Cortex-clustering.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Clustering results of scPI-based methods (Cortex)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import torch\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "plt.style.use('ggplot')\n",
    "import time\n",
    "from scPI import scPI_FA\n",
    "from scPI import scPI_ZIFA\n",
    "from scPI import scPI_ZINB\n",
    "from utils import *\n",
    "from sklearn.decomposition import FactorAnalysis\n",
    "from sklearn.metrics import silhouette_score\n",
    "from sklearn.cluster import KMeans\n",
    "from sklearn.metrics import normalized_mutual_info_score as NMI\n",
    "from sklearn.metrics import adjusted_rand_score as ARI\n",
    "def cluster_scores(latent_space, K, labels_true):\n",
    "    labels_pred = KMeans(K, n_jobs=8, n_init=20).fit_predict(latent_space)\n",
    "    return [silhouette_score(latent_space, labels_true), NMI(labels_true, labels_pred), ARI(labels_true, labels_pred)]\n",
    "n_celltype = 7\n",
    "n_repeat = 10\n",
    "c_train = np.loadtxt(\"/Users/jiazhao/Downloads/Results/label_train\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### example of performing scPI-FA, scPI-ZIFA, scPI-ZINB (vary n_epoch and linearity)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "When evaluating the impact of the number of epochs, we vary n_epoch to be $\\{250,500,10000\\}$. When evaluating the impact of the linearity and non-linearity, we vary non_linear to be False or True. We repeat the experiments ten times to draw boxplots."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(1234)\n",
    "torch.manual_seed(1234)\n",
    "file_name = \"scPI_FA\"\n",
    "for i in range(n_repeat):\n",
    "    result, result_test = scPI_FA.fitModel(Y=expression_train, K=10, batch_size=128, n_epoch=500, lr=5e-4, non_linear=True, Y_test=expression_test)\n",
    "    scores = cluster_scores(result[\"latent\"], n_celltype, c_train)\n",
    "    np.savetxt(str(i) + file_name + '_scores.txt', scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(1234)\n",
    "torch.manual_seed(1234)\n",
    "file_name = \"scPI_ZIFA\"\n",
    "for i in range(n_repeat):\n",
    "    result, result_test = scPI_ZIFA.fitModel(Y=expression_train, K=10, batch_size=128, n_epoch=500, lr=5e-4, non_linear=True, Y_test=expression_test)\n",
    "    scores = cluster_scores(result[\"latent\"], n_celltype, c_train)\n",
    "    np.savetxt(str(i) + file_name + '_scores.txt', scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(1234)\n",
    "torch.manual_seed(1234)\n",
    "file_name = \"scPI_ZINB\"\n",
    "for i in range(n_repeat):\n",
    "    result, result_test = scPI_ZINB.fitModel(Y=expression_train, K=10, batch_size=128, n_epoch=500, lr=5e-4, non_linear=True, Y_test=expression_test)\n",
    "    scores = cluster_scores(result[\"latent\"], n_celltype, c_train)\n",
    "    np.savetxt(str(i) + file_name + '_scores.txt', scores)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Impact of the number of epochs on clustering scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_repeat = 10\n",
    "models_plot = []\n",
    "asw_plot = []\n",
    "nmi_plot = []\n",
    "ari_plot = []\n",
    "file_models = [\"/Users/jiazhao/Downloads/Results/Cortex/FA/\", \\\n",
    "               \"/Users/jiazhao/Downloads/Results/Cortex/scPI_FA/10K/\", \\\n",
    "               \"/Users/jiazhao/Downloads/Results/Cortex/scPI_FA/500/\", \\\n",
    "               \"/Users/jiazhao/Downloads/Results/Cortex/scPI_FA/250/\"]\n",
    "name_models = [\"CORTEX_FA_scores.txt\", \\\n",
    "               \"_CORTEX_FAVAE_nonlinear_scores.txt\", \\\n",
    "               \"_CORTEX_FAVAE_nonlinear_scores.txt\", \\\n",
    "               \"_CORTEX_FAVAE_nonlinear_scores.txt\"]\n",
    "models = [\"FA\", \"scPI-FA-10K\", \"scPI-FA-500\", \"scPI-FA-250\"]\n",
    "for i in range(len(name_models)):\n",
    "    for j in range(n_repeat):\n",
    "        res = np.loadtxt(str(file_models[i]) + str(j) + str(name_models[i]))\n",
    "        asw_plot.append(res[0])\n",
    "        nmi_plot.append(res[1])\n",
    "        ari_plot.append(res[2])\n",
    "        models_plot.append(models[i])\n",
    "\n",
    "res = np.loadtxt(\"/Users/jiazhao/Downloads/Results/Cortex/ZIFA/CORTEX_ZIFA_scores.txt\")\n",
    "for j in range(n_repeat):\n",
    "    asw_plot.append(res[0])\n",
    "    nmi_plot.append(res[1])\n",
    "    ari_plot.append(res[2])\n",
    "    models_plot.append(\"ZIFA\")\n",
    "\n",
    "file_models = [\"/Users/jiazhao/Downloads/Results/Cortex/scPI_ZIFA/10K/\", \\\n",
    "               \"/Users/jiazhao/Downloads/Results/Cortex/scPI_ZIFA/500/\", \\\n",
    "               \"/Users/jiazhao/Downloads/Results/Cortex/scPI_ZIFA/250/\"]\n",
    "name_models = [\"_CORTEX_ZIFAVAE_nonlinear_scores.txt\", \\\n",
    "               \"_CORTEX_ZIFAVAE_nonlinear_scores.txt\", \\\n",
    "               \"_CORTEX_ZIFAVAE_nonlinear_scores.txt\"]\n",
    "models = [\"scPI-ZIFA-10K\", \"scPI-ZIFA-500\", \"scPI-ZIFA-250\"]\n",
    "for i in range(len(name_models)):\n",
    "    for j in range(n_repeat):\n",
    "        res = np.loadtxt(str(file_models[i]) + str(j) + str(name_models[i]))\n",
    "        asw_plot.append(res[0])\n",
    "        nmi_plot.append(res[1])\n",
    "        ari_plot.append(res[2])\n",
    "        models_plot.append(models[i])\n",
    "\n",
    "for j in range(n_repeat):\n",
    "    res = np.loadtxt(\"/Users/jiazhao/Downloads/Results/Cortex/ZINB/\" + str(j) + \"CORTEX_ZIFA_scores.txt\")\n",
    "    asw_plot.append(res[0])\n",
    "    nmi_plot.append(res[1])\n",
    "    ari_plot.append(res[2])\n",
    "    models_plot.append(\"ZINB\")\n",
    "    \n",
    "file_models = [\"/Users/jiazhao/Downloads/Results/Cortex/scPI_ZINB/10K/\", \\\n",
    "               \"/Users/jiazhao/Downloads/Results/Cortex/scPI_ZINB/500/\", \\\n",
    "               \"/Users/jiazhao/Downloads/Results/Cortex/scPI_ZINB/250/\"]\n",
    "name_models = [\"_CORTEX_ZINBVAE_nonlinear_scores.txt\", \\\n",
    "               \"_CORTEX_ZINBVAE_nonlinear_scores.txt\", \\\n",
    "               \"_CORTEX_ZINBVAE_nonlinear_scores.txt\"]\n",
    "models = [\"scPI-ZINB-10K\", \"scPI-ZINB-500\", \"scPI-ZINB-250\"]\n",
    "for i in range(len(name_models)):\n",
    "    for j in range(n_repeat):\n",
    "        res = np.loadtxt(str(file_models[i]) + str(j) + str(name_models[i]))\n",
    "        asw_plot.append(res[0])\n",
    "        nmi_plot.append(res[1])\n",
    "        ari_plot.append(res[2])\n",
    "        models_plot.append(models[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = models_plot + models_plot + models_plot\n",
    "score = list(\"ASW\" for _ in range(len(asw_plot))) + \\\n",
    "        list(\"NMI\" for _ in range(len(asw_plot))) + \\\n",
    "        list(\"ARI\" for _ in range(len(asw_plot)))\n",
    "value = asw_plot + nmi_plot + ari_plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,1,'ASW (Cortex)')"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ASW\n",
    "model = models_plot\n",
    "score = list(\"ASW\" for _ in range(len(asw_plot)))\n",
    "value = asw_plot\n",
    "d = {'model': model, 'score': score, 'value': value}\n",
    "df = pd.DataFrame(data=d)\n",
    "fig = plt.figure(dpi=100)\n",
    "grid = plt.GridSpec(1, 10, wspace=0.2, hspace=0.2)\n",
    "ax = plt.subplot(grid[:, 2:10])\n",
    "sns.boxplot(y=\"model\", x=\"value\", data=df)\n",
    "ax.artists[0].set_facecolor('tab:red')\n",
    "ax.artists[1].set_facecolor('tab:red')\n",
    "ax.artists[2].set_facecolor('tab:red')\n",
    "ax.artists[3].set_facecolor('tab:red')\n",
    "ax.artists[4].set_facecolor('tab:blue')\n",
    "ax.artists[5].set_facecolor('tab:blue')\n",
