3. Plots in paper.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"name":"3. Plots in paper.ipynb","provenance":[],"collapsed_sections":[]},"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"language_info":{"codemirror_mode":{"name":"ipython","version":3},"file_extension":".py","mimetype":"text/x-python","name":"python","nbconvert_exporter":"python","pygments_lexer":"ipython3","version":"3.5.5"}},"cells":[{"cell_type":"code","metadata":{"id":"XJ1Vabgm7wLv"},"source":["# # LATEX compatible\n","\n","# !mkdir tmp\n","# %cd tmp\n","# ! sudo apt-get install texlive-latex-recommended \n","# ! sudo apt-get install dvipng texlive-latex-extra texlive-fonts-recommended  \n","# ! wget http://mirrors.ctan.org/macros/latex/contrib/type1cm.zip \n","# ! unzip type1cm.zip -d /tmp/type1cm \n","# ! cd /tmp/type1cm/type1cm/ && sudo latex type1cm.ins\n","# ! sudo mkdir /usr/share/texmf/tex/latex/type1cm \n","# ! sudo cp /tmp/type1cm/type1cm/type1cm.sty /usr/share/texmf/tex/latex/type1cm \n","# ! sudo texhash\n","\n","# !apt install cm-super"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"jKXJmuEn6Mbl"},"source":["__author__ = \"Lisette Espin\"\n","__copyright__ = \"Copyright 2021, GESIS\"\n","__credits__ = [\"Lisette Espin-Noboa\", \n","               \"Fariba Karimi\",                \n","               \"Claudia Wagner\",\n","               \"Markus Strohmaier\"]\n","__license__ = \"GPL\"\n","__version__ = \"1.0.1\"\n","__maintainer__ = \"Lisette Espin\"\n","__email__ = \"lisette.espin@gesis.org\"\n","__status__ = \"Development\""],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"FoOSkdHA6Mb1","executionInfo":{"status":"ok","timestamp":1636857226290,"user_tz":-60,"elapsed":256,"user":{"displayName":"Lisette Espin","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GjDac6esfgXyAYCvmgTKR1CnSZMB2A0Bt49sHtuDA=s64","userId":"15904188957194365368"}}},"source":["import warnings\n","warnings.filterwarnings(\"ignore\", message=\"numpy.dtype size changed\")\n","warnings.filterwarnings(\"ignore\", message=\"numpy.ufunc size changed\")\n","warnings.simplefilter(action='ignore', category=FutureWarning)\n","warnings.simplefilter(action='ignore', category=UserWarning)"],"execution_count":2,"outputs":[]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"6CHaCqio6Mb4","executionInfo":{"status":"ok","timestamp":1636857227730,"user_tz":-60,"elapsed":1043,"user":{"displayName":"Lisette Espin","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GjDac6esfgXyAYCvmgTKR1CnSZMB2A0Bt49sHtuDA=s64","userId":"15904188957194365368"}},"outputId":"01a149cb-83ce-400c-a484-b60748da02d2"},"source":["import sys\n","sys.path.append(\"/mydrive/\")\n","\n","%load_ext autoreload\n","%reload_ext autoreload\n","%autoreload 2\n","%pylab inline\n","\n","import os\n","from org.gesis.lib import io\n","from org.gesis.lib import paper\n","from org.gesis.lib import viz"],"execution_count":3,"outputs":[{"output_type":"stream","name":"stdout","text":["Populating the interactive namespace from numpy and matplotlib\n"]}]},{"cell_type":"code","metadata":{"id":"UKJPMGmF6Mb9","executionInfo":{"status":"ok","timestamp":1636857228916,"user_tz":-60,"elapsed":3,"user":{"displayName":"Lisette Espin","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GjDac6esfgXyAYCvmgTKR1CnSZMB2A0Bt49sHtuDA=s64","userId":"15904188957194365368"}}},"source":["ROOT = 'results/'\n","DATA = 'data/'\n","\n","MODEL = 'DPAH'                                                     # main model\n","MODELS = [\"Random\",\"DPA\",\"DH\",\"DPAH\"]                              # all models (Null)\n","ALLKIND = ['empirical'] + MODELS                                   # empirical and fit\n","MODELS_H = [m for m in MODELS if m not in ['Null','Random','DPA']] # models that rely on homophily\n","MARKERS = [\"x\",\"x\",\"x\",\"x\"]\n","QTYPE = {'mae':'qae', 'me':'qe'}                                   # quadrant type (based on horizontal ineq. metric)\n","SMOOTH = 0.05                                                      # smooth for quadrant assignment\n","METRICS = ['pagerank', 'wtf']                                      # 'pagerank','wtf',indegree','outdegree',\n","DATASETS = ['aps','hate','blogs','wikipedia']\n","\n","PLOTS = os.path.join(ROOT, 'plots')\n","viz.plot_setup(latex=True)                                         # Latex environment"],"execution_count":4,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"kPCOsTDx6Mb_"},"source":["<h2>Data</h2>"]},{"cell_type":"code","metadata":{"id":"DEC3qws26McD","scrolled":true,"executionInfo":{"status":"ok","timestamp":1636857232481,"user_tz":-60,"elapsed":2124,"user":{"displayName":"Lisette Espin","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GjDac6esfgXyAYCvmgTKR1CnSZMB2A0Bt49sHtuDA=s64","userId":"15904188957194365368"}}},"source":["df_node_metadata_empirical = paper.load_node_metadata(os.path.join(ROOT,'empirical'))\n","df_network_metadata_empirical = paper.load_network_metadata(os.path.join(ROOT,'empirical'))\n","df_rank_empirical = paper.load_rank(os.path.join(ROOT,'empirical'), df_network_metadata_empirical, SMOOTH, DATASETS, ALLKIND)"],"execution_count":5,"outputs":[]},{"cell_type":"code","metadata":{"id":"s7XwhdKG6McG","scrolled":true,"executionInfo":{"status":"ok","timestamp":1636857233639,"user_tz":-60,"elapsed":1174,"user":{"displayName":"Lisette Espin","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GjDac6esfgXyAYCvmgTKR1CnSZMB2A0Bt49sHtuDA=s64","userId":"15904188957194365368"}}},"source":["df_rank_fit = paper.load_rank_all_models(os.path.join(ROOT,'fit'), MODELS, SMOOTH, DATASETS)\n","df_rank_fit = df_rank_fit.query(\"dataset in @DATASETS and kind in @ALLKIND\")"],"execution_count":6,"outputs":[]},{"cell_type":"code","metadata":{"id":"LZrriww16McT","executionInfo":{"status":"ok","timestamp":1636857233641,"user_tz":-60,"elapsed":25,"user":{"displayName":"Lisette Espin","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GjDac6esfgXyAYCvmgTKR1CnSZMB2A0Bt49sHtuDA=s64","userId":"15904188957194365368"}}},"source":["df_rank = df_rank_empirical.copy()\n","df_rank = df_rank.append(df_rank_fit[df_rank.columns], ignore_index=True)\n","df_rank = df_rank.query(\"dataset in @DATASETS and kind in @ALLKIND\")"],"execution_count":7,"outputs":[]},{"cell_type":"code","metadata":{"id":"qHDTn5Y46McU","scrolled":true,"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1636857239582,"user_tz":-60,"elapsed":5960,"user":{"displayName":"Lisette Espin","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GjDac6esfgXyAYCvmgTKR1CnSZMB2A0Bt49sHtuDA=s64","userId":"15904188957194365368"}},"outputId":"3b156c5c-4d31-4a78-8857-fb7b9ab6a8ac"},"source":["df_rank_synthetic_all = paper.load_rank_synthetic_all_models(os.path.join(ROOT,'synthetic'), MODELS, SMOOTH, False)\n","df_rank_synthetic_all = df_rank_synthetic_all.query(\"kind in @MODELS\")\n","df_rank_synthetic = paper.remove_extra_synthetic(df_rank_synthetic_all)\n","\n","print(df_rank_synthetic_all.kind.unique(), df_rank_synthetic_all.shape)\n","print(df_rank_synthetic.kind.unique(), df_rank_synthetic.shape)"],"execution_count":8,"outputs":[{"output_type":"stream","name":"stdout","text":["['Random' 