGgtIQm21kmXz6vsqHxkZGYSGhnLlyhV++OEHOnToIMm0EEIIYaWS\nOdQFBQX06dOHhQsXMn78eLy9vdm9e3eZsidOnMDZ2ZmmTZvi4+NDfHw8BoOhjiL/Pw899BAnTpzg\n7NmztGzZssw1BwcHiouLsbOzY+fOnURFRbFv3z6cnJwICQmhsLAQAEdHR+zt7QHzi4w+Pj7s27ev\nQl+jRo1iy5YtGAwGVq5cWeblwBIxMTG89tprFc47OTmxd+/eMufc3NzIzs7Wj7Ozs3Fzq/j+moOD\nA3PnztWPg4KC9F9M+vXrR79+/QBYunSp/j0ACgsLa+RdMpnyYaVbYWOXo0eP0qNHD7KysvD09OSe\ne+6p65CEEEKIW5KTkxMRERF88sknmEwmhg8fzp49e/RR3ytXrjB+/HjeeOMNACZPnsz777+vjx4X\nFxezZMmSP+3D09OTzMxMMjIyAFi9ejU9evQoUyYwMJBdu3aRl5fHtWvX+PLLLyttKyMjQ18N46ef\nfuKPP/6gRYsWlfZ54sQJAC5evEjz5s1xcnLi6NGj7N+/v8o4z5w5oyfU165dIzk5GTC/r+Xq6sq1\na9dYs2ZNpfVLRqjLf8on0wCurq40bdqU/fv3o2kaq1aton///hXKFRQUcPnyZQAiIyNxcHDA29sb\nMM9VBzh//jyLFi0q84LmsWPH8PX1rTROW0hCbaX6vg51UlISISEh/Prrr4SEhLBjx44beoNXCCGE\nEGV17NgRf39/1q1bR6NGjdi6dSvvvvsunp6e+Pn50aVLF30JO39/f+bNm8ewYcPw8vLC19dXT1yr\n4ujoyIoVKxg8eDB+fn7Y2dkxZsyYMmVcXV2ZMWMG3bp1Izg4uMrdjTdt2oSvry9Go5GXX36ZDRs2\nVPpi4uOPP66PIj/66KOYTCa8vLyYOnWq/iJgeQ0aNGDjxo1MmTIFg8GA0WjUk+GZM2cSGBhIcHCw\n/kKmrUqSYA8PD9q1a6e/kLht2zbeeecdwJw0d+rUCS8vL2bNmsXq1av1+hMmTMDb25vg4GCmTp1a\nZt+N2NjYGpnGoypb268+CwgI0Cpbl7GmnS04y5nLZ2jVuBUtnVpev0Ite+KJJ9iyZQu9evVi69at\nNr1hK4QQQtSF1NTUKhNGUT1yc3MZMWIEkZGRdR1KrTt06BBz5swpk3yXVtnPn1IqXtO0gOu1LSPU\nVmrp1BKvVl71MpkG+Pzzz5k8eTJfffWVJNNCCCGEqJSrqysvvPBCvd7YpaacPXuWmTNn1kjbklBb\n6duMb5mzbw5Jp5PqOhTd/v37GTBgAAUFBTRt2pTZs2fj6OhY12EJIYQQoh4bMmTIbTkttHfv3rRp\n06ZG2paE2krrk9cz8buJHMg5UNehALBnzx569+7N1q1bmTdvXl2HI4QQQghx25KE2koly+bVh3Wo\nY2Ji6NOnD/n5+QwbNkx/w1gIIYQQQtQ+SaitVF+WzYuMjKRv374UFBQwcuRIVq9ejYODLCcuhBBC\nCFFXJBOzUn0Zob7rrrto2LAhzzzzDEuWLMHOTn4nEkIIIYSoS5KNWamu16FOS0sDoHPnzsTHx0sy\nLYQQQtQAe3t7jEYjvr6+9OvXjwsXLujXkpOTCQ0NxdPTk/bt2zNz5kxKLz/8zTffEBAQgLe3Nx07\ndmTixIkV2v/jjz/o1asXRqORDRs2VBlHSEgIlS0TvHLlSn3t68ocPHgQBwcHNm7cWOl1TdMIDQ2t\n16t8xMfH4+fnh4eHB+PHj6eyJZ6vXr3Ks88+i5+fHwaDocwOjRs2bMDf3x8fHx+mTJmin1+wYAHL\nly+vkZglI7PS0n5LSXkphZA2IbXe95dffomvry+zZs0CoF27dpJMCyGEEDWgZOvxpKQk7rrrLhYu\nXAiYd0YMCwtj6tSppKWlkZiYyN69e1m0aBFg3mBt3LhxfPHFF6SkpBAXF4eHh0eF9g8dOgRAQkIC\nQ4cOrdbYi4qKmDJlCo888kiVZbZv347BYLihVT6KioqqIzyrjR07lmXLlpGenk56ejo7duyoUGbZ\nsmUAHDlyhMjISCZOnEhxcTF5eXlMnjyZ6OhokpOTOXXqFNHR0QCEh4czf/78GolZpnxY6Z4m93BP\nk9rfynvNmjWMGDGC4uJizp8/X+v9CyGEEHXik4q7/FWLidZvaNetWzcOHz4MwNq1awkODtaTVScn\nJxYsWEBISAgvv/wys2fPZtq0afpugfb29owdO7ZMe6dPn+bpp5/mzJkzGI1GNm3aRGZmJpMmTcJk\nMtGlSxcWL15Mw4YNy9RbsWIFH3zwAc2aNcNgMFS4XmL+/PkMHDiQgwcPVvmd1qxZw+jRo/XjAQMG\nkJWVRWFhIRMmTNCvOTs78+KLLxIVFcXChQtp1KgRr7/+Ovn5+bRs2ZKVK1fi6urKsmXLWLp0KVev\nXsXDw4PVq1fbtB9Gbm4uly5d0ndtHDFiBFu2bNF3SyyRkpJCaGgoAC4uLjRr1oy4uDiUUrRv355W\nrVoB0KtXLzZt2kTPnj1xcnKiTZs2HDhwgK5du950jJWRYU4rrTm8hjn75vDr77/WWp8rVqzgmWee\nobi4mOnTp/PBBx/UWt9CCCHE7ayoqIjo6GjCwsIA83SPzp07lynTrl078vPzuXTpEklJSRWul+fi\n4sKnn35K9+7dSUhIwM3NjVGjRrFhwwaOHDmCyWRi8eLFZerk5uYyffp0YmNj2bNnDykpKZW2nZOT\nw+bNmysk8eXFxsaWiXP58uXEx8cTFxdHREQEeXl5AFy+fJnAwEASExMJDAzklVdeYePGjcTHxxMe\nHs60adMAePLJJzl48CCJiYl4eXnx2WefVegzJiYGo9FY4RMUFFTp93B3d9eP3d3dycnJqVDOYDCw\nbds2TCYTJ0+eJD4+nqysLDw8PEhLSyMzMxOTycSWLVvIysrS6wUEBPDDDz/86T26GTJCbaWIAxEc\nyDlA0L1BtTJSnZSUxHPPPYemabz33nu8+eabNd6nEEIIUW/cwEhydbpy5QpGo5GcnBy8vLzo3bt3\njfWVlpZG27ZteeCBBwAYOXIkCxcu5NVXX9XL/Pjjj4SEhOgjrkOHDuXYsWMV2nr11VeZNWvWdaeE\nnjt3jiZNmujHERERbN68GYCsrCzS09Np0aIF9vb2DBw4UI8zKSlJvxdFRUW4uroC5nzlrbfe4sKF\nC+Tn59OnT58KfT788MMkJCRYfV+sER4eTmpqKgEBAbRu3ZqgoCDs7e1p3rw5ixcvZujQodjZ2REU\nFMTx48f1ei4uLhw9erRaYwFJqK1W26t8+Pr68uGHH2JnZ8ekSZNqpU8hhBDidlcyh7qgoIA+ffqw\ncOFCxo8fj7e3N7t37y5T9sSJEzg7O9O0aVN8fHyIj4/HYDDUSdxxcXE89dRTgHmL7e3bt+Pg4MCA\nAQPKlHNwcKC4uBg7Ozt27txJVFQU+/btw8nJiZCQEAoLCwFwdHTE3t4eML/I6OPjw759+yr0O2rU\nKLZs2YLBYGDlypVlXg4sERMTw2uvvVbhvJOTE3v37i1zzs3NjezsbP04OzsbNze3CnUdHByYO3eu\nfhwUFKT/YtKvXz/69esHwNKlS/XvAVBYWEijRo0qtGcrmfJhpdpahzoiIkL/gX3jjTckmRZCCCHq\ngJOTExEREXzyySeYTCaGDx/Onj17iIqKAswj2ePHj9c3V5s8eTLvv/++PnpcXFzMkiVL/rQPT09P\nMjMzycjIAGD16tX06NGjTJnAwEB27dpFXl4e165d48svv6y0rZMnT5KZmUlmZiaDBg1i0aJFFZLp\nkj5PnDgBwMWLF2nevDlOTk4cPXqU/fv3VxnnmTNn9Pzk2rVrJCcnA/D777/j6urKtWvXWLNmTaX1\nS0aoy3/KJ9MArq6uNG3alP3796NpGqtWraJ///4VyhUUFHD58mXAvEeHg4MD3t7egHmuOsD58+dZ\ntGgRzz//vF7v2LFj+Pr6VhqnLSShtlJtjFC///77TJgwgb59+3Lu3Lka60cIIYQQ19exY0f8/f1Z\nt24djRo1YuvWrbz77rt4enri5+dHly5d9CXs/P39mTdvHsOGDcPLywtfX189ca2Ko6MjK1asYPDg\nwfj5+WFnZ8eYMWPKlHF1dWXGjBl069aN4OBgvLy8bPpOjz/+uD6K/Oijj2IymfDy8mLq1Kn6i4Dl\nNWjQgI0bNzJlyhQMBgNGo1FPhmfOnElgYCDBwcH6C5m2KkmCPTw8aNeunf5C4rZt23jnnXcAc9Lc\nqVMnvLy8mDVrFqtXr9brT5gwAW9vb4KDg5k6dao+cg3mOeQ1MY1HVba2X30WEBCgVbYuY027/9/3\nc/LCSTJeyaDdXe2qtW1N0/jnP//JP//5T5RSfPbZZzz77LPV2ocQQghR36WmptqcMIo/l5uby4gR\nI4iMjKzrUGrdoUOHmDNnTpnku7TKfv6UUvGapgVcr22ZQ22l70d+z5VrV7j3znurtV1N03jzzTf1\n+dKff/45Tz/9dLX2IYQQQggB5hHvF154gUuXLt3QWtR/BWfPnmXmzJk10rYk1FZq06xNjbRrMpmI\nj4/H3t6etWvXMmTIkBrpRwghhBACuG1zjZpcsUUSaitF/BhBUXERY7uMxdHB0eb2iouLKSgowNnZ\nmS1btnDgwAFCQkJsD1QIIYQQQtQqeSnRSm99/xavf/c6V4uu2txWcXExY8aMITQ0lIsXL+pL1Qgh\nhBBCiFuPJNRW0pfNs3GVj6KiIsLDw1m2bBlHjhwhKSmpOsITQgghhBB1RKZ8WKlkZNqWdahNJhMj\nR45k7dq1ODk58dVXXxEcHFxdIQohhBBCiDogI9RWKNaKKdaKAbBX9tcpXbXx48ezdu1anJ2d2bFj\nB6GhodUVohBCCCGqgb29PUajEV9fX/r168eFCxf0a8nJyYSGhuLp6Un79u2ZOXMmpZcf/uabbwgI\nCMDb25uOHTsyceLECu3/8ccf9OrVC6PRyIYNG6qMIyQkhMqWCV65cqW+9nVpO3fu5M4778RoNGI0\nGvnXv/5VabuaphEaGsqlS5f+9D7Upfj4ePz8/PDw8GD8+PFUtsTz1atXefbZZ/Hz88NgMJTZoXHD\nhg34+/vj4+PDlClT9PMLFixg+fLlNRKzJNRWKNnUpYF9A5RSN93Oyy+/jIeHB5GRkXTv3r26whNC\nCCFENSnZejwpKYm77rqLhQsXAuadEcPCwpg6dSppaWkkJiayd+9eFi1aBEBSUhLjxo3jiy++ICUl\nhbi4ODw8PCq0f+jQIQASEhIYOnRotcbevXt3fRfCkg1Qytu+fTsGg+GGlswrKiqqrhCtMnbsWJYt\nW0Z6ejrp6ens2LGjQplly5YBcOTIESIjI5k4cSLFxcXk5eUxefJkoqOjSU5O5tSpU0RHRwMQHh7O\n/PnzayRmSaitcIf9HRwbd4zEMYk3XLewsJDly5ejaRo+Pj6kpqZWuROREEIIIcyUqpnPjejWrRs5\nOTkArF27luDgYB555BHAvDX5ggUL+PDDDwGYPXs206ZN03cLtLe3Z+zYsWXaO336NE8//TQHDx7E\naDRy/PhxoqOj6dixI35+foSHh/PHH39UiGPFihU88MADdO3aldjY2Bu9lWWsWbOmzFbeAwYMoHPn\nzvj4+LB06VL9vLOzMxMnTsRgMLBv3z7i4+Pp0aMHnTt3pk+fPuTm5gLmxLZLly4YDAYGDhxIQUGB\nTfHl5uZy6dIlHnzwQZRSjBgxgi1btlQol5KSov+l38XFhWbNmhEXF8eJEydo3749rVq1AqBXr15s\n2rQJMD+zNm3acODAAZtirIwk1FawU3a0b9GeDi1vbEvNgoICwsLCeO655/jggw8AcHCQaetCCCFE\nfVdUVER0dDRhYWGAebpH586dy5Rp164d+fn5XLp0iaSkpArXy3NxceHTTz/VR5Ld3NwYNWoUGzZs\n4MiRI5hMJhYvXlymTm5uLtOnTyc2NpY9e/aQkpJSZft79+7F39+fxx57jOTk5ErLxMbGlolz+fLl\nxMfHExcXR0REBHl5eQBcvnyZwMBAEhMTCQwM5JVXXmHjxo3Ex8cTHh7OtGnTAHjyySc5ePAgiYmJ\neHl58dlnn1XoMyYmRp+KUvoTFBRUoWxOTg7u7u76sbu7u/5LTWkGg4Ft27ZhMpk4efIk8fHxZGVl\n4eHhQVpaGpmZmZhMJrZs2UJWVpZeLyAggB9++KHKe3izJLuzQv7VfBYdXEQzx2aM7jzaujr5+fTr\n14+dO3dy99136/8ghRBCCHF9lUybrRVXrlzBaDSSk5ODl5dXjW4GkpaWRtu2bXnggQcAGDlyJAsX\nLuTVV1/Vy/z444+EhIToI65Dhw7l2LFjFdrq1KkTv/zyC87Ozmzfvp0BAwaQnp5eody5c+do0qSJ\nfhwREcHmzZsByMrKIj09nRYtWmBvb8/AgQP1OJOSkvR7UVRUhKurK2Ce6vLWW29x4cIF8vPz6dOn\nT4U+H374YRISEm7qHlUlPDyc1NRUAgICaN26NUFBQdjb29O8eXMWL17M0KFDsbOzIygoiOPHj+v1\nXFxcOHr0aLXGApJQW+VC4QWmRE3hnib3WJVQX7p0ib59+xIbG4urqyvff/+9/icgIYQQQtRfJXOo\nCwoK6NOnDwsXLmT8+PF4e3uze/fuMmVPnDiBs7MzTZs2xcfHh/j4eAwGQ53EXXpOdN++fXnppZc4\ne/YsLVu2LFPOwcGB4uJi7Ozs2LlzJ1FRUezbt0/fE6OwsBAAR0dH7O3NCzGUTFvdt29fhX5HjRrF\nli1bMBgMrFy5sszLgSViYmJ47bXXKpx3cnJi7969Zc65ubmRnZ2tH2dnZ+Pm5lahroODA3PnztWP\ng4KC9F9M+vXrR79+/QBYunSp/j3APBW3UaNGFdqzlUz5sELJS4nWrkH9/fffExsby7333svu3bsl\nmRZCCCFuMU5OTkRERPDJJ59gMpkYPnw4e/bsISoqCjCPZI8fP5433ngDgMmTJ/P+++/ro8fFxcUs\nWbLkT/vw9PQkMzOTjIwMAFavXk2PHj3KlAkMDGTXrl3k5eVx7do1vvzyy0rbOnXqlL4axoEDBygu\nLqZFixaV9nnixAkALl68SPPmzXFycuLo0aPs37+/yjjPnDmjJ9TXrl3Tp5T8/vvvuLq6cu3aNdas\nWVNp/ZIR6vKf8sk0gKurK02bNmX//v1omsaqVavKzPkuUVBQwOXLlwGIjIzEwcEBb29vwDxXHeD8\n+fMsWrSI559/Xq937NgxfH19K43TFpJQW0Hf1OU6a1CX/CAPGDCAVatWsWvXrkrf8BVCCCFE/dex\nY0f8/f1Zt24djRo1YuvWrbz77rt4enri5+dHly5d9CXs/P39mTdvHsOGDcPLywtfX189ca2Ko6Mj\nK1asYPDgwfj5+WFnZ8eYMWPKlHF1dWXGjBl069aN4OBgvLy8Km1r48aN+Pr6YjAYGD9+POvXr690\nZbLHH39cH0V+9NFHMZlMeHl5MXXq1CoXTWjQoAEbN25kypQpGAwGjEajngzPnDmTwMBAgoODq20A\nsSQJ9vDwoF27djz22GMAbNu2TV+95PTp03Tq1AkvLy9mzZrF6tWr9foTJkzA29ub4OBgpk6dqo9c\ng3kOeU1M41GVre1XnwUEBGiVrctYk5JPJ+O72Bevll6kvFz5ywCnT58mLCyM9957j549e9ZqfEII\nIcRfQWpqapUJo6geubm5jBgxgsjIyLoOpdYdOnSIOXPmlEm+S6vs508pFa9pWsD12pYRaiuU7JLY\nwL5BpddPnTrFww8/zI8//sgbb7xBcXFxbYYnhBBCCGEVV1dXXnjhhXq9sUtNOXv2LDNnzqyRtuWl\nRCt4t/Lm2Lhj2NtV3CUxJyeH0NBQjh07ho+PD9u3b8fOTn5PEUIIIUT9NGTIkLoOoU7U5IotklBb\noaFDQ9q3aF/h/KlTp+jRowfHjx/HYDAQGRmpL2sjhBBCCCFuD5JQW+HnCz/zv8n/S9vmbRnkPUg/\n36JFC/z8/GjWrBnfffcdd911Vx1GKYQQQggh6oIk1FZIy0vjjag36HV/LwZ5DyIjI4MmTZpw9913\ns2HDBq5cucKdd95Z12EKIYQQQog6IJN9rVB6HerU1FQeeughevXqxdmzZ2nQoIEk00IIIYQQt7Ea\nTaiVUo8qpdKUUhlKqamVXFdKqQjL9cNKqU41Gc/NKlmH+srlK4SEhJCbm0urVq1qZKcdIYQQQtQd\nZ2dnm9sICQmhtpf4FXWrxhJqpZQ9sBB4DPAGhimlvMsVewxob/mMBhbXVDy2KBmh3vvDXk6fPk3v\n3r35+uuvady4cR1HJoQQQggh6lpNjlB3BTI0TTuhadpVYD1Qfu/I/sAqzWw/0Ewp5VqDMd2UP4r+\nAODqlav07duXbdu24eTkVMdRCSGEEH9tSqkKn9GjR9/09Zs1atQoNm7cqB+XHsWeNWsWfn5+GAwG\npk4t+8f44uJiRo0axVtvvXXTfYtbQ02+lOgGZJU6zgYCrSjjBuSWLqSUGo15BJv77ruv2gO9nv6e\n/YkKi2Llbyv5dO6nNGzYsNZjEEIIIUT98s0337B161Z+/PFHnJycOHfunH7NZDIxfPhwfH19mTZt\nWh1GKWrDLbHKh6ZpS4GlYN56vLb7b9KwCT079qTnItlSXAghhKgtmvbn/5dv63VbRUVF8eyzz+p/\ntS69fO6LL77IkCFDJJm+TdTklI8c4N5Sx+6WczdaRgghhBCizjg4OFBcXAyYp3FcvXr1unWCgoKI\niYmhsLCwpsMT9UBNJtQHgfZKqbZKqQbAU8C2cmW2ASMsq308CFzUNC23fENCCCGEEHWlTZs2xMfH\nA7Bt2zauXTMvVtC7d29WrFhBQUEBQJkpH8899xx9+/ZlyJAhmEym2g9a1KoaS6g1TTMB44BvgVTg\nfzVNS1ZKjVFKjbEU2w6cADKAZcBLNRWPEEIIIcT1FBQU4O7urn/mzJnDCy+8wK5duzAYDOzbt09f\n5evRRx8lLCyMgIAAjEYjH3/8cZm2Xn/9dTp27Mgzzzyjj3CLvyZV0/OLqltAQIAmazsKIYQQfz2p\nqal4eXnVdRjiNlXZz59SKl7TtIDr1ZWdEoUQQgghhLCBJNRCCCGEEELYQBJqIYQQQgghbCAJtRBC\nCCGEEDaQhFoIIYQQQggbSEIthBBCCCGEDSShFkIIIYSwUErx9NNP68cmk4lWrVrxP//zPzXa76hR\no2jbti1GoxGDwUB0dLR+7erVq7z66qt4eHjQvn17+vfvT3Z2tn791KlTPPXUU7Rr147OnTvTt29f\njh07VqGPK1eu0KNHD4qKimr0u9hix44deHp64uHhwYcfflhpmfPnz/PEE0/g7+9P165dSUpK0q/N\nnTsXHx8ffH19GTZsmL5T5aRJk/j+++9rLG5JqIUQQgghLBo3bkxSUhJXrlwBIDIyEjc3t1rp+6OP\nPiIhIYF58+YxZswY/fybb77J77//TlpaGunp6QwYMIAnn3wSTdPQNI0nnniCkDnlnlMAABHmSURB\nVJAQjh8/Tnx8PB988AG//fZbhfaXL1/Ok08+ib29vVXxaJpWqxvSFBUV8fLLL/PNN9+QkpLCunXr\nSElJqVDu/fffx2g0cvjwYVatWsWECRMAyMnJISIigri4OJKSkigqKmL9+vUAvPLKK1Um6NVBEmoh\nhBBC1Evqn6rCZ/RXo2/6urX69u3Lf//7XwDWrVvHsGHD9GuXL18mPDycrl270rFjR7Zu3QpAZmYm\n3bt3p1OnTnTq1Im9e/cCsHPnTkJCQhg0aBAdOnRg+PDhXG9TvW7dupGTkwOYd25csWIFc+fO1RPh\nZ599loYNG/L9998TExPDHXfcUSYBNxgMdO/evUK7a9asoX///gDk5+fTs2dPOnXqhJ+fX5nv4enp\nyYgRI/D19SUrK4vvvvuObt260alTJwYPHkx+fj4A//rXv+jSpQu+vr6MHj36ut/reg4cOICHhwf3\n338/DRo04KmnntLjKi0lJYXQ0FAAOnToQGZmpv4LhMlk4sqVK5hMJgoKCrjnnnsAaN26NXl5eZw6\ndcqmGKsiCbUQQgghRClPPfUU69evp7CwkMOHDxMYGKhfe++99wgNDeXAgQPExMQwefJkLl++jIuL\nC5GRkfz0009s2LCB8ePH63UOHTrEvHnzSElJ4cSJE8TGxv5p/zt27GDAgAEAZGRkcN9999G0adMy\nZQICAkhOTiYpKYnOnTtf9ztdvXqVEydO0KZNGwAcHR3ZvHkzP/30EzExMUycOFFPiNPT03nppZdI\nTk6mcePGvPvuu0RFRfHTTz8REBDAnDlzABg3bhwHDx7UR/S//vrrCv2uWbMGo9FY4TNo0KAKZXNy\ncrj33nv1Y3d3d/0Xi9IMBgP/+c9/AHMS/vPPP5OdnY2bmxuTJk3ivvvuw9XVlTvvvJNHHnlEr9ep\nU6fr3vub5VAjrQohhBBC2Eib/ucjnrZer4q/vz+ZmZmsW7eOvn37lrn23XffsW3bNj7++GMACgsL\n+eWXX7jnnnsYN24cCQkJ2Nvbl5nD3LVrV9zd3QEwGo1kZmby97//vUK/kydP5s033yQ7O5t9+/bd\nVOxVOXv2LM2aNdOPNU3jzTffZPfu3djZ2ZGTk6OP8rZu3ZoHH3wQgP3795OSkkJwcDBgTsy7desG\nQExMDLNnz6agoIBz587h4+NDv379yvQ7fPhwhg8fXq3fZerUqUyYMAGj0Yifnx8dO3bE3t6e8+fP\ns3XrVk6ePEmzZs0YPHgwX3zxhT4n3sXFhV9//bVaYykhCbUQQgghRDlhYWFMmjSJnTt3kpeXp5/X\nNI1Nmzbh6elZpvyMGTO4++67SUxMpLi4GEdHR/1aw4YN9f+2t7fHZDJV2udHH33EoEGDmD9/PuHh\n4cTHx9OuXTt++eUXfv/9d5o0aaKXjY+P11+U3Lhx43W/T6NGjfQX9MA8cnzmzBni4+O54447aNOm\njX69cePGZb5v7969WbduXZn2CgsLeemll4iLi+Pee+9lxowZZdov3c9HH31U4byHh0eFuN3c3MjK\nytKPS0ady2vatCkrVqzQ42vbti33338/3377LW3btqVVq1YAPPnkk+zdu1dPqAsLC2nUqNGf36ib\nJFM+hBBCCCHKCQ8PZ/r06fj5+ZU536dPH+bPn69Pjzh06BAAFy9exNXVFTs7O1avXm3TShrjxo2j\nuLiYb7/9lsaNGzNy5Ehef/11vc1Vq1ZRUFBAaGgooaGh/PHHHyxdulSvf/jwYX744YcybTZv3pyi\noiI96b148SIuLi7ccccdxMTE8PPPP1cay4MPPkhsbCwZGRmAeQ75sWPH9HZatmxJfn5+lUn98OHD\nSUhIqPCprHyXLl1IT0/n5MmTXL16lfXr1xMWFlah3IULF7h69SoAn376KQ899BBNmzblvvvuY//+\n/RQUFKBpGtHR0Xh5een1jh07hq+vb+U33UaSUAshhBBClOPu7l5mHnSJt99+m2vXruHv74+Pjw9v\nv/02AC+99BKff/45BoOBo0ePlhnlvVFKKd566y1mz54NwAcffICjoyMPPPAA7du358svv2Tz5s0o\npVBKsXnzZqKiomjXrh0+Pj784x//4G9/+1uFdh955BH27NkDmBPduLg4/Pz8WLVqFR06dKg0llat\nWrFy5UqGDRuGv78/3bp14+jRozRr1owXXngBX19f+vTpQ5cuXW76+5ZwcHBgwYIF9OnTBy8vL4YM\nGYKPjw8AS5YsYcmSJQCkpqbi6+uLp6cn33zzDf/+978BCAwMZNCgQfqLlsXFxYwebX5J9dq1a2Rk\nZBAQEGBznJVRtr6RWdsCAgK0uLi4ug5DCCGEENUsNTW1zIiiqF4//fQTc+fOZfXq1XUdSq0reQFz\n5syZVZap7OdPKRWvadp1s3AZoRZCCCGEuA106tSJhx9+uF5v7FJTTCYTEydOrLH25aVEIYQQQojb\nRHh4eF2HUCcGDx5co+3LCLUQQggh6o1bbSqq+Guw9edOEmohhBBC1AuOjo7k5eVJUi1qlaZp5OXl\nlVnq8EbJlA8hhBBC1Avu7u5kZ2dz5syZug5F3GYcHR31zXduhiTUQgghhKgX7rjjDtq2bVvXYQhx\nw2TKhxBCCCGEEDaQhFoIIYQQQggbSEIthBBCCCGEDW65nRKVUmeAyjecr3ktgbN11LeoHfKMbw/y\nnG8P8pxvD/Kc//rq8hm31jSt1fUK3XIJdV1SSsVZs/2kuHXJM749yHO+Pchzvj3Ic/7ruxWesUz5\nEEIIIYQQwgaSUAshhBBCCGEDSahvzNK6DkDUOHnGtwd5zrcHec63B3nOf331/hnLHGohhBBCCCFs\nICPUQgghhBBC2EASaiGEEEIIIWwgCXU5SqlHlVJpSqkMpdTUSq4rpVSE5fphpVSnuohT2MaK5zzc\n8nyPKKX2KqUMdRGnsM31nnOpcl2UUial1KDajE/YzppnrJQKUUolKKWSlVK7ajtGYTsr/jf7TqXU\nV0qpRMtzfrYu4hQ3Tym1XCl1WimVVMX1ep1/SUJdilLKHlgIPAZ4A8OUUt7lij0GtLd8RgOLazVI\nYTMrn/NJoIemaX7ATG6BFyJEWVY+55Jys4DvajdCYStrnrFSqhmwCAjTNM0HGFzrgQqbWPlv+WUg\nRdM0AxACfKKUalCrgQpbrQQe/ZPr9Tr/koS6rK5AhqZpJzRNuwqsB/qXK9MfWKWZ7QeaKaVcaztQ\nYZPrPmdN0/ZqmnbecrgfcK/lGIXtrPn3DPAKsAk4XZvBiWphzTP+f8B/NE37BUDTNHnOtx5rnrMG\nNFFKKcAZOAeYajdMYQtN03Zjfm5Vqdf5lyTUZbkBWaWOsy3nbrSMqN9u9Bk+B3xToxGJmnDd56yU\ncgOeoJ6NdAirWfNv+QGguVJqp1IqXik1otaiE9XFmue8APACfgWOABM0TSuunfBELanX+ZdDXQcg\nRH2mlHoYc0L997qORdSIecAUTdOKzQNb4i/IAegM9AQaAfuUUvs1TTtWt2GJatYHSABCgXZApFLq\nB03TLtVtWOJ2IQl1WTnAvaWO3S3nbrSMqN+seoZKKX/gU+AxTdPyaik2UX2sec4BwHpLMt0S6KuU\nMmmatqV2QhQ2suYZZwN5mqZdBi4rpXYDBkAS6luHNc/5WeBDzby5RoZS6iTQAThQOyGKWlCv8y+Z\n8lHWQaC9Uqqt5WWGp4Bt5cpsA0ZY3jZ9ELioaVpubQcqbHLd56yUug/4D/CMjGTdsq77nDVNa6tp\nWhtN09oAG4GXJJm+pVjzv9lbgb8rpRyUUk5AIJBay3EK21jznH/B/FcIlFJ3A57AiVqNUtS0ep1/\nyQh1KZqmmZRS44BvAXtguaZpyUqpMZbrS4DtQF8gAyjA/FuxuIVY+ZzfAVoAiyyjlyZN0wLqKmZx\n46x8zuIWZs0z1jQtVSm1AzgMFAOfappW6bJcon6y8t/yTGClUuoIoDBP5TpbZ0GLG6aUWod5hZaW\nSqlsYDpwB9wa+ZdsPS6EEEIIIYQNZMqHEEIIIYQQNpCEWgghhBBCCBtIQi2EEEIIIYQNJKEWQggh\nhBDCBpJQCyGEEEIIYQNJqIUQfzlKqSKlVEKpT5s/KdtGKWXzMmqWra3TlFKJSqlYpZTnTbQxpmRr\nbKXUKKXUPaWufaqU8rY1TktbG5VS91v++z2lVJZSKv86dcKVUkeUUoeVUklKqf7VEUup9vXvp5Qa\nrJRKVUrFKKUClFIR16m7XSnVzPJ5yYq+/JRSK6spdCGEkGXzhBB/PUqpfE3TnK0s2wb4WtM0Xxv7\n3AlM0jQtTik1GvgfTdPCqqM9W+KqpF0f4F1N056wHD8I/AykV3XPlFLuwC6gk6ZpF5VSzkArTdNO\nVmdspfrbYYlxzw3Wa4OVz1IpFQWEa5r2y00FKYQQpcgItRDitmAZif5BKfWT5RNUSRkfpdQBy6j2\nYaVUe8v5p0ud//+UUvbX6W434GGp21MpdcgyurtcKdXQcv5DpVSKpZ+PLedmKKUmKaUGYd4WfY2l\nz0aWEfAAyyj2R6ViHqWUWnADcQ7HvHsgAJqm7bditzEX4Hcg31InvySZtsT1b0ufSUqprpbzjS3f\n94Dl+/e3nLdXSn1sKXtYKfVKqXYClFLvAH8HPlNKfaSUClFKfW0p46yUWlFqpHyg5XymUqol8CHQ\nzhLLR0qpVUqpAaXu1ZpSI+tfYd5xTwghbCYJtRDir6hRqekemy3nTgO9NU3rBAwFKptGMAb4t6Zp\nRswJbbZSystSPthyvghzUvpn+gFHlFKOwEpgqKZpfph3px2rlGoBPAH4aJrmD7xburKmaRuBOGC4\npmlGTdOulLq8yVK3xFBg/Q3EGQzEXyf+8hKB34CTloS2X7nrTpY+XwKWW85NA77XNK0r8DDwkVKq\nMTAaaAMYLd99TemGNE37F//33SeX6+dtzNsN+1nqfl/u+lTguOWeTQY+A0YBKKXuBIKA/1rKxgHd\nb+w2CCFE5WTrcSHEX9EVS4JX2h3AAqVUSbL5QCX19gHTLFMc/qNpWrpSqifQGTiozNvQN8KcnFdm\njVLqCpAJvAJ4Aic1TTtmuf458DKwACjEPAr7NfC1tV9M07QzSqkTlqka6UAHINbSrjVxugJnrO3P\n0meRUupRoAvQE5irlOqsadoMS5F1lnK7lVJNlVLNgEeAMKXUJEsZR+A+oBewRNM0k6XOuRsIpRel\nRpU1TTt/nbh3KaUWKaVaAQOBTSX9Yr4391RdWwghrCcJtRDidvEa5lFWA+a/zhWWL6Bp2lql1I/A\n48B2pdSLgAI+1zTtH1b0Mbz0nGel1F2VFdI0zWSZGtETGASMA0Jv4LusB4YAR4HNmqZpypxFWxPn\nFczJbZUsU0VKRrG3aZr2jmZ+4eYAcEApFQmsAGaUfKVyTWiY79tATdPSyrV9nfCq3SrgacyJ+LOl\nzjtivhdCCGEzmfIhhLhd3AnkappWDDwDVJhfrMwrX5zQNC0C8zxjfyAaGKSUcrGUuUsp1drKPtOA\nNkopD8vxM8Auy0t9d2qath1zom+opO7vQJMq2t0M9AeGYU6uuYE4U7HM766KpmlFlmkTRk3T3lFK\n3aOU6lSqiBHzi4wlhlr6/DvmKRkXgW+BVyyJPkqpjpaykcCLSimHkjj/LJZyIjGPxGOp27zc9cru\n2UrgVcv3Sil1/gHA5tVdhBACJKEWQtw+FgEjlVKJmKdJXK6kzBAgSSmVAPgCqyxJ2FvAd0qpw5iT\nOldrOtQ0rRDzqOiXSqkjQDGwBHPS97WlvT3A65VUXwksKXkpsVy75zEnxq01TTtgOWdtnP8FQkoO\nlFKzlVLZgJNSKlspNaOSOncAHyuljlruzVBgQqnrhUqpQ5bv9pzl3ExLvcNKqWTLMcCnwC+W84nA\n/6ukv6q8CzS3vNCYiHlutk7TtDwg1nL9I8u53zDfqxXl2nqY/5tPLYQQNpFl84QQ4jZiSc5jML+8\nWFQN7e2kBpb3qy5KKSfgCJYl/yznGmJeBvDvpeZUCyHETZMRaiGEuI1YVgyZDrjVdSw1TSnVC/Po\n9PySZNriPmCqJNNCiOoiI9RCCCGEEELYQEaohRBCCCGEsIEk1EIIIYQQQthAEmohhBBCCCFsIAm1\nEEIIIYQQNpCEWgghhBBCCBv8//8lbrsfTdWTAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x158a399f630>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAbQAAAGmCAYAAAAUKnHWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIpKE/wN/q6WSGRz8eQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x158a3996940>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Run logistic Regression with cross-validation and plot ROC curves\n",
    "\n",
    "mean_tpr = 0.0\n",
    "mean_fpr = np.linspace(0, 1, 100) # `linspace` returns evenly spaced numbers over a specified interval.\n",
    "roc_kfolds = []\n",
    "cm_folds = []\n",
    "logit_actual = []\n",
    "logit_prob = []\n",
    "best_mean_cutoff = 0.0\n",
    "best_mean_tpr = 0.0\n",
    "best_mean_fpr = 0.0\n",
    "best_mean_ba = 0.0\n",
    "\n",
    "# For each cross-validation batch, fit model 80:20 split, plot ROC, and get cutoff\n",
    "# TODO - add feature selection per CV\n",
    "# train, test are of type <type 'numpy.ndarray'>\n",
    "i = 0\n",
    "for (train, test) in cv.split(X, y):\n",
    "\n",
    "    test_actual = y[test]\n",
    "    test_X = X[test]\n",
    "    train_X = X[train]\n",
    "    train_y = y[train]\n",
    "    \n",
    "    # train Logit on training dataset\n",
    "    logit = LogisticRegression()\n",
    "    logit.fit(train_X, train_y)\n",
    "       \n",
    "    # test Logit on test dataset\n",
    "    # predict probability returns <type 'numpy.ndarray'> (n_samples, n_classes)\n",
    "    prob = logit.predict_proba(test_X)\n",
    "    \n",
    "    # get predicted values: 1/0; default threshold/cutoff is 0.5\n",
    "    predicted = logit.predict(test_X)\n",
    "    train_predicted = logit.predict(train_X)\n",
    "    \n",
    "    # Compute ROC curve\n",
    "    # fpr = false positive rate (1-specificity)\n",
    "    # tpr = true positive rate (sensitivity)\n",
    "    fpr, tpr, thresholds = roc_curve(test_actual, prob[:, 1]) # returns <type 'numpy.ndarray'> x 3\n",
    "    \n",
    "    # add mean Sensitivity\n",
    "    mean_tpr += interp(mean_fpr, fpr, tpr)\n",