    "ax.artists[6].set_facecolor('tab:blue')\n",
    "ax.artists[7].set_facecolor('tab:blue')\n",
    "ax.artists[8].set_facecolor('tab:green')\n",
    "ax.artists[9].set_facecolor('tab:green')\n",
    "ax.artists[10].set_facecolor('tab:green')\n",
    "ax.artists[11].set_facecolor('tab:green')\n",
    "plt.xlabel('ASW', fontsize=15)\n",
    "plt.ylabel('', fontsize=15)\n",
    "plt.xticks(fontsize=12)\n",
    "plt.yticks(fontsize=12)\n",
    "plt.title('ASW (Cortex)', fontsize=18)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,1,'NMI (Cortex)')"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgkAAAGPCAYAAADWY0OgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAAPYQAAD2EBqD+naQAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzs3XtYE1f+P/D3BMIlAnJHQKQEEEVDWyu4YiuIVFnERVZF7SLFbSuudautrtt655EV7dZ2bV0Xbb+0+oVVWQuVeumWLuFirZdq8xVky1ZFoRUKKLeAhJDM7w9+mU1MAuEaLp/X8/CUmTlz5swEm8+c+ZwzDMuyLAghhBBCHsMzdgMIIYQQMjRRkEAIIYQQnShIIIQQQohOFCQQQgghRCcKEgghhBCiEwUJhBBCCNGJggRCCCGE6ERBAiGEEEJ0oiCBEEIIITpRkEAIGbbi4uLAMAwyMzON3ZQRq6KiAubm5vDx8UF7e7uxm0MGGQUJhIwwu3btAsMwYBgGY8aMwf379/WWvXv3Llc2Pz9fa3toaCi33dTUFD/99FOXx5bJZHBwcOD2eeKJJ7o85ieffNLDs/uvb7/9Fn//+98xdepULF26tMuyNTU12Lt3L55//nmMHz8elpaWGDNmDJ544gksWrQIR44cQUNDQ6/b0hcNDQ3YtWsXdu3aZbQ2dGXChAlYtWoVbt++jb/+9a/Gbg4ZZBQkEDKCtba2IikpqV/qUigUOHbsWJdlPvvsMzx8+LBfjtedjRs3gmVZ7Ny5EwzD6CzDsiz27NkDLy8vvPXWW/jqq6/w008/wdTUFCYmJrh37x5Onz6NxMREPPHEE0hLSxuUtqtraGhAUlISkpKShmSQAABbtmwBn89HcnLykG0jGRgUJBAywqWlpeE///lPn+pQ9Qh8/PHHXZZTbdfVg9CfLl26hMLCQri4uCAmJkZnGZZlsXLlSmzduhWtra2YMWMGPv30U9TX16O5uRlNTU1obGxEdnY2Fi5ciMbGRuTk5Axou4erCRMmIDIyEg8fPsRHH31k7OaQQURBAiEjlIeHBwICAtDR0YEtW7b0qa7Zs2fjiSeewA8//IALFy7oLPPjjz8iNzcXVlZWWLx4cZ+O153U1FQAwIoVK2BiYqKzzNtvv42MjAwAwIYNG/DNN9/g17/+NWxtbbkyNjY2WLRoEXJyclBQUIDx48cPaLuHs7i4OADAkSNHQC8PHj0oSCBkhOLxeEhJSQEAfPrpp7hy5Uqv62IYBgkJCQD09yZ88sknUCqVWLp0KaysrHp9rO40NTVxiYovvPCCzjJ1dXXYvXs3AGDu3Ll499139T6SUJk9ezbef/99ndvy8/OxdOlSuLu7w9zcHI6Ojpg7dy4+/vhjKBQKnfuockNCQ0MBdH4G8+bNg7OzM3g8Hnbt2oXQ0FB4eXlx+3h5eXH5Gur7qlMoFPjkk08wf/58uLi4wMzMDE5OTpg/fz5OnDih8wv8d7/7HRiGga2tLe7evauzvYcOHeJyT4qKirS2L1y4ENbW1vjhhx905q+QkYmCBEJGsMjISISEhAAA3nzzzT7VlZCQwI0kaGlp0dquSkJctWpVn47TnYKCAjx69AhjxozBtGnTdJb5+OOPuTaqvqwNweNp/y/xjTfewJw5c3Dq1ClUVVVBIBCgoaEBeXl5+O1vf4t58+ahubm5y3o3btyIJUuW4KuvvkJHRwd3HHt7ezg6OnLlHB0d4eLiwv3Y29tr1PPzzz9j1qxZWLVqFb788kvU1NRAIBCgrq4OX375JVasWIFFixZpjUJ49913MWXKFDQ2NuKFF15AR0eHxvaSkhJs3LgRALB161Y899xzWudgbm6OoKAgAMAXX3zR5fmSkYOCBEJGuH379gEAxGJxn/7n7unpibCwMEilUvzjH//Q2FZQUIDbt2/Dx8dH5xdMfyosLAQATJs2Te+jhn/9618AOr90n3322V4f6+DBg3jvvfcAAKtXr8b9+/dRX1+PxsZGvPfeezA1NUVeXh5eeeUVvXVcu3YN7777LjZv3oyff/4ZDx8+REtLC1atWoWsrCxcvXqVK3v16lVUV1dzP1lZWdy29vZ2LFy4EJcvX8a0adNw9uxZtLS0oKGhAVKpFEePHoWzszNycnLwxz/+UaMNlpaWOHnyJCwtLfHNN99g586d3LZHjx5h+fLlaGtrw6xZs7Bjxw695zJjxgwAnZ83GR0oSCBkhJsxYwaX3PfWW2/16Xmyqpfg8UcOqlEBA92LAACXL18GADz55JN6y9y8eRMA8PTTT/f6OI8ePeK+TFesWIHDhw9j3LhxAIAxY8Zgw4YNePfddwEAJ0+exLfffquzHqlUijfeeAP79u2Dk5MTgM67ck9Pzx6158MPP8TVq1cxZcoU5OfnIzIyEgKBgGtPfHw8zp07B4ZhcOjQIdTU1GjsP2XKFK69e/fuhVgsBgC8/vrruHnzJmxtbZGRkaE38AL+ez2vXbtGcyaMEhQkEDIK7NmzByYmJpBIJDh+/Hiv6/n1r3+NsWPHorCwELdv3wYANDc349NPPwWPx8OLL77YX03WSzXvg+oLV5cHDx4AgFZ3fU/k5uZywzl37dqls8zatWvh6uoKAHqvK4/H07qz7w3VqIK1a9fC2tpaZ5lnnnkGU6ZMQXt7OxcEqFuzZg0WL14MpVKJuLg4HDlyBIcPHwbQmZDYXeCiejTS0dGB2travpwOGSYoSCBkFJg0aRJ3l799+3bI5fJe1WNpaYnly5cD+G9vwsmTJ9HS0oJ58+bB3d29fxrcBdWXkyEBgKG5CLqoegY8PDwwceJEnWVMTEwQFhamUf5xPj4+cHZ27nU7gM5A7MaNGwA6P79x48bp/SkrKwMA3Lt3T2ddH374ISZMmID79+8jMTERAPDyyy93OyEVoHnNKUgYHShIIGSU2LVrFywtLXHnzh1uCGFv/Pa3vwUAHDt2DEqlkgsWVOsHWltbG4DOLnt9HBwcAPy3R6E3VN313QU+qmGTj3fvq/Q1QACA6upqKJVKAMDDhw/x888/6/1RBYCtra0667Kzs9OYOVEoFOLAgQMGtcPS0pL7XfU5kJGNggRCRgl3d3f8/ve/BwAkJydDKpX2qp6goCD4+/ujsrISf/3rX3Hx4kXY29vjV7/6VX82Vy9VAFBfX6+3zJQpUwAAEomkz8cztDdCX7munvEbSn2Y5aVLl8CybLc/+h6RANCYEOmnn37CrVu3DGqH+myaqs+BjGwUJBAyirz11luws7NDTU0N9u/f3+t6VI8uNm3aBKBzvoKu7uz7kyoXoavpn+fOnQugs0tc3+RP3VH1AFRWVnZZ7scff9Ro10BwcXHhfi8uLu5TXQcPHsTp06dhYmICf39/yGQyLF++XG/Pgzr1az6Q50uGDgoSCBlFbG1tufkS9u/fr7eLvDsrV66Eqakpl+E+WI8aAMDf3x8AcOfOHb1lVq1axWX+79q1y+ARHaoufQCYPn06gM4gQN+01gqFgksQDAwMNOgYj1Ofm0FfO+3s7LjzPnHiRK+OA3QGGH/4wx8AADt27MC5c+dga2uLf//733j99de73b+8vBwA4OrqqjFzJRm5KEggZJR57bXXMH78eDQ3NyM5OblXdbi4uOC9997Dxo0bkZSU1Kehhj01e/ZsAOhyBklHR0ds27YNQOecCaqXQXXl66+/xvr167nl559/nutS19d1f/jwYW60xYoVKww+B3U2Njbc7129PGn16tUAOs+nu0BBVy/Lo0ePsGLFCrS1teHZZ5/F1q1b4enpiSNHjgDoHN3w6aefdlmvavip6jMgowBLCBlRdu7cyQJgPT099Zb56KOPWAAaP2KxWKtcSEgIC4B98cUX+60N5eXl3DE//vjjHtXLsixbWlrK7V9dXa23nFKpZJctW8aVnTlzJpuVlcU2NjZyZZqamtjPP/+cjYmJYRmGYaOjozXq+OCDD7j9ExMTueO1tLSw77//Psvn81kA7LJly/Reg5CQkG7Pyd3dnQXA/v73v2flcrnOMm1tbeyMGTNYAKypqSm7detWtqKigtve0tLCisVi9tVXX2VtbW219k9MTGQBsLa2tuy9e/c0tr300kssANbOzk6jzsf5+fmxANhDhw51e05kZKAggZARxpAgoaOjg500adKwDBJYlmWffPJJFgB75MiRLssplUo2KSmJtbS01DhXa2tr1traWmOdvb09e+zYMa06Xn/9da4MwzCsnZ0da2pqyq2bM2cO29TUpPcaGBIk7N69m6vP3Nyc9fDwYD09PbWCj9raWjYsLEyj3TY2NqytrS3LMAy3ztTUVGO/Tz/9lNuWmZmpdfyWlhbu7+G5555jOzo6tMqUlZWxAFgzMzO2pqam23MiIwM9biBkFDIxMcGePXuM3YxeU43vV73lUR+GYbBjxw7cuXMHe/bsQVhYGNzc3NDe3o6Ojg54enpi0aJF+Oijj3D37l2sXLlSq453330XeXl5WLx4MVxcXCCVSmFtbY05c+YgLS0Nubm5eic3MtSWLVtw4MABTJ8+HXw+Hz/++CPu3buH6upqjXKOjo746quvcPr0aSxZsgQeHh6QyWR49OgR3N3d8ctf/hIHDx7UeIlTZWUlXn75ZQDASy+9pHM+BIFAgOPHj8Pc3BxFRUU6H0OprnVMTAwlLY4iDMvSOz8JIcNLc3Mzl1dRXl7e4ymOSc+wLAtfX1/cvn0bBQUFlJMwilBPAiFk2LG2tsabb74JlmW5F1iRgZOZmYnbt29j/vz5FCCMMtSTQAgZltra2uDn54fq6mrcunULHh4exm7SiKRUKiESifD9999DIpFAJBIZu0lkEJkauwGEENIbFhYWOHr0KPLz81FRUUFBwgC5f/8+li5dCi8vLwoQRiHqSSCEEEKITpSTQAghhBCdKEgghBBCiE4UJBBCCCFEJwoSCCGEEKITjW4gBqmvr0dHR4exmzEkOTk5oba21tjNGBZG87Vqb2/vdpbLLVu2wMzMDMDovlY9Rdeqe6amprCzs+v5fgPQFjICdXR0QC6XG7sZQw7DMAA6rw8NFOraaL9W7e3tuHXrVrdlGIYZ9deqJ+haDSx63EAIIYNsd+WP2F35o7GbQUi3qCeBEEIGmRnd8ZJhgoIEQggZICzLor29nfu9J/u0tbVR9zkxOnrcQAghA6S9vR1r1qzBmjVruGDBkH0SExMRHR1t8D6EDBTqSRhm8vPzcejQIZ3boqKiEB8fzy13dHQgMTERzc3N2LRpE4KCggarmYQQQkYAChKGqbVr18LNzU1jnb29vcby1atX0dzcDADIy8ujIIEQQkiPUJAwTHl4eMDb27vLMnl5eeDz+Zg0aRIkEgnq6+t7NU6WEELI6EQ5CSNUXV0dbty4gcDAQERFRUGpVKKgoMDYzSJkVFFPPJTJZP9dr6e8TCbTLEeJi8TIqCdhmFIqlVAoFBrrTExMuN/z8/PBsizmzJkDkUgEBwcHiMViLFq0qMt65XK5xqRJDMPA0tKyfxtPyCihnni4efNm7nc5w+gcBrl+/Xqt/c3NzQeugYR0g4KEYWrr1q1a644fPw4TExOwLAuxWAxHR0eIRCLweDyEhIQgKysLpaWl8Pf311tvdnY2Tp06xS17eXlh3759A3IOhBBChjYKEoapdevWwd3dXWOdqiehuLgYtbW1WLJkCXi8zidKoaGhyM7Ohlgs7jJIiImJQVRUFLesmvKUENJzqvcwAMDbb7/N9Sbw9TxGOHDgAID/9iio70+IMVCQMEy5u7vrTVzMy8sDAAQFBaGlpQUAYG1tDV9fX1y6dAmrVq2CQCDQuS+fzwefzx+YRhMyyqgH2eqPDfSF3o8/WqAgnRgbBQkjjFQqxdWrVwFoPgNVd/HiRYSHhw9mswghhAxDFCSMMEVFRZDL5VixYgUmTpyosY1lWezfvx9isZiCBEIIId2iIGGEEYvFsLa2RlRUlM7HBrNnz8b58+dRWVkJDw8PI7SQEELIcEHzJIwgd+7cwd27dxEaGqo3r0DVg6DKWyCEDBwzMzOkpqYiNTXV4CREMzMzHD58GKdPn6bERWJ01JMwzISGhiI0NFTnNqFQiMzMzC739/Dw6LYMIaR/MAzDJSOqT5JkyD4WFhZgGIYmVCJGRUECIYQMsnYatUCGCQoSCCFkkG33GG/sJhBiEMpJIIQQQohO1JNACCGDQJXEqMKyLPduBzMzMzAMQ4mKZMihIIEQQgaBehKjioWFhZFaQ4hhKEgghAxb6nfjvd1PdRdviJ6UJWQkoCCBEDJstbe3Y82aNYN2vNTUVHp1MxlVKHGREEIIITpRTwIhZETYXfkjzAyceKidYbhhiN3tp16WkNGGggRCyIhgxrIw78XshL3dj5DRgIIEQsiAUyUKtrW1jdhphnubDEnIUEY5CYSQAdfe3o7ExERER0f3ajTCcKBKolyzZs2IPUcy+gz7noTMzEycOnWKWzYxMYG9vT2mT5+O2NhYjBkzBgCQn5+PQ4cOISUlBd7e3l3WqSqrS1RUFOLj4zXWdXR0IDExEc3Nzdi0aROCgoIMbn9rayuysrJQXl6Ou3fvorm5GcuWLcPixYt1lr99+zYyMjLwww8/wNTUFFOmTEF8fDycnZ25MtXV1Xjttdfw4osvYsGCBdx6pVKJQ4cOobCwEC+88AIWLVpkcDvJ6CCRSJCeno64uDg89dRTxm4O6SX6HEl/GfZBgsqWLVsgEAjQ1taG69ev4/z587h16xaSk5N73e23du1auLm5aayzt7fXKnf16lU0NzcD6HwFc0+ChKamJnz11Vfw9PREYGBgl69wrqysxK5du+Dt7Y033ngDMpkMJ0+exM6dO/H222/D2tpa775yuRzvvfcerl27htWrV3OvjCZERSaT4dixY6ivr8exY8cwefJkGu43DNHnSPrTiAkShEIhbGxsAAABAQGQSqUoLCxEWVkZJk2a1Ks6PTw8uu11ADoDAz6fj0mTJkEikaC+vh52dnYGHcPFxQWffPIJAKChoaHLIOHkyZOwsLDAm2++yc3U5uXlhQ0bNuDMmTNYsWKFzv3a2trw9ttv4/vvv8f69esRHBxsUNvI6HL27Fk0NDQA6PxbPHfuHGJiYozcKtJT9DmS/mTUIKGpqQnHjx+HRCJBY2MjLC0t4ebmhqVLlyIgIABAZ7dZTk4Obt++DYVCAScnJ8yePbvbP3pfX18UFhairq5uQM+hrq4ON27cwMyZMxESEoLi4mIUFBQY3JVvaC+HXC7Hd999h7CwMI2pXF1cXDB58mRcuXJFZ5DQ3NyMPXv24Mcff8TmzZup65Ho9PPPP+PcuXNcUiHLsjh79iyCg4Ph4uLS5/rVkxVlMlm/JS/KZLL/HqNfatSmXq/68bpsi5GSMwf6cySjj1GDhA8++ADl5eVYvnw53Nzc0NLSgvLyckilUgCdd+iHDx+Gv78/XnnlFYwdOxZVVVWoqKjotu7q6moA4HoXekOpVEKhUGisMzEx0VjOz88Hy7KYM2cORCIRHBwcIBaL+/15f3V1NeRyOTw9PbW2TZgwAaWlpejo6ICp6X8/0ocPH2Lnzp2or6/Htm3b4Ofn1+1x5HI55HI5t8wwDCwtLfvnJMiQxLIs0tPTtb7YVOvfeOONPmfqqyfyvfbaa32qSx85w8BiAL6c5Wrnvn79eoP2aW9vH/T3MgzG50hGH6MGCWVlZQgLC9N4Ph4YGAigs4v86NGj8PPzw44dO7g/bpFIpLMu1Re6KichNzcXDg4OvX7UAABbt27VWnf8+HEuUGBZFmKxGI6OjhCJRODxeAgJCUFWVhZKS0vh7+/f62M/TpXzYGVlpbXNysoKSqUSra2tGkHR559/DgDYsWOHQQECAGRnZ2skgnp5eWHfvn19aToZ4qqqqlBSUqK1XqlUoqSkBFVVVVq5OWTooc+RDASjBgk+Pj4oKCiAtbU1RCIRhEIhdydcVlaGR48eYd68eQZFv6tXr9ZY9vPzQ2Jiot5Xr7IsC6VSqbHu8V6CdevWwd3dXW+Z4uJi1NbWYsmSJeDxOkeThoaGIjs7G2KxmAsSDDnWQHjqqadQUlKCo0ePYvv27Qb1qsTExCAqKopbpjuPkc/V1RVTp05FaWmpxt8pj8eDv78/XF1d+3wM9X+H77//fr+9Elkmk3F39/wB6uJXr/fAgQN6kwDV22KMVz4PxudIRh+jBgkbNmxAVlYW8vLyuKS8oKAgxMXFoampCQDg4OBgUF3bt2+HQCCAiYkJHBwcusz0B4DS0lIkJSVprDt48KDGUEJ3d/cuExdVSYZBQUFoaWkBAFhbW8PX1xeXLl3CqlWrIBAIUFxcjOTkZI19//a3vxl8bqp6AXCPYtRJpVLweDwIBAKN9U8++STmz5+P/fv3Y/fu3QYFCnw+H3w+3+B2keGPYRjExcVp9ZwxDIOVK1f2S6CoXoe5ufmAfIkOVDirXq+5ublBIwWMEVwPxudIRh+jBgk2NjZISEhAQkIC6urq8O233yIjIwONjY3c3eyDBw8MqsvT07NH+QdCoRApKSka63QNb9RHKpXi6tWrAIDNmzfrLHPx4kWEh4fD19dX61hjx441+FhA510Cn8/XmY9RWVkJV1dXjXwElWeeeQabNm3C/v37kZSUhJ07d/YpT4OMTC4uLoiMjMSZM2fAsiwYhsGCBQs0gmYy9NHnSPrbkBkC6ejoiIiICBQXF6OsrAx+fn4QCATIzc3FrFmz+j0KtrS0NGh4oz5FRUWQy+VYsWIFJk6cqLGNZVns378fYrEY4eHhfT4WAJiamuLpp5/GpUuX8MILL3BJUTU1NSgtLUV0dLTefadNm4ZNmzbhnXfeQVJSEnbs2NHjIIWMfAsWLMCFCxdQX18PW1tbREZGGrtJpBfocyT9yWjTMre2tuKPf/wjcnJycO3aNZSWliInJwcSiQQBAQGwsLBAfHw8/v3vf2P37t34+uuvUVJSgq+++gr/8z//Y6xmc8RiMaytrREVFYUpU6Zo/EydOhWzZ8/GDz/8gMrKym7run79Oi5duoRr164B6OwZuHTpEi5duqSRFb5s2TI8evQI+/btg0QiweXLl7F3717Y2tpqzKyoy9NPP40//OEPqK6uRlJSEjeOmhAVc3NzxMfHw8HBAfHx8TQBzzBFnyPpT0brSeDz+fDx8UFRURFqamqgUCjg6OiI6Oho7q44LCwMdnZ2OH36NFJTUwEATk5OCAkJMVazAQB37tzB3bt3sXDhQr3P78PDw3H+/Hnk5eXhxRdf7LK+I0eO4OHDh9zyxYsXcfHiRQCauQseHh7YuXMnMjIysH//fvB4PIhEIsTHx3ebgwF0JjJu3rwZb7/9NvfowdbW1tDTJqPAU089RXNpjAD0OZL+wrAj9ZVspF/V1tZqzJ9AOjEMA1dXV1RVVY3Ytxv2B5ZlIZfLMW7cOI2AuK9kMhnWrFkDANhXUWnwK59lDIM/TvAwaD/1sqmpqXrvzPvzLZD0d2U4ulaG4fP5cHJy6vF+QyYngRAycjEMA3Nzc1hYWIBhmBH5P3PVORIyklCQQAgZEdp7cOeuXra7/XpSLyEjDQUJhJARYbvH+EHdj5DRwGijGwghhBAytFFPAiFk2DIzM+NGPvVEb5MMjTHdMiHGREECIWTY6kuy4GC/pZGQ4YiCBEIIIYNKvSenN/uo9/4wDIO2tjbIZDKtUTN9HYpKKEgghBAyyNrb27n5LQZSV/NaEMNQ4iIhhBBCdKKeBEIIIUazu/JHmBkwuVY7w3DDVbvaR70c6TsKEgghhBiNGcsaPJ12X/YhvUNBAiGEjHL9+d6JkWq0XqMRESRkZmbi1KlT3LKJiQns7e0xffp0xMbGYsyYMQCA/Px8HDp0CCkpKfD29u6yTlVZXaKiohAfH6+xrqOjA4mJiWhubsamTZsQFBRkcPtv3LiB5ORkndt0tfX//u//kJmZibt378LCwgLPPPMM4uLiYGNjo9WmrKwsFBQUoKGhAc7OzoiIiMD8+fMNbhshZORTTySkZD/dRus1GhFBgsqWLVsgEAjQ1taG69ev4/z587h16xaSk5N7HfWtXbsWbm5uGuvs7e21yl29ehXNzc0AgLy8vB4FCSq/+c1vMHnyZI117u7uGsvFxcVISUnB9OnTsWzZMjQ0NCAjIwPl5eVISUmBqel/P9IjR47g66+/xvLlyyEUCvHdd98hLS0NbW1t3Ou4CSGEDG0SiQTp6emIi4sb9FeAj6ggQSgUcnfTAQEBkEqlKCwsRFlZGSZNmtSrOj08PLrtdQA6AwM+n49JkyZBIpGgvr4ednZ2PTqWq6srJk6c2GWZ9PR0eHh44I033gCP1zk4xdHREbt27UJ+fj7Cw8MBAHfv3kV+fj7i4uKwcOFCAMCUKVPQ1NSEU6dO4fnnn4dAIOhR+wghhAwumUyGY8eOob6+HseOHcPkyZMHtRfD6EFCU1MTjh8/DolEgsbGRlhaWsLNzQ1Lly5FQEAAgM4oKicnB7dv34ZCoYCTkxNmz56NmJiYLuv29fVFYWEh6urqBvQc6urqcOPGDcycORMhISEoLi5GQUEBFi1a1K/Hqa2tRXl5OeLi4rgAAQD8/f3h4uKCq1evckHC1atXAQChoaEadcyZMwf5+fmQSCQIDg7u1/YRQgjpX2fPnkVDQwMAoKGhAefOnev2u68/GT1I+OCDD1BeXo7ly5fDzc0NLS0tKC8vh1QqBdB5h3748GH4+/vjlVdewdixY1FVVYWKiopu666urgYArWf1PaFUKqFQKDTWmZiYaCzn5+eDZVnMmTMHIpEIDg4OEIvFPQ4SPvzwQ7z77rswNzfHxIkTsWTJEvj5+XHbKysrAQCenp5a+06YMAF37tzRKGtnZ6d17qp9VXURQoj6TIUymWzAj6d+jP4eo6BeX3+ei0abB2lkxc8//4xz585xx2NZFmfPnkVwcDBcXFwGpQ1GDxLKysoQFhbG3QEDQGBgIACgra0NR48ehZ+fH3bs2MHlFYhEIp11qb7QVTkJubm5cHBw6PWjBgDYunWr1rrjx49zgQLLshCLxXB0dIRIJAKPx0NISAiysrJQWloKf3//bo8xZswYLFiwAP7+/rCyskJVVRVycnKwa9coELHSAAAgAElEQVQubNmyhTtfVc6DlZWVVh1WVlZcYKUqq6ucQCAAj8fj6nqcXC6HXC7nlhmGgaWlZbfnQAgZvtSnSF6/fv2gHlvOMLDoxy9duVr+2UCdS3t7+4C/+4NlWaSnp2sFJKr1b7zxxqCMsDB6kODj44OCggJYW1tDJBJBKBRyyXdlZWV49OgR5s2bZ9DFWL16tcayn58fEhMT9b65jWVZKJVKjXWP9xKsW7dOK3lQvUxxcTFqa2uxZMkS7hFAaGgosrOzIRaLuSChq2N5e3tr5D1MnjwZQUFB2LhxI9LT07Fv3z6N/fr6h8GyrN46srOzNUaKeHl5aR2fEELIwKqqqkJJSYnWeqVSiZKSElRVVWkl1Q8EowcJGzZsQFZWFvLy8nDy5ElYWFggKCgIcXFxaGpqAgA4ODgYVNf27dshEAhgYmICBwcHWFtbd1m+tLQUSUlJGusOHjwIZ2dnbtnd3b3LxMW8vDwAQFBQEFpaWgAA1tbW8PX1xaVLl7Bq1SoIBAIUFxdrDXP829/+pvfcrKys8PTTT0MsFqOjowOmpqbc+ejqBZBKpRo9B9bW1rh//75WudbWVrAsq7OXAQBiYmIQFRXFLY+WscCEjGbqN1IHDhwY8MQ4mUzG3eXz+7nrXr2+/jwX9TYPxivDXV1dMXXqVJSWlmrcYPJ4PPj7+8PV1XXA2wAMgSDBxsYGCQkJSEhIQF1dHb799ltkZGSgsbGR+7J68OCBQXV5enr2KP9AKBQiJSVFY52u4Y36SKVSLkFw8+bNOstcvHgR4eHh8PX11TrW2LFjDT4W0Jl3AAAVFRV48sknNbZVVFRw21VlL1++jKamJo1rosrl8PDw0HkMPp8PPp/fo3YRQoY39ZsBc3PzQc2e7+/bEPX6BupcBuPmiWEYxMXFaT3yZhgGK1euHLQbOKMHCeocHR0RERGB4uJilJWVwc/PDwKBALm5uZg1a1a/XxRLS0uDhjfqU1RUBLlcjhUrVmgNXWRZFvv374dYLEZ4eHiPjyWVSnH9+nWNxy+Ojo7w8vJCYWEhFixYwD3e+P777/Hzzz9rJEoGBgYiMzMTBQUF3BBIoDPJ0tzcfNDH2hJCCOkZFxcXREZG4syZM9xj4gULFmj0dg80owYJra2tSEpKwqxZs+Du7g5LS0vcunULEokEM2bMgIWFBeLj45Gamordu3dj7ty5GDt2LKqrq3Hv3j289NJLxmw+xGIxrK2tERUVpfPue/bs2Th//jwqKyv13rkDwHvvvQcXFxcIhUJYW1ujqqoKn3/+OZqbm/Haa69plI2Li8Of/vQnvPfee5g3bx43mZKnpydCQkK4cp6enggNDcWJEyfAMAw3mZJYLMYLL7xAcyQQQsgwsGDBAly4cAH19fWwtbVFZGTkoB7fqEECn8+Hj48PioqKUFNTA4VCAUdHR0RHR3MzAoaFhcHOzg6nT59GamoqAMDJyUnjC9EY7ty5g7t372LhwoV6u+fDw8Nx/vx55OXl4cUXX9Rbl6enJ7755ht8+eWXaGtrg5WVFSZPnoz169dDKBRqlBWJRHjzzTeRmZmJvXv3akzLrD7bItCZyOng4ICzZ8+isbERzs7OWLVqFSIiIvp+AQghhAw4c3NzxMfHczMuDvZ00Aw7WAM+ybBWW1urMTSSdGIYBq6urqiqqhq0sdPDFV0rww32tRrslxfJZDLuPQj7KioNeqOjjGHwxwke3e6jXq4/37Ew3F/wxOfz4eTk1OP9hlROAiGEkMHHMMyoeWFRb43Wa0RBAiGEEKNpN/COXL1cV/sYWh8xDAUJhBBCjGa7x/hB2Yf0Dq/7IoQQQggZjagngRBCyKAyMzPjRqsZSl/iIMMwGDduHKqrq7WSPAdjZsSRjoIEQgghg6q3SYC6XqrEMAwsLCxgbm5Oo2YGAAUJhBAySqnfnfdHHX0dGjgchxaOdBQkEELIKNXe3s7NVzAU9Oe8BqR/UOIiIYQQQnSingRCCCHYXfkjzHrxTL+dYbghib2pQ31/MvRQkEAIIQRmLGvQ9MgDXQcZWihIIISQfjDc5/Yfiegz6bsRESRkZmbi1KlT3LKJiQns7e0xffp0xMbGYsyYMQCA/Px8HDp0CCkpKfD29u6yTlVZXaKiohAfH6+xrqOjA4mJiWhubsamTZsQFBRkcPtv3LiBCxcuoKysDA8ePICVlRWEQiGWLl0KLy8vjbLvv/8+Lly4oFWHh4cH9u/fr9WmrKwsFBQUoKGhAc7OzoiIiMD8+fMNbhshxDDqSYCUgDc00GfSdyMiSFDZsmULBAIB2tracP36dZw/fx63bt1CcnJyryPItWvXws3NTWOdvb29VrmrV6+iubkZAJCXl9ejIOGf//wn2trasGDBAowfPx6NjY34/PPPsWXLFmzfvh3+/v4a5S0sLLBt2zaNdbr++I8cOYKvv/4ay5cvh1AoxHfffYe0tDS0tbVxr+ImhBBC9BlRQYJQKISNjQ0AICAgAFKpFIWFhSgrK8OkSZN6VaeHh0e3vQ5AZ2DA5/MxadIkSCQS1NfXw87OzqBjrF69GmPHjtVY99RTT2HdunXIzs7WChJ4PB4mTpzYZZ13795Ffn4+4uLisHDhQgDAlClT0NTUhFOnTuH555+HQCAwqH2EEEJGJ6MHCU1NTTh+/DgkEgkaGxthaWkJNzc3LF26FAEBAQAAiUSCnJwc3L59GwqFAk5OTpg9ezZiYmK6rNvX1xeFhYWoq6sb0HOoq6vDjRs3MHPmTISEhKC4uBgFBQVYtGiRQfs/HiAAgKWlJcaPH48HDx70qk1Xr14FAISGhmqsnzNnDvLz8yGRSBAcHNyrugkhhIwORg8SPvjgA5SXl2P58uVwc3NDS0sLysvLIZVKAXTeoR8+fBj+/v545ZVXMHbsWFRVVaGioqLbuqurqwGA613oDaVSCYVCobHOxMREYzk/Px8sy2LOnDkQiURwcHCAWCw2OEjQRSqV4u7du3jyySe1trW1teGVV15BU1MT7O3tERgYiNjYWFhZWXFlKisrYWdnp3Xunp6e3HZCSP9RnxJYJpP1qS6GYdDW1gaZTDagUw2rt9NYYxLUj9ub69bVtdI4Pxp10StGDxLKysoQFhaG8PBwbl1gYCCAzi/Do0ePws/PDzt27ODyCkQikc66VF/oqpyE3NxcODg49PpRAwBs3bpVa93x48e5QIFlWYjFYjg6OkIkEoHH4yEkJARZWVkoLS3VelRgqI8++gjt7e349a9/rbHey8sL3t7e8PDwAACUlJTg3LlzuHnzJvbs2cPlJjQ3N2sEDSoCgQA8Ho/Ln3icXC6HXC7nlhmGgaWlZa/OgZDRRH164/Xr1xuxJb0jZxhYGOGLVK6WLzaQ1629vV3nux9I14weJPj4+KCgoADW1tYQiUQQCoUwNe1sVllZGR49eoR58+YZlHi4evVqjWU/Pz8kJibqfRMYy7JQKpUa6x7vJVi3bh3c3d31likuLkZtbS2WLFkCHq9zAsvQ0FBkZ2dDLBZzQYIhx1L5+9//josXL+Lll1/GE088obFNlV+gEhAQgCeeeAJ/+ctfkJeXh1/+8pc661THsqze65mdna0xUsTLywv79u3rtk5CCCEjj9GDhA0bNiArKwt5eXk4efIkLCwsEBQUhLi4ODQ1NQEAHBwcDKpr+/btEAgEMDExgYODA6ytrbssX1paiqSkJI11Bw8ehLOzM7fs7u7eZeJiXl4eACAoKAgtLS0AAGtra/j6+uLSpUtYtWoVBAIBiouLkZycrLHv3/72N61zO3HiBD777DP85je/wbx587o/aQC/+MUvwOfz8cMPP3BBgrW1Ne7fv69VtrW1FSzL6uxlAICYmBhERUVxyzSumBDDqN+MHDhwoE/D7bp6/XF/kslk3N0730jd8erH7c116+paqZ8fvTa6d4weJNjY2CAhIQEJCQmoq6vDt99+i4yMDDQ2NnJfVoYm73l6evYo/0AoFCIlJUVjna7hjfpIpVIuQXDz5s06y1y8eBHh4eHw9fXVOtbjCYsnTpxAVlYWli1b1qshiupf6BMmTMDly5fR1NSkcU1UuRyqxxWP4/P54PP5PT42IaOd+r8/c3PzPgcJg/36Y2PdDqgftzfXzdBrRTc8vWP0IEGdo6MjIiIiUFxcjLKyMvj5+UEgECA3NxezZs3q9w/Z0tLSoOGN+hQVFUEul2PFihVaQxJZlsX+/fshFosRHh7e7bEyMzORlZWFpUuXYvHixT1qx8WLFyGXy+Hr68utCwwMRGZmJgoKCjQeUeTn58Pc3BxPPfVUj45BCCFk9DFqkNDa2oqkpCTMmjUL7u7usLS0xK1btyCRSDBjxgxYWFggPj4eqamp2L17N+bOnYuxY8eiuroa9+7dw0svvWTM5kMsFsPa2hpRUVE6775nz56N8+fPo7KyUu+dOwB89tlnOHXqFKZNm4Ynn3wS//nPf7htDMNwX/7V1dU4dOgQgoODMW7cOLAsi5s3b+L8+fPw9PREWFgYt5+npydCQ0Nx4sQJMAzDTaYkFovxwgsv0BwJhBBCumXUIIHP58PHxwdFRUWoqamBQqGAo6MjoqOjue72sLAw2NnZ4fTp00hNTQUAODk5ISQkxJhNx507d3D37l0sXLhQb/d8eHg4zp8/j7y8PLz44ot667p27RoA4Pr167h+/brGNh6PhxMnTgDoHJlgbW2Nzz//HA0NDWBZFk5OTliwYAEWLVqk9cxt9erVcHBwwNmzZ9HY2AhnZ2esWrUKERERfTl1QgghowTD0uBRYoDa2lqNoZGkE8MwcHV1RVVVFY3D7sZIv1b9+TKhwbpWMpmMe7fBvorKXr3BUcYw+OMEj17Xob5/b96v0NW1ohc8/Refz4eTk1OP9xtSOQmEEDJcMQxDLxAaYugz6TsKEgghhKC9l3fZ6vv1po7eHpcMDgoSCCGEYLvH+CFRBxlaeMZuACGEEEKGJupJIISQUcrMzIwbNdZb/ZkcSLMiDj0UJBBCyCjVX4l99OKkkYuCBELIkEND1wgZGigngRAy5LS3t2PNmjVYs2aNxiuYCSGDi4IEQsiQI5PJdP5OCBlcFCQQQgghRCcKEgghhBCiEwUJhJAhR6lU6vydEDK4KEgghAw5LS0tOn8nhAwuGgI5yGJjY7sts2TJEsTGxiI2Nhbz58/HSy+9BACoqanBunXrdO4jFAqxd+9erfX79u3DtWvXEBkZiYSEhD61nRBCyOhCQcIgS05O1rleqVTi4MGDePjwIaZNm9ZlHREREXj22Wc11umazKS+vh4SiQQAUFRUhLi4OJia0kdOCCHEMPSNMcgmTpyoc31aWhpqamqwevVq+Pj4dFmHo6Oj3nrUFRQUQKFQYNq0abh+/TquXLmC4ODgXrWbEELI6EM5CUNAYWEhvvjiC4SFhSE8PLzf6hWLxbCzs8PatWvB5/MhFov7rW5CCCEjHwUJRlZeXo4jR47A29ubyz3oDsuyUCgUGj8sy2qUKS0tRVVVFUJCQmBjY4PAwEDcuHEDdXV1XdYtl8vR2trK/Tx69KjX50YIIWR4o8cNRtTU1IR33nkHFhYW2LhxI/h8vkH7ZWRkICMjQ2Pdtm3bEBAQwC3n5eUBAMLCwrj/Xrx4Efn5+ViyZIneurOzs3Hq1Clu2cvLC/v27TP4nAghhIwcFCQYiVKpxIEDB/DgwQNs27YNjo6OBu8bGRmJ5557TmOdm5sb93traysuX76MyZMnY9y4cQAAkUgEJycniMViLF68WO8Lc2JiYhAVFcUt04t1CCFk9KIgwUjS09NRXFyMlStXYurUqT3a197eHt7e3nq3f/3115DJZJg5c6bGGPOZM2ciJycHJSUlEIlEOvfl8/kG92gQQggZ2ShIMIILFy7gzJkzCA4OxsKFC/u9ftWjhrS0NKSlpencri9IIIQQQlQoSBhk9+7dw+HDh+Hh4YE1a9b0e/0VFRW4ffs2Zs6ciXnz5mltP3XqFK5cuQKpVAorK6t+Pz4hhJCRg4KEQSSVSvHnP/8Zcrkc0dHRqKio0FnOxsaGyyXoKVUvQnR0NIRCodb2lpYW3Lx5ExcuXEBERESvjkEIIWR0oCBhEN27dw81NTUAgIMHD+otFxISgldffbXH9Xd0dKCoqAje3t46AwQAeOaZZ2BnZwexWExBAhmyxowZo/N3QsjgYtjHB9gTokNtbS3kcrmxmzHkMAwDV1dXVFVVac1VQTT15Fo1NTVh/fr1AIADBw7AxsZmMJo4ZNDfleHoWhmGz+fDycmpx/vRZEqEEEII0YmCBEIIIYToREECIWTIMTc31/k7IWRwUeIiIWTIMTMzQ2pqKvc7IcQ4KEgghAw5DMNQDwIhQwAFCYSQYYFlWbS3tw9K/WZmZn1+b0l/1EGIsVGQQAgZFtrb2wdkltKBkpqaSr0hZNijxEVCCCGE6EQ9CYSQYefW1Jeh5PXv20oZpRy+JR8BAH6Y+jLYXtTPU8rh8//rIGQkoCCBEDLsKHl8sCYD90pztpf1KwegLYQYEwUJhJBe6e9EP2Ic9DmSrlBOAiGkV1SJhGvWrBnQUQdkYNHnSLpi9J6EzMxMnDp1ils2MTGBvb09pk+fjtjYWO4NcPn5+Th06BBSUlLg7e3dZZ2xsbHdHnfJkiVcudjYWMyfPx8vvfQSAKCmpgbr1q3TuZ9QKMTevXu11u/btw/Xrl1DZGQkEhISuj2+us8//xzff/897t69i9raWohEImzfvl1n2YaGBqSnp+O7776DTCaDl5cXli1bhqlTp2qUW7NmDYRCITZv3qyxPjc3Fx999BECAwOxfv168PkD12VLCCFkeDN6kKCyZcsWCAQCtLW14fr16zh//jxu3bqF5OTkHnd/JScn61yvVCpx8OBBPHz4ENOmTeu2noiICDz77LMa6ywsLLTK1dfXQyKRAACKiooQFxcHU1PDL+2XX34JgUAAkUiEq1ev6i3X3t6OpKQkyGQyrFq1CjY2Njh//jz+9Kc/YefOnZg0aVKXx/nss8/w97//HaGhoVizZg14POpIIoQQot+QCRKEQiH3OtiAgABIpVIUFhairKys2y+/x02cOFHn+rS0NNTU1GD16tXw8fHpth5HR0e9dakrKCiAQqHAtGnTcP36dVy5cgXBwcEGt/fAgQPcF/b333+vt9xXX32Fn376CXv27OHa7+/vj02bNiEjIwO7d+/Wu296ejpycnKwYMECxMfH03NHQggh3epTkNDU1ITjx49DIpGgsbERlpaWcHNzw9KlSxEQEAAAkEgkyMnJwe3bt6FQKODk5ITZs2cjJiamy7p9fX1RWFiIurq6vjSRU1hYiC+++AJhYWEIDw/vlzpVxGIx7OzssHbtWvzud7+DWCzuUZBg6B391atX4eHhoRHgmJqa4tlnn0VmZiYaGhpga2ursY9SqcSRI0eQl5eH2NhYLFmyxOB2EdIVlmW532UyWbflGYZBW1sbZDKZxr6G0jhGL/YfFD28Jvr09Vr1hHo7B/pYZPjpU5DwwQcfoLy8HMuXL4ebmxtaWlpQXl4OqVQKAMjLy8Phw4fh7++PV155BWPHjkVVVRUqKiq6rbu6uhoAuN6FvigvL8eRI0fg7e3N5R0YgmVZKBQKjXU8Hk/jLry0tBRVVVVYtGgRbGxsEBgYiG+++QZ1dXVwdHTsc9vVVVRUcMGXOk9PTwDAjz/+qBEkKBQK/OUvf8Hly5fx0ksvYf78+d0eQy6XQy6Xc8sMw8DS0rIfWk9GGvUkt/Xr1w/qsRm2AyyG3oufGLaD+32wr0l/aG9v1/lIlYxefQoSysrKtO7MAwMDAQBtbW04evQo/Pz8sGPHDu6LVSQS6axLqVRCoVBwOQm5ublwcHDo8aOGxzU1NeGdd96BhYUFNm7c2KNEvYyMDGRkZGis27Ztm8YXdV5eHgAgLCyM++/FixeRn5/f73ftUqkUVlZWWutV65qbmzXWf/fddwA6kzQNCRAAIDs7WyOR1MvLC/v27ettkwkhhAxjfQoSfHx8UFBQAGtra4hEIgiFQi5hr6ysDI8ePcK8efMMev69evVqjWU/Pz8kJibqfU0sy7JQKjWnLjExMdFYViqVOHDgAB48eIBt27b1+M4+MjISzz33nMY6Nzc37vfW1lZcvnwZkydPxrhx4wB0BkFOTk4Qi8VYvHgxd+7d9UgYqqt9Ht/m5eWFpqYmnD9/HtOmTTMoDyMmJgZRUVEGHY+Mbur/Ng8cONDtewoYhsG4ceNQXV3d68cNqrtzlhky6VQa1NtlyDXRp6/XqifUryu9lps8rk//0jZs2ICsrCzk5eXh5MmTsLCwQFBQEOLi4tDU1AQAcHBwMKiu7du3QyAQwMTEBA4ODrC2tu6yfGlpKZKSkjTWHTx4EM7Oztxyeno6iouLsXLlSq0hgoawt7fvcrjl119/DZlMhpkzZ6KlpYVbP3PmTOTk5KCkpAQikQgKhQIrVqzQ2HfdunWYPXt2j9pjZWWl1VsAgHu883gvg4ODA15//XUkJSUhOTkZ27Zt6zZQ4PP5NCySGEQ9gDQ3NzcoSLCwsIC5uXnfv/iGavDaw2uiv5p+vFY9PC4h6voUJNjY2CAhIQEJCQmoq6vDt99+i4yMDDQ2NnJ3ow8ePDCoLk9Pzx7lHwiFQqSkpGiss7e3536/cOECzpw5g+DgYCxcuNDgentC9aghLS0NaWlpOreLRCKYmJhotdXFxaXHx5swYYLOfA7VuvHjx2ttGzduHHbt2oWkpCTs3r0bW7duNWjEBiGEENJvfXaOjo6IiIhAcXExysrK4OfnB4FAgNzcXMyaNavfI1RLS0u9d/n37t3D4cOH4eHhMWCvlq2oqMDt27cxc+ZMzJs3T2v7qVOncOXKFS6PoLsJoAwRFBSEjz/+GHfu3IFQKATQ+RjjwoUL8PPz0xrZoOLi4sIFCn/6058oUCCEEGKQXgcJra2tSEpKwqxZs+Du7g5LS0vcunULEokEM2bMgIWFBeLj45Gamordu3dj7ty5GDt2LKqrq3Hv3r0ejTLoCalUij//+c+Qy+WIjo7WO5LCxsaGyyPoDVUvQnR0NPeFra6lpQU3b97EhQsXEBER0WVdt27d4oZ6Pnr0CI2Njbh06RKAzrwPVS7F3Llz8eWXX2L//v1YsWIFbGxs8MUXX6C6uho7d+7s8hjOzs7YuXMnFyhs2bIFfn5+PT5vQggho0evgwQ+nw8fHx8UFRWhpqYGCoUCjo6OiI6ORnR0NIDOTH87OzucPn0aqampAAAnJyeEhIT0T+t1uHfvHmpqagB05ijoExISgldffbVXx+jo6EBRURG8vb11BggA8Mwzz8DOzg5isbjbIOHcuXO4cOECt1xfX493330XgGbugpmZGXbu3In09HSkpaVx0zJv3brVoFEgzs7OXI/Cnj178NZbb/V59AgZvczMzLh/15TwNnzR50i6wrA0ewYxQG1trcb8CaQTwzBwdXVFVVUVTUTTjb5eK5lMxj0+/E/A7/r9VdGMQo6JN/7Wp/rV60hNTe1T4iL9XRmGrpVh+Hw+nJycerzf0BxHRAghXeAp5VB2X6xHGKVc5+89wevlfoQMVRQkEEKGHZ+Sjwa0ft8Brp+Q4YJeA0gIIYQQnagngRAyLKgn2A0ElmW591GYmZn1edg2JQGSkYCCBELIsMAwTK8TAQ1FLzciRBMFCYQQQgBo9qboW29IL0t/9MSQoYGCBEIIIQA6XxXdH7PU9mX4JxlaKHGREEIIITpRTwIhhBAtt6a+DCWvc0IpRinnhoX+MPVlsDztiaZ4SvmAD00lg4+CBEIIIVqUPL7OWSdZPev7e3IrMjRQkEAIIcNQfw/Z7IcGqf1K0yOPFJSTQAghw5AqyXDNmjU6RyQMNobt4H4fCu0h/cPoPQmZmZk4deoUt2xiYgJ7e3tMnz4dsbGxGDNmDAAgPz8fhw4dQkpKCry9vbusMzY2ttvjLlmyhCsXGxuL+fPnc6+vrqmpwbp163TuJxQKsXfvXq31+/btw7Vr1xAZGYmEhIRuj6+iUCiwYsUKndvi4uLwq1/9SmNddXU1jh07hps3b4JlWUycOBFxcXF44okntPYvKipCTk4O7t+/DysrKwQHB2PZsmU0FpwQMuxJJBKkp6cjLi4Orq6uXW5/6qmnjNDCkcHoQYLKli1bIBAI0NbWhuvXr+P8+fO4desWkpOTe9yNlpycrHO9UqnEwYMH8fDhQ0ybNq3beiIiIvDss89qrNP1BVtfXw+JRAKg84s5Li4OpqY9u7TBwcGIjIzUWPf4G7saGhqwY8cO2NjY4NVXX4WpqSmys7Oxc+dO7N27V+MfiiqoCg8Px4svvoj79+8jIyMD9+/fx1tvvdWjthFCyFAik8lw7Ngx1NfX49ixYwgLC+ty++TJk2lIZi8NmSBBKBTCxsYGABAQEACpVIrCwkKUlZVh0qRJPapr4sSJOtenpaWhpqYGq1evho+PT7f1ODo66q1LXUFBARQKBaZNm4br16/jypUrCA4O7lGbbW1tuz3W6dOn0dLSgpSUFDg4OAAA/Pz8sG7dOvzjH//Aa6+9BqCzdyIjIwNPP/00Vq9eDQCYOnUqzM3N8de//hU3btxAQEBAj9pHCCFDxdmzZ9HQ0ACg8+bp5MmTmDdvnt7t586dQ0xMjFHaOtz1KUhoamrC8ePHIZFI0NjYCEtLS7i5uWHp0qXcl5BEIkFOTg5u374NhUIBJycnzJ49u9sPzNfXF4WFhairq+tLEzmFhYX44osvEBYWhvDw8H6pU0UsFsPOzg5r167F7373O4jF4h4HCYa4cuUKRCIRFyAAwJgxYxAUFIRvvvkGSqUSPB4PZWVlaGxsxJw5czT2nzVrFj788ENcuXKFggRChjn15ECZTNYvdWrUM0STD3/++WecO3eOO3+WZZGZmYknn3wSzs7OOrefPXsWwcHBcHFxMWbTh+xdJFgAACAASURBVKU+BQkffPABysvLsXz5cri5uaGlpQXl5eWQSqUAgLy8PBw+fBj+/v545ZVXMHbsWFRVVaGioqLbuqurqwGA613oi/Lychw5cgTe3t5c3oEhWJaFQqHQWMfj8TQef5SWlqKqqgqLFi2CjY0NAgMD8c0336Curg6Ojo4GH6ugoAC5ublgWRYTJkxAREQEQkNDue1tbW2ora3FrFmztPb19PSEWCxGbW0tXFxcuOs7YcIEjXKmpqZwc3NDZWWl3nbI5XLI5XJumWEYWFpaGnwehJDBoZ4cuH79+n6vn2E7wGJovaSKZVmkp6drjZ5gWRb/+7//i9dff13v9vT0dLzxxhvGHwUyzPQpSCgrK9O6Mw8MDATQ+aV29OhR+Pn5YceOHdwHIxKJdNalVCqhUCi4nITc3Fw4ODj0+FHD45qamvDOO+/AwsICGzduBJ+vPb5Xn4yMDGRkZGis27Ztm8ZdeF5eHgBwz8TCwsJw8eJF5OfnY8mSJd0eg2EYPPvss5g2bRocHR3R0NCAf/3rXzh06BBqa2uxdOlSAOACLysrK606VOuam5vh4uLSbVlVN5wu2dnZGomkXl5e2LdvX7fnQQghA62qqgolJSVa6xUKBUpKSnDjxg2d25VKJUpKSlBVVQU3N7fBaOqI0acgwcfHBwUFBbC2toZIJIJQKOQS9srKyvDo0SPMmzfPoMhN9excxc/PD4mJiXpft8qyLJRKzek7TExMNJaVSiUOHDiABw8eYNu2bT26sweAyMhIPPfccxrr1P/AWltbcfnyZUyePBnjxo0D0BkEOTk5QSwWY/Hixdy56+uR4PF4XC6Byi9+8Qvs2bMH2dnZiIiIgLW1Nbetq2v5+DZ9ZbuqIyYmBlFRUQaVJYQYj/r/Gw8cONAviXkymYzrlWCZIZOyxnF1dcXUqVNRWlqq8f9/ExMTTJ48GQEBATq383g8+Pv76xwFQbrWp7+CDRs2ICsrC3l5eTh58iQsLCwQFBSEuLg4NDU1AYDG8/OubN++HQKBACYmJnBwcND4YtSltLQUSUlJGusOHjwIZ2dnbjk9PR3FxcVYuXIlpk6d2sOzA+zt7bscbvn1119DJpNh5syZaGlp4dbPnDkTOTk5KCkpgUgk0jnMcd26dZg9e7beup977jlIJBKUl5cjICBAo7fgcY/3HKiXffw6SqVSnT0MKnw+v0e9LYQQ41AP4M3Nzfs/e38I3iAwDIO4uDhs3bpVa318fDx4PJ7e7StXrqSbnl7oU5BgY2ODhIQEJCQkoK6uDt9++y0yMjLQ2NjI3Y0+ePDAoLo8PT17lH8gFAqRkpKisc7e3p77/cKFCzhz5gyCg4OxcOFCg+vtCdWjhrS0NKSlpencLhKJYGJiotXW7hJoVM/UVH/UFhYWcHZ21pnPUVFRAQsLC27IpCoXoaKiQqPno6OjA/fv39fIdSCEkOHExcUFkZGROHPmDFiWBcMwiI2NhbOzM1iW1bl9wYIFGjeQxHD91p/k6OiIiIgIFBcXo6ysDH5+fhAIBMjNzcWsWbP6PYKztLTUe5d/7949HD58GB4eHv3y2lNdKioqcPv2bcycOVNj6I3KqVOncOXKFe7OvbsJoB5XVFQEU1NTeHl5ceuCgoLwz3/+Ew8fPuQCotbWVly5cgXTp08Hj9c5gaafnx/Gjh2L/Px8/OIXv+D2v3jxItrb2xEUFNSbUyaEkCFhwYIFuHDhAurr62Fra4tly5ahvr5e7/bH56Ahhut1kNDa2oqkpCTMmjUL7u7usLS0xK1btyCRSDBjxgxYWFggPj4eqamp2L17N+bOnYuxY8eiuroa9+7d69Eog56QSqX485//DLlcjujoaL0jKWxsbLg8gt5Q9SJER0dDKBRqbW9pacHNmzdx4cIFRERE6K3ns88+Q1VVFaZMmQIHBwc0NjbiX//6F4qLi7F8+XKNRwO/+tWvUFRUhJSUFCxdupSbTEmhUHAJjkDn87nf/OY3OHToED766CPMnDmTm0zp6aefpuGPhJBhzdzcHPHx8dyMio9Pcvf4dppIqfd6HSTw+Xz4+PigqKgINTU1UCgUcHR0RHR0NKKjowF0Zvrb2dnh9OnTSE1NBdA5i2BISEj/tF6H/9fevYc1daX7A//uQEKC3ML9IqCAYBFQrIpCRYY6YtUWrUDpeCmeOepY/Wlb7Ti1YsVLqY+2fazYImOdo+LxUm/gDYcWuVUdpgWq1hmsDI22ElBQMAgJJvn9wcluIgkkXBN4P8/DI1l77bXXXonk3WuvtbZIJEJtbS2AtjEKukyZMgXLly/v0jGePn2KoqIi+Pr6ag0QAOD555+HUCjEpUuXOgwSPDw8UFpaiu+//x5NTU3g8XgYPnw43nrrrXZrLdjZ2WHTpk04ePAgdu/eDYVCgYCAAKSkpLQbkBMVFQUzMzNkZWUhLy8P1tbWiIqKQmJiYpfOmRBiXHg8Hvt3VdcA776kPtCxL+ozZswYjBkzRmcvtWo76R5GSY/rInq4f/++xvoJpA3DMHBzc0N1dTU9+a4T1Fb666+2kkql7C3aWyHL2EdCM/JW+F/7ol26Rp3V8qSnp/fZ1Tt9rvTD5XLbLfWvD+Ob40IIIaTfcRStUE0iZBRqC6wptF8scHSkE9NGQQIhhJB2/G7s1Zo+Qkc6GZg4/V0BQgghhBgn6kkghBACQHMwpDqlUsk+K4LH43U6pd0YBlKSnkFBAiGEEABtgwB1DTh8dpohGRwoSCCEEMJS7zUwJG9nPQz69EAQ40NBAiGEEJZMJuuVlWr7clok6Tk0cJEQQgghWlFPAiGEEK1uB/03FBzdT4VlFK3slMifgv4bymfychStOqdSEtNAQQIhhBCtFByu1tUVtVFqyavQkZeYDgoSCCFkgDN0CqOpUSqVkEqlUCqVA/L8+hONSSCEkAFONRjxT3/6k94zF0yJVCrF0qVLB+z59ad+70k4duwYjh8/zr42MzODvb09xo0bh4SEBAwZMgQAkJ+fj88//xypqanw9fXtsMyEhIROjxsXF8fmS0hIQExMDPv46traWqxYsULrfj4+Pvjoo4/apW/btg3ff/89ZsyYgaSkpE6Pr/Lrr7/i66+/xs2bN1FTUwMOhwMPDw/MmjULYWFhGnm/+eYb7NmzR2s5e/fuhY2NjUZaUVERsrOzce/ePVhZWSE8PByvvfYazXcmxARcvXoVn332GebPn09PM+xl5eXl7GOlqa019XuQoLJu3TpYWlqipaUFpaWluHDhAm7fvo0tW7YY3HW0ZcsWrekKhQJpaWmor6/H2LFjOy1n+vTpeOGFFzTStH3BPnz4EOXl5QDavpjnz58Pc3P9mra8vBzl5eWIjIyEr68v5HI5iouL8fHHHyMxMRGvvvpqu31WrFgBV1dXjTRVMKWiCqqmTp2KN954A/fu3cOhQ4dw7949vPfee3rVjRDSP6RSKT777DPU1dXhwIEDeO6552j6YC+RSqU4cOAAHj58SG2thdEECT4+PuyVcEhICCQSCQoLC1FRUYGRI0caVJa/v7/W9H379qG2thZLliyBn59fp+U4OjrqLEtdQUEB5HI5xo4di9LSUpSUlCA8PFyvuk6ePBkzZszQCIRCQ0PR2NiI06dP45VXXmkXcHh5eWHYsGE6y5TL5Th06BBCQ0OxZMkSAEBQUBAsLCywe/duXLt2DSEhIXrVjxDS986ePYv6+noAwKNHj3D+/HnMmTOnn2s1MJ07dw6PHj0CQG2tTbeChMbGRhw+fBjl5eVoaGiAQCCAu7s74uPj2S+h8vJyZGdno7KyEnK5HE5OToiMjOz0TRgxYgQKCwvx4MGD7lSRVVhYiJycHERHR2Pq1Kk9UqbKpUuXIBQK8eabb2LZsmW4dOmS3kHCs7cIVPz8/PDDDz/gyZMnOvPoUlFRgYaGBvzud7/TSI+IiMBf//pXlJSUUJBAiJGqqanB+fPnoVQqAbQNyjt37hzCw8Ph4uLSpTJVZQFtV84d0diutl8XD6z3cbuKYRi0tLSoHVL/OvdGWw803QoSdu3ahaqqKiQmJsLd3R1NTU2oqqqCRCIBAOTl5WHPnj0IDAzE4sWLYWtri+rqaty5c6fTssViMQDdX6KGqKqqQkZGBnx9fdlxB/pQKpWQy+UaaRwOR+Oq/+bNm6iursbs2bNhY2OD8ePH48qVK3jw4AEcHR27XOcff/wRtra2sLKyardt69ataGxshKWlJUaNGoXExEQMHTqU3a5qXy8vL439zM3N4e7ujrt37+o8bmtrK1pb1Z4dzzAQCARdPg9CiP6USiUyMzPbfdGp0t95550ujdxXH8y3atUqvfdjlE+hRNcf1sQon3bpuN0hk8n0GnfVW2090HQrSKioqGh3ZT5+/HgAQEtLC/bv34+AgABs2LCBbezg4GCtZSkUCsjlcnZMQm5uLhwcHAy+1fCsxsZG7NixA3w+H6tXrwaXq9+cXwA4dOgQDh06pJG2fv16javwvLw8AEB0dDT77+XLl5Gfn4+4uLgu1fnvf/87/v3vf+OPf/wjOJzfJqAIhUK8+uqrGDFiBAQCAUQiEbKysrBu3Tps2bKFDQpUQZq2AMPKyortWtPm1KlTGgNJhw8fjm3btnXpPAghhqmursaNGzfapSsUCty4cQPV1dVwd3fvh5oNPNTW+ulWkODn54eCggJYW1sjODgYPj4+7P3ziooKNDc3Y9q0aXpFY6p75yoBAQFYunSpzkeOKpVKKBSaS3WYmZlpvFYoFNi5cyfq6uqwfv16g6/sZ8yYgcmTJ2ukqX9onjx5gn/84x947rnn2IGEwcHBcHJywqVLlzB37lz23DvrkVD5/vvv8be//Q3h4eGIiYnR2DZ27FiNAZeBgYEIDQ3Fu+++i2PHjmHNmjUa+XW1e0fvx5w5czBr1iy98hJCepabmxuCgoJw8+ZNjb9vHA4HgYGBcHNz61K56n9Hd+7c2eHAPKlUyl71K5nuDVtT37+z43YVwzCws7PDa6+9BkD/x1T3VlsPNN36BLz11ls4efIk8vLycPToUfD5fEyYMAHz589HY2MjAMDBwUGvspKTk2FpaQkzMzM4ODjA2tq6w/w3b95ESkqKRlpaWhqcnZ3Z15mZmbh+/ToWLFiAoKAgA88OsLe373C65bfffgupVIpJkyahqamJTZ80aRKys7Nx48YNBAcHQy6X4/XXX9fYd8WKFYiMjNRIKy0txSeffILQ0FCdUzCf5erqCn9/f/z0009smqoH4fHjx+3aUSKRaO1hUOFyuQb1thBCeg7DMJg/fz7ef//9dukLFizoctCuvp+FhYX+X9bdvUjo6nENOgSjcXtB3zbqrbYeaLoVJNjY2CApKQlJSUl48OABvvvuOxw6dAgNDQ3s1WhdXZ1eZXl7exs0/sDHxwepqakaafb29uzvxcXFOHv2LMLDw/Hyyy/rXa4hVLca9u3bh3379mndHhwcDDMzs3Z1fXZQTGlpKT7++GMEBQXhnXfe0XsKpYr6bQnVbYc7d+5o9Hw8ffoU9+7dQ1RUlEFlE0L6jouLC2bMmIGzZ89CqVSCYRjMnDlT4wKI9Axq68712BRIR0dHTJ8+HdevX0dFRQUCAgJgaWmJ3NxcRERE9HhUJhAIdF7li0Qi7NmzB56enr3yyFOg7Qu4srISkyZNwrRp09ptP378OEpKStgr9456JMrKyvDxxx8jMDAQa9asMShAEIvFuHXrFkJDQ9m0gIAA2NraIj8/HxMnTmTTL1++DJlMhgkTJuhdPiGk782aNQuXL19GXV0d7OzsMGPGjP6u0oA1c+ZMFBcX4+HDh9TWWnQ5SHjy5AlSUlIQEREBDw8PCAQC3L59G+Xl5QgLCwOfz8fChQuRnp6OzZs348UXX4StrS3EYjFEIpFBswwMIZFIsH37drS2tiI2NlbnTAobG5t2CxIZQtWLEBsbCx8fn3bbm5qa8OOPP6K4uBjTp0/XWc7NmzexY8cO2NvbY/bs2aiqqtLY7unpyc4uSElJwahRo+Dl5QVLS0uIRCJkZ2eDw+ForDJpZmaGefPm4fPPP8fevXsxadIkdjGl0NBQmv5IiJGzsLDAypUr2RUXaXGf3mNhYYGFCxeyKy5SW2vqcpDA5XLh5+eHoqIi1NbWQi6Xw9HREbGxsYiNjQXQNtJfKBQiKysL6enpAAAnJydMmTKlZ2qvhUgkQm1tLYC2MQq6TJkyBcuXL+/SMZ4+fYqioiL4+vpqDRAA4Pnnn4dQKMSlS5c6DBKuXbuG1tZW1NTUtBtjAQCbNm1iZ3h4enri22+/xZkzZyCTyWBjY4Pg4GDExcW1C3iioqJgZmaGrKws5OXlwdraGlFRUUhMTOzSORNC+tbEiRPh7e1t0Lx/0jVjxoyh5Zh1YJT0CSR6uH//vsb6CaQNwzBwc3NDdXU1/THvBLWV/nq6rQx5CqRUKmVv094KWdbho6IZeSv8r32hM6/69vT09F4buOjq6gqRSERPgewAl8uFk5OTwfsZzbLMhBBCegfDMAO6G111fhR89jwKEgghhGjFUbRC0cF2RtGq9Xf1/YlpoyCBEEKIVn439uqdd4QBeYnp4HSehRBCCCGDEfUkEEIIYfF4PHY2WmcMGRCp73LJxLhQkEAIIYRl6CBHfZ64SEwXBQmEEGIE1K/KAeDRo0dobGwEl8vVa0ofTf0jvYGCBEIIMQIymaxby8j31joEZHCjgYuEEEII0Yp6EgghxMjUvFIDl+y2J8WKXxVDaa59kSDmKQPXk11/Bg0hnaEggRBCjIx6UKA0V+oMEgjpbRQkEEJIP1MqlZBKpb1Wtr7TFAl5FgUJfUD9Mc66xMXFISEhAQkJCYiJiWEfpV1bW4sVK1YAAFatWoWIiAiN/Y4dO4bjx49j7969sLGxAQDs3r0bBQUFbB6GYSAUCuHv74+4uDh4eXn11KkRQnqATCbDqlWreq1s1YBIGtxIDEVBQh/YsmWL1nSFQoG0tDTU19dj7NixnZZz5MgRhIWFwdy887eNx+Nhw4YN7HHEYjFOnDiB5ORkfPrpp7C3tzfsJAgZZMrLy5GZmYn58+drfYxwZ9sJGQhodkMf8Pf31/pz+fJl1NbW4r/+67/g5+fXYRmhoaGoqalBbm6uXsdkGIY9zsiRIxEVFYWlS5eiubkZpaWlPXFahAxYUqkUBw4cQF1dHQ4cONDuVkBn2wkZKChI6CeFhYXIyclBdHQ0pk6d2mn+UaNGYfTo0Thx4gSam5u7dExLS0sAgJmZWZf2J2SwOHfuHB49egSgbVGj8+fPG7SdkIGCgoR+UFVVhYyMDPj6+rJjD/Qxb948PH78GNnZ2Xrll8vlkMvlkMlkuHPnDg4ePIghQ4bodWuDkMGqpqYG58+fh1LZNqNAqVTi3LlzqKmp0Wt7V6jK+i1B3x1/+1Uqler80XkcQjpBYxL6WGNjI3bs2AE+n4/Vq1eDy+Xqve+wYcMQERGBs2fPIiYmBnZ2djrzSqVSvP766xppQqEQa9euha2trc79Wltb0dqq9ox4hoFAINC7joSYMqVSiczMzHZfpqr0t99+u8Pt77zzTpdmD6gvxwwAjFy/MtTz6TPwUSaT0bMWiEEoSOhDCoUCO3fuRF1dHdavXw9HR0eDy0hMTMTVq1fx1VdfYfHixTrz8Xg8pKSkAGj7A1ZfX4/z588jNTUV77//Pvz9/bXud+rUKRw/fpx9PXz4cGzbts3gehJiiqqrq3Hjxo126QqFAjdu3MC1a9c63F5dXQ13d/e+qCohfYKChD6UmZmJ69evY8GCBQgKCupSGc7Ozpg2bRouXryIWbNm6czHMAx8fX010kaPHo1ly5Zh//792Lp1q9b95syZo1Euzakmg4mbmxuCgoJw8+ZNKBQKNp3D4SAwMBAhISEdbndzc+vScZ99jLLSTL/bAur5du7cqXV6o1QqZXsZ6HHNxFA0JqGPFBcX4+zZswgPD8fLL7/crbLmzp0LHo+Hw4cPG7SfhYUFXFxcIBKJdObhcrmwtLRkf+hWAxlMGIbB/Pnz2wXHDMNgwYIF4HA4HW7valDdbj99i1HLZ2FhofNH53EI6QQFCX1AJBJhz5498PT07NZT3lSsra0RGxuLq1evorKyUu/9WlpaIBaLOxyTQMhg5+LighkzZrBfqAzDYObMmXB2dtZrOyEDCd1u6GUSiQTbt29Ha2srYmNjcefOHa35bGxs4Oqq/4NaZs6ciYsXL6KsrEzrdqVSiVu3brG/19fX48KFC2hqakJ8fLzhJ0LIIDJz5kwUFxfj4cOHsLOzw4wZMwzaTshAQUFCLxOJRKitrQUApKWl6cw3ZcoULF++