'DPA' 'DH' 'DPAH'] (1620017, 22)\n","['Random' 'DPA' 'DH' 'DPAH'] (541501, 22)\n"]}]},{"cell_type":"markdown","metadata":{"id":"E_Q2dJf56McW"},"source":["<h1>Disparity: Inequality vs. Inequity</h1>"]},{"cell_type":"markdown","metadata":{"id":"nxA_6F5_6McZ"},"source":["<h2>1. Synthetic networks</h2>"]},{"cell_type":"code","metadata":{"id":"QKOpxBaA6Mce","executionInfo":{"status":"ok","timestamp":1636857239585,"user_tz":-60,"elapsed":9,"user":{"displayName":"Lisette Espin","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GjDac6esfgXyAYCvmgTKR1CnSZMB2A0Bt49sHtuDA=s64","userId":"15904188957194365368"}}},"source":["metric = 'pagerank'\n","vtype = 'me'\n","mean = True"],"execution_count":9,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"PD10IoNr6Mcg"},"source":["<h3>Global Disparity</h3>\n","<p>Gini Coefficient vs. MAE Fraction of minorities across top-k ranks</p>"]},{"cell_type":"code","metadata":{"id":"0deFKUcY6Mch","colab":{"base_uri":"https://localhost:8080/","height":464},"executionInfo":{"status":"ok","timestamp":1636857241525,"user_tz":-60,"elapsed":1948,"user":{"displayName":"Lisette Espin","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GjDac6esfgXyAYCvmgTKR1CnSZMB2A0Bt49sHtuDA=s64","userId":"15904188957194365368"}},"outputId":"c076da28-d359-4ec6-d469-e797227787c6"},"source":["fn = os.path.join(PLOTS, \"vh_ineq_{}_{}_{}.pdf\".format(vtype, metric, 'mean' if mean else 'all'))\n","viz.plot_inequalities(df_rank_synthetic, MODELS, MARKERS, vtype=vtype, mean=mean, metric=metric, fn=fn)"],"execution_count":10,"outputs":[{"output_type":"stream","name":"stdout","text":["results/plots/vh_ineq_me_pagerank_mean.pdf saved!\n"]},{"output_type":"display_data","data":{"image/png":"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\n","text/plain":["<Figure size 600x400 with 1 Axes>"]},"metadata":{}}]},{"cell_type":"markdown","metadata":{"id":"rtdM5ic06Mci"},"source":["<h3>Global Disparity as a function network structure</h3>\n","<p>Gini Coefficient vs. MAE Fraction of minorities in each top-k rank</p>"]},{"cell_type":"code","metadata":{"id":"-Qje5Nt-6Mcj","colab":{"base_uri":"https://localhost:8080/","height":269},"executionInfo":{"status":"ok","timestamp":1636857245372,"user_tz":-60,"elapsed":3854,"user":{"displayName":"Lisette Espin","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GjDac6esfgXyAYCvmgTKR1CnSZMB2A0Bt49sHtuDA=s64","userId":"15904188957194365368"}},"outputId":"df56ad97-30bd-4724-f1fc-d13d3f83af24"},"source":["model = MODEL\n","qtype = QTYPE[vtype]\n","fn = os.path.join(PLOTS, \"vh_ineq_{}_{}_{}.pdf\".format(metric, model, qtype))\n","viz.plot_synthetic_quadrant_homophily(df_rank_synthetic, qtype=qtype, model=model, metric=metric, fn=fn)"],"execution_count":11,"outputs":[{"output_type":"stream","name":"stdout","text":["results/plots/vh_ineq_pagerank_DPAH_qe.pdf saved!\n"]},{"output_type":"display_data","data":{"image/png":"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\n","text/plain":["<Figure size 1000x250 with 6 Axes>"]},"metadata":{}}]},{"cell_type":"markdown","metadata":{"id":"CafaSIgc6Mcl"},"source":["<h3>Local Disparity as a function network structure</h3>\n","<p>Gini Coefficient vs. MAE Fraction of minorities in each top-k rank</p>"]},{"cell_type":"code","metadata":{"id":"-z2RJ4dG6Mcm","colab":{"base_uri":"https://localhost:8080/","height":617},"executionInfo":{"status":"ok","timestamp":1636857251248,"user_tz":-60,"elapsed":5886,"user":{"displayName":"Lisette Espin","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GjDac6esfgXyAYCvmgTKR1CnSZMB2A0Bt49sHtuDA=s64","userId":"15904188957194365368"}},"outputId":"433c7db3-4944-4c05-87e4-2135fd6e41da"},"source":["model = MODEL\n","qtype = QTYPE[vtype]\n","all = False\n","fn = os.path.join(PLOTS, \"vh_ineq_topk_{}_{}_{}{}.pdf\".format(metric, model, qtype, '_all' if all else ''))\n","viz.plot_synthetic_quadrants(df_rank_synthetic, qtype=qtype, model=model, metric=metric, all=all, fn=fn)"],"execution_count":12,"outputs":[{"output_type":"stream","name":"stdout","text":["results/plots/vh_ineq_topk_pagerank_DPAH_qe.pdf saved!\n"]},{"output_type":"display_data","data":{"image/png":"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\n","text/plain":["<Figure size 600x600 with 10 Axes>"]},"metadata":{}}]},{"cell_type":"markdown","metadata":{"id":"Vx_KrUKo6Mcy"},"source":["<h3>Correlation</h3>\n","<p>Spearman's test</p>"]},{"cell_type":"code","metadata":{"id":"0S5FaYeR6Mc0","scrolled":true,"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1636857251249,"user_tz":-60,"elapsed":21,"user":{"displayName":"Lisette Espin","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GjDac6esfgXyAYCvmgTKR1CnSZMB2A0Bt49sHtuDA=s64","userId":"15904188957194365368"}},"outputId":"e0039e44-f519-410a-c4af-6bd1ac1f8026"},"source":["for local in [True,False]:\n","    print('Local: ' if local else 'Global: ',\n","          paper.spearmancor_inequality_inequity(df_rank_synthetic, model, metric, local))"],"execution_count":13,"outputs":[{"output_type":"stream","name":"stdout","text":["Local:  SpearmanrResult(correlation=0.21465234847452905, pvalue=0.0)\n","Global:  SpearmanrResult(correlation=0.4111008799006741, pvalue=6.888599844888209e-250)\n"]}]},{"cell_type":"markdown","metadata":{"id":"WyqDLiMa6Mc2"},"source":["<h2>2. Empirical networks</h2>"]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":300},"id":"u9Xewrsk6Mc3","executionInfo":{"status":"ok","timestamp":1636857253069,"user_tz":-60,"elapsed":1836,"user":{"displayName":"Lisette Espin","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GjDac6esfgXyAYCvmgTKR1CnSZMB2A0Bt49sHtuDA=s64","userId":"15904188957194365368"}},"outputId":"2ac3edab-d949-4c86-f12a-2ab4b08ae2e7"},"source":["group = False\n","all = True\n","kind = ['empirical'] \n","kind = kind + MODELS if all else kind\n","metrics = 'pagerank' # pagerank, all, wtf\n","vtype = 'me'\n","\n","fn = os.path.join(PLOTS, \"vh_ineq_empirical_fit_{}{}{}{}.pdf\".format(vtype,\n","                                                                    \"\" if metrics=='all' else '_{}'.format(metrics),\n","                                                                    '_all' if all else '_empirical', \n","                                                                    \"_group\" if group else ''))\n","\n","tmp = df_rank.copy()\n","viz.plot_vh_inequalities_fit(tmp, x=vtype, group=group, kind=kind, metrics=metrics, fn=fn)\n"],"execution_count":14,"outputs":[{"output_type":"stream","name":"stdout","text":["results/plots/vh_ineq_empirical_fit_me_pagerank_all.pdf saved!