pymare_example.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Run a set of Estimators on a simulated dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "In /Users/tsalo/anaconda/envs/python3/lib/python3.6/site-packages/matplotlib/mpl-data/stylelib/_classic_test.mplstyle: \n",
      "The savefig.frameon rcparam was deprecated in Matplotlib 3.1 and will be removed in 3.3.\n",
      "In /Users/tsalo/anaconda/envs/python3/lib/python3.6/site-packages/matplotlib/mpl-data/stylelib/_classic_test.mplstyle: \n",
      "The verbose.level rcparam was deprecated in Matplotlib 3.1 and will be removed in 3.3.\n",
      "In /Users/tsalo/anaconda/envs/python3/lib/python3.6/site-packages/matplotlib/mpl-data/stylelib/_classic_test.mplstyle: \n",
      "The verbose.fileo rcparam was deprecated in Matplotlib 3.1 and will be removed in 3.3.\n"
     ]
    }
   ],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/tsalo/anaconda/envs/python3/lib/python3.6/site-packages/statsmodels/tools/_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.\n",
      "  import pandas.util.testing as tm\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200612-11:28:29,866 nipype.utils INFO:\n",
      "\t Running nipype version 1.4.2 (latest: 1.5.0)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:nipype.utils:Running nipype version 1.4.2 (latest: 1.5.0)\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy import stats\n",
    "import seaborn as sns\n",
    "\n",
    "from nimare.transforms import t_to_z\n",
    "\n",
    "import pymare"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def var_to_ci(y, v, n):\n",
    "    \"\"\"Convert sampling variance to 95% CI\"\"\"\n",
    "    term = 1.96 * np.sqrt(v) / np.sqrt(n)\n",
    "    return y - term, y + term"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Simulate some data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_studies = 40\n",
    "between_study_var = 200  # population variance\n",
    "between_study_sd = np.sqrt(between_study_var)\n",
    "true_effect = 20\n",
    "sample_sizes = np.round(np.random.normal(loc=50, scale=20, size=n_studies)).astype(int)\n",
    "within_study_vars = np.random.normal(loc=400, scale=100, size=n_studies)\n",
    "study_means = np.random.normal(loc=true_effect, scale=between_study_sd, size=n_studies)\n",
    "\n",
    "sample_sizes[sample_sizes <= 1] = 2\n",
    "within_study_vars = np.abs(within_study_vars)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Convert data types and match PyMARE nomenclature\n",
    "y = study_means\n",
    "X = np.ones((n_studies))\n",
    "v = within_study_vars\n",
    "n = sample_sizes\n",
    "sd = np.sqrt(v * n)\n",
    "z = y / sd\n",
    "p = stats.norm.sf(abs(z)) * 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot variable distributions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/tsalo/anaconda/envs/python3/lib/python3.6/site-packages/ipykernel/__main__.py:15: UserWarning: Matplotlib is currently using module://ipykernel.pylab.backend_inline, which is a non-GUI backend, so cannot show the figure.\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABHgAAALICAYAAAAE6EcMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOzdeZCcd37f98/v6fvununpuTAH7oPgDRK73BUpcbkWJa28VjmKJGYr1ioulVOuxHKcw5LKSSWuJJWKLVlKLJUkW9baJUqWZTlSJG28u+QuqeUusQTI5QkQAAHM4BjMffR9/vLH80zPDAiSIIhBT8+8X1Vdz9HPM/2bwYM+Pv39/X7GWisAAAAAAAB0L6fTDQAAAAAAAMAnQ8ADAAAAAADQ5Qh4AAAAAAAAuhwBDwAAAAAAQJcj4AEAAAAAAOhyBDwAAAAAAABdjoAHAAAAAACgyxHwAAAAAAAAdDkCHgAAAAAAgC5HwAMAALYVY8x/Z4z59zfs+3VjzK91qk0AAACbzVhrO90GAACAO8YYMyjpvKRha+2SMcYv6ZqkH7HWnups6wAAADYHFTwAAGBbsdZOSXpR0k96u56WNEe4AwAAtjMCHgAAsB19RdKXvPUvSfo3HWwLAADApqOLFgAA2HaMMWFJU5J+QNLLko5Yayc72yoAAIDNQ8ADAAC2JWPM70g6Lrd71pOdbg8AAMBmoosWAADYrr4i6V7RPQsAAOwAVPAAAIBtyRgzKumMpAFr7Uqn2wMAALCZqOABAADbjjHGkfTfSPpDwh0AALAT+DvdAAAAgDvJGBOTNC1pQu4U6QAAANseXbQAAAAAAAC6HF20AAAAAAAAutzH6qKVzWbt+Pj4JjUFAAAAAABg5zp16tSctbbvds79WAHP+Pi4Tp48eTuPAwAAAAAAgA9hjJm43XPpogUAAAAAANDlCHgAAAAAAAC6HAEPAAAAAABAl/tYY/AAAIDu8uyJyY4+/jPHRzv6+AAAADsFFTwAAAAAAABdjoAHAAAAAACgyxHwAAAAAAAAdDkCHgAAAAAAgC5HwAMAAAAAANDlCHgAAAAAAAC6HAEPAAAAAABAlyPgAQAAAAAA6HIEPAAAAAAAAF2OgAcAAAAAAKDLEfAAAAAAAAB0OQIeAAAAAACALkfAAwAAAAAA0OUIeAAAAAAAALocAQ8AAAAAAECX83e6AQAAANienj0x2bHHfub4aMceGwCATqCCBwAAAAAAoMsR8AAAAAAAAHQ5Ah4AAAAAAIAuR8ADAAAAAADQ5Qh4AAAAAAAAuhwBDwAAAAAAQJcj4AEAAAAAAOhyBDwAAAAAAABdjoAHAAAAAACgyxHwAAAAAAAAdDkCHgAAAAAAgC5HwAMAAAAAANDlCHgAAAAAAAC6HAEPAAAAAABAl/N3ugEAgO3t2ROTHX38Z46PdvTx0Vk7/frr9O8PAADuHip4AAAAAAAAuhwBDwAAAAAAQJcj4AEAAAAAAOhyBDwAAAAAAABdjoAHAAAAAACgyxHwAAAAAAAAdDmmSQcAAFtSq2WVrzS0XK63b4VqQ+V6Q8VqU+VaU8Vao70s1ZqqNVqqNVqqN1uqNVu6ulhWs2XVaFk1W1YtK1lrJUlW7rq1a+vGGBlJxkjGGDlGMjIyRnKMkc9xb36fUcBx2ut+x/H2Gfm8db9jVKo1FAr4FPI5CgUcRQI+xcN+xUNrt1jIr2jQJ2NMJ//cAACgyxHwAACAu6LaaGquUNN8oaq5QlVzhZrmClXNr1sulWtumFOqK19tyMtiPlTI7yga9Cka9CsUcBT0OQr6HQV8jowxCgecdjDjGDescUOcdWGOvJ1aC31a1nrBj7du5YVFLTVaVtV6w11vugFSo2XVaLbagZIkfe2d6Vv62zhGigX9iofdwGct/PEpHgooHnKDoVQkoHQ0qEw0qEw0oEzMXU9FAvI5BEQAAOxkBDwAAOATqzdbWi7VtVSua6Vc13LFrbh57vS0ppYrml6paL5Yu+m50aBP2XhIPbGgcomw9vXFlYoElIoElPSWq7fYuoqXaNCnSMAnv++De5w/e2Jys37lD9WybsXQTzw0rGrdrSaq1N2qo0K1oUKloWKtoXyloWK14e67yf6ZfEXFalP5Sl3FWlPN1s0TL2OkZDignlhQ6WhAmai7nF6uKBbyK+IFYPGQX4mwX4mQX6GA7y7/VQAAwGYi4AEAAB+pZa1WynUtlGpaLNa1WKppsVjztmvKVxq6MXqIBn0a741pIBXWA6NpDSTDyiVCysZD6o0H28tocPu9HXGMkeMzSoYDUvjO/ExrrQrVxtrfv1TTUmnt32LRW18q1TW9UtGZqRXNFqqqN28eCgV9juJe2JMI+xUPB5QK+zeEaslIQIEPCdAAAMDWsf3eUQEAgNtirVWp1tzQbWq1K9V8cWNQYCSlIm4XoX25hDKxgHqiQaWiAaXCa8HAM8dHO/cLbTPGGCXCASXCAY32Rm/pnGdPTKrebKlUa6pUW6sSylcaylfcbnCFSkPTK1Wdny2oUm+972dEgz6lvbAnHQ2qZ13XsJ5YUGEqgQAA2BIIeAAA2IFKNfdD/fRKxbu56+V6s32MY6SemFtpsy8XV2/c/UC/GuT4HSo7ukHA5ygVcZSKBD7y2GqjqZXyxoGtl71ud0ului7OFVVtbAyBIgGfMjG3W1g2HlK2XZ0VUozBowEAuGsIeAAA2MaqjaZm1gc5eXc9X2m0jwkHHPUnwjo6nPK6ULkf0NPRIAP37jAhv099CZ/6EqGb3m+tVbne1GKx3u6et9pdbHqlqjNTeTXXjYwdDjhe6ONeV32JsPqTIfXGQlxbAADcYQQ8AABsA/VmS+dnCnr3el7vTud19npeZ2fyurxQbh8T8BnlEmHtz8XVnwy3b8mwnyoL3BJjjKJBv6JBv4Yzkffd32xZLZVq7RnSVrv7XZor6vuXl9rH+RyjvnhIuWRIA951mEuElIkF5XAtAgBwWwh4AADoMpV6U2eu5/X2tWW9dXVFb19b1pnredW8rjMBn9HevrgeGMno0EBS/V7VBB+esdl8jlGv1z3roBIb7qs1WporbOwWOLlQ0htXltvHhPyOBlNhDaYjGk5FNJgOK5cIU+0DAMAtIOABAGALK1QbeufayoYw59xMoT1ddioS0NHhpL782LiODCV1ZDCp8WysPfNRp6YJX9Xpx8fWEfQ7GkpHNJTeWPlTqTc143UdnFou69pSRScvLei73qDefseoPxnWUDqiXemIdvVE1J8ME1YCAHADAh4AALaIpVJNb19b0VtXl/XWtRW9fXVZF+eLWh3SpC8R0tGhpD5/pF/3DKV0z1BSuzIRulehq4UDPo32RDXaszYzWMtazRWqurZU0dRSWdeWy3rr6rJeubQgyQ2LhtMRjWSiGulxl8lbGEQaAIDtjIAHAIAOKNeaevvasl6/sqzXLy/p9StLmpgvte8fTkd0dDipn3hwWEeH3TAnlwx3sMXA3eMYd7yoXCKsB0bSktwBnheKNU0ulHR5sawriyW9dH6uPahzKhLQaE9U49mYxnujarWsHLp2AQB2EAIeAAA2WaPZ0rmZQjvIef3yst6dzre7WQ2lwrpvV1o/9ciI7htO656hpDKxYIdbDWwtxqyN7/PgaEaSO7j41HJFlxdKurxY0sR8SW9edcf0+b3vXNKxsYwe2d2jR8d7dO+ulEJ+Xyd/BQAANhUBDwAAd5C1VpcXyl6Q4wY6b11dUbnelCQlw37dP5LWf3lor+4fSev+XSkqc4DbFPA5G7p3WWu1VKrr0nxRfp/R9y4u6JvvzkpyB3B+cDStx/Zm9djeXt0/km6PVQUAwHZAwAMAwCewUqnr9ctLenViSa9dXtTrl5e0WKpLcscJOTqU1E89MqIHRtK6fySt8d4oY+YAm8QYo0wsqEwsqGeOj0qS5gtVvXJpUa9cWtDLF+b1q984q1/5uhQN+vTIeI8e29urx/ZmdWQoyWxdAICuRsADAMAtarWszs8W9OrEol6bdAOdczMFWSsZI+3ri+vzR/p13660HhhJ6+BAggoBoMN64yE9fXRATx8dkCQtFms6cXFe33nPvf3vXz0jya2u+8y+rJ440KcnDvZpMBX5sB8LAMCWQ8ADAMAHWCrV9NrlJb02sajXLi/p+5NLylcbktwBXR8cTesL9w3pwVG3OicZZhYfYKvLxIJ6+uignj46KEmaWanouxfm9dL5Ob14dk5ffeu6JOlgf0JPHOzTEwf6dGw8w/g9AIAtj4AHAABJzZbVu9fzenVyrTrnwmxRkuQY6eBAUj/+wJAeGs3owdG09mRjdLUCtoFcMqwvPjCsLz4wLGutzk4X9MLZGb1wdlb/6qWL+u0XLyga9Omxvb164kCffvBgTiPrpnQHAGCrIOABAOxIlXpTr19e0smJRX3v4oJenVhsV+f0xIJ6aDStv/nQLj04mtZ9u9KKh3jJBLY7Y4wODiR0cCChn398r4rVhr773rxeODurb52d0TdOz0h6W3uyMT3udeX69J5ehQNU9wAAOo93qwCAHWGxWNOpCXeg1VcuLejNq8uqN91pyvfn4vrC/UN6ZDyjh8cyGu1hIGQAUizk11NH+vXUkX5Za3VxruiGPe/O6g++N6nf+84lhfyOju9xq3uePJTT7mys080GAOxQBDwAgG3HWqvFUl0T80W9eXVZJy8t6NxMQZIU8BndO5zSz312tx4Z69HDYxllYsEOtxjAVmeM0Z6+uPb0xfXlz+xWpd7UiYsLeuFdt7rnH//5O