XO9yLSwsEB8fj4yMDK3bZTIZ1q9fz762tbWFh4cH1qxZgwkTJuh9HEIGIwsLCyxcuJBdUfHZe/2dbSdkoGCUNHGW6OH+/fsaUyNJG4Zh4ObmhurqapqD3glqK92USiUeP37MDjAUvypmHwFdnVDd4aOi3Y61DZbU9VyGgf6AJ/pc6YfL5cLJycng/agngRBC+hnDML3WG9GbZZOBj4IEQggxMsxTRuvvHeUjpDdQkEAIIUbGJduF/V1124GQ/kBTIAkhhBCiFfUkEEKIEeDxeEhPT2dfC4VC1NTUgMvl6jXYkFZTJL2BggRCCDEC6gMMGYaBnZ0dmpubacQ+6VcUJBBi5NSnsOmbz9imuhlbfQgh+qEggRAjJ5PJemQ57/6Unp5OjygmxATRwEVCCCGEaEU9CYSYEPGr4g5X31NNl+soX19Rrw8hxDRRkECICVGaK/X68tc3HyGEdISCBGJyjHmAHiG60OeWmKJ+CRKOHTuG48ePs6/NzMxgb2+PcePGISEhAUOGDAEA5Ofn4/PPP0dqaip8fX07LDMhIaHT48bFxbH5EhISEBMTgz/+8Y8AgNraWqxYsQIAsGrVKkRERGit8969e2FjYwMA2L17NwoKCtg8DMNAKBTC398fcXFx8PLy6rROAHDu3Dn861//QlVVFe7fv4/AwEBs3LhRa96GhgZkZmaitLQUUqkU3t7eSExMRHBwsEa+5cuXw9PTE3/5y1800r/55htkZGRg7NixePvtt01ybrX6QD5dD7UhxNjQ55aYon7tSVi3bh0sLS3R0tKC0tJSXLhwAbdv38aWLVsMjrK3bNmiNV2hUCAtLQ319fUYO3asXmUdOXIEYWFhMDfvvHl4PB42bNjAHkssFuPEiRNITk7Gp59+Cnt7+07LyM3NhYWFBYKCgvD999/rzNfa2opNmzbhyZMnSEpKgq2tLXJycvDhhx8iOTkZgYGBHR4nOzsbmZmZiIyMxLJly2BmZtZp3QghvaO8vJx91PSYMWP6uzqEaNWvQYKPjw97VR4SEgKJRILCwkJUVFRg5MiRBpXl7++vNX3fvn2ora3FkiVL4Ofn12k5oaGhKCsrQ25uLl566aVO8zMMo3HskSNHwtHREZs2bUJpaSmmTp3aaRmffPIJOJy2iSarV6/WmS8vLw93797Fli1b2GOOGjUK7777LjIzM/Hhhx/q3Pd///d/cfr0aUyfPh2LFi2irk5C+pFMJsOBAwfw8OFDHDhwAM899xz1LBCjZHCQ0NjYiMOHD6O8vBwNDQ0QCARwd3dHfHw8QkJCALRFyNnZ2aisrIRcLoeTkxMiIyMxZ86cDsseMWIECgsL8eDBg66dzTMKCwuRk5OD6Ohovb6sgbYvXYVCgRMnTiAqKgoCgcDg41paWgKA3lfqqgChMyUlJXB3d9cISszMzDB58mQcPnwY9fX17XouFAoFvvzyS+Tm5mrcbiGE9J+cnBw8evQIAPDo0SOcP3++07+PhPQHg4OEXbt2oaqqComJiXB3d0dTUxOqqqogkUgAtF3t7tmzB4GBgVi8eDFsbW1RXV2NO3fudFq2WCwGALZ3oTuqqqqQkZEBX19fdtyBvubNm4e1a9ciOzsbr732Wqf55XI5+69YLMbBgwcxZMgQvW9v6Ovu3btae1i8vb3Z7epBglwux2effYYrV64gKSkJM2bM6NH69Bf1ZWqlUmk/1qStJ6mlpQVSqbTXls/VOEdTmrCgVlepVNonbWXM1N/HnJwctg2USiXOnTuH8PBwuLi46NqdkH5hcJBQUVHR7sp8/PjxAICWlhbs378fAQEB2LBhA9ul/eygOhWFQgG5XM6OScjNzYWDg4PBtxqe1djYiB07doDP52P16tXgcrkG7T9s2DBERETg7NmziImJgZ2dnc68UqkUr7/+ukaaUCjE2rVrYWtr26X66/L48WNYWVm1S1elPX78WCP9hx9+AADMmTNH7wChtbUVra2t7GuGYbrUm9Kb1JcoXrVqVT/WpO8xcgZKrml8wTLy325pDbb3yVBKpRKZmZl455136FYgMSoGBwl+fn4oKCiAtbU1goOD4ePjww7wq6ioQHNzM6ZNm6bXB33JkiUarwMCArB06VKdI+6VSiUUCoVG2rNd+gqFAjt37kRdXR3Wr18PR0dHQ06PlZiYiKtXr+Krr77C4sWLdebj8XhISUlh61dfX4/z588jNTUV77//PntrQNXboMLhcLr0x6CjfZ7dNmzYMEgkEuTk5OD555/XOW5D3alTpzRmngwfPhzbtm0zuJ6EEN2e/TumUChw48YNVFdXw93dvZ9qRUh7BgcJb731Fk6ePIm8vDwcPXoUfD4fEyZMwPz589HY2AgAcHBw0Kus5ORkWFpawszMDA4ODrC2tu4w/82bN9kvZJW0tDQ4OzuzrzMzM3H9+nUsWLAAQUFBBp7db5ydnTFt2jRcvHgRs2bN0pmPYZh20zNHjx6NZcuWYf/+/di6dSsAtOttePPNNxEVFWVQnaytrdv1FgBgb/U828tgb2+PNWvWICUlBVu3btUIWnSZM2eOxvka41WNehC5c+fOfh3wxTAMXF1dIRaLe/V2g+pKXGlmGr0IgGZdd+7cCT6f3+tt49CijQAAIABJREFUZczU30cOh6MRKHA4HAQGBsLNza2/qkeIVgYHCTY2NkhKSkJSUhIePHiA7777DocOHUJDQwP75VJXV6dXWd7e3gaNP/Dx8UFqaqpGmvo9+OLiYpw9exbh4eF4+eWX9S5Xl7lz5+LSpUs4fPgwhg4dqvd+FhYWcHFxgUgkYtOerbd6YKMvLy8v3L17t126aryHp6dnu23Ozs744IMP2EBh3bp1CAgI0HkMLpdr8O2ZvqYeuFhYWPR7kMDn82FhYdE3X3zGF7PpplZX1fvUp21lQhiGwYIFC4wyKCeDW7ce8OTo6Ijp06cjJCQEVVVVCAgIgKWlJXJzc3vlj4BAIICvr6/Gj+pWh0gkwp49e+Dp6dljT8yztrZGbGwsrl69isrKSr33a2lpgVgs1hiT8Gy9O+s10WbChAn49ddf8dNPP7FpcrkcRUVFGDFihM41GZydnbFx40ZYW1vjww8/xL///W+Dj00I6TnTp09nAwKGYTBz5swuXTgQ0tsM6kl48uQJUlJSEBERAQ8PDwgEAty+fRvl5eUICwsDn8/HwoULkZ6ejs2bN+PFF1+Era0txGIxRCKRwbMM9CWRSLB9+3a0trYiNjZW50wKGxsbuLoa9sCZmTNn4uLFiygrK9O6XalU4tatW+zv9fX1uHDhApqamhAfH6/XMSorK3H//n0AbW2sVCpx9epVAG3BhZOTEwDgd7/7HS5evIhPPvkE8+bNg42NDf7+97/j3r17SE5O7vAYTk5O2LhxI1JSUvDhhx/ivffew3PPPadX/QghPWv69Om4cuUKHj58CDs7uwEz84gMPAYFCVwuF35+figqKkJtbS3kcjkcHR0RGxuL2NhYAEB0dDSEQiGysrKQnp4OoO0LasqUKT1f+/8jEolQW1sLoG2Mgi5TpkzB8uXLDSrbwsIC8fHxyMjI0LpdJpNh/fr17GtbW1t4eHhgzZo1mDBhgl7HyMnJ0VjeGWhbYAnQHLvA5XKRnJyMzMxM7Nu3D1KpFMOGDcN7773X6WqLQFvPj+rWgypQ0Gc/QkjP4vF4WLhwIbviIi2kRIwVo6Sbg0QP9+/f15ga2Z+M6UE5DMPAzc0N1dXVvTpwUXULrTqhusNHRbsdc+s0X19Rr096ejr4fH6vt5UxM+Rz2xefq4GC2ko/XC6X7ZU2BD0FkpgchmHoyouYHPrcElNEQQIhJoR52sHVp9q2jvL1FWOoAyGkeyhIIMSEuJ7Ub+CtvvkIIaQj3ZoCSQghhJCBi3oSCDFyPB6PnSnUEWMa0PksXUutE0KMGwUJhBg5Qwa88fn8Xq4NIWQwoSCBENIj1HsytOmPR0X3dO+KsfXQENLbKEgghPQImUzWY0uiG6v09HSaxkgGFRq4SAghhBCtqCeBENLjxK+K+33FR6BtrQbVdNCu1km9DEIGGwoSCCE9TmmuNIogQZ0x1okQY0dBAiGEZczTKE0ZtSsxVTQmgRDCUg0+/NOf/tThTAViGGpXYqr6pSfh2LFjOH78OPvazMwM9vb2GDduHBISEjBkyBAAQH5+Pj7//HOkpqbC19e3wzITEhI6PW5cXBybLyEhATExMfjjH/8IAKitrcWKFSsAAKtWrUJERITWOu/duxc2NjYAgN27d2s84plhGAiFQvj7+yMuLg5eXl6d1qmjuv/hD3/A7NmzNdJqampw8OBB3LhxA3K5HP7+/pg3bx58fHza7f/tt9/i9OnTuHfvHqysrDBp0iQkJibSXHpCCCF66dfbDevWrYOlpSVaWlpQWlqKCxcu4Pbt29iyZYvB3XFbtmzRmq5QKJCWlob6+nqMHTtWr7KOHDmCsLAwmJt33jw8Hg8bNmxgjyUWi3HixAkkJyfj008/hb29vV7HnDhxImbNmqWR5ujoqPG6sbERGzZsgJWVFZYtWwYul4vTp08jJSUFqampcHd3Z/MWFRVh165diI6OxhtvvIHq6mocOnQIv/zyC9avX69XnQghhAxu/Rok+Pj4sFflISEhkEgkKCwsREVFBUaOHGlQWf7+/lrT9+3bh9raWixZsgR+fn6dlhMaGoqysjLk5ubipZde6jQ/wzAaxx45ciQcHR2xadMmlJaWYurUqXrV39bWVuc5qGRnZ6OxsRFbtmxhnws+cuRI/L//9/9w9OhRvP322wDagpXMzEyMHj2anbceFBQEgUCAzz77DGVlZQgNDdWrXoQQQgYvg4OExsZGHD58GOXl5WhoaIBAIIC7uzvi4+MREhICACgvL0d2djYqKyshl8vh5OSEyMhIzJkzp8OyR4wYgcLCQjx48KBrZ/OMwsJC5OTkIDo6Wu8v61GjRkGhUODEiROIioqCQCAw+LiWlpYA2m6j9KSSkhIEBQWxAYLqWGFhYSgsLIRcLoeZmRlu3bqFhw8fYuHChRr7T5w4ERkZGSgpKaEggWilvhKiVCo1aF+N/ANpEoHauRjaJtr266vVJgnpCQYHCbt27UJVVRUSExPh7u6OpqYmVFVVQSKRAADy8vKwZ88eBAYGYvHixbC1tUV1dTXu3LnTadlisRgA2N6F7qiqqkJGRgZ8fX3ZcQf6mjdvHtauXYvs7Gy89tprneaXy+Xsv2KxGAcPHsSQIUP0vr0BAMXFxcjLy4NSqYSnpyemT5+O3/3ud+x2mUyGmpoaTJgwod2+Xl5e7HZ3d3fcvXsXAODt7a2Rz9zcHB4eHux2bVpbW9Ha2sq+ZhimS4ESMU3qg+pWrVrV5XIYOQMld2B8GTLy3259dqdNVGQyGY0LIibD4CChoqKi3ZX5+PHjAQAtLS3Yv38/AgICsGHDBnZcQXBwsNayFAoF5HI5OyYhNzcXDg4OBt9qeFZjYyN27NgBPp+P1atXg8vlGrT/sGHDEBERgbNnzyImJgZ2dnY680qlUrz++usaaUKhEGvXroWtra1ex3vhhRcwduxYODg4oKGhAZcuXcIXX3yBmpoaJCYmAgAkEgmUSiWsrKza7a9KUwVqjx8/1kh/Nm9tba3Oupw6dUpjUOnw4cOxbds2vc6DEELIwGJwkODn54eCggJYW1sjODgYPj4+7AC/iooKNDc3Y9q0aXoNPFyyZInG64CAACxdulTnY2WVSiUUCoVG2rNd+gqFAjt37kRdXR3Wr1/fbvCfvhITE3H16lV89dVXWLx4sc58PB4PKSkpbP3q6+tx/vx5pKam4v3332fHGah6G1Q4HA7bRitXrtTYNnHiRHz00UfIysrCjBkzeqRnRV1H782cOXM0BlDSfO7BRf3/3s6dOw16ToFUKmWvtJVmA6MXAdA8F0PbREW9beix2cSUGBwkvPXWWzh58iTy8vJw9OhR8Pl8TJgwAfPnz0djYyMAwMHBQa+ykpOTYWlpCTMzMzg4OMDa2rrD/Ddv3mS/kFXS0tLg7OzMvs7MzMT169exYMECBAUFGXh2v3F2dsa0adNw8eLFdrMO1DEM02565ujRo7Fs2TLs378fW7duBYB2vQ1vvvkmoqKidJYbGRmJ0tJSVFZWIjQ0FFZWVmAYhu0tUKdKU/UcqNpRIpG06wWRSCRaexhUuFyuwT0vZOBQDwotLCy6/jCjgRRbqp1Lt9pEVRwF3sSEGBwk2NjYICkpCUlJSXjw4AG+++47HDp0CA0NDeyXaV1dnV5leXt7G3SV7OPjg9TUVI009SmGxcXFOHv2LMLDw/Hyyy/rXa4uc+fOxaVLl3D48GEMHTpU7/0sLCzg4uICkUjEpj1bb/XARhvV4CbVHxQejwdXV1etYzvu3LkDHo8HFxcXAGDXZ7hz545GveVyOX799dd2a0AQQggh2nRrCqSjoyOmT5+O69evo6KiAgEBAbC0tERubi4iIiJ6PGIWCAQ6F1USiUTYs2cPPD09e+xxtdbW1oiNjcWRI0cMmg3Q0tICsVisMSahs8WgnlVYWAgzMzONRZLGjx+P8+fP48GDB+xtlObmZpSUlGDcuHHsrZcRI0ZAKBQiPz8f4eHh7P5Xr15FS0sLwsLCDKoLIYSQwcmgIOHJkydISUlBREQEPDw8IBAIcPv2bZSXlyMsLAx8Ph8LFy5Eeno6Nm/ejBdffBG2trYQi8UQiUQGzzLQl0Qiwfbt29Ha2orY2FidMylsbGzg6mrY09xmzpyJixcvoqysTOt2pVKJW7dusb/X19fjwoULaGpqQnx8fKflZ2dn45dffkFQUJDGwMUffvgB8fHxGj0tr7zyCoqKivDRRx8hISGBXUyptbVV41gcDgfz5s1DWloaMjIyEBERwS6mFBISgjFjxhjUBoQQQgYng4IELpcLPz8/FBUVoba2FnK5HI6OjoiNjUVsbCwAIDo6GkKhEFlZWUhPTwcAODk5YcqUKT1f+/8jEonYEftpaWk6802ZMgXLly83qGwLCwvEx8cjIyND63aZTKaxgqGtrS08PDywZs0ardMVn+Xu7o7vvvsOpaWlaGpqAo/Hw7Bhw7QuDW1jY4OUlBQcPHgQu3fvhkKhgL+/Pz744AN4eHho5I2MjASHw0FWVhby8/NhZWWFyMjIdmMjCFHH4/HY/7c0wK7nULsSU8UoaWUPoof79+9rrJ9A2jAMAzc3N1RXVw/6RXKkUil7q686odooHsvMPGXgdswNQNfrpF5Genp6twcu6nVM+lzpjdpKP1wuV2MhPn3Ro6IJIT2OeWocI/jV69HVOhnLuRDSHyhIIIT0ONeTho396QvGWCdCjB2nvytACCGEEONEPQmEkB6hPjhPG4Zh4OrqCrFY3Gf3jpVKJfs8Ch6P1+1p2TTokAw2FCQQQnoEwzAdDupjGAZ8Ph8WFhZ9OsCMHqZESNdRkECMhvpVnyl59OgRGhsbweVyTXrJ3Z640iaEDCwUJBCjIZPJemy1TGK4vpreRwgxHTRwkRBCCCFaUU8CMUriV8VGsRhPZ5inDDu1zlTqrE69/oQQ8iwKEohRUporTe4L1xTrTAghHaEggfSqnp6CRkh30OeREMPQmATSq1SDEf/0pz+Z5MwFMrDQ55EQw/RLT8KxY8dw/Phx9rWZmRns7e0xbtw4JCQkYMiQIQCA/Px8fP7550hNTYWvr2+HZSYkJHR63Li4ODZfQkICYmJi2MdX19bWYsWKFQCg9QmMqjrv3buXfXzz7t27UVBQwOZhGAZCoRD+/v6Ii4uDl5dXp3W6d+8evv76a/z444+oqakBwzAYOnQoZs6ciYkTJ2rkVbWHNhkZGbCzs9NI+/bbb3H69Gncu3cPVlZWmDRpEhITE2neOCGEEL306+2GdevWwdLSEi0tLSgtLcWFCxdw+/ZtbNmyxeBuwC1btmhNVygUSEtLQ319PcaOHatXWUeOHEFYWBjMzTtvHh6Phw0bNrDHEovFOHHiBJKTk/Hpp5/C3t6+w/2vXbuGsrIyTJ48GX5+fpDL5bh8+TI++eQTJCQkIC4urt0+b775Jtzd3TXSrKysNF4XFRVh165diI6OxhtvvIHq6mocOnQIv/zyi8ajrQkhhBBd+jVI8PHxYa/KQ0JCIJFIUFhYiIqKCowcOdKgsvz9/bWm79u3D7W1tViyZAn8/Pw6LSc0NBRlZWXIzc3FSy+91Gl+hmE0jj1y5Eg4Ojpi06ZNKC0txdSpUzvcPzw8HDExMRpBUWhoKB4/foysrCzExsaCy+Vq7OPp6dlhz4pCoUBmZiZGjx7NrjsQFBQEgUCAzz77DGVlZQgNDe303AghhAxuBgcJjY2NOHz4MMrLy9HQ0ACBQAB3d3fEx8cjJCQEAFBeXo7s7GxUVlZCLpfDyckJkZGRmDNnTodljxgxAoWFhXjw4EHXzuYZhYWFyMnJQXR0dKdf1iqjRo2CQqHAiRMnEBUVBYFAYPBxLS0tAbTdRumMKkh6lp+fH0pLSyGRSCAUCg06/q1bt/Dw4UMsXLhQI33ixInIyMhASUlJnwUJ6svvSqXSDvNqbKdJAn1DrZ07e3+6i2EYtLS0QCqV9umyzOrUz7G/6kCIKTE4SNi1axeqqqqQmJgId3d3NDU1oaqqChKJBACQl5eHPXv2IDAwEIsXL4atrS2qq6tx586dTssWi8UAdH9xGqKqqgoZGRnw9fVlxx3oa968eVi7di2ys7Px2muvdZpfLpez/4rFYhw8eBBDhgzR+/aGNj/++CNsbGxga2vbbttHH32ExsZGWFpaYtSoUUhISNAY/3D37l0AgLe3t8Z+5ubm8PDwYLdr09raitbWVvY1wzBdCpRU1AeHrVq1Su/9GDkDJZf+iPc2Rv5bD5Yh789AIJPJaHwOIZ0wOEioqKhod2U+fvx4AEBLSwv279+PgIAAbNiwge1CDw4O1lqWQqGAXC5nxyTk5ubCwcHB4FsNz2psbMSOHTvA5/OxevXqdt31nRk2bBgiIiJw9uxZxMTEtBsQqE4qleL111/XSBMKhVi7dq3WL3h9fPPNN/jxxx+RlJQEDue3CSh2dnZ49dVXMWLECAgEAty5cwdZWVl4//33sXnzZgwbNgwA8PjxYwDtxymo0mpra3Ue+9SpUxqDSocPH45t27Z16TwIIYSYNoODBD8/PxQUFMDa2hrBwcHw8fFhB/hVVFSgubkZ06ZN02vg4ZIlSzReBwQEYOnSpTofx6pUKqFQKDTSnu3SVygU2LlzJ+rq6rB+/Xo4OjoacnqsxMREXL16FV999RUWL16sMx+Px0NKSgpbv/r6epw/fx6pqal4//332fEKqt4GFQ6Ho7WNysrK8OWXX2LixIntxkSMGTMGY8aMYV8HBgZi7NixWLNmDY4dO4Y///nPep1bR+/NnDlzMGvWLL3y6kP9vdy5c2eHzwaQSqXs1azSjHoR+oJ6O3f2/nRXfzwq+lnqnzF67DMhnTM4SHjrrbdw8uRJ5OXl4ejRo+Dz+ZgwYQLmz5+PxsZGAICDg4NeZSUnJ8PS0hJmZmZwcHCAtbV1h/lv3rzJfiGrpKWlwdnZmX2dmZmJ69evY8GCBQgKCjLw7H7j7OyMadOm4eLFixpfms9iGKbdIMLRo0dj2bJl2L9/P7Zu3QoA7Xob3nzzTURFRWmklZeXY8eOHQgJCcHKlSv1+oJ2dnbGyJEj8dNPP7FpqnaUSCTtekEkEonWHgYVLpdrcM9LR9TPwcLCQv8vIVrjpm+otbNB709XDtVPj4ruqD6EkI4ZHCTY2NggKSkJSUlJePDgAb777jscOnQIDQ0N7JdpXV2dXmV5e3sbNP7Ax8cHqampGmnqUwyLi4tx9uxZhIeH4+WXX9a7XF3mzp2LS5cu4fDhwxg6dKje+1lYWMDFxQUikYhNe7be6oEN0BYgbN++HYGBgVi9erVe0y9VlEqlxh881fiEO3fuaNRbLpfj119/bbcGBCGEEKJNt6ZAOjo6Yvr06bh+/ToqKioQEBAAS0tL5ObmIiIioscjdYFAoHPqn0gkwp49e+Dp6dljjxu2trZGbGwsjhw5YtBsgJaWFojFYo0xCR1NWfzhhx+wfft2jBw5Eu+++65BV/K1tbWoqKjQGPcxYsQICIVC5OfnIzw8nE2/evUqWlpaEBYWpnf5hBBCBi+DgoQnT54gJSUFERER8PDwgEAgwO3bt1FeXo6wsDDw+XwsXLgQ6enp2Lx5M1588UXY2tpCLBZDJBIZPMtAXxKJBNu3b0draytiY2N1zqSwsbGBq6thT7ybOXMmLl68iLKyMq3blUolbt26xf5eX1+PCxcuoKmpCfHx8Z2W/+9//xvbt2+HnZ0d5syZg59//llj+9ChQ9kplZs3b8Zzzz0HLy8vWFpasgMXGYbRmIXB4XAwb948pKWlISMjAxEREexiSiEhIRrjGgghhBBdDAoSuFwu/Pz8UFRUhNraWsjlcjg6OiI2NhaxsbEAgOjoaAiFQmRlZSE9PR0A4OTkhClTpvR87f+PSCRiR+ynpaXpzDdlyhQsX77coLItLCwQHx+PjIwMrdtlMpnGCoa2trbw8PDAmjVrMGHChE7Lv3btGmQyGe7fv49Nmza12/7BBx9g1KhRANoWUbp8+TLOnDkDmUwGGxsbBAUFYe7cue1WYIyMjASHw0FWVhby8/NhZWWFyMjIdmMjehuPx2M/BzRQjPQ3+jwSYhhGaQwjiIjRu3//vsb6Cb1BKpWyt4qqE6pN4rHLzFMGbsfcAJhOndWp1z89Pb3XBy66ubmhurraKAYuGjNqK/1RW+mHy+XCycnJ4P3oUdHEKDFPTWPkuXo9TaXO6kyxzoSQvkNBAjFKricNGztiDEyxzoQQ0hFO51kIIYQQMhhRTwIxGuqDykyJUChETU0NuFyuSS/QQwP5CCHPoiCB6MWQxZ26wxS/qOzs7Hp9UOdA0lefpYGA2kp/1FYd62r70OwGQgghhGhFYxII6Ybm5masXbsWzc3N/V0Vo0dtpT9qK/1RW/UuChII6QalUomqqiqan60Haiv9UVvpj9qqd1GQQAghhBCtKEgghBBCiFZmGzdu3NjflSDElHE4HIwaNQpmZmb9XRWjR22lP2or/VFb9R6a3UAIIYQQreh2AyGEEEK0oiCBEEIIIVpRkEAIIYQQrWgdSzLotbS04MiRI7hy5QokEgnc3d0xe/ZsREREdLjfxo0bcfPmTZ3bMzIyYGdn1y5dJpPh3XffRXV1NebPn49XXnml2+fQV/qqrVpaWpCVlYXLly/j/v374PP58Pb2xpIlS+Dm5tZj59Nb+qKdWltbceHCBRQUFKC2thZ8Ph/Dhw/H3LlzERAQ0KPn05u62lYAcOPGDZw6dQoikQhSqRQuLi6Ijo7G9OnTweFoXgNfu3YNR48ehUgkgoWFBcaOHYv58+fD1ta2t05tQKAggQx6O3bsQGVlJf7whz/Azc0NxcXF2LlzJ5RKJV544QWd+/33f/83njx5opEmk8mwdetW+Pj4aA0QAODIkSNoaWnp0XPoK33RVi0tLUhJSUF9fT1mz54Nb29vPHnyBBUVFZBKpb12bj2pL9ppz549KCoqwpw5cxAUFASJRILTp09j48aN2Lx5M/z8/Hrt/HpSV9vq2rVr2Lp1KwIDA7F06VJYWFjgu+++w//8z/+gpqYGixYtYvPevHkTqampCA0NxZ///Gc0NDTg0KFD2LRpEz766CNwudy+OFWTREECGdRKS0tx7do1rFy5kv2DFBQUhAcPHuDgwYMIDw9vd0WiMnTo0HZp+fn5kMvlePHFF7Xuc/v2beTk5GDlypX45JNPeu5E+kBftdWRI0fwyy+/YMeOHXBxcWHTx40b14Nn03v6op1aW1tRXFyMF154AYmJiWx6QEAAli5diqKiIpMIErrTVvn5+TA3N8fatWvB5/MBACEhIaiurkZ+fr5GkHDw4EG4ublh9erV7DRJZ2dnJCcn49KlS5g2bVovn6npojEJZFArKSkBn8/HpEmTNNKjoqLw8OFD/PTTTwaVd+nSJfD5fISHh7fb9vTpU3zxxReIiYmBj49Pt+rdH/qiraRSKb755htMmjRJI0AwJX3RTgzDgGEYWFpaauQVCARgGMZkroy701bm5uYwNzdv9+RYS0tLjbT6+npUVlYiMjJSYx2FgIAAuLm5oaSkpIfOZmCiIIEManfv3oWHh0e7RVi8vb3Z7fqqrq7Gv/71L4SHh7NXNuqOHz+OlpYWvPbaa92rdD/pi7b6z3/+A6lUCldXV/z1r3/FokWL8Prrr+Mvf/kLSktLe+ZEellftJO5uTliYmJQUFCAkpISPHnyBLW1tdizZw8sLS0xderUnjmZXtadtvr973+Pp0+f4m9/+xvq6+vR1NSEwsJClJSUaIzzuXPnDgDAy8urXRne3t4GvR+DEd1uIIOaRCKBs7Nzu3QrKysAwOPHj/UuKy8vDwAQHR3dbtvPP/+MrKwstmu0sbGxizXuP33RVvX19QCArKwseHl5Yfny5eBwODhz5gy2bduG9957D2PGjOnqKfSJvvpMvfHGG7C0tMTHH3/MPtzI0dERH3zwAVxdXbtS9T7XnbYaMWIENmzYgE8++QQXL14E0Lby4h/+8Ae8/PLLGsdQL/PZ4xjyfgxGFCSQQY9hmC5tUyeXy1FQUABPT0/4+/u32/bFF18gPDzc6L/gOtPbbaX6sjM3N8e6desgEAgAAKNGjcLKlStx4sQJk2jD3m4nADh58iTOnDmD+Ph4PPfcc2hubkZOTg42b96M9evXY/jw4V2uf1/qalv95z//wY4dO+Dn54epU6fCwsICN27cwJEjRyCTyRAXF6dXWfq+H4MVBQlkUNN1JdHR1Yc2ZWVlePToEWJjY9ttO3fuHGpqavD222+jqakJANDc3AygbQBaU1MTBAKBzgFaxqIv2kpVRkBAABsgAICFhQUCAwPxz3/+sytV71N90U6//PILjh07hnnz5ml0rY8ZMwbvvPMODhw4gA8++KCLZ9B3utNWX375JWxtbfHuu++y/3eCgoLA4XDw1VdfYfLkyXBxcemwV0Iikej9fgxWxv1XiZBe5uXlhV9//RVyuVwjXXUf09PTU69y8vLyYG5ujsjIyHbb7t69iydPnmDlypVYtGgRFi1ahHfffRcAcPToUSxatIg9njHri7ZS3YvWRqlUmsRVX1+0k0gkglKpbDeDwdzcHN7e3ibxeQK611Y///wzfHx82gXXvr6+UCqV+PXXX9ljqJf57HH0fT8GKwoSyKA2YcIEtLS04B//+IdGekFBAYRCIUaMGNFpGY8ePUJZWRnGjx8Pa2vrdttnz56NDz74QONn1apVANoGX5nKPeS+aCuhUAh/f39UVFRorBcglUrxr3/9S2u3u7Hpq3YCgFu3bmmkt7a2oqqqCg4ODt04g77TnbYSCoWorKyEQqHQSFe1ib29Pfuvn58fioqKNPLeunUL9+7dw4QJE3rqdAYkelQ0GdTc3NxQUVGBr7/+GlZWVnjy5AlOnTqFK1euYPHixRg2bBgA4IsvvsDHH3+MKVOmYMiQIRpl5OTk4Nq1a0hKStLZwHTuAAAEsUlEQVQ6bc/GxgbOzs4aP3w+H+fPn0d4eDgmT54Mc3Pjv/PXF20FAB4eHvj6669x/fp1WFtb4969e/jyyy9x7949rFixwui/APuinRwcHFBWVoYrV65ALpdDoVDg559/xt/+9jf8/PPPmD9/foe9MsaiO23FMAwKCgpQWVkJPp+Puro6fP311zhz5gyCgoI0btO4urri3LlzEIlEsLKywk8//YSMjAw4Ojpi8eLF9IjpDhj/XyZCetmaNWtw+PBhHDt2DBKJBB4eHli1apXGsrAKhQIKhQLanqx+6dIlODk5ITg4uC+r3S/6oq0CAgKwYcMGHDlyBLt27QLQNpJ948aNJtGTAPR+O3E4HCQnJyM7OxtXrlzBmTNnwOfzMXToULz33nsIDQ3ttXPraV1tq5deegn29vY4d+4c0tPTIZPJ4OzsjLi4OMycOVPjGKNGjcJ7772Ho0ePYtu2beyyzAsWLDCZNSX6C6PU9gklhBBCyKBHYxIIIYQQohUFCYQQQgjRioIEQgghhGhFQQIhhBBCtKIggRBCCCFaUZBACCGEEK0oSCCEEEKIVhQkEEIIIUQrChIIISYtISEBCQkJWLRoEfuUzWcdO3YMCQkJOH36tNb0hIQEbN26tcPjvP3222ze/Px8jW27d+/Wmk6IqaMggRAyIDQ1NeHcuXNd3v/69et49OiR1m3/+c9/2KcKEjKYUJBACDF5DMPA3Nwc586dg0QiMXj/4cOHQ6FQ4Ntvv9W6vaioiM1HyGBCQQIhxOSZm5sjOjoazc3NOHv2rMH7P//88xAIBCguLm63TaFQ4PLly3Bzc4Ovr29PVJcQk0FBAiFkQHj11VfB5XJx4cIFg3sTeDwewsLCUFlZiXv37mlsu379Oh4+fIjJkyf3ZHUJMQkUJBBCBgR7e3u8+OKLaG5uxpkzZwzeXxUEqG4tqKheU5BABiMKEgghA8bs2bPB5XKRk5ODx48fG7TvqFGjYG9vr3HLQSaT4Z///Cf8/f3h4uLS09UlxOhRkEAIGTDs7e0xderULvUmcDgcREREoKamBrdu3QIAlJSUoLm5mXoRyKBFQQIhZEBR701obGw0aF9VMFBYWAig7VaDmZkZwsPDe7yehJgCChIIIQOKUCjE73//e7S0tCA7O9ugfYcNGwYvLy9cuXIF9fX1uHbtGsaMGQNra+teqi0hxo2CBELIgDN79mzweDxcvHjR4N6EF154AY8fP0Z6ejrkcjndaiCDGgUJhJABx87ODtOmTYNUKkVWVpZB+06ePBkMw6C8vBwCgQDjxo3rpVoSYvwoSCCEDEixsbGwsLDA3//+dzQ0NOi9n4ODA0JDQ2FtbY3JkyeDx+P1Yi0JMW7m/V0BQgjpDba2tpg2bRrOnDmDgoICg/b9y1/+0ku1IsS0UE8CIWTAUvUmyGSy/q4KISaJggRCyIBlY2ODmJiY/q4GISaLUSqVyv6uBCGEEEKMD/UkEEIIIUQrChIIIYQQohUFCYQQQgjRioIEQgghhGhFQQIhhBBCtKIggRBCCCFaUZBACCGEEK0oSCCEEEKIVhQkEEIIIUQrChIIIYQQohUFCYQQQgjRioIEQgghhGj1/wFTJOjacKWiyQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# NMI\n",
    "model = models_plot\n",
    "score = list(\"NMI\" for _ in range(len(asw_plot)))\n",
    "value = nmi_plot\n",
    "d = {'model': model, 'score': score, 'value': value}\n",
    "df = pd.DataFrame(data=d)\n",