\n"]},{"output_type":"display_data","data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAyMAAAEKCAYAAAAB74oSAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAAPYQAAD2EBqD+naQAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzde1xTd5438E+4KxCoF9RKQJQqJWBra6sEpO2gJLjbzmqrMK/d2bGtaO2M1XmEzm6fSqfU6XQKzg7a1gu12t3X85SIveksBEafzihJtK1rKwRtCxVycLy2kHARQ8h5/kjPaUISCCGcBPi+X6+8JCcn5/cDvzk53/O7iViWZUEIIYQQQgghAgvwdQUIIYQQQgghExMlI4QQQgghhBCfoGSEEEIIIYQQ4hOUjBBCCCGEEEJ8gpIRQgghhBBCiE9QMkIIIYQQQgjxCUpGCCGEEEIIIT5ByQghhBBCCCHEJygZIYQQQgghhPgEJSOEEEIIIYQQn6BkhBBCCCGEEOITQUIV1NbWBpVKhaqqKnR1dQEAIiMjIZFIsHLlSmRnZwtVFUIIIYQQQogfELEsy45mAV1dXdi7dy9EIhFycnIgkUgQGRnJv97Z2QmNRoPq6mpKSgghhBBCCJlARjUZ6ezsxOHDh/H000+7tb9Wq0VbWxvWrFkzWlUihBBCCCGE+IlRbxkZrs7OTruWE0IIIYQQQsj45BcD2LkxJAAoESGEEEIIIWSCEGwAu622tjZ0dnbyzysqKvDyyy/7oiqEEEIIIYQQHxG8m1ZRURG0Wq1dC8jly5dx5swZIatBCCGEEEII8THBW0ZSUlJQXFxst62mpkboahBCCCGEEEJ8TPAxI9HR0Q7b4uLihK4GIYQQQgghxMcEbxnR6/XYunUrUlNT+W1VVVV4//33ha4KIYQQQgghxIcEbxlRKpWIjY1Fe3s7//Cz2YUJIYQQQgghAhB8ALtWq0VaWprdtsbGRiQnJwtZDUIIIYQQQoiP+WzRQ61WCwAOiQkhhBBCCCFkYhB8zAjDMNiyZQv/vKurC++88w5iY2OFrgohhBBCCCHEhwRPRmpra/HBBx/Ybdu5cye2bdsmdFUIIYQQQgghPiT4AHZnLSASiUToahBCCCGEEEJ8TPBkpK2tzWEbwzBCV4MQQgghhBDiY4J300pOTsZTTz0FqVQKANBoNCgoKBC6GoQQQgghE4JSqYRer0dhYaGvq0KIg8Df/va3vxWyQIlEgvvuuw8NDQ0AgI0bN+K+++4TsgpkCBqNBiqVCvfff7/ddp1OhzfeeAPPPPMMwsLC0NTUBK1Wi48//hihoaF8dzudToePP/4YN27cQFNTE86ePYu2tjYkJib64tchhBAArs9h7777LkQiEX+O0mg0ePLJJxEaGoqUlBQf15r4C5VKhWeeeQb/+Z//icTEREgkEn7b7t27ERcXx8dQUVERfve73yEmJgaJiYlYvXo1oqKikJiYCI1Gg9WrV9vtb8t2X29JSUnB1q1bkZeXh9DQ0FErhxCPsH6AYRhfV4HY2L59O5uVleX0Nb1ez86fP99um8FgYOfPn8/q9XqWZVl21apVdq+//vrrbHV19ehUloxZDQ0N7Pbt29n58+ezr7/+OqvX61mDwcDu37+fnT9/Prt582ZWrVa7dSwu9ggZCne+cratoaGB3/b666+zFRUVQleP+Lnq6mqH78fq6mp28eLFdtsMBoPd+UutVrMGg4F/vn37dpffiwP39ZasrCy7445WOYQMlyBjRmpra9HV1QUAqKysdHi89NJLQlSDuMFoNEKhUMBgMECn0zm8HhUV5bBNLBYDsN5NdDb+Z+PGjd6vKBnzpFIp8vPzAQCFhYWQSCQQi8XIz8+HWCzGypUrIZPJhjyOTqdzGquEuEssFkMsFtP4RTIkhUIBhmHszjkGgwEA7LZpNBq785dMJuO/K4cynH1HQqhyCBmKIGNG9u7di8jISKSlpeG9995zuMBob28XohrEDRqNBgqFAhs2bIBSqURxcbFb7xGLxcjJyYFYLIbRaMSTTz6J/Px8pKSkQCwWQ6FQCFB7Ml5ERUW59SVpNBpRWlqK3NxcAWpFxiuNRoO0tLRBz1MajQY6nQ4SiQQMw0ChUPBdU1UqFRiGQUdHBwAgLi4OAJCTkwOlUsnvxzAMn4CTsUsul6Oqqoof+wpY/68HbuPodDps374dubm5Ts9Vzz33HBobG1FQUACJRGK3r0ajQVFREWQyGX9snU5n993MxZ9EIkF9fT0/LkSn06GqqgqpqakwGAx80uSqTlysckkVxSoRiiDJiO26Ijt27EBycrLd642NjUJUg7iBuwDMzc3FAw884DIZUSqV/M9GoxEnTpzg3/vBBx+gtLQURUVFYBgGcrkcu3btGv3Kk3HL1ZekRqOBwWBAVVUVDAYDnxC7+nImhKNSqQBYEwSNRjPohRfDMCgvL8fBgwf5batXr+a/27Zv347PPvsMAPDAAw/wP5eXl0MqlfI34GzPm2TsWrlyJbZv347CwkLodDrIZDJIJBJs2bIFhYWFMBqNdksWSKVSpKWlOT0WwzBITU21+4603Vcmk/HHt00aioqKUFxcDJ1OB6VSycemwWCAUqlETk4OtmzZguPHj/PHKi0tHbROSqUShw4dgkwmw5NPPmkXu4SMJsFn0+LuGAHWJKStrY2C3U9w3V24rgoSiQRKpdLpnZzB7kSLxWI+iTEajdiyZQtKSkrogpC4NPAizfYOHve6sy9JhULBd4fg7mq7+nKm1hNiy7YVJDc3F1lZWSgoKHAaJxUVFQ430QBrMiyRSOy6r0ZFRfGJsEwmw7p165CSkgKZTEZ3mscJhUKBLVu2QKfToaGhAbm5uXYTuBgMBreua+rr61FaWuqwELQzti3FMpmMT0aqqqoAWGPRllKpdKiDs27Wtg4dOoTq6mr+udFoHLJehHiD4OuM2AZ6cnIysrOz6W6Rn+C6EHDNtgUFBcP+vxnYf59LTKj1iwyGiznuMfBLk/uS5OJxsC9J2y/ngV/QhDgjFouxdu1aj76LBi7aKxaL+W0SiQSfffYZ8vPzwTAMnnvuOa/Ul/ieXC53iBeuq5a7oqOjUVBQgC1btoyoLlziK5PJXHYFG4rRaMS6dev4Y9Bi1ERIgiQjDMOgtrYWtbW1UKvV/M+1tbWorKyEXq8XohpkmBQKhV1LCfDjHevBLgb37dtn95xrxibEE8P5kuQSYW98OZOJpbGx0WVXmpUrV0Kr1dptMxqN/HlNJpNBqVTyLXicffv28fu5M/6OjB15eXkOrQ+5ubk4fPiw2xfyEokECoUCkZGRQybCtt+5KpWKP6fl5eU53HRRqVR8q7GtwSZo0Gg0iIqKshvfZDAY+O6MhIwmQbppSSQSdHZ2orS0FB0dHairq+NfE4vFWL9+vRDVIC5oNBq+L6lUKuVPRtzJsaioCPn5+YiKiuK37du3D3l5eU5Puunp6Q5jSqh7AhkOg8HAf/kO9iXJfZFzSTLDMMjLy8OTTz5pdzxuXzKxcV34AOt4Dm7Cjfr6eiQnJ9sN/NVqtWhsbOQHDhcUFKCkpARxcXHQ6/V240e4cSfAjxeKCoUC0dHRqK6uRlRUFAwGA/Ly8oT/pcmokMlkkMvlTseGDPxeHBhPXLx0dHRAKpUiNTUVRUVFfOJquy93rPr6eqhUKhgMBnR0dPDJrUQiQXFxMR+bwI/dqLnt3AB2iUSCF198Ebt27XKok0wmQ1VVFZRKJaKiopCbm4uqqiqKWSIIEcuyrFCFdXZ2QqPRQC6XC1UkIcRPcReGSqUS69evR15eHp/wlpaW8n3sU1JS8OKLLyI9PZ3vvsV9SXJf7OXl5fwAT7FYDI1GA7Va7fDlTIi3cVOaczHGjZMrKChwOrMSIcNVVFQEqVRK5zEybgmajLjS1dWFiIgIX1eDEEIIGZaSkhKkp6fbdddxto0QT1EyQsY7wWfTunDhgsO2ffv24U9/+pPQVSGEEEJGpLCwEOXl5XxLnF6vR1xcHCUixCu4iTgYhrHrtkXIeCJ4y0hSUhKkUim4YhsbG5Geno4DBw4IWQ1CCCGEEEKIjwneMlJcXIy1a9fyz41Go8MsJYQQQgghhJDxT/B1RmwTEcA6m5btYj6EEEIIIYSQiUHwlhFn3bH0er3L+d0JIYQQQggh45PgLSMVFRVob2/nHyzLoqCgQOhqEEIIIYQQQnxM8AHsWq12TLSCfP/996irq0NsbCxCQ0N9XR0yxty+fRttbW3IyMjAlClT3H4fxR0ZCU/jDqDYIyNDsUd8xd3Ys1gs8IPVLCYEkUjEP9wheDetsZCIAEBdXR2/Gi8hniopKcFjjz3m9v4Ud8Qbhht3AMUe8Q6KPeIrrmLPZDJBr9ejr6/PB7WauEQiEaKjoxETE4OAgME7YgmejADWBXwOHz4MkUiEtWvX4uWXX/ZFNQYVGxsLwBrc8+bN82ldrl69ipkzZ47b8saj5uZmFBYW8nHkLoo7iruR8DTuAIo9ir2RodgbG+WNR0PFnl6vR2RkJKZOner2nXoycn19fbh27RpaW1uRkJAw6L6CJyOlpaVIT0/n74JoNBrs3LkT27ZtE7oqg+KaiufNmwepVOrTukRERCA+Pn7cljeeDbfLAcUdxZ03eNLVhWKPYs8bKPb8u7zxzFnsWSwW9PX1YerUqQgK8sn99wkrMDAQs2fPxjfffAOLxTJo64jg/zOpqamQy+X8c7lcDoPBIHQ1/Iqpx4SQySEut0dERIz4GMPhTnlk/BM6DijuCIdij/gKxd74wo0RoRYR3+D+7kON1RF8Ni1nAREdHS10NfyGqceEik1H0XKGsdvecobBe898DFOPCZMmTXJ4z8DnymedH0P57FGH/YcysDwyMQkdBxR3hEOxR3yFYo8Q4Y16MnLhwgXU1tbyD7VajQMHDqCyshKVlZU4cODAhGgZ6brZ7XR793c96O/rx/GSU3wy0XKGwV9eP4Vew220fn4ZN27c4Pd3lmCETA7Bsk1LcGJnnd0xTuysw7JNS4bdMmJbHpm4hI4DijvCodgjvkKxR8ab1atXQ6fTjXgfZ3Q6HZYvXw6VSuVp9QAI0E1LrVZDo9HYDSxqbW212ycqKmq0q+FTXTe78X/Wf4hFj0vx4M8X8ds//a9zOPe+Dpm/XIK6vZ/heMkpJGYmoOnkJQDAoiekOPnGacheWATE/5hgZG3LcEgw5iyRIGtbBk7srMO8jDlormtB1rYMzFkiEfR3JYR4l21XS3d+JoQQQjgbNmyARDL4taA7+zgjlUohk8k8rRpv1JMRqVSK9evXj3Yxfi1iWjgWPS7FuSPWrPPBny+yJiJHdAgKDcIkcRhWPL8Mta+dxFcnmiESAcGTg3Hv41LEzJ8G7Tuf41piu0OCMfACZM4SCeZlzMFXJ5qxIGuex4lITEzMyH9pMuYJHQcUd464LpjLNi3Bnakz+J8B4NSeM8h96zH8vf4a//N4QbFHfIVij9hiWRZqtRpNTU1ITExEenr6mBt/olAovLKPK5GRkR6/lzPqyYg764q0tbV5NB3gWMK1iJw7okOzWg/jlU4sesKabJzYWYfUR5P4fVkWkObMBwDcmToDEZJwuwTD1GOyuwDhEpKWMwya61qwIGsemutaMGdJrEcJSU9PD/VjJYLHwXiPO08mmbDtgpn5q6VYtmkJjpecAssCK55fhr/XX3PZWjqWUewRX6HYI5zW1lbI5XJcunQJISEhMJlMSEhIQE1NjVdnQFMqlTAajRCLxTAajcjPz4dGo0FRURHy8/MBgH89KioKBoMBarUaO3bsgFgs5veVy+WIjo5GR0cHoqOjkZ+fD51Oh+3bt6OgoAAymQwajQZbtmzBK6+8Ao1GA4lEAplMZrcPYJ3pVq1WIy4uDjqdDsXFxSgpKUFqairq6+uRmpo6ogRmIEFm06qtrYVMJkNERAQqKysdXlepVDhw4IAQVfGpB3++iE9ExLMi+QQl9dEka6uJCBAFiCASAV9+1Ahd9TeQ5tyFqLkRuKK5jqaTlzD5jjDoqr9Bf58ZywuW8ce27cI1Z4kEc5bE2j0fjq6uLkydOtWrvzsZe4SOg/Ecd7YtHLafx5YzjMNNhYHmLJEg81dL8f92qjF74UywLCASAS1n2tB08hICQwJwZ+oMoX4VQVDsEV+h2COAtUVELpejubkZZrMZJpN1nG5zczMUCgUaGxu90kKi0+mgVquxa9cuAMBzzz0HnU4HmUwGmUwGvV7PL4WxYMECHD9+HBKJBEajEdXV1cjNzYVMJkNycrJdgvDkk09CIpFAoVAgJSUFRqMRACCTyZCWlgaGYVBQUACGYSCVSu32YRgGpaWl+OCDD/g6GY1GNDY2orCwEAqFAsuXL/dqMiLIbFp79+5FfX09AOC9995Da2ur3aO9vV2Iavjcp/91jk9EjFc68el/nYOpxwRd9TeACAALzH9kLlb8JhOACGaTGeeO6BA5PRzBk0LQb7bg3BEd+m718YmI8tmj6LrZjVN7ztglHtwYklN7zgx7Nq2x1gRJRofQceCruHP1+Rju52YwI51kIn7xbAQEBeDy+auYlRyDxMwEfHWiGf1mC3wwKeKomyixR/wD91k39ZjsYsF2+2ih2PNParUaLS0tMJvNdtvNZjO+/fZbqNVqr5RTVVUFwHpTXqVSOYzbSE1NtXvOvc61otgSi8X8zwqFgj+2s25UEokEYrGYX9fHdh+VSoWUlBT++a5duyAWi1FWVgalUgmlUgmDweBQ/kgI0jLCZVcAsGPHDiQnJ9u93tjYKEQ1fIobI7LoCandmJH+vn6wLIuAwADc9VAC373qkS1pOLFTjdSfJuF/3mjEjAXTcfn8VQBAQFAAzLfN+Nsbp5G1LQMR08Kd3lmds0SCO1NnDLv7RlxcnNd+bzJ2CR0Hvoi7kbRYDNdIJpn4e/01ANYLl8vnr1ovYH64gSHNuWtcddECJkbsEf/AnQPSnrofmgNnkfnsEiDux3PA0qfuh/bAWeTt8d65wBbFnn9qampCcHAwbt++7fBaSEgImpqakJGR4ZWyuBYMYPCxG7bJhjfKHA6GYbBlyxYcOnQIYrEY5eXlXqsL4INbavv374dWq7XbNjA5GW+6bnbj3Ps/JiKAtcvWoiekOH/0Ivr7zFjx/DI8/Fwaf7ESFBqEfz6wCrKnFuOeDUnWROSHblz9ff346y6t3YXMYH3Oh4thmKF3IuOe0HHgi7jz9rTYQ7GdZGJexhy3EhGuPiueX4Z7H7eeK1mWhQgiLHpCivpjFx3WGBrrJkLsEf/AnQP+uluLvltmtDS3QPvO5zixsw4Lsubhb7u06O/rh6mnb1RaSCj2/FNiYiLfNWsgk8mExMREr5SzcuVKu2tihmHcjomOjg6757bvU6lUyMvL86hOCoUCGo3G7rjHjh1DSkoKnxBxS3KMdEpfjuDJSEpKisOg9oHJyXgTMS0c//z2KrtpfQHg3selCBOHYkVhptPuVSGTg9FyhkGP4RZEASIEBgVgVsoMgAUs/YOvZjkSFotl1I5Nxg6h40DI8mwvKmxbLE78UT3kWKuRdOsaOMnEUEmEqcfEd8G8M3XGj106YU1IpsRHe9wd05+N59gjvuVs0eA5SyRI3/AALP0WBAQH4PzHFxExPRznP77ATyhT+dyfUbHpqMs1wzxFseef0tPTkZCQgKAg+w5EQUFBmDt3LtLT071SjlQqxYYNG1BUVASlUskPKtfpdNBoNPzg9vLychiNRiiVSjAMA5VKBa1Wa7c2iE6ng1KpRElJCRQKBWQyGXQ6HbRaLaqqqmA0GqHRaNDY2Ih9+/bxycvAfSQSCYqLi+3q9C//8i/o6Ojgu2lt2LABpaWlCA4OtnuvpwTppmVLLBbjpZdeQkpKCr++iFKpdGvWrbEsYlq4w7aQySH42d6fuuxexc2UM0s2Hdn/lonrX9/EuSM63LlwJq5duI5PyjT457dXef3ubXi4Y13JxCN0HAhVnrOuWXOWSDAtcSqa/nYJiQ8l2CUiA9fz8LRblyeTTIRMDuGP2XKGQV9vHwKDApCYmYBv/nYJn5RpseL5ZV7tTuYPxmvsEd8a+PntutmNDwuqMS9zDuqPXsSs5BgYLnUBADrarBdWQaFB+PIj62Blac58fFhQ7dXPG8WefxKJRKipqXGYTWvu3Lmoqanx6lgfhULh0D1LKpXi+PHj/PP8/Hx+Zi0AOHjwoNPjDFzzQyqV2g2VkMlkdsd1tg+338BjcYPsB8rKynK6fTgEbxkpLy8Hy7Kor69HXV0d6urqJnQz5WAnNO6O6LJ11nUF6o9dxKInpOhgOvDwc2mwmC18X3JvioiI8PoxydgjdBwIVZ6zrlnadz7HVd113BEfjZbTersuW8pnj/J3Uz3t1mXqMeHkW6edTjJx8q3Tg7ZqhEwOganHhE/KNAgIEGF5obVL54rnl0EkAj4p07h871g1XmOP+Jbt5/ebk5fwYUE1Ztw9HfUfX0TMgmm4oruOnqv2YwTMt80QiURY+NO7UX/sote7b1Ls+a/4+HhcuHABJ06cwO7du3HixAk0NjbSOJ9RIHjLSHFx8YTrpuUJ2zuil769hLo9Z/kLmXsflyJkcgiCQoNwas8ZjwapD+batWtenUObjE1Cx4GQ5dl2zeImh0iQScCc/TtSH03i1/6pP3bRYQ0Pzweiu7qT5t4dtsDgQDyyRWaXzCwvXIaTb51x6/1jyXiOPeJbAz/7lzQMYpKm4frFmwiLCsWstKnoaLbvbhISEcqfCzxdTNgVij3/JhKJkJGR4bXB6qOB63pVUVFhN65jLBE8GbGdpqyxsRFtbW1eWUp+POIugAICA+yahbl/PZ0tixBi/fwkLJXgm7+1YPbCmcj+zUN8K8eMBdNx7ogOP9mW7vTiw3YgOrcY6WBCJocg89kldt2ybLttDfUZDpkc4nQ2HzoHEDJ8tp/fWdIYXNFdR1h0GHo7ep3uf6v9FmZKY7yeiBDiDc66Xo01gnfTqq6u5n9OTk5GdnY2lEql0NUYU6ZPn+7V2bLcKY8QoeNA6PJazjC4dJrBXQ/NwbWvbqDlDMNfpFw+fxV3PTQHd2UmuHzvcAaiA/Z3ZP+6SzvsRUmFPAf42niPPeJbtp/fG03fIVoSxScibSd/7Poc/+BsBARaWy6v6q7j0/865/W6UOwRIlDLCMMwuHDhAgDrQjLcwHXAOj2YXq8XohpjVm9vLyZPnjxuyyP+aTzH3cDB5Nzz1EeTHJKMgcmCJwPROcNtUZmoxnPsEd8a+Pk13TLhkoZBxPRwdN3oRvjMSejU9yAmaRpaP72MRU9IMSU+Gn974zTOHdFhSnw0El3cpPAExR4hArWMSCQSxMbGoqKiAnq9nh+4XldXh9bWVqxfv16IaoxZnZ2d47o84p/Ga9zZTpdrO/4i9dEknDuiw0O/Wmq35o9tq4er97o7va4nLSoT0XiNPeJbAz+/TScv4ZKGQfyDs9F1sxsLf5qEKUlRkK68C9e/uomFP03CVyeaEbd4Nn5+6HFkbUuH9p2zXp1Gm2KPEAHHjCQnJ6OsrAwajQZyuVyoYgkhxI7t5BAcU48JX51oRta2dP6up22SwY3LcPZebt+hxm6MpEWFEDJytp9fU48J2nfO8p/5rpvdiJgWjgtfXMTnxxqwdvc/4g5JNO7PW8h/rhMzExC3ePa47BpJiC8JOoA9MjISDMOgsrISa9aswdatWwEAubm5436dkZEQeqYNmtmDAOM77gZeTAwnyfBk7MZQLSo0CN3eeI494lu2E8HYfua5tcDuvjcJ896a6zBhzMD3ewvFHiE+GMAukUiwZs0avP3224iKisKf/vSnCb3OiDva2trGdXnEP020uBvNAeLchc/AFpA5SyTjbsFCb5hosUd8w9nnrq2tTdDPI8UeIT5IRrhVK7VaLb/ipO2AduKov79/XJdH/BPFnXdNpNmwRopij/gKxR4Zr8rLy1FeXu7rajgl+Dojer0eWq0W9fX1SEtLQ2dnJ4xG49BvnMCEnmmDZvYgAMUd8R2KPeIrFHtESDqdDvv27UNNTQ0KCgrQ0dEBAIiOjkZ+fr5Xy5LJZNi+fbvXj+sNgreM5ObmQqfT4dChQ+js7MS+ffsoGRmC0KtpjsXVO4n3UdwRX6HYI75CsUeEJJVKsXHjRkgkEuTn56OwsBCFhYVgGAZFRUVeLUsi8d+JUgRPRiIjI7F+/XokJycjMjISBQUFNLvWEK5evTquyyP+ieKO+ArFHvEVZ7HAsizq6upw6NAh1NXVgWXZUS2P+AeLi2mXXW33JqlUioaGhlEvx18I0k2rtrYWMpkMERERqKysdHhdpVLhwIEDQlSFEEIIIcQtra2tkMvluHTpEkJCQmAymZCQkICamhqaCWscs3R24lrmw4h+7VVMsrlhfqumBh3/9gJmnPwrAiIjR618tVqNDRs22G0rKSlBamoq6uvrkZqaCoVCAY1Gg6KiIr7rlV6vR1xcHHJzcwEAGo0GarUacXFxDr2QNBoNdDodxGIxjEYj8vPzHY5nNBohFosRFRUFg8EAtVqNHTt2eL1FT5BkZO/evYiMjERaWhree+89yGQyu9fb29uFqMaYNXXq1HFdHvFPFHfEVyj2iK/YxgLLspDL5WhubobZbIbJZF3ssLm5GQqFAo2NjfykPN4oj/iPgMhIRL/2Ktp/uRl4czcmyeW4VVOD9l9uxh1v7vZ6ImIwGKBSqWAwGKBUKlFQUOBwrdzY2IjCwkIoFAosX74cCoUCMpkMMpkMer0ehYWFMBqNWL16NXJzc8EwDEpLS/HBBx8AABiGQXV1NQDrWJWKigrs2rWLf15UVITi4mK74wHAggULcPz4cUgkEhiNRlRXV/PJjrcIkoxwfwgA2LFjB5KTk+1eb2xsFKIaY1ZfX9+4Lo/4J4o74iuDxQLLslCr1WhqakJiYiLS09NHfEFIsUc4trGgVqvR0tICs9lst4/ZbMa3334LtVqNjIwMr5VH/MskuRx4czfaf7kZvY89iltHj+GOHxITb4uKiuJnmJXJZFi9ejVOnDhh1wJRVlYGpVIJwJq8cK0WAJCamgrAOgbJYDAAsPY6sl3Dz3bm2qqqKv49gCkoYPoAACAASURBVLVbmFKpRHFxsd3xONx4E64VxdsEHzMyMBFxtY38SOgB/jShAAEo7ojvuIqF1tZW3H333cjKysLmzZuRlZWFu+++G62traNSHpl4bGOhqakJwcHBTvcLCQlBU1OTV8sj/meSXI5Jjz2KHuVhTHrs0VFJRAaSSCSIioqCRqPhtzEMg3Xr1iEnJwe5ubmCLokhxCQLgiQjXV1dgz64ldgJIYQQZ2y7zJhMJnR1dcFkMvFdZrw5qJgQAEhMTOS7Zg1kMpmQmJgocI2I0G7V1ODW0WOYnLsWt44ew62amlEph2vN4IjFYtTX1wOwJiIajQYpKSl8YmDb+uGKQqGAVqvlnzMMwye/K1eutEt2NBqN25NJcdMPe9Ood9MqLS3F4cOHER0dDZZl+T9EVFQUWJYFwzCIi4tz+3g6nQ4ajcZuwI0r5eXl/H+cTCbz62nNBiN0vcfq34l4F8Ud8RVnsTCaXWYo9gjHNhbS09ORkJDAjxnhBAUFYe7cuUhPT/dqecS/2I4RmSSXI0yebTeGxBsYhuGXuCgvL+evaQsKCvhFChUKBXJycqBWq/luWhs2bEBpaSkeeughaDQaMAwDmUwGpVIJo9EIpVKJ3NxcbNiwASUlJYiLi4NerwfDMPxrubm5dq/t2LGDv8Z2djyZTMaPa9HpdJBKpV75GwAA2FGmVCoHfe5qmyvr1q3jf66oqGArKiqc7vf666+zarWaZVmW3b9/P7t//363y2BZlm1oaGDnz5/PNjQ0DOt9o+Hy5cvjurzxyNP4obgjIzGS+PH32Dt48CAbERHBAnB4REREsAcPHvRqeWR4xmvstbS0sAsWLGBDQkLYiIgINiQkhE1KSmJbW1tHpTwyfIPFj9lsZhsbG1mz2TysY/Ybjezf772P7VGp7Lb3qFTs3++9j+03GkdU54nC3b//qLeMrF271u55dHS0wz7Otjmj0WgQaTODQUpKCkpLSx1G9RuNRtTU1PAzAfjjapPDQQOJiS9Q3BFfcRYLo9llhmKPcAbGQnx8PC5cuOD1SRNclUf8Q0BkpNPpeyfJ5QiVyUZ1Wt+JSPAB7OfPn3fYxvWLG4pOp7NLXMRisdNFYRoaGiCRSKBSqaBUKlFSUuJ5hf1AWFjYuC6P+CeKO+IrzmKB6zITFGR/D80bXWYo9gjHWSyIRCJkZGRg3bp1yMjI8Foi4qo84h9cJRyUiHifIFP72lq5ciVWrFjBz5+s1Wrx8ssve7UMbrDPwYMHAYBPSLiWkuH44osv0N7ejiVLlqC+vh49PT0Qi8W46667cPbsWQDA3LlzYbFY0NLSAgB44IEHcPHiRXR2diIiIgJ33303PvvsMwDWuyyBgYH49ttvAQD33XcfmpubYTAYMGnSJNxzzz04ffo0AGtf0rCwMDQ1NeGbb77BvffeC71ej++//x6hoaG4//77+QFId955JyIjI/HVV18BsE7LduXKFdy8eRNBQUFYunQp1Go1WJbFjBkzMHXqVH5KZalUips3b+LatWsICAjA4sWLcfr0aZjNZkyfPh0zZszgk76kpCQYDAZcuXIFAJCRkYFPP/0UJpMJU6dORWxsLL788ksAwPz589HT04O2tjYAQFpaGr744gvcunUL0dHRSEhIwLlz5wAA8+bNQ19fH/R6PQDgwQcfhE6nQ3d3NyIjI7FgwQJ8/vnnAICEhAQAwKVLlwAAixcvxldffYXOzk6Eh4dDKpXi008/BQDExcUhODgYzc3NAIBFixbh0qVL6OjowKRJk3DvvffyA7xiY2MxefJkfP311wCAe+65B21tbfjuu+8QEhKCBx98EHV1dQCAWbNmISoqChcvXgRgbaW7du0abty4gaCgILsWPE9Q3FHc+SLuAP+Overqajz88MO4cuUKgoODYTabMXPmTOzYsQNGo5Fij2KPznsUe2QsEqDLmAOj0ciWl5ez5eXlrF6vd/t9+/fvZzdv3sw/b2hoYBcvXuywX3V1Nbtq1Sr+uVqtZrOysoZVR3/qw9rS0jKuyxuPxsOYEYq7sWe89NsfLBYsFgt76tQp9uDBg+ypU6dYi8UyquUR90yE2BsP5Y1HozFmhHiH34wZcSYyMhLr168f9vukUqldly6DwYCUlBSH/Wh2CkIIGZ+4LjMjXWyOEEKIfxB8zMhIyGQyvgkSsHbH4las1Gg0/LTBA6cb0+l0bs+f7I+mTJkyrssj/onijviKs1gw9bgYvO5i+0jLIxMTnfcIEd6YSkYA4JVXXkFJSQmUSiX0ej0/k1Z5ebndAi6vvPIKioqKUF5ejo6ODo/Gi/gLi8Uyrssj/onijvjKwFgw9ZigfPYoWs4w/HNTjwktZxgonz3KJyTc9pGWRyYuOu8RIjyfdNMaCalU6nShFW6wuu1+xcXFQlVrVHV0dCAqKmrclkf8E8Ud8ZWBsRAyOQTLNi3BiZ11yPzVUmgPnEXfrT6wFhYPPZcGAGg6eQmn9n6GwGAR8vb8lH+fqceEkMkhwyqPTFx03iNjHcMwY264guDJSG1tLbKzs+22HThwAE8//bTQVSGEEDJG3Jk6AwFBAfjbLi1m3h2Dy+evAgHAyTfPgLVY0G+yAAGAfMtD+Hv9NZzacwZLn7ofp985i1WlOYiYFu7rX4EQQuzodDrs27cPNTU1KCgoAGBNUDs7O1FQUACxWDys/QCgqKgICoXCYQ0+fyZ4N63S0lK754cPH3bYRuzFxsaO6/KIf6K4I77iLBZCJofgkS0yWPpZXD5/FbMXzkSASARzr9maiACABfj6r9/ixM46JGbOwck3TmNB1jx8WFA9aPctij3CofMeEZJUKsXGjRshkUiQn5+P/Px8FBYWoqCgAOvWrbMbC+1sP6lUinXr1vHHMxqNSE5Ohkql8tFv5BnBk5GkpCQcOHAAXV1deOqpp6BWq7Ft2zahqzGmXL9+fVyXR/wTxR3xlcFiQfTDt9bl+quw9LM2LwCiQBEuaRiERYXh/McXMT9rLuqPXcSyTUsG7apFsUc4dN4j/kAsFiMnJwcvvvjioPtJJBLodDr+uUajQWFhIRoaGvhEZiwQPBnZtWsXsrOzsXr1auTl5aGsrIyfEYs4ZzJ5NltMd695WNtHWh4ZX4SOA4o7wnEWCy1nGJzYWYcVz2fizoUzAXbADizA9rMICQ9B1/VuhISHoLHqG2T+ainmLBm8/zTFHuHQeY9wPL2G8haFQoGamppB91Gr1U67Y6WlpaG6unq0quZ1giQjtbW1do8LFy4gOzsbDMOgtraWumkNITQ0dNjv6e41Y+3uUzh50f6uy8mL17F296lBP0yelEfGH6HjgOKOcAbGgqnHhFN7ziBrm3VtkauNP57XRIEiQGSzb7cJQWFBMHWbMCU+GndlJgy7PDJx0XmPACO7hvIWbmID2xYOg8EApVIJpVKJ8vJypKam8pM1GY1GfuB6Xl4elErlqNfRWzwawD7cAecvvvgicnJywLL2t7KMRiNaW1vR2NjoSTUmjGnTpg36enevGeFhQQ7PN2bdhaIj51H8xEJkJsWgtv4KXv1Yhxd+KrXbf7jlkYlB6DiguCOcgbEQMjkEuW89hr/XX8Px0lMICg0EQgMRMT0c37d0QCQCQqNC0Wu4DQAw95oREROO71s78Ol/ncODP180rPLIxEXnPQIA4WFB+M2jUrtrqJMXr/PPB7uG8haDwQAAdoPTAbgcmF5dXQ2j0ch30dLpdGNmZi2P/poVFRUQi8VIS0tza/BVWVkZ0tLSXL6u1Wo9qcaEcfnyZcTHxzt9jcvet8iTkL1wFrp7zViz6xTkC2eh4nQrHl8swfbKL5Eqicbnl77Hz2TxKFNdRPpd0wHA6QdqsPLIxCF0HFDcEY6rWDi15wyWFyzDnakz0Pr5ZZx84zQCggJgMVvQa7iNkPBgmLr7ED49HJa+fqT+NAnnjugwJT4aiYO0kFDsEQ6d9wgnMykGxU8sRNGR81ieMhPHG67yiYkQNBoNZDKZ2/tHRUXZJSoMw0ClUiE/P380qudVHnXTeuedd7BmzRoYjUZUVlaitrZ20P0HS0QAIDU11ZNqEFiTiS3yJBS9fx57jn8NAOi5bcZ72lZIY6Nw5DMGff0WfH7pewSIgA8/Y/CbR6U42/K9YE2NhBAyUlzryJwlEvy9/hpOvnEaWdsy8PCWNAQEBUAUKIKpuw93xEej+0Y3EjPnQPbUYmRtS4f2nbNeWamdEGffmd29ZpfbydiWmRSD5Skz8edzl7E8ZaZgiYjRaIRSqXR7vTyGYRxaUBQKxZjpquVRMsI1+SQnJ0MsFmPv3r3Izs5GZWUl2trahnx/W1sbLly4wD9KSko8qcaEcccdd7h8rbvXjOyFs/AzWTzePXUJBe/9D3rNFogANDAGxIhDYfmhd5yFBVY9YP2/KzpyHr951Hl3rcHKIxOH0HFAcUc4rmKBW8SQGz9yZ+oMnH7nLO75p7shEokQEh6Mf3otG4uekOL8xxfRdPISEjMTkPvWY4POpkWxRzhDfd8+seuk3TiC7l4z/uk//obH/vhXu+TD3bEFFHv+7eTF6zjecBX/uGg2jjdcdRhDMlLc+iEMw6C8vBzl5eUoKSlBaWkpysrK+OtthmGwb98+GI1GlJeX2x1Do9HgySefdNiuUqnAMAxKSkr8fmYtj7ppbd26FVFRUVCpVEhLS0NhYSHf+tHY2Oh0YUNOUVERtFotIiMj+W2XL1/Gyy+/7ElVJjTbLlq1568gcUYEzrW0IzIsCJ0/nACvG2/z+4sAHNbqceQMg9+tvUewDJ8QQryFayHhkotVpTn4sKAaK563dt8KmRyCB3++CFPio6F95yziFs8ecgV2QtzRfduMvn4W//vwl/x3qPqbG+i+bYaFBdRf30D2wlk/jC34EsVP3CPI2AIyOmzHiGQmxSAzKcbuuTdIpVLs2rVryP0kEonL/WQyGY4fP+6wvbi42O2WFV/z6FNSV1eHTZs24fjx42AYBgzDoKurCxEREQB+HHTjTEpKisMfZ6ipyya69vZ2h+Y3wH6AVdpdU/FJ43U+EZkcEoAebiEwAIkzInDpRjf6LSyCRCLcP2fKsMsjE4vQcUBxRzhDxYJtchExLdxpy0diZoLbiQjFHuG4ioXuXjN+sU8Dc78FLGtNSH4inYGa81cAAKFBAXj1aAM+/fY71Jy/gpCgwb9nhyqP+FZ3rxl/OKazSzy4MSR/OKbD/XOmUKLpRR5101q1ahWefvppREZGIjk5GXK5nO+XptFo+OnInImOjnbYFhcX50k1CKwfjtylcfik8TpmRoWhs9eMsGD7REQEoOVGNwAgMEAEC8vibMv3PqoxIYR4l6uEg1pEiLeEhwXh14q70dtngYVl0W+x8IkI5+G7Z+DP5y7DbLEgQCRycSQyFoSHBeHw5mUOLSCZSTE4vHkZJSJe5vZfk2EYXLhwAQBw48YNu0HrBoMBer0eALB+/fpBj6PX67F161a7QetVVVV4//33h1XxiWT27NkuX+vuNePP5y5jXkwEmq934Y7wYLR39/GvJ0wPh/67HpgtLCYFB+LfH5Oi5L8bsb3yS7yyxnlXrcHKIxOH0HFAcUc4FHvEVwaLhfT50zEpOBC3+vodXrtttuAv9VchEgEsC6x+QOLWBSvFnv9y9f9HiYj3ud0yIpFIEBsbi4qKCuj1etTV1fGP1tbWIZMQjlKpRGxsLNrb2/nHwPVHiL2bN2+6fC08LAgbf3IXmq93IXFGBNq7+xDwww2ZtMQp6OgxYVJIIAIDRAgMsJ5MP/r1Q3hlzT34wzGd08F1g5VHJg6h44DijnAo9oivDPV9u3ap654c/SwLEYCw4AAoT7e6NdiZYo+QYY4ZSU5ORllZGTQaDeRyuUcFFhcXO0z1S4seDu727dsuXzt58Tr+WH0RP0uLx4eft+FnafF4T9uKR5JjoP3mO7zwmBTp86fjbMv3+P3RBgDWE2pmUozLPo+DlUcmDqHjgOKOcCj2iK8MFgvdvWYoT7cO+n4LC+QujYc0NtqtsQUUe4R4MGYkMjLSaSLizpS+gOOaI7W1tYMOeCdASIjzfs/cAKsXHpOitv4Kip9YiC2KJBQ/vhDn9R144adSlNVcBGDt53jkuUy7k6KrE6Sr8sjEInQcUNwRDsUe8ZXBYkH9zQ309lkQIHJ+8RQoEiEwQIT/q7EmLO6MLaDYI2QYLSO1tbWQyWSIiIhAZWWlw+sqlQoHDhxw61harRadnZ0AgI6ODlRVVQ25MOJEFhPjfAo5boBVeFgQ0udP50962Qtn8c/T7/pxu7v9HF2VRyYWoeOA4o5wKPaIr7iKhe5eM/6j+gLCggLQZ2ERECDCCpvZtABgRepM/EQ6E//78Jd45cN6fPTrhzwuj5CJxO2Wkb1796K+vh4A8N5776G1tdXu0d7e7tZx3n77bZSXl0OpVKKurg4qlQorV670rPYTxGCtTq4SjeEmIO6WRyYOoeOA4o5wKPaIr7iKhfCwILy7UYagQBECRCL8bu09eP4fkhEWbL2MCgoQ4a8XrgEAfrf2HgQHund5RbFHyDBaRj744AP+5x07diA5Odnu9eGM+3jnnXdw+PBhrF27FgDsZuYihBBCCPE34aFBCAkKRNFqKTKTYvgB6r9YloBj/3MZWxVJ/KJ4lc/R9K+EuMujT8rARASA24v2SKVSdHV1AbCuvD579mwaMzIEZ2uzjKfyiH+iuCO+QrFHfGWwWAgPC+KTjB8XxbNOkf+vGXMRHhaEsJBA/OGYDoc3LxtxeYRMFG4nI87Gidhyd8wIwzAoLS3FoUOH8Itf/AJ33HEHIiMjsWbNGnerMuEEBHi0NuWYKY/4J4o74isUe8RXhooF2y7QtgPUuX8Hm6nSk/LI+KbT6bBv3z7U1NSgoKAAgHUsdWdnJwoKCvgb/QP3y8/PB8MwKCkpsds2VrmdjOzfvx8KhcLlmiDujhlZu3Yt3z3r3XffhUajgUwmc7caE9L333+PyMjIcVse8U8Ud8RXKPaIrwwnFryxKB7F3sQmlUqxceNGNDY22iUTRqMR69atw6FDhyAWi/n9tFotv59EInHYNla5/Ylxtj6ILXfHjGzduhW5ublIS0tzOU0wIYQQQgghvtLT14PJwZPd3u5NYrEYOTk5ePHFF7Fr165RLcsfuN0+OFgiMpy1QlJSUhyOpdVq3a3GhDRr1qxxXR7xTxR3xFco9oivUOwRwJpwbDq+Aaev2F+fnr6ixabjG9DT1zPqdVAoFKipqRn1cvyBx1M9eLpWiFgsxksvvYSUlBRERUUBAJRKJa0zMoj29nbMmDFj3JZH/BPFHfGVgbHQ3Wt22vXluqEXMVFhbm93dRyKPcKh8x4BgMnBk7Hp3l9i5+cl2La4EEtnpeH0FS3/fLRbRgDw18hGo5EfO2I0GqFUKvl99Hr9qNdDCB4lI2+//TY0Gg1EIhFmz56NtrY25OXlufXe8vJypKWl8WuWANZB7cS13t7eEb2fZVmo1Wo0NTUhMTER6enpEIlEo1YeGR+EjgOKO8KxjYXuXjPW7j6F3zxqnU6Vc/RsG35/TIeXVy9E9sJZDtv//VEpHrs/lt9+8uJ1/OGYDgc3pDkkKhR7hEPnPcJZOisN2xYXYufnJciYnYm6yyf5xEQIXI8j29lqxWIxcnNz+ec6nQ6HDx8WpD6jyeOWEU/XCnE29oS6aQ0uODjY4/e2trZCLpfj0qVLCAkJgclkQkJCAmpqahAfH+/18sj4IXQcUNwRjm0shIcF4TePSvn1GzKTYlB7/gpeParDI8kxePVoA8JCApGZFIM9x7/Gu6cu4YkHJHj1qA6X23uwafl8nLx4HUVHziN3aRye3K+1mwVpYHlkYqPzHrG1dFYaMmZn4oT+L8iKWyFYIgJgQk3w5FEyMpK1Qpx1x6IuWoObOXOmR+9jWRZyuRzNzc0wm80wmUwAgObmZigUCjQ2NjptIfG0PDK+CB0HFHeEMzAWMpNiUPzEQhQdOY/lKTNxvOEqfpYWjw8/t7aq/+/DXyJFEoVzLe34WVo8Fs2Zgj+fu4x3T13CeX0HLvzdiNylcVCe1uOFx6QOXbUo9giHznvE1ukrWtRdPomsuBWou3wSS2YtESQh4bpjlZWVjXpZ/sCjCa4ZhsEvfvEL5OTkYPPmzXj66aehVqu9XTfyA0+7sanVarS0tMBsNtttN5vN+Pbbb13+n1G3OQIIHwcUd4TjLBYyk2KwPGUm/nzuMh6+Owa19VeQuzQe/Ragr9+Ccy3tmBkdhj+fu4ztlV8iOCgAM8RhONfajqkRIVCe1iN3aRzKai6iu9c8ZHlkYqLzHuHYjhHZct9WvsvWwEHtI8GtH8IwDMrLy1FeXo6SkhKUlpairKwMEonEbj+j0Yjy8nIA1tgZuG2s8qhlhNYKGRuampoQHByM27dvO7wWEhKCpqYmZGRk+KBmhBAyPCcvXsfxhqv4x0WzcbzhKt/SETtlEi7d6AYAXO3oRYAIeC1vEf6z7ls0MAZEhAXhcvstzIuJgPK0HsVPLBzWOhCEkImnp68He754026MCDeGZM8Xb2LhtHu8MohdKpW6NXWvs/0kEsm4mfZ3xGdkbq0QrtsW8T7bwUvDkZiYyHfNGshkMiExMdGr5ZHxReg4oLgjnIGxwI354MaMZCbFoOjIeSTOjEADY0BkWBA6f2jtsLDAH6su4KqhF3HTJkN/swfTIkLQfL0LjyTH2A2Cd1UembjovEcA62xae5bvd0g4ls5K81oiQn7kUTetCxcuODxefPFFb9eN/MDTAW7p6elISEhAUJB9zhkUFIS5c+ciPT3dq+WR8YUGchJfsY2F7l4z/nBMxycigLXL1qrFsWhgDEiYHo7OXjNsR79dNfQiRhyK64bbWJwwBTe7TJgXE4FPGq+jtv7KoOWRiY3Oe4TjKuGgRMT7PGoZWbVqFaRSKViWBWBdfd3Vha0zhw8fhkgkwpo1a1BTU0OrsA/hu+++Q0RExLDfJxKJ+L+v7Wxac+fORU1NjcvpfT0tj4wvQscBxR3h2MZCeFiQw+xX3b1m1NZfwRMPSPDx/1xGYIAIASIRpkaG4GqHdarU68bbeCQ5BtpvvuO7dv0sLR5lqotIv2u63fEo9giHznuECM+jZKS4uJgfMwJYR/27Oz1vaWkp4uLi+OdyuRyVlZVYs2aNJ1UhQ4iPj8eFCxeGtc4IIYT4k4FjPMLDgrBFkYTffdSAsOAAbF+ViuovL+OTxutIiAlH23e3wLIWfNJ4Hb9YloBNy+fzXbte+KnjbFqEEEJ8x6NuWraJCGDt8+huv8fU1FSsXbsWsbGxQ+9MAIx86j+RSISMjAysW7cOGRkZQyYiNNUgAWiKS+I7Q8VCd68ZZaqLeGXNPfjo1w+h19SPTxqv45HkGBh7+lD4D3fDbAEeSY7Bn89dRnevmZ8euEzlOJsWxR7h0HmPEOF5dHvowIEDDtv0er1b64W0tbU5bGttbXW7bJ1OB41GA7FYDKPRiPz8/CHf89xzz43pGQeMRiOmT58+bssj/slVHFg6OxEQGen29pGWRyaeoWLBtutWd68ZZTUXUfy4dSX27l4zwsOCEBYciLKaizi4IY1vCclMisH9c6Y4tIxQ7BEOfd8SIjyPWkYqKirQ3t7OP1iWRUFBgVvvjY2NxVNPPYW3334bL730ErKzs4c13qS0tBT5+fnIzc2FWCyGUqkcdP/y8nKnCdBY0tPTM67LI/7JWRxYOjtxLfNh3Kqpsdt2q6YG1zIfhqWz0277SMsjE5M7scAlFFxikr1wlt327IWzcHjzMsREhTl933DLIxMDfd8SIjyPx4x4umq6XC6HRCJBVVUVAGsrC7eoy1A0Gg0ibe68pqSkoLS0FLm5uU731+l0bh/bnwUGBo7r8oh/chYHAZGRiH7tVbT/cjPw5m6EymS4uiQN7K1bmLL3Lb5l5FZNDTr+7QXMOPlXt1tLKO4IZ7ix4GoMiLtjQyj2CIe+b8lYxzDMmLv29ahlJDU1FV1dXS4fO3fudPnerVu3wmAwoKCgAAUFBcP6g+l0OkRHR/PPxWIxGhoaXO7f0NAwLhZjFHp8DY3nIYDrOJgkl+OON3ej/ZebYXjpt2Bv3bJ7/VZNDdp/uRnRr73qlfLIxEPnPOIrFHtESDqdDs899xwWLFhgtwJ7UVERjEbjsPcDgKKiIqe9hgYeA7AmLgO3+YLH3bQWL16M1atXY9WqVVi8eDGysrKwevVq/OQnP8Hbb7/t8r0pKSkOrSruzsQ1HEql0mWLyVgznDE1Y7E84p8Gi4NJcjlCly9Hj/IwwuRyiCZNwvfPPIv2/7UN3z/7K0SXvg721q1hdd2iuCMcOucRX6HYI0KSSqXYuHEjJBIJ8vPzkZ+fj8LCQhQUFGDdunV8ouFqP6lUinXr1vHHMxqNSE5OhkqlclmWWCzmx1tLJBKHbb7gUTctiUSCzz77zK7LlO16IYcPH3b5XrFYjJdeegkpKSmIiooCYE0c3O321dHRwf88MBvkjMUmKkLGkq73KtB77BiC5s9H77FjCPuHHPSqatGjtH722wueB27dwh1v7h5R1y1CCCFkohGLxcjJycGLL7446ARMEokEOp2Of67RaFBYWIgHHngARqPR7Zlufc2jlhGj0WiXiACwmy524NS/tsrLy8GyLOrr61FXV4e6ujowDONWuVKp1O65wWBASkqKw34ajQYMw0CpVGLfvn38z2PVwL/1eCuP+CdXcXCrpgbG7UUI+4ccmL/+GggMRO9/VwMWi3UHkQgY0HWr56OP+K5brhIRijvCoXMe8RWKPcIx9ZiGtd3bFAoFamwmi3FGrVY77QWUlpaG6urq0aqa13nUMlJfX++wSOH58+eRnZ095HudDX53t5uWTCZDaWkp/5xhGCgUCgDWBCQlJQVisdjuP0an00Gr1Y7pLlthYWFD7zSGyyP+yVkcWDo70fFvL+CON3djTtzS2QAAIABJREFUklyOjuJidO/7oZ8py9r/GxqK9m2F6H6vAqY6Nf+e4ZRHJiY65xFfodgjgDXhUD57FMs2LcGcJT/2tGk5w+DUnjPIfesxhEwOGdU6cL2HbFs4DAYDf3PdaDQiNTWVvw42Go18r6C8vDyXEzwZjUa7G/R6vX5Ufw93eJSM5OXlYcWKFXxLRWNjI15++WW33uusO9ZwulS98sorKCkpQVxcHPR6PQoLCwFYW1xyc3P5/xTAmojs27cPOp0O5eXlXusP19PXg8nBk93ePlI3btxAfHy814/rL+UR/+QsDgIiI+26WYl//Wt0lx/4sVXEVmAgghemwlSnRkhGOib/0z8NuzwyMdE5j/gKxR4BgJDJIVi2aQlO7KxD1rYMzFkiQcsZhn8+2okIYE08ADh0tXJ1c726uhpGoxENDQ0wGo3Q6XROhy04u2k/2PAKIXiUjCQnJ+Odd97hm4+2bds2aEJRW1sLmUyGiIgIVFZWOryuUqmcLqTojFQqdeiuBQAHDx50uq+3FjvkEo2evh48czwfz977KyydlcZvP31Fi7e+eAN7l5ePSkJCiL+w7WbV+eab1kRk8mTAdr78gADg1i30nT6DkIx09J39H9yqqRm0ZYQQQgjxF3OWSJC1LQMndtZhXsYcNNe18ImJEDQazbBmhI2KirJLMhiGgUql8unAdHd5NGYEAE6fPo2oqCisX78eFy5cGHTfvXv3or6+HgDw3nvvobW11e7R3t7uaTUE0dPXg03HN+D0FS16+nrQ19+Hks/+gAPny7Hp+Ab8jfkrXv/0NfT19/H7e9OMGTO8ejx/K4/4p6HiwPDaa+ja/SZC5SvsExGRyNpVi2WB4GBMfbucnwr41iD9XynuCIfOecRXKPaIrTlLJJiXMQdfnWjGvIw5giUiXFeq4uJit/ZnGMahBUWhUIyZ8dIetYyUlpYiLi6OH7SenZ2NyspKh3EknA8++ID/eceOHUhOTrZ7vbGx0ZNqCGZy8GRsuveXKPn0DwgJCkVOwkq8/80RfPztR4iPnIP/OLsTFliw5Z5f4/zNL7HnizexZ/l+r7WQdHV1CdqvVOjyiH8aLA7MV66g6809CN+Qb51BKygICAhA0Ly5MF+4aN0pMBAICMBtjcbaIvLmbnT82wsIlcmcDmKnuCMcOueRgYTqHk2xR2y1nGHQXNeCBVnz0FzXgjlLYr2akHDDCRiG4df56OjoQGdnJ8rKyvheRwzDYN++fTAajQ7DDjQaDYqKiiCRSOxaUlQqFRiGQUlJCTZu3Oj0GIMdV0geJSOpqamQy+UerQ8SFxfH/9zY2Ii2trYxsTDhwmn3ICQoFLfMPfig6X0EiALQz/ajtbMFAJAdL0doUCh2fl6CbYsL+S5d3936DhKxY+AO5wTa3d2NadOmuXzd0tnp9OLO1faRlkcmhsHiIGjWLMz49DQCIiLQU1EBBARgyt630G0GzM8+AwAQTZ6E6N9bV2vvLtuNaf8gd5mIDFUemViEjgWKPf/29fdf47d12/HcA1uxdNaP407/3HwU/1f3f/B2zkGvJSQUe4RjO0ZkzhIJ5iyJtXvuDe4OJ5BIJC73k8lkOH78uMP24uJiu5YVZ2UNdlwhedRNq62tzWGbuwv32E41lpycjOzs7DHRjDQ5eDK23LcVAGBhLbBwMwb9oLa1Bn/49PfYtrgQS2el4fQVLX5R/S/45f97Bn9pqbXb9/QVLTYd3+B2d66AANf/TZbOTlzLfJjv/sItKnerpsZuwbnBFpsbTnlk4hgqDoJmzQIA3AoXY8ret2C+/0H8/KQRja/txZT9e3ErXIywrCzoXtmNn580ovNm+6DJMcUd4QgdCxR7/osxMig4+Wt0Wbrw6pkdOH3FehP0+b9tw/76feiydKGpvclr5VHsEcA6m9apPWfsEg9uDMmpPWcEm953ovCoZSQ2NhZPPfUURCIRVCoVtFrtoLNpMQzDjytRq9X8dGWAdbYAf5hWzF0BogBrd3hYZw+6I/QOtN+2jnnpZ/vx9fdfAQD+8Onv0c/2QzZLhv31exEZGsknKbatJ+4YbHKAgMhIRL9mvftseeVlGF8vRdjKHNxSHuYXnOv56CMYXn4FUw9XIOSuu0ZUHpk43ImDW8GT8Ks1v8Ov70xG9rQ78Ju8B/Hb//4GudPjcHTN77DpawP+2ByI3+Y9iMhpd7g8DsuyaG1txYkTJ5CYmIj09HS7tYvIxCL0OYjOef5r6qSpCA4IRp/FOibz1TM7MD0sBjd6rwMAZoXPwsKYhV4rj2KPANbZtJxN3ztniQR3ps4QZDaticSjZEQul0MikaCqqgoAcODAgUE/UBKJBJ2dnSgtLUVHRwfq6ur418RiMdavX+9JNQTFJRFbF/0v/PFsKbh2EaPJiEBRIPrZfgDAkW8qIYIILFg8cdca/Kt0Hf/ejNmZqLt8km89cZder7fr3jbQJLkc/a8Uo6OgEIHz5qHn0LsIuvce6zoQLxeje385QrNX4MYjWZj+yYkhE5KhyiMTg7tx0NfPoujIeTRf68Sm5fOR+91tvHvqEgJFwKtHdSh+fCEeXjjL5ftbW1shl8vx7bffIjQ0FCaTCQkJCaipqaEpLycooc9BdM7zX5ODJ6Pwgd/gtTOvwvLDTUAuEZkSOgX7Vrzt1fIo9gjHVcJBiYj3eZSMANYuVrYD0bu6uhARETHo/mVlZdBoNJCPsek9e/p6sOeLN7FtcSEab+pggQXx4jlo62TQz/YjKCAYTySuxkdNH8LMmsGCxazwO/Gv0nUAgKWz0pAxOxMn9H9BVtyKYSUigPWu8WAsnZ0wvPp7AEB/czMgEsH8xZe4nJQM/NA963btX4CQEHy3Zi1mnDo5aHeZocojE4M7cRAeFoSi1Qvx78ov8O6pS9C1GfClvgMi4P+3d+/xTdV53sA/Kb0RmqTcylSbOkgRaMNFC0pTZcah0uK6KzJqmb0JSBHGgTprcXxG7YxVd1ZadgRHuVRE2X0eGxBGn3HatDIysjTFC8ilJ6AWkSZYaYE2SS+Ulmb/yOTQtGkbejknST9vX32ZnHNyfr+cfHvoN78brjqB2RPHYEEviYjT6UR6ejpOnz6N9vZ2tLW5vv08ffo0MjIyYDab2UIyDEl9D+I9z7/NjU3B03f8Gv/+yYse21fN+vmgl8XYI5JevzsrWq1WnDx5UvzJz8/v8zUqlcprItLY2NjfakhCGabE5rRtaG1vxXun/4hFkx7Ay3flI3JEJEJDwrBo0iKUfFuCDue1hd9qmr7DTuEtAK5WlYPnDmB+/D04eO6A2OfVV70leYCrq1bUP/3s2gb3za3zOJG/9UuNfvk/+hzU3ld5NDz0FgdNl9vFx/OmxuB3mbMQogA+P3MJbVc74IQrEam02nDgVG2P5ykvL8e3336L9vZ2j+3t7e345ptvUF5ePuD3QYFH6nsQ73n+b+9X73bb1nkMyWBh7BFJr18tI7m5uaioqICq0x+1586d83kVdqvVCkenP5SLiop8fq2c3qx8A08mr8OPtD8GAGxPfwvHLxzDxiOvoLmtSeyadZP6h3jlyH/i3a9347vGczhce1jsmnVH7B3imBFfW0iUyt7HlrSUlqLpjTcRteZxNG7eCnT5ww4AEBqKMVte92nRub7Ko+GhpzhoutyOh1/9H/zq75Mwb2qMmJh0/n5vUkwUXl5yKw5/ewm57x5H3oMzMG9qTLdzVVVVISwsDK2trd32hYeHo6qqCnfeeeegvB8KHFLfg3jP828vHcrDqXrXlOHKUCUut18Wu2z97pOX8M7f7Rq02bQYe8HF3bLOFih5uK97Xz0c+pWM6HS6bguxlPaymFlnA01k5OJuHel8w1OGKTFj3EyMQIjHGBEArml+P8uHqcaENbOyxcRjbmwKnpy9DpuPvoYZ42b6dAOtra3tse98h8OBhqd/jdGvvYqR6em4vO8v19Z56CQkZrzPq1/3Vh4NHz3FwajIUPzq75OQ++5x/Pr+JOR/YEZTazs63+tP1zbivoK/4oOcHyPvwRl4+U8Ckn84BqMiPW85CQkJuHLF+6wkV65cQUJCwqC+JwoMUt+DeM/zXxa7BZ98/wkA1xiRtxb+Fw7VVIhjSJxwoqq+atAGsTP2gktISAjCwsJw8eJFjB07lt1+JdTW1obz588jMjKy79k5+1NAdHR0t22+DsAaSCIjN2+JgzJMiW0LtqO5rRnjlNfmCp8bm4K3F/6313VG5sam+JyI9CVEpcKEA39FiEqFhrw8r4kIAHRYz+HC8uUY9+abAy6TaN7UGOQ9OAO57x7DlfYOdDiBEAXw25/OwOnzDrz9P2fQ0nYV5V/VYcGMWK+JCACkpqZi4sSJ4pgRt9DQUNx8881ITU2V8m0RkZ8ZO3Is1GFq3HfzP2DJNFd3ZPcYkj8c2YS1t/5yUGfTouATHx+P6upqXLp0Se6qDCsKhQLR0dGIieneK6KrfiUj1dXVeOKJJzB9+nRxW3FxMfbs2dPnaweSyPgrZZiyx0Slp4TjehKR8ePH97rfPX1v09bCaxsVimtjR8LDgfZ2tJZ+iKZ33sGon/3M+4l8LI+Cj7dFOMePH9/r4pzzpsbgR9MmoPR4DUaEACEKBT49fRH7Kr/HI3dNxJ5PLXjFeAqzbhqNGI33FYYVCgVKS0uRnp6OM2fOIDw8HFeuXMHNN9+M0tJSfos1TEl9D+I9z3+5v/Dreh+aG5uCGQsG50u9zhh7wSc8PBwJCQno6Ohgdy2JKBQK8ccX/UpGDAYD0tPTUV9fL27z9QMeSCIzXLW0tPTar7TD4UD9c78BFAqMylqBpt3vQtHRgVH/+s9o3L4DaGlBdP7LaHg+D/aX12Pkfff1Ooi9r/IouDS3NWP1vpVYPetxj3FMJ8+bsenUK926J7qVnajBxydrkZY0AQdO1aIDwAdfnEPGjFisTrsFP50Tj6PV9Vi2rQK71tzltWUEAG666SacPHkSxcXFqKur4zojJPk9iPc8/zYYX+r5irEXvLjApP/qVzKSl5eHlBTPwddms9mn1w4kkRmuGhsbMXbs2B73h6hUiDUdREdjI0JjY6H+t1+K21WPPy5uH3nffeL2gZRHwUUZpsTqWY97TKxwqKYC52rPYfWsx3tMRHLfPY6fpdyEshM1SJ0yHvvNtVAAMB6vweiocHxw5BzarnYg78GZPSYibgqFAjqdjn2nCYD09yDe88iNsUckPZ+TEavViri4OADologArsULfTGQRGa48uUb4hCVSkwyOicbPW0faHkUPJrbmjE3NgVrZmV7LM75fyY/g9tik7t11Wq63I6NxlN45K6JMByqRsrksdhvroVOq0GlxYYQBfCO6SxGhCjwu8xZXmfR8oZxR25SxwJjj9wYe0TS87nNymAwDGi/W0pKCnbv3o3du3cDAMrKyjwWT6TupB5TE+hjeMg3zW3NuNB8Aav3rcQBy8fYXlmIKWOm4i/VH2Jy9C3YeOb3+NjyV6zetxLNbc0er9215i6sTrsFv74/CfvNtZgUE4VKiw233jQaCrj+cQ0NUSD5h2N8rg/jjtx4zyO5MPaIpOdzMlJYWIhp06Z5/Zk6dSreeOMNn86zYcMGj25ZCxYsEBMT8s5isQR1eSS95rZmrPowC2v2P455N/4Irx7diGljpuF43TEooEDlxRPInPAz/OHoJo+uWu41Rg5/65qVZMH0WKTPiMXp2kaEKIAvztYDCuC+W28EAPE4XzDuyI33PJJLX7HQ0XkxYR+2D7Q8ouHA525aDz/8MLKysnrcX1hY2OO+znQ6HdLT01FRMbirpg6Vo0ePor6+HnfccQdOnDiB5uZmqNVqTJ48GYcPHwYA3Hzzzejo6MC3334LAJgzZw5OnToFh8OBqKgoTJs2DZ999hkA12DdESNG4JtvvgEA3HbbbTh9+jRsNhtGjhyJmTNn4tChQwAArVaLyMhIXLhwAWfPnsWsWbPE6ekiIiKQnJwMk8kEALjhhhugUqnw5ZdfAgCmT5+OmpoaXLhwAaGhoZg7dy7Ky8vhdDoxYcIEjB07Vuwel5SUhAsXLuD8+fMICQnBjTfeiEOHDqG9vR3jx4/HhAkTUFlZCQCYOnUqbDYbampqAAB33nknPv30U1y5cgVjx45FXFwcjh07BgC45ZZb0NzcDKvVCsDVKnb06FG0tLQgOjoaEydOxBdffAEAmDRpEtra2lBdXQ0AuP322yEIApqamqBSqTBlyhR8/vnnAICJEycCAM6cOQMAmD17Nr788ks4HA6MGjUKSUlJ+PTTTwG4vnUKCwvD6dOnAQC33norzpw5g4aGBowcORKzZs0SYzEuLg5KpRJfffUVAGDmzJmwWq24ePEiwsPDcfvtt+PgwYMAgNjYWGg0Gpw65ZpKWafT4fz586irq0NoaKjHOjr9IUXc/ePof4H5nBnffVWDSeMmIbw6Ej/GT3A25FsowhSIbB2JB0b8FDrVdJjNZjHunrovERWmctSfUSNCMw4nv/kO98e3AE6gvDYSN6uvYnrEecy4Iwy57x7HL2eHYGxUOOMuAOJOqtjjPY+x542/x175gQNoPXgQNyYkYNz8+TglCACAKc3N+Pa993