/rHf/6O9vTF9PnDbij00GiGmbkAAHcNAQ8AoOu1rNX15Yom5ou6NF/SxHxRKxW3u1Ui5NfD4xn9jQeHdWwso/tH0nSnAPCJhQM+d8atA336H3VElxdKeu70tJ47M6PffemifuvFC8pEA/qhQzl9/nC/fuBAH109AQCbilcZAEDXqTdburxY0oQX5kzMl1RttCS5Ux2PZ2Ma641pvDeqX3jqAN+gA9h0Iz1R/exndutnP7NbK5W6Xjw7q+dOz+i50zP6k1evKuhz9Km9vfr84Zw+d7hfQ2mmYQcA3FkEPACALa9UbWhioaRL80Vdmivq2lJFTeuOn5NLhHT/SFrjvVGN9caUjgQ2jJ9DuAPgbkuGA/rCfUP6wn1DajRbOjWxqG+cntY3Ts/oH/3p2/pHf/q2jgwm3fF9Dud0dCglh+cqAMAnZKz3BvlWHDt2zJ48eXITmwMA2G6ePTH5sY5fP37OJa/L1Wy+KskNa3alI+3qnNHeqKLBrf1dxTPHRzv6+B/37w9gc83mqzo9taIz11c0MV+SlVt5eGggqUODCe3tiyvgcz7x43T6uQcAcHuMMaestcdu59yt/a4YALDtrY6fc8nrarV+/JxwwNFYT0wPjqQ11hvTrkzkjnzwAYBO6UuE1Jfo0+MH+lSsNvTudF5nplb0/StL+t6lBQV8RvtzCR3ypmtPhAOdbjIAoEsQ8AAA7qpao6Uri6X2YMiTC2vj56QiAY1nYxrvdW+5ZEgO05UD2KZiIb8eGs3oodGMGs2WLs4Vdfr6ik5P5fXO1IqM3LF9Dg0kdHgwqVwitKELKgAA6xHwAAA2VbHaaFfmXJpfGz/HSOpPhvWAV50z3htVOsp05QB2Jr/P0f7+hPb3J/Tj91ldX6l4Xbny+to70/raO9PKRANuV66BhHZnY/JT0QgAWIeABwBwx1hrdXmhrFcuLbRv780WJXnj52Qi+uz+rMZ6oxrriSkSZLpyALiRMUaDqYgGUxE9eahfK+W6zlzP68z1FZ2cWNB3L8wr6He0PxfXoYGkDg4kmIIdAEDAAwC4fc2W1empFb1yaUEnLy3qlUsLmvEGRE6G/To23qN9fXGN9cY0zPg5AHBbkpGAHt3do0d396jebOm92YLOXM/r3et5vX3N7cq1KxPRoUG3umcgGe50kwEAHUDAAwC4ZeVaU69dXmyHOa9NLqlQdQdEHk5H9Om9vXpkvEePjPdofy4uxzHM4gQAd1DA53jdtJKy1mpqueKFPSv6xjvT+vo700pFAnpnakVPHe7Xp/f2KhygWhIAdgICHgDAB5paLuvUxKJOTSzq1YlFvX1tRY2WlTHSwf6EfuLBYR0bz+jYeI+G05FONxcAdhRjjIbSEQ2lI3ryUE75Sl1np/M6cz2v/+e1q/r9E5MKBxx9dl9Wnzvcrx86mNNAiuoeANiuCHgAAJKkRrOl01N5nZpY0KnJJb06sairS2VJ7nTlD4yk9fOP79Ej4z16aDSjVJSpewFgK0mEA3p4rEcPj/Xobz48rBMXFvT8mRl94/S0vnF6RpIbzv/gwT49caBPD49nFPJT3QMA2wUBDwDsUMulul6dXGxX6Hz/8pLK9aYkaTAV1sNjGf3tH9ith8cyOjyYZPwcAOgiIb9Pjx/o0+MH+vQ//fgRnZsp6FvvzuiFs7P63Zcu6rdevKBo0KfH9vbqiQN9euJATqO90U43GwDwCRDwAMAO0GxZnZ8p6PUrbmXOqYlFnZspSHJntzoymNRPPTKih8cyengsoyG6WwHAtmGM0YH+hA70J/Tzj+9VsdrQd9+b1wtnZ/WtszNedc/bGu+N6jP7svrsvqwe25ulUhMAugwBDwBsM9ZaXZov6Y0rS3rjyrLeuLKkt66utKtzUpGAHhpN64sPDOnhsR7dP5JSNMjLAQDsFLGQX08d6ddTR/plrdXFuaK+9e6sXjo/1x67xzHSvcMpN/DZn9XDY3TnAoCtjnf0ANDFVmdQeePKkl6/sqw3vUBnpeLObBXyO7pnyK3OuW9XSvftSmtPNibHMR1uOQBgKzDGaE9fXHv64vq5z+5WvdnS9y8v6dvn5vTS+Tn91osX9Bvfek/hgKNjYz361J4efWpPr+7blVbQT9ddANhKCHgAoIvMFap688qyXm9X5yxrrlCVJPkdo4MDCf3YfUO63wtz9vfHGTsHAHDLAj5Hj4z36JHxHv39zx9QvlLXiQsL+vb5Ob18YV7/5GtnJWlD4HN8T6/uJ/ABgI4j4AGALajRbOniXFHvTK3o9FTeW65oNu+GOcZI+/rievxAVvfvSuu+XSkdHkwqHKB8HgBw5yTCgXZ3LklaKNb0vYsLevnC/PsCnwdG0jo21qOHxzN6aITZFgHgbiPgAYAOstbq2nJF56bzOjdd0NnpvM5cz+vsdF7VRkuSFPAZ7csl9AP7szoymNTR4ZSODqcUD/EUDgC4u3piQT19dEBPHx2QJC0Wa/reJTfwOXlpUb/5wntqftNKkg70x73B+3v08FhG471RGUMXYQDYLMZae8sHHzt2zJ48eXITmwNgszx7YrLTTeiYZ46PdroJajRburJY1oW5gs7PFHR2uqBzMwWdn86rWGu2j8vGgzo0kNThwYQODyZ1eDCpvX3xT1T2vpP/7QEAd1et0dLlxZIy0YBOTizq1YnF9rhwmWhA9+5K6/5dKbf6dCSlXCLc4RZvjk6/9m6F9z4Abo8x5pS19tjtnMvXvwBwh7RaVtP5ii7OFd3brLecL2pyvqRGay1Qz8ZDOtAf108eG9G+XFwH+hPan4srEwt28DcAAOCTCfod7e2LtwOGVsvq/GxBJy8t6vXLS3r9ypL++TdntfqSOJgKtycBuHfY7W7clwh18DcAgO5FwAMAt8haq6VSXZcXS7q8UPaWJU0ulHRlsayri2XVmq328SG/o93ZmA7kEvrhewa0OxvT7mxM+/oIcgAAO4PjGB3oT+hAf6Id+pRqDb19bUWvX16dMGBJ//Ht6fY52XhIR4bcatYjg0kdGUxqdzYmP5MGAMCHIuABAEnNltVCsaaZfEXXlyuaWq5oarnsLpcqur7iblfqrQ3npaMBjWSiOjyY0F870q9dPVHt7o1pd19Mg8kw05EDAHCDaNDfnqlr1XKprrenlt2JBa65Ewv87ntzqjfdUp+Q39H+/rj25xLal4trXy6u/bm4RnuiBD8A4CHgAbAtWWtVqbdUqjX0/ctLWirVtFSqa6FY01yhqtl8VbOFqmZW3OV8oarWDUOS+Ryj/kRIg+mIjgwl9dThnAZSEe3KRDSSiWqkJ6JEmBlCAAD4pFLRgB7bm9Vje7PtfbVGS+/NFtqBz9mZgk5cmNd/eO1q+5igz9Gevpj25eLa0xfX7mxUY70x7e6NKR0NMKgzgB2FgAfAlmWtVa3ZUrXRUq3uLquNpkq1psq1pkq1hkq1pnfbuF6uN9uBzT/9+tkNP9fvGPUlQupLhNp9//sSIeUSIWXjIQ2kwhpKR5SNh+SjAgcAgI4I+p32hAPr5St1vTdb1LnpvM7PFnR+uqA3rizrL9+c2vBlTTLs13g2pvHemMZ7oxrpiWrY+5JmIBVWgMofANsMAQ+AO8Zaq3rTqtpouqFMYy2UWQtpmt6+1fs/eLveaOmj5vnzO0bRoE/RoF/RoE/9yZCiwZi3z93/I/cOKB0NKhMNKBMNKhUJ0HUKAIAulQgH9MBIWg+MpDfsrzaaurxQ1sS8O8nBxHxJl+aLeu3yov78jWsbwh/HSAPJsIYzEQ2nI9qViWowHdZAMqx+79YbC/J+AUBXIeABdjBrrRot6wYq9fVBy0eEMN6xtWZL1XWhTe0WAplVAZ9R0O9TyO+0b/GQXz2xYHu7fX9g9Rifgn5HkcBaeHMr04d/7nD/J/tDAQCALS/k97XH57lRtdHU1FJFV5fKurJY0tXFsq4slnVlqaxXLi3qz16/9r6u2gGfUS4RVn8ypP5kWH2JkHpiQfXGguqJeetxdzsdDVL1C6DjCHiALmStVbneVKHS0EqloUK1oXylrkKloXylofwN24VqQ2en86qsC3GqjaZqjdb73sx8EL9jFPSvBS0hv6No0Kd0dH0gs3bf+0Iav6NQwKegzz2ON0EAAOBuCfl9bnetbOym9zeaLc0Wqrq+XNH0ijvhwnS+qulld6KFd6fzeun8nFYqjZueb4yUiQbVE3NvxWpDsZBfsaBfsZBPsZBf0YBP4fbNUTjgo5sYgDuKgAe4y2qNlhu+VL0wptLYsF2oNrRyQziTr9Tbxxaq7q15C8lMNOhTPORXIuxXtdFSOOBTMhK4SQizVh3jVsusv99HIAMAALY1v8/RYCqiwVTkQ4+rN1taLNY0X6xpYXVZqK6te8uZfFXFuaLKteaHVjf7HaNQwKeIF/iE/Wvhz/ogyL3PUcDvKORzv0ALeu/hgj5HAZ9hQGkABDzArWq27FqlzGoYU/HCmHXb+Urdq6DxtqtuWOMGNw3VGq2PfKygz1Ei7Fc87IYz8ZBfIz1RJbywxt0faIc37jGB9rHJcECxkG/DtKHPnpjczD8PAADAthfwOcolw8olwx963Or7rpa1KtWaKlYbKteaqjSaqtSbqtRb3rKp8rr1Sr2p5Uq9vb46TfxHMdK68MfRv3l5QrGgT9GQX7GgT5GgT7GgX9GQT9GAW1UU9aqLIgG3wigSdAOmSNDdFwn4FA66ARLhEdAdCHiwrbW7MlUbKlab7aDF3b7JeqWhYq2xoVpmtatTsdb8yMdzjN4XvOQSYe3JroU1bkizdkw87AYy67dDft9d+OsAAABgMznGKB5yv4C7Hc2WbYc97UkommvjItaa3tLrer86+2guEVKp1tRyua6ppXJ7ltFirXlLXzZu/B2k8GrgE1gLgFariyIb9vnW7XPHTQwF1gKjSNC37mc5G84L+QmSgE+KgAdbhrVWlXqrPd11pd5sT3u9tt5YF9i4oU2+4q17wcyNwc2tjDFjjNovvrF14cxwOrIheEmEA0qENlbWJMKB9vGRgI8XJgAAANwRPse4Y/l8zIDomeOjH3hfo9lSqd5UqdpUseZWFhWqjXUVRU2Vay2V11UVlWve/npT1XrLO8Z9Hz6br647z61OqjU/Xogkue/H11cQrQZAq/vC64KicMBROHhDcOT3tSfm2DgupG/DhB2hdUMSMAQBthsCHkhyvx2oN1tqtKwaTfdJudG0ajSt6i33m4JG0z2m3rSqNVrrvklotmdhqjQ2zqpUbbgvApV1My+tvlCUamsvBKvrH4fPMYp5Y8zEw2vBzGAqrJgX1qy/Lx7yKR5yuy4lvOXq/QQzAAAA2An8PkdJn6NkOLBpj7FaebQW+rjBT3ndvmpjY3BUqbmfJd6/z+3iNleo3RBCuZ8tPgm/Y9oTgawff/KDAqH1E4fceHzA57QnJfE77rhIAZ+3v72+tvQ77phKAcfccAxjX+L27biAZ65Q1fOnZyRJVlbWSlbylmvbstY7xrvP2nXHudurbjx3/fbaMRvvb1mrlrVqtqxa3ra7btVqWTWtt7/1AcesP/cmxzdabhDTDmZaGwOaRquleqOluhfo3OpMSrfK75h2qeXNngTT0aAGU+5U15Ggtwz4FAn61637blj3t9fjIb/CAco4AQAAgK3mdiuPPq5Wy6rSWAt71n/BfNP1dV86f/D9a+vlelNL5doHHnOnP0OtcowbxAXXBT9+x8gxRn6fkc8YOY5p7/M5627rth3HyGckn+PI58jb78hn1D7ft+5nGEnGGBnjdi80co8zkrR+n3eMMavnbDxe645xjGS08XjHW3e8k1d/hvdIWv8Rb/2nvY37zYYDsvGgnjzUf8f/LbqNWR9UfOTBxsxKmti85tyyrKS5TjcC+ABcn9jKuD6x1XGNYivj+sRWxvWJrYzr89aNWWv7bufEjxXwbBXGmJPW2mOdbgdwM1yf2Mq4PrHVcY1iK+P6xFbG9YmtjOvz7nA++hAAAAAAAABsZQQ8AAAAAAAAXa5bA57f7nQDgA/B9YmtjOsTWx3XKLYyrk9sZVyf2Mq4Pu+CrhyDBwAAAAAAAGu6tYIHAAAAAAAAHgIeAAAAAACALteVAY8x5h8YY6wxJuttG2PMrxtjzhtj3jDGPNTpNmLnMcb8n8aYM941+B+MMel19/2id32+a4z54U62EzuXMeZp7xo8b4z5h51uD3Y2Y8yIMeabxph3jDFvG2P+nre/xxjzdWPMOW+Z6XRbsXMZY3zGmNeMMX/ube82xpzwnkf/rTEm2Ok2YmcyxqSNMX/svfc8bYz5NM+f2EqMMX/fe31/yxjzB8aYMM+hm6/rAh5jzIikvyZpct3uH5G037v9vH0higgAACAASURBVKTf7EDTgK9LOmqtvU/SWUm/KEnGmCOSflrSPZKelvQbxhhfx1qJHcm75v653OfLI5J+xrs2gU5pSPoH1tojkj4l6e961+Q/lPSctXa/pOe8baBT/p6k0+u2/w9Jv2qt3SdpUdJ/0ZFWAdKvSfr/rLWHJN0v9zrl+RNbgjFmWNJ/LemYtfaoJJ/cz0M8h26yrgt4JP2qpP9e0vrRob8o6V9b18uS0saYwY60DjuWtfZr1tqGt/mypF3e+hcl/aG1tmqtvSjpvKRHO9FG7GiPSjpvrb1gra1J+kO51ybQEdbaKWvtq956Xu6Hk2G51+VXvMO+IulvdKaF2OmMMbsk/Zikf+FtG0lPSvpj7xCuT3SEMSYl6XFJ/1KSrLU1a+2SeP7E1uKXFDHG+CVFJU2J59BN11UBjzHmi5KuWmtfv+GuYUmX121f8fYBnfJzkr7qrXN9YivgOsSWZYwZl/SgpBOS+q21U95d1yX1d6hZwD+T+6Viy9vulbS07sscnkfRKbslzUr6V14Xwn9hjImJ509sEdbaq5L+idxeN1OSliWdEs+hm87f6QbcyBjzDUkDN7nrlyX9ktzuWUBHfNj1aa39U++YX5bb9eD372bbAKAbGWPikv69pF+w1q64RRIua601xtgPPBnYJMaYL0iasdaeMsb8YKfbA9zAL+khSf+VtfaEMebXdEN3LJ4/0Une+E9flBtGLkn6d3KHqsAm23IBj7X2qZvtN8bcK/cCed1787dL0qvGmEclXZU0su7wXd4+4I76oOtzlTHmZyV9QdLnrLWrL6pcn9gKuA6x5RhjAnLDnd+31v6Jt3vaGDNorZ3yulvPdK6F2ME+I+mvG2N+VFJYUlLumCdpY4zf+waa51F0yhVJV6y1J7ztP5Yb8PD8ia3iKUkXrbWzkmSM+RO5z6s8h26yrumiZa1901qbs9aOW2vH5T6xPWStvS7pzyT9595sWp+StLyuPBG4K4wxT8st5f7r1trSurv+TNJPG2NCxpjdcgcD/14n2ogd7RVJ+73ZC4JyB7r7sw63CTuYN57Jv5R02lr7K+vu+jNJf8tb/1uS/vRutw2w1v6itXaX957zpyU9b639zyR9U9J/4h3G9YmO8D7/XDbGHPR2fU7SO+L5E1vHpKRPGWOi3uv96jXKc+gmM2tFBt3FGHNJ7qjcc95F83/LLfsqSfqytfZkJ9uHnccYc15SSNK8t+tla+3f8e77Zbnj8jTkdkP46s1/CrB5vG+i/5ncmQx+11r7v3a4SdjBjDGflfRXkt7U2hgnvyR3HJ4/kjQqaULSf2qtXehIIwFJXhet/9Za+wVjzB65g9T3SHpN0pestdVOtg87kzHmAbkDgAclXZD0Zblf3vP8iS3BGPM/S/opuZ9/XpP0t+WOucNz6Cbq2oAHAAAAAAAArq7pogUAAAAAAICbI+ABAAAAAADocgQ8AAAAAAAAXY6ABwAAAAAAoMsR8AAAAAAAAHQ5Ah4AAAAAAIAuR8ADAAAAAADQ5Qh4AAAAAAAAuhwBDwAAAAAAQJcj4AEAAAAAAOhyBDwAAAAAAABdjoAHAAAAAACgyxHwAAAAAAAAdDkCHgAAAAAAgC5HwAMAAAAAANDlCHgAAMC2Yoz5H4wxf3zDvl8zxvx6p9oEAACw2Yy1ttNtAAAAuGOMMWOSTkvqt9bmjTE+SVck/YS19uXOtg4AAGBzUMEDAAC2FWvthKRXJf2Et+tJSSXCHQAAsJ0R8AAAgO3oWUk/460/420DAABsW3TRAgAA244xpk/SpKT9kt6S9Glr7enOtgoAAGDzEPAAAIBtyRjzVUl+SVlr7YOdbg8AAMBmoosWAADYrp6V9JTongUAAHYAKngAAAAAAAC6HBU8AAAAAAAAXY6ABwAAAAAAoMsR8AAAAAAAAHQ5Ah4AAAAAAIAu5/84B2ezWTs+Pr5JTQEAAAAAANi5Tp06NWet7budcz9WwDM+Pq6TJ0/ezuMAAAAAAADgQxhjJm73XLpoAQAAAAAAdDkCHgAAAAAAgC5HwAMAAAAAANDlPtYYPAAAAMB6z56YvOuP+czx0bv+mAAAbHVU8AAAAAAAAHQ5Ah4AAAAAAIAuR8ADAAAAAADQ5Qh4AAAAAAAAuhwBDwAAAAAAQJcj4AEAAAAAAOhyBDwAAAAAAABdjoAHAAAAAACgyxHwAAAAAAAAdDkCHgAAAAAAgC5HwAMAAAAAANDlCHgAAAAAAAC6HAEPAAAAAABAlyPgAQAAAAAA6HIEPAAAAAAAAF2OgAcAAAAAAKDLEfAAAAAAAAB0OQIeAAAAAACALkfAAwAAAAAA0OUIeAAAAAAAALocAQ8AAAAAAECXI+ABAAAAAADocgQ8AAAAAAAAXY6ABwAAAAAAoMsR8AAAAAAAAHQ5Ah4AAAAAAIAu5+90AwAAAIBu8OyJybv+mM8cH73rjwkA6E5U8AAAAAAAAHQ5Ah4AAAAAAIAuR8ADAAAAAADQ5Qh4AAAAAAAAuhwBDwAAAAAAQJcj4AEAAAAAAOhyBDwAAAAAAABdjoAHAAAAAACgyxHwAAAAAAAAdDkCHgAAAAAAgC5HwAMAAAAAANDlCHgAAAAAAAC6nL/TDQAAAABwc8+emOzI4z5zfLQjjwsAuH1U8AAAAAAAAHQ5Ah4AAAAAAIAuR8ADAAAAAADQ5Qh4AAAAAAAAuhwBDwAAAAAAQJdjFi0AAADcddZa1ZtWtWZLtUZL1UZTzZZVy0qtllXLeuvWW29JxkiOMfr2uTn5fUZ+x8jvc7ylUTTgVzzsVzzkV9DP95gAgJ2FgAcAAAC3zVqrUq2plUpdxWpTpVpDpVpTRW9ZqnrLWlPVRtMLc9xQx97mY37lu5c+8pig31EitBb4xEN+ZaJBZRNB9cXDyiaCysZDysZD6ouHlE0EFQ3y1hgA0L14FQMAAMBNNZotTeerurpY1pXFkq6vVDSzUtX0SsW7VXV9paJm6+ZRTTjgKBr0Kxr0KRbyqTceVNDnKOR3FPQ7Cvp9CvodhXyOAn5HAcfIcYwcY+R41TqOkRzHyBgjWalprZ48lFOj1VKjadVotVRvWjWaVuV6U4VKXYVqQ/lqQ4VKQwVvma80dH62oJcvVrVUqt+0valIQLsyEY1kotqViXi3qEZ63G0AALYyAh4AAIAdylqr2XxVl+ZLujRf1JWFkq4slXV1sayrS2VNLb8/vEmE/epPhtWfDOn47h7NFWpKRvxKhAOKhXyKeYFOJOiT39mcblKP7u75ROfXGi0tFGuaK1Q1m69q1ltOLZd1ZbGs87MFfevsjCr11obzEiG/+hIh9SVCyiVC6kuElUuElAj73QAKAIAOIuABAADYxlZDnPMzBV2cL2rSC3Mm5kuamC+pXG+2jzVGGkiGtSsT0bGxjHZlohrORDScjmg4E9FgKvy+bkzPnpi827/SJxb0OxpIhTWQCn/gMdZazRVqurJY0uXFsi4vlPTc6RnN5iv6/uUlVRtr4U844KgvHtJAKqzBVERD6YgGkmHGAQIA3FUEPAAAANtAq2V1dams8zMFnZvJ6/xMoX1bqTTaxwV9jkZ7oxrvjeqxvVmNZ6Ma641prCeqoXSEUMJjjGlX6zw4mpEkZaJBSW74k680NJOvajZf0Uy+qpl8VW9dXdErlxbd8yX1JUIaSrvB2FDaDcrCAV+nfiUAwDZHwAMAANBF6s2WJuaLbpAzXdD5WTfEeW+2sKFLUTYe1N6+uH78/iHtz8W1L5fQ7r6YBpJh+Ry6E30SxhglIwElIwHty8Xb+621WirXNbVU1rXliq4tlXVhtqDvX15yz5Mb+oz2RDXa447t05cIyaF7FwDgDiDgAQBsKZ3o7vHM8dG7/pidtJP+xt38u5ZrTb03W9hQiXNuJq+J+ZIa68bFGU5HtDcX1/HdvdrfH9e+XFz7+uLKxIJ3pB1b0VbtFmaMUSYaVCYa1JGhVHt/odrQtSV3oOrJhZLevraikxNupU844GjEG8h5rDeqsZ4YVVQAgNtCwAMAANBBy+W6F+Dk1wU5BV1dKst6OY7PMRrriWpvLq4fvmfADXFyce3tiysW4u3cVhcP+XWgP6ED/QlJUstazRdqmlwo6fKCG/p888yMrCTHuKHd7mxMu7MxjfXG6NYFALglvCMAAADYZNZazRaqOr+uS9Vq96rZfLV9XNDvaE82pgdG0vrJh0faQc54NqqQnw/524Wzbnyfh8fc8X2q9aYmFkq6OFfUxbmiXjo/rxfPzclIGvICnz1e6BMi8AEA3AQBDwAAwB3SslaXF0obulTdbKDjeMivvbm4njjQ1+5StS8X10hPlPFxdqhQwLehyqfWaGnSC3wuzRf18oV5ffv8nBwjjfRE29fMrgzXDADARcADAADwMdUaLc0VqpotVDWX37isN9fGx+mNBbU35w50vFqNsz+XUH8yJMPAuvgQQb/TvmYkd3DtSS88fG+2oOfPzOi5MzPtqq+9fXHtz8XVl+DaAoCdioAHAADgJqy1Wqk0NJt/f5CzVK63jzOS0tGAsvGQdmdj+rH71sKcnm080DHuroDP0d4+d9wlSSrVGrowW2wPxH3mel6SlIkGdKA/oYP9Ce3pizNgMwDsIAQ8AABgRyvXmpovVrVQrG0Mcgo11Rpr044H/Y764iGNZ2PKxoPqS4SVjQeVjYcU8K19iN5ps7KhM6JBv44Op3R02J2ta7FY09mZvM5ez+u1ySWduLggn2O0OxtrBz7ZeJDqHgDYxgh4AADAttayVvlKww1xCjUtFGuaL7rLhWJN5Xpzw/HpaEB98ZDGxmLqi7sD4WbjISXDfj4cY8vKxII6vrtXx3f3qtFs6dJ8SWen83p3Oq+/fHNKf/nmlDLRgA4OuOP87MlS3QMA2w0BDwAA6HrlWlNXl8q6vFDSxHxREwslTc6X9MbVZS0Wa2q01sbFcYyUjgbVEwvq3kxKvTF3vScWVG8sxIdedD2/b238nh+9d1CLxZrenc7r7HRepyYW9fKFBflvqO7ppboHALoeAQ8AANjSWi2ruUJVV5fKurZU0bWlsrde1rVld99CsbbhnEjAp7HeqPriIR3sT3jhjRvipKNBZh3CjpKJBfWpPb361J5e1ZstXZov6uz1vN6dLugv3pzSX7w5pZ5YUAf7Ezo0kNDubKzTTQYA3AYCHgAA0DHWWi2W6prJVzS9UtWUF9xcXRfkTC2XN8xMJUmxoE/DmYiG0hHdtyut4XREQ+mwRjJRjXrBjjFGz56Y7NBvBmxNAZ+j/bmE9ucS+jFJC6vVPdfzeuXSgr57YV5Bn6Nvn5/T5w7n9EMHc8olw51uNgDgFhDwAACAO85aq3KtqZVqQ/lyXf/u5GXN5KuaWXGDnOl8RTMrVc3mq6o1WxvOdYzUnwxrOB3RAyNp/ei9gxpOhzWUjrRvjIcD3Bk9saA+vadXn97Tq1qjpQtz7oxcb11d1tfemZYk3Tuc0pOHcnryUE73DqfkUAEHAFsSAQ8AALglLWtVqTVVqDZUqDVUrLrrxWqjvcxXGspX6lqpNNRs2ff9jGTYr/5kWLlkSMd39yiXDCuXCLX3DabC6k+GN8xKBeDuCPodHRpI6tBAUj/z6IjOXM/r+TMzev7MjH79+XP6tefOKRsP6clDfXryUE6f3d+neIiPEwCwVfCMDADADmWtVa3ZumlQU/BupXX3FWsN3SSzkZEUCfoUD/mVCPs11htTMuxXIhxQIuxXMhzQlz41plwypHDAd9d/TwAfnzFGhweTOjyY1N/9oX1aKNb0wtkZPXd6Rl9967r+6OQVBXxGx3f36slDOX3ucE5jvYzdAwCdZKy9yTu1D3Ds2DF78uTJTWwOgG7UqTEunjk+2pHHxebaSWOmbMY1XGu0tFCsaa5Q1XyxpvlCVfOFmuaK7nK+UNW5mUI7tLlxbJtVIb+jWMiveMjvLX03bK8to0GfnI/oLtWp/6876XoC7qQP+z9bb7Z0amKxXd1zfqYgSdrTF9PnDuX05KF+HRvPUIkHALfBGHPKWnvsds6lggcAgC3MWqvlcl1zhapm8zXNrwtq5oo1LRTW9s0VqlqpNG76c4I+R73xoHrjQUWDPvXFQzcNa2JekMMHMwAfJOBz2rNy/dKPHtbEfLEd9nzlOxP6nb+6qETYr8cP9Olzh3L6wYM59cSCnW42AGx7BDwAANxlLWtVqjV15vqK5vJuMDNXqGq2UG1vzxfd9fli9aZVNo5xB0d1p/8O6chQUtl4SL2xoHrjIfXGg8rG3ft640HFQ2uDElPVAuBOGuuN6cuf2a0vf2a3itWGvn1+Ts+fntHz787oL96YkjHSgyNp/eDBnB4/0Kd7h1PyMVAzANxxBDwAANwBLWvdcWsqa+PXbFhft12sNmQl/W9/ufFnBH2OG8rEQ+qLh3R4IKlsIqRsPKRsPKi+eEjZhBvipKNBPiAB2HJiIb9++J4B/fA9A2q1rN6+tqLnzkzr+TMz+tVvnNWvfP2sMtGAPrMvqycO9OnxA33qZxp2ALgjCHgAAPgI9WZL+UpDy+W6Vsp1d1lZv+7OHHWzAYgDPqO41/0pEw1opCfS3n766KCy8WA7xGHqbwDbieMY3bsrpXt3pfQLTx3QfKGqb5+f0wtnZ/VX5+b0529MSZIODST0+IE+Pb6/T8fGMwzGDgC3iYAHALBjWWtVqbe07IU1K+X6uvWGVipugFOqNd93btDvKBUOKBnxa082plQkoGQk0J5JajXECfqdDwxtfuy+wc3+FQFgy+iNh/TFB4b1xQeGZa3V6am8Xjw3qxfPzur3Xrqk337xgsIBd3yfx/e71T17+2IE3wBwiwh4AADbUstaFVarbtrVNmuhzYq3/2bj28RCfqXCfqUiAY30RJUMB7wAx99e5xtmALh9xhgdGUrqyFBSf+eJvSrVGnr5wrxeeHdWL56b0//y7juSpOF0RI8fyOrx/X16bF9WqUigwy0HgK2LgAcA0HXqzdb7qm02VOGU6ypUG+/rMuUzRomIX6lwQEPpiA5HkkqG/UpGvAAnHFAi7JefGaQA4K6KBv168lC/njzUL0m6vFDSC2fd6p7/9/Up/cH3Lssx0tHhlD69p1ef3turR8Z7FAvxcQYAVvGMCADYMqy1KtVWx7p5f7XNStm9r1x/f5epkN9xg5pwQPty4Q3VNqsBTjTok0OpPwBseSM9UX3pU2P60qfGVG+29Nrkkr7z3py+8968fveli/qtFy/I7xjdP5LWY3vdwOehUcbvAbCzEfAAAO6KRrOl2UJV15cr7m3Fu3nb0ysVTS1XVG20Npxn5HWZigSUiQY01httV9skvW5TqXBAId7UA8C2FPA5enR3jx7d3aNfeEoq15o6NbHYDnx+41vv6f96/ryCPkf37kq5x4736KGxDF26AOwoBDwAgE/EetODz+Srml5xg5rry1VdXy6vBTgrFc3mq+/rMhX0OepPhTSQDOveXWl9/khIV5cqXoDjdp1KhP3yO3SZAgC4IkGfPrs/q8/uz0qS8pW6Xrm0oBMXF/TKxQX9zosX9Jvfek/GSIcGknp0PKNHdvfokfEepmQHsK0R8AAAPlCh2miHNrNegDOzUtW0t76672azTCXCfg0kwxpIhXWgP6HBVFj9qXB730AyrJ5Y8H2zozx7YvJu/XoAgG0gEQ5sGL+nXGvqtcuL+t7FBb1yaUF/dPKKvvLdCUnSUCqsB8cyenAkrYfGMrpnKKmQnwpQANsDAQ8A7CDuGDdNzRdqmitWNZevar5Y03yhqrlCTfPFmmZWKu1qnJsFN5GAT/3JkHKJsO4ZSurJQznlEiH1J8Pu0gtvGPgSANAJkaBPj+3N6rG9boVPvdnSW1eX9erkkl6dXNT3J5f0F29MSXIrSe8ZTuqh0YzuH0nrvuGUxnqjTM0OoCsZa98/PewHOXbsmD158uQmNgdAN9pJFRfPHB/tdBM2sNYqX21ouVTXUqmuxVJNC8Wa5go3BDftAKeqSr1105+VCPvVGwsqlwgrl1wX2HjLXDKs/mRI8ZB/U9/47qTrCQC2qq32enenTa9U9Nrkol6dXNJ/fPu6ri6W1fD6EYcDjobSEe1KR9xlJqpMNHDHX/u2+98YwO0xxpyy1h67nXP5ehUAtoBqo6l8paFCpaGlcl1LpZqWy25os1Sqa6lcc0OcshvirK4vl+tq3jiwjSfgM+qNhdQbD6o3HtLevriyiZB6Y+52bzyorHd/TyzIzCMAgB2jPxnW00cH9fTRQY33xtRotTSzUtXVpbKuLpZ1damsl96bb7/GRgI+DabCGkx53YxTEeUSIQV8jBEHYOsg4AGA22CtVaXeUqnWUKnWVKXeVKnm3orVhvLVugqVhvJVN7TJVxoqVN1lvlJXobq2Xag0VGvevKpmVSLkVyoaUDoaUCYa1HA6onQ0oHQkqHTUnQI8HXXXe2JBZeMhJcObW2kDAMB24Xfcqp2hdESPjLv7Gq2Wpleq7cBnarms711aUL3phj6OkbLxkAZSYQ0m3XHmcomw0tGAHF5/AXQAAQ+AballrRpNq1qzpXqj5S6bLdXa63Zt/cald1x99fx15/3Tr72rUq2pcv39Y9N8EL9jlAj7lQgHFA/5FQ/7NZgKt9fjoYB3v1/xkN8LbIJegONOBc43hAAA3F1+x9FwOqLhdKS9r2WtFgo1Ta1UNLVc1vXliibnS3rjynL7mIDPqC/udm3OJULKJULqS7gTC/gcgh8Am4eAB0DHNFot1RtW1UazHaCsX9abLdWa9gMDmPcFNevOW/127eMI+IwCPkdBn6OA31v6HEWDPqX+f/buPEiS7KDz/O95eNxHHhGZWVWZVZVZVd1dfV+l7pa6VyMk0AhGK8EsA6LBhmsG213MgDEwDJi1HYNd/hhbbGDYGbDFgIFhkYARAjQCsWiEGAlBt+iqvrv6qDPryvuK+/K3f7hHZGTW1VVdmZGR8f2YQu7+3CPzZbuXR8Qv3hEPtwdijIdDikdcJSKhYD20YT0dDQfBjR/aRF2HljQAAOwCjjHKpaPKpaN6cHygXV6uNTWX9ycpaE1WcG6hqJcurHQ8VxpK+K1ss6mIGp6nyWxSU7mk9g3GCX8AvGcEPABuylqrasNrdzMqVv1uScVaQ6VqU8fPL/khS8NTtRW+dAQ11U3brUfzFgZ5l64fwCSjIQ2F/FYuEXfTftdRpPU812k/v70eHOeGzLtqTs2AiAAAYLN4JKSD2aQOZpMbyqv1puYLVc2tVbVQqGohmADh7EJRf3d6sX1cJORoYiiuieGEJobi2j+