    "fig = plt.figure(dpi=100)\n",
    "grid = plt.GridSpec(1, 10, wspace=0.2, hspace=0.2)\n",
    "ax = plt.subplot(grid[:, 2:10])\n",
    "sns.boxplot(y=\"model\", x=\"value\", data=df)\n",
    "ax.artists[0].set_facecolor('tab:red')\n",
    "ax.artists[1].set_facecolor('tab:red')\n",
    "ax.artists[2].set_facecolor('tab:red')\n",
    "ax.artists[3].set_facecolor('tab:red')\n",
    "ax.artists[4].set_facecolor('tab:blue')\n",
    "ax.artists[5].set_facecolor('tab:blue')\n",
    "ax.artists[6].set_facecolor('tab:blue')\n",
    "ax.artists[7].set_facecolor('tab:blue')\n",
    "ax.artists[8].set_facecolor('tab:green')\n",
    "ax.artists[9].set_facecolor('tab:green')\n",
    "ax.artists[10].set_facecolor('tab:green')\n",
    "ax.artists[11].set_facecolor('tab:green')\n",
    "plt.xlabel('NMI', fontsize=15)\n",
    "plt.ylabel('', fontsize=15)\n",
    "plt.xticks(fontsize=12)\n",
    "plt.yticks(fontsize=12)\n",
    "plt.title('NMI (Cortex)', fontsize=18)\n",
    "# plt.savefig(\"cortex-nmi-10K.pdf\", dpi=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,1,'ARI (Cortex)')"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ARI\n",
    "model = models_plot\n",
    "score = list(\"ARI\" for _ in range(len(asw_plot)))\n",
    "value = ari_plot\n",
    "d = {'model': model, 'score': score, 'value': value}\n",
    "df = pd.DataFrame(data=d)\n",
    "fig = plt.figure(dpi=100)\n",
    "grid = plt.GridSpec(1, 10, wspace=0.2, hspace=0.2)\n",
    "ax = plt.subplot(grid[:, 2:10])\n",
    "sns.boxplot(y=\"model\", x=\"value\", data=df)\n",
    "ax.artists[0].set_facecolor('tab:red')\n",
    "ax.artists[1].set_facecolor('tab:red')\n",
    "ax.artists[2].set_facecolor('tab:red')\n",
    "ax.artists[3].set_facecolor('tab:red')\n",
    "ax.artists[4].set_facecolor('tab:blue')\n",
    "ax.artists[5].set_facecolor('tab:blue')\n",
    "ax.artists[6].set_facecolor('tab:blue')\n",
    "ax.artists[7].set_facecolor('tab:blue')\n",
    "ax.artists[8].set_facecolor('tab:green')\n",
    "ax.artists[9].set_facecolor('tab:green')\n",
    "ax.artists[10].set_facecolor('tab:green')\n",
    "ax.artists[11].set_facecolor('tab:green')\n",
    "plt.xlabel('ARI', fontsize=15)\n",
    "plt.ylabel('', fontsize=15)\n",
    "plt.xticks(fontsize=12)\n",
    "plt.yticks(fontsize=12)\n",
    "plt.title('ARI (Cortex)', fontsize=18)\n",
    "# plt.savefig(\"cortex-ari-10K.pdf\", dpi=100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Impact of the linearity on clustering scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_repeat = 10\n",
    "models_plot = []\n",
    "asw_plot = []\n",
    "nmi_plot = []\n",
    "ari_plot = []\n",
    "\n",
    "file_models = [\"/Users/jiazhao/Downloads/Results/Cortex/FA/\", \\\n",
    "               \"/Users/jiazhao/Downloads/Results/Cortex/scPI_FA/500/\", \\\n",
    "               \"/Users/jiazhao/Downloads/Results/Cortex/scPI_FA_linear/500/\"]\n",
    "name_models = [\"CORTEX_FA_scores.txt\", \\\n",
    "               \"_CORTEX_FAVAE_nonlinear_scores.txt\", \\\n",
    "               \"_CORTEX_FAVAE_linear_scores.txt\"]\n",
    "models = [\"FA\", \"scPI-FA\", \"scPI-FA-linear\"]\n",
    "for i in range(len(name_models)):\n",
    "    for j in range(n_repeat):\n",
    "        res = np.loadtxt(str(file_models[i]) + str(j) + str(name_models[i]))\n",
    "        asw_plot.append(res[0])\n",
    "        nmi_plot.append(res[1])\n",
    "        ari_plot.append(res[2])\n",
    "        models_plot.append(models[i])\n",
    "\n",
    "res = np.loadtxt(\"/Users/jiazhao/Downloads/Results/Cortex/ZIFA/CORTEX_ZIFA_scores.txt\")\n",
    "for j in range(n_repeat):\n",
    "    asw_plot.append(res[0])\n",
    "    nmi_plot.append(res[1])\n",
    "    ari_plot.append(res[2])\n",
    "    models_plot.append(\"ZIFA\")\n",
    "\n",
    "file_models = [\"/Users/jiazhao/Downloads/Results/Cortex/scPI_ZIFA/500/\", \\\n",
    "               \"/Users/jiazhao/Downloads/Results/Cortex/scPI_ZIFA_linear/500/\"]\n",
    "name_models = [\"_CORTEX_ZIFAVAE_nonlinear_scores.txt\", \\\n",
    "               \"_CORTEX_ZIFAVAE_linear_scores.txt\"]\n",
    "models = [\"scPI-ZIFA\", \"scPI-ZIFA-linear\"]\n",
    "for i in range(len(name_models)):\n",
    "    for j in range(n_repeat):\n",
    "        res = np.loadtxt(str(file_models[i]) + str(j) + str(name_models[i]))\n",
    "        asw_plot.append(res[0])\n",
    "        nmi_plot.append(res[1])\n",
    "        ari_plot.append(res[2])\n",
    "        models_plot.append(models[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = models_plot + models_plot + models_plot\n",
    "score = list(\"ASW\" for _ in range(len(asw_plot))) + \\\n",
    "        list(\"NMI\" for _ in range(len(asw_plot))) + \\\n",
    "        list(\"ARI\" for _ in range(len(asw_plot)))\n",
    "value = asw_plot + nmi_plot + ari_plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([0, 1, 2, 3, 4, 5]), <a list of 6 Text yticklabel objects>)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ASW\n",
    "model = models_plot\n",
    "score = list(\"ASW\" for _ in range(len(asw_plot)))\n",
    "value = asw_plot\n",
    "d = {'model': model, 'score': score, 'value': value}\n",
    "df = pd.DataFrame(data=d)\n",
    "fig, ax = plt.subplots(figsize=(12, 7))\n",
    "sns.boxplot(y=\"model\", x=\"value\", data=df)\n",
    "ax.artists[0].set_facecolor('tab:red')\n",
    "ax.artists[1].set_facecolor('lightcoral')\n",
    "ax.artists[2].set_facecolor('lightsalmon')\n",
    "ax.artists[3].set_facecolor('tab:red')\n",
    "ax.artists[4].set_facecolor('lightcoral')\n",
    "ax.artists[5].set_facecolor('lightsalmon')\n",
    "plt.xlabel('Model', fontsize=20)\n",
    "plt.ylabel('ASW', fontsize=20)\n",
    "plt.xticks(fontsize=15)\n",
    "plt.yticks(fontsize=15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([0, 1, 2, 3, 4, 5]), <a list of 6 Text yticklabel objects>)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ASW\n",
    "model = models_plot\n",
    "score = list(\"ASW\" for _ in range(len(asw_plot)))\n",
    "value = asw_plot\n",
    "d = {'model': model, 'score': score, 'value': value}\n",
    "df = pd.DataFrame(data=d)\n",
    "fig, ax = plt.subplots(figsize=(12, 7))\n",
    "sns.boxplot(y=\"model\", x=\"value\", data=df)\n",
    "ax.artists[0].set_facecolor('tab:red')\n",
    "ax.artists[1].set_facecolor('lightcoral')\n",
    "ax.artists[2].set_facecolor('lightsalmon')\n",
    "ax.artists[3].set_facecolor('tab:red')\n",
    "ax.artists[4].set_facecolor('lightcoral')\n",
    "ax.artists[5].set_facecolor('lightsalmon')\n",
    "plt.xlabel('Model', fontsize=20)\n",
    "plt.ylabel('ASW', fontsize=20)\n",
    "plt.xticks(fontsize=15)\n",
    "plt.yticks(fontsize=15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([0, 1, 2, 3, 4, 5]), <a list of 6 Text yticklabel objects>)"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x700 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ARI\n",
    "model = models_plot\n",
    "score = list(\"ARI\" for _ in range(len(asw_plot)))\n",
    "value = ari_plot\n",
    "d = {'model': model, 'score': score, 'value': value}\n",
    "df = pd.DataFrame(data=d)\n",
    "fig, ax = plt.subplots(figsize=(12, 7), dpi=100)\n",
    "sns.boxplot(y=\"model\", x=\"value\", data=df)\n",
    "ax.artists[0].set_facecolor('tab:red')\n",
    "ax.artists[1].set_facecolor('lightcoral')\n",
    "ax.artists[2].set_facecolor('lightsalmon')\n",
    "ax.artists[3].set_facecolor('tab:red')\n",
    "ax.artists[4].set_facecolor('lightcoral')\n",
    "ax.artists[5].set_facecolor('lightsalmon')\n",
    "plt.xlabel('Model', fontsize=20)\n",
    "plt.ylabel('ARI', fontsize=20)\n",
    "plt.xticks(fontsize=15)\n",
    "plt.yticks(fontsize=15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}