    "    mean_tpr[0] = 0.0\n",
    "  \n",
    "    if output_data_to_excel:\n",
    "        output_excel('logit-fold-%d-test' % (i+1), prob, test_X, test_actual, predicted, 'TestSet%d' % (i+1))\n",
    "        output_excel('logit-fold-%d-train' % (i+1), \n",
    "                     logit.predict_proba(train_X), train_X, train_y, train_predicted, 'TrainSet%d' % (i+1))\n",
    "    # /if...\n",
    "    \n",
    "    # get_cm_row(method, foldnum, cutoff, actual, prob)\n",
    "    # append confusion matrix rows for output to Excel\n",
    "    for cutoff in test_cutoffs:\n",
    "        cm_folds.append( get_cm_row('Logit', i+1, cutoff, test_actual, prob) )\n",
    "    \n",
    "    \n",
    "    logit_actual.extend(test_actual)\n",
    "    logit_prob.extend(prob)\n",
    "    \n",
    "    roc_kfolds.append([tpr, fpr])\n",
    "    \n",
    "    #output_cutoffs(i, tpr, fpr, thresholds, prob)\n",
    "    \n",
    "    best_tpr, best_fpr, best_cutoff, best_ba = get_best_cutoff(actual=test_actual, prob=prob)\n",
    "    print('Fold %d, TPR: %0.5f, FPR: %0.5f, Cutoff: %0.5f, BA: %0.5f' % (i+1, best_tpr, best_fpr, best_cutoff, best_ba))\n",
    "    best_mean_cutoff += best_cutoff\n",
    "    best_mean_tpr += best_tpr\n",
    "    best_mean_fpr += best_fpr\n",
    "    best_mean_ba += best_ba\n",
    "    \n",
    "    i += 1\n",
    "#/for (train, test)...\n",
    "\n",
    "best_mean_cutoff /= float(num_cv)\n",
    "\n",
    "print('Means:  TPR: %0.5f, FPR: %0.5f, Cutoff: %0.5f, BA: %0.5f'%(best_mean_tpr/float(num_cv),\n",
    "                                                    best_mean_fpr/float(num_cv), \n",
    "                                                    best_mean_cutoff, \n",
    "                                                    best_mean_ba/float(num_cv)))\n",
    "\n",
    "show_roc(roc_kfolds, mean_tpr, mean_fpr, 'ROC Curves for Logit with K-fold Cross-validation', 2, 'logit-roc')\n",
    "\n",
    "if output_data_to_excel:\n",
    "    output_cms('logit-cms', cm_folds, 'Logit_Test_Folds') # output to excel\n",
    "  \n",
    "#Determine the best cutoff points and output confusion matrix\n",
    "best_tp = get_confusion_matrix(best_mean_cutoff, logit_actual, logit_prob)\n",
    "#best_accuracy =  get_confusion_matrix(0.5, logit_actual, logit_prob)\n",
    "show_confusion_matrix(best_tp, ['0','1'], (6,6), 12, 'logit-best-tp-cm')\n",
    "#show_confusion_matrix(best_accuracy, ['0','1'], (6,6), 12, 'logit-best-accuracy-cm')\n",
    " "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}