U0P7h1nJrBnkGsPsQ8AC7XKPpdYz9sj7+Sz4YHyZfqW8aI6bePra1Xag23nWLGCP5oUkQnESD9VYIE3FDirh+ebhjf8R1FO0IZm43gAEAANgpouGQJoYSmhhKbCi31uqb7xvT2YWizi0UdXaxqAtLJV1YKuuViytaKdU3HJ+MhLRvMK69g3HtCwZ63jcQ197BWDD4c1zJKB/tgH7HXQDYoTzPqlRvtgOYtc7ApXKTgKZj8N53M1ZMKBgjxu9SFFY6GowR0x4XZuMYMcmoq2TEVSIaUjLi6ktvzLZDG9cxfMMEAABwA8YYjWViGsvE9NSh7FX785W6Li6XdWGp5C+XS7q8UtaV1YpOXlnTfL561XNSUVcj6ahyqUiwjGok5Xcnay+D/VGXmTOB3YiAB7jDWsFMIQhcCsEsSoVqvR3CFNrhzMYWNYXKekBTqDb0bnowtcZ5aS0HEhFNDCeUjm4MZ1IxNygLt4Ob1nYs/N7GiDl+fvm2nwsAAICN0rGw7t0b1r17M9fcX2t4ml2r6MqqP9jz5ZWK5vIVzef9rmBvzeT19cKiVsv1az4/E3M3Bj+pqIaTG2fmHIj7kz0MJsJKx8KMEQT0AAIeQH4oU2n4U1yXa80NY83kOwKazmmv2+FN53aloULt3QUzUdfxW8t0tIzJ5RIbWstsDmgym7ZTEVcOL7YAAAB9JeI62j/sj89zI9VGU4uFWjv42bj0y09eXtNX81Xlq43r/hxjpHTU1WBiPQQaCMKfwWDmz3TMb+Wdiq63+O5cj7jMCApsNQIe9ARrbXvK6krdD2FKtaZKtcb6er2pUjD4b7nu72sFNsVaU+Vgu1VWqvvPLVbf/ZTXxkipSPCC1dFqZk9mfcrrdPTq/alouL3eerHjRQ4AAABbKer6Y/fs65jq/XpqDU9rlbpWSnWtlmvB0t9eKde1Wqr520HZpeVysF6T9y6+3Iy4ThD2+F38W++Lk1H//XMi4ioecRQPhxQLZiaNh/1HrGO9Vd46JuY6ckO8rwYkAh4ErLVqelYNz6re9NRoWtU9zy9rBmXBvqbnhy2NoKzh+ev1plXD85/rz5zUVDWYQana2q6vz6i0YX+9qVrTC/Y3O45Z377FCZfkGCkZcdensA6mtU7HXI1losGLSEiJ8Mb98UhImY4gph3QxFwlwiFazAAAAGDXibiOcil/7J5b4XlWhZo/3ECxur5stYS/er3ZPmaxUNP0Ykn5akOV4Avb5rtJizbXvWPijvYkHR0Td3RO8NFZtnlSkM4JPjqf15q5NRQych2jkOPP0Bpybm+b8SqxVfou4Lm4XNL/+9y0JMnKKvifrLWy7fVgX7Cu1v6Ofe3y9jHB8zv2t25N7Z/X+bs6f3ZHmYJjPU9qWtsOXjwredbKa217/nbTBvu89X3WKii3QbmCn9E6Rmp66yFOo+mHNFst2r5JhhR1HUXDwQ017G+noq6yyWBfsD/aMeNS1F3fboU2/sNtr8cjfggTj/g/h5snAAAAsHUcxygTCysTC9+Rn1dveirXm6oErfLLQev9cr0ZtOT32uWVdsv9pqqNpmrBl8T1pqdac/0L41rDn1V2w/6Gf0xrf615618o3y7HSG7IaYc//tKRYyTHGDnGH4jbcaSQ8WeSNe19Ro5jrjo25HQ872bHBj/btI4J9kv+fH0pmgAAIABJREFUjLgKPkIZ+b+3VWQ2la9/1LrJca2jOj6b+eWmY73jecao40dvOO56xgfj+r6nDt7KadiVjL2Fq9gYMy/p/NZVB5vkJC10uxLoGs5/f+P89zfOP7gG+hvnv79x/vsb57+/5SQlrbUjt/PkWwp4sL2MMS9Ya491ux7oDs5/f+P89zfOP7gG+hvnv79x/vsb57+/vdfzz2hUAAAAAAAAPY6ABwAAAAAAoMcR8Oxsv9HtCqCrOP/9jfPf3zj/4Brob5z//sb572+c//72ns4/Y/AAAAAAAAD0OFrwAAAAAAAA9DgCHgAAAAAAgB5HwLMDGWM+Zox5yxhzyhjzM92uD7aGMea3jTFzxpjXOsqGjTFfMsa8EyyHgnJjjPnV4Jp4xRjzWPdqjjvBGLPfGPMVY8wbxpjXjTE/HpRzDfQBY0zMGPMNY8zLwfn/+aB8yhjzfHCe/9AYEwnKo8H2qWD/ZDfrjzvDGBMyxrxojPlCsM357xPGmHPGmFeNMS8ZY14Iyrj/9wljzKAx5rPGmDeNMSeNMe/n/PcHY8w9wb/71mPNGPMTnP/+YYz5V8F7v9eMMZ8J3hPesdd/Ap4dxhgTkvQfJX2rpPskfY8x5r7u1gpb5HckfWxT2c9I+rK19i5JXw62Jf96uCt4/IikX9+mOmLrNCT9pLX2PklPSfrR4N8610B/qEr6sLX2YUmPSPqYMeYpSf9W0i9ba49IWpb0w8HxPyxpOSj/5eA49L4fl3SyY5vz31++yVr7iLX2WLDN/b9//HtJf2mtPSrpYfn3Ac5/H7DWvhX8u39E0uOSSpL+RJz/vmCMGZf0Y5KOWWsfkBSS9Cndwdd/Ap6d5wlJp6y1Z6y1NUl/IOmTXa4TtoC19quSljYVf1LS7wbrvyvp2zvK/7P1PSdp0Bizd3tqiq1grb1irT0RrOflv7kbF9dAXwjOYyHYDAcPK+nDkj4blG8+/63r4rOSPmKMMdtUXWwBY8yEpH8i6TeDbSPOf7/j/t8HjDEDkj4o6bckyVpbs9auiPPfjz4i6bS19rw4//3ElRQ3xriSEpKu6A6+/hPw7Dzjki50bF8MytAfxqy1V4L1GUljwTrXxS4WNLd8VNLz4hroG0H3nJckzUn6kqTTklastY3gkM5z3D7/wf5VSdntrTHusF+R9NOSvGA7K85/P7GS/soYc9wY8yNBGff//jAlaV7Sfwq6aP6mMSYpzn8/+pSkzwTrnP8+YK29JOmXJE3LD3ZWJR3XHXz9J+ABdihrrZX/BhC7mDEmJemPJf2EtXatcx/XwO5mrW0GTbQn5LfePNrlKmGbGGM+LmnOWnu823VB1zxjrX1MfveLHzXGfLBzJ/f/Xc2V9JikX7fWPiqpqPXuOJI4//0gGGPlE5L+y+Z9nP/dKxhb6ZPyg959kpK6esiO94SAZ+e5JGl/x/ZEUIb+MNtqdhks54JyrotdyBgTlh/u/L619nNBMddAnwma5n9F0vvlN712g12d57h9/oP9A5IWt7mquHOelvQJY8w5+V2xPyx/TA7Of58IvsWVtXZO/vgbT4j7f7+4KOmitfb5YPuz8gMfzn9/+VZJJ6y1s8E2578/fLOks9baeWttXdLn5L8nuGOv/wQ8O88/SLorGEk7Ir/p3ue7XCdsn89L+v5g/fsl/VlH+T8PRtJ/StJqRzNO9KCg/+xvSTpprf13Hbu4BvqAMWbEGDMYrMclfYv8cZi+Iuk7g8M2n//WdfGdkv46+IYPPcha+7PW2glr7aT81/m/ttZ+rzj/fcEYkzTGpFvrkj4q6TVx/+8L1toZSReMMfcERR+R9IY4//3me7TePUvi/PeLaUlPGWMSwWeB1r//O/b6b3h/sPMYY75Nft/8kKTfttb+YperhC1gjPmMpA9JykmalfRvJP2ppD+SdEDSeUnfZa1dCm4A/0F+E76SpB+01r7QjXrjzjDGPCPpa5Je1foYHD8nfxweroFdzhjzkPxB80Lyv2z5I2vtLxhjDslv0TEs6UVJ32etrRpjYpJ+T/5YTUuSPmWtPdOd2uNOMsZ8SNJPWWs/zvnvD8F5/pNg05X0aWvtLxpjsuL+3xeMMY/IH2A9IumMpB9U8Fogzv+uFwS705IOWWtXgzL+/fcJY8zPS/pu+TPqvijpX8gfa+eOvP4T8AAAAAAAAPQ4umgBAAAAAAD0OAIeAAAAAACAHkfAAwAAAAAA0OMIeAAAAAAAAHocAQ8AAAAAAECPI+ABAAAAAADocQQ8AAAAAAAAPY6ABwAAAAAAoMcR8AAAAAAAAPQ4Ah4AAAAAAIAeR8ADAAAAAADQ4wh4AAAAAAAAehwBDwAAAAAAQI8j4AEAAAAAAOhxBDwAAAAAAAA9joAHAADsKsaYc8aYnzLGvGKMWTXG/KExJtbtegEAAGwlAh4AALAbfZekj0makvSQpB/oam0AAAC2mNvtCgAAAGyBX7XWXpYkY8x/lfRIl+sDAACwpWjBAwAAdqOZjvWSpFS3KgIAALAdCHgAAAAAAAB6HAEPAAAAAABAjyPgAQAAAAAA6HHGWtvtOgAAAAAAAOA9oAUPAAAAAABAjyPgAQAAAAAA6HEEPAAAAAAAAD2OgAcAAAAAAKDHubdycC6Xs5OTk1tUFQAAAAAAgP51/PjxBWvtyO0895YCnsnJSb3wwgu383sAAAAAAABwA8aY87f7XLpoAQAAAAAA9DgCHgAAAAAAgB5HwAMAAAAAANDjbmkMHgDAzvHp56e3/Xc+++SBbf+dAAAAAG6OFjwAAAAAAAA9joAHAAAAAACgxxHwAAAAAAAA9DgCHgAAAAAAgB5HwAMAAAAAANDjCHgAAAAAAAB6HAEPAAAAAABAjyPgAQAAAAAA6HEEPAAAAAAAAD2OgAcAAAAAAKDHEfAAAAAAAAD0OAIeAAAAAACAHkfAAwAAAAAA0OMIeAAAAAAAAHocAQ8AAAAAAECPI+ABAAAAAADocQQ8AAAAAAAAPc7tdgUAAAC2yqefn9723/nskwe2/Xdid+B6BQC8F7TgAQAAAAAA6HEEPAAAAAAAAD2OgAcAAAAAAKDHEfAAAAAAAAD0OAIeAAAAAACAHkfAAwAAAAAA0OMIeAAAAAAAAHocAQ8AAAAAAECPI+ABAAAAAADocQQ8AAAAAAAAPY6ABwAAAAAAoMcR8AAAAAAAAPQ4Ah4AAAAAAIAeR8ADAAAAAADQ4wh4AAAAAAAAehwBDwAAAAAAQI8j4AEAAAAAAOhxBDwAAAAAAAA9zu12BQBgt/j089PdrsKuxH/XrfHskwe6XQX0sG78u+SaBQDgxmjBAwAAAAAA0OMIeAAAAAAAAHocAQ8AAAAAAECPI+ABAAAAAADocQQ8AAAAAAAAPY5ZtAAAPcWz1n94/nrTs2paK8+z8uzGY03HiuksN2bjMWZ9PWSMQs76o/NYAAAAYKci4AEAbKlyranlUk3LpZpWS3WtVRoqVBsqVOoqVBvKVxsqVBoqVv3yUq2pSr2pSt1TpdHUUqGmetNT3bNqNL2rQpyttjnw6Xy419h2HUfhkFE45CjsOgo7jsKuUSTk+GWtfdfcDo5zHbmESwAAALgFBDwAgFvieVZLpZrm81XN5auaW6tovlDV3FpV84Wqlgp+mLNSqmu5VFO14d3w50VcR6moq1TUVTLqKhEJKR4JaSgRUSwc0pXVstyQo7DjByEhx8hxjELGXzpGfsBi/EDEGElWWs+BrGxHKGTb/+cvbOeRtqNVkGfV8NbXNzxs5z6vXV5qeGo0G34g1fRUa3pqNP1jb8fm8Ke1HnEdRTYt18tNezvq+mFRJLTx+LDryFpLgAQAALCLEPAAANqstarUPa2U/dY2y+W6Vks1rZTrWinV9X//9Tuaz1evGViko65y6aiGkxFNDMX1wPiAhhJhDSYiGkpE2uuZuKt0NKxUzFUyGlLUDd2wTp9+fnqr/txt41kbhD7BsuG3SKo3vHYY1NrXCoVqzav31Rv+/mq9qUKloVrTUy0oqzc83UqM9AtfeEOJsB+m+aGaH64lIiHFw+tlyU37W8f7x/nnsL0/+HlR1yE8AgAA2GYEPADQZ6y1ylcbWizUtFSsarFQ02KxpsVgfXOLm5BjNBgPayAR1vsmc9ozENVoOqaRdFSj6ahGgkciwkvK9TjGKOqGFN3C/0TWWtVbwVAQ+tQ2LTvL7x5Lq1RrqlRrBMumynW/m9x8vtouK9UaKtebG1pB3fzvlRIRtyMMunZ4lNi0f3N41N4fJTwCAAC4Gd6NA8Au1fA8LRZqfjeqfEXz+arm81UtFKqqN9c/rTtGGkpElE1FdGA4oaFERIOJiAbjYQ0mwkpGXTnBB+pnnzzQrT8HN2GMUcT1u2cpevPjb+Vctlp2tcKgcj0If4Ixk0r1psqdQVFHOHT98Gj9Z72X8KgVGCWjruLhkN/dL+a2l29eySvqOoqF/XAo2loGZYx1BAAAdgsCHgDocZ61WirWdGW1opnVcjAuTlWLxeqGAYmHEmGNpKM6lEtqOBVVNhlRNumHOSGHD7i4PmOM4kELm+wd/tm3Ex4Va42OEGk9PJpbq6pY89fzlYaa72LsI8dIUTekWNgJlv7fGQ+HFA87wd/tBtsd+4Il/3YAAMBOQcADAD2k2mhqZrUShDkVXVkta3atqlrT71ZlJGVTEY2mY7p/X8bvRpWJaSQV9Vt2ADvMVoVH1lpVG55+9+/Oqdrw/Ee9qWrDUyVYVutNVRqeqo2mqnVPlWDfcrGmy3U/TGr927qeiOtsCH8SkZBev7yqbBCiDgdB6nAqomwyqqFEWG6If4sAAODOI+ABgB2q1vB0ZbWsSytlXVou6+JKWQv5ansg3VjY0Z5MXI9PDmlvJqa9A3GNZqIK8+ERkDFGsXBI6VhY6ffwcxqep0rdUzloXVQOxiRa3964nMtX9RevXtFKuX7NrmfGSAPxsIaTEeWS/qDkfvjjP0bSMY1l/HGuRjNRxcI3HoQcAACghYAHAHaApmc1u1bR9FKpHejM5SvtLlbpqKvxobgeGh/QvsG49gzENBgPM3YIsMVcx1Eq6ih1CyNkP/vkATU9q+VSTUvFmhYKVS0V/XV/UPNqe/30fEHfOFfTcql2zUBoIB7WWCaqsYw/sPlYJqaxYDnaEQTdbDY6AACw+xHwAEAXFKoNXVgqaTp4XFout7uCJCIhTQzFde/eUU0MxTU+GFcmHu5yjQHcipBjlEtFlUtFdffYzdsQtQKh+XxVc/mqZtcqmluraHbNHyR9dq2q03MFzeWralxjbKGhRFhjmZj2DMS0b9C/b+wbjGnfQFzjQ3GNZWK07gMAYJcj4AGALWat1UKhprMLRZ1bLGp6qaSlYk2SP8DrvsG4Hj84pAPDCe0fTmgoQcscoN90BkL37r3+cV4QBM2uVTWb90OguWB9ZrWqmbWyXrm42r7HtDhGGsv44c++IPwZH4xr30C8HQhl4i73HgAAehgBDwDcYZ71u1udXSjq3EJRZxdLKlYbkqRU1NXBbEJPTg1r/1BC40NxvlUH8K45jvEHcE5FdZ8y1z2uXGvq8mpZl1f8x6WVii4t++uvXFzR//da5aoBpJORkCaG/KDZD5zjOhCsTwwlFI/QDQwAgJ2MgAfAlvv089Pb/jufffLAtv2uRtPT65fX9LV35tutdCp1/4PTYDysu0dTmswlNZVNKpuK8A05+lY37gXdsN1/57Xud/FISIdHUjo8krrmczzPaqFY1eWVSjsEurjsPy4slfR3pxdUqjU3PGckHfWDn6G4VssNDScjGkqGNZyIKBMPy+HeBgBAVxHwAMAtqjaaeuXiqr5xdknPnVnUifPLKgYfhHKpiB7YN6CpXFKTuaSGEpEu1xYAruY4xh+gOR3TI/sHr9pvrdVisabppZIuBA9/vax/OLesyytldY4EFHKMBuNhZYPp4LOpiHLBVPGDiYhCDuEPAABbjYAHAG6i0fT0yqVVff2dBX399IJenF5RteG30LlnLK3/6fEJPTE1rIvLZWViDIYMoPcZsz4m0GMHhq7a/3t/f14rpZqWSjUtF+v+LGGlmpYKVZ1bLKnWWO/+5RhpKBEEPq0p4YOfPRAPE/4AAHCHEPAAwCbWWp2eL+jrpxb1t6cW9NzpReWrDRkj3bc3o+976qCenBrW+yaHNZRcb6HTL91PACDUMRbQZtZaFaqN9pTwC4WaFgtVLRb9weY7x/4JGaOhZFjZZFS5lB/8tAKgwQTdvgAAuBUEPAAgaXatoq+fWtDfnlrQ108taHatKkk6MJzQxx/ep2eO5PT+w1kNJ+lyBQA3YoxROhZWOhbWZC65YZ+1VvlW+BOEPguFqhYLNZ1ZKKjeXO/4FXKMhhORdqufprU6lEvq0EhSezIxxjMDAGATAh4AfWmtUtfzZ5baoc6puYIkaTgZ0QcOZ/X0kZyePpzTgWyiyzUFgN3DGKNMLKxMLKypa4U/lYYWitV2ALRQqGmpWPNbVZ5ebB8bD4c0FYQ9h0ZSOjyS1KFcSlMjSaWivL0FAPQnXgEB9IVqo6kXp1f0d0Gg8/LFVTU9q1jY0RNTWX3XsQk9fSSne/dk5DAeBABsO2OMMvGwMvGwDuU27vOs1UfuHdXZ+aJOLxR1Zr6gM/NFvXxxRX/+6hXZjhGfR9PRdvBzKJfU4ZGUpnJJTQzF5Yac7f2jAADYRgQ8AHYlz7N6cybfbqHzjbNLKtebcoz08P5B/a8fOqynj+T06IFBRd1Qt6sLALgBxxjtHYhr70BcHziyMf2p1JuaXirpzHxBp+eLOjNf1NmFgv7i1StaKdXbx4VDRgezyaCbV6rd3evQSIrutwCAXYGAB8CusVys6dR8QafmCvqlv3pLS8WaJOnIaErf/b79+sDhrJ46nGWmKwDYRWLhkO4eS+vusfRV+5aKtXZrn9MLBZ2dL+rMQlFfeWtuw3g/g4mwDuWSmsqldGgk6Xf5GknpYDbBlwAAgJ5BwAOgZ5VqDZ2ZL7ZDnVagk465+pZ7x/xxdI7ktGcg1uWaAgC6YTgZ0XByWMcmhzeUN5qeLi6XdWbBD3/OBN2+vvbOvP74xMX2cY6RJoYSG8f7CVoAjWWiDPQMANhRCHgA9Ix609P0Ukmn5vxA5/JKWVZS1HU0lUvqA4ezOjyS0mg6qu996mC3qwsA2KHckKPJXFKTuaQ+fHTjvnylrrMLRZ1dKAZdvvwQqNXVtyURaQ303NHdi4GeAQBdxKsPgB3Ls1YzqxU/0Jkv6NxCUQ3PyjHS/uGEPnx0VEdGU5oYSijEwMgAgDsgHQvroYlBPTQxuKHc86xm8xW/xU9rvJ+Fol6cXtYXXrm8YaDnsUxUh4LuXp0BEK9XAICtRMADYEfpHEfn9HxBpZr/beloOqonp4Z1eDSlqWxS0TBjIgAAto/jrA/0/PQ1Bno+v+gP9HxmoajT8wWdXSjqC69c0Wp5faDnSMjRgWzCH+9nJKnDQYufqVxSWQZ6BgC8RwQ8ALqqVGvo9HxRp4NWOq1xdDIxV0f3pHV4JKXDoykGRgYA7FixcEj37Enrnj0bB3q21voDPbemdl8oXneg50zMVSYe1kgqqmwqqlwqopF0VNlkVBGX6d0BADdHwANgW9Wbns4vlnR6/upxdA4F4+gcGUlpJM3glQCA3maMUTYIbN53jYGeL62UO0Kfgv7+9KLf7evCyoZjB+Jh5VIR5VLR9mMkHdVgIiyH10oAQICAB8CWqtSbOhNMTXt2oajppVJ7HJ0Dwwl9+N5R3TWS0jjjEgAA+ogbcnQwm9TBbFLfdI9f9unnpyVJtYanxWJV8/mqFgo1LRSqWihU9fLFFVXqXvtnhByjbLIV/HQEQOmokpEQX5QAQJ8h4AFwR1XqTb04vaLnzy7quTOLOjG9olrDk5G0dzCmJ6eGdWQ0pclcUlGXcXQAANgs4jrt8X46WWtVrDW1kK+2Q5+FQk3zharemsmr2THScyzs+C19ghZEI2k/BKLLFwDsXgQ8AN6TSr2pE9PLev7Mkp47s6gXL/iBjmOk+/cN6Pvff1DVuqeD2aTiEQIdAABulzFGqairVNTVZC65YV/Ts1op1Ta0+FkoVG/a5atSb7YHfB4fitOaFgB6mLGdczrexLFjx+wLL7ywhdUBsNO1Ap3ngkDnpekV1Zp+oPPA+ICeOpTVk1PDOjY5rIG4PzByq8k5AAC95NknD2zr79uq18taw9vQ4qczAOrs8hUOGe0fTmgqm9RkLqnJbCJYJrVvsLfDn268F9nu6wfA7mCMOW6tPXY7z6UFD4Abmlmt6MT0so6f9x+vX15VvemPofPg+IB+8OlJPXnID3SY6QoAgJ0n4jraNxjXvsGru3z94wf26Gwwy9fZhZLOLRR1brGor59e2BD+REKO9g/HNRUEPgdzySAISmjfQFxOD4c/ALBbEPAAaKs3PZ28sqYT55d1fHpFJ84v69JKWZI/y9XDE4P6oWem9NRUVscmh5Qm0AEAoGcZY9oDM2+e5ctaq9m1qs4uFHV+saizi0U//Fko6WvvLKja6Ah/XEcHhhOazCY1lVtv9TOZS2pvJkb4AwDbhIAH6GNLxVoQ5izrxPnlDbNz7B2I6bGDQ/rhZ6b0+MEh3bs3w6CMAAD0CWOM9gzEtGcgpvcfzm7Y53lWs/lKEP74rX7OBi1/vvbO/IbwJ+o6OphN6GDW7/J1YDih/cP+cnwozoQLAHAHEfAAfaLe9PTWTF4vX1zRifMrenF6WWcWipIk1zG6f19G3/PEAT1+cEiPHRi6qhk3AACAJDmOac/y9YHDG/d5ntXMWiXo6lXSucUg/Fko6qtvbwx/jJH2ZmI6EAQ/neHPgeGEhpMRpnoHgFtAwAPsQk3P6vR8QS9fWNGrl1b18sVVnbyyplrwpmo4GdFjB4b0z47t1+MHh/Tg+AAzXAEAgPfMcUx7vJ8PHNm4z/OsFgpVTS+VNL1U0vnFki4E63/z1rzm8tUNxycjoQ2Bz4GsHwAdpPUPAFwTAQ/Q42oNT2/P5vXGlTW9cXlNr19e1euX11SqNSX5b44eGPenK39wYlAPjQ/oYDbBN2IAAGBbOY7RaCam0UxMxzaN+SNJ5VpTF5dL7QBoeskPgM4uFPXfr9P6pzMAmhiOa3zQD3/G0lG5IbqWA+gvBDxAD1kt1fXmzJreuLKm1y/7j1NzedWbVpKUiIR0dE9a33Vsvx6aGNBDEwOayqV6elpTAADQH+KRkO4aS+uusfRV+6y1ms9XN4Q/rQDoq+/Ma3ZtY+ufkGO0JxPT+FBcE4NxLZVqGopHNJgIazDhL8MEQAB2GQIeYAeq1Jt6Z7agt2bzemtmTW/NFvTWzNqGNy+5VET37RvQh+4Z0X17M7p/X0YHs0nCHAAAsOsYc+PWP5V6UxeXy7q0Utal5bIurZSCZVnPn13S5ZWy7KbnJKOuhhJhDcbXQ59MLKyBuP9IxVw5tHgG0EMIeIAuWqvUdXquoFNzBZ2eLwbLgs4vFuUF70IirqO7RlN6+nBO9+xJ6+49ad2/N6PRTKy7lQcAANghYuGQjoymdGQ0dc39v/f357VWqWulVNdKqablYLlSrmtmraI3Z/JqeBsjIMdI6VhYmZjbDn0y8fCG9XTMlevQEgjAzkDAA2yxWsPTheXS+mwSC+tBTudgguGQ0WQ2qaN70vrEw/t0z5607tmT1sHhBH3IAQAA3oOQYzSUiGgoEZGUvGq/tVbFWlNr5brWynWtVupaba2X65pZq+qt2fVu8S1GUirqKhP3g6BUzA990jFXI+moRtJRjaajyqWiiri8nwOwtQh4gDsgX6nrwlJZF5dLurBcDsIc/3FpuazOL4TSMVeHR1L64N0jOjzif9N0eCSpAwQ5AAAAXWGMUSrqKhV1tW8wfs1jrLWq1D2tVtaDn9ZytVzXcqmu6aWSisFEF3/20uUNzx9KhDWajmk0E9VIKqqRTNTfDoKgXCqqXCqiTCwshy73AG4DAQ9wE9ZaLZfqurxS1pXViq6slnVxuawLSyVdWC7p4nJZK6X6huekY66mckk9un9I3/HIuCZzSR3MJjWVS2ooEWYGKwAAgB5jjFE8ElI8EtKeG3SVb3pWhWpDTx/Jam6tqvlCVXNrVc3lK5rPVzWXr+rMfFHz+apqTe+q54cco+FkRNlkxF+mosq2tlMRZZNRZVORoCyqTNzlvSUASQQ86HOlWkPz+arm81UtFKrtF93LK36Q0wp0KvWNL75R19HEUFwTQwk9sn9Q+4cSmhhKaP+wX0aIAwAA0J9CjtFAPKyHJgZveJy1VqvluubyfgC0WKxqsVDTYrGqpWJNC4Waloo1vXpxRYvFmvKVxjV/jhsEQsPJiAbiYQ0mwhpKRDSQCGswHvEHkk6ENRCPaCjplw0mwoqFQ1vx5wPoIgIe7Dq1hqfFYrUd3HSGN/OFjeWtJrSdHCONpmPaOxjTfXsz+sjRUe0djGt8MKa9A3HtHYwpl4zSdBYAAAC3zRgTzN4V0d3XmBp+s2qjqeViXQsFPwDyQyB/fbFQ03KpppVSXWcXinqxtKKVUv2aLYRaYmGnHfZsHji6Nbh0JhhTqFWeia3vZ0whYOch4MGOZK1Vud7USsnv0+wv/RetlfI1ykrr/Z8L1Wt/u5HpGOzuwYlBv+9zuuMRbA8nI0w1DgAAgB0l6oa0ZyCkPQPvbibVzvfTKx2zhvnvp2vrZcH+6aWS1sp15SsN5a/zfrpTLOwoHdsY/CQiISWDsYwSEVepqL+djLpKRlwlo6GOff52Muoq6jq0fgfuAAIe3BHWWlUbnir1pip1T8VaQ8VqQ4VqQ8VqU8Vqo6Ms2K42VKw1O45b379WvvE3DuGQ0UDwjcNgPKy9AzEd3ZtufwuR2xTeZJMRmqECAACgbxhjlIj4Ycr1Bo6+ntY4QvmKH/i0gp+1YDtfqWutY7lWrqtYbWhurapCtaFSzf8McKP3851CjlEiHFI9Ub6VAAAgAElEQVQsElI8HFIs7CgeDika9rfbZZGQYmH/0XlcZ5l/jLOxLBxSNOwoHHLkOoYwCbsWAc8O5nlWTWvV9IKHtWo2/aXnWTWCcs/6663jG01/u970VG94qjU9NZr+dq3pqR6sN5qeas314+od29WGH9S0Aptqo6lq3VOl0WyX+cumqg1P1ca7u3m3JDvT/WhIyYirPZlYO+Fv9R8eDJqLtvoQDwZ9iOPhEDdmAAAAYAu0xhEaiIff08+pNTyVautf+q6HP/6Xuuv7GirXPJXrzfZnjHK9qXKtqZVyXTOrlfa+1nLzlPXvljFSOOQoGnIUdh1FQo7CrvGXIUdR11HE9dc7l5FQ57EhhV3j/4yO49yQUcgxch2jkOMES7O+DK2Xu5u2NxznOAqFzFXljvEfxihYV3ubz0aQ+jDgeXs2r//jC2/IWsnKylrJszbYltTalt8qxV9uWpeV56l9jILyd/s8SeuBTSukaQU2HeFNN0SCG1MsHFLM9ZPvaJCOx9yQhpMRxdyNqXg02NdaxsIhv7llxG2HOO3mmFFXiXCI8WsAAACAXS7iOoq4/jhDd1qj6anS8FSubQyFKvVNQVHNL681PNWCL7WrzfV1f2n9/ZvKi9WGqu0vwj3VGzZYrv+MncJphz7XDoBaAZHpKHeCYMhx/O3Qhud2rDuSURAktX5hECiZjZsd29fbv/HA6+1vb1+nfPPvuGs0pf/t4/fdzn+6XcW0Aop3dbAx85LO3+LvyElauMXnYPfiesBmXBPoxPWATlwP2IxrAp24HtCJ6wGb9eo1cdBaO3I7T7ylgOe2foExL1hrj23pL0HP4HrAZlwT6MT1gE5cD9iMawKduB7QiesBm/XjNcHcdgAAAAAAAD2OgAcAAAAAAKDHbUfA8xvb8DvQO7gesBnXBDpxPaAT1wM245pAJ64HdOJ6wGZ9d01s+Rg8AAAAAAAA2Fp00QIAAAAAAOhxBDwAAAAAAAA9bssCHmPM/2WMedMY84ox5k+MMYMd+37WGHPKGPOWMeYfb1UdsLMYYz4WnPNTxpif6XZ9sL2MMfuNMV8xxrxhjHndGPPjQfmwMeZLxph3guVQt+uK7WOMCRljXjTGfCHYnjLGPB/cJ/7QGBPpdh2xfYwxg8aYzwbvH04aY97PPaJ/GWP+VfB68Zox5jPGmBj3iP5ijPltY8ycMea1jrJr3hOM71eDa+MVY8xj3as5tsJ1rgc+c/apa10PHft+0hhjjTG5YLtv7g9b2YLnS5IesNY+JOltST8rScaY+yR9StL9kj4m6deMMaEtrAd2gOAc/0dJ3yrpPknfE1wL6B8NST9prb1P0lOSfjS4Bn5G0pettXdJ+nKwjf7x45JOdmz/W0m/bK09ImlZ0g93pVboln8v6S+ttUclPSz/2uAe0YeMMeOSfkzSMWvtA5JC8t8/co/oL78j//NCp+vdE75V0l3B40ck/fo21RHb53d09fXAZ87+9Tu6+nqQMWa/pI9Kmu4o7pv7w5YFPNbav7LWNoLN5yRNBOuflPQH1tqqtfaspFOSntiqemDHeELSKWvtGWttTdIfyL8W0CestVestSeC9bz8D27j8q+D3w0O+11J396dGmK7GWMmJP0TSb8ZbBtJH5b02eAQroc+YowZkPRBSb8lSdbamrV2Rdwj+pkrKW6McSUlJF0R94i+Yq39qqSlTcXXuyd8UtJ/tr7nJA0aY/ZuT02xHa51PfCZs39d5/4gSb8s6acldc4m1Tf3h+0ag+eHJH0xWB+XdKFj38WgDLsb5x1txphJSY9Kel7SmLX2SrBrRtJYl6qF7fcr8l+AvWA7K2ml440a94n+MiVpXtJ/Crrt/aYxJinuEX3JWntJ0i/J/wb2iqRVScfFPQLXvyfwXhN85uxzxphPSrpkrX15066+uR7eU8BjjPlvQb/ozY9Pdhzzr+V3zfj991pZAL3PGJOS9MeSfsJau9a5z1prtTFtxy5ljPm4pDlr7fFu1wU7hivpMUm/bq19VFJRm7pjcY/oH8G4Kp+UH/ztk5TUNZrio79xT0ALnzlhjElI+jlJ/3u369JN7nt5srX2m2+03xjzA5I+LukjwQ1Yki5J2t9x2ERQht2N8w4ZY8Lyw53ft9Z+LiieNcbstdZeCZpKznWvhthGT0v6hDHm2yTFJGXkj78yaIxxg2/ouU/0l4uSLlprnw+2Pys/4OEe0Z++WdJZa+28JBljPif/vsE9Ate7J/Bes0/xmROBw/K/FHjZ7/WvCUknjDFPqI+uh62cRetj8pvef8JaW+rY9XlJnzLGRI0xU/IHOvrGVtUDO8Y/SLormP0iIn/Qs893uU7YRsH4Kr8l6aS19t917Pq8pO8P1r9f0p9td92w/ay1P2utnbDWTsq/H/y1tfZ7JX1F0ncGh3E99BFr7YykC8aYe4Kij0h6Q9wj+tW0pKeMMYng9aN1PXCPwPXuCZ+X9M+D2XKekrTa0ZULuxSfOdFirX3VWjtqrZ0M3l9elPRY8P6ib+4PZj3kvMM/2JhTkqKSFoOi56y1/3Ow71/L7yPZkN9N44vX/inYTYJv6n9F/kwYv22t/cUuVwnbyBjzjKSvSXpV62Ou/Jz8cXj+SNIBSeclfZe19loDpmGXMsZ8SNJPWWs/bow5JH8Q9mFJL0r6PmtttZv1w/Yxxjwif9DtiKQzkn5Q/pdR3CP6kDHm5yV9t/z3iy9K+hfyx0zgHtEnjDGfkfQhSTlJs5L+jaQ/1TXuCUEQ+B/kd+UrSfpBa+0L3ag3tsZ1roefFZ85+9K1rgdr7W917D8nfybGhX66P2xZwAMAAAAAAIDtsV2zaAEAAAAAAGCLEPAAAAAAAAD0OAIeAAAAAACAHkfAAwAAAAAA0OMIeAAAAAAAAHocAQ8AAAAAAECPI+ABAAAAAADocQQ8AAAAAAAAPY6ABwAAAAAAoMcR8AAAAAAAAPQ4Ah4AAAAAAIAeR8ADAAAAAADQ4wh4AAAAAAAAehwBDwAAAAAAQI8j4AEAAAAAAOhxBDwAAGBXMcZ8tzGm0PGoGmP+ptv1AgAA2EoEPAAAYFex1v6htTZlrU1J2ifpjKTPdLlaAAAAW8pYa7tdBwAAgDvOGONI+rykC9ba/6Xb9QEAANhKtOABAAC71S9KSkv6sW5XBAAAYKu53a4AAADAnWaM+ZSk75H0Pmttvdv1AQAA2Gp00QIAALuKMeZRSX8l6VustS91uz4AAADbgS5aAABgt/mkpCFJf9sxk9YXu10pAACArUQLHgAAAAAAgB5HCx4AAAAAAIAeR8ADAAAAAADQ4wh4AAAAAAAAehwBDwAAAAAAQI9zb+XgXC5nJycnt6gqAAAAAAAA/ev48eML1tqR23nuLQU8k5OTeuGFF27n9wAAAAAAAOAGjDHnb/e5dNECAAAAAADocQQ8AAAAAAAAPY6ABwAAAAAAoMfd0hg8AAAAePc+/fx0t6ugZ5880NXf3+3/Bt3++wEA2C604AEAAAAAAOhxBDwAAAAAAAA9joAHAAAAAACgxxHwAAAAAAAA9DgCHgAAAAAAgB5HwAMAAAAAANDjCHgAAAAAAAB6HAEPAAAAAABAjyPgAQAAAAAA6HEEPAAAAAAAAD2OgAcAAAAAAKDHEfAAAAAAAAD0OAIeAAAAAACAHkfAAwAAAAAA0OPcblcAAIDd6tPPT3f19z/75IGu/n4AAABsH1rwAAAAAAAA9DgCHgAAAAAAgB5HwAMAAAAAANDjCHgAAAAAAAB6HAEPAAAAAABAjyPgAQAAAAAA6HEEPAAAAAAAAD2OgAcAAAAAAKDHEfAAAAAAAAD0OAIeAAAAAACAHkfAAwAAAAAA0OMIeAAAAAAAAHocAQ8AAAAAAECPc7tdAQAAgJ3KWqtSramlYk3LpZrylYbylYYK1YYKlbq/rDZVqNZVqDRUqXuqNpqqNT1V656urFbU8Dw1mlYNz8pae8Pf5xgjN2TkOo7ckFHIMQo7jkKOXx51Q4qHHcXCIcXCIcWDZSziKB4OKRlxlYy6Cjlmm/4LAQCAnYKABwAA9JWmZ7VYrGpurar5fFVz+Yq/XqhqsVjTcrHWDnSWi3XVmt4Nf144ZJSOhZWM+oFL1A0p4jqKuo6S0ZBcJxyENkaOuXHw4lmretOq6dl2MFRremrUPdWbVrWGp0q9qWrj+nUykhJRV5mYq3TM1YnpZY2moxpNR7VvMK4D2YT2DyWUjPI2EACA3YRXdgAAsGt4ntV8oarLK2VdXqnoC69c1kq5rtVSXavlutYqdRWrDXnXaEgTD4eUjLpKRkJKRF1NDCV0z1hIiYirRMRfxiMhRV2/BU3EdRRzHbmh7e/x3vSsqo2mKnVP5XpTlXpT5VozaFHUUL5Sb7c2+tt3FrRQqKqx6Y/OJiPaP5zQgeCxfziuQyMp3T2a1kAivO1/EwAAeG8IeAAAQM+w1mq1XNf0Uqn9uNCxPrNaUb25MciIuI4G42ENxMPaMxBTOuYqHQsr3W7lElYq5irchaDmdoUcEwRPNz/22ScPyPOslko1XVou+//Nltf/u714YVl//uoVNTsCoNF0VHePpXVkNKW7x9K6eyylu8bSGogT/AAAsFMR8AAAgB1nrVLX2fmizi4UdWbBX55dKOj8Ykn5SmPDsa2WKI/sH9LEQ3HtG4xrfDCmvQNx/d2pRcXCjsxNukbtdo5jlEtFlUtF9fD+wav2N5qeLq9UdHq+oLdn83p7tqB35vL6w3+4oHK92T5ufDCuhyYG9ODEgB4aH9SD4wO09gEAYIcg4AEAAF1hrdWV1YremSvondm8Ts0VdGbeD3QWCtX2ccZIE0NxTWaTeuzAUNCdKNFepm4wlsyL0yvb8af0PDfk6EA2oQPZhL7p6Gi73POsLq2U26HP65dX9eqlVX3xtZn2MQezCT04PqCHJgb06IEhPTQxoKgb6safAQBAXyPgAQAAW8paq9m1qk7OrOmd2bzemS3o7bmCTs8VVKiut8bJJiM6PJLSR46OamokqalcUodySe0fTigWJjDoBscx2h8EaR+5d6xdvlKq6bVLa3rl0opevbiqF6dX9IVXrkjyu8Q9PDGg900O632TwyrXmopHOH8AAGw1Ah4AAHDHlGtNvT2b15szazp5xV++OZPXSqnePmYkHdXdYyl95+MTOjKa0l2jKR0ZTSmbinax5rgVg4mInrkrp2fuyrXLFgtVHT+/rBfOL+sbZ5f0G189o1/7m9MyksYyMU3mEprKpXR4JKlEhLegAADcaby6AgCA27Jaquv1y6t67fKqXr20ptcvrersYlE2GKs3EQnp6J60vu3Bvbp3T1r37Mno7rGUBt/NyMDoOdlUVB+9f48+ev8eSX7Y9+KFZf32357VucWSTkyv6LkzSzKSxofiOjziB3sHhxNdmYkMAIDdhoAHAADc1Eqpppcvruq1S8Hj8qouLJXb+8cH47p/X0afeGSfju7J6L69GU0MxeU4/T24cT+LR0L6wOGczi2UJPlTu19cLunUXEGn5gr62jvz+u9vzyscMprMJtszdo2mo30/KDYAALeDgAcAAGxQqTd18sqaXrqwopcvrOjli6s6u1Bs75/MJvTQxKCefeKgHhjP6P59AxpO0ioHNxZyjA5mkzqYTeoj946pUm/q7EKxHfh88bUZffG1GQ0lwjq6J6Oje9KayiVp3QMAwLtEwAMAQB+z1p8l6fj5ZZ04v6yXLqzojStrqjf9flZjmage2T+of3ZsQo9MDOr+8QENxJkWG+9dLBzSvXszundvRpLfSuzt2YLenFnTP5xb0t+fWVQk5OjIaEpH96R1z5600jGuPQAAroeABwCAPlJvejp5ZU0vnFvW8ellHT+3rJm1iiR/zJyHJwb1w88c0iP7B/TI/iHtGYh1ucboF4OJiJ6YGtYTU8OqNTydWSjozZm83prJ640ra5Kk/UNx3b9vQA+M02oMAIDNCHgAANjFKvWmTkwv6/kzS/rG2SW9dGFF5XpTkj9uzvumhnXs4JAePziko3vSdIfBjhBxnaCbVkbWWs2sVXTySl5vXFnVX74+o798fUZ7B2K6f5/fRZBxewAAIOABAGBXqTU8nV8q6uxCUX/y4kW9fGFVtaYnY6T79mb03e/br8cPDunY5JD2DsS7XV3gpowx2jsQ196BuD58dFTLxZpev7yq1y+v6csn5/TfTs4pl4rq/n0ZPTg+oL0DMcIeAEBfIuABAKCHNZqeppdLOj1X0On5oi4ul+RZyUh6aGJAP/D0pJ6cGtaxyWHGzsGuMJSM6Jm7RvTMXSNaq9T1xuU1vX55tT0rVy4V0UMTg3pwfEBjGboYAgD6BwEPAAA9xLNWM6sVnZor6PR8QecWi6o3rYyk8aG4njkyoqlcUgezCf3QM1Pdri6wpTKxsJ46lNVTh7IqVht64/KaXr60oq+8Oae/fnNOezIxLZdq+vhDe3Uwm+x2dQEA2FLGWvuuDz527Jh94YUXtrA6AADsHp9+fvqO/JzVcl3vzOb19lxBZ+YLKtX8MXRG0lEdGUnp8EhKU7mk4pHQHfl9QK9bq9T12qVVvXpxVeeXSpL8Fm2feHifPvHIPo2madkDANiZjDHHrbXHbuu5BDwAAGyN2w146k1P5xaLeme2oLdn85rLVyVJmZirI6NpHR5J6vBIShm6XAE39Y/uGdFfvHJFn3/5sl69tKqQY/Q/3JXTdzw6ro/et4dgFACwo7yXgIcuWgAA7ABLxZrenFnT27N5nV3wu12FHKOpbFKPHxzSXWNpjTFTEHDLxgfj+pcfPKR/+cFDOjWX1+dOXNKfvnhJP/4HLykVdfWtD+zRdzw2rqemsnIc/n0BAHoXAQ8AAF3gWasLSyW9OZPXyStr7VY62WRExw4O6+6xlKZyKUVcpi0H7pQjo2n99MeO6qc+eo+eP7ukz524qC++NqP/cvyi9g3E9O2PjuufPjauI6PpblcVAIBbRsADAMA2qdabenuuoDevrOmt2bxKtaYcI01mkzo2Oax796SVTUW7XU1g13Mco/cfzur9h7P6hU8+oC+dnNXnTlzU//PVM/q1vzmtB8cH9E8fG9f/+PA+5fg3CQDoEQQ8AABsoeViTSdn1vTmTF5n54tqWqt4OKS7x1I6ujeju0fTjAECdFE8EvIHX354n+bzVX3+5cv63ImL+vn/+ob+zz8/qX9094i+49Fxfct9Y4qF+bcKANi5CHgAALiDmp7VSxdW9OWTs/rjExc1u+Z3vcqlov9/e/cdJdl12Hf+d1/l1F1d1TlN6JnpwWAGGMQBQBBRICFKoqgVFUjLory0VytLXsle6ax1tOdoj1dea3e1ljboyJYtrSXtQom0Ra4okgDJAQHQwBA5TM6hZzrnrq5894/3urp6Yk/orq7u7+ecOu/Ve7eqbs+8V+H3btBjfWnt7GhQbyoqH2N9AGtOSyKkLz6+RV98fIuODS2O1/OdI8NKhPz6oXs69NkHuvXApibGwwIArDnMogUAwG3KFkr63olRfeOjQX3nyLDG5vLyOUa9qajuak9oZ0cD3TyAGvn8vt7benypbPXGqTF9+Z0L+vqHg5ovlLSlOaYfv79LP3Z/t7qSkTtUUwAAmCYdAIBVN5crav/RYX3jo0HtPzKsuXxJibBfT/e36tm7WvXUjlZ97cNLta4msOHdbsBTbTZX1Nc/vKQvvX1BB06PyxjpY33N+uwD3frk3Uy5DgC4fUyTDgDAKpjM5PWtw26o88rxEeWLZaVjQX16b6ee392hR7emmfUKWMfiIb9+4sEe/cSDPTo3ltGX37mgL79zQb/yl+6U6z9MFy4AQA0R8AAAcB3DM1m9eHBI3zw4qNdPjqlYtupsDOvv7evV83e368HNKcbTATag3nRU//S5HfrlZ7frwOlxfentC/rq+xf1F2+epwsXAKAm6KIFAMBlzo9n9M2Dg/rmwUG9dXZC1kpbmmN6fne7nr+7Xfd0Ny7r6vwLB86tQm0BXM+d7KJ1I3O5ov6OLlwAgNtAFy0AAG7TieFZffPgoL7x0aA+HJiSJN3V0aBfeXaHnt/drh1tcbpcALiuGF24AAA1RMADANiQrLU6eHFa3zw4qK9/NKgTw7OSpPt6k/r1H9ypT97drs3NsRrXEkC9ul4Xrs3pqD77QDdduAAAdxRdtAAAG0a5bPXu+Ql946NBfePgoM6Pz8sx0r4taT2/u12fuLtNHY137scWXbSA2lvNLlo3Mpcr6usfDepLb5/XG6fowgUAuBJdtAAAuIZCqazvnx7X1z+6pBcPDml4JqeAz+jxbc36pae36QfualM6Hqp1NQFsALGQX599oFuffaBb58fdLlxfenuxC9en9rTrM3u7tG9rmsHbAQA3jRY8AIB1J1so6X/62mF9dHFahy9Na75QUsBntKMtobs7G7WzPaFwgCvlwEawFlrwXK81X9lanRmb0ztnJ3Xw4pRyxbIawn7d053U3p6kOhrDtz1ez1r4NwAALA8teAAAG95MtqDvHhvRNz4a1P4jw5rLlxQOONrZ3qDdnQ3a1ppQ0O/UupoAsIRjjLY2x7W1Oa5P39upI4PTeu/8pP7zyVG9dmJUrYmQ9vYkdW93Uk2xYK2rCwBYwwh4AAB1a2g6q5cODemlQ0N6/eSY8qWy0rGgPr23UyG/T1tbYvI7hDoA6kPQ7+ie7qTu6U4qkyvqw4tTeu/cpF48NKQXDw1pUzqqvT1J7elsVDTE13gAwFJ8MgAA6oa1VseHZ/WS92Pn/fOTkqRN6ai+8NgmPberXQ9sapLPMQxwDKCuRUN+7duS1r4taU3M5fX+hUm9e35SX3nvov72/Uva0RbXvT1J3dXRoICPIBsAQMADAFjjSmWrt89O6KVDg3rp0JDOjGUkSff2JPVrn+zXc7vatL01fttjVADAWtUUC+qp/lY9uaNFl6ayeu/8pD64MKnDgzMK+R3d3dmgvT1N2toSk8N7IQBsWAQ8AIA1Zz5f0qvHR/TSoSF9+8iwxufyCviMHutr1j/8+FY9t6tNbQ3hWlcTAFaVMUadyYg6kxE9v7tdp0bm9P75SX10cUrvnJtULOTX7s4G7e5q1JZmwh4A2GgIeAAAa8LwdFYvHx3RS4eH9OrxEWULZSXCfj2zs1XP7WrTkztalAgHal1NAFgTHGO0rTWuba1xfXpvp44MzujDgSm9c25CB06PKxby6+7OBu3palSxVJafblwAsO4R8AAAaqJUtnrv/KRePjqs/UeH9dHAtCSpszGsn3qwR8/tate+rSnGlgCAGwj4HO3patSerkbli2UdHZrRRwNTevfchL5/elx/8+6APrm7XZ/a3cH7KgCsYwQ8AIBVMz6X1yvHRrT/6LC+e2xEk5mCHCM9sKlJv/bJfj3V36JdHQ2MpwMAtyjoXxr2HBua0VS2oP/0zoBeOHCu0jLyE7va9WR/i+LMxgUA6wbv6ACAFZMrlvT2mQm9emJUrx0f1UcXp2StlI4F9czOVj3d36ontreoMUrXKwC404J+R7u7GvX5fb2az5f02olRvXRoUN86PKyvvHdRQZ+jj21L6xN3t+vZu1rVmmBsMwCoZwQ8AIA7xlqrY0OzevX4iF49Pqrvnx7XfKEkv2N0X29Sv/LsDj3V36I9XY1yHFrpAMBqiQR9em5Xm57b1VaZnfDFg4P65qFB7f+PH8oY6d7upJ7ub9XTO1u0u5P3aQCoNwQ8AIBbZq3V6dE5vXFqXG+cGtPrp8Y0MpOTJG1tieknH+zWx7e36JG+NN0AAGCN8DlGD29J6eEtKf3GD92lo0MzevHgkL5zZFi/9+1j+t1vHVNzPKgnd7hhz8e3t6gxQktLAFjr+LYNAFi2ywOdN06NadgLdFoTIT2yNa3Ht6X1+PYWdSUjNa4tAOBGjDHa2d6gne0N+m+e3a6x2ZxeOT6i/UdG9K3DQ/ryOxfkc4we6G3SEzua9bFtzdrT1cisXACwBhHwAACuKVcs6aOBab1zdkJvn53Q2+cmKi10WhMhPdqX1iNb3dvmdJTBkQGgzqXjIf3Yfd36sfu6VSyV9f6FSe0/4g6O/zsvHtPvvHhMiZBfj/Sl9fi2Zn1sW1p9LXHe/wFgDSDgAQBUjMzk9M65iUqg88HAlPLFsiSpNxXV49ua9fCWFIEOAGwAfp+jBzal9MCmlH71k/0am83p9VNj+t6JUX3vxJheOjQkSWprCOljfc16pC+thzan+HwAgBox1tplF37wwQftW2+9tYLVAXAnvHDgXE1f//P7emv6+hL/BssxMZfXhwNT+nBgSh9cmNSHF6Z0cSorSQr6HO3pbtQDm5p0f2+T7t+UvOnZVWr9fwAAWFnjc3mdHJnVieFZnRyZVSZfkiQlQn5tSke1uTmmTemYOhrD+plHNtW4trW1Fj4T6+G7CQDJGPO2tfbBW3ksLXgAYJ2z1mpoOqfDg9M6fGlaBwem9cHApM6Pz1fKbG2O6aEtKe3patR9vU3a3dWgkN9Xw1oDANa6VCyoVCylhzanVLZWIzM5nRmb09mxjM6Mzumji9OSpJDf0YuHhvRAb5Pu7WnUvd1JNcWCNa49AKw/BDwAsI7M50s6NjSjI4PTOnzJXR4ZnNFkplAp05uK6p7upH5m3ybt6W7U7q5GNYSZHQUAcOscY9TWEFZbQ1j7tqQlSZOZvM6MZXRmbE6DU/P6vW+PaKHzwKZ0VHt7krq3O6l7e5K6u7NB4QAXFgDgdhDwAECdsdZqZDank8NzOjky693mdHJ4VgOTi61yokGf+tsT+sHdHbqrI6Gd7Q3qb0uoMUqYAwBYecloUHujQe3tSerz+3o1ky3ow4EpvX9+Su+fn9T3T4/rK+9dlCT5HaNtrXHt6mjQrs4G3dXh3lK09AGAZSPgAYA1qFy2GpzO6tx4xr2Nucuz4xmdGpnVTLZYKRsJ+NTXGtODm5v0k8096m9P6K6OhHqaonIcBrkEAKwNiRcroeYAACAASURBVHBAj/U167G+5sq2oems3j8/qffOT+rgxWm9dmJU//Hdgcr+toaQdnlhT397Qn0tcW1tiSka5GcMAFyOd0YAqIFsoaSLk/O6NJWtLC9NzeviZFbnJzK6MD6vfKlcKe9zjDqTYfWmovrM3i71tcTU1xpXX0tc7Q1hghwAQF1qawjrE3e36xN3t1e2jc3mdPjSjA5fcseOO3RpWq8eH1WxvDg5TFcyom3e5+C2Vve2uTmqlniIGbwAbFgEPABwh1hrlS+VNZst6s0z4xqezmlkJquR2Zy77i0vTc1rompMnAXpWFAdybB2tif03K429aai2pSKqTcVVUcyrIDPqcFfBQDA6krHQ3p8e0iPb19s6ZMvlnVmbE4nhhdn7ToxPKsDp8eULSxeEIkEfOpNRdWbjrqfo+moelLuelcywjg/ANY1Ah4AuEzZWuWLZWULJWULZc0XSsoVSsoWS5ovlJXJF5XJlTRXvcyXNJcrLl5dfGnx+XyOUXM8qNZEWO2NYe3tTaqzMayOxog6kmF1NkbU3hjmSycAANcQ9Dva0ZbQjrbEku3lstXA5LxOjMzq3FhGZxe6NI/N6dXjI0vCH2nxYkpHY0Qd3mdxp3e/vSGs5kSQ7l8A6hbvXgDqXrHkhjDz+ZK7LJR0bjyjQqmsfLG8dFmyynllcpUQxw1ysl6IkyuUZW/wmuGAo1jQr2jQp8ZIQB2NEcVCPkWDfiVCfn16b6daEiG1JkJqigbpQgUAwApwHKOelNtK53ILkxIsBD8XJ+d10esSfW4sozdOjS0Z025BLOhTSyK0eIuH1BwPKRUPKhkJqjESUDIaqCzjIT/dwgCsCQQ8AG6ZtValslVx4VYqq1S2OjY0o3yxrFyxrFyxpHzRDVjyXtCSW7jvbVu4f7Wy1eVz3rasF+Zk8kVlC+UlY9Ush2OkcMDn3vyOQgGfUrGgwgF3Pez3KRJwFA743PsBRxFveyjgKBr0y3eDwOaJHS23808LAABukzFGrYmwWhNhPbg5ddUys7miLnnBz/C02616dCavkVm3m/XRwRm9NjOq6asEQQt8jnHDnkhAjQvBTySgZHQxDDp0cVrhgE9Bv6Ogz1HI7yjodxTyu9tu9L0CAJaDgAdYB0plu9gSpVjWyExOhVLZu7nBS6Fsr9zmrVcvi6XyFYHNwv2S9xylqvtX87+9dOym/4agb+GLjrtc+AIU9C9ubwwGFPQ5igTdACYa9Csc8Cka9CkS8Hnb3ftvnBpX0O8o4DMK+hwFvOcL+NxtXGkDAADxkF/b2xLaflnXr8tlCyVNzRc0mSloMpN31+cLmsoUNDmfr+ybmi9obDavUyNzmszkrxsMVfM5ZvE70FUCoIXtfp9xv8s4Rv6q7zV+x1t69wPePr/PKODts9by/QdY5wh4gBVQLltli1XdfrwuQdlCuTKWy8K++aruQTkvoJnPL4Y1C4/PeWPBZC97fO4WWrAsMNLiF4WqLwh+n5HfMQoFfPI7Rj7Hve93HPl8RgHHyFdVrrLf58jvGD3Z33JZOONb8qXlakHOnf7CcbVBjAEAAG7FQsvftobwTT2uVLaayRb0p6+fvbJlc6W180Lr5dLSlsvFsuZyeeWKbrfyoncx7kbdyK/lN/+/gwr73ZbJC39PyO8sWVb2VZVbaM280JLZ3Vf12Cv2ed/9Au73PUIlYPUQ8GBDKJft4ngrXjgyn19Yd0OSy8OWhXK5yvbqfdVjtywNcm6ly1C1Snegyz54wwFHzfHgYtci78MzHPB55RfLvX12csnVm8oVHWfpNp+zMi1Zfviezjv+nAAAAPXG5xglo0E1x0N35PmstSpZWwl7Fltge/fL5Sv3ea2y+9sT1/w+myuUNZHJX/G9diGAuh0h78LetUKl6qAo5K/+jru4vN7jlgZK7v2VuHgI1AMCHqyqUtlWxlrJFcvKFarWvcFtc5fvL7khy8IHTK46VClePWCp3j5fcK+G3KqrXclY+DBJxYJX2V599cNdRoK+JR9QC2UqQc4dvspxG/kSAAAA1ihjjPzGyO/opmff/Py+3lt6zXLZKl8qVwKfhe/YueL1l5WAqOpxly8z+aImMtXbFn8T3M4FU2N07SCpsryy9dHSVkhXb6EUukY4Ffb7GAYANUfAs4ZYa2WtZOUGIWVvANuStSqXq9d1lW3uslRevLmPv/pzFUpWxXL1eCyL47MUy1aFojtmS/U4LcVyWYWiveLKwNLncZdLQpyqN/biNcZsuRnVaf1CQBLywpRkNHhlC5iqpqSRywKWyoC6wcWmp9UBDc1KAQAAsJE5jlHY8d10oHS7qlvg32h5+barhUrV4dNcrqix2fySUCrnXSQulG7994pjdM1WSEG/UxkKIeANaxDwV4+ntDhUwsJ4S9VjK/kdc9lzLO4P+NyBuh1jlix9jq6yrWrdGDmOqtarltX7jfsYY8RvozVuwwU8Hw1M6RdfeMcLUrxAxTuHrbWy0pJ9kirbpMUAZklZb11X21f1GF12v7rcWuX3uvQsjtOy9I2keiA3v2MUDfrl95klLVIWxl8JBRwFfVffXln3ZjS66jqpOAAAALAhOI5xJ9AIrm6wVCrbJYFPrtI7oFwZL/NqyyvCp8sen/fGUcoWS1dcXF9yv7R4Qf1OXBxfKca443lWgh+5G4y3zzHGW3eXi/uMHLO43Xg73ce4z2OqylZea+E1Ko/x6uEt7+5q1O9//v5V+dvXMmNvIl0wxoxIOrty1QHWlGZJo7WuBLBOcX4BK4tzDFhZnGPAytrI59gma23LrTzwpgIeYCMxxrxlrX2w1vUA1iPOL2BlcY4BK4tzDFhZnGO3xql1BQAAAAAAAHB7CHgAAAAAAADqHAEPcG1/WOsKAOsY5xewsjjHgJXFOQasLM6xW8AYPAAAAAAAAHWOFjwAAAAAAAB1joAHAAAAAACgzhHwAB5jTMoY85Ix5ri3bLpKmb3GmNeNMQeNMR8YY36qFnUF6oUx5nljzFFjzAljzD+/yv6QMeYvvf0HjDGbV7+WQP1axjn2z4wxh7zPrG8bYzbVop5AvbrROVZV7seNMdYYw7TOwE1YzjlmjPlJ77PsoDHmhdWuYz1hDB7AY4z5XySNW2t/23tzabLW/neXldkhyVprjxtjOiW9Lekua+1kDaoMrGnGGJ+kY5Kek3RB0puSPmetPVRV5h9Lusda+18bY35a0o9ZawlOgWVY5jn2tKQD1tqMMeYXJD3FOQYsz3LOMa9cQtLXJAUl/ZK19q3VritQj5b5ObZd0l9JesZaO2GMabXWDtekwnWAFjzAoh+V9Cfe+p9I+szlBay1x6y1x731i5KGJbWsWg2B+vKwpBPW2lPW2rykv5B7nlWrPu++JOlZY4xZxToC9eyG55i1dr+1NuPdfUNS9yrXEahny/kck6T/UdL/LCm7mpUD1oHlnGP/SNLvW2snJIlw5/oIeIBFbdbaS976oKS26xU2xjws90rNyZWuGFCnuiSdr7p/wdt21TLW2qKkKUnpVakdUP+Wc45V+6Kkr69ojYD15YbnmDHmfkk91tqvrWbFgHViOZ9jOyTtMMZ8zxjzhjHm+VWrXR3y17oCwGoyxnxLUvtVdv1G9R1rrTXGXLP/ojGmQ9KfSfqCtbZ8Z2sJAMCdZYz5GUkPSnqy1nUB1gtjjCPpX0v6uRpXBVjP/JK2S3pKbivUV4wxexgi4+oIeLChWGt/4Fr7jDFDxpgOa+0lL8C5avM/Y0yD3H7Wv2GtfWOFqgqsBwOSeqrud3vbrlbmgjHGL6lR0tjqVA+oe8s5x2SM+QG5FzKetNbmVqluwHpwo3MsIWm3pJe93sXtkr5qjPk04/AAy7Kcz7ELcseSK0g6bYw5JjfweXN1qlhf6KIFLPqqpC9461+Q9JXLCxhjgpL+k6Q/tdZ+aRXrBtSjNyVtN8Zs8c6dn5Z7nlWrPu8+K+k7ltH/geW64TlmjLlP0r+V9GnGLQBu2nXPMWvtlLW22Vq72Vq7We44V4Q7wPIt57vi38htvSNjTLPcLlunVrOS9YSAB1j025KeM8Ycl/QD3n0ZYx40xvx7r8xPSnpC0s8ZY97zbntrU11gbfPG1PklSd+UdFjSX1lrDxpj/oUx5tNesT+SlDbGnJD0zyRdcwpaAEst8xz7XyXFJf2195l1+RdnANewzHMMwC1a5jn2TUljxphDkvZL+jVrLa29r4Fp0gEAAAAAAOocLXgAAAAAAADqHAEPAAAAAABAnSPgAQAAAAAAqHMEPAAAAAAAAHWOgAcAAAAAAKDOEfAAAAAAAADUOQIeAAAAAACAOkfAAwAAAAAAUOcIeAAAAAAAAOocAQ8AAAAAAECdI+ABAAAAAACocwQ8AAAAAAAAdY6ABwAAAAAAoM4R8AAAAAAAANQ5Ah4AAAAAAIA6R8ADAADWFWPMGWPMrxtjDhljJowx/7cxJlzregEAAKwkAh4AALAe/T1Jn5TUJ2mHpP++ttUBAABYWQQ8AABgPfq/rLXnrbXjkv6lpM/VukIAAAAriYAHAACsR+er1s9K6qxVRQAAAFYDAQ8AAFiPeqrWeyVdrFVFAAAAVgMBDwAAWI9+0RjTbYxJSfoNSX9Z6woBAACsJAIeAACwHr0g6UVJpySdlPRbta0OAADAyvLXugIAAAAr4E1r7b+qdSUAAABWCy14AAAAAAAA6hwBDwAAAAAAQJ0z1tpa1wEAAAAAAAC3gRY8AAAAAAAAde6mBllubm62mzdvXqGqAAAAAAAAbFxvv/32qLW25VYee1MBz+bNm/XWW2/dyusAAAAAAADgOowxZ2/1sXTRAgAAAAAAqHMEPAAAAAAAAHXuprpoAQAA3KwXDpyr2Wt/fl9vzV4bAABgNdGCBwAAAAAAoM4R8AAAAAAAANQ5Ah4AAAAAAIA6R8ADAAAAAABQ5wh4AAAAAAAA6hwBDwAAAAAAQJ0j4AEAAAAAAKhzBDwAAAAAAAB1joAHAAAAAACgzhHwAAAAAAAA1DkCHgAAAAAAgDpHwAMAAAAAAFDnCHgAAAAAAADqHAEPAAAAAABAnSPgAQAAAAAAqHMEPAAAAAAAAHWOgAcAAAAAAKDOEfAAAAAAAADUOX+tKwAAALAevXDgXE1e9/P7emvyurVWq39vqbb/5rX8u2tpox7nAHA9tOABAAAAAACocwQ8AAAAAAAAdY6ABwAAAAAAoM4R8AAAAAAAANQ5Ah4AAAAAAIA6R8ADAAAAAABQ5wh4AAAAAAAA6hwBDwAAAAAAQJ0j4AEAAAAAAKhzBDwAAAAAAAB1joAHAAAAAACgzhHwAAAAAAAA1Dl/rSsAAACw0VhrVbJWxZJVsWxVLJVVLFuVylbWSlZWZeuWq14aScZIxpirrvuM0dmxOQV8joJ+x1166z7H1PivBgAAK4mABwAA4CbM50uayOQ1kclrKlPQRKbgrs8XNJ0taC5X1FyupCOXppUrliu3fLGkQsmqWC6rWLKyK1S/3/v28atud4wUCfgUDfkVC/oUDfoVC122DF5/f0M4oGTUvUUCPhlDaAQAwFpBwAMAADa8YqmswemshmdyGp7OaWQ2p5HprLucyWl4xl2Oz+WVK5av+TxBv6N4yA1ECkWrkN9RLORTKhZUyGtR4/cZ+Z2FpZHf5yjgGPm8mzFGjpGMvKVZXEpuax67sLSqrJetVLZWD21OqVAqq1Byg6VCySpfLCtfKilbKCuTdwOoheVkJq+ByZIyuaLm8iXN5Yoqlm8cPwV9jhoiXuDjLRsjQW8ZqFoGK/ubYkElQn6CIQAAVgABDwAAWPcy+aLOjWc0MDGvi5PzGpjM6uLkfOU2OJ3V5ZmGMVI6FlJLIqTWREjbWxNKx90AoykaVFPUCy+8+42RgMIBX+XxLxw4t8p/pevHH+i+7efIF90gKJNfDILmckVNZwuazBQ0NV/Q5PzCel6TmYIuTmZ1+NKMJjN5zeVL13zuoM9ROh50b7GQ0vGgmuMhpWNBpePefW97KhZc8m8KAACujYAHAACsC9lCSaOzbiubsbm8xmbz+pt3B3RmbE7DM7klZYM+Rx3JsDobI3q0r1ldybA6khG1N4QrgU4qFpTftzHnowj6HQX9QSWjt/b4fLFcFQa5AdBEpqCJubxG53Iam81rbDansbm8TgzPamQ2p/w1WkYlQn4vEAqpOR5UayKs1kRIrQ0L4Zt7v2ytHFoGAQA2MAIeAABQN6y1mskVq7pNZSvdp2ayxSVlE2G/7mpv0JM7WrQpHVVvOqaepoi6khE1x0NyGHR4xQT9jprjITXHQ8sqb63VXL6ksdmcRqvCn8r9ubxGZ3I6NTKnN06Na2q+cMVzGEnxkF+JsF+JcEDx8OJ6IuRXQ9ivRCSgRNgvv7MxgzsAwPpGwAMAANakXKGkwemsBqezujSV1eBUVsMzWWULiy09Qn5HLYmQtrfG1ZIIq9nr9pOKBRX0L/0RP5st6vClGR2+NLPaf8qqqlXXsJWQioWUil0ZEhVKZc1mi5rJFjSdLWomV9RstqCZbNG7FXRxcl6zueIVg1kbSbGQX42RgBrCfjVEAt56QA2RgBoifjWGAwrRNQwAUGcIeAAAQE1ZazU17/4gv+gFOYPTWY3P5StlwgFH7Q1h3dOddLvnJNyuVA1hBuzdiAI+R02xoJpiweuWK1ur2dxi6DMzX9RUtqBpb8az8UxeZ8Yymi9cOWZQyO9cFv74lfQGkU56g0dfHiICAFBLBDwAAGDVWGs1nS1qYGJeA5MZDUzOa2BivjIor5GUjofUmYzo/t4mdTSG1d4YVjISIMjBTXOMccOZcEBS5JrlFsYMWgh+puaLmp53B5OezhY0PJ3VTPbK1kDRoE/JaEDfPTaszqTb/a+7KVJZT8WCHLcAgFVDwAMAAFbMTLagE8OzOj+R0fnxjC5MuN1mJDfMaWsIq7+9QV3e2DjtDWFaRWDVLWfMoFLZasYbOHrSGzh6Yf3UyJxePT6qzGWzh4UDTiXs6UpG1JOKalM6qs3pmHrTUS94AgDgziDgAQAAd0S5bHVyZFbvnpvUO+cm9O65SR0bnpH1mj20xEPa0RZXZzKi7mRE7Y0RwhzUDZ9jlIwGlYwGJcWW7Pv8vl5ZazWZKbit0ryWaRe99YuT8zp8aVqjs/klj0vFgupNRbXZGwR8Uyqqzc1R9aZiao7T+gcAcHMIeAAAwC3JFUv64MKU3jwzrjdPj+utsxOVmawawn7d19ukH9zTrqlMQd1NUUWCDFqL9csYUxkXaHdX41XLzOWKOjuW0bnxOZ0Zy1TW3zwzoa+8f7EShkpSLOhTbzrmhT9uq58tzTH1tcQJfwAAV0XAAwAAlmU6W9DbZyfcMOfMhN67MKl80Z3Rqq8lph/a06EHNjXpvt4mbW2OVaYhX0+zOgG3Ixbya1dng3Z1NlyxL1cs6cLEvM6NZXRmbE5nxzI6Ozano0Mz+vbhYeVLi7PHNYT96muNq68lrtlsUS2JkFriITXFgvI5BD8AsFER8AAAgKuayRb05plxvXFqXK+fHNPBi1MqW8nvGN3d1agvPLpJD25O6cFNTUpfZ+wSADcW8vvU1+KGNpcrla0uTs7r9OicTo7MurfhOb1ybETDM7lKOZ9j1JoIqa0hrLaFJYOUA8CGQcADAAAkSbO5ohvonBzTG6fG9OGAG+gEfY729ib1S89s1yNbUtrbm1Q0yFcIYLX4HKOeVFQ9qaie2NGyZN8fvXpao7M5Dc/kNDyT1dB0VqdH5/Te+clKmaDfUVsipPbGiDoaw+psdIOfkJ9ukwCwnvDtDACADapYKuv9C1N69fiIXjs+qnfPT6pUtgr4jPb2JPVLT2/TI1vTun9Tk8IBfggCa1Ek6KuEP9WyhZKGprMams5paDqrwemsPhpwx8yS3Fns0vGgOrzQp6Mxos5kWAlm9gKAukXAAwDABmGt1ZmxjF47PqJXj4/q9VNjmskWZYy0p6tRP//EVn1sW7Pu721iQGSgzoUDPm1Kx7QpvTjjl7VWU/MFXZrK6tLUvC5OZnVhIqMPB6YqZRojAXUlI+pu8qZ3b4rQYg8A6gTv1gAArGOTmby+d2JMr51wQ50LE/OSpK5kRD+0p0OPb2/Wx/qa1RQL1rimAFaaMYtTvd/VsTjQ83y+pEvTi4HPwMS8Dl2aruxPxYLqSkbU0xRRbyqqzmREfp9Tiz8BAHAdBDwAAKwj+WJZb5+d0Gsn3G5XHwxMyVopEfLrkb60fv6JrXp8e4s2p6MMugpAktvNa2tzXFubFwd4ns+XNDA5r4HJeV2YyOj8+GJLH59j1NkYVq/XNaw3FVUySkgMALVGwAMAQJ07P57R/qPD2n9kWG+cGtd8oSSfY3RfT1K//Ox2fXx7s+7tTnLFHcCyRYI+bWuNa1vrYugznS3o/HhG57zbgdPj+t7JMUnu1O2b0jFtTke1uTmmtoawHEJkAFhVBDwAANSZQqmsN8+M6+WjI/rOkWGdGJ6VJG1KR/UTD3br8W3NeqQvrQYGSwVwBzWEA7q7s1F3dzZKkorlsganspXA5+zYYiufcMDRplRMm5vd0KerKSK/Q8gMACuJgAcAgDowPJPVy0dH9PLRYb16bFQzuaICPqN9W9L63MO9erq/RVtb4jd+IgC4Q/yOo+6mqLqbonqsz902kcnrzOiczozN6cxoRkeHZryyRr3pqLa1xNXXEldnMiKfQwsfALiTCHgAAFiDymWr9y9Mav+RYe0/OlK5Kt7WENIP3dOhp3e26mPbmhUP8VEOYO1oigbV1BvUfb1NkqTZXFFnRud0dmxOJ0fm9OKhIUlDCvkdbW2Oqa/VDXxaEyHGBQOA28S3QgAA1oipTEGvHB/R/iPD+u6xEY3N5eUY6b7eJv3aJ/v1VH+LdnU08CMIQN2Ih/za3dWo3V1ut67ZXFGnRmZ1cmRWJ0fmdHjQbeGTCPm1tSWmPq+FDzP7AcDNI+ABAGwoLxw4V+sqVFhrNTSd09HBaR0dmtG58YzKVooEfOpvT+jZu9q0ozWuqNdK5/3zU3r//FSNaw0Aty4e8uue7qTu6U5Kkibm8l7Y4wY+719w3+NSsaD6vMBna0uc1ooAsAy8UwIAsIryxbJOjszq6OCMjg7NaGq+IEnqbAzriR0t2tmWUHcqyuwzADaEplhQD8ZSenBzStZaDc/k3LBneFYfXJjSm2cmJEntDWFta41rR1tCm9PRGtcaANYmAh4AAFbY2GxOR4dmdHRwRqdG51QqWwX9jra1xPXszlbtaEuoIcKMVwA2NmOM2hrCamsI67G+ZpXKVhcn53VyZFYnhmf1+skxvXZiVEGfo1eOj+rJ/hY9taNFPSkCHwCQCHgAALjjiqWyzoxlKl2vRmfzkqTmeEiPbk27V6Cbo0wZDADX4XOMelJR9aSieqq/VbliSadG5nRsaEZHBqf1rcNDkqS+lpie3NGqp/pb9PCWlMIBX41rDgC1QcADAMAdMDVf0DGvlc6JkVnli2X5HaMtzTE9sjWt/raE0vFQrasJAHUr5Pfpro4G3dXRoM893KOTI3P67rERvXx0WP/PgbP64++dViTg06N9aT25o0VP9bdoUzpW62oDwKoh4AEA4BaUrdX58UxlLJ1LU1lJUmMkoL09SfW3JdTXElfQTysdALjTjDHa1hrXtta4vvj4Fs3nS3rj1JhePjqsl4+N6DtHhiVJW5pjenJHi57sb9GjW9O07gGwrhHwAACwTJlcUceGZ3V0cFrHhmY1XyjJMVJvKqZP3t2u/vaE2hIhpjEHgFUWCfr09M5WPb2zVZJ0ZnSuEvb8+ffP6T/85zMKBxw91tesp/tb9PTOVnU3MXYPgPWFgAcAgGsoW3eAz6NDMzo2OKMLE/OykmJBn3a2J9TfntD21oQiQa4IA8Basrk5pp9r3qKf+9gWZQslHTg9rv1HhvUd76avHNSOtrie3tmqZ/pb9cCmJvl9tLgEUN8IeAAAqJLJFXV8eFZHh2Z0fGhGc/mSjKSupoie3tmq/raEupoiTGMOAHUiHPC53bR2tOg3f2SXTo3OVcKeP3r1tP7td0+pIezXEzta9HS/O1gzY6YBqEcEPACADe1arXSiQZ92tCW0oy2uba0JxUN8ZAJAvTPGqK8lrr6WuP7hx7dqJlvQa8dH9Z0jw9p/dER/+8ElGSPd253UMztb9czOVt3d2UDXWwB1gW+rAIANh1Y6AABJSoQD+sE9HfrBPR0ql60OXpx2u3EdHdbvfuuY/vVLx9SaCOmp/hY9s7NVj29vIfAHsGbx7gQAWPeKpbLevzCl146P6ktvn6eVDgDgCo5jtKe7UXu6G/XLP7Bdo7M5vXx0RPuPDOvrHw7qr966oIDP6OEtKT3d77bu2doSr3W1AaCCb7IAgHXHWqvTo3N67cSoXj0+qjdOjmkmV5QxUleSVjoAgBtrjof02Qe69dkHulUolfX22YnK2D2/9bXD+q2vHdbmdFRPeWHPvq0phfwMug+gdgh4AADrwthsTt87OabXjo/oteOjujiVlSR1N0X0w/d26PFtLXqsL62vfzRY45oCAOpNwOfoka1pPbI1rV//1F06P57R/qNu2LMwDXs06NPHtjXrmZ2terq/Ve2N4VpXG8AGQ8ADAKhL2UJJb54Z12vH3VY6hy5NS5Iawn491tesf/x0sz6+vVmb0rEa1xQAsN70pKL62Uc362cf3az5fEmvn/IGaj4yopcODUmS7upo0DM73bF79vY0yefQYhTAyiLgAQDUhVLZ6vClab16fFSvnRjRm2cmlC+WFfAZ3d/bpF/9xA49vr1Fe7oa+RINAFg1kaBPz+xs0zM722St1bGhWS/sGda/+e4p/f7+k2qKBvTx7S36+PZmfXx7C617AKwIAh4AwJpUKJX14cCUvn96XAdOjemtMxOayRUlSf1tCf39Rzbp8e3NenhzSjEGRwYArAHGGPW3J9TfntAvPNWnqUxBJ2uUKwAAEP5JREFUrxx3B2p+5fiIvvr+RUnSjrZ4JfDZtyWtSJCxewDcPr4RAwDWhGyhpPfPT7qBzulxvX12QvOFkiSpryWmH763U/u2pPRYX1qtDVz5BACsfY3RgH7k3k79yL2dKpetjgzO6NXjI3r1+Kj+7I2z+qPXTivoc/TQliZ9fHuLHt/WrF0dDXJoiQrgFhDwAABqIpMv6p2zkzpwekwHTo/rvfOTyhfLMsZtofNTD/Xo4S0pPbQ5pZZEqNbVBQDgtjiO0a7OBu3qbNDPP9mn+XxJ3z8zrte8wOe3v35EkpSOBfW415Xr49ub1cZFDQDLRMADYIkXDpyr2Wt/fl9vzV57o1rN/+9MrqhzExmdGZ3T6dE5XZrKqli2coy0u6tRX3h0kx7ektZDm5uUjAZXrV4AgPpTy+8rd9qW5ri2NMc1PV/QiZFZnRie1bcOD+sr77ndudoaQtremtAXH9+ih7akFKdbMoBr4N0BAHDHla3V0HRW58YzOj+e0bnxjEZn85IknzHqaorov3piqx7ektIDm5qUCAdqXGMAAGqrIRLQ/b1Nur+3SWVrNTiV1YnhWR0fntEbp8b02olR+RyjPV2NerQvrUe3pvXg5iZFg/ykA+Di3QAAcFustZrOFjUwkdGFiXmdG8/owuS88sWyJCkW9Kk3FdUDvU3qSUXV3RRV0O/QYgsAgGtwjFFnMqLOZERP7GhRvljWtta4Xj81qtdPjunfvXJKf/DySQV8Rvd2J/VoX1r7tqR1X2+SiQeADYyzHwBwU2ZzRQ1MzOvCZEYDE/MamJivzG7lGKm9Maz7e5PqTUXVm4qpKRqQMQwWCQDArQr6HT2+vVmPb2+WJM3linrr7IRePzmm10+N6ff3n9D/+Z0T8jlGuzsb9NDmlB7cnNJDm5uUjjOOHbBREPAAAK7KWqvJ+YIuTWZ1aWpel6ayujg1r8lMQZJkJDXHQ9rWGldXU0TdyYg6khEFfE5tKw4AwDoXC/n15I4WPbmjRZI0ky3o7bMTeuvMhL5/Zlx/+sZZ/fvXTktyZ6J0u0SndF9vUlubY1x4AdYpAh4AgPLFskZmchqadsOci1PuMltwu1kZSel4SD1NUT2yJaLuJrfZeDjgq23FAQCAEuGAnupv1VP9rZKkXLGkDy9M6ftnxvXWmQl97YNL+vPvn5ckJaMB7e1J6r6eJt2/Kal7e5JqYCw8YF0g4AGADSRfLOvU6KyODc3q2OCMvn1kWMPTWY3P5WW9MgGfUXtDWPd0JdWRDKujMaL2hrCCflrmAABQD0J+nx70umlJUrlsdXJkVu+cm9C75yb1zrkJfffYiKyVjJG2tcS1tyepPd2N2tPVqLs6GriIA9QhAh4AWGestRqdzevUyKxOjc7p1MisTo64y/MT8yqV3SjH5xilYkF1JCPa25NUW0NYrQ0hpWMh+RyabgMAsF44jtH2toS2tyX0Uw+5kxxMZwv64PyU3jk3oXfOTeg7R4b1129fkCT5vfL3dDVWQp/+9gShD7DGEfAAQB0ql60Gp7M6O7Y4DfnZ8YzOjc3p1OicZrLFStmQ39GW5ph2dTboh+/p1Pa2uPrbE9rSHNOX3x6o4V8BAABqpSEcWDJws7VWF6ey+vDClD4cmNQHF6b04qFB/eVbbtcun2PU1xLTXR0NVbeEWhPhWv4ZAKoQ8ADAGpQvljU0ndXA5LwuereByawuTs7r/ERGF8bnlS+VK+V9jlFXMqLeVFSf2dulrS0xbW2Ja2tzTF3JiBxa5AAAgOswxv0u0ZWM6Pnd7ZLc0OfCxLw+HJjSoYvTOnxpWm+eHtdX3rtYeVxzPKid7Q3qb09oR1tc21oT2tYaV2OEcX2A1UbAAwCrqFy2Gs/kNTyd09BMViPT7sDGw94Ax0MzOV2anNfIbE7WLn1sOhZURzKsne0JPberTb2pqDalYupNRdWZDMvP7FUAAOAOMsaoJxVVTyqqT+3pqGyfzOR1+NKMDl+a1pHBaR2+NKP/98DZyuQMktTWENKONjfs2d6aUF9LTFuaY2pJhJjFC1ghBDwAcIustcrkS5qaL2gyU9DUfEFT83mNzeU1Pusux+byGpvNadxbH5/LV8bAqZaMBtSaCKk1EVZ/f4s6kxF1NrozVXV6Ax1HgvR7BwAAtZeMBvVoX1qP9qUr20plq4GJeR0fntHx4VkdH5rV8eEZ/eWb55XJlyrlYkGfNjfHtLk5pq3NMW1Ox7SlJaZNqahSsSDhD3AbCHgAbCjWWuWKZc3misrkSu4yX9Rsrqi5XElz+aLmct4tX/LWvaVXbnp+IcwpqFC6MqxZkAj7lY4F3enFU1Hd15tUKhZUayLshjkN7rIlEWLQQgAAUNd8jlFvOqredFTP3tVW2V4uWw1Mzuv06JzOjM3p1Ii7/GhgSt/4aHDJha9o0Keepqh6UhF1N0XV3RRxWxA1RdXVFFFD2E8ABFwHAQ+Am2atVdlKZWvdW1kqVda9fWW7uM27X7betrKqylqVvP3hgKNi2apUtiqWyovrl93PF8vKFcvKFUvesry4rVBSvlRWruDuX1xfKFPSVRrQXFXQ5yga8ikW9CsW8ikW8ise8quzMaLGaECNkYCSEXfZGAlUtqVjITXFAgr5CW0AAMDG5jiL3byeUMuSffliWRcmMjo9OudOHDGR0fnxeV2YyOj1k2Oaq2r5I7mtfzqSEXU0hr3bYkvn9sawWuIhJaMBQiBsWAQ8QI1Za5UvVQUUV4QUXlBRFWoUilaFclnFklXBCz4KxbIKVUFIoVR295WsCiWrolc+XypfVsZ9zEKZ8bm8ytZtZlsd2JSqApnlBiQ3a2GWhusxxp0VKuhzFAr43HW/o5DfXQ/5HcVDfqVjPoUCjkI+x136fV45R9GFwKYquImF/Iv3g+79oJ8xbQAAAFZK0O+4k0K0xK/YZ63VRKag8+Nu8ONOOpHVpal5DU5ldWRwRqNXGbMw6HPU4rWQbq0sw0rHg0rFgkpGA0rFgmqKuutckMN6QsCDDW0hXHFDlbIXqlzeKqS02AKkVLpG6HL1UCZ3nf35qv13kt8x8vuMAo6jgN+R3zEK+Bx3m++y+46jcMCRP+RXwGfkdxyF/PPyOUaOMfI5kmOMd1Nlu+N49yvr7n3HMe42Y+RUPdbnlXeqyvu88tXP95m9XfJ79fA5pvK3VN9nNigAAID1zxijVMwNZe7tSV61zMKso5emspVJK4ZnshqZyWlkJqezYxm9dXZC43P5a75OLOhTkxf4uMuAux4Nqim2uN4YCSgedi8GJkIBhQMOLYWw5hDwYNVZ63a5yVd1q8l74Um2ErKUq1q1lK5S1gtiSpdtry5/xfOUr+jak78D4cpCi5KFFiQLrURCfrcFSdDnqDESUCgRurK1SaWFyWLrk8v3B31eSxT/YiuUQFVY4/ctvX+7HzQvHDh32/8mt2pzc6xmrw0AAID6EvQ7le5f15MvljWRybu3uULVel4TmYK3zGs8U9DZsTmNz+U1ky1e9zl9jlEs6FMiHFAs5FM85Fc8HFB8YT3krYfdluHRoE+RgE+hgE9hv0/hgKNwwOfdHG+b+xuAC5q4VRsu4CmX3RYbC6yVrKy3XNhmZb198vYvlr1sv7z9lX3XKF/VdPDy17TezsXntCp5Y5QsdJNZXLpdZ6w3lslyti8ZC8Ubw2RhzJOl2xe74Sx0/ymWqrr9lGxlvVDpGrTYxcft8rO0i1DR6/ZTKV9yn/vyppS3ajFMWeyyE/QthixBv6N42H/VfQvddkJVz1EdylTvD14twPHegO9EqAIAAABgZQT9jtoawmprCC/7MYVSWZOZgiYz7iyo01l3Eo4ZbzKO2aw7+castz6XdyfiuDg5v7g/X7yl3z1Bv6Owf2kAFPL7Khd13Qu+Xst873fQwr7AZetB/+JFYd9Ci/yFVvfesnrbQgt+v2+xJX6lrHP5NlXWjdei3xgjI/ciuJHRws8kc419RpIuu395ORldc9/C8zvGMLyCNmDA8+75Cf34H7xe62qsaY6R2yqkqnWI33G79AS9rj1+x9vuvbFEg/5KF6Dq8gFnadeghedbCEquFcpcLVBZKL+wL+AjWAEAAABw5wWqxvK5VeWy1XxhYdbWkrKFhVtZ2WJJ2XzJXRbKi9sL7rZcoaz5yn63h8LCRfO5XLGyXqi6+F7wei5Uj7W5UTywqUlf/oXHal2NmjP2JiJFY8yIpLMrVx2sY82SRmtdCeAO47jGesMxjfWGYxrrDcc01iOO66U2WWtbblzsSjcV8AC3yhjzlrX2wVrXA7iTOK6x3nBMY73hmMZ6wzGN9Yjj+s6hkxoAAAAAAECdI+ABAAAAAACocwQ8WC1/WOsKACuA4xrrDcc01huOaaw3HNNYjziu7xDG4AEAAAAAAKhztOABAAAAAACocwQ8AAAAAAAAdY6AB3eUMeZ5Y8xRY8wJY8w/v0aZnzTGHDLGHDTGvLDadQRuxo2OaWPM7xpj3vNux4wxk7WoJ3AzlnFc9xpj9htj3jXGfGCM+VQt6