F58QMIj47G7UlJMB05AowYwdgLkNgjeSmcPo4eN5vNvXan6mu/2/bt2/Hoo4+ioqJCHDtSUFCAnJwcH6ssDUEQsHjxYuzduxdJSUmy1qWurk7S6f+kLi8Y9Td+pI67A5aPUXB4vce2UaGj0NTehHnRP8LxlmPYklbYbRD7gVO1eGbXMXS4V1cFEDpCgasdwD/qb4LhUDXyHnTN/f/yn4ReZ9NyY9wN3EDih/c8xt5ADIfYa6+pwfk5dwBOJyIW3IMrR76As6kJaG0FOlwrwkcXrIfthZegCA3F+NIShERF9Tpek7E3cP4UP9Q/PnfT6ivR8HXcR1xcHJYvX4433ngDv/nNb7BgwQIubNYHqbN+fsswfMz+wRzMGD9TfB4+IhxN7U0AgKqWr/GoLsvrbFrJPxyDEAVwtcOJ0BAF/j1zFv6cczdeengmPvjiHH59f5K46roviQjAuKNreM8jufQWC6GxsYj6xc8BAK1lH0IRHg60tIiJSFjybWj49bNwNjWho6UFtT9Jw/d3zuu1Cxdjj6ifU/sORHp6OrRaLYqLiwG4Wkq0Wm0frxrezp8/L+mUp1KXR/JobmvGo2XL0NzWJG67cvWK+Pif4v4Fm46+gojQCMwY57m4mHssyH233ogPT7i6EIyKDMW8qTHiauupk8f7lIS4Me7Ijfc8kktfsaB5+mkAQOOrr6Hju+889rUddnXNgkIBXLkCZ3s7xrzyn722jDD2iGRIRgBXKwpn0CKS1/ELx9DS1gwnnJgcPRlfN3wt7ps6eiouXr6IX8xai42Hf4+wEWFid60Dp2qR++5xvPDQTMybGoN5U2OQ++5x5D04A/OmxogJyPUkIkREgSL81ltdCYe7l3tYGNDW5np89SoQEgKEhmLMltcxMj1dvooSBQguRxkAxo0bF9TlkfSa25rx+hd/wMgwJW7/wR34uuFrKP72HwCcqj+F6DGjEREagXZnO34+6xdQhinRdLkdL/9JEBMPAJg3NQZ5D87Ay38S0HS5vd91YtyRG+95JJe+YqGltBSXVq66logA1xIRt44OKB9Y5FMiwtgjYjISEFpbW4O6PJKeMkyJLfcU4tW7X4NwUYACCjjhxE/i03D/pEUAgIPVB1Dw2XqsmZWNubGuySZGRYZi15q7ui1kOG9qjM9jQ3rCuCM33vNILr3FQofDgUu/WAu0u750ify7ha5WEC+a9+xFS2npgMojGi4kSUY2bNiAsrIyKYoKSo5+zl8eKOWRPJRhSijDlHB2dGBESCjCQ8JxwPpXJI3TYcb4mZiinIrJo2/B9spCj0UPe0o4Btoti3FHbrznkVx6i4WOxkbXgHUAo1ZmofWTz6CIjHR12fqb8Pl3ux60t+PSY6v7TEgYe0QSJSMajQZ6vR4AvC5wGChrjhAFm+MXjuEqruKpOb9Czpyn4IQTL3/6O5gvClCFqSBcrMT8+DSvM2oREQ0nobGxmPDZJ5jw+aeI/k0uYozFcEZGQqFSAVFRiC5Yj/E7d2L0a68CajUUI0ei/qmn+70gItFwIckI04aGBmzduhXx8fEoLy/vtr+oqAh79uyRoioBKT4+PqjLI3k0tzVj89HX8OTsdWI3rEWTHsC7X+9GKEKx3VqIByc/hD998/9xy5gp4jFDhXFHbrznkVz6ioXQ2FiPx7Gmg+Jz96xZykWLEDl/frft/SmPaDiQpGUkJycHGo0GZ8+ehd1ux9mzZz1+fFx3cdj6rsv0gcFWHslDGabE5rRtYpLR3NaMv1Tvw4xxM9HubMeKmx7DvyYtxZOz12Hz0dc8umoNBcYdufGeR3K53lgIUanEH1+2D7Q8omAk2dybK1asAOB9pXaz2SxVNQJSe3v/ZygKhPJIPp27XynDlFiuW4E/HN2E+fH3oPGyA4dqKjA3NqXbOiNDgXFHbrznkVwYe0TSk3w2LXciUlFRIY4V4ZojvRs5cmRQl0f+4VBNBf5wdBOenL0O2bc9gR+Om4gNn+fjUE2FJGNGGHfkxnseyYWxRyQ9yVcls1gsyM7OFp83NjbizTffRFxcnNRVCRgajSaoyyP5eRs/oovT4Umlq4uWFC0jjDty4z2P5MLYI5Ke5C0jZWVl2Lt3r/hTVlYGg8EgdTUCyvfffx/U5ZH8uo4fAVxxMDc2BZvTtknSMsK4Izfe80gujD0i6UmejHhrAdFqtVJXg4i66Cnh4LS+RERENFQkT0asVmu3bRaLRepqBJSxY8cGdXnknxh3JBfGHsmFsUckPcnHjCQmJmL58uVISkoCAJhMJuTk5Pj8ekEQYDKZoFarYbfbkZWV5fU4k8kEQRDQ0NAAh8OBvLy8Qam/HDi7B8mBcUdyYeyRXBh7RNKTvGUkJSUFzz//PJxOJ5xOJ1544QWkpPi+mFpBQQGysrKQmZkJtVrtdbyJ3W6HIAjIysrCunXr0NDQgPz8/MF8G5Ky2WxBXR75J8YdyYWxR3Jh7BFJT/JkBHCNEcnJyUFOTs51TetrMpmg6rSAkE6ng9Fo7HZcZWUlSkpKxOepqaniNMJEREREROQfZElG+ksQBERHR4vP1Wo1Kisrux2n1+uxceNG8Xl1dfV1tb74G6kH+HNCAQIYdyQfxh7JhbFHJL2ASkauR+df8NLSUixZskTG2gwMpxokOTDuSC6MPZILY49IegGXjDQ0NIiP7XZ7n8evXbsWeXl5Af3tQ1tbW1CXR/6JcUdyYeyRXBh7RNILqGTEPQOXm81mg06n6/H4wsJCPPbYY9Dr9QE9fXBkZGRQl0f+iXFHcmHskVwYe0TSC6hkRK/Xe6xTYrFYkJGRAcA1uL1zS4nRaIRWq4VWq4XFYvE60D1QjBkzJqjLI//EuCO5MPZILow9IukFVDICAC+88ALy8/NhMBhQXV2NzMxMAK5WEJPJBMCVpGRnZyM7Oxtz5sxBWlqaR/euQPPdd98FdXnknxh3JBfGHsmFsUckPckXPezKYrHg0KFDeOihh3w6PikpqVt3LQDYsWOH+Fir1eLLL78ctDoSEREREdHgk71lxOFw4ODBg3JXw6+x2ZjkwLgjuTD2SC6MPSLpSZKMbNiwAWVlZV73JSYmeqwJQt11dHQEdXnknxh3JBfGHsmFsUckPUmSEY1GA71eDwDYvXt3t/1cHb13Uo93CeTxNTR4GHckF8YeyYWxRyQ9ScaMNDQ0YOvWrYiPj0d5eXm3/UVFRdizZ48UVSEiIiIiIj8hSTKSk5ODN954A2fPnoXdbsfZs2c99judTimqEbBuvPHGoC6P/BPjjuTC2CO5MPaIpCfZbForVqwAAJjNZiQmJnrsM5vNUlUjINXV1SE2NjZoyyP/xLgjuTD2SC6MPSLpST6bVmJiInbv3i2OHSkrK+uWnJCnK1euBHV55J8YdyQXxh7JhbFHJD3Jk5GCggKPblkLFizwOqidromIiAjq8sg/Me5ILow9kgtjj0h6kicj06dPx8MPP4y4uDipiw5Y48aNC+ryyD8x7miwNbc197jdva+5rdkjFjpvHyqMPXLjfY9IepInI1artdu2rgPaydO5c+eCujzyT4w7GkzNbc1YvW8lDtVcm8r9QvMFHKqpwGMfrsBjH67Ax5a/YvW+lTj69VE0tzXjY8tf8diHK3DA8jFW71vZLWkZLIw9cuN9j0h6kg1gd4uLi8Py5cuhUChgNBpRUVGB559/XupqEBGRhJRhSqye9Tg2fJ6PNbOykTg2CY+WLYVCEYKRIyLR3tGOjV+8ggTNJJyxf4N///MLaEc7AOCVL36Pp+b8Cscsn2DzyUIsv+VfsaPqv7E5bRuUYUp0OBzoaGxESFQUQlQqscwOhwMAPLb1tp2IiKQneTKSnp4OrVaL4uJiAMD27duh1WqlrkZAiY6ODuryyD8x7miwzY1NwZpZ2Sg4vB4zxs9EiCIEV51XMXP8THzy/Se46ryKU/WnEKVRi4lIiCIEgBNffl+JP329F3d/VIc//GQj1kb9PZRhSrSUluLSL9YCLS3A6NGYUGZESFQUAOD8XfPQcfkyxmx8BRF/W3i31WRC/S+fBMLC8IODBxCiUjH2SMT7HpH0JE9GANeMWpxBy3cKhSKoyyP/xLijoTBP+yOUVZfieN0xzBg/E7dE34J3v/acxKSxzeHxPHFMEvacfQ+6yHjsvxtYvqMak8z5sK22o3HzVqDdlbhEzE5Gbca9QFsbRv9+A8Jvn4PLfy7BpazHgIgIKEJHwHm5FQAw5vcbxJYRxh658b5HJD3Jx4y4VVRUoKKiou8DCfX19UFdHvknxh0NhUM1Ffjy0inMGD8Tx+uO4av6r8R94SHhAIA5mtsBADPGzwQAHL9wDLGjbkBlhxV6pQ4zzc1AezsaX305VGZgAAAVw0lEQVTNlYiEhiLy7xaitexDRNwxB86WFlxasRKX/1yCsOTbgKtXgeZmOBubAABjtryOkenpYrmMPXLjfY9IepK3jFgsFmRnZ4vPGxsb8eabb3J2LSKiIHeopgIbPs/Hk7PXYW5sCp49+Gscv3AMADAqNApN7Y3isQooIFyohAIKKBCCmqbvMGP8TFRcOoWZK+/B1NdLxGOVP12M0f+5Abb/+A80vvoaQmJi0FFbCygUaDt6DAgJATo6gI4OKB9Y5JGIEBGRvCRvGSkrK8PevXvFn7KyMhgMBqmrEVBuuOGGoC6P/BPjjgZTc1szNh99TUxEDtVUQLhYKe7vnIhUth4Xx5MoFCEYERKCyBGROHXxJBYqdNicYMGxGepr596zFy2lpdA8/TRGTPwhOmproYiKApxOV6sIAIwYAYwYgeY/voeW0lKPujH2yI33PSLpSZ6MeGsB4QD23l28eDGoyyP/xLijwaQMU2Jz2jYxESn4bD06nB24f9IiKEcoAQAPTn4I99+8COMxAREhEYgIicDVjnY8ceu/4a2M/8La0Hvw0cVD+Of/a8H/W3IjRmQ/BoSGAu3tuLRyFS6uXImrZ75FSEwMnI3Xkht0dCDq56swpnArAODSqp97JCSMPXLjfY9IepJ30/K2zojFYpG6GgGltbU1qMsj/8S4o8GmDFOKLSQ5c55CgmYyxinH4WdT/wmff/8ZtlcWYnPaNpyznMPylEcB/G0RROU4dDgcSHh6K37bbMNIRQTumvALjH3kH9Ayc45rNq3mZlz+cwki0u9B6/6PAYXC1TIS6vpnrnFrIcZseR1jtryO+l8+ifqnnkaEXo8QlYqxRyLe94ikJ3kykpiYiOXLlyMpKQkAYDKZkJOTI3U1AkpYWFhQl0f+iXFHQ8HdQqIMU3psm6f9EWb/YA6UYUqMihwl7nf/P0Slwg/KD3ZbT2RkejrGF3+Auh//BKOWPoLmD/4MxciRGL3ldbR+8gmatr2B6PyXYXvhJTT86mlM+J8D+MEnFeI5AcYeXcP7HpH0JE9GUlJSEBcXJ44TeeGFFzjNbx8mTJgQ1OWRf2Lc0VDpnIh4295TLISoVF4XKgyfPBkTPv8UobGxUD/9K/HYkenpiMrKQmhsLEbed5+4vSvGHrnxvkckPcnHjFitVpw8eRI5OTnIycmBzWaTugoBx1vXtmAqj/wT447k0p9YCI2NBdA9Yelp+0DLo+DE+x6R9CRvGTEajbBYLFiwYAEAV0vJ7t278dBDD0ldlV65+3GePn1a5poANTU1aOw8GDPIygtG7ri53v7AjDvG3UD0N+46v4axR/3B2AuM8oLRQGKP/IPkyUh8fDxWrFghdbHXzf1txbp162SuCQUyq9WK22677bqOBxh3NDDXG3fu1wCMPRoYxh7JpT+xR/5B4XQ6nVIWWFBQ0G3Aurdtcrt06RIOHjyIuLg4REREyF0dCjCtra2wWq248847MWbMGJ9fx7ijgehv3AGMPRoYxh7JZSCxR/5B8mTEbDYjOztbnE3LbDbj+eefR0pKipTVICIiIiIimUmejACAw+EQZ9NKT0/noodERERERMOQLMkIERERERGR5FP7EhERERERAUxGiIiIiIhIJn6RjHDRHyIiIiKi4UeSdUbKysqg1+sRFRWF3bt3d9tvNBqxfft2KapCRERERER+QpJkZMuWLVCpVEhJScE777wDvV7vsb++vl6KahARERERkR+RZZ2RxMTEPrcREREREVFw84upfRsbGxEVFSV3NQKOIAgwmUxQq9Ww2+3Iysoa0HH+xtd6m0wmCIKAhoYGOBwO5OXlSVzT4Yex58LYkx5jz4WxJy3GnQvjjoaEUwYWi8VpNpvFn9zcXDmqEfCWLl0qPi4qKnIWFRUN6Dh/40u9bTabc9u2beLzNWvWONevXy9J/YYzxh5jTy6MPcaeHBh3jDsaOpLPppWbm4tly5bhmWeeEX+MRqPU1Qh4JpMJKpVKfK7T6bxeR1+P8ze+1ruyshIlJSXi89TUVFRUVEhSx+GKsefC2JMeY8+FsSctxp0L446GiiQD2DvT6XTdmvVKS0ulrkbAEwQB0dHR4nO1Wo3Kysp+H+dvfK23Xq/Hxo0bxefV1dVISUmRpI7DFWPPhbEnPcaeC2NPWow7F8YdDRXJW0Y6B7xbfHy81NWgIKLVasXHpaWlWLJkiYy1oeGEsUdyYeyRHBh3NBQkbxmprq7GE088genTp4vbiouLsWfPHqmrEvAaGhrEx3a7fcDH+ZvrrffatWuRl5fncbOkocHY88TYkw5jzxNjTxqMO0+MOxpMkreMGAwGxMXFob6+Xvxxyj+hV8BJSkryeG6z2aDT6fp9nL+53noXFhbiscceg16vh8ViGerqDWuMPU+MPekw9jwx9qTBuPPEuKPBJnkykpeXh5ycHI+fF198UepqBDy9Xg+r1So+t1gsyMjIAOAajOb+ZqO34/yZr+8PAIxGI7RaLbRaLSwWS0AMGAxkjD3GnlwYe4w9OTDuGHc0tPxinRGr1Yq4uDi5qxFwBEFAcXEx4uPjUV1djXXr1gEAli1bhszMTPFm0tNx/s6X92exWJCWlubxuhUrVgTMewxUjD3GnlwYe4w9OTDuGHc0dCRJRsrKyqDX6xEVFYXdu3d32280GrF9+/ahrgYREREREfkRSQawb9myBSqVCikpKXjnnXeg1+s99tfX10tRDSIiIiIi8iOSd9Mym81ITEzscxsREREREQU3vxgzQkREREREw4/ks2m5cYwIEREREdHwJkkykpubi2nTpuGJJ57AyZMnAQDHjx/H1KlTMW3aNCYmRERERETDkCQD2BcuXIglS5Z4jAvRarXYu3cvLBYL1Gq1FNUgIiIiIiI/IknLiMPh6DZAffr06UhMTER6ejocDocU1SAi8muCICA3NxdTpkxBfn6+rKsbL168uNuCZlxtmYiIBpskyYjNZuu2LT09vdf9RETDTVJSErKysgAA69atg1arla0uOTk5HtOwC4IAQRBkqw8REQUnSZIRu90+oP1ERCQtvV4vdqG12+0oKCiQuUZERBSMJElG6uvrcerUKa/7Tp48yUUPiYh6YDKZkJaWhsLCQhiNRhgMBuTm5nocYzQaxf35+fnidne3L4PBIB5jt9s9zgkABoMBU6ZMEVs+BEHA4sWLYTAYxDrYbDYUFxfDYDBg3759mDNnDtauXQvA1X2r8/mIiIh8JckA9pycHNxzzz3IyMiAXq+HVquFxWKByWRCaWkpysrKpKgGEVHA0ev1SE9Px4kTJ8QuXGlpabBYLNBqtRAEAQaDATt27ADg6vZqMBig1+uRnZ2Nffv2AXC1bsyZMweZmZniOd0yMzPFxANwdRdLSUkRn2dkZMBkMkGv1yMjIwMAcObMGXEMiVarRVZWFjIzM4f2YhARUdCRbJ2RPXv2oLq6GsuWLcM999yDZcuWwWw2Y8+ePVJVgYgoYE2fPl18rFarxe6txcXFAFytFyaTSTzGaDR6TByiVqsHdQxKZmYmSkpKALhaUjqPLyEiIvKVJC0jgOsfwo0bNwIAzGZzt9m1iIiof7RabbdkYCi7TAmCILaeGI1G2Gw2tooQEVG/DGnLiNVq9bq9t0TEvSgiERH1bcmSJR4tIoCrVSQjIwNms9lje+epeaOjo8XHdru9z5myVCqVOPOh+zxLlizBtm3boNFoBvQeiIho+Brx29/+9rdDdXK1Wo2CggKMHj0a48eP7/P4Xbt2ISIiQtbpLImI5CIIAnbu3AlBEHD58mXEx8fDarVi586dqKqqgk6nw5EjR/Dee++htbUVycnJiImJQUJCAoqKimC1WiEIAhYtWgSNRiNudzgcqKqqgslkwiOPPIKIiAhMnjwZRqMRERERqKqqgt1ux4kTJ5Camoqqqirs3LkTVqsVycnJ0Gg0iI+Ph9FoRFVVFRYuXCjeqwsLC5GXlyf3pSMiogClcDqdzqEuZNeuXTCZTOJCh9HR0VCpVHA4HLBYLDh48CAUCgXWrVuHqKiooa4OEdGwlJaWhr1794pT9g4Gg8HALlpERNRvkiQjbhUVFRAEAdXV1bDZbNBoNJg+fToyMjKgUqmkqgYR0bA0Z84c/OUvfxlwMlJYWCiOURnsgfFERDS8SDaAHQBSUlI8poskIiJpGAwG2O12bN26FevWrRvQubRaLSorK6HRaMSpfomIiPpD0pYRIiIiIiIiN8nWGSEKdJ1nIiIiIum419UhouDDZITIB/n5+ewXT0RBx263w2AwwGg0dttnNBr95ksYdzdDIgo+Qzq1L1EwMBgMSE1NRUxMDCwWCzZs2IBVq1Zh8uTJSEhI6HZ8Wloa3nvvPQCATqeTurp+LT8/H2azGR999BFSU1P7fUx/WCwWj/UwFi9eLE5/K7XBfI9D/T68nd+frqU/MZlMWLZsGSIiIqDT6bo970nX6znYdVq8eDHi4+O9fj6rVq3CM888g9raWo9poA8fPoydO3di5cqVQ1Kv661ncnIyNm3aNKj3BCLyD2wZIepDeXk5kpKSALgG7ubk5CA9PR0Gg6HbsSaTCYmJiUhJSeF0p124VwTPzMzE9OnT+31MfwiC0G1Rv5ycnG6rlkthsN/jUL+Pruf3p2vpb/R6PdLT03t87o236znYdVq4cKHXfRaLRZzJUq/XIz4+HhkZGcjIyIDBYMDGjRuHrF7XU8/O/KWlhogGD5MRol6418fp6t5770VlZWW3fxjtdjuio6M9Vrcml4aGBsTHx0OtVvc4A5Mvx1wvu92OgoKCbtv1ev2grrfhq8F+j0P9Pjqf39+uZaDr6XrKxd06k5+fj8zMTL/rmnrvvfd67U5GRIGN3bSIemE0GpGUlOTxj3JrayssFgsmT56Mzz//XOw24O7PLAgCNBoNkpOTPc6zf/9+1NXV4Y9//KP4Gnc/aKPRiCNHjoivcXfvaG1tRV1dHQ4fPoxdu3bh7rvv7rWuXcvo3E2koqICb7/9NjQaTbdtCxcuhMlkgtFoRF1dHfbv34+YmBhoNBqv5+jpG8yeziEIAt5//32cP38edrvda5eVrsfY7Xav5fZ0zdznePvtt+FwOMTyjxw5ApPJJJ43Pj4eVVVVWLVqFYBrXen6ev/X81kM5Dp0Lq+1tRUmkwm7du1CQkICBEEQV1rX6XQQBMHjffhS1+v5nOPj4z3Ov3//fp+upbdYtNvt2LlzJ+rq6lBVVYX9+/d7fHZu1/v5CoJwXfHsrR6TJ0/2qW6+1NFkMkGj0XjEVefnnXm7nhEREX1+RlVVVaitrYUgCH3GorscrVaLhIQErF27Fhs2bEBMTAySk5Oxa9cuLFq0CIIgQKvVwmq14qOPPupz+uee7gsDuaf1VE93t63IyEgUFBSw1Zko2DiJqEfPPfecs7Ky0mObzWZzlpSUOG02m3P27Nni9pKSEvE127ZtE7dXVlY6ly5dKj4vKipyFhUVOZ1Op/OBBx5w2mw2p9PpdC5dutRZXl4uHrd+/XrnmjVrxOfz5893VldXe61nb2U899xzzueee87pdDrF83fdVl1d7fF6d906X4eu5+iqr3OsX79erFNPuh7jrdyerpnNZnPOnz/f47WdPxP3Y29l+VJ3Xz+LwboOXctzv6Zr3HU9X2917c/n7O0z6e1a9hSL27Zt84idnq5Bfz9fX+PZWz18rVtfdex6Lbw976rr9fTlM+p8fykqKhLfe19lVFdXe7zWXV5RUZF4n1u6dKnTZrOJ23v6fXef19ffT6ez79+j3urZ+TVEFFwkXfSQKNA0NDT02P1ErVZDp9PBaDQiIyOjx+OKi4sBuL4Z7Oqtt95CSUmJ+LzrbDGdu4ip1eoeZ5PprQwA4piXzv36O2/Lz89HYmJit9eZTCbxNd7O0VlRUVGf5+iPruX2dM0MBoNH+Xl5eT6X4Uvdff0sBus6dC3P/c16b2X39Fr38b7Ura/PuS89xaJer8fSpUuh0+mg1+uRlZXl9fX9/XyvJ5671kMQBJ/q1lcdB4Mvn1Hne41er0dubm6f8X7ixAkUFBRg7969Htu1Wq3Y0pCfn4+srCyo1WpkZ2djx44dMJlMEARBvL5d+fr76dbX71FP9SSi4MVkhKgX0dHRvf6hkZWVhcLCQqjV6l7/eNNqtd322+12LF26FBs3boRWqx3wIFZvZbh5m6nnemfvGarZfq6n3MG4Zr39YUXX9zn3dC17ivfPPvtM7IK0du1abNq0qdsx/f18fa23VqvtVo8XX3yxz7oNRh37MtBzLVu2TBzHlpiY6PEeoqOjkZOTIyYZPZWt1+s9xsLp9XoYDIYef2cG+/ezr3pybBJR8OEAdqJeJCUl9Tp7i16v9zqQvaGhQXy8ZMmSbt8SG41GsS+5ezyKxWKBzWbr1wDNnsrw1b333ouKigqPbXa7/bq+HR+Mc/Slt2uWkZEBs9nc7XgAUKlUsNls4muGsu5SXIf+Goy69XUte4rFrVu3imX19C1+fz/frnp7n97q4UvdfKljf3S9nr58Rp2/IDEajWLLxo4dO7Bv3z7s27evWzKl1WqRkZEBlUrldSbAgoKCHseJ+JroDca16a2edrsdcXFxPp+LiAIDB7AT9SI+Ph4Gg0EcZCkIAvLz88WBllqtFgqFAsnJyYiJiYHBYMD777+Puro6cd0F9//dA48FQcCiRYsQExOD8vJy2O12cUDyRx99hOTkZHGQbVVVFXQ6HY4cOYL33nsPra2tSE5ORkREhEc9eyrDZDJh586dsFgsiImJgVar9brN/X/360tKSvDss8+Kg2a7Hu9Nb+cQBAE7d+4UB8l6O0fXYywWi9d69nTNdDpdt2vgHmgfHx8Po9GIqqoqLFy4EFVVVdi5cyesViuSk5ORkJDQZ919/SwG6zp0Le/8+fNITU3Fpk2bcOTIEURGRiIyMtLjfVit1l7r6i7T18/ZXRf3+TUaTZ/X0j0AuWssms1mWCwWWK1WHD58GKmpqd3ef38+3+uNZ2/1cDgcfdbNlzq6f2+7fh6dr5+3e0zn69nbZwS4BnlXV1dDoVDg8OHDOH/+PJ599lmvdQUgXp+GhgbodDo4HA689NJLiIyMFAeX5+bmIicnRyxDo9Hg/fff9xjY3rXuPV33/t7T3Oua9FZP92QDw309G6Jgo3A6nU65K0Hkz3zpj01EJIXc3FwkJSUN6oxS7panzux2O0pKSnrt/im13rrPEVHgYjctoj7k5OR47dZARBQMvK13o1arkZmZ6TeJiMlkwr333it3NYhoCDAZIeqDWq0e9IGqRETXy2QyiYPsh9NK5Ha7HRaLZdAWQiUi/8JuWkREREREJAu2jBARERERkSyYjBARERERkSyYjBARERERkSyYjBARERERkSyYjBARERERkSyYjBARERERkSyYjBARERERkSyYjBARERERkSyYjBARERERkSz+F5E6PebBZgzoAAAAAElFTkSuQmCC\n","text/plain":["<Figure size 800x200 with 4 Axes>"]},"metadata":{}}]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":173},"id":"uFbO-z506Mc4","executionInfo":{"status":"ok","timestamp":1636857253078,"user_tz":-60,"elapsed":34,"user":{"displayName":"Lisette Espin","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GjDac6esfgXyAYCvmgTKR1CnSZMB2A0Bt49sHtuDA=s64","userId":"15904188957194365368"}},"outputId":"a9eb3fce-270d-4c4f-e757-41323669d392"},"source":["vtype = 'me'\n","metric = 'pagerank'\n","paper.best_fit(df_rank_fit, df_rank_empirical, vtype=vtype, metric=metric)"],"execution_count":15,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>dataset</th>\n","      <th>metric</th>\n","      <th>kind</th>\n","      <th>gini</th>\n","      <th>mae</th>\n","      <th>me</th>\n","      <th>dif</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>1</th>\n","      <td>aps</td>\n","      <td>pagerank</td>\n","      <td>DPA</td>\n","      <td>0.530178</td>\n","      <td>0.012927</td>\n","      <td>0.002313</td>\n","      <td>0.098323</td>\n","    </tr>\n","    <tr>\n","      <th>5</th>\n","      <td>hate</td>\n","      <td>pagerank</td>\n","      <td>DPA</td>\n","      <td>0.321931</td>\n","      <td>0.011134</td>\n","      <td>-0.005928</td>\n","      <td>0.075220</td>\n","    </tr>\n","    <tr>\n","      <th>11</th>\n","      <td>blogs</td>\n","      <td>pagerank</td>\n","      <td>DPAH</td>\n","      <td>0.559996</td>\n","      <td>0.013472</td>\n","      <td>-0.004671</td>\n","      <td>0.070994</td>\n","    </tr>\n","    <tr>\n","      <th>15</th>\n","      <td>wikipedia</td>\n","      <td>pagerank</td>\n","      <td>DPAH</td>\n","      <td>0.695099</td>\n","      <td>0.045760</td>\n","      <td>-0.045760</td>\n","      <td>0.088897</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["      dataset    metric  kind      gini       mae        me       dif\n","1         aps  pagerank   DPA  0.530178  0.012927  0.002313  0.098323\n","5        hate  pagerank   DPA  0.321931  0.011134 -0.005928  0.075220\n","11      blogs  pagerank  DPAH  0.559996  0.013472 -0.004671  0.070994\n","15  wikipedia  pagerank  DPAH  0.695099  0.045760 -0.045760  0.088897"]},"metadata":{},"execution_count":15}]},{"cell_type":"markdown","metadata":{"id":"Y0V2Fm6i6Mc4"},"source":["-----"]}]}