ZDHHC9_connectome_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Populating the interactive namespace from numpy and matplotlib\n"
     ]
    }
   ],
   "source": [
    "%pylab inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import bct\n",
    "import nibabel as nib\n",
    "import numpy as np\n",
    "import os\n",
    "import pandas as pd\n",
    "import re\n",
    "from scipy.stats import ttest_rel\n",
    "from scipy.stats import sem\n",
    "from scipy.stats import shapiro\n",
    "from scipy.stats import wilcoxon\n",
    "import seaborn as sns\n",
    "from statsmodels import robust\n",
    "from statsmodels.formula.api import ols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from matplotlib import rcParams  \n",
    "rcParams['font.family'] = 'serif'  \n",
    "rcParams['font.serif'] = ['Computer Modern Unicode']  \n",
    "rcParams['text.usetex'] = True  \n",
    "rcParams['axes.labelsize'] = 9  \n",
    "rcParams['xtick.labelsize'] = 9  \n",
    "rcParams['ytick.labelsize'] = 9  \n",
    "rcParams['legend.fontsize'] = 9  \n",
    "rcParams['figure.autolayout'] = True\n",
    "mm2inches = 0.039371\n",
    "single_column = 86*mm2inches\n",
    "double_column = 180*mm2inches"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "sns.set_style(\"whitegrid\")\n",
    "sns.set_style({'axes.grid': False,\n",
    "               'font.family': [u'serif'],\n",
    "               'font.sans-serif': [u'Computer Modern Unicode'],})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "outfolder = '/imaging/jb07/ZDHHC9/connectome_revisions/figures/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def plot_adjacency_matrix(network, filename):\n",
    "    import matplotlib.pyplot as plt\n",
    "    from mpl_toolkits.axes_grid1 import make_axes_locatable\n",
    "    rcParams['xtick.labelsize'] = 6 \n",
    "    rcParams['ytick.labelsize'] = 6  \n",
    "\n",
    "    ## Loading labels\n",
    "    labels = pd.read_csv('/home/jb07/joe_python/GitHub/ZDHHC9_connectome/DK_labels.txt', header=None)[0].values\n",
    "\n",
    "    plt.figure(figsize=(double_column, double_column), dpi=300)\n",
    "    im = plt.imshow(network, \n",
    "               cmap='afmhot',\n",
    "               interpolation='none',\n",
    "               vmin=0,\n",
    "               vmax=1)\n",
    "    ax = plt.gca()\n",
    "    ax.set_yticks(np.arange(0,network.shape[0]))\n",
    "    ax.set_yticklabels(labels)\n",
    "    ax.yaxis.tick_right()\n",
    "    ax.set_xticks(np.arange(0,network.shape[0]))\n",
    "    ax.set_xticklabels(labels, rotation=90)\n",
    "    ax.grid('off')\n",
    "    \n",
    "    rcParams['xtick.labelsize'] = 9\n",
    "    rcParams['ytick.labelsize'] = 9  \n",
    "    divider = make_axes_locatable(ax)\n",
    "    #cax = divider.append_axes(\"right\", size=\"5%\", pad=0.05)\n",
    "    #cb = plt.colorbar(im, cax=cax)\n",
    "    #cb.ax.yaxis.set_label_position('right')\n",
    "    plt.tight_layout(pad=0, w_pad=1, h_pad=0)\n",
    "    plt.savefig(filename, dpi=600)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def load_data(folder, model, threshold, weight, subject): \n",
    "    import numpy as np\n",
    "    \n",
    "    if weight == '_FA':\n",
    "        weight_switch = str(0)\n",
    "    elif weight == '_RD':\n",
    "        weight_switch = str(1)\n",
    "    elif weight == '_GFA':\n",
    "        weight_switch = str(2)\n",
    "        \n",
    "    matrix = folder + '_subject_id_' + subject + '/_model_' + model + '/_threshold_' + threshold + '/calc_matrix/mapflow/_calc_matrix' + weight_switch + '/' + subject + weight + '_matrix.txt'\n",
    "    matrix = np.loadtxt(matrix)\n",
    "    \n",
    "    # Exclude label entries \n",
    "    entries_to_exclude = [1, 2, 3, 4, 10, 11, 15, 17, 18, 19, 20, 21, 22, 23, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43]\n",
    "    matrix = np.delete(matrix,entries_to_exclude, axis=0)\n",
    "    matrix = np.delete(matrix,entries_to_exclude, axis=1)\n",
    "    \n",
    "    # Removing NaN values\n",
    "    matrix[np.isnan(matrix)] = 0\n",
    "    \n",
    "    return matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "folder = '/imaging/jb07/ZDHHC9/connectome_revisions/connectome/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "subject_list = sorted([subject.split('_')[-1] for subject in os.listdir(folder) if re.search('subject', subject)])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Comparing nuisance variable"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "RMS displacement"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df = pd.read_csv('/imaging/jb07/ZDHHC9/motion/Displacements_Results.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ZDHHC9\n",
      "mean: 2.1\n",
      "SE: 0.292\n",
      "median: 1.97\n",
      "mad: 0.648\n",
      "control\n",
      "mean: 1.8\n",
      "SE: 0.304\n",
      "median: 1.53\n",
      "mad: 0.314\n"
     ]
    }
   ],
   "source": [
    "for group in ['ZDHHC9', 'control']:\n",
    "    print(group)\n",
    "    print('mean: ' + str(round(np.mean(df[df['group'] == group]['Unnamed: 1']), 2)))\n",
    "    print('SE: ' + str(round(sem(df[df['group'] == group]['Unnamed: 1']), 3)))\n",
    "    print('median: ' + str(round(np.median(df[df['group'] == group]['Unnamed: 1']), 2)))\n",
    "    print('mad: '+ str(round(robust.mad(df[df['group'] == group]['Unnamed: 1']), 3)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ttest_relResult(statistic=-0.58985872368309544, pvalue=0.57680805175906857)"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ttest_rel(df[df['group'] == 'control']['Unnamed: 1'], df[df['group'] == 'ZDHHC9']['Unnamed: 1'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "WilcoxonResult(statistic=10.0, pvalue=0.49896229860376107)"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "wilcoxon(df[df['group'] == 'control']['Unnamed: 1'], df[df['group'] == 'ZDHHC9']['Unnamed: 1'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Number of corrupted voxels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "in_folder = '/imaging/jb07/ZDHHC9/connectome_revisions/connectome/'\n",
    "d = []\n",
    "\n",
    "for subject in subject_list:\n",
    "    filename = in_folder + '_subject_id_' + subject + '/dwi_preproc/eddy/eddy_corrected.eddy_outlier_map'\n",
    "\n",
    "    sums = list()\n",
    "\n",
    "    with open(filename) as f:\n",
    "        next(f)\n",
    "        for line in f:\n",
    "            values = np.asarray(line[:-2].split(' '))\n",
    "            sums.append(np.sum([int(value) for value in values]))\n",
    "\n",
    "    total = np.sum(sums)\n",
    "    group = 'control' if re.search('c', subject) else 'ZDHHC9'\n",
    "    \n",
    "    d.append({'subject': subject,\n",
    "              'outliers': total,\n",
    "              'group': group})\n",
    "df = pd.DataFrame(d)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ZDHHC9\n",
      "mean: 12.71\n",
      "SE: 2.254\n",
      "median: 11.0\n",
      "mad: 5.93\n",
      "control\n",
      "mean: 10.0\n",
      "SE: 3.078\n",
      "median: 8.0\n",
      "mad: 5.93\n"
     ]
    }
   ],
   "source": [
    "for group in ['ZDHHC9', 'control']:\n",
    "    print(group)\n",
    "    print('mean: ' + str(round(np.mean(df[df['group'] == group]['outliers']), 2)))\n",
    "    print('SE: ';  str(round(sem(df[df['group'] == group]['outliers']), 3)))\n",
    "    print('median: ' + str(round(np.median(df[df['group'] == group]['outliers']), 2)))\n",
    "    print('mad: '+ str(round(robust.mad(df[df['group'] == group]['outliers']), 3)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ttest_relResult(statistic=-1.0921223729660698, pvalue=0.31667963027925633)"
      ]
     },
     "execution_count": 131,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ttest_rel(df[df['group'] == 'control']['outliers'], df[df['group'] == 'ZDHHC9']['outliers'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "WilcoxonResult(statistic=8.0, pvalue=0.31049443431723489)"
      ]
     },
     "execution_count": 132,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "wilcoxon(df[df['group'] == 'control']['outliers'], df[df['group'] == 'ZDHHC9']['outliers'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plotting overview matrices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "control_networks = np.rollaxis(np.asarray([load_data(folder, 'CSD', str(10), '_FA', subject) for subject in subject_list if re.search('c', subject)]), 0, 3)\n",
    "ZDHHC9_networks = np.rollaxis(np.asarray([load_data(folder, 'CSD', str(10), '_FA', subject) for subject in subject_list if re.search('z', subject)]), 0, 3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 565,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfcAAAH3CAYAAACmdRYnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f836fb76c90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_adjacency_matrix(np.mean(ZDHHC9_networks, axis=2), outfolder + 'ZDHHC9_adjmat.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 566,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfcAAAH3CAYAAACmdRYnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzsvXt4JEd97/0d9vX6prks5hq6R4AJOZqR1jbe2OoZCI5N\ndrWWSeyD90JICGTtw3NI8oZknSe8BwzH5Fy4OYRjyItf7wIngWDJOOZiLI2AkGBm5mBWthdperML\nNljdi4lvSDOL7Riv5/1DVKm7umamNZoezWi/n+fRI1VNdVVJK21N1e9b31+sXq/XQQghhJANw/PW\newKEEEII6Sxc3AkhhJANBhd3QgghZIPBxZ0QQgjZYPxf6z0BQgghGx/HcfDRj34Ue/fuRaVSwdjY\nGAzDaNjetm3EYjEMDQ01fH1paQkAUKvVsH379kjm3a9wcSeEEBI5pmkinU7DsiwAy4tzs8U9Ho9j\ny5YtDV+vVCrYuXMnBgYG4LouarUa5ufnZf+nOi0X99nZ2W7MgxBCyAbgwgsvbPia4zg4cuQISqUS\n9u/fj0KhAMdxkE6nEY/HASwv2tdccw1qtRps2wYAzM3NYXBwEJZlyTcEY2Nj+PCHP4xqtYrrrrsO\npVIJrutiZGQEc3NzcrxkMimfTyQScF0Xe/bswcDAQMQ/ifUl1M5927Zt8uvZNwZfd5xg3cUX+8uH\nNO8Rkolg3Wtf5y8fuCXY5vTTg3V3/MhffuvWYJtf/u74cF1/+e0PBdu0y/M1dd+6yl9++ulgm69N\nBet+73f95c2bg210P5evf91fvvrqYJszXhz8hzj2vaqv/OoLg38Ijn0iUPf6z/jLt44Fx7sovylQ\n9+hPTvrK998ffO748WDd224Y9JUf+W7wH/DF570kUFc/4Z/7QiX4veh+nurv+mteE2zzvOC3h0cf\n8Ze/9a1gmxe/OFj305/6y888E2zzBz8K1qm8SlP3+Z3Buk8ov3t/37prAMCU8ve+87vhnrsh5i//\nncZ14wHNcw/8vr+8XTNR3XM/eIu//E//FGwz+3CwTt0/fkjT919o6j6ilOv/I/iHq7Ma+fP3/MJX\nfseeYN9PPRWsy2b95cv/Z7CN7ueiziDsf4PqvvtLhw41b28YGBoagmEY+OhHP4rx8XEkEglYliV3\n3s4v/8jEYp/95TeVTqfhOA5KpZI8hv/ABz4AAJicnMTIyAgSiQTq9Tqmpqbw5je/GbVaDZlMRj6/\nuLiITCaDhYUFWb9R4bE8IYSQyHEcB67rolwuo1KpIJ/PY2FhAa7rwrIsTExMIJ/PI5FI4MiRI1hc\nXITrunK3HYvFsLS0hN27dwNYXtDdX+7ORkZGYBgGSqUSUqkUdu7cicXFRcTjcTmueB6Ab9HfqHBx\nJ4QQ0hWq1SpGRka0cfFrrrkGBw4cwMjIiBTRiXaWZWFychKmaco3B9dcc02gj3379vnKhUIBiUQi\nUC8ol8uoVqtIJBKoVqvYsWPHWr/FnoFX4QghhESOaZowTbNhrFvE3nO5HMrlsu+1crmMVCoFy7Jg\nWRbS6XToMZthGAZisRhM00QsFmvatt+ItfKWn52dBW5Yibn/+leDbY69JVj3SiXIt+mGYBvdBYdv\nXOsvv0wTcw/DJ18UrPujR4J1/cLblHJJ00YjM8AXOz+VVXG+pk4TTo8U3d/syePv8JWf9ys3t9XX\nyV+8L9DmQwMfCNT9PxptBTn10C0f918VrDvvjvb62quUvxBiTp3k0KFDTQV173//+3HDDSuLgVjE\nHceBaZoolUoYHh4OiN7EFTqvut77bDKZDAjzHMfB8PAwSqWSjNtbluV7LpfLwbZtZDIZVCoV2ffS\n0hJM00S1Wu1bMR537oQQQrqC2Es6joNarYapqSmkUimkUikYhoHh4WGk02lkMhnU63VMTk7i4MGD\nyOfzvsXXtm1MT0/LZ03TRDabxfbt2319ApDX70ql5S2R9/V6vS4/gBXx3sjICAAgk8lgZGQEw8PD\nSCQSUozXDzDmTgghJHIcx8Hx48dRLpcxMTGBv/mbv/EJ31zXheu6yOfzKBaLSKVSUjwHALfddpvc\ndZumibGxsYBozrIsX59CRFcul5HL5QBAO6bYlefzeTiOg2p15bZQv4rxuLgTQgjpCvV63SeoU4V1\nlUoFpVIpIJYTYrpbbrkF1157Laanp2WcXHWmE32Wy2U4joNdu3ZpX1fL4rM6dr+a4vBYnhBCSOSs\nRVCXy+VgWRZM04RlWRgbG5NH941o5n53KhBq5+56jDuOasRzr/58sO7L24J1Kkc0dRe3KaBT6Wfx\nnI7PhmjzXJt979LU3aaURzRt5kL0HVY8d7ZS/nnI58Kg+/v/yKvCCehURpS+fvDWoHjuuxTPkQbo\nlqIw4rmwfXVbQLdWVEHd3Nwc6vW6dJoTbwTUhdowDDiOg7m5OVSrVYyOjso7795rcvV6XY4h6guF\nQkAkt3v3bhmTF2I6AEgmk1haWurL3Tt37oQQQrpCu4I69XmBELu5rovh4WHZtxe1PpPJYOvWrT6R\nnOM4ATEdANx1110d/gl0D8bcCSGERM4999yDQqGASy65BF/5yle0grrbb78dV1xxBR555JGAoO5T\nn/oUKpUKbr31Vpw4cQLZbNYndrNtG7lcLuBw9573vAc7duzw1YuF3iuSsywrIKYbHx/H4uJid39Q\nHYKLOyGEkMi56KKLMDY2hksvvRSXXnopAL9YzXEcXH311cjlcpibm/Olei2Xy3jlK1+J22+/HQAw\nMzMjDW1UvM+Je+2pVArj4+O+el0WOZ3rXb8SanH3JoF5cTAHB76k8Sz4HSV/wB9p+v0HTZ2rqSPh\n+GGbz1U0decqZV18/W81iU7e+W/tzUGNsX/uV4Nt/vIHwTpNLplQ/M8n23vuOuUH86UvBdtoqqDk\nQ8LdmjYv09S1+/0R0g+EjbkXi0Xs3bti0bN9+3ZtLB3wx8wByB292r5RFjnvaUE/w5g7IYSQrtBJ\nExtdLB1YiZmXSiVYliX7UtubpilfU41vNgI8lieEEBI5nTSxMQwDk5OTvhi7UNCLmLnYrdu2jUQi\nIY1sRHuRhEbNIrdR4OJOCCGkK7RrYuM4DqampnDttdeiUqlgZGREZnrzxth1MXN1DG/7RtniNgI8\nlieEEBI5azGxEc9aloVsNivvpHup1WqB505lQu3cvze78vVvHwq+rstO9E6l/MnwcyJdxtbUNU+U\nuMyH2xTPheF9HRTP6VCzZ30q5HP3POAvD70w2GZQI9bTCehUKJ4jpxphBXUAsLi4iHK5jFqthu3b\ntwcEdNVqNSCQa2Zcs2fPHp+QLplMyjaWZfW9wx137oQQQrrCWkxskskkLMuSXvKqgE48B4QzrllY\nWPBllstms9IURzXC6UcYcyeEEBI5axHUiWdPnDghd/OpVAqu60oBncjdnkqlUCqVAoY2OuMaXWY5\nbwa4foaLOyGEkK7QrqBOPDs3N4dKpYK9e/ciHo/7BHHlchmGYWBxcRGGYaBer+Pyyy+XAjpx9K9m\nkfPSjx7yjeCxPCGEkMiJWlBnGAZisZhMBZtIJHxxc6+xzalAqJ170nP179c0rx/V1K13diKNaR5m\nNXVET5iI048jHP/BCPsGwgvoVD6hVjy6xokQcgqzFkGd6jinUqvVUKlUkEgkAKw41emEeDqxXaM3\nIf0Cd+6EEEK6QicFdboscPV6XX6IxVl1qlOzv2UyGSmk84rt+h3G3AkhhEROpwV1ExMTyOfzgSxw\nYvctPqtOdWNjY4Hsb6qQrlarSeV9v8LFnRBCSFdYq6AOAA4cOIClpSVcfvnlAID9+/fLNmq/3v6b\nZX/bSEI6QajF/XW/sfL1N18VfP3iW4J1623IcSrG11+hqftR12fRH3zpAn/5yvs60w8A/L+avgrt\ndU/IhmG1gjrvgut9NpvNolarIR6PI5lMdmv6fQd37oQQQtaF1QjqHMdBuVxGpVJBOp2WgjnxWjqd\nRjweR7Va9ZVFTvdkMgnXdWFZFubn5wEsi+zE8XsrN7tisRgYB/CnmE0mk1haWuqJkwAK6gghhHSF\ntQjqDMOQV+G8grl0Oo10Oo3t27ejWCz6yuLKXLFYRCaTgWEYKJVKUmS3b9++0G52pmn6xlGFeQBw\n1113decHGQLu3AkhhEROpwR1lUoFS0tLcldtmiYWFhZQLpeRz+dRr9dlOZfLAQDy+TzK5TJc18XO\nnTuRSCQwMzODeDwur8i1crMzDMM3TjKZDAjzxsfHsbi42OWfrB4u7oQQQrpCu4I60zRx8OBBvO99\n78MHPvABWW9ZFgqFAmKxmOxLLevGEWWRE35oaMiXChZYXtB37doly7p+dcK/XiHU4n6LRzD3jp9E\nNRXSjKxSrmjaUDwXnnYFdFH1Q8hGZy2COkEsFsPMzIzPQnbHjh2+Nmq51Zx0GIaBI0eOtN1vL8Cd\nOyGEkHVhNYK6crmM6667Du9973tlytdGznKFQkEK3cTxeSKRQLFY9B2diyN5ndtdvV5v6YLXy1BQ\nRwghpCusRVDnOA7m5+dx4sQJAI2Fb8DyLlt8nclkpAXt+Pg4hoaGAq51Orc7QO+C1y9w504IISRy\n7rnnHhQKBVxyySX4yle+ohXU3X777bjiiivwyCOP+AR1tm0jFothamoKuVwON954I3K5nFb4Jq62\nbd261XfsXiwWMTg4iKGhIblbv+OOO/D85z8fL3zhCwOiOtd1ZUxe1Ktx+V6GizshhJDIueiiizA2\nNoZLL70Ul156KQC/0M1xHFx99dXI5XKYm5vzLaSZTAaZTAbz8/MwTVOmem10n3xychKmacqscLVa\nDePj43AcB0eOHJHPqffSvWN6++6nRV0QanE/4/Sop0FaoRPQEUJIP7OWmLt41mtEY9u2L84OQGZ9\nE9fnSqWST9gnjHC8OI6DZDIpY/qjo6NwXbenTGpawZg7IYSQrtDJmHuxWJRGNF5TGrHLDpv1LZ1O\nw7IsFItFTE9Py/lks1n5nOu6AHrLpKYVPJYnhBASOWsxsRExd8uyYNs2brzxRuTzeRQKBSQSCZmn\nPZFIYHh4eFVZ35aWlqQxDQA5H8dxAs/1kklNK2J18VaqAbOzs6js2ibLf8DL1IQQQjQcOnQIF154\n4XpPgyDkzv0OLuiEEELWgOM4eP/7349rr71W3hsXO3JVsCbc4LxmNTqEcA4Ifxdd7PAbxc3L5TKq\n1SoSiQSq1WrfmdcIGHMnhBASOcKhzrIsebUskUhoXeJEzF1Qq9WkgE5QLpeRSqVgWRYsy0I6nQ49\nj2YYhoFYLAbTNBGLxUL12Ysw5k4IIaQrLC0t4ciRI1hYWMDu3bth27ZUq3sd5AYGBjA3N4dqtYrR\n0VHYth1Q0BeLRezdu1f27VXQC7W7N0WrqB8eHvYp5EVKWPG6SDYDLAsAC4WCnLtpmlJ9r3PG6yW4\ncyeEENIVEokEhoaGpNhNLLqqgxwAn1Jdp6DP5/O+Rdq2bZ/a3TRNZLNZbN++3afKB1YU8iIlrPf1\ner0uPwAEUru2Ut/3Cty5E0IIiRzhUCcU8rfddhtOnDiBhx9+GL/5m7/pc5C74447cPz4cVx22WVY\nWlrCyMgISqWST0E/OTmJBx98EDfffLNUsZ9zzjn4xje+gW3btkm1u2VZckxRF4/HfSlhdap9sSvP\n5/OrUt/3CqHU8g+9fUUt//BPg23++NH2BtdpKmfb6yoU390ZrBvZ6i+f9aEIJ0B6hpP3X+Urbzr/\njrb6OfymYN15t7fVVd/wFk3dxef4y598PNjmaCSzWR2vV8r/EvK5i5XydzswF8HJe64I1G266M6W\nzwWfAr7yVf//qvVkKtj3b3wz9NxWSyu1/Pvf/37ccMMNmJycxMjICOLxOFKplDSWEWY2AwMDsG1b\nitlqtRrm5+d9IjjXdWEYhuzTdV3U63XfcyrCoa6VUG8jwJ07IYSQrtAs5r64uIjvf//7UvDWKuYu\nrGUFhmHAcRzfc8J8plW2N12Gud27d8tjexFvB4KWtb0KY+6EEEK6Qidj7gL18Nn7XNhsb7oMc47j\nBOLtQP+41HHnTgghJHKEQx2wvFjedttt2LJlC1zXRSaT8cXcRdxbxLV1MXdvnyJjm/qcyNfeKtub\niMcD/gxzlmUF4u394lLHxZ0QQkhXEEfiS0tL0sTmZS97GRzH8WVtq1arME3TFxsXmeC8lMtlXHPN\nNVhcXMSBAwdwzTXXaI/L1Wxvk5OTAID9+/f76nWEMcbpRUIt7r88MQEAvGmuc4NHKZ77Pxrx3OhU\nsO7ezf7yr2n66gUhEOks7QroVDa6eE7H5zV1/64I6HR2ImH+jv46Gaz786Vg3ZhSng7RNxBeQKfy\n0jafC8NfX9paPKfjP2p+yEvf8P+vev3H2+o6ErwmNmIRTyQSSKVScvEVn4WgTqAT1HlNbESb1cxF\nh26cfoU7d0IIIV2hWyY2XqGcGNe7oIvjerV9qVSS48zNLe9kHcfxhQL6BQrqCCGEdIWoTWx0QjnR\nFwCZGlb0pbY3TVO+phrf9BvcuRNCCImcKAR1t912m9x9G4YREMqJ63HizYTYrdu2jUQiIY1sRHvT\nNOVxv9fYph8Jtbj/8qpgX3He1mDd7OZg3Wu+7C//J01fjLmTRugic2VN3Ubnix3qRxdfP0/TTo2x\nf+5Xg21+7wedmJF+Dt/StNFMPRSvOldTebj1c/dqHE9P+5q//Opzgm2gMRjqFp0U1DmOg6mpKZll\nbmRkRLbxCuh0gjg1pu5trxPu9SPcuRNCCImcTgvqvP0By0fuqvPcRhLIrRYu7oQQQrpCJwV1wLKr\nXblcRq1Ww/bt27UZ3FSBXDNXuj179viEdMlkUraxLCvgitfLUFBHCCGkK3TaoS6ZTMKyLLljVwV0\n4jkgnCvdwsKCL7NcNpuV81Bd7nod7twJIYRETqezwon+LrnkEjz44IPYu3cvUqkUXNeVbx6efPJJ\nFAoFPPTQQ7j33nvx+te/Hs8884x0pbvzzjtx/vnnyzcZtVoNY2NjUkgnssh5M8D1C6EW97c/FPU0\nOo8uu5vOoEYV0N0cxWTIhuVUFM91mxDaMvyXDorndKh51HT/zV+gqbsvRN//K8w3qOEBTZ37Q395\nXtNGZ8jzcHtTWBUXXXQRxsbGZMx9eHjYlxXOK6i76qqrYNu23JGLlKpe4Zvo79JLL8WZZ54pY+5e\nQZzjODh58iQymQye//znI5PJ4Dd+4zd8CnhvzF9Hv8bruXMnhBDSFTqZFU484425q6Y0KrVaDZVK\nRYYBhJmNLlavi8d7x+51GHMnhBDSFaKOueuywNXrdfkhFmfVzEbN/pbJZOT43nh8P8GdOyGEkMiJ\nKub+zne+E3feeSeWlpbw9NNP4/zzzw9kgRO7b/FZNbMZGxsLZH9TY+0iNNAvcHEnhBASOVHF3F/y\nkpcgm83KNslkEuPj47KdiJmrn9Wv1WP8fo21Czbs4l7X1P1ryDpCSH8R9YHpd0K0CSOe03Gxpk7n\ngKeiSXIZyMYX07TphniuEZ2OuTuOg3K5jEqlgnQ6LWPq4rV0Oo14PI5qteorW5aFcrmMZDIJ13Vh\nWRbm55flh7Ztyx16qzvxxWIxMI74PsX3kUwmsbS01PU3C4y5E0II6QqdjrkbhgHLspDNZn0x9XQ6\njXQ6je3bt6NYLPrKpVIJwHJWuUwmA8MwUCqVZBx+3759oe/Em6bpG0eN3QPAXXfdFenPtBEbdudO\nCCGkd+h04hjR34kTJ1CpVLC0tCR31aZpYmFhAeVyGfl8HvV6XZZzuRwAIJ/Po1wuw3Vd7Ny5E4lE\nAjMzM4jH41JF743di4XeG4M3DMM3TjKZDMTux8fHsbi42M0fNQAgVq/XdSfYktnZWWzbtq1b8yGE\nkFOKd2vqPthmX2GO5aO0LTl06BAuvPDCVT83MzMD0zThOA5M02x67zwMhUIBruvKO+9quVNE1W9H\nqLfg0KFDdSyHsPnBD37woz6k+Xih8rHec+TH+nwcOnSo4VqysLBQf/vb314vlUr1W265pV6v1+uV\nSqVu23ag7fT0dL1QKLRanur1er1+/fXXh2oXhomJCe18SqVS6Pn0Coy5E0IIiRxvFjdx3J1IJKTw\nzIuIrwtqtZo0qFGJxWKYmZnp2Bx19FPCGAFj7oQQQrpCp7PClctlXHfddXjve98rs8I1UrULBbuI\ni3vHEzFxEWvXOd3Vf5mLXq3vVbhzJ4QQ0hU6rZZ3HAfz8/M4ceIEAL2q3XEc7NixQzrMZTKZwHiq\nY53O6Q7QO+D1Kty5E0IIiZxOO9TZto1YLIapqSnkcjnceOONyOVyWlX70NAQRkZGfMfu3vG8jnWO\n4+C5557Dm9/85oDT3eTkpE9Fv1bhX6S0CspTUMcPfvCDH/wI89FMUFev1+vve9/76vX6inDNcZx6\nrVarFwqFum3b8vPCwkJ9enpaPletVuulUqlhn63EbhMTE77n1fG8VCqVhmP1E9y5E0II6QqddqhT\nY+4iJu51mbNtW2Z6m5ychGmaqNfr+NrXvobBwUHU63WUSiVf1jcxJy+O4yCZTMqY/ujoKFzXXTcH\nulYw5k4IIaQrRB1zLxaL0mXO6zjn7RNonfUtnU7DsiwUi0VMT08jlUohlUohm8365gWsnwNdK7hz\nJ4QQEjlf/vKXcfjwYRQKBTzvec8LONR97nOfkwY2YRzqDh48iEcffVSeANx4443I5/MoFApIJBIy\nT3sikcDw8PCqsr4tLS1J1zlg+VQhHo/DcZzAc+vlQNcKOtQRQgjpCM0c6kTGt0wmg0qlgh07dsjX\nZmZm8Itf/AIPPvgg3vCGN4QSqtm2Heo4vFarYX5+vueOzaMm1M59i+frn0U0EUIIIRufer0u4+mW\nZcEwDGzfvh22bePhhx/G4uKiz7BGl5ltdHQUAFCtVnHw4EHs2LFDxsTFjlp8Xa1WZbx+bm6uZ2Pk\nnYYxd0IIIV1BHBSLuLXjOL4YujfWLeLjusxsIt5dLBZhmqZ0kJuaWkmEK2Lh3ni9t36jw5g7IYSQ\nyBFxdOEcJ+LW3nvr3lj3xMRE08xstVoNIyMjcoF3XReXX365vJMuYuEipWsqlfLVb3RCxdx/yxNz\n57E8IYQQHe1mhSOdJ9TO/VtXrXx9/h1RTYX0G4OaujDpJNt9rl2e/Mtg3Vkfav3cI38UrHvRJ/3l\nQ28Mttn21XDzIuRUolwuS7V6LBbD9u3b19zn5OQkRkZGAgK8crmMWq3WkTGEE15YN7rVthd0cs4A\nY+6EEEK6gGEYiMViyGazaHFgHJq1ZnFrlm1OEI/HG47TifaCRnMOM0cdjLkTQgjpGjq1PACtu5xX\nKQ9AZnYTrDaLm6q49zrfzc3NAVhxonMcB+l0GoZhwLZtJBIJJJNJuK4Ly7JQKpVkG2HGI+Zj2zYA\nBMbyZqITpjniezIMQzvnUqmknaPQKjSCO3dCCCGRc++99+K+++7Dt7/9bTz33HM+tfyBAwd8jnJq\n9jWx2AsFvWj7yCOP4MEHHwy0v/fee3H48GFtP1u3bsVZZ52FYrEolfT/+I//iH/4h3+QTnRHjhzB\nwMAA4vE47rvvPjlmJpOBYRj4q7/6K7iui2w2i+3bt/uc8YSFrepmp7rweb8n27Zx9913AwhmnjNN\nU6r9p6am5BxbwZ07IYSQyInFYnjiiScwNDSExx57DI7j+NTy5XIZMzMziMfjcgcslPKGYfgc5sRu\n/amnnsIDDzyAn/70p772sVgMjz/+eCCLm1Dcn3nmmTh58iRM00SpVMLZZ5+NfD4vneie97zn4cEH\nH0Q+n8ehQ4fw9NNPyzG///3v47WvfS1OP/102LYNy7KQz+fl3AFobwU8//nPx5133olcLoehoSGp\n3q9Wq3j22Wfx+OOP46mnngrM2TRNlMtlpFIpmVFPjNOMUIv7vz/d7j8nWW9eoZR/1MG+dSK4MaU8\nHfK5KDlbI557aJ+/PHgw2EYVz+l42a8E687VtHugdVc9wZVK+UvrMgvSLie/8Ru+8qY3fHudZhLk\nggsuwBlnnIGhoSEcO3ZMayQjFvZMJoPFxUUkk0kMDQ2hUCjANM3Asfzg4CByuRwWFxd97QcGBnD6\n6acDgEwQs7CwANd1sWfPHiwsLMBxHMzNzSGTycg3E+KK3Nlnn43f+q3fAgC8+tWvlrtux3GwdetW\nlEolDA8Po16vy0XdMAx5ZF+tVmVKWbHb//CHP4wXvehF2Lx5M2699VaZtGbXrl3Svc8rpvMK8vbt\nU/7DCgF37oQQQrqC18QmnU7DcRyUSiVUq1UsLS1h//79sCxLWsZ6j9Nt28bIyAgWFxdRKpWwf/9+\n2efw8LCvvSCbzQJYNrJZXFyUCWLEzlftRy0LpqensXfvXtRqNRiGgeHhYZimiUQiAcuycOONN8rn\nyuWyb1zHcWAYBq644grpk1+v1wPz7ZTIUMDFnRBCSOS0MrFp51heiNTK5XLg+F03FrCsPo/H43Bd\nN9CPWha78bGxMXkc7k1q4zgOLMuSz7mui507d2Jubi4wrvd4vVQqaefbSUKZ2Jy8fsXE5uKpJo1J\nzxHlsbyOMMfy3SamqftxiGP5MDz8jmDda28O1vFYnnSD9T6Wb2ZiE8U9d9IYquUJIYREThT33Elj\nQh3L38ndet8S9U59vcd7mVI+rmmj+2+k3Z26yks1u/R+hjv1/qaXBHSNCHPP3TRN1Ot1KaCrVqu+\nO+PiCFvcSQf8d+BPlcxvzeDOnRBCSFdolBVOveculOLiXnsmk/HdGZ+enpb3vYV4TbQFTp3Mb82g\noI4QQkjkzMzM4OjRozh69CiOHz+Oyy67LCCoKxQKSCQScuEXMXoAPoGaV+Am3ORE20qlgnQ6LZ8L\n6/XeaW/39YaLOyGEkMjZvn27dFurVCqBRVQ9Qhc2sgJx3U23+HrbCqtX0V88HseWLVtazs8wDBw5\nciT099PrhFrcf+/NK18vfCH4+v/u1Gw0ZDV1lQjH6xd0kaQwqQXO09QdXuNcvBztYF9h0MXYCSG9\nS6uYu4ijixi7aKN6rDfyeR8aGpL9j46Oolarab3eRdze60e/keDOnRBCSFdoZmLjOI40ihG+614T\nGGEao5r8f/RMAAAgAElEQVTNlMtln6GM4ziyf9d15ZsHr7GM67pa45uNpOLn4k4IISRywpjYiDi6\n4zhNTWBU05hisSh38V6TGfGsblzVKCcKI5n1JJSJTfzGFROb/8Fj+Z6gV4/lCSGnLms1sTlw4ADS\n6XRDUZtXHNeq7akOr8IRQgiJnFYmNiJm3myxFglawrQ91Ql1LL9588rXYXaHnYS79GXUnfr/0bRx\nrw3W7bvFXy5onnubpu6zIeake38+G+K5TvLYn/jLL7ipyxPQcIfmeOShBX/5XT8L19cFSvm+tmZE\nSO/QSFDnui7m5uZkG50pTTKZlL7vc3Nz8iqcTlgnxhkcHJRH8nv27JH51jc63LkTQgjpCo1MbA4e\nPAjTNDE8PCzFdDpTGpHNTbQtlUoypaoQ1mWzWWzfvh2ZTAZbt27F8PAwEomEzAh3qkBBHSGEkMgJ\nI6hzXRf5fD5gYDM+Pi4Fb94+WgnrhBLemxFOvHnY6HBxJ4QQ0hUMw0C1WoVpmoF4eTKZRC6Xw8DA\nQODeuVc8J8xpxGfbtpFKpTAwMIAdO3bIZyzLastbPqyjnaBXne14LE8IISRyWgnq4vG4jLF7aSWe\na/Rcu+j6q9Vq0mRHRWgGeo1QO/czzlj5ekTz+klNXbv5q7udf7xfUAV0jkY8Z9wSrAvDizV16r/z\nnKbNvZq6xT/zl1Mfa29OYWlXQPd+pXzDmmeywvnnB+t+++pNvvK7rtf91QShgI6shV/X1H2v67Pw\n00hQ18xJTojngBUHO7HoG4YhRXbNxHXia9EPsOJOVygUtP15x/O643mf7VW4cyeEENIVmgnqxKKb\nzWZ9GeAMw8Dw8LAvE5xXOCeeayauM01T7v6LxaIU64mYvK4/73jpdFrG6oeHh33Per+vXoIxd0II\nIZETVlCnvi6e9WaCEw52Xke6ZuI6AFhYWEC5XEY+n8fExITPnU7Xn3c84W2fSqVQKpX6wtmOizsh\nhJCuEEZQNzQ0FBDCVSqVhqI173NevOI6x3EQi8UCYjzxjPfZRkK8ffv2NX3GSy+I7HgsTwghJHLW\nU1CXy+W6egWuF0R2oXbuM19f+fr2qGbyS041AZ0qIAT0PwP1T+GaNsVzOj6sqfto3F/eXwu20UWZ\nboxYQNeKj58TrPvTx4N17/vgWb7yDe9+smNz+KLmj+QvToQT0BHSSdZbPKdjvQR1qiBOTS3rdbEr\nFos+AZ5lWSiXyw37F3PrJZEdd+6EEEK6wnoK6jKZDOr1OqampmQ/3rG8LnZeAV6pVAIAFIvFhv33\nosiOMXdCCCGRs56COm+62J07dzZNLVur1WAYhhTg5XI5AEA+n2/Y/8TEBPL5fE+J7Li4E0II6Qrt\nCuosy0KhUNCmim0kgPMK6uLxuBTECYQTnVoPQI7l7Vd8XS6XUSqVfP2Lo/hmIrtuE2pxf9ObVr7+\ng49ENZVTk7AaAzXR2HQH56A7PNLF2MPwV2uaydp5lya+/okXBus2hYix6/40VY+qk/dcEWhz7kV3\nBuuUcrsmT4T0K4Zh4MiRI8hkMqhUgvk+4/E4tmzZ4qur1WqYn5+HZVnIZDIyJt8JdOMJvAu3ivg+\nVLxz7QW4cyeEENI1WgnqhDBuaWkJ1WpVCuEAyOdGR0dh23ZADLd7924ZI9eljBXYto1cLqcV8Ik5\nFItFjI+Py0xyoj/DMFCv16UdrRDRlUolrWhPhB26DQV1hBBCukIYQR2wkuLV6wznfc51XWQymYAY\nznEcZLNZ2VYwNTUFYFl0Z1kW9u3bJ/O6qwI+YTs7Pj6OoaEhbQpaIOhU10i0t15w504IISRyZmZm\ncPToURw9ehTHjx/HZZdd5hPUHTx4EMeOHcO2bdtkilevM5wQqalCO1UMZ1kWvvOd72DTpk1yob/8\n8suxuLiIXC6HQqEgF3CxmM/OzuL48ePSja5YLGJwcBD1eh1PPvkkXNf1paB1XReTk5PI5XKo1Wr4\n9Kc/jV27dmlFe+tFrN5Crz87O4v/MLFNlgcYc18X1Jj74XWZRe8T09TdpIm5//GjrfvqZMx9k1Jm\nzJ1sRA4dOoQLL7xQ+5rjOL6YuxrXFq+rgrl2YtmN+mrWfsuWLRgYGMDMzAxM04TjONi8eTO2bduG\ner3ecA6rHatbhNq5n/mShKdUDby+S/PMvFIOyg/IalAX87dp2uiyu32ozfHU/v/sTZo2GrOWTmUx\ne4em7mZNXZg3PWEWch36BI9+Ri8OLuQzbwm2e/Xn/WWdn5ajqSNko9Eo5i5eCxvL9hrZ6ExkRF9e\n45n5+eWVybZtedyuxt/FMXs8HkcymUSxWAQAZoUjhBBCdDSLuQvCxrJbmcgAfmObUqnki7nr4u/p\ndNqXPU4c3zMrHCGEEKIhbMxdGNIIQ5hWBjQ6E5lSqYRjx47hyiuv9BnPJBIJzMzMIB6P47TTTsNN\nN92EbduWw86zs7N49tln8fDDD8vscWKMfD4vY/8f+9jHMDo6umbDmqiTy3BxJ4QQEjleG9hKpRJY\n1MTrol4YwtRqNWQymUDWNy+qiczAwABM09Sa4QjE9ThRZxgGZmdn8cwzz/jais/79u2D4zi4+uqr\ntUY6OprpBRrdl+8UXNwJIYR0jfWMuXuTvRiG4TPTSSaTOPPMM3HuuefCtm05nvd5kcSmXq/7xlbH\nEPfmn3nmGTz99NPrEqsPtbgfuycoovNym6buVUqZAqLO8llN3VZNnZrd7TqN89wfhOj/OY147ou/\nH6x7+a+d5itveu8vNL23Riee0znNtSuWa5dBpbxZE2q78vPBOhXd7/7FmrrvhpjTRud6Td16OyGS\n9vDG3NPpNBzHQalU8h1PDw8PY35+3hdzTyQSMub+5je/Wfq/JxIJWJYld8i6mPv+/fsBQCZ7Ec/Y\nto10Og3LsvDRj34Ue/fuBbAcXxfjqc8PDQ1heHjY18+NN97oa5PNZhGLxWAYBu6//35frF6dY5Sx\negrqCCGERI5I8CI+RDx79+7d2Ldvn6yfnJxEKpXyxdxt25Zxc2/MXSjcJyYmfM+oyWQqlQpyuZzv\nGWDZdU6Nr3vHU5/3zl/0oxtD9PXss8/Ctm0sLCwEvi/RT1SEuud+9kdW7rkPTYTrWN25/7umDXfu\nnUW3c39rmzv3/632o2nz/gh37jp6ceduaNr8TFMXxhGbO3c93Ln3D83uuZPuwpg7IYSQyCmXy9Ll\nTZfdTbxumiZisVhTAV3Y8aJUo/c6oRb3V28bWClMnAi8PhKoAebanRFpm+9r6nQ7dZX9GoMaNcb+\nd5rn7L8P1h1C53bqKlHu0t+vqbtBU/dQizIAXKKpC7Nzb3eXHsZJr1/QnV6cirv01ynlu9dlFp2l\nVVY48bpIHNOp8U5VuHMnhBDSNZqp5YHl62Ni8V9NtjZBrzvHdQsu7oQQQrpCK7V8vV73ZWsDltXr\numxtAwMDsG0bw8PDWFpa6roavdfh4k4IISRyWjnUCSc5kZtdfBZZ37zZ2ubm5vDKV74Ss7OzOHr0\nKF760pdKj/hEIoFbb70Vjz76KM4666yG8xHZ4YTBjG3boWL9Bw4ckBa1ayVKXQAXd0IIIZHTyqHO\nsqyAK5zX2W1mZgbj4+NwHAdPPPEEzj77bLz1rW9FKpWSu31g2aUuTKY20zRlqlhgOVHMli1bmrrK\nCeOa1SzG6+VSF2pxd+2giM6LTjz3t0qKsg/9W7CNToy00VGvUrX7M9BdNplts6+3awxqJpRrbnrx\nXHToMtxpfoU6lgpXJ557m6ZuRilPacSI52l+nuq/l+7fqt2rcP0qntPBq3/LqAK6X9O0Oaqpc6/1\nl41bOjShDtIo5l4oFGSMXeza9+zZI53dxP12cccdAKrVKiqVijyyX6tDnbi77nXEW4tDndAKrEdG\nOZrYEEII6QrNssJls1lZn0gkkMlksLCwgOnpaaRSKSSTSZnJTWSJE7azYbLCCYc6kUkOgHSo02WA\nEw513ucNw5AOdaIftY33+3Bdd90yyvFYnhBCSOQIRzY1lq7G3EU9sHykPTY2hsXFRSQSCekGJ5Tz\nhmFIVztvJjmRqU20F+52xWJR7rABvUOdNwOc+vzc3Jyco+M4sCyraZulpSWMjIzI/kqlknaeUcDF\nnRBCSFcwDEMa1TSLuXtRhW5iIRdlx3EwMjKCxcVFDA8Py3pdXzt27JBfZzIZuaP2zgFYzgCnQzdH\nb1kI5NTnRVnMzSvaa5RRbq3wWJ4QQkjkGIaBWCyGbDa7qqPoeDzuu8fu/dpbFv17qdVqMstcN1Dv\n7K8noXbur//M6jt+p079RAICujFNmx9p6lTxzL2aNot/Fqy78WP+ss7tS9fXK/6D3yN+NkLnOR1h\nf33aFdB9Q7EAe4PGAuyzIfr5HY14TkcYsSPFZKQROvGcjl4U0KmsVlBXq9XkMTwAeeVNLNqiLPoW\n9SItbKN0sbqxisWiTyhnWVZTUZ4u1WyvwJ07IYSQrtCOoE4soKqYTi0LhAAvlUr5xHFTU1OyvtFY\npmnKq26lUgkAUCwWG4ryuimQWy2MuRNCCImcsCY2qqAuHo/7xHFiF+8tO46Dc845B2eddRbGxsZw\n+PDymZ7YaadSqUC6WN1YhmFgYWEB5XJZngbk8/mGoryJiQnk8/mmArn1SmATKuXr1dtWUr7+OOoZ\nnWK0eywf07T5WZvH8rq+nv3vSurW93T3WD5qwhzLh+Hlmroft9cVIX1Ps5SvwlhGmNh4xW1rReRp\nF306joMtW7b4zG3CUCgU4LpuQ0Fdu3NrZagTBaF27v/g+Tf4z4Xg6+3GPQkw3eZzundkqY9pKtvs\na6Mt5irtLuYqP+5MN4ScMjSLuauxbGAlnq173RvrFn2Ojo7KOD0A3zPVajXQhzeeLo7oDx48KJX0\n3vHDxOnVea8XjLkTQgjpCs1i7ul0Wsa7i8UiRkaWk4mLeHarWLfXOEYssmqfatlrQFMqlWQcf9++\nfYExMpkMtm7d2jROr5u39/vuJoy5E0IIiZxWJja2bct4dz6fx8TEhM/wRcTJm8W6vTF013Vhmqav\nz3q97isD8MXTE4kEZmZmEI/HfYloxBhiwW4Up9fNO0qjmmaEirk/856VmDuP5QkhhOhoFnMvl8uo\nVqsAgFgsFohBRxHvVvsMO8ZqRHBRzLsTcOdOCCEkckQGNCGoU+mkwK5Rn7lcDvPz8y2fW022tijm\n3QlCLe4Xv3aT/Ppw4WRkkzkV6FRWOBKeyzR1r1LKN7fZ98e3BOse/lmw7oNt9k+iZY+mbqLrszi1\niEJQ5zW0EUI41axGzdCmE8Kttwiuk1BQRwghpCtEJagrFotSCNfIrMaboa1er/eNGU27cHEnhBAS\nOUJQJz6EQG737t3Yt29fQOw2MTGBVCrlE9SJ623qa/l8HoVCAeVyWWtWIz6bpil37836E3PsZxhz\nJ4QQ0hWaZYVzHAexWExmSROfdZnUxNG5t64RatY1NUObrj9hM9sMNVtdr8GdOyGEkMhplRVux44d\nPaM4D5PdTc1W12uE2rk/9pPmIrqzNXU/b2c2pwAU0IVDZ8vbrpvfN0PWhWFaSb38UU02yTCHeW/U\n1H21nQmRNdHP4rnf19T9fddnsXo6JahLJpPyvnmrrG3AisBOfc22bZimiXq9jqWlJZimCcMwfBnm\ndHMwDEP63Iu5RJWbvR24cyeEENIVOiWoA4CpqSkA8LnMlcvlQD/eLHHePrwiPHG0LsYUqEI77/Pe\nNw533XVXVD+ytmHMnRBCSOS0ygq3Goc613Vx+eWXY3FxUWaHEy5zCwsLsp/NmzcjnU5LgV2pVMIj\njzwCy7KkCC+RSMg3GtVqFQ888ADOOussnDhxArOzs7jqqqsazkE47o2Pj2NxcXGdf8J+uLgTQgiJ\nnO3bt8M0TWlisxZB3a5duxqO4+1HPS43DANbtmxBrVYD4DegEaI6b/a6ZDKJeDyunYNlWT11DK8S\nanG///7mr+vi65/7VX/5+h8E2+hSm5Lu8w5NXRhTlxdr6v6tzTmoMfZ24+tRc+YZ/vIPNW1+HKIf\nxtf7i5cp5ePrMgs/1fWeQJs0irmL14S3u2C1MXdvP4Zh+NzwksmkvP4mzGzm5uZ84whqtRoqlUrP\nxtRbwZg7IYSQrtCNmLvaLp1Ow7IsFItFJBIJWSec7BoZ2HhzwfdiTL0VXNwJIYRETidNbETMHYCM\nuVcqFeRyOV87YDmDm+hTNbMRby50Bjbej/Hx8XX5ma0FxtwJIYR0hU6Z2DSLuYs4ujCZUduKflUz\nGzWeLj6Xy2U4jtM0Q1wvGtpw504IISRyojKxqdVq8j66l06ZzPSroU2onXs7FrvvVgR0/e3Su7HR\niec+8UJ/+Y8fDbZpVzyno9sCukuU8j+Hfe5b/vLs7wTbvObLq58P6W16QUCnovs1+68xpdyDuU86\nbWJTrVa1md68JjNqP2qmuEQigWKxiPHxcSwsLABA3xvacOdOCCGkK0QhqGuU6U28QVAFc5lMBlu3\nbpWZ4sTiPz4+jqGhISm063dDGy7uhBBCIicqQZ3YNauZ3sQ4OsGc4zhyfGA5Zez8/DwAIJVKwXVd\nzM/P++ar9qMbq5fEd7F6i6S1s7OzOHzVNlne5zRp7EGNUvBYvr8Icyzfz1yilP855HPKqSeP5UlP\nsd7H8ocOHcKFF16ofa1cLqNaXb6dH4vFAgK1QqEA13VDxd3DCNjK5TJqtVpTIRyw7JxnmiZKpRJe\n8pKX9MzivFaolieEEBI5hmFI5zevsYzA6xbXing8ji1btoQarxVi8R8YGGjZZz8RanF/2w2D8ut9\nfxgurxl36u2jumEBQUHPY38SbPOCm9ob7zxNXZiduu65w+1Noev8c5vPqRuhCzW79Mc1/zYp69d8\n5U2/ezTUeCNKeS7UU2Q9uFJT96Uuz6EXBXQqYQV14hoaEMzoporlbNuWsXKv05xXCKdroxPFAfAJ\n7sTRveM40ge/H2DMnRBCSFcIK6grlUoAlkVzuoxuQsBWKpVkZjed01yrNro+s9msnJ/rur6scv0E\nj+UJIYREzmqywuVyOQDAzp07ZUY3XUa2XC4nM7vZtu3LIre4uNiyTaMsb8ViEYcPH8YrXvEKjI2N\nyTn0E1zcCSGERM5qs8IBaHhX3LIsDA8PY35+XsbqVUc73fPimf3792vberPHHTlypKUYr5cJtbg/\n8t1wcXbSGcIYZoSNr3/8HH/5Tx8Ptmk3Tt4v8XUdJ/7CXx74SHv9XK6pO0f3b3NTuBi7CmPs/YMu\nvv4HSvl/a9rs1dTduvbp9Cxhs8KJmLs349v8/LzWvEZkdvPG1UWcXjWs0RneiF25mhmun2HMnRBC\nSFdYbcy9WCz6Mr4BKyYxXvMab1zdG6dX4+eNDG8aZYbrZ3gsTwghJHK8MW3XdRGLxVrG3PP5PMrl\nMlzXxc6dOzE1NYXx8XEsLi7CMAyUSiWkUilfXF0XpxdjjYyMyGe88faJiQnk8/lAvL6f4eJOCCGk\nK6wmKxwQjJmrGd5aGd4IAZ03tq5mgwPguzonxl2NN3xYw5xuwmN5QgghkdPprHCNssF5WW22tnaz\nuzXKHBdmjlERauf+ovNf4in9NKKpnBo8+Zf+8lkfina8dykCOtVWFui+texlmrpvdncKGA0hoFNF\nd0BQePeV/y/4R535T8HjvPbkdKTfUQV0Z2vadFI891tK+esd7LtTrNbEZjWCumZmN95+SqVSQEwn\ndvk6I5t2M8eVSiXtHLthhsOdOyGEkK4QlaAujNmNtx9vm2KxKMV4AwMDAIJGNmvJHKebYzdgzJ0Q\nQkjkRCmoC2N24+2nWCzKNvl8Xl6/E/NU52gYBorFIgYHBzE0NCQzx4lEOOK5yclJn0mOaZool8uB\nOXaDUFnhLvjeFbK86T/zWH4tdPtYXs1idhOP5QEAw0p5XtMmzLH8SR7Lk1WgO5b/eQf7X+9j+W5l\nhes0rQRxInOc4zgwTbNpNrpegTt3QgghkdPJrHBhqdVqmJ+fb6l8b5VBTiz6/bCoC0It7s/7+cp7\ny5M/eUfg9Y+ce3Og7oNP+cs6F6ZPaeq+dIG/fOV9ISao4eT9VwXqNp1/R3uddZCzlZ36Q5o3qYMH\n2+v7/Zq6933wLF9507ufbK/zkHzjdf7yG+4OtnmV5rkwO/dLNHX/rJR1u+2LNeI5dac+puk7rnnu\nZ3/qL//7D4O79Pn3BJ973gv9qSQ3vetnmhHJRuOTyknZzZpTsvu+o+63gU2vbW/PPT3vvyq2afi2\ntvqJkrCCumq1GhC9iZ0z0NyNTvTrFbRtdEc6FQrqCCGEdIWwgrpisYh0Oi1Fb0KMlkwmQ7nR6QRt\nG92RToXH8oQQQiJnNVnh8vk8FhYWpOhNiNESiURoNzoAPkHbRnekU+HiTgghJHLayQonCOsWp7aL\nx+NNHemaZZDrd0Kp5c+5aZssv0KX1ohseHTx/Bu6PovOocbYpzVtbn5psO6Wh/3lQ52aEDllialX\nWgB4LigBALZ9tTtzWSvN1PKO4+DIkSMYGhrCxMSELzYOtBdzN00T9XpdmslUq1Wf+YzYjYuYPLBi\nPANAGuJstIUdYMydEEJIl+h0zF3svIWZTCaT8ZnPTE9P+2Ly3rbAiiHORoTH8oQQQiJnNSY2YWPu\n4g2CaiYj+h4bG5MxeWFL620rDHE2IlzcCSGERI4305ruGLzdmLt6lW01bTcyXNwJIYRETiuHuk6b\n2HQqDatt24jFYqENbFbbXtDptLGhFvczzujIWATAI3/kL7/ok90dX/eeNkxCQp147m2aus+uZjJd\nIoyN7Kc04rl3PBysU/l1Td33Qs2KkGV0kuYwArprNXW3rHk20dHKoS4snXKdC9tXPB7Hli1bGr6+\n1vaCRvMN+/2qcOdOCCGka4R1qAuT6rWV65wuDWuhUJCK+meeeQZPP/30hkwbS7U8IYSQyLn33ntx\n33334dvf/jaee+45rVpeqOPVFK2APtVrM9e5e++9F4cPHw7UZzIZbN26FWeeeSb+7d/+Tfb1iU98\nAk888YRMG/voo4/CNE384Ac/wP333x+Y06FDh/DUU091JG2sbdu4++5lr+5OpY3l4k4IISRyYrEY\nnnjiCcTjcTz22GPSMW737t3Yt28fFhYW5I5XpGitVCrI5XJwXRfj4+MAgFQqhampKbiuC8dx5DMT\nExNIpVLSdS4Wi+Hxxx/H5OSkr148V61WEY/HYds2XNfFpZdeilqtJt3unnrqKd+8vXMaHBzEaaed\nhrPOWs7dIdLGirSyYgzv5yeffBKTk5OYn5+X34frupifn8ezzz6Lxx9/XKaN9c7XNE05x9WkjeWx\nPCGEkMi54IILcMYZZ2BoaAjHjh3TmsfU63WZWz2ZTKJWq6Fer8sjawD4/ve/j8HBQZimiYWFBflM\nNpvF4uIikskkhoaGMDAwgNNPPx0A8LWvfQ2Dg4Pyet2ePXuwsLAAx3GQSCTkdbjzzjtPXrkbHByU\n8xb9iDk+8cQT2LRpEwzDwIEDBwKhhN27d6NUKuFnP/sZEokERkdH8YUvfAEvetGLsHnzZtx6660Y\nHBxEvV7Hrl27pMGPV0znFeS1kwY31OL+y9BAQ7Zq6q4711/+7gPBNl3WkvUE3RbQqYQRzwHAoFJ+\nSNNmZo1z8VJQtCKna0Scl3yrvb512d2eULK7bf94e33ftDNYNzoVrHuLUv58e8ORdeJipfzddZmF\nn5vLQXX5px5/wlfedEVvyTu9JjbpdBqO46BUKqFWqyGXyyGRSMCyLNx4443Yv38/gGUVeTwex113\n3YV8Po90Oo1EIiGP5cUzQnjmTQYDQJrXpNNpLC4uIpPJYGFhQe5+S6WSHKterwfKgunpaezduxe1\nWg2GYWB4eNg3vjpn77iu6+KNb3wjFhcXMTIygnq9HphrpxPXcOdOCCEkclqZ2BQKBSlWE0fg4ih6\nampKGs6IVK5hk8Go4wGQx++u68qxhOBNLRuGAdd1fYY44nuJxWJwHEc757m5Od+4IyMjct6lUkm+\nmYkqcU0ob/mT1694y1+s2ZVw577xCLNz/xVN3U/aHC/KnbvGuhuPh9i5h/GN/z/cuZ8S9OLO/TnN\nzr2+zjv3Zt7yre65d4qw98zFcfpa59Hp++mdgoI6QgghkWMYBmKxGLLZbKS50+PxuEwM0whxhS3s\nglyr1aRqX0Vc5es1Qh3Lv8bzRuzkV98XeP2Hv/+BQN2Xv+wvZ16o6fjRYNWXLvCXr7wvxAQ1HH5T\nsO6824N1qi1A2Jh0u8y+0V9+qWb7+ys3Rzf+yXuuCNRddNGdgTp146zbuU9pfsa/rfyMdc99XOPv\n8BHlB6856MG9vxOse43yexb87oAv3xL843v6mP8ITLdLv0hT97+Unbql2aXP5IN1B4uazhTU0xJA\n//OLkpOf8P+hbvpjzR9pD3LPeLDuoq91rn91p57VtAljy5LR1M2VgjvwTblCy77cW4JtjGwyxCzW\nl0b33Mvlsu8uOwBftjfvnXLLsuRi2+699Lm5OdTr9YZ360ulUtP78N779L0Id+6EEEK6QrOscMDK\nXXY1g5s3Y1ypVAIA351v7313r1DOey/d28YriGt0tz6bzWLr1q3IZrOIxWLyXrp6D937ffUSFNQR\nQgiJnFaCuttuu02K5tQMbslkUmaMy+VyAOC78+0V1olxVIFbsVgMtGkliBML+LPPPgvbtpFMJuUc\nohTDdQIu7oQQQrqCYRioVqswTTMQ7961a5ev7D3yLhQKgYxxjbzWk8kkcrlcQFDnTUzjzVDXqD/R\nxrZtmKYp+xM7eG//3md7RWDHY3lCCCGRsxZB3Y4dO0IbuYQR1K0GXX/9ILALtXPftGnl6w8PBMVz\n3306+MwdSvnlT4ab0KfaFNCp6MRzOqIW0KlcqGR7elWb/dxxfrDuvPOCdbd/URnv4qB4rqDe0wJw\npXJX6xLNHML+jFUe/lmwzlHKP9I8d+GXg3WP/Ym//IKbgm2G/1PwyOzwf2k4Pck9mjpVQFfQiOd2\naMwfbJAAACAASURBVMRzz/61X+g08edLgTZhxHOf+9Vg3e/9IMSDIekXAZ1KJ8VzYWg3p5mtqXtD\nCPGcjsFPB+teDf/v1d9oxKvv0vz9dZNeENSpQjqBN7mMrj/veGoyl14T2HHnTgghpCv0iqDONE3Z\nX7FYlON4k7Wo/U1PT8vxRPIaoHcFdlzcCSGERI4QsamJVUTiGG9yGLETn5+fh+u6qNfrLQV1IoGM\nKqgTyWe8bQDI/vL5fCDpjK4/r0Od2M2LNydqYppeENhRUEcIIaQr9IqgznEcX3/is3hGFcvpxhIa\ngEbPCNZLYMedOyGEkMjpJUFdLpeTx+pRs14Cu1A790cfWfn63RrxnI7XKeW7Q05oOmS7jcIP23zO\n0WTq++03bQrU/cXPT/rK5wZaAK8OYXSuEwLpHKRnW3eFD4Zoo0P338GW3K/5K246Gmjzr5oHN71I\nVRqFUxn9rlL+TAjxHABs0gjowvB7Svk1r9E06qCgjnSfNlMmaP8e1N/+9RbP6egVQZ0qiBP9CFc6\ncR9/z549KBaLAQFeM/c7YP0Fdty5E0II6Qq9JKjLZDKo1+u+frLZrJxbIpGQ6WFVAV6zvntFYMeY\nOyGEkMiZmZnB0aNHcfToURw/fhyXXXbZujnUmaaJcrmMVCql7cd1Xfz85z/HS1/6Upm/3SvAA5ZP\nGv7lX/4FmUwmVOrZbsPFnRBCSOR4d8yVSiUyQV0jAZxXUBePxxvG8C3LguM42LJlCwYGBhqODyzH\n048cOeL7XsS8GwnsukWoxf2f2ggKhY2xk/b4v5/Q1F1/MlipoMu2ppOeqKYyOsLE18OiJMvDV7Wt\ngmx6czDGrqI7ENv0p+0FI8PkYf9Cm/F1HZ9TyxMd65qQdaGdmLt4bmZmBvF4PBCDB/wGNLq4uWo2\n4zWqqdfrSCQSKBaLGB8fR71el1fhRF/bt2+Xcf71jqeHgTF3QgghXaETMfdisYh0Ot0wxp3JZLB1\n69ZA3Fxt542TC/e58fFxDA0NyTcM3hi867o9E08PA4/lCSGERE6nssLl83ksLCw0jXGLxXdpafkU\nTXjBe7O5eePkAFAsFjE4OIihoaGGc7Vtuy8ywgFArN7ibcfs7Cx+8HvbZPnN/xr5nEiXafdYvpO0\neyxPCOkdDh06hAsv1F2SXT52F4t1LBbTmrocOHBA7tC9FAoFuK4b+q67GC+MeUy5XMbdd9+NN77x\njXAcx5cBrp/hzp0QQkjkCPGZENSpiPi3bjH2iuG81Go1zM/PawVrYrww8zr//PMxNDS0IRZ1QajF\n/SUv9hRC7txfppSPh3uMrAO6XfrFSvm7Ido0aheGft2pD2rqwmR3I+RUpZGgznVdzM3NBTK2qWYx\nQuA2OjoK27abZmer1+s+4ZzOgEa085riiFBBP0NBHSGEkK7QTFBnmiaGh4d9QjfVLEYVuLXKzgYE\nxXhqnwJv1reNAI/lCSGERE4rQZ0wnInFYnAcB5ZlBYxo5ubmfM+OjIygVCohlUqhVCppxW6maWoN\naNQ+vVnfNgKhBHW1/SuCut/8l3Ad81i+v+n2sXy/wmN5QlZYq6AuChqJ8cII7tYro1sn4LE8IYSQ\nyGmVFU5cV+s0jTLKhcnWtl4Z3TpBqGP5n/509R1zpx4tF2jq7utg/2F24L2wSx9RynNdHj/sLv0t\nSjmM0x0hG5FGgrpSqbSmTG0iU9xqsrWpgruFhQV5j1542Per2I47d0IIIV2hmaBOiOPaydQmMsUV\ni8VVZWsD/IK7r3/96xgaGpKJZQT9KLajoI4QQkjktMoKV6vVpDhOl6lNiOhEW2+mNrHLzufzDTPB\nNcrW5hXcCf/4bojtoo7nc3EnhBASOa2ywgHLvvA6IxmdSU2YTHFe85tG2dq8/QwPD2N+fl5mqGuU\nYa4ThDXZaZdQi/szz0Q2PmmTTsbX+5lux9jD8PlXB+suUEQSn2d2N3KKEibmrsbTAX3WN7HT9maT\nE5nkRB2wErtXzXGWlpbkUf/k5KSM17cb9wewqph/lDDmTgghpCuEjbk3i5V7s76JpC0im9yOHTuw\nsLAg23lj994+AWBqagqAP56+lri/ziCnVcw/yoxyPJYnhBASOa1MbLwxd2/GtjBZ30Q2OQAYHByU\nwjdd7N6yLLiui8svvxyLi4u+eLphGG3H/XUGOWFi/lERysSmsmvFxOYPfhTZXAjZEIQ5ls/wWJ5s\nQNZiYmPbNmKx2KqSt6y3yUw72eq6BY/lCSGERE4rE5t4PA7T1CWgbt7netLIIKcXCHUsz9060elF\nO+8ltTF4yzFNpa6OkFOQRoK6Wq0G27YBwJf5zXVdn2AOgBTNGYbhM5npllitH+DOnRBCSFdoJqgT\nanI18xuwIpjLZrPyeS+NDGpOZSioI4QQEjleQd03v/lNvOAFL8ArXvEKX1a4b37zm3JB94rXhGBO\nXGcTO/jvfe97eOyxx+C6ri8j3Gri9huVUIK6bdu2NWtCTgF4LE8IaUUzQZ0Xx3Fg27bPZAZYFtUt\nLS2FNo4J275Wq2F+fj4yQ5pehDt3QgghkeNN0DI0NKSNuwPLcXMVnYHN6OgoAKBareLgwYPYsWOH\nNJwRO35vrN5rTiPqN/Jiz5g7CUVZ80GI4HXKR7e5TPNBgF/VfKwXqomMLu4OLCdysSwLxWJRxtab\nGdgUi0UprgNWzGmAlVi915zGW7+R4c6dEEJI5KgmMjojG3HMLkxhJiYmfLF01cCmVqtJu1rTNH3m\nNK7ryli915zGW7+RYcydELJm1N363V0eX7dT/2aX59CL6HbqP4hwvLAxdxI9PJYnhBDSVQqFAmZm\nZjrS1+TkpDa7Wrlc7tgYtm2vKoPbatsLOjnnUMfy53q+fqAjwxJCNhLd3qmr9PMu/VxNXaf+n41y\nl74WMpmMNKxZK41c7cKmVA2jpI/H49iyZUvoOa22vaDRnNtR+zPmTgghJHLCqOWF05xt28hkMgD8\nSnlgxZ1OYNs2crmc1qVO516nKu5t226Y5lXM1zAM2LaNRCLRMqWrmI/ObU/0USwWMT4+LjPYtXLc\n86bD9c5RaBV08FieEEJI5LRSyx84cAClUgmWZWHfvn3aVK/iOQCyrVDBN3Kpa5UytlGaV+98xcLd\nLKVrsViUcx8YGAAQdNtLJBIAlk15hoaGAt9TozmbpqmdYzO4cyeEEBI5XrX8HXfcgePHj+Oyyy7z\nqeVFzDkej8sdsFDKG4bhc6cTu/UvfOELeNWrXoWzzz7b175UKuHYsWMB9zqhuHccBz//+c/xh3/4\nhyiVSrjnnnuQyWRkmteJiQmcfvrpAJad8ET/IqXro48+ivvvvx/nnHMOLMtCPp+XcwegTW9rGAaK\nxSIGBwcxNDQk1fv33Xcfnn76aezYsQOu62JyctI3Z9M0US6XA6lomxFKLb/Ho5ZnzJ0QQjpHlDH3\nbrNWh7p2EIueajnrOA6OHDnSMB2s4zjYsmWL3GWr/YjywMBAoJ9Wfeuo1Wr47Gc/ize84Q1wHAem\nacqx2umvFaF27p/fufL16FTjdoQQQhrzzHuCdZv/e+vn5ncF6956W7Duvef5y//xcLh5dYP1irl7\nY+xiJ71nzx6ZhU4clav9iLLoR9SLhXlubg71et0Xc1fj8Wq8/ayzzoJpmlhcXMTi4iIOHDgQWRY7\nxtwJIYREznrF3L1x70QigUwmg4WFBV8c3duPWhZMT0/LWLdhGBgeHvZ9T2o8Xo23i9g+oM9i1+IQ\nfdUw5k4IISRyooq5i913uVwOxNZ1cW9g+Yg8Ho/LeLy3H7VsGAZc18XY2JiMdXsd9hzHgWVZvnj8\nzp07MTc35xt3ZGREuuSVSiXtXDtJqJj7s9evxNx5LE8IIe2x0Y/l1yPmTvRw504IISRyVhtzN00T\n9Xpdxtir1aovhi12uuJOOuCPx58Kmd+aEWpx/1vu1iPjSk3dl7o+C9IuJz/xwkDdpj9+dB1m0lv8\niabupq7PIjr+VFP38RDPna7Zpf/47cG6l3/GXx7W7NJjmv7VNCCXaXbu6+XmZ5omEokELMuC4zgY\nGRlBOp2G4zgolUqoVqtYWlrC/v37ZRvbtjEyMoLFxUUZr06n03BdF9PT09i7dy9qtRoymQwqlYps\nCyxnfsvn8+v03a4/3LkTQgiJnDBZ4crlMgqFAhKJhBTbiRg7AN8z3hi4cJPztj0VMr81g4s7IYSQ\nyPHG16vVKkzTDNzrVo/Q1WtizY7YRdtCoYBYLCbvkNu27Ss3olwuo1ardfSu+XrCq3CEEEK6inrN\nLAy1Wk3G5FfTdzweb5hcxouI+28UQu3c/y7qWZzCML7e3zC+rqcX4usjmrq5DvUdJr6uQ7ecqfH1\ntfSVPtheX91gNYI6IZJrluClUQIXb9+jo6PSrAaAVpDnNb3ZSPBYnhBCSOSEEdQ5juMTyQErArp0\nOo1EIiGNZvbv3w8AMoGLrm/XdeUbh2w26+tveHgY8/PzkRrJrCdc3AkhhEROWEGdEMmJ9joTGNUw\nplgsNuwb0CdxUU1yojCSWU9CmdhsU+9XEEJIjxPlsTzRE9bERojedOK1AwcOIJ1ONxS2eQVyrdqe\nylBQRwghpKs0EtSJuHmzxToej0srWi7sjdmwx/Jv0dR9voP9q5cqjnSwb0L6jf92mr/83l+szzy8\ncJfeW4QR1Lmui7m55X85rzsdsOI4l0wmpe/73NycNhubGKder2szwok0rxuZDbu4E0II6R1aCepq\ntRp27NiB4eHhpo5zIpubaZoYHh5uKa6LxWIyzarICCfEehsZLu6EEEIiJ6ygznVd5PP5ho5z4lmx\nE28lrhNqeG9GuFNhcd+wgjoeyxPSPXrxWJ50n7UK6pq5yTUTz4V1oQvLavvrRXc7CuoIIYR0lUaC\nukZucq3Ec2Fd6MKi66+ZQ14vutuFOpa/6+KVry//blRT6SwXnxOse/rxYN3tbfb/WJvPEdINXqeU\n7+5g3+ouHQCuV3bq9p5gm/GJYN2PlPJnBoNtvvxQsE51dvzVYBP8QFMXJR88M1j37qe6PAmFw28K\n1n1O85/eR0L0pfmngeafpiFhBHXN3OSEeA5YcbAT/RmGIUV2zcR14mvRD7DiUOedn7c/73hehzzv\ns70Id+6EEEIixzRNZLNZufseGRmRmd8mJydx8OBBuehms1n5unCZGx4exvT0NFKpFFKplK8/8Vyp\nVEImk4FhGFJcJ9qYpil3/8ViESMjy04IIiav6887XjqdlrH64eFh37NA77nbcXEnhBASOSI/OwDc\ncccd+Kd/+icpstu9ezf27duHYrGI2dlZWS8+u66L22+/HS94wQt8DnaiP/GcENdVKhXkcjlfGwBY\nWFhAuVxGPp/HxMQEUqmUdKjzthVjetPKit28eDPifVa091IulzEzM9Oln24QquUJIYREjjfl61VX\nXQXbtgMx9Hw+jyuuuAIDAwO+9K6O4+Dqq69uGHP3PtdozGPHjmFhYUEeo4v+hWjOK56zLEubXnbf\nvn1Nn/FiGAaOHFk/qXWoxb1f4uxe/lYTX++c3AJgLjDSy7QbY/8TpazL7qbG1wGgosTYM5r4ehhe\n/vJg3e9o2n1JCfb+UNPmiT8N1j2/3XRuCrr4861txtcPvTFYt+2rrZ8b19Sp/y9tSQbb3PCXwbqP\nfKj1eLr4+mq06esdcz/zzDMxODiIEydOyJi5mn3Oa3RTLBZ9MXrLslAulxv2L+bWK3F4HssTQgiJ\nnF6IuQuV/tTUlOzHO1YikZBGN94YfalUArB8/N+o/16Lw3NxJ4QQEjm6mLYac1frvTF3NQau669Z\nzN00TZkTfufOnYHsc+pYwEqMPpfLAVg+/m/UvxrD18Xhu8mGNbH5D5o63bH816OeCCF9RJhj+Zim\nrlPH8t96fbDuxz8O1r1dOSPWzenxLh/Lb9HU3R+iryiP5b/4h8E2L3hhsO6sEMfyOtRj+b8PaWLT\nDxQKBbiuK+Ps/QYFdYQQQrpKOw51zZ5rl2bjecV4Kr3oSKeyYRf3fw1ZRwhZQbdTV9Ed9V3R5k5d\n5dJ/Cdbpdraquks3p3d1aJeuQycu09X9V+VI4b9qJvrFELt0HV8L0eadnw7W3dnecFra1YJnMhnf\nFTVBPB7Hli3+M5BarYb5+XlYltXwuXbRjReGRkp471zXmw27uBNCCOkdVqOWF6r3paUlVKtV6QwH\nQD43OjoK27YDSvfdu3dLAZwuZazAtm3kcjmtOl/MoVgsYnx8HAsLC77+DMNAvV6XdrRCIV8qlQIu\ndo7jyMQ43YSCOkIIIZGzGrU8sJziFYDPGc77nOu6yGQyAaW74zjIZrOyrWBqagrAsqLesizs27dP\n3otX1fnCdnZ8fBxDQ0NyfG9/QNCprpEifz3gzp0QQkjkeFO+3nHHHTh+/Dguu+wyX8rXgwcP4tix\nY9i2bZtM8WoYBkqlElKplEz5KlLFAgiUa7UaLMvCd77zHWzatEku9JdffjkWFxeRy+VQKBTkAi4W\n89nZWRw/fly60RWLRQwODqJer+PJJ5+E67q+FLSu62JychK5XA61Wg2f/vSnsWvXLpTLZaRSKZ8i\nfz3YsGp5Qkj3eKVSfrDNfnSqd13M/cIQceq3aur+brUTWiNhYu7v1jz3wQ6Nf4WmrpMxd5WwKV/F\nFTJVtOY4Do4cORIQqrUTy27UV7P2W7ZswcDAAGZmZmCaJhzHwebNm7Ft2zbU6/WGc1jtWN0g1M7d\n+wuq++UkhJzatLuYq+j+ewmzkOvo9kKuQ/3/UnPTr2MLOQA88kf+cu6THex8jYSJuQNYVSzb61Kn\nc4gTfXld5ebn5wEsx9zFcbsafxfH7PF4HMlkEsViEQCYFY4QQgjxEibmLggby27lEAf4XetKpZIv\n5q6Lv6fTaV/2OHF8329Z4RhzJ4QQEjmribkLtznh9maapjaW7e1zYmIC+XxePlMqlXDs2DFceeWV\nKJfL0pkukUhgZmYG8Xgcp512Gm666SYZep6dncWzzz6Lhx9+WGaPE2Pk83kZ+//Yxz6G0dFROZbQ\nAqyWKO/Lh4q53/nrKzF3HssTQkh76I7lNVf72ybMsbwuyU6n6LeYu23bWFpakv06joPZ2Vn85Cc/\nwQUXXNCR+HqzuUcZqw+1c//cBlnQ/1qTIenPl4J1JMjFmro+TBZIIuIzL/eXBzX+rDqDmjA89PZg\n3WLVX/66Jm32dbX2xtPxLqX8N5o2Yf5GOrmQ6zzc3qgs5v98bbDNZbcE69T4bBSJSnsl5u7N5GYY\nBiqVihw7mUzizDPPxLnnngvbtuV43udFhrp6ve4bWx1D3Jt/5pln8PTTT3c9Vs+YOyGEkMjphZi7\nmskNWI6lW5alja/X6/XA8yJDnbcftU02m8XWrVuRzWYRi8XWJVbPxZ0QQkjkrCYr3OTkpC/DWrOM\nbq2ysjXLFAcsu86p8XXveOrz3sxxoh/dGKKvZ599FrZtY2FhIfB9RZk5LlTMfa/nnnuU8Zqo4bF8\n+/BYnjSDx/Ld/xvRHcsnlPLtXT6WDxtzJ9FDtTwhhJCu0ii7W7lcRrVahWmaTbPDhaUfsrdFRajF\nvZ936150u/TzlPLhrsyk97leKf/VusyC9Atf+bG//DuaNmHylv9Ys0t/+WeCdW9Ryjf8frDNn70+\nHajbdM2CZmatUXfqmg0xNBviSNEdOj6llD+hmdTREH3r8tXrst61S6PsbiLbmkgcs1YaZW87FeDO\nnRBCSOSEVcsDy9fHhIp9NdnaBL3sHNctKKgjhBASOV51+be//W0899xzWrV8vV4Pla1tbm4Omzdv\nhuM4OPvsswH41ejlchl33313UzV6oVCQ1+6A5XvvYXb6Bw4cwMyMRuixSsrlckf60cGdOyGEkMjx\nusnF43E89thjcBzH51AnnOREbnbxWWR982ZrM00Tp512GqrVKk4//XT8/+y9fZRc1XUn+mtkIQlU\nVS2zJHnGdbuNEShdH4AEz6G6XzwYYlpYENtrliSW/ZyxA05mvUmIJ8SZN/FL/DFO1vMD2Y7JSzxj\nKYknTqCblwiTCKkbLCBQVXIsBLa6q42QjOkqHBt9VXe1ERKEen/020fn7rvPrVOfXS2d31osVLfv\nPXffc/fd55y9f2fv1157TeWIj0ajOHPmDE6cOIHTp3mw4hw8z/PVeI9EIli1alXNxDOUntYGYW21\nM2zgBncHBwcHh46gWq1ifHwcGzZswOHDh33Z4QBgdnYWp06dQqlUQjweR6FQwLZt23Do0CFEIhHs\n3bsXAwMD+OlPf4ojR47gQx/6EDZt2oRisahW9dlsFvfccw+KxSLOnDmDaDSqtrXZJLHRt+st5iQ2\ndoQ6jayy7q/bJUr7sUk4trfjUiwOOAKdQz3YxX//WDhJOsYwOxs89hHhvG/x35Jd+uvGyHM2aDd5\n7tPs973COT+xaKdRh28ryXMEz/MQjUaRyWRQLBaRTqfR19eHYrGIXC6HSqWi0tH29fWhXC4jkUhg\nenoae/fuxR133IHBwUFMTEzgnnvuQT6fR7VaVZnpUqkUJiYmxCQ299xzDwCoJDYkR6FQUEls7rvv\nPtxxxx3q/tFoVCWx0a8fGBhQSWyone3bt/vOoeQ18Xgczz//vC+JDZexXUls3MrdwcHBwaHt0N3y\nlLyF3O3cLU/HgXm39qZNm1AulxGNRlXCGCLXxeNxlfhGLzZDxVzofEqAk81mlRyAnMRGLxLDrz90\n6JCSsVgsIpPJiOfQAE5JbGKxGPL5vChjO2CVxKb3T84lsXErdwcHh3bhB/8+eOxLfxc89jftF2VB\nYbNyt8FG4djBBtuygW0SG9M+dxMKhYJv3/vo6CjS6XTgd7lcxuzsbKAgTSfRLXvrHVvewcHBwaGj\noLzttohEIr6tbvq/9d/xeBw9PT2+v1UqFR8jvt3g2/oWClZu+U2LeLWuQ1qlf+tK/+//+mLwnGLw\nkIMDAOCfNwePvWd35+XguJn9/k6b78c+IwifkRUefyx4TEpQ8zddaJN4HwDBfpDWtM8Kx/hK/beF\nc04Jx36T6aOkix8Xrvsr9vszwjl/JByrBzb73MfGxhQRjZjy27ZtQ6VSUW54AIoVT4M2/Qb8VeWo\ncpypopx0r2w26yPKZTKZ0MpyUjW6boBbuTs4ODg4tB02VeH0fe3RaFQR6mgAzeVyyGQyvopt+m/C\n3r17VeU4vcKbXlHOdC/P89RWt1wuB2CegW+qLNfJSm/1wBHqHBwcHBzajmYIdZFIxEeOo1W8iVxH\nBLxIJIJ4PK7IcbyinHSveDyO6elpRX4DgKGhISMpb2RkBENDQx0hydUDN7g7ODg4OLQdOsmNisNw\n0lkmkxGTvdDfpH/T77GxMVQqFdx5550AzpHwIpGIOhbWJoHIfmH305+FXPF6kRt+/kKQ7Jxb3sHB\nwcGho6iXUNdIm5yEZ4vh4WHjZKBRLATJ7oKqCifhf2uU+ePggO4gz0loFYEuLRw7JBzjNuKkwAD7\n7T8JHuO8uHuEJDa/fWOwuls7E9TY4ArhmI0pkchz+28NHrthj/+30HUi9lvo419ZtNMseU6CLaGO\nE9WAc2Q16e86kY3avOGGGxQJD4DvmtnZ2UAbOlluYmICwPzKn5LP6Pe3IeFxuRcCbuXu4ODg4NB2\n2BDq+vr6FJktm80inZ6fXhJZrRaRTS8yQ4Msb5P/zuVyiiyXy+UUSe/OO+8M3CORSODqq68OJeFJ\ncgOdJ9m5mLuDg4ODQ9thQ6grFAqKzDY0NISRkRFfRje9DRORTSfIlUoleJ7na7Narfp+A/CR5aLR\nKMbHxxGJRHyFaOgeNGCbSHiS3AtBsnODu4ODg4ND22FDqCsWiz4yG/2fyGo6ac2GyEZEO71NiTDH\nrwHO5aEfGBgIvYfUJpfbdF074QZ3BwcHB4eOIpFIqHi4jnakjeVtUvGZWqinHOtCprs1wWpwP/zR\nc/++apEkdf53wrGnOi6Fg8PihkSek8CjiW+3ZYDVaAcAlty5sOQ5CUdb2BYnzzWDw61rquVoN6FO\nz1ZHRDieie6GG25Q5V/T6bRIhOu2THONwhHqHBwcHBzajnYT6rLZrCLCmTLRlUolX8a6xZBprlG4\nwd3BwcHBoe0oFovKFb9r1y7s27dPEeS2bt2KO++8M0B2GxkZQW9vr49QR23wvw0NDWFsbAz5fF7M\nREf/9zxPrd7D2hsfH8e+fftCn6lQKFi77jsNq5Kv0a+cK/nq3PIODg4ODhJsS77SoMpj1WNjYyiV\nSi1PItMIisUipqamQrPKFYtFrFq1CitXruygZHawirk/8US7xWg93EDevdgmHBvpuBQODp3HX10e\nPPZfXwoe+5cG2/82G1e/ImTNebLBtpuFTcwdmHeF01Y0ginmHovF1Ja0WlXbgHMxeP63QqEAz/NQ\nrVYxMzMDz/MQj8d9FeYkGeLxuMprT7J0mhVvgnPLOzg4ODi0Ha2OuQPAnj3zbEQ9EU0+nw+0o1eJ\n09vQ4/S0bY3uSeCxeP16feLw6KOPtqvrGoIb3B0cHBwc2g49vk1JbJqJuZdKJXzgAx8AAFUdbnJy\nEoODg4F29Cpxeht6nL63txfFYhETExNKvlKphNHR0VAZ6L/NmzcvQK+a4fa5Ozg4ODi0Ha1OYrNl\nyxbjvfR2xsbGEI1GVVuFQgF9fX2+9gm0BS6fz6uQQU9Pjy+RDU9M0y1ueA43uDs4ODg4dBTtTmKj\nt+N5norLA/PV4latWoVKpYKJiQlxcKYENhRLX4ywGtwP/KTdYjhcSHDkOYcLFR8XyHMCx84Kz384\neOyru/y/l1q29R/Z7683IlANtCOJjQ2hLh6PY3JyUrUVi8V8bvV0Oo1Dhw757kOoVCqYnJzsSsJc\nLbiYu4ODg4ND29FJQh0/r6+vD5lMBtlsFtFoVB2jTHamBDb6FrduI8zVghvcHRwcHBzajk4S6vTz\ngPkKbtQmT2ZDkwv9PjpRrlsJc7VglcTmv19/LonNN9oukoODg8OFA8ktL3jvA7Bxy78iXPeYncr9\nCAAAIABJREFUcKxVbnnbJDZURY0T6lqdxKZQKChCXDPI5/OoVCqhCW1ada9Wwa3cHRwcHBw6Csrt\nzjE8PNzQwF6pVFSyGR2RSASe5zUkow49yY4JrbpXq2BFqHt7u6VoA35ROPZvhGPXsN+PC+dkmxfH\noU68UzgmrUI6Da5X310QKc5ffEo49tWOS9FZ2KzSJZw4HjwWYb+lVfoG4Vg7CHQc7STUzc7OipXe\n9AxyvB1eKS4ajSKbzWLz5s2Ynp6vRLiYs9W5lbuDg4ODQ9vRTkKdqdIbTRA4YS6RSODqq69WleJo\n8N+8eTMGBgYU0W4xZ6tz+9wdHBwcHNoOIrJlMhns2rULr7zyCm6++WbMzMxg69atAObj1pxQNzg4\n6CPUURv//M//jF/4hV8AMO82z+VyKsscnTM+Po7Dhw+jVCr52imXy2qAnpmZQTweRzabxdKlS1V7\npVIJs7OzSn7KVmeSh4h30WgUmzdvRrlc7nAP+2FFqHtII9R9qe0itQbOLb+44dzyFyYuRLd8o/jO\nLwWPPfy0//f9wnWSW/65VgiEzlaFs6nIZlPZDQDGx8fheR6ee+45rF+/3uoZuh1WK/fFMqDrsDW6\nT7LfCzvXciB0w0AugetVUjhnUji2WPDb7PefNNhOv3DsZYvrpIH8k8Ixm107VwjHjlpc12n82drg\nsf/9Z7Wv++Wng8cO3O7//cw/BM9p1UBeLxqpCpfJZFScm1d04/H0QqGg3Ol6Mho9Vi6dQ3IBwLXX\nXotisRiIyZdKJSUDeRm6HS7m7uDg4ODQdtQbc8/lcgDm4+pSRTeKcedyOVXZTUpGU+scqc1kMqnk\nK5VKvqpyiwUu5u7g4ODg0HZI8emenh5jzH1wcBAAcOutt/oquvE2BgcHVXGYQqEQiK3XOscUN9fl\n06vKLRa4wd3BwcHBoe1opCocEKzcph/nf+NV5KTrwyrNUZuUwrZWrL6b4dzyDg4ODg4dRauS2JiS\n1zR7vk3Smm6H1cr909q/722TIAsFTqBrJ5PUofX4EPv9cIfvv5jJcxIaJdB9aYX/99+eDp5jQ6iT\ndrnYkOeuFI69aHFdN8CGPCdho3DsHkag+w9CiPjA51YFji351KnGhKgDjSSxoVW0XvFtYmJCTF5D\nld100hyR8Igcd/bsWbz++uuBZDfkbudV4RYz3MrdwcHBwaHtaJRQl81mfRXfgHNJYvTkNTppTifh\nJZNJXH311Ugmk+jp6RGT3Ziqwi1muMHdwcHBwaHtqLcqHBHqhoaGfBXf9ApttB1OJ83l8/kACY/u\n8+abb6rzdXl49TmSbzHDEeocHBwcHNqOVhHqtmzZ4vtdK0ZP7HherU3/Ta74gYEBFAoF9Pb21l3A\nxqZyXCfhVu4ODg4ODh1FJwl19VZra7S6m4mEVy/pr1WwWrmfbyS6MDjy3OICJ9D9B+GcbwrH/nS1\n//dvHmuRQBco/otAoOP4XI9wjNn3RtP5LhbynISSkILvdsYilOySxMN7i/3+lJBy81MdIM9J6BSh\nLiyTnd5OLpcLEOpolU/u+lZUjsvlcqKM7c5051buDg4ODg5tR7sJdTaZ7PR29HOy2awi41Guep6l\nrpnKcZKM7YYb3B0cHBwc2g6dwLZr1y7s27evpYQ6TqLj5L3BwUFfO/o5Q0NDuP/++3H//fcHyH70\nf2B+ojExMQEA6O3tRalUwsTEhLqGKsfp5DzP80QZ2w2rqnDXa1XhHBy6Gc4t372wcctfiGjULS9F\neJladzzM2MmqcLaoVCqYmJgwZrrTZQqrIEeV44rFIjzPCxD0ug1WMfd//eJS9e8l/+cbbRPGwaFZ\nSAO5VBDyf7RoMJc+76nWNN12nP1M8NjFf9S++0kD+b9jv59qsG1pOHnW4rq/ujx47OMvNShEg4hb\nZOn5o6XBYy8LpvjPvxzz/V7yOzMNStV6NFIVbnZ2NhAXp8EVCE9YQ+3qMe9mEtfQoN/tgzrBueUd\nHBwcHNqOemPu2WwWfX196hqKV8disdCENXq7dN8LJXGNDrfP3cHBwcGh7ai3KtzQ0BCmp6fVNRSv\njkajvipvUtU4aheYH9zz+Tx6e3t9MoyMjGBoaChQQe58gRvcHRwcHBzajkaT2BBqxcxN50UiERXD\nNyWuqfceiwFucHdwcHBw6CgSiYRiquvgBDuHxmE1uPdoLNfnPxz8+7W7WiVO6/Cv/3xb4NiXb/rH\nwLErrvD//tr3g23dILT/fzUol0N7cYdwbEQ49twz7/f9XvK/PtbQ/SZyQWN08+BY4NiTDbXeOlwh\nHFsmkOd+/An/73f9ZWP3O3B78Nj/+w/BY636jiTy3P5bg8du2OP//fsCeU7g2IGf9mdrg+dI1d04\nE96GPAcAn2W/JR7zPs5GBDDzYz+B7rX/Ejznmi8Fj3UiCVAjSWxsCHWe56FarapkMrOzs77kM+Rq\nJ8IdcC7xDACVEOd8WrUDjlDn4ODg4NABtItQR251SiaTSCR8yWf27t3rI9zp5wLnEuKcb3BueQcH\nBweHtqNdhDqaIMzOzqp76W1v2rRJEe4oLa1+7ubNm1EuC3l6Fznc4O7g4ODg0Hbo8fRMJiO6wRsl\n1PF96vWce77CDe4ODg4ODh3F2NgYenp6Amz5VhPqWlWGtVAooKenxzqBTb3nE1pZNtYq/ey3tPSz\nX236lg6LEb8kHHu641K0Dj0sFer3gvxLXC8QwBy6A59mv7u1cuVG9vugcA4nzwHAF9hvlngOAPBd\nIfncg3ZitQ3Npp+1RatSytq2VSwWsWrVKlVUphbqPb+WvLbPq8Ot3B0cHBwc2o5G2PI2pV5rpZSV\nyrCOjY0pRv3Zs2fx+uuvn3dlYx1b3sHBwcGh7dDZ8v/0T/+Et956S2TL0zm8RCsgl3oNSyl78OBB\nfP/73w8cTyQSuPrqq7FixQr87Gc/U2396Z/+KU6ePKnKxh47dgye5+HFF1/E888/H5DpwIEDOH36\ndEvKxhYKBTz99Lw/tBVlY93g7uDg4ODQduglViORCI4fPx4o+To9Pa3O4SVa9VKvvb292LNnjyrJ\nSteMjIz4yq329PTgxIkTgTKsdN3s7CwikYgqyXrTTTehUqmoVLanT58GAPT09ODkyZM+mfr7+7F0\n6VJccsklAObLxo6PjyOfzxvLxr722msYHR0Vy8a++eabOHHiRMvKxlq55X9927l/f1XKCOJw3kOK\nr68Xjr3QbkFaBM40uRDj6xNbgsdSD3VejkbAY+y/LZzzJ50QpAZ4jP2LQnW3PxAS1GxnMfb/LMTX\npeRa29jvbjPXVPFtw4YNOHz4sJg8Rq8KF4vFUKlUUK1WlcsaAH7wgx+gv78fnudhenpaXZNMJlEu\nlxGLxTAwMICVK1di2bJlAIDdu3ejv79fba/btm0bpqenUSwWEY1G1Xa4a665Rm256+/vBwBs2LBB\ntUMynjx5EkuWLEE8HseOHTsCoYStW7cil8vh1KlTiEajuOGGG/DAAw9gzZo1uPjii/Hggw+iv78f\n1WoVW7ZsEePtOiGv3jK4Lubu4ODg4NB2eJ6HaDSKTCaDYrGIdDqNvr4+FItF5HI5VCoVDA4OqnO2\nb9+Oe+65B8A8izwSieDRRx/F0NAQ+vr6EI1GlVueriHimV7pDYBKXtPX14dyuYxEIoHp6Wm1+s3l\ncupe1Wo18Juwd+9e3HHHHahUKojH40ilUr77c5n1+5ZKJdx+++0ol8tIp9OoVqsBWVtZlc4N7g4O\nDg4ObYdNEpuxsTF1DrnAyRW9Z88elXCG6rTbVnqLRqOBanHkfi+VSupeRHjjv+PxOEqlki8hjv4M\nxWJRlPnQoUO++6bTaSV3LpdTk5l2VKWz2gp3yb3ntsIlus3P47BgWMxueYfF7Zbn6Fa3PEejbvnf\nsXTL97PfnTbXtlvhTPvcWwXbfebkTm9WjlbuT28VHKHOwcHBwaGjIOZ3uxCJRFRhGBNoC5vtgFyp\nVBRrn4O28nUTrNzyp19rtxith5CTBB8W3vVBf2gGR4Xr9rZCoPMQrVylf4z9nhXO+XYL72eDXxeO\n/TmrAlf8RrACXKOV1DqNX23jKn2zcGy3xXXSOkvK+v0v7PdJi7Zt8X2h8uWx4/7fvywwTKX1Ki8U\nNy2s0r8jVHf7H0/5f/+i0PZ3hWNfYx0/InS61MdTwrFWw2afez6f9+1lB+Cr9qbvKc9kMmqwbXRf\n+qFDh1CtVo1763O5XOh+eH0/fbfBrdwdHBwcHNoOm6pwwLm97LyCm14xLpfLAYBvz7fevk6U0/el\n6+fohDjT3vpkMomrr74ayWQSPT09al8634cOtJYM1wq4wd3BwcHBoe3Q96PzfeC0z13fy04r8YmJ\nCZRKJVSrVVUxbnBwEAB8e76l9vleeemcUqlk3FtfLBaVjG+++aaqWiftm28lGa4VcGx5BwcHB4e2\nQ88jPzs7C8/zAvHuLVv8LE/d5U0kPH1fvCnXeiwWw+DgYIBQV6synfQ7k8mgUCjA8zzVHq3g9fb1\na7uBYOdW7g4ODg4OHUUjhLrh4WHrRC42hLp6ILXX7QQ7q5X7/x9eAADcIWTyWuhKRBIe+YcgvWXm\n8WcDx5b+o/93SWDUSSpSFI41gn99/L2BY0t++Z9a1Hp7Ufpk8Fj8G4219dcW53yuRzhmYR+kufMe\ntg9sibAHTHqU/37KT93qS0UD51wp0AFfDJVwYfAH1waPbWSfTf/Oxtp+1fa8/+T/vfn/CZ5jw+f9\nLYHBt18gk/H38G2BBfeVXcFj/C0fuD14zj2CbeTq+fWvBMu7nfxRcJ/bg4xQd0ewadwvPPMvsmdu\ndIvgkHAsa3FdGLqNUMeJdAS9uIzUnn4/Xsylmwh2buXu4ODg4NB2dBuhzvM81V42m1X30Yu18Pb2\n7t2r7kfFa4DuJNi5wd3BwcHBoe3QyWy7du3Cvn372kaoy2azePbZZ0MJdQBUe0NDQ4GiMxL5Ts9Q\nR6t5mpyEEezy+TzGx8c70s8ER6hzcHBwcGg7dDLbhz/8YRQKhbYR6oaGhnDbbbep0quSDIcPH8b0\n9LS6B7VFJDlOlpPuRRwA0zWEeDyOqalOZBM4B6v0s//H9efSzz7edpFag7eevjlw7O5f+k7g2FWX\n+X9/7USwLSHvBF5uUK524n8Rjn2v41JcWBBoAPjqquCx3z7VdlF8uJL9to35/zL7bfu9/+Df+39H\ng1QE/Cchuc9h9vsJgcfxpwL5gacOei5MuBC8TzgmrXgeY783Cud8LBhOD1Rz6xEU5ue/Fzx2yZeE\nGzBIunc3+/014Zw3Pxs8tuTzte8ngb+u3whJP8tj7iMjI+jv729bzJ2qvYXF3Kk6HMXMqR1KXEM5\n6bdt24ZsNhuI0YclyAHOxeClim/thnPLOzg4ODi0Hd0YcyfWvt5OMplUskWjUVVBjsfow9ruhhi8\nG9wdHBwcHNqObkti43keCoWCquCmt6PLpsfP9Rh9WNs8fr8QSW5czN3BwcHBoe3oZBIbU4xclyES\niRj3zUuJa6T7A/OhhFwu52ub5DbF4DsBN7g7ODg4OHQUiUTCx1q3gT54dgKRSASrVp0j0JjuvxBk\nORtYDe5H2i1FG3CRQJ4TwQh0/0Y4hVeg6lY48lznIUXROk2ek9Bo0pxGCbPf+jv/78/9l+A5NlXh\nflkgz/3Q4rqPC8f+yuK6JyzOAYAN7LdE4Pve54NMyv/8Kb8ySGHXay3IcxJ+QTjGCXStJM+tEY5t\nlbLdGNDqJDb8N+BPQCOR4niyGV2marWKaDSKbDaLzZs3o1qtqgkItXXDDTco93o3JayR4GLuDg4O\nDg5tRysJddlsFn19fUYCWyKRwNVXXx0gxfHzdJko+9zmzZsxMDCgJgw6wa5UKnUFWc4Gzi3v4ODg\n4NB2EEEtk8kogllPTw9mZmawdetWAMBDDz2EzZs3o1wuK0IdreBjsZiP0DY9Pa3aGxkZwdDQkCKw\nlctlNfjOzMzvR6Rc8IODg75ENdQGMJ/8pr+/HwMDA0pGWvmTrIVCwddGuVzuuopwgBvcHRwcHBw6\nAFtC3Y4dO5SrvRahjmAisNE+eH4fnnQmn8/j6aefxu233672pOukPOmeC0mWs4Eb3B0cHBwcOgoT\noY7i31KyFxOhrVKpYGJiQhxgbclu8Xgc1157LQYGBgJlYhcrrAb37osmtA+LhTzn4NBN+L/57wZJ\nYlLWNRv8VYPXSfiPwrGvW1y35FONMSl5lj5b2PCzGyXPSZAq/b2flYo7EHK9DaGuVCrh0KFDgYpt\nPBOcTnCjveqm6mzVatV3bym7HJ2nZ7yjUMFihSPUOTg4ODi0HTaEOs/zkEqlfOfyTHCc4FarOhsQ\nJOPxNgl61bfFDueWd3BwcHBoO3Ty2q5du/DKK6/g5ptv9hHq8vm8Iq8Vi0VkMhmVCY4yyR06dMhH\ncEun08jlcujt7UUulxPJbp7n+ch4dK+nnnoKiURCrPpGyOfzqFQqHc0L3wpYFY7591rhmG4smOLg\n4HB+QIp2djo9SKNueQfgQEjhGB2UqpXH0cPi581gbGwMpVIpkJHOpqDLQhR9aQWsVu5uQHdwcOgE\npIG8XzjGbdJnhHP+qEEZFvNAznPKZMWzFgY2MfdcLheIn9dTqY0Y8qZqbeSCD4vJT09Pq612lMN+\nMcbjXczdwcHBwaHtsIm5U/y8kUptVCkum83WVa0N8MfkH3vsMQwMDMDzvEUdj3cxdwcHBweHtsMm\niU2lUlHxc6lSG51P58bj8UCluKGhIV+MPpvNhia74TH5W265RYzxS/H4boZVzP16Lebu4ODg0El0\n2i2/mLHQbnnbmDslpGk2jm2KpS9UO/WinWQ9t3J3cHBwcOgoTEls6iXUtapSHLXTLkKfCe2sKOcG\ndwcHh6bBV9etJOHatHUhrtIl2KzUVwvHjrVaEAH1Euo4WQ6Qq75RpTa9mhxVknvppZdw+eWXK1Ie\nT4xDLn7gHHGvnYQ+/hzthCPUOTg4ODi0HfUS6sKIcHrVNyraQtXkhoeHMT09DQB44YUXFAmO33/P\nnj0A/ES5hSD0tauinBvcHRwcHBzaDtrbDgC7du3Cvn37FFFu69atuPPOOzE3N4cnnngCpVLJd/7I\nyAh6e3sVEY7+XiwW8dRTT2FsbAybN29W9+rv70dvb6+PBKe3VyqV8IEPfAAAfOfE43GVzlYi9NH/\n6T8ARkLf5OQkBgcHaz5HuyrKOUKdg4ND02inW96htWinW77ZJDbFYhGrVq3CypUrre+5kElmFoqI\nZwMXc3dwcHBwaDtsYu6VSkWtcsNi6gBUXD0ej/uSzHQink1oFaGvHbAa3N+p/fuVNgni4OCweMFX\n6t2QRtYB+KRwbAvfLwfglg7smfM8D9FoFJlMBsViEel0Gn19fSgWi8jlcr4tYclkEsB8chk9pj40\nNIRkMonJyUmk02mUy2XVfiqVwsTERCBBzYUKt3J3cHBwcGg7bJLYUPIYYqTrSWs2b96McrmsWO+0\nggfmY+ijo6O+ojHnS132RmEVc/+gFnN3K3cHB4dacCv37kCnV+62MXeH9sOx5R0cHBwcHM4zWLnl\nv33gQLvlcHBwcHBYIDgLf/6hplvewcHBwcHBYXHBueUdHBwcHBzOM7jB3cHBwcHB4TyDG9wdHBwc\nHBzOM7jB3cHBwcHB4TzDokpic6ElJhgfH1f/rlarXZvDuFOg9JIA6qq3PDo6it7eXlWSsV1w7+v8\nxWLTvX379uHmm29Wv0ulEjzP853j9PP8RksGd674tgo9NzeH6elplEollXZQ+oh27NiBYrGIaDRa\n1+AutSXJplf72bJlS0CuWh+2adJhum50dBQ//vGPsWLFCqxfv974oVGVIkI0GrV+9kYGGttrpPO4\n4ajXuPC+kvouFovh0Ucf9VV/spHd8zz09vaiXC7X3S+mdyU9C5V5JESjUSuja2Ng65E7TO96e3vx\nwgsvYP369aFtNTpRqXWdyTaYrtOPP//88zhx4gRuvvnm0Pdge+zOO+8U+8pG9/TnqFQqxmcm3Xv8\n8cfVebb92YzulUoln00i22GyJ2E2201aFydCB3dp8JWOccXXjSmBD6DAfPm7RCKh0gtKbQHzeYaT\nySR6e3sBmI0Xl01qS5Itn88jlUqpj4/LZRpU+KSDy2W6zvM8vPXWW7jkkkvUBygN3Pwjm5ubE59V\nR9hAQzAZ+Ww2i09+8pOBa/jHbTqvGePC+0rqu5mZGQwNDWFiYkK1Tc9SLBZ9VaF4fz7wwAO49dZb\n654whb0r3o7pfdn0i/4suoG1fT7g3CQzTO+oLUkGXX/D+ilM90y6wWXgg53pOl2O06dPo6enp+Z7\nqOeY1Fc2uqfbkVo69cADD6C/v9/43k32rBndGxwc9E0C5ubmAgsQ3Z5wu6jrYq3nu9A8qosFoYO7\nNPhKxySjS8aUwAdQwK+gYW0B80UBRkdHQ40Xl83UFpft5ZdfVrmMJblM7fBJB5fLdB0AFAoF3Hrr\nreoD1D80/rFLngv9WePxuPVAA4QbeW4QANnQ1zIc9RoX3lf6b5KZcknrOaXpWSKRiCg7AKTTaaRS\nqcDAqp8X5p2R3lXYs+jv65577qnZL/xZ9ImnzfNx/QjTO9J9qS1df8Oer5bumeTUZZAGO+k6PthN\nTU3VfA+2xwD5G7XVvQcffBCbNm0K/dZ03TP1S5hHqhndGxgYEG2HyRNKz0PPTLpoer5GPaoOnUHo\n4C4NvtIxDl2hCXwABeY/lL6+Pt+Ar4NWjNVqFblcTh03Ga9GZUun00gmk2pGWksuHfqkg8uVTqdR\nrVbV4C/JIH0g/GPnkwj+rGHGxfQBhhl5fk09H3e9xgXwG88XX3wRlUrFZ0yr1SpOnToFz/MQiUQC\nKwfdKEnelEKhgEQigUKhYJTTZGBrvSvpmPS+bPqFP4vN80n3s9E7qS3pu5LuZ6N7puerNdiZrtOP\nZ7NZq/cQduzs2bP4zGc+49OzqakpxONxK92Lx+PYtGmTz47Y6J4kk8metUL3JF3Uj+k2dnBwEIVC\nQb1Trov8flLbDt2D0MFdGuSkY1S+r1QqGY0pH0ABecDX2zK5g+LxeMAlKMmmtwXAKNvMzAwKhQIi\nkYgoF2/HNOnQ5TLdix+XPhDpY9cnEab3YDvQ1DLypo/W5rx6jQs3nqdOncLMzAz6+vqwcuVKRKNR\nJBIJ1e+33HKLb1DgRpbfL5VKYXZ2FrFYTFWeks6T+tzmXZn6ir+vsH6RDKzustaNbq37lctlK72z\nNc7SOTa6J11nM9iZZNKPl8tlq/cQduzkyZOBQbtarVrr3v79+wOeSBvdsx0QW6l7qVQKIyMjoieU\nh+8OHz6M8fFxcbCX7sfv5dA9CN0K9/LLL2N6etpXH1c6FovFMD09jZmZGaRSKcTj8YAxpQFURzqd\nRiKRUB8QbyuTycDzPPUfKdX+/fvR29uLWCwWKq/eFgCjbNS+SS7eTiQSwcDAgDqHztPlonv19vb6\n7iXJkEqlfJMEjng8jomJicDEqFgs+t5DMpnEpk2b8IEPfMB3fSqVQjY7X/Ypn89jZGQEExMT6p7S\ndZJMtufp96P+NJ2TSCTgeZ5aya1atQqFQgHFYhGFQsGnM9PT08jn89i5c6c6JumCfv9IJILh4WEk\nEonAhJDLyWH7rvixdDqNnp4eDA4Oht6PrpP0yaRjejvj4+MYHx/H7Oysun89esdlLxaL4qDDn89W\n96Tn5TLY6hQ/bqN3gFn3/uVf/sWnd1Rj3Fb3JDvIZTDpHpdT6vdW6V6lUkEkEsFdd92lztHtCdnY\n3t5eTE1NIZlMGvWO7veXf/mXAb1z6D6Ertyl1TYNfNw1mkwmMTg4qBQa8BMtpHgirY650aW2CHx2\nL634TbLpbZlkA/wuKEkuvR2TK1KX65ZbbkE0GkW5XPbFrbgM6XQaAHzPyz0F0iohnU5jYGBAeRsI\nfCbNP27TSla/jmSSdgZI5+myS8bk61//OoaGhkLPIeMZRsjyPA+xWMy30uC6ILUteTT4ebzPAbt3\nxfvK5LHh9+MGls7J5/MYHBzEypUrA8d4O/l8PuDZolVoLb2T3h+fxHI59WeupXvSezB9f7a6R3JI\n50j3A2rrHukdvTMTcYzrnmQbbXRPOkfq91bpnv58xHCX7Anp0uTkJG666aaA3unP9+EPf7jhXTwO\nnUPo4M7d1YA88A0ODqrZ7i233GIdT5QMajQaxYMPPoihoSGjO0j6sCTZuFyAbOjT6TRisZjRDS/J\nJLXF5arFsq9UKkgmk4GBgH/s0mSmp6dH9DjQcxP4x13LyHMDwWfuZGBNgxi/X6VSwe/+7u/6CGuS\nwSHjOTQ0ZCQekdw9PT2+Z9b7XGpbcify8yQDW+tdSX1Fkye+cub3CzOwe/bsUTtKcrkcPM9Tx6Tn\no74l/bDZ3WFy3/OJtSSnje5JcnIZ6tG9YrGIVCqFp556ShH69HOk+9nonj5oh3FXSPdeeukldXxy\nchKlUkldZ6N7pn7h/a73VTO6ZzMhBuYncToZmesdl13friqFSB0WHqGDu7TalgZkit2QYpjivPoA\nCsgzVumDBILENT7pkGTjcplkk7a+6XKZZJLitrpctVj2S5cuxYkTJwKraDqHPnZpMsONhGmwlT7u\nMCO/detW/OhHP8JLL72EpUuXBvqXYvW6Idm2bZs6h99venoa4+PjmJ6eFg0qgQ/ctkQ8wG9kpbaB\noEdDOk8ysPxd6c+bSqUCfRWJRDAzM4NisehbOfP72RhYAOjr61M6b5Kb673N7g4ijunPo+svyS3J\naaN7pvegy8CvC9O9WCym3v3y5csD50j3s9E9acJYi4h39dVXi+ROG92TzpH6Xe+rZnSPJk2690Ky\nJzMzM9i+fbvSf6530vNJY4FD96Bm+tkHHnjA99vkuuPgsaCRkRHMzMwEVkbJZNI3ONAHOTk5iT/+\n4z/G1NSU2ralG3V91WCSzcSe57JJ5+ly6TLp8V7elkkuU/98//vfVzE5faJCH/vu3btj0uVqAAAg\nAElEQVR9qwQ9Bsg/NFOsdXBwEKVSyfdx63FO/bpt27YhEong4osvxrXXXot3vOMdPpnJwM7OziIS\niSCRSKBcLmPPnj3G+3meh3g8jsHBQSUrrc77+vrUdTt27MBnP/tZPProowEZJbl1Ip4eG9TblmLS\nBC4D9Tltf5LeFX9eU19JesD7hQysvjolA6sjm81i27Zt6t1L79PkdQj7LqX3J8ktyWmje5KcXIZ6\ndA8AEokEPvaxj4nnSPez0T2ud0C47v3qr/5qIFZPsNE9SfdNdoP6qhnd071BU1NTRnsCzJM3CVzv\npD426Z1DdyB0cE+n0/j0pz8tMlfvuOMO9ZtWAJVKRRyMAb+hmZqawtTUFPr6+jA7O4u9e/eqv+kf\n5JkzZ1Aul5HL5TAzMxPYYypBl43kisVi6p6SbFx+Lhc3EmFt6XLFYjHkcjnVf9IHTzsMHnroIV8f\n0AcqDWBA8EPTZ+36AKV/3DpSqRR27dqlVizT09MYGRkBML9ao6RDHIlEQhlwiczG70fylUolDA8P\ni4Q+em+2ZL1cLhcg4pGbUG/bREqTZOBGUXpX0vOa+orrJ/XLX/zFX1gbWGB+BUWeJal/9f6jyaiN\n3pmeh+QOGwhsdI/LyWX42c9+Vpfu9fX1qdXywYMHA+dI/WKje2Ek1DDyGifY2eieSff1fje9r0Z1\nj7xB8XjcOCEGznkn9f7W9c7Ux3xx5tA9MLrluctsenpafYzkBqWY0csvv4xYLIbnn38eAwMD6jp9\nMNa3zxw6dAie54nnDQ0NqexXn//859VxHg/TV7pjY2OibCQXualM96wlvy7T8PAwRkdHjW3xFfjg\n4KByj0pb+6TkPsA5foLneSqzn+7+kuKzUhhFd/VKW/joY49Go2qSQsREHvYgAwvMG38b1zIPeZjc\n+fQ3iaxn2noI+F3EW7du9bVNpDQeGzSRCnVOiPSustls4Hn1vqIEK7Ozs4FtRNQvP/jBD4x79k0x\nYt21nMlkAq77aDSq2OuZTMZK7wA5eRNtK4zFYkY5bXSP6wGX4dSpU3Xp3u7duwHMcz5WrVoVOEcK\nadjqHg/Z2Oie5L3Q3eQ8lW2pVMKmTZtEvePbOVupe7q7nQZqOpeSAhHxTtezarUaCGnwPuZ659Bd\nMA7unJwxMTGhDHu1WjWyZ+kl85etD7Q6kYSfJ2XAA+CLBwHwGUE9u5UumySXdM9a8nOZ9A+Tt6XL\nRSv+qakp37n6B28aIPWPXSI1ScYMCJIWZ2ZmcN9992FoaMiYaS6fz+PRRx9VKx2K0/H4s25gY7GY\nMXZH9yM5dZA7uFQqqXYAv/GUyEI2LGbeNg08PDaokwpNfS69K+l59b6anp42Doi6kdUzrYUZWIoR\nkx4Vi8VAbBSQ0z/X0jsgGHvV9YwGAmliaaN7XA+4DEuWLKlL94aGhlAul/Hyyy/jpptuQqFQ8J3D\n70dy6tD1Y2RkBIODg4FB21b3eKye3OTJZDI0bwfXO97vUl81q3vcG8S/Y+IC6HoGIPCb93FYZj2H\nhYfolh8fH8fMzAySySQOHTqEUqmE2dlZtfeT7wEdHh727UMHEIgd0gDK2df8PClO2Nvbq9J4joyM\nYHJyEuPj4+ojMskmySXds5b8YZnvtm/fjkKhgHw+H5BrbGwM4+PjARetvq9VimmSC+6ll17C2NiY\nb08trST0+Cy5UPVZOyGTyeB3f/d3kclkFOGMjLW+fUx3yZlcv7qB9TxP3YtWOvx+gD80wp8vFouJ\n7kseizXJTe8ukUioGKbkvqTYoO5h4eEQU+4E/V1Jz6v3lRQq0KEb2d27d2NqagoPPfSQz5hSv544\ncUK9a/JO0MDAjTVPl2qjd9LzSHu39X6pR/e4HphksNU93Yu2b9++wDnS/cJ077LLLhNDNra6J8Xq\nJTc51z0pDGfaM98K3eP9S31I33GpVPLp2Te/+U1fuETXO97HXO8cugviyp1WeJ/61Kfw1a9+Ffl8\nHtFoFPF4HD09PYGVu8QupbgcgRRkbm7Ol+YxlUr53JdS9iuaaY6NjSnjB5wbdE2ymbbkcdl0Juzv\n//7vB+QypaOlScfY2BhOnjyp5Dp58qT68J555hnfBwn4P/hNmzYFViq0mj9+/LjP8ESjUTWBOHTo\nkE/GMBdqLpfD4OCg6mO+krVx/dK7iUajGBoaUu447p7l99M9Nvz59NAOrVK+973vqf6anp5GLpdT\n50jsXNscCIlEwtfHPBxiuk5/V5I7WlrF2mzL0g3s29/+dp+BLZVK+KVf+iVxxSi57ikz4tGjR631\njvpAfx49TwRlKdP7hVy1Nron6Z0kg63u0Yq7VCrhtttuC5wj3S9M9y655BLlOQDmB109b30t3TPt\nCJJCCrruSWE4U+6QVuie5A3Sv+MDBw74bGOpZC5sxPvYlCnUoTtgdMvPzc35XqxUMpXAFb2WEb71\n1lsDaR4J0gdJhnDlypUYHBwMpDo0ySZ9gJJsekpKSS5JJuDcpIPiaLorMxaLIRaL4Y477hA/APrg\npY9W/9g9z/OlvqRMWlRIpFZ61kqlgnvuucc3GHADa+P6JZnJwG7ZskW5LflWRv1+UsITycVI70Un\n/eixWElu6R1LbXOugDQJMBlY/V0Vi8XA80p9ZbMtKx6PGw2sntiGQOQ2yViTW/eiiy6y1jsgODjo\neSJIr/SJZT26J+mdJIOt7umua+kc6X42uqfbg3p1j8fqpS26NjwVUzIvva8a1T3qW9293tPTYwyf\nhhU24n0shQsdugfi4B6LxbBnzx58+tOfDhAmJNIWEKzaxrdI6EZYJ+dRiVJqX/og+YrR87xA2VmT\nbPwDNG3fSKVS+G//7b+pPaO6XKaMXPrqa/PmzaoUZk9PDxKJhIpX8nvpH/wbb7whEqT0j12fjRNj\nv1AoWCXRkAYDYN5oXHbZZb5Bg1bS4+PjgWpmQPBjloyL6X68Ahw3gvReLr30UjEWS+ArcP6Opbal\nGCM38iYDq7+rH/zgB9izZ484qOiQ9k5zIxtmYMMK/0jGmgaMyy67zFrvpqamAu9PH+joGfRBmzxu\nNrpn0gOSYWpqCq+99pq6tpbu6YOi53mBc6T76d5CAtcP3R7oOxJq6Z7uZdDfMbc/NjwV037xVuge\nJwNL7wqwK2zE+/jUqVOit8uhOyAO7vqMfXR01EeAA85VFAtjMYcZYYoP0bW6oksfJG3dIOzcuVN0\n3ZFsq1evRqVSEeWSZCP35caNG9V9dLkkmYDgpIO2aJFcppmt/sH/9Kc/DQyQ/GPXZ+PValUx/22S\naEiDARmNSqWCAwcOBDwVpixaUqpXblz4/SQjIRlBqtNtIk3qcgPnVuDcyEpt80kYEDTyJgOrv6vp\n6Wn83u/9nu/vNoaSZNCN7I4dO6yu454JyViHEeNMz0JxaP39SRNfnpXMVvckvdNlGB8fx+rVq7Fh\nwwYr3dOfcXJyUsxwyO9no3uS3gG1dY/0TV/JA8Fy0lz3Dh06ZJzYcnvWCt2TCMqSnQgrbKRfp/dx\nPB4XM4U6dAdCM9QBCKxWiZgFnIvZEfhgqOeP1reX0MwPCK7cTSUR8/m8T7kkkGwmuSTZdHbswYMH\n8fa3vz0glyl2zycdY2NjPnlM7nweb/2t3/ot39/5x65nmdK3d3FjIK0YpcGAjMbk5KTaE33nnXfi\n4Ycf9mV5MyU0CTPE/H75fF6sXMWNYLFYxC233KImbpOTk8oFS5BW4JKR5W3zSRitgHUjbzKw+rt6\n88031fN6nqdWtLyvbLIg2hhYgl7VS29HjxHrK2kbvdu8eTOy2azv/UkDnS6nvrKtpXuS3ukyvPHG\nG9iwYUNN3SNvAQD87d/+LS677DKxz6X7mfpT1w/SO30gt9G9SCQSmJQXCgV84Qtf8LUleR1tJrb8\nfTWqe5KtlOwEULuaHPXx/v37sXXrVmSz2YAHx6F7EJrEhlxgOrMzn89j+/btyOfzxqptnElKyRsO\nHTqEXC6nBkmdrUqQkkoUi0Vks1nFSpdYsLpsAES5JNl0duzw8LAolynRBU06dNa+Lpdph4DO/P3h\nD3/oS4YBQLn1aS+uxCCOxWLIZrOhzwbImQE5613nCoyPj2NiYkKFY4B5A0urwMnJSTz88MOqj7lx\nke7HE4KQEdRBMhw/fhyxWEyFH5LJpFFuwvbt29WEI6xtHsbRsy/yPifMzs7iwIEDGBoawtve9jb1\nvOSi5H1F4NXIuJHlLH9TX0kFZwimpCS19G7ZsmVqoNHfHx3T2c+mjHi1dM+UkVLXfYoj072l/tTL\nsF500UV47bXXxD433a+W7umTQ+LOhOnek08+qXYPEalQaosf03WPZ/2U+p3aaVb3JFsp2QmbanLU\nx5VKxZgMx6F7ELpyNyVY4cSVWjWZ+Z557g3g7m4pTk7bOUZHRzE9PW2M9eiy2bg59SpaJ0+eNJJJ\npLrFxWIR09PTSKfTKBaLAbe1yZ2fSCTw8MMP4x3veAduvvnmwACpryaIlSqRaXj+cGmlIs3c+WqC\nViFU7Yl2S9CAyetcP/PMM8q4AP7CEfx+UmzS5NK877778N73vheJRELFek1y/8M//AOKxSKq1aqP\nuS21Lbmp+SqL97ku6zPPPIOxsTFEIhEVz77rrrtU3+l9ReDVyPiOC1MBIyC88I/eDg9VkfckTO9K\npRJefPFFrF+/PvD+pL6TCJ82umfyrnFSJk3ITbpH/ATqn4mJCTX4mjJgEmx1j/Zu03sP072NGzdi\n+fLlPqKhDr6a57onre5N30MrdE8i8PF3JdUGkLxy1Mf/9t/+W2WvdL1z6C6EDu66YpAySi4raVDR\nB0N9ANWTi0gGTvogdbfe1q1bMTY25mM2m2QLc3PqAzWPt3K5JJmA4KSDM65N7vz9+/ejUqng9OnT\nWLJkCQD/AMk/dhODWEI6nfa50qRtfNzA7ty5U02KyNOhb8/TDWy1WsW73vUuX+Yz3bjw+9Gz8Aki\nTzjC9+NKxkWX27RVDAgaWMnA8T42Gdj9+/fjIx/5CJ577jlcfvnlvuflyWj4s+gGnMtgY2DpHrUK\nzgD+8FmY3qVSKdx2223G98f7zpSwRoKue6bto9JgF6Z7tM2VEI1GkUqlAttHTXoOhOuetA88TPei\n0aiyX4B/ApVOp1VhJQJ/XyY9M+0QaFb3TBU0dRsoZe6TwkZSH/OFmkP3oGZueUqwwl1WqVTKd24q\nlcJXvvIVjI+Pqy0z+uyXJ68AZHe39EHOzc2hUCgoAh9nNofJRm65HTt2GGXj21y4XCYjMTQ0hKmp\nKYyPj2Pr1q0BuUzu/JdffhnXXXcdlixZotJCSsl96INMp9P49V//9cCWPt6fwPysXW+rVmIS6fl4\nGEAKo3ieh1Qq5as1bbrf9u3bfUaAkmFIJVgpPCEZF11uU0hI6ispURDvY+m3/jynT59GsVhUWd9M\nctMxXmhIkkF3GfPkKQSbgjM8fBamd/RuPM8LPE8mk1F9S5CK2djonikxCz9eS/f090zfrxSmMt2v\nlu5J76+W7gHm0CPPGy+9L65nxWIxYE/1Z2pG94g4qoM8jTQpkIoISeEY3sdS2NaheyAO7lLRCIoR\nTkxMoFQqBfKXT0xM4N3vfrcaNGmmS5D2iQLzRi2bzfqO8Q+Sx9O4cTHJRnJFo1Ekk0mjbHRMTyvL\n5eIykVzlctkoFxCMvdIxKi8qfbS6kTUZjbByudK9ag34+vPRao2Mm2RgAZkLwO+XSqWQSqV8EwAp\n5scNqsm48Bij3hbvq7DiJ9ygmgys/jx8lSTJTcficX81Mm5kbQys1MeSsc7n84G4apje6TsL+PPw\nyY00WbLRPdOEyjTYmXRPR9g5pslTLd2T3p+N7vEJFK+MZ9I9Sc88zwtwL/gzNap70nUSP4FPXmy+\nbUnvHLoH4uAeVj1IUhZSjEsvvTT0ZjSAzs3NqUGCkzakD5J/6CZmM5fNlFKUQHJQ+tmenh5RLkkm\nW7n01Yz0wUv9CZwzsmQ0qNgEeS96enoC1fmkFaNp5h42AEurNYJUrEPvB34/crfySlJ8ZTkzM6Pc\nnnSdZFw4UcxUvnbr1q1Kj/v6+kQ95mU39d8m48xTukoDAZHzEomECinx92xjYKU+1tsZHx8PpCcm\n1NI7gv48fCXGJ0v16B7XA5MMtrqnyyadY5r01NI9rnd0nY3u6Z4XHno06Z5pINdJdq3WPU7gkyZO\nXIfCvu0DBw4Y9c6heyAO7jRTpP9LW0mAoHF58cUXVW71sbExlEqlwAAaiUTw4IMPIhKJKGOhQ/og\nOePTxGwm2STlIzauLhuXwySXJJOtXPpqxvTB849WN2RUbCKRmE9rSUb00KFDmJycVKUyAXnFqA8G\nJqPBn09arXG5ANkQ80HMtBLjnhGb60y1AvQa5bqBJf3N5XLiBIobPf23/q6OHz+O1157TczdLw0E\nkruUt29jYIFzxTqkdqrVqiq5yScvYXr385//XJUU1Z+Hr8R4qKAe3ePv06T7trqnyyadI02SbXTP\n9jque7RXXe9zfRdBmO5xvatUKvjCF76gPGKt1D2pbLfEoOcTfknvqK+q1aqyc44t370IJdRJxBxS\nRACK2MaZow8++CCGhoZUso5kMuljoCaTSeVyBuZX0LR/WyLw8H27JmYzySbJRffTZaPrSY53v/vd\nolwmUpFJLtNebZJn586diMfjOHz4MK666qpAeVC+S4GKUXCmrOd5vhUHTS70ewHn9vXqRkNPHKI/\nn4nYJcklEYj0+wEyGVDaZkN6o1cV49fZEDClSlUSZ0LXY+m3/q5oYiUxpCXdkNylvK8kcpJEftLz\nSvB2qGASMD8g0IAVpnd33nknLr30Ut97PXz4sJoM6+QvXmWvHt3L5/M+PeAy0D1q6R7PcZ/NZtUW\nOq6fPL+Bre5JlRRr6Z5EjJN2EXDd43oGBAl8rdI907cs5UHg1fi43vE+phAdT3ft0D0IHdwltrmu\niLRy0FeJwLxi7Ny5E5lMRhzIv/GNb2Dv3r0qvvjKK6/43MicSc3Tv5oYpyTb3/3d3xnl0mXjcoyN\njWHTpk0BuSSZwuTS92oDwVUanUfpQoH5j1Yq1gEEB1WaQfNtKBLLVx8M6OPmBlY3ZhJzVpLLZDik\nAZ9XMTNlIeOcBs4elvRR2lnADSwQZOdL6VIlmRKJBHp7e5XXg5IbEaSBQE86RO+P95WNgaVrqVgH\ngNBypC+88IKV3vX09CgPTrVaxYkTJ8RJOhBkt9vqnmniJ72vMN2jVSJd97a3vQ1vvvlmgHhoup+N\n7kmMbxvd4++dWPz6BAPw6550f/IISO+rGd2rVqti/XgpnTbf8qjrnZ4aXO9j0zfj0B2omaGOZynS\nZ4Z8m4qekY4URxrIP//5zwf2dROkGSPFjkZHR9XfpK0jJFs6nTbKpcvG5bjtttvwnve8JyCXaRZr\nkov2agPmfaCPPPIIrrnmGt9HS8Q/bmRttj/pz0/PbhqAqf2nn35apelNpVJ44oknxIpsNInT5ZIM\nsXQ/aVuYtLLl51F9A874lvIN6LogGViJaMgNqsnAUvuZTEalF56dnVXbkKTKhtxQUl/pRtbGwJLs\ntFWsVr6Iu+++W/3bpHf33Xcf3njjDZTLZfWudLIkX4k1onth3h/qz4svvlhVcQzTPT3Hved5+MVf\n/EVks1nfOab72eiedE6Y7t1///3W711vX78/17NoNIrdu3cHFg/UV83onu5RIUh5ELgXgBeJkfpY\n+o4dugehg7vkwpqZmUFPTw+i0agauIH5GTKtdOfm5nDrrbdiampKHMj139wI8Rnj+Pg4du/ejbm5\nOd/qRao4RbI98cQTOH78eEAuAD7ZdKPD5dB/S7NYW7mkVUGlUsGGDRtUTI4+2mg0Kqblldy1evyb\nDAlfMUqDCmH79u2qpjXh1KlTxqpYulzPPvusaIj54EN/NxX00NNf8vMSifn8/5RohYwZEZhMBk4y\nsNIKgxtUk4GtVCrYsWOH0gMeFyUeSZi7NBKJBIysjYGV+oWSmZTL5dCUnya945McvQ2T56de3TOt\nGAkUhrLVPZLt5MmTePTRR3Hrrbf6zpH0Tuo7Sfekc2ZmZkJ1z/a9S30q6Zk0waDjzeqebe0D7gXg\n/bJ169ZAH1Nfuph7dyJ0cA9zIZXLZXzyk5/0GYdiseib4U1OTvomBtJqghshfcb4h3/4h+jp6cEX\nv/hFTExMqHuZ3PIk2+bNm333oRV42IqCy6H/5rNYW7kkI0j9mkwmlSx6H0qJWaRKZ1L2QD6wSYMK\nQS+5SbjmmmvgecGKbNxDYzLEkUgEMzMzvsGHVoXxeO2CHvp5lUpFcTSoApVuzEwGzjYGzg2qycDq\n30AikVAGkPpUWu2aquXp35KNgSXZ4/FzxTqi0ahYlVFHmN7x71m/n9RPjeqeSe/0VbGN7un94Hke\nVq1ahZmZGd85kt7p14TpnnROLd0zeTmk8J2Np8BUQa8Vuie9UynEwItAcb0z9bFD92LJ5z73uc+Z\n/njs2DGsXbsW1WoV69atAzBvOB5++GGsW7cOxeJ83XIyrN/97ndx/fXXY2ZmBmvXrsXQ0BAuvvhi\n1d6RI0d8+Zrn5uawbt06/OhHP8Lq1asBQK26jxw5gt7eXvzO7/wODh48iGKxiEKhgGQyiVKphJGR\nEVx55ZU+xSbZzpw5g8svv1zJRazQ3t5eJdvg4KCSjcvBf+syPffcc7jnnnvwvve9r6ZcY2Nj6Ovr\nw5kzZ3DFFVf4+nXNmjUA5vN8j46OwvM8tb+d/qNnm52dxdq1a/H6669jamoKR44cweHDh7F8+XKU\nSiVs3LgRAHDmzBnlBaBrn3jiCZw5cwbj4+PqvEqlotygq1evxo4dO/Dtb38b09PTanvNtm3b1CBB\nBoL+S6VSiMViOHLkCPbs2eOrmKXL6nkejh07hu985ztYs2YNbrzxRiXnunXr1P0BYO3atZicnEQq\nlUIikfDJPTw8jFgshueeew4bN26E53lYvXp14Bjp2MjIiE+mtWvXolAo+PSRv9OZmRnfb+qrY8eO\n4ec//znm5uZw5ZVXYt26dTh69KjSY13OFStWYMWKFThy5AjWrFmDo0eP+tqhd75u3TqxD6ampvCx\nj33Mp9PveMc71Hat4eFh3/dnIjLV0rtSqYQVK1Zg+fLlvudZt24dVq9erb71ZnRP0jt6P57nKVtQ\nS/cIumzSOVzv6Hlr6R7XO/5OJd0zfVee5wXqoXPd43q3ceNGHDt2zKev1P+t0D1J90ulElKpFPbs\n2aNk58/D9c7Uxw7dC3FwL5VK2L9/P4aHh/HNb34TV155JcbGxnD8+HHceOONWLt2LdasWYNMJqMG\nUCKezc7OYmJiAsuWLUM+n1fKwwfMfD6PJ598Url5dSPIFd3zPCSTSfT09GD16tXqIxodHcXg4CB2\n7Njhk+0973mPmqnu3LlTtb1s2TIsW7YMr776Kg4ePIhkMumT4+DBgzhz5ozvd9jHZ5LrS1/6EpYs\nWSIaQWD+gx8fH8e73/1uRCKRwEcrvQ8y6GfOnFEz79WrV2PZsmW+gY0mCrqB0wcVOk83sGTwqtUq\nKpUKJicncfz4cRw7dswoE8nFjSwffGjA4uek02ns3btX9ctTTz2FeDyOXC6HZDIpyi0ZVH5MMrC8\nbUnPYrGY+I7Xrl2Lhx56CNdeey0OHTqk4s/SoB2NRvHWW29h+fLliMfjvnfDjayNgQWAv/7rv0Y0\nGlXvolQq+fRPx/j4uHHw1WU4ffo0Lr30Uhw8eDDwPLV0L5FIIJVK1dQ96f1Re8D8QFGv7uXz80Wa\npHOkSY+N7pl0o5buSefk83mMjo6ip6dH3U/Sa65n69atwyOPPILh4WFfX7VC96Tno5K1GzduxPj4\nOK644grVVk9PD6644oqA3pn62KF7Ibrl9+7di2Qyie3bt+M3fuM3MDIyoljvwLmtLrqrJxKJYOvW\nrWJBA+6qLxaLGBwcVDEcPRmGHpPSc8qTS1+KdYfJxt3tnP3LiWEUL9SJYiaZTHJddtllIjGOkM/P\nV7Z77LHHsHr1arXiMSGTyWBychKZTAZjY2OBuHxvb6+RYMPj8CakUin8/d//PdLptJFtrcsP+Ldc\njY2NKYKUvrXP5HLkhVW4O12Ky0suRolEyPkRkqt+eHhYvTtOftTdyfl8Htu2bcPk5GQghAL4OShh\n7lLuttb5LPr5vIiQlChJInsBQWa5pHfxeBwzMzPqOWq5V3XdI+jymrgPJr3jdRwAe90jr4B+jknv\nALO7W9c9k24A4bqnv3cCD99J7Ut6B8gciVboXi0eQLVaFYtAmb59fcugQ3dDHNxpsOzt7UUkEgld\nvQFBklo+n1fGa3x8PDCQU9IWUigiZAHm2GckEjHGum3kogpitfbybtmyJfDbJJNJLr7/mNc7jsfj\nvhrSNvErMiJSIY0wgo1EMOMGlt7VvffeC0CuiqVDMrIzMzMiX8AU0+T9yWOFUkyayFq60eVGVjKw\npsI7YTwLAjeOUkycx8BNFbVMRCeTgaW+1vfD62QvzkLXmeXS/mPJ0NsUJJLIbSR7JpMRdU/SOyCY\nxbEe3ePPR/0h6R1gp3u2+RS47pneu0TO4+1zPTNxJFqheyYeBzH2h4eHxdwFUh6GsImlQ/dBdMvf\nd999KBQK2LdvHyYmJpDP5xGLxfDKK68odzu5ayT3uh6bOXLkSMBVf9ddd+Hiiy9GLBbDFVdc4XPv\nSTEpcun39/djy5YtgVh3Pp8PyDY7O+tzr9PsVnIpjo+P+2KU/LckEwCjXFNTU7jpppuwevXqwL0A\nvwsTQMCVzkHGnkIlPC5vij8Dchy+VCphZmYGr776KjZu3BhwOZripQSKfz733HO44YYbFE/h7Nmz\nOHv2LM6cOYNnnnkGyWQSxWJRkbLIcEn9qbtOPc8LxEZ5+ITk4nFAve29e/diw4YNYgxc71Nyd+u/\n9Xf1Z3/2Z+jr68Pp06fFdnbt2uXjf5hCIdxtTXyWrVu3qu9h3bp1Pn3Zs2ePz6MccIMAABucSURB\nVMU+NTWFG2+8MXTCHRbvHh0dxVtvvYV3vetdYr+E6Z7ktiaeCNc9Se8AqPDTddddh2XLltWte/z5\nrrzySp/eHT16FBdddBFWrFhhpXsUTiO9I9lr6Z7pvZPLvVQq4aqrrgr0sdSfJo5EM7pnuj+d853v\nfEddR3bkoYcewq/8yq+Iekfy6OE8h+6GOLinUils2rQJmzZtwuDgoCpy8uqrr2LPnj248sorMTw8\njPHxcZFApyvdBz/4QXEgj8ViyOfzmJub8xkzHpM6duxYYPLAY93JZDIgGyfPJRIJxGIx/OQnP8EV\nV1wBz/PwF3/xF3j11VdVjHLfvn146623AjFLKU4mffAk19vf/nasW7cuQIybmppSpJezZ8+iVCrh\n0ksvrWlgTR8/h2QYpTg8N7B8MDLFS8PuVy6XsWzZMjWRoux7UvxZ6k9O8uOxUf19ZjIZLFu2DEDQ\nwOltnzx5Eu985ztFGXif8t/6u1qyZAl+/OMf47rrrgu043ke3ve+9/kGsIMHD+Ls2bOIRqOYm5sT\niU42BhaYX2kmk0l1T5vBj78//VnWrFmDQqFgfDc2uicR77hckt4B83HkY8eOoVAo+GK99eiefs70\n9LRP70qlEuLxuJH7wHVvzZo1Pr0jnaqle9J712Pny5cvF2XQ+/Po0aMiR6IVuvev//qvWL16tdgH\nEhcB8BOeud7RdbUmlg7dA3Fw54VW6De5YvVZtkSg8zwP119/vTIGpoFcX3XNzc1h+fLlvtl2LBYz\nsu937tzpi39z2STyHAB885vfxKpVq/Dkk0/iJz/5CW666Sa1mn/llVfw3ve+V/3+6U9/iv7+fnEF\nYGLec7l0HDt2TH3wS5cuxfLly1WdZsnA1iJISe2T0XvjjTeMzG1uYPXByPM8kWFb634f/OAH1Yo7\nk8ngbW97G5LJJF5//XVVBpfegUQg4qsCGrSvu+46eJ6nJiGvv/46Xn31VaOB09veuHFjYFXJ+zRs\nQkfviiaGV111FY4dO4bjx48r3ZZIVLpeL1u2TDTyNgYWmDfWlUpFXWcz+PHBV3+WeDyuOCErVqwI\nPE+zuhe2YgSAb33rWz6iViO6p59D+SJI786cOaPeu43uSatRG92T3rv+/rg3TdppYPImtkL3li9f\njmQyKXrzTJ4z4JxHgesdXVdrYunQPQit524CxYzuuusuDAwMiDXJy+UyPvWpT6nfJrezXkRhZmbG\nV1iCEqRMTExgenraV6hCio/y41KVraGhIZTLZbz88su4/fbb4Xke5ubm4Hke7r77bszNzanf11xz\nDZ599lmxqpipRKdJLuBcQZ5SqYSbbroJv/Irv6LiWrxWPBBeGESCXuEqrLLf6OioryIZfzd6vDQM\nUlUqAL5iMqbKfFJ+g0wmY8zsBcxPEnO5nG9/bj6fx/bt20MTB+nFNHifvvOd7xT7WH9Xnje/TVGq\n5Mb3BxNIr6V2SCZewY8X5gGC70aqYlYLJhmk5yHUWxyE5Lr00kuNegcEiVqN6J7pHF7EyEb3uN7R\nsVq6J7136f2R7j3++OOB/qRiNERK1EnBrdA9/f68Xj0thnQiqt5Xkr0mD5PLSLc4ELrPXQLf0qav\nyt966y1cfvnlABBwSUvxIX2F/z//5//ERRddhNnZWRw7dgzPPvssfvM3f1N06Uv743XZ/uZv/sbn\nbtdXHYVCAcuWLUNfXx+OHTuGK664wreap3jxqlWr8OUvfxmvv/46LrnkEiWTTkzhsXmTXBz6DHhy\ncjKw5YRw9OhRHD9+XG37CXPdA/4Vo7RqJRw7dgzLli1T25H4uzHFSzn0bTYXXXQRjh8/rvZOP/LI\nIzhz5oz4DqS4oxR75asLaZXFV8C1YsQAfH36/e9/H5FIxNjHuly08qZ+o79zOSV3qbSViocmJBc4\nfzfN7DWWZODPQ6Cw0ne/+92A/kiwWTECwHPPPed7xkZ0Tz9nbm7Op3dHjx7FK6+8YvQ6cP0wba+s\npXvSOdL7I9277LLLsHHjRrE/TSviVuie5OkhL8TRo0exfPlysa8ke829LA7djZq55XXwLW16drHe\n3l585jOfwVNPPQXgXNpXfctNKpUKrKrK5TL+4A/+QP1tcnISyWTSN4PU2fcmGfTjP/zhD/H+97/f\ntyWIZsfAvCEqlUqKyayv5jdv3oxIJIJyuYzXX38dfX19KqMcyVRPIRUJtjNgiRkvQWcP8wpXEnOb\nM2F5QQgTu5xDZ/MeOnQInuepPnjttdeM27KkDGe8T2jlYMocpj+fzlCmvPd625xRr/epXlBH6mNd\nLolBLMkpbcfjz0elaUulEmKxmFiVTXo3QLD6mS24DNLzcNQqDmLSPdN1/Bkb0T39HK53lMbZVvek\nb9FG9/Rz/vEf/9H4/rjumbJ+SvagFbonbdnTd8uYtk9Kekd2Piw7okP3wHpwl7a06XjggQfw8Y9/\n3EcG0gvCAOcG8i9+8YuqxKvnedi2bRvS6bQ6RvnUCaRUjz/+OD7xiU+IMuj71f/wD/8QK1euVFvz\nPM/D7t271XaPSCSCJUuWqGvJIAwNDaktc9FoFJ/4xCfw5ptvor+/3yeT9EFIhVRM0D+4UqlkNLB8\ni5xptqzX2eZ7fSXDYZqUPPzww7j55ptVMQ+9HdNzkJHVc3rr9ybMzc0Zq94Bwb3F0n5gk2H0PA/F\nYhHHjx/HqVOnAm3rRok/Dx+weB/r70pPJELQB2mp2IvUDhA0sFJeBGnCGIvFAiVDbcFlMBWF4f0R\nNhE16Z7punQ6jUQigWg0WrfujY+PA4DvHP39mfbuh+metKfdRvf0c6gErTSZ4AOi1C+mHO2t0D1p\nQNa3I9J1+vZJ00LlpZdewgsvvOD2ui8SWA/u09PTAOS96aZVObnQ+UB+9913KyNKK/xqtaqMNC8D\nC8xPHqrVqorBcxmk/eq7d+9WMsdiMVQqFfUB7d+/P7Caf/bZZ3H33Xcjm81idnYWX/7yl3HDDTcE\nZKKPUZ88UNYn/f4m6B+cjYGtBb3Otp48CJCTbegGVi8uc8stt+Cqq65S54V5CgAYDbFpv3oqlTIm\n95H2FteqegecM8S0Z5dKfHLZm1nt0ruSiqhISVmk/c7cyOoG1lQwSCqIQnXPe3t7Q71DtZ4FkIvC\ncNQqDmLSPUnvqB0iuW3atKku3aN7mM4J269O6WW57kl6R/fSIekenVMrv4Cue/UUW2mF7iUSCVH3\naQeUfh0N6lKRGAC4/PLLcfnll/s4Lw7dC+vBPWwAAc6tyr/61a/6jksDuankq3QM8E8ewmTg2ed0\nd/vZs2dRLpexe/duzM7OIp1Oo1qtolwuq9X8T37yE9/vT3ziEz7vAPdE8MnDzMxMwCVsAn1wNgbW\nBjq3oZYrTTewmUzGt3rhq20JkpE1ZSrTi1SQwQGCyX14MQspsYe0yiJ5ABjbNtX6tsW9996LDRs2\n4KqrrsLk5KSvbZ6UBZDdpUBwgkEGdmJiQhy0IpGIb4UIzOvg9PQ0+vv76x7cSYbVq1crHeXP0wgk\n3TO5cPVvlJ4PsNM9aUUeVv5U1ynqKz4AS0VUbHRPOsdUVW8hdU8qVx2NRlEsFtXf6TqavEh6BzSv\new6dRV0xdx7/llblHLVKvtK1OqTqcfrkQZcBkGPggN/dfuDAAWzdujVQ6pVW3ACwcuVK3+ped91L\nMtHk4cknn8TGjRtFl7CEdDqNV199FVNTU3jnO9/ZEgML+A1nWByeT4L46sW0miFIRjYsS14qlVIp\njMNWXul0GiMjIyrLIZ8oSXKRkX3qqafw0Y9+NNB2PVwICel0Gh/96EexZMkSRZLicvNVkOQu5UZW\nMrAE/j3oq8hYLBZgctfzLKlUCrt37w7lGTQC0qmnn35aHdP1zvSN1qt7/Jpbb73VqHfAOd2jtkwD\nsH6Oje7Z8Ea6Qfck/kcsFsOjjz6KzZs3q+skjwL3XjSjew6dR12DO+CfkUur8lqGyoZlSUplmjzw\nVQGPgZ86dUrN/Ik899xzz+Gzn/2scoOSrJQ3fnZ2FuVyGdVq1be65zLpoMnD+vXrsWLFCjWYhhlL\n+uCB+QnL+9///prX2EI3nJTSVI+FVioV0cDy1Yu0mpGgGzzakjM3N4fp6Wll9PQc6hxcF77+9a9j\n2bJlRnKSJBcZWc5Apn+ban3bgBtnqiymryqlQVpyl3IjKxlYCXyFSLFqwC5tsfQsP//5z608NPWA\n3teRI0fEGLyJuNWI7kkrcq53gKx7knudcs2HcULonlNTU0b95ANkN+ieif8xNDQkpjAmSJ6JRnXP\nYWFQ9+AOnHMvmtzrzUDPB3/vvfcaJw9SHItW5VdffTUOHDjgI8/xWFixWFSrZT0nvv5bkmlgYCAQ\nq7/00kvxkY98JBCDlz4c+uBpGxRNXlphYHUXfyaTCYQwZmdnjcSfVCqF0dFRJTNf8UiQDPHIyIgx\nh7rEtibQ3uKvfe1roUV3aJVlqhXAQau5crlcdx3qVCqFH/3oR3jppZewdOlScVUpDdKSu5QbWRsD\nCwRXkbaTAulZOOHTZpVsC133yLWrh87oGyU91/W9Ed3j13C9A2Td4wOwnmu+VuGddDqNXbt2BXbi\nmNANuifxPzKZjOiJ1CF5JhrVPYeFQd2Du4k8B9itysPAt7mZJg8mGSgGTtXKaLa7f/9+0YPw2c9+\nFnv37lWr+bGxMWzatMm3ur/rrrsCrjUeqyf3PY/BS8aJE/9mZmaaNrD6ZEN38fNYKM3AybA+9thj\najWfy+VUe2GrbQ5uZMO284SBjFCxWMT999+vmNM8XEGrLNrva2NkGzVKkUgEF198scoFQEWU9MmM\naZCWCEvcyIYZWNMq0nZSID0LJ3zaemjCYNI9IMj/MLHjbXVPj6/r1wByFTMb3dPfS09Pj0qrLRHj\nvv71r+PDH/4wPM8znsOx0Lpn4n+Ykm2FeS8a1T2HhUFDK3cTea4ZSFvtaFVL0D8kSQaKgT/yyCO4\n/vrrcd999wXc6zr45OG2227De97zHvW7WCyit7c3sOJJJBK+WP3+/ft996f98iboMe94PN60gQ2L\nd/P4um5gf/zjH6uwAHBugKy12g4zsnw/rokEx0FG6LrrrsNbb72ljIf+LLyiV1jlPR3NGKWXX34Z\nsVhMbXfiq0ppFWQiLHEjG5bN0LQ9Lh6PN0zO4roABCdn9cKke6Z70bdkYseH6R6fVOu6Ie0Dt9E9\n6b1IMnDds/VILbTuSfwPyeVOCPNeNKN7Dp2H9eBuQ55rBmFb7WxloBj4jTfeqD7aWslkdOgDu/53\nvuLZsWOHL1ZPkwe+X950b+mjbcbA6iQY/rHzGbhuYD/96U+rAUa/rtaKJ8zI1iJMmSCxc0kWgr7K\nmp6exic/+Un1t7CVeyqVQj6fb2grXDqdxsDAACKRiHFVKRGP+GqNG9kwA0uQJi/SoGkL0jsKWfX2\n9qp30yhMuiet/HQ39SWXXKLera3u9fT0oFKpqP/rbUv3s9E9aU+7JAP3vNh6pBZa90z8D5MOhW3r\na0b3HDoP68HdhjzXDGpttQuTgcfAR0dHccMNN/jc6/qe+XrBlZrq3esxeE7gC5vNp9NpLF26FJde\nemlLDKwO/rHzrXamOKB+nWmPMiHMyPL71eP6DVvJAv5VFv1b/5sJptriNtBjxZOTk4HVmjRIS6s1\nbmTDDCxgzk6oTwzrBQ0Wzz77LMrlsi+rWyugvz/TFk9p4mOre1I2urD72eoe1zuJOc5X+Lb71Rda\n9/QYfJjLnUPyTDSjew6dh/Xg3g7yHAffamcrA19J1tqfbgvTx/CNb3zDF6t/4YUXsH379kAM3oTh\n4WGMjo4CmPdQtMrA6h97KpUyxkK5geVGwrRHmRBmZKUVON9qVEt2kxGUVlk20NPk1gueCIWv1qRB\nWnKXkpH93ve+h2g0WtPAmnad6Ku5ekGDxYc+9CGfa7cVoPf3ta99Db/2a78GQN5DzwefenRPyoKo\nP5vk+amlezZ6x9tfbLpHMXiyNTY8Fckz0YzuOXQe1oN7I1vaGkHYx22SgcfA+eDaqKym+JM0yeDZ\n78Kgf7T6/tNmoX/sYXF4bmAlIxGW0S3MyALBlRCR4GxlD0MjhmVmZgb33XdfTRkk6Ks/z/NQrVbR\n09NTcxUkyZlMJnH69GlrIqAEzjqvB5Kb2qaGgA3o/elV4XS903kaANT/69U9giS31Oe1dM9W70zt\n18JC654eg6+VSU+H5JloRvccOo+GCHXtRr2pQk0x8GZhSjPLVwFjY2N13V//aMvlctMGVvrY9SIX\nNJPnBvbAgQPiKrKerFpcdr4S4kQkG9lNsF1lcUhZuupBKpXCvffei/7+frVrgoiQ0iBt2iNMGb4u\nueQSALWJgBKa2b4muampbG+j4O8PODeZ1leQfLJ5+PBhcb+4je4Vi0WkUinfManPw3SvHr0ztW+D\ndumeaQXO5ZTCILZkQI5Wbp10aD+6bnAP22pnAo+Btxq1trg1cn8i0SUSiaYNrIlZTaC+5Ab2hRde\nCFxXT1YtycjylZC0Bawe2cPatoWUpasW9Pz5uVwO73jHOxCPx1VCklwuZ2Trh+0RjsfjTWWHa3b7\nmk7elMhW9SKMXa1/w5x4d9VVVwWus9U9ybsn9XmY7tXa087RTbqnZ+bkK3Aup8T/sCUDcrRi66RD\n59B1gztQ/1Y7HgNvlkDHUWuLm+39+UcLyGzWRmAaaEwrjp07d2L9+vWB66hoRCwWq5lVSzeyppWQ\naZ+tjeyEeldZHFKWrlowFSkhIqJEeLPZI0zvg9BIyMgmyYsOSe8AOdlOvTC5esNWunpOef26SqVi\nrXuU2jasz8N0z9ZF3Y26R9+HvgIntz2XU4rB11O8hqNe3XNYOHTN4N7MVrt2k/1qbXGzvb/00c7O\nzjZtYE3MasC84qB4KL9ON7w2nggysqYVeC0iUpjshHpW9xJqeQ8kmPLnk4uT94vneRgbGzPKaZMV\nzAb1JBgihFVTI89Rs+CuXpPe8UFfKqUKhOteOp1GLBZTaZZNK3AbElwtF3W9K3yp/VbrHkFfgdt6\nUJpBI7rnsHDomsG9ma127SL78S12pi1utveXPlpKCduMgZWY1WErDjKwfJVMctvG1nQjC5hX4GHG\nxbZmvW3CGgk23gMTOPmxVkKSMDlbYWQbiZdKegcEOSXNgAaaEydOGHOvA8HJJncR2+ienmpWqkeu\nw2Z7ZZiLuh4Smqn9dumetAKvx4NSLxqN1TssDLpmcO/EVrt6YUoz2yykgiut5guErXZrxQ9tY2u6\nkY3H4+LqsBXGxWZ1H4ZGtzEB/r6Ix+OhCUnC5GyVkW00Xgo0VoHNFjTQUGImrnemySYfoGx0T3p2\naeCx3V6p39+ERge2TumeSc5GuQISmtE9h86jawb3Tm21qwemNLPNQv9o28lA5atIU1U4CTZZ8/SP\n3LQCb4VxsV3dh6GZVTPFGfv6+kITkoTJ2SojOzQ0pLZE1YtGq//VA9NKtx73di3dk1LNSgNPtwxs\nndA9LmctD0ojaEb3HDqPrhncuxHt2mIHnDNg7WKgSqvIWsVWTMQrEyQjq7fVDBGplWhm1azHGYnJ\nXQ8/opX90Gx9cECuwNZM6mMTTLF0k3u7Ht2rVW61HbrXKAmt07pXy4PSKFqhew6dhRvcQ9DqLXZT\nU1MAEEg52w4DK60ix8bGasaE63F/hxnPZklwrUQzKzh9kEokEnXnJGhlPzRaH9ykd1RrvRHWdC2Y\nVrom93Y9umdKbUtolgTXSiyU7jXLFeBopja9w8LADe4haPUWOyltazsNLEet2LU+0Nswu2sZ2WZI\ncK1AK1Zw+iBF25rqRav6gXYy1Mu6N6UL1muatxqmla5p0K9H92r1QasHtkbQLbrXKhJco7rnsHBw\ng3sIWk3yC0vb2olKS/XErm3ihGEfe7MkuFagFatmPc6YzWbrNpTt6Id6Y7hheteq9LO2sHFv2zyf\nzTkLye5uhfegWd0DWk+Cc3nlFw+WfO5zn/vcQgvRrYjFYqG/G0WpVMLo6KgaXEdHR7Fu3bquIBEC\nUKuM/fv3I5lMhp575MgR8RwyKvTfQjzb0aNHcfz4cRw/fhzHjh3D0aNHsXHjRuvr8/k8nnzySVSr\nVRw8eBCe52Ht2rWoVqtYt26dVRut7od63o10Lekd/VcoFPDBD36wKZlaCZvns+2DY8eOYc2aNQBg\n/b5ahXXr1iEWi+G73/0uNm7cWPd7b4XukRyrV69uyfM3o3sOncdFCy3AhQpKOZvJZBCLxTrusg5D\nPp9Hb29vzcmMThbqRtAOhIGBAd8qyhapVMqX9nN4eBi5XE6tPhcCtu/GBJ7qeCHj0RJsns+2D4aH\nh5FIJBb0fZH3oN5w3vmoew6dhXPLLwB4ytluMbD1xglbudWoHWh2Cx1l8dPTfi6UW7IVMVyud61M\ncNIsbJ6vm3Zg2KJRt/j5pnsOnYcb3BcAek7vbjKwtjHqC+lj16tqLeS7agV/gOeS76bJmU2MuptY\n8LZoJo97N+neYut3Bze4Lxgo5ezevXu7xsACdszubtrm1m7oVbUALOi7agXrPpFI4PXXX295gpNm\nYcNw7wYWfCfRLbp3ofX7+QI3uC8AhoeHMT4+jmKxiFOnTnWNga2H2b3Q29w6hVQqhT//8z/H+vXr\n8eKLLy7Yu2oF654b6W6cnNkw3C+UHOfdonuEC6Xfzxe4wX0BsGPHDuRyOfT19amPpBsMrG2Muhu2\nuXUK+Xwel112GX72s5+pSoUL8bytSMGrpzqm4kLdthKziVFfKDnOSfd++tOfLqjuES6Ufj9f0FOt\nVqsLLcSFhnw+j+9973tYuXKlKtdoqjjnsLAoFAooFos4fPgw1q9fv6jfFdUU7+3tVYVjksmkW4l1\nKSjpUD6fR6VSwS233LLQIjksIritcAuEZDKJU6dONbxNy6EzmJmZwdjYGPr7+8+Ld5VKpVTudrcS\n627MzMxg+/btABrfUudw4cK55RcAtFK65JJLlEu0m1yjDudQqVTw5S9/2Xdssb4rnuq4GSa3Q/uh\np5ydnZ11EzGHuuAG9wVAO3N6O7QW5xOJyOnd4oKke24i5mAL55ZfIGzfvt3laV4E8DwPvb29582K\nyend4sH5pnsOnYUb3BcA3Zhy1kFGN6T9bBWc3i0unE+659B5uMF9gXA+bx8733A+rXSd3i0unE+6\n59BZuMF9AdDtBVcczuF8elfn07NcCHDvy6EZuMF9AeCqKy0enE/v6nx6lgsB7n05NAPHlu8gLqSC\nK4sd59O7Op+e5UKAe18OrYAb3DuIC6ngymLH+fSuXFWvxQX3vhxaAZd+toMYGxtDT08PACzqNKYX\nAs7Hd0W55B0WB9z7cmgGbuXeQVxIBVcWO87Hd3U+JeS5EODel0MzcIN7B9GKyl4OncH5+K5cLvnF\nBfe+HJqBc8s7/H/t2zENAAAAgKD+re3hoISXAMxY4QBgRtwBYEbcAWBG3AFgRtwBYEbcAWAmk5PB\nwcOHxBMAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f83743c84d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_adjacency_matrix(np.mean(control_networks, axis=2), outfolder + 'control_adjmat.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Creating files for visualization with BrainNet Viewer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "control_network = np.mean(control_networks, axis=2)\n",
    "control_network = control_network[17:, :]\n",
    "control_network = control_network[:, 17:]\n",
    "np.savetxt(outfolder + 'control_mean.txt', control_network)\n",
    "#np.savetxt(outfolder + 'ZDHHC9_mean.txt', np.mean(ZDHHC9_networks, axis=2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 721,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([2], dtype=int32)"
      ]
     },
     "execution_count": 721,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.unique(roi_data)[1:2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def make_BrainNet_file(filename, network, out_file, labels_file):\n",
    "    import bct\n",
    "    import nibabel as nib\n",
    "    import numpy as np\n",
    "    import pandas as pd\n",
    "\n",
    "    # Loading the data\n",
    "    img = nib.load(filename)\n",
    "    data = img.get_data()\n",
    "    affine = img.affine\n",
    "    labels = sorted(pd.read_csv(labels_file)['label_number'].values)\n",
    "\n",
    "    # Calculating the centroids - the crucial part is to apply the affine to get from voxel to real-world coordinates\n",
    "    get_centroid = lambda label: nib.affines.apply_affine(affine, np.mean(np.where(data == label), axis=1))\n",
    "    centroids = np.asarray([get_centroid(label) for label in labels])\n",
    "\n",
    "    # This creates a .node file to use with BrainNet viewer\n",
    "    degrees = bct.degrees_und(network)\n",
    "\n",
    "    df = pd.DataFrame({'x': np.round(centroids[:, 0], 2),\n",
    "                     'y': np.round(centroids[:, 1], 2),\n",
    "                     'z': np.round(centroids[:, 2], 2),\n",
    "                     'colour': degrees,\n",
    "                     'size': degrees})\n",
    "\n",
    "    df['label'] = labels\n",
    "    df[['x','y','z','colour','size', 'label']].to_csv(out_file, header=None, index=False, sep='\\t')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "filename =  '/home/jb07/Desktop/temp/FreeSurfer/CBU_MNI152/mri/aparc+aseg.nii.gz'\n",
    "out_file = '/home/jb07/Desktop/temp/nodes.node'\n",
    "labels_file = '/home/jb07/Desktop/temp/labels.csv'\n",
    "network = np.mean(ZDHHC9_networks, axis=2)\n",
    "network = network[1:,:]\n",
    "network = network[:,1:]\n",
    "np.savetxt(outfolder + '/ZDHHC9.edge', network)\n",
    "make_BrainNet_file(filename, network, out_file, labels_file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "filename =  '/home/jb07/Desktop/temp/FreeSurfer/CBU_MNI152/mri/aparc+aseg.nii.gz'\n",
    "out_file = '/home/jb07/Desktop/temp/nodes.node'\n",
    "labels_file = '/home/jb07/Desktop/temp/labels.csv'\n",
    "network = np.mean(control_networks, axis=2)\n",
    "network = network[1:,:]\n",
    "network = network[:,1:]\n",
    "np.savetxt(outfolder + '/control.edge', network)\n",
    "make_BrainNet_file(filename, network, out_file, labels_file)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Comparison of edge weight by region"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "d = []\n",
    "threshold_range = np.arange(0,100,10)\n",
    "\n",
    "for subject in subject_list:\n",
    "    for threshold in threshold_range:\n",
    "        \n",
    "        # Loading the adjacency matrix\n",
    "        network = load_data(folder, 'CSD', str(threshold), '_FA', subject)\n",
    "        \n",
    "        # Getting the labels for the matrix\n",
    "        labels = pd.read_csv('/home/jb07/joe_python/GitHub/ZDHHC9_connectome/DK_labels.txt', header=None)[0].values\n",
    "        \n",
    "        subcortical_labels = [np.where(labels == label)[0][0] for label in labels if re.search('(Left|Right)', label)]\n",
    "        lh_labels = [np.where(labels == label)[0][0] for label in labels if re.search('lh', label)]\n",
    "        rh_labels = [np.where(labels == label)[0][0] for label in labels if re.search('rh', label)]\n",
    "\n",
    "        mean_connection_strength = list()\n",
    "        \n",
    "        # Selecting subcortical, left and right cortical, and cortical interhemispheric connections\n",
    "        for labels in [(subcortical_labels, subcortical_labels), (lh_labels, lh_labels), (rh_labels, rh_labels), (lh_labels, rh_labels)]:\n",
    "            subconnections = network.copy()\n",
    "            subconnections = subconnections[labels[0], ...]\n",
    "            subconnections = subconnections[..., labels[1]]\n",
    "            mean_connection_strength.append(np.mean(subconnections))\n",
    "        \n",
    "        # Setting group variable\n",
    "        group = 'ZDHHC9' if re.search('z', subject) else 'control'\n",
    "        \n",
    "        d.append({'ID': subject,\n",
    "                  'group': group,\n",
    "                 'threshold': threshold,\n",
    "                 'subcortical': mean_connection_strength[0],\n",
    "                 'lh': mean_connection_strength[1],\n",
    "                 'rh': mean_connection_strength[2],\n",
    "                 'interhemispheric':mean_connection_strength[3]})\n",
    "        \n",
    "df = pd.DataFrame(d)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fc571146e50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/imaging/local/software/anaconda/2.4.1/2/lib/python2.7/site-packages/matplotlib/figure.py:1718: UserWarning: This figure includes Axes that are not compatible with tight_layout, so its results might be incorrect.\n",
      "  warnings.warn(\"This figure includes Axes that are not \"\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAagAAAEYCAYAAAAJeGK1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nERGFEhsUERGFEhsUERGFEhsUERGF0v8BiggjG6K0SsgAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fc571111610>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(single_column, single_column), dpi=600)\n",
    "fig,axs = plt.subplots(2,2, sharex=True, sharey=True)\n",
    "ordering = [[1,0],[0,0], [0,1], [1,1]]\n",
    "titles = ['subcortical', 'left hemisphere', 'right hemisphere', 'interhemispheric']\n",
    "counter = 0\n",
    "\n",
    "for measure in ['subcortical', 'lh', 'rh', 'interhemispheric']:\n",
    "    for group in ['control', 'ZDHHC9']:\n",
    "\n",
    "        copy_df = df.copy()\n",
    "        copy_df = copy_df[df['group'] == group]\n",
    "        median = copy_df.groupby('threshold').median()[measure].values\n",
    "        mad = copy_df.groupby('threshold').agg(robust.mad)[measure].values\n",
    "        colour = '#56B4E9' if group == 'control' else '#D55E00'\n",
    "        marker = 'o' if measure == 'clustering_coefficient' else 'D'\n",
    "        \n",
    "        axs[ordering[counter][0], ordering[counter][1]].errorbar(x = threshold_range,\n",
    "                     y = median,\n",
    "                     yerr = mad,\n",
    "                     marker = marker,\n",
    "                     markersize = 2,\n",
    "                     color=colour,\n",
    "                     linestyle='solid',\n",
    "                     linewidth=1)\n",
    "        axs[ordering[counter][0], ordering[counter][1]].set_title(titles[counter])\n",
    "    counter += 1\n",
    "\n",
    "    if measure == 'subcortical':\n",
    "        plt.legend(['control', r'\\textit{ZDHHC9}'], loc='best')\n",
    "        plt.xlabel('streamline threshold')\n",
    "        plt.ylabel('mean node strength')\n",
    "        \n",
    "plt.savefig(outfolder + '/regional_node_strength.png', dpi=600)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "subcortical\n",
      "\n",
      "ZDHHC9 AUC: mean=0.08; SE=0.011\n",
      "ZDHHC9 AUC: median=0.07; mad=0.033\n",
      "control AUC: mean=0.18; SE=0.018\n",
      "control AUC: median=0.18; mad=0.062\n",
      "Wilcoxon: W=0.0; p=0.018\n",
      "Paired t-test: t(6)=5.32; p=0.002 corrected-p=0.007\n",
      "\n",
      "lh\n",
      "\n",
      "ZDHHC9 AUC: mean=0.19; SE=0.012\n",
      "ZDHHC9 AUC: median=0.18; mad=0.034\n",
      "control AUC: mean=0.33; SE=0.012\n",
      "control AUC: median=0.31; mad=0.014\n",
      "Wilcoxon: W=0.0; p=0.018\n",
      "Paired t-test: t(6)=11.68; p=0.0 corrected-p=0.0\n",
      "\n",
      "rh\n",
      "\n",
      "ZDHHC9 AUC: mean=0.16; SE=0.01\n",
      "ZDHHC9 AUC: median=0.18; mad=0.015\n",
      "control AUC: mean=0.3; SE=0.022\n",
      "control AUC: median=0.31; mad=0.094\n",
      "Wilcoxon: W=0.0; p=0.018\n",
      "Paired t-test: t(6)=7.02; p=0.0 corrected-p=0.002\n",
      "\n",
      "interhemispheric\n",
      "\n",
      "ZDHHC9 AUC: mean=0.01; SE=0.004\n",
      "ZDHHC9 AUC: median=0.01; mad=0.008\n",
      "control AUC: mean=0.06; SE=0.005\n",
      "control AUC: median=0.06; mad=0.012\n",
      "Wilcoxon: W=0.0; p=0.018\n",
      "Paired t-test: t(6)=8.76; p=0.0 corrected-p=0.0\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/jb07/python_modules/scipy/stats/morestats.py:2388: UserWarning: Warning: sample size too small for normal approximation.\n",
      "  warnings.warn(\"Warning: sample size too small for normal approximation.\")\n"
     ]
    }
   ],
   "source": [
    "for measure in ['subcortical', 'lh', 'rh', 'interhemispheric']:\n",
    "    control_area = [np.cumsum(df[df['ID'] == subject][measure].values)[-1] for subject in subject_list if re.search('c', subject)]\n",
    "    ZDHHC9_area = [np.cumsum(df[df['ID'] == subject][measure].values)[-1] for subject in subject_list if re.search('z', subject)]\n",
    "    print measure + '\\n'\n",
    "    print('ZDHHC9 AUC: mean=' + str(round(np.mean(ZDHHC9_area), 2)) +\n",
    "         '; SE=' + str(round(sem(ZDHHC9_area), 3)))\n",
    "    print('ZDHHC9 AUC: median=' + str(round(np.median(ZDHHC9_area), 2)) +\n",
    "         '; mad=' + str(round(robust.mad(ZDHHC9_area), 3)))\n",
    "    print('control AUC: mean=' + str(round(np.mean(control_area), 2)) +\n",
    "         '; SE=' + str(round(sem(control_area), 3)))\n",
    "    print('control AUC: median=' + str(round(np.median(control_area), 2)) +\n",
    "         '; mad=' + str(round(robust.mad(control_area), 3)))\n",
    "    print('Wilcoxon: W=' + \n",
    "    str(wilcoxon(control_area, ZDHHC9_area)[0]) + '; p=' +\n",
    "    str(round(wilcoxon(control_area, ZDHHC9_area)[1], 3)))\n",
    "    print('Paired t-test: t(' + str(len(control_area) -1) + ')=' + \n",
    "    str(round(ttest_rel(control_area, ZDHHC9_area)[0], 2)) + '; p=' +\n",
    "    str(round(ttest_rel(control_area, ZDHHC9_area)[1], 3)) + ' corrected-p=' + \n",
    "    str(round(4*ttest_rel(control_area, ZDHHC9_area)[1], 3)) + '\\n')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Creating a $Latex$ version of the table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\\begin{tabular}{lrrrrrrrrrrr}\n",
      "\\toprule\n",
      "          measure &  ZDHHC9\\_mean &  ZDHHC9\\_SE &  ZDHHC9\\_W &  ZDHHC9\\_p &  control\\_mean &  control\\_SE &  control\\_W &  control\\_p &  ttest stat &  ttest p &  ttest corrp \\\\\n",
      "\\midrule\n",
      "      subcortical &         0.08 &      0.011 &      0.97 &      0.92 &          0.18 &       0.018 &       0.93 &       0.54 &        5.32 &    0.002 &        0.007 \\\\\n",
      "               lh &         0.19 &      0.012 &      0.98 &      0.95 &          0.33 &       0.012 &       0.82 &       0.07 &       11.68 &    0.000 &        0.000 \\\\\n",
      "               rh &         0.16 &      0.010 &      0.82 &      0.06 &          0.30 &       0.022 &       0.90 &       0.36 &        7.02 &    0.000 &        0.002 \\\\\n",
      " interhemispheric &         0.01 &      0.004 &      0.84 &      0.11 &          0.06 &       0.005 &       0.91 &       0.38 &        8.76 &    0.000 &        0.000 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n"
     ]
    }
   ],
   "source": [
    "d = []\n",
    "for measure in ['subcortical', 'lh', 'rh', 'interhemispheric']:\n",
    "    control_area = [np.cumsum(df[df['ID'] == subject][measure].values)[-1] for subject in subject_list if re.search('c', subject)]\n",
    "    ZDHHC9_area = [np.cumsum(df[df['ID'] == subject][measure].values)[-1] for subject in subject_list if re.search('z', subject)]\n",
    "\n",
    "    d.append({\n",
    "    'measure': measure,\n",
    "    'ZDHHC9_mean': round(np.mean(ZDHHC9_area), 2),\n",
    "    'ZDHHC9_SE': round(sem(ZDHHC9_area), 3),\n",
    "    'ZDHHC9_W': round(shapiro(ZDHHC9_area)[0], 2), \n",
    "    'ZDHHC9_p': round(shapiro(ZDHHC9_area)[1], 2),\n",
    "    'control_mean': round(np.mean(control_area), 2),\n",
    "    'control_SE': round(sem(control_area), 3),\n",
    "    'control_W': round(shapiro(control_area)[0], 2), \n",
    "    'control_p': round(shapiro(control_area)[1], 2), \n",
    "    'ttest stat': round(ttest_rel(control_area, ZDHHC9_area)[0], 2),\n",
    "    'ttest p': round(ttest_rel(control_area, ZDHHC9_area)[1], 3),\n",
    "    'ttest corrp': round(4*ttest_rel(control_area, ZDHHC9_area)[1], 3)\n",
    "    })\n",
    "    \n",
    "result_df = pd.DataFrame(d)\n",
    "print(result_df[['measure', 'ZDHHC9_mean', 'ZDHHC9_SE', 'ZDHHC9_W', 'ZDHHC9_p', 'control_mean', 'control_SE', 'control_W', 'control_p', 'ttest stat', 'ttest p', 'ttest corrp']].to_latex(index=False))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Graph theory analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Correlation of graph metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "d = []\n",
    "\n",
    "for subject in subject_list:\n",
    "    network = load_data(folder, 'CSD', str(threshold), '_FA', subject)\n",
    "    network[np.isnan(network)] = 0 # removing NaN\n",
    "\n",
    "    # Calculating graph metrics\n",
    "    d.append({'subject': subject,\n",
    "             'degree': np.mean(bct.degrees_und(network)),\n",
    "             'strength': np.mean(bct.strengths_und(network)),\n",
    "             'clustering_coefficient': np.mean(bct.clustering_coef_wu(network)),\n",
    "             'global_efficiency': np.mean(bct.charpath(bct.distance_wei(bct.weight_conversion(network, 'lengths'))[0])[1])})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\\begin{tabular}{lrrrr}\n",
      "\\toprule\n",
      "{} &  degree &  strength &  clustering\\_coefficient &  global\\_efficiency \\\\\n",
      "\\midrule\n",
      "degree                 &    1.00 &      0.99 &                    0.86 &               0.96 \\\\\n",
      "strength               &    0.99 &      1.00 &                    0.89 &               0.98 \\\\\n",
      "clustering\\_coefficient &    0.86 &      0.89 &                    1.00 &               0.94 \\\\\n",
      "global\\_efficiency      &    0.96 &      0.98 &                    0.94 &               1.00 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(pd.DataFrame(d)[['degree', 'strength', 'clustering_coefficient', 'global_efficiency']].corr().round(2).to_latex())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1. Mean node degree & strength"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Because the node degree has a large influence on all other graph measures, differences in mean node degree are investigated first. The following section calculates the average node degree in the network for participants in the *ZDHHC9* and control group over a range of streamline thresholds. The groups are compared on the area under the curve of mean node degree across thresholds. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "threshold_range = np.arange(0,100,10)\n",
    "d = []\n",
    "\n",
    "for threshold in threshold_range:\n",
    "    for subject in subject_list:\n",
    "        network = load_data(folder, 'CSD', str(threshold), '_FA', subject)\n",
    "        network[np.isnan(network)] = 0 # removing NaN\n",
    "        \n",
    "        # Calculating graph metrics\n",
    "        mean_degree = np.mean(bct.degrees_und(network))\n",
    "        mean_strength = np.mean(bct.strengths_und(network))\n",
    "        \n",
    "        # Setting group variable\n",
    "        group = 'ZDHHC9' if re.search('z', subject) else 'control'\n",
    "\n",
    "        # Collecting the data\n",
    "        d.append({'ID':subject, \n",
    "                  'threshold': threshold,\n",
    "                  'group': group,\n",
    "                  'mean_degree': mean_degree,\n",
    "                  'mean_strength': mean_strength})\n",
    "        \n",
    "df = pd.DataFrame(d)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mean_degree\n",
      "\n",
      "ZDHHC9 AUC: mean=22.23; SE=0.921\n",
      "ZDHHC9 AUC: median=21.98; mad=2.163\n",
      "control AUC: mean=33.18; SE=1.056\n",
      "control AUC: median=32.94; mad=3.384\n",
      "Wilcoxon: W=0.0; p=0.018\n",
      "Paired t-test: t(6)=10.14; p=0.0 corrected-p=0.0\n",
      "\n",
      "mean_strength\n",
      "\n",
      "ZDHHC9 AUC: mean=9.57; SE=0.487\n",
      "ZDHHC9 AUC: median=9.86; mad=1.648\n",
      "control AUC: mean=16.34; SE=0.645\n",
      "control AUC: median=15.64; mad=1.785\n",
      "Wilcoxon: W=0.0; p=0.018\n",
      "Paired t-test: t(6)=13.08; p=0.0 corrected-p=0.0\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for measure in ['mean_degree', 'mean_strength']:\n",
    "    control_area = [np.cumsum(df[df['ID'] == subject][measure].values)[-1] for subject in subject_list if re.search('c', subject)]\n",
    "    ZDHHC9_area = [np.cumsum(df[df['ID'] == subject][measure].values)[-1] for subject in subject_list if re.search('z', subject)]\n",
    "    print measure + '\\n'\n",
    "    print('ZDHHC9 AUC: mean=' + str(round(np.mean(ZDHHC9_area), 2)) +\n",
    "         '; SE=' + str(round(sem(ZDHHC9_area), 3)))\n",
    "    print('ZDHHC9 AUC: median=' + str(round(np.median(ZDHHC9_area), 2)) +\n",
    "         '; mad=' + str(round(robust.mad(ZDHHC9_area), 3)))\n",
    "    print('control AUC: mean=' + str(round(np.mean(control_area), 2)) +\n",
    "         '; SE=' + str(round(sem(control_area), 3)))\n",
    "    print('control AUC: median=' + str(round(np.median(control_area), 2)) +\n",
    "         '; mad=' + str(round(robust.mad(control_area), 3)))\n",
    "    print('Wilcoxon: W=' + \n",
    "    str(wilcoxon(control_area, ZDHHC9_area)[0]) + '; p=' +\n",
    "    str(round(wilcoxon(control_area, ZDHHC9_area)[1], 3)))\n",
    "    print('Paired t-test: t(' + str(len(control_area) -1) + ')=' + \n",
    "    str(round(ttest_rel(control_area, ZDHHC9_area)[0], 2)) + '; p=' +\n",
    "    str(round(ttest_rel(control_area, ZDHHC9_area)[1], 3)) + ' corrected-p=' + \n",
    "    str(round(4*ttest_rel(control_area, ZDHHC9_area)[1], 3)) + '\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\\begin{tabular}{lrrrrrrrrrrr}\n",
      "\\toprule\n",
      "       measure &  ZDHHC9\\_mean &  ZDHHC9\\_SE &  ZDHHC9\\_W &  ZDHHC9\\_p &  control\\_mean &  control\\_SE &  control\\_W &  control\\_p &  ttest stat &  ttest p &  ttest corrp \\\\\n",
      "\\midrule\n",
      "   mean\\_degree &        22.23 &      0.921 &      0.96 &      0.85 &         33.18 &       1.056 &        0.9 &       0.34 &       10.14 &        0 &            0 \\\\\n",
      " mean\\_strength &         9.57 &      0.487 &      0.97 &      0.86 &         16.34 &       0.645 &        0.9 &       0.31 &       13.08 &        0 &            0 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n"
     ]
    }
   ],
   "source": [
    "d = []\n",
    "for measure in ['mean_degree', 'mean_strength']:\n",
    "    control_area = [np.cumsum(df[df['ID'] == subject][measure].values)[-1] for subject in subject_list if re.search('c', subject)]\n",
    "    ZDHHC9_area = [np.cumsum(df[df['ID'] == subject][measure].values)[-1] for subject in subject_list if re.search('z', subject)]\n",
    "\n",
    "    d.append({\n",
    "    'measure': measure,\n",
    "    'ZDHHC9_mean': round(np.mean(ZDHHC9_area), 2),\n",
    "    'ZDHHC9_SE': round(sem(ZDHHC9_area), 3),\n",
    "    'ZDHHC9_W': round(shapiro(ZDHHC9_area)[0], 2), \n",
    "    'ZDHHC9_p': round(shapiro(ZDHHC9_area)[1], 2),\n",
    "    'control_mean': round(np.mean(control_area), 2),\n",
    "    'control_SE': round(sem(control_area), 3),\n",
    "    'control_W': round(shapiro(control_area)[0], 2), \n",
    "    'control_p': round(shapiro(control_area)[1], 2), \n",
    "    'ttest stat': round(ttest_rel(control_area, ZDHHC9_area)[0], 2),\n",
    "    'ttest p': round(ttest_rel(control_area, ZDHHC9_area)[1], 3),\n",
    "    'ttest corrp': round(4*ttest_rel(control_area, ZDHHC9_area)[1], 3)\n",
    "    })\n",
    "    \n",
    "result_df = pd.DataFrame(d)\n",
    "print(result_df[['measure', 'ZDHHC9_mean', 'ZDHHC9_SE', 'ZDHHC9_W', 'ZDHHC9_p', 'control_mean', 'control_SE', 'control_W', 'control_p', 'ttest stat', 'ttest p', 'ttest corrp']].to_latex(index=False))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAOwAAADsCAYAAAB300oUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fc5713d67d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAOwAAADsCAYAAAB300oUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRA75//LKD6+76CT6AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fc5710137d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "labels = ['mean node degree', 'mean node strength']\n",
    "counter = 0\n",
    "\n",
    "for measure in ['mean_degree','mean_strength']:\n",
    "    fig = plt.figure(figsize=(single_column, single_column), dpi=600)\n",
    "\n",
    "    for group in ['control', 'ZDHHC9']:\n",
    "\n",
    "        copy_df = df.copy()\n",
    "        copy_df = copy_df[df['group'] == group]\n",
    "        median = copy_df.groupby('threshold').median()[measure].values\n",
    "        mad = copy_df.groupby('threshold').agg(robust.mad)[measure].values\n",
    "        colour = '#56B4E9' if group == 'control' else '#D55E00'\n",
    "        marker = 'o' if measure == 'clustering_coefficient' else 'D'\n",
    "\n",
    "        plt.errorbar(x = threshold_range,\n",
    "                     y = median,\n",
    "                     yerr = mad,\n",
    "                     marker = marker,\n",
    "                     markersize = 2,\n",
    "                     color=colour,\n",
    "                     linestyle='solid',\n",
    "                     linewidth=1)\n",
    "\n",
    "    if counter == 0:\n",
    "        plt.legend(['control', r'\\textit{ZDHHC9}'], loc='best')\n",
    "        plt.xlabel('streamline threshold')\n",
    "    plt.ylabel(labels[counter])\n",
    "    plt.savefig(outfolder + '/' + labels[counter] + '.png', dpi=600)\n",
    "    plt.show()\n",
    "    counter += 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3. Clustering coefficient & Global efficiency\n",
    "The mean node degree analysis indicated significant differences between the groups. Differences in $D_{G}$ may influence other graph measures. Consensus thresholding is applied to equalize the number of connections in both groups"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def consensus_thresholding(in_matrices, percentage_threshold):\n",
    "    # ===============================================================#\n",
    "    # Consensus thresholding of connectivity matrices \n",
    "    # for method see: http://www.ncbi.nlm.nih.gov/pubmed/23296185\n",
    "    # ===============================================================#\n",
    "    \"\"\"\n",
    "    inputs:\n",
    "    in_matrices: numpy array with connectivity matrices (dimensions: ROIs x ROIs x participants)\n",
    "    percentage_threshold: ratio of connections to be retained, e.g. 0.6\n",
    "\n",
    "    outputs:\n",
    "    matrices only retained connections that occured in the percentage threshold of participants\n",
    "    \"\"\"\n",
    "\n",
    "    import bct\n",
    "    import numpy as np\n",
    "    \n",
    "    connection_consensus = np.sum(np.asarray(in_matrices > 0), 2).astype('float64')/in_matrices.shape[2]\n",
    "    connection_consensus = bct.binarize(bct.threshold_absolute(W=connection_consensus, thr=percentage_threshold))\n",
    "    consensus_matrices = in_matrices * np.reshape(np.repeat(connection_consensus,in_matrices.shape[2]), newshape=in_matrices.shape)\n",
    "    \n",
    "    return consensus_matrices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "d = []\n",
    "\n",
    "for threshold in np.arange(0,100,10):\n",
    "    \n",
    "    # Consensus thresholding to equalise the graph density between groups\n",
    "    networks = np.rollaxis(np.asarray([load_data(folder, 'CSD', str(threshold), '_FA', subject) for subject in subject_list]), 0, 3)\n",
    "    networks_before = networks.copy()\n",
    "    networks = consensus_thresholding(networks, 1)\n",
    "    \n",
    "    # Calculating graph measures in the matched networks\n",
    "    for i in range(0,len(subject_list)):\n",
    "        subject = subject_list[i]\n",
    "\n",
    "        # Calculating graph metrics\n",
    "        mean_degree = np.mean(bct.degrees_und(networks[...,i]))\n",
    "        clustering_coefficient = np.mean(bct.clustering_coef_wu(networks[...,i]))\n",
    "        global_efficiency = bct.charpath(bct.distance_wei(bct.weight_conversion(networks[...,i], 'lengths'))[0])[1]\n",
    "        clustering_coefficient_before = np.mean(bct.clustering_coef_wu(networks_before[...,i]))\n",
    "        global_efficiency_before = bct.charpath(bct.distance_wei(bct.weight_conversion(networks_before[...,i], 'lengths'))[0])[1]\n",
    "        \n",
    "        # Setting group variable\n",
    "        group = 'ZDHHC9' if re.search('z', subject) else 'control'\n",
    "\n",
    "        # Collecting the data\n",
    "        d.append({'ID':subject, \n",
    "                  'group': group,\n",
    "                  'threshold': threshold,\n",
    "                  'mean_degree': mean_degree,\n",
    "                  'clustering_coefficient_before': clustering_coefficient_before,\n",
    "                  'clustering_coefficient': clustering_coefficient,\n",
    "                  'global_efficiency_before': global_efficiency_before,\n",
    "                  'global_efficiency': global_efficiency})\n",
    "\n",
    "df = pd.DataFrame(d)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAOwAAADsCAYAAAB300oUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nKxDEEEKwAkEMIQQrEMQQ/w824Snzc4al8wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fc57514fd10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "labels = ['global clustering coefficient', 'global efficiency']\n",
    "counter = 0\n",
    "\n",
    "fig = plt.figure(figsize=(single_column, single_column), dpi=600)\n",
    "\n",
    "#for measure in ['clustering_coefficient', 'clustering_coefficient_before']:\n",
    "for measure in ['global_efficiency', 'global_efficiency_before']:\n",
    "\n",
    "    for group in ['control', 'ZDHHC9']:\n",
    "\n",
    "        copy_df = df.copy()\n",
    "        copy_df = copy_df[df['group'] == group]\n",
    "        median = copy_df.groupby('threshold').median()[measure].values\n",
    "        mad = copy_df.groupby('threshold').agg(robust.mad)[measure].values\n",
    "        colour = '#56B4E9' if group == 'control' else '#D55E00'\n",
    "        marker = 'o' if measure == 'clustering_coefficient' else 'D'\n",
    "        linetype = 'dashed' if counter == 1 else 'solid'\n",
    "        \n",
    "        plt.errorbar(x = threshold_range,\n",
    "                     y = median,\n",
    "                     yerr = mad,\n",
    "                     marker = marker,\n",
    "                     markersize = 2,\n",
    "                     color=colour,\n",
    "                     linestyle=linetype,\n",
    "                     linewidth=1)\n",
    "        counter += 1\n",
    "        \n",
    "plt.xlabel('streamline threshold')\n",
    "plt.ylabel(labels[1])\n",
    "plt.savefig(outfolder + labels[1] + '.png', dpi=600)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "clustering_coefficient\n",
      "\n",
      "ZDHHC9 AUC: mean=0.2; SE=0.003\n",
      "ZDHHC9 AUC: median=0.2; mad=0.007\n",
      "control AUC: mean=0.21; SE=0.003\n",
      "control AUC: median=0.21; mad=0.006\n",
      "Wilcoxon: W=0.0; p=0.018\n",
      "Paired t-test: t(6)=5.91; p=0.001 corrected-p=0.004\n",
      "\n",
      "global_efficiency\n",
      "\n",
      "ZDHHC9 AUC: mean=0.18; SE=0.003\n",
      "ZDHHC9 AUC: median=0.18; mad=0.005\n",
      "control AUC: mean=0.2; SE=0.002\n",
      "control AUC: median=0.2; mad=0.009\n",
      "Wilcoxon: W=0.0; p=0.018\n",
      "Paired t-test: t(6)=5.81; p=0.001 corrected-p=0.005\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for measure in ['clustering_coefficient', 'global_efficiency']:\n",
    "    control_area = [np.cumsum(df[df['ID'] == subject][measure].values)[-1] for subject in subject_list if re.search('c', subject)]\n",
    "    ZDHHC9_area = [np.cumsum(df[df['ID'] == subject][measure].values)[-1] for subject in subject_list if re.search('z', subject)]\n",
    "    print measure + '\\n'\n",
    "    print('ZDHHC9 AUC: mean=' + str(round(np.mean(ZDHHC9_area), 2)) +\n",
    "         '; SE=' + str(round(sem(ZDHHC9_area), 3)))\n",
    "    print('ZDHHC9 AUC: median=' + str(round(np.median(ZDHHC9_area), 2)) +\n",
    "         '; mad=' + str(round(robust.mad(ZDHHC9_area), 3)))\n",
    "    print('control AUC: mean=' + str(round(np.mean(control_area), 2)) +\n",
    "         '; SE=' + str(round(sem(control_area), 3)))\n",
    "    print('control AUC: median=' + str(round(np.median(control_area), 2)) +\n",
    "         '; mad=' + str(round(robust.mad(control_area), 3)))\n",
    "    print('Wilcoxon: W=' + \n",
    "    str(wilcoxon(control_area, ZDHHC9_area)[0]) + '; p=' +\n",
    "    str(round(wilcoxon(control_area, ZDHHC9_area)[1], 3)))\n",
    "    print('Paired t-test: t(' + str(len(control_area) -1) + ')=' + \n",
    "    str(round(ttest_rel(control_area, ZDHHC9_area)[0], 2)) + '; p=' +\n",
    "    str(round(ttest_rel(control_area, ZDHHC9_area)[1], 3)) + ' corrected-p=' + \n",
    "    str(round(4*ttest_rel(control_area, ZDHHC9_area)[1], 3)) + '\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\\begin{tabular}{lrrrrrrrrrrr}\n",
      "\\toprule\n",
      "                measure &  ZDHHC9\\_mean &  ZDHHC9\\_SE &  ZDHHC9\\_W &  ZDHHC9\\_p &  control\\_mean &  control\\_SE &  control\\_W &  control\\_p &  ttest stat &  ttest p &  ttest corrp \\\\\n",
      "\\midrule\n",
      " clustering\\_coefficient &         0.20 &      0.003 &      0.90 &      0.34 &          0.21 &       0.003 &       0.95 &       0.75 &        5.91 &    0.001 &        0.004 \\\\\n",
      "      global\\_efficiency &         0.18 &      0.003 &      0.85 &      0.12 &          0.20 &       0.002 &       0.90 &       0.32 &        5.81 &    0.001 &        0.005 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n"
     ]
    }
   ],
   "source": [
    "d = []\n",
    "for measure in ['clustering_coefficient', 'global_efficiency']:\n",
    "    control_area = [np.cumsum(df[df['ID'] == subject][measure].values)[-1] for subject in subject_list if re.search('c', subject)]\n",
    "    ZDHHC9_area = [np.cumsum(df[df['ID'] == subject][measure].values)[-1] for subject in subject_list if re.search('z', subject)]\n",
    "\n",
    "    d.append({\n",
    "    'measure': measure,\n",
    "    'ZDHHC9_mean': round(np.mean(ZDHHC9_area), 2),\n",
    "    'ZDHHC9_SE': round(sem(ZDHHC9_area), 3),\n",
    "    'ZDHHC9_W': round(shapiro(ZDHHC9_area)[0], 2), \n",
    "    'ZDHHC9_p': round(shapiro(ZDHHC9_area)[1], 2),\n",
    "    'control_mean': round(np.mean(control_area), 2),\n",
    "    'control_SE': round(sem(control_area), 3),\n",
    "    'control_W': round(shapiro(control_area)[0], 2), \n",
    "    'control_p': round(shapiro(control_area)[1], 2), \n",
    "    'ttest stat': round(ttest_rel(control_area, ZDHHC9_area)[0], 2),\n",
    "    'ttest p': round(ttest_rel(control_area, ZDHHC9_area)[1], 3),\n",
    "    'ttest corrp': round(4*ttest_rel(control_area, ZDHHC9_area)[1], 3)\n",
    "    })\n",
    "    \n",
    "result_df = pd.DataFrame(d)\n",
    "print(result_df[['measure', 'ZDHHC9_mean', 'ZDHHC9_SE', 'ZDHHC9_W', 'ZDHHC9_p', 'control_mean', 'control_SE', 'control_W', 'control_p', 'ttest stat', 'ttest p', 'ttest corrp']].to_latex(index=False))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Comparison of topographic variation in graph metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from statsmodels.sandbox.stats.multicomp import multipletests"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/jb07/python_modules/ipykernel/__main__.py:55: VisibleDeprecationWarning: converting an array with ndim > 0 to an index will result in an error in the future\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-142-048ea8a4ad48>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[0;32m     19\u001b[0m                 \u001b[0mnewdata\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnewdata\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbct\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_components\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnetwork\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mno_depend\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mTrue\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     20\u001b[0m                 \u001b[0meccentricity\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mzeros\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnetwork\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 21\u001b[1;33m                 \u001b[0meccentricity\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mbct\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_components\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnetwork\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mno_depend\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mTrue\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mbct\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcharpath\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mbct\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdistance_wei\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mbct\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mweight_conversion\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnewdata\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m'lengths'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     22\u001b[0m                 \u001b[0mgraph_metrics\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0meccentricity\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     23\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m/home/jb07/python_modules/bct/bct.pyc\u001b[0m in \u001b[0;36mdistance_wei\u001b[1;34m(G)\u001b[0m\n\u001b[0;32m   1722\u001b[0m                                 \u001b[0mW\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mwhere\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mG1\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mv\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m    \u001b[1;31m#neighbors of shortest nodes\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1723\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1724\u001b[1;33m                                 \u001b[0mtd\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0marray\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mD\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mu\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mW\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mflatten\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mD\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mu\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mv\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m+\u001b[0m\u001b[0mG1\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mv\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mW\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mflatten\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1725\u001b[0m                                 \u001b[0md\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmin\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtd\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0maxis\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1726\u001b[0m                                 \u001b[0mwi\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0margmin\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtd\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0maxis\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "for metric in ['node_degree', 'node_strength', 'clustering_coefficient', 'local_efficiency']:\n",
    "    AUC = list()\n",
    "\n",
    "    for subject in subject_list:\n",
    "        graph_metrics = list()\n",
    "\n",
    "        # Calculating the graph metric across a range of thresholds\n",
    "        for threshold in np.arange(0,100,10):\n",
    "            network = load_data(folder, 'CSD', str(threshold), '_FA', subject)\n",
    "            if metric == 'node_degree':\n",
    "                graph_metrics.append(bct.degrees_und(network))\n",
    "            elif metric == 'node_strength':\n",
    "                graph_metrics.append(bct.strengths_und(network))\n",
    "            elif metric == 'clustering_coefficient':\n",
    "                graph_metrics.append(bct.clustering_coef_wu(network))\n",
    "            elif metric == 'local_efficiency':\n",
    "                # This requires to identify the fully connected subcomponent - otherwise some distances are infinite\n",
    "                newdata = network[bct.get_components(network, no_depend=True)[0] == 1, :] \n",
    "                newdata = newdata[:, bct.get_components(network, no_depend=True)[0] == 1] \n",
    "                eccentricity = np.zeros(len(network))\n",
    "                eccentricity[bct.get_components(network, no_depend=True)[0] == 1] = bct.charpath(bct.distance_wei(bct.weight_conversion(newdata, 'lengths'))[0])[2]\n",
    "                graph_metrics.append(eccentricity)\n",
    "            \n",
    "        # calculating the area under the curve across thresholds\n",
    "        AUC.append(np.cumsum(np.asarray(graph_metrics), axis=0)[-1,:])\n",
    "\n",
    "    AUC = np.asarray(AUC)\n",
    "    \n",
    "    pvals = list()\n",
    "    tvals = list()\n",
    "    ZDHHC9_mean = list()\n",
    "    ZDHHC9_SE = list()\n",
    "    ZDHHC9_W = list()\n",
    "    ZDHHC9_p = list()\n",
    "    control_mean = list()\n",
    "    control_SE = list()\n",
    "    control_W= list()\n",
    "    control_p = list()\n",
    "    \n",
    "    for area in np.arange(0,AUC.shape[1]):\n",
    "        tvals.append(ttest_rel(AUC[0:7, area], AUC[7:14, area])[0])\n",
    "        pvals.append(ttest_rel(AUC[0:7, area], AUC[7:14, area])[1])\n",
    "        \n",
    "        ZDHHC9_mean.append(np.mean(AUC[7:14, area]))\n",
    "        ZDHHC9_SE.append(sem(AUC[7:14, area]))\n",
    "        ZDHHC9_W.append(shapiro(AUC[7:14, area])[0])\n",
    "        ZDHHC9_p.append(shapiro(AUC[7:14, area])[1])\n",
    "        \n",
    "        control_mean.append(np.mean(AUC[0:7, area]))\n",
    "        control_SE.append(sem(AUC[0:7, area]))\n",
    "        control_W.append(shapiro(AUC[0:7, area])[0])\n",
    "        control_p.append(shapiro(AUC[0:7, area])[1])\n",
    "\n",
    "    if np.isnan(pvals).any():\n",
    "        pvals[np.where(np.isnan(pvals) == True)[0]] = 1\n",
    "    corrected_pvals = multipletests(pvals, alpha=0.05, method='fdr_bh')[1]\n",
    "    \n",
    "    np.savetxt(outfolder + '/regional_comparison/' +  metric + '_p.txt', pvals)\n",
    "    np.savetxt(outfolder + '/regional_comparison/' +  metric + '_t.txt', tvals)\n",
    "    np.savetxt(outfolder + '/regional_comparison/' +  metric + '_pcorrected.txt', corrected_pvals)\n",
    "    \n",
    "    np.savetxt(outfolder + '/regional_comparison/' + metric + '_ZDHHC9_mean.txt', ZDHHC9_mean)\n",
    "    np.savetxt(outfolder + '/regional_comparison/' + metric + '_ZDHHC9_SE.txt', ZDHHC9_SE)\n",
    "    np.savetxt(outfolder + '/regional_comparison/' + metric + '_ZDHHC9_W.txt', ZDHHC9_W)\n",
    "    np.savetxt(outfolder + '/regional_comparison/' + metric + '_ZDHHC9_p.txt', ZDHHC9_p)\n",
    "    \n",
    "    np.savetxt(outfolder + '/regional_comparison/' + metric + '_control_mean.txt', control_mean)\n",
    "    np.savetxt(outfolder + '/regional_comparison/' + metric + '_control_SE.txt', control_SE)\n",
    "    np.savetxt(outfolder + '/regional_comparison/' + metric + '_control_W.txt', control_W)\n",
    "    np.savetxt(outfolder + '/regional_comparison/' + metric + '_control_p.txt', control_p)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\\begin{tabular}{lrrrrrrrrrrr}\n",
      "\\toprule\n",
      "                  label &  ZDHHC9\\_mean &  ZDHHC9\\_SE &  ZDHHC9\\_W &  ZDHHC9\\_p &  control\\_mean &  control\\_SE &  control\\_W &  control\\_p &  t-value &  p-value &  corrected\\_p \\\\\n",
      "\\midrule\n",
      "           Left-Caudate &        26.71 &      3.220 &      0.87 &     0.193 &         44.57 &       2.861 &       0.88 &      0.241 &     4.14 &    0.006 &        0.031 \\\\\n",
      "          Right-Caudate &        19.29 &      2.661 &      0.83 &     0.088 &         33.86 &       2.586 &       0.92 &      0.472 &     4.85 &    0.003 &        0.020 \\\\\n",
      "          Right-Putamen &        17.71 &      4.927 &      0.91 &     0.392 &         36.86 &       5.938 &       0.92 &      0.493 &     4.34 &    0.005 &        0.028 \\\\\n",
      "    lh-inferiortemporal &        14.29 &      2.466 &      0.90 &     0.336 &         28.57 &       1.850 &       0.86 &      0.156 &     4.33 &    0.005 &        0.028 \\\\\n",
      " lh-medialorbitofrontal &        23.43 &      1.412 &      0.96 &     0.800 &         38.00 &       2.837 &       0.91 &      0.405 &     3.83 &    0.009 &        0.035 \\\\\n",
      "      lh-middletemporal &        10.71 &      1.248 &      0.86 &     0.157 &         18.71 &       1.267 &       0.87 &      0.195 &     4.00 &    0.007 &        0.034 \\\\\n",
      "     lh-parsopercularis &        11.43 &      2.369 &      0.92 &     0.499 &         22.14 &       1.752 &       0.90 &      0.331 &     5.16 &    0.002 &        0.018 \\\\\n",
      "       lh-pericalcarine &        22.00 &      2.000 &      0.82 &     0.070 &         36.14 &       2.198 &       0.97 &      0.913 &     3.87 &    0.008 &        0.035 \\\\\n",
      "          lh-precentral &        60.86 &      4.803 &      0.95 &     0.723 &        107.43 &       4.064 &       0.83 &      0.078 &     7.36 &    0.000 &        0.008 \\\\\n",
      "     lh-superiorfrontal &        76.29 &      5.402 &      0.86 &     0.140 &        135.71 &       6.643 &       0.91 &      0.425 &     8.55 &    0.000 &        0.006 \\\\\n",
      "    lh-superiorparietal &        50.14 &      5.049 &      0.83 &     0.079 &         85.29 &       5.126 &       0.83 &      0.088 &     5.15 &    0.002 &        0.018 \\\\\n",
      "    lh-superiortemporal &        50.57 &      2.125 &      0.95 &     0.728 &         65.29 &       2.222 &       0.93 &      0.569 &     4.39 &    0.005 &        0.028 \\\\\n",
      "             rh-lingual &        13.43 &      2.742 &      0.96 &     0.819 &         34.86 &       3.453 &       0.91 &      0.366 &     6.51 &    0.001 &        0.009 \\\\\n",
      "      rh-middletemporal &        22.00 &      2.370 &      0.93 &     0.525 &         39.14 &       4.636 &       0.96 &      0.800 &     3.75 &    0.010 &        0.035 \\\\\n",
      "     rh-parahippocampal &         7.43 &      1.232 &      0.88 &     0.249 &         20.43 &       1.395 &       0.87 &      0.205 &     7.17 &    0.000 &        0.008 \\\\\n",
      "       rh-parsorbitalis &         7.29 &      0.837 &      0.93 &     0.555 &         13.71 &       1.822 &       0.83 &      0.089 &     3.72 &    0.010 &        0.035 \\\\\n",
      "           rh-precuneus &        54.00 &      3.748 &      0.95 &     0.735 &         98.86 &       6.296 &       0.91 &      0.423 &     9.40 &    0.000 &        0.006 \\\\\n",
      "    rh-superiorparietal &        62.71 &      5.317 &      0.97 &     0.923 &        125.57 &       5.559 &       0.86 &      0.151 &     6.93 &    0.000 &        0.008 \\\\\n",
      "    rh-superiortemporal &        40.57 &      4.418 &      0.82 &     0.063 &         71.29 &       2.917 &       0.96 &      0.844 &     5.31 &    0.002 &        0.018 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n",
      "\\begin{tabular}{lrrrrrrrrrrr}\n",
      "\\toprule\n",
      "                       label &  ZDHHC9\\_mean &  ZDHHC9\\_SE &  ZDHHC9\\_W &  ZDHHC9\\_p &  control\\_mean &  control\\_SE &  control\\_W &  control\\_p &  t-value &  p-value &  corrected\\_p \\\\\n",
      "\\midrule\n",
      "                  Brain-Stem &       117.17 &      5.576 &      0.94 &     0.594 &        144.22 &       5.829 &       0.91 &      0.401 &     4.37 &    0.005 &        0.018 \\\\\n",
      "               Right-Caudate &         9.20 &      1.258 &      0.89 &     0.262 &         17.23 &       1.552 &       0.97 &      0.920 &     4.43 &    0.004 &        0.018 \\\\\n",
      "               Right-Putamen &         7.09 &      2.028 &      0.90 &     0.336 &         16.58 &       3.006 &       0.95 &      0.717 &     4.21 &    0.006 &        0.020 \\\\\n",
      "      lh-caudalmiddlefrontal &         9.15 &      0.583 &      0.97 &     0.910 &         13.81 &       1.210 &       0.99 &      0.989 &     3.48 &    0.013 &        0.035 \\\\\n",
      "                 lh-fusiform &        14.26 &      2.449 &      0.92 &     0.495 &         23.41 &       1.717 &       0.94 &      0.626 &     3.39 &    0.015 &        0.037 \\\\\n",
      "         lh-inferiortemporal &         6.38 &      1.339 &      0.90 &     0.341 &         14.66 &       0.787 &       0.97 &      0.925 &     5.38 &    0.002 &        0.011 \\\\\n",
      "         lh-isthmuscingulate &         5.23 &      1.205 &      0.96 &     0.855 &         14.28 &       1.174 &       0.92 &      0.502 &     4.84 &    0.003 &        0.014 \\\\\n",
      "                  lh-lingual &         8.22 &      0.383 &      0.87 &     0.175 &         13.08 &       0.790 &       0.91 &      0.368 &     6.28 &    0.001 &        0.008 \\\\\n",
      "      lh-medialorbitofrontal &         8.76 &      0.465 &      0.91 &     0.406 &         17.92 &       2.140 &       0.86 &      0.139 &     3.68 &    0.010 &        0.031 \\\\\n",
      "           lh-middletemporal &         4.72 &      0.576 &      0.81 &     0.051 &          9.90 &       0.613 &       0.85 &      0.121 &     5.40 &    0.002 &        0.011 \\\\\n",
      "          lh-parsopercularis &         5.17 &      1.095 &      0.94 &     0.632 &         10.51 &       1.149 &       0.92 &      0.482 &     4.69 &    0.003 &        0.015 \\\\\n",
      "            lh-pericalcarine &        10.18 &      1.085 &      0.82 &     0.058 &         20.07 &       1.291 &       0.94 &      0.618 &     4.99 &    0.002 &        0.014 \\\\\n",
      "               lh-precentral &        30.02 &      2.804 &      0.93 &     0.515 &         57.94 &       2.438 &       0.86 &      0.138 &     9.04 &    0.000 &        0.003 \\\\\n",
      " lh-rostralanteriorcingulate &         6.98 &      1.126 &      0.93 &     0.531 &         15.70 &       1.952 &       0.92 &      0.466 &     4.13 &    0.006 &        0.021 \\\\\n",
      "          lh-superiorfrontal &        36.07 &      3.302 &      0.93 &     0.576 &         70.83 &       3.858 &       0.98 &      0.939 &     9.98 &    0.000 &        0.002 \\\\\n",
      "         lh-superiortemporal &        22.68 &      0.827 &      0.92 &     0.489 &         33.89 &       1.677 &       0.87 &      0.195 &     4.87 &    0.003 &        0.014 \\\\\n",
      "                  rh-lingual &         5.38 &      1.233 &      0.97 &     0.932 &         18.49 &       1.984 &       0.92 &      0.499 &     6.99 &    0.000 &        0.005 \\\\\n",
      "      rh-medialorbitofrontal &         5.76 &      0.739 &      0.93 &     0.561 &         11.10 &       0.872 &       0.99 &      0.990 &     4.39 &    0.005 &        0.018 \\\\\n",
      "           rh-middletemporal &         8.15 &      0.933 &      0.91 &     0.364 &         19.10 &       2.706 &       0.98 &      0.943 &     4.39 &    0.005 &        0.018 \\\\\n",
      "          rh-parahippocampal &         3.27 &      0.613 &      0.84 &     0.100 &         10.74 &       0.728 &       0.89 &      0.288 &     8.32 &    0.000 &        0.003 \\\\\n",
      "              rh-postcentral &         3.99 &      1.483 &      0.86 &     0.168 &         14.66 &       2.370 &       0.93 &      0.528 &     3.36 &    0.015 &        0.037 \\\\\n",
      "                rh-precuneus &        26.26 &      2.302 &      0.94 &     0.637 &         52.14 &       3.591 &       0.96 &      0.807 &    12.42 &    0.000 &        0.001 \\\\\n",
      " rh-rostralanteriorcingulate &         5.60 &      0.727 &      0.95 &     0.761 &         11.87 &       2.223 &       0.82 &      0.069 &     3.17 &    0.019 &        0.044 \\\\\n",
      "         rh-superiorparietal &        29.04 &      2.939 &      0.96 &     0.789 &         64.64 &       3.058 &       0.87 &      0.169 &     7.39 &    0.000 &        0.004 \\\\\n",
      "         rh-superiortemporal &        18.20 &      2.316 &      0.89 &     0.258 &         35.87 &       1.937 &       0.95 &      0.716 &     6.02 &    0.001 &        0.008 \\\\\n",
      "              rh-frontalpole &         2.40 &      0.683 &      0.81 &     0.054 &          4.45 &       0.565 &       0.85 &      0.111 &     3.74 &    0.010 &        0.031 \\\\\n",
      "                   rh-insula &         2.73 &      0.208 &      0.83 &     0.077 &          5.68 &       0.951 &       0.88 &      0.214 &     3.13 &    0.020 &        0.044 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n",
      "\\begin{tabular}{lrrrrrrrrrrr}\n",
      "\\toprule\n",
      "               label &  ZDHHC9\\_mean &  ZDHHC9\\_SE &  ZDHHC9\\_W &  ZDHHC9\\_p &  control\\_mean &  control\\_SE &  control\\_W &  control\\_p &  t-value &  p-value &  corrected\\_p \\\\\n",
      "\\midrule\n",
      " lh-isthmuscingulate &         1.07 &      0.433 &      0.84 &     0.098 &          3.62 &       0.287 &       0.90 &      0.333 &    10.01 &    0.000 &        0.005 \\\\\n",
      "  lh-parsopercularis &         0.75 &      0.157 &      0.90 &     0.341 &          1.81 &       0.162 &       0.91 &      0.409 &     6.39 &    0.001 &        0.027 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n",
      "\\begin{tabular}{lrrrrrrrrrrr}\n",
      "\\toprule\n",
      "                   label &  ZDHHC9\\_mean &  ZDHHC9\\_SE &  ZDHHC9\\_W &  ZDHHC9\\_p &  control\\_mean &  control\\_SE &  control\\_W &  control\\_p &  t-value &  p-value &  corrected\\_p \\\\\n",
      "\\midrule\n",
      "  rh-caudalmiddlefrontal &        76.81 &      5.653 &      0.90 &     0.345 &        120.94 &       4.369 &       0.91 &      0.401 &    10.00 &    0.000 &        0.005 \\\\\n",
      " rh-rostralmiddlefrontal &        83.09 &      4.541 &      0.92 &     0.504 &        124.14 &       6.102 &       0.95 &      0.725 &     6.56 &    0.001 &        0.017 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n"
     ]
    }
   ],
   "source": [
    "labels = pd.read_csv('/home/jb07/joe_python/GitHub/ZDHHC9_connectome/DK_labels.txt', header=None)[0].values\n",
    "\n",
    "for metric in ['node_degree', 'node_strength', 'clustering_coefficient', 'local_efficiency']:\n",
    "    \n",
    "    # Loading all values\n",
    "    ZDHHC9_mean = np.loadtxt(outfolder + '/regional_comparison/' + metric + '_ZDHHC9_mean.txt')\n",
    "    ZDHHC9_SE = np.loadtxt(outfolder + '/regional_comparison/' + metric + '_ZDHHC9_SE.txt')\n",
    "    ZDHHC9_W = np.loadtxt(outfolder + '/regional_comparison/' + metric + '_ZDHHC9_W.txt')\n",
    "    ZDHHC9_p = np.loadtxt(outfolder + '/regional_comparison/' + metric + '_ZDHHC9_p.txt')\n",
    "\n",
    "    control_mean = np.loadtxt(outfolder + '/regional_comparison/' + metric + '_control_mean.txt')\n",
    "    control_SE = np.loadtxt(outfolder + '/regional_comparison/' + metric + '_control_SE.txt')\n",
    "    control_W = np.loadtxt(outfolder + '/regional_comparison/' + metric + '_control_W.txt')\n",
    "    control_p = np.loadtxt(outfolder + '/regional_comparison/' + metric + '_control_p.txt')\n",
    "    \n",
    "    pvals = np.loadtxt(outfolder + '/regional_comparison/' + metric + '_p.txt')\n",
    "    tvals = np.loadtxt(outfolder + '/regional_comparison/' + metric + '_t.txt')\n",
    "    corrected_pvals = np.loadtxt(outfolder + '/regional_comparison/' + metric + '_pcorrected.txt')\n",
    "    \n",
    "    significant_regions = np.where(corrected_pvals < 0.05)[0]\n",
    "    normal_regions = np.where((ZDHHC9_p > 0.05) & (control_p > 0.05))[0]\n",
    "    d = []\n",
    "    \n",
    "    # Collect the results for regions that survive correction for multiple comparisons\n",
    "    for significant_region in significant_regions:\n",
    "        if significant_region in normal_regions:\n",
    "            d.append({'label': labels[significant_region],\n",
    "                      'ZDHHC9_mean': round(ZDHHC9_mean[significant_region], 2),\n",
    "                      'ZDHHC9_SE': round(ZDHHC9_SE[significant_region], 3),\n",
    "                      'ZDHHC9_W': round(ZDHHC9_W[significant_region], 2),\n",
    "                      'ZDHHC9_p': round(ZDHHC9_p[significant_region], 3),\n",
    "                      'control_mean': round(control_mean[significant_region], 2),\n",
    "                      'control_SE': round(control_SE[significant_region], 3),\n",
    "                      'control_W': round(control_W[significant_region], 2),\n",
    "                      'control_p': round(control_p[significant_region], 3),\n",
    "                      't-value': round(tvals[significant_region], 2),\n",
    "                      'p-value': round(pvals[significant_region], 3),\n",
    "                      'corrected_p': round(corrected_pvals[significant_region], 3)\n",
    "                      })\n",
    "    \n",
    "    # Produce a latex table with the results\n",
    "    print(pd.DataFrame(d)[['label', 'ZDHHC9_mean', 'ZDHHC9_SE', 'ZDHHC9_W', 'ZDHHC9_p', 'control_mean', 'control_SE', 'control_W', 'control_p', 't-value', 'p-value', 'corrected_p']].to_latex(index=False))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Correlation between gene expression and regional graph metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              graph_metric    gene   control    ZDHHC9\n",
      "0              node_degree    FMR1  1.000000  1.000000\n",
      "1              node_degree   GAPDH  1.000000  1.000000\n",
      "2              node_degree  GRIN2A  1.000000  1.000000\n",
      "3              node_degree   FOXP2  1.000000  1.000000\n",
      "4              node_degree  ZDHHC9  1.000000  1.000000\n",
      "5            node_strength    FMR1  1.000000  1.000000\n",
      "6            node_strength   GAPDH  1.000000  1.000000\n",
      "7            node_strength  GRIN2A  1.000000  1.000000\n",
      "8            node_strength   FOXP2  1.000000  1.000000\n",
      "9            node_strength  ZDHHC9  1.000000  1.000000\n",
      "10        local_efficiency    FMR1  1.000000  1.000000\n",
      "11        local_efficiency   GAPDH  1.000000  1.000000\n",
      "12        local_efficiency  GRIN2A  1.000000  1.000000\n",
      "13        local_efficiency   FOXP2  1.000000  1.000000\n",
      "14        local_efficiency  ZDHHC9  1.000000  0.439246\n",
      "15  clustering_coefficient    FMR1  1.000000  0.620279\n",
      "16  clustering_coefficient   GAPDH  1.000000  1.000000\n",
      "17  clustering_coefficient  GRIN2A  1.000000  1.000000\n",
      "18  clustering_coefficient   FOXP2  1.000000  1.000000\n",
      "19  clustering_coefficient  ZDHHC9  0.444018  0.003820\n"
     ]
    }
   ],
   "source": [
    "expression_folder = '/home/jb07/joe_python/GitHub/ZDHHC9_connectome/expression_data/'\n",
    "metrics = ['node_degree', 'node_strength', 'local_efficiency', 'clustering_coefficient']\n",
    "genes = ['FMR1', 'GAPDH', 'GRIN2A', 'FOXP2', 'ZDHHC9']\n",
    "\n",
    "control_pvals = list()\n",
    "ZDHHC9_pvals = list()\n",
    "\n",
    "for metric in metrics:\n",
    "    df = pd.DataFrame()\n",
    "    df['ZDHHC9Mean'] = np.loadtxt(outfolder + '/regional_comparison/' + metric + '_ZDHHC9_mean.txt')\n",
    "    df['controlMean'] = np.loadtxt(outfolder + '/regional_comparison/' + metric + '_control_mean.txt')\n",
    "    df.index = pd.read_csv('/home/jb07/joe_python/GitHub/ZDHHC9_connectome/DK_labels.txt', header=None)[0].values\n",
    "\n",
    "    for gene in genes:\n",
    "        gene_df = pd.read_csv(expression_folder + gene + '_expression_lh.csv').append(pd.read_csv(expression_folder + gene + '_expression_rh.csv'))\n",
    "        gene_df.index = ['-'.join(label.split('-')[1:]) for label in gene_df.index]\n",
    "        gene_df = gene_df[['normalizedExpression']]\n",
    "        gene_df.columns = [gene]\n",
    "        df = df.merge(gene_df, left_index=True, right_index=True)\n",
    "        control_pvals.append(ols('controlMean ~ ' + gene, data=df).fit().pvalues[gene])\n",
    "        ZDHHC9_pvals.append(ols('ZDHHC9Mean ~ ' + gene, data=df).fit().pvalues[gene])\n",
    "\n",
    "results_df = pd.DataFrame()\n",
    "results_df['graph_metric'] = np.hstack([np.repeat(metric, 5) for metric in metrics]).tolist()\n",
    "results_df['gene'] = np.hstack(np.tile(genes, 4)).tolist()\n",
    "results_df['control'] = multipletests(np.asarray(control_pvals), alpha=0.05, method='bonferroni')[1]\n",
    "results_df['ZDHHC9'] = multipletests(np.asarray(ZDHHC9_pvals), alpha=0.05, method='bonferroni')[1]\n",
    "print results_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that these values have to be multiplied by two to account for the two group"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 442,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPcAAAD3CAYAAADBjMJTAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f837549aed0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(double_column/2, double_column/2), dpi=600, tight_layout=True)\n",
    "\n",
    "sns.regplot(x='ZDHHC9',\n",
    "            y='ZDHHC9Mean',\n",
    "            color='#D55E00',\n",
    "            robust=True,\n",
    "            data=df)\n",
    "\n",
    "sns.regplot(x='ZDHHC9',\n",
    "            y='controlMean',\n",
    "            color='#56B4E9',\n",
    "            robust=True,\n",
    "            data=df)\n",
    "\n",
    "plt.legend([r'\\textit{ZDDH9}', 'control'])\n",
    "plt.ylabel('clustering coefficient')\n",
    "plt.xlabel(r'\\textit{ZDHHC9} expression')\n",
    "\n",
    "plt.savefig(outfolder + 'clustcoef_ZDHHC9expr.png', dpi=600)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Analysis of influential data points"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 489,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "expression_folder = '/home/jb07/joe_python/GitHub/ZDHHC9_connectome/expression_data/'\n",
    "metric = 'clustering_coefficient'\n",
    "gene = 'ZDHHC9'\n",
    "\n",
    "df = pd.DataFrame()\n",
    "df['ZDHHC9Mean'] = np.loadtxt(outfolder + '/regional_comparison/' + metric + '_ZDHHC9_mean.txt')\n",
    "df['controlMean'] = np.loadtxt(outfolder + '/regional_comparison/' + metric + '_control_mean.txt')\n",
    "df.index = pd.read_csv('/home/jb07/joe_python/GitHub/ZDHHC9_connectome/DK_labels.txt', header=None)[0].values\n",
    "\n",
    "gene_df = pd.read_csv(expression_folder + gene + '_expression_lh.csv').append(pd.read_csv(expression_folder + gene + '_expression_rh.csv'))\n",
    "gene_df.index = ['-'.join(label.split('-')[1:]) for label in gene_df.index]\n",
    "gene_df = gene_df[['normalizedExpression']]\n",
    "gene_df.columns = [gene]\n",
    "df = df.merge(gene_df, left_index=True, right_index=True)\n",
    "control_results = ols('controlMean ~ ' + gene, data=df).fit()\n",
    "ZDHHC9_results = ols('ZDHHC9Mean ~ ' + gene, data=df).fit()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 490,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Container object of 3 artists>"
      ]
     },
     "execution_count": 490,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAagAAAEYCAYAAAAJeGK1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n0v8DtiOQW/LMYN4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f83748f5f10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "(c, p) = ZDHHC9_results.get_influence().cooks_distance\n",
    "plt.stem(np.arange(len(c)), c, markerfmt=\",\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Influential data point:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 491,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Index([u'lh-bankssts', u'lh-caudalanteriorcingulate', u'rh-bankssts'], dtype='object')\n",
      "[ 0.13630182  0.10965772  0.17549007]\n"
     ]
    }
   ],
   "source": [
    "print(df.index[np.where(c > 4./(68-2))[0]])\n",
    "print(c[np.where(c > 4./(68-2))[0]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 495,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.001254678194023045"
      ]
     },
     "execution_count": 495,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = df.drop(df.index[np.where(c > 4./(68-2))[0]])\n",
    "ols('ZDHHC9Mean ~ ' + gene, data=df).fit().pvalues['ZDHHC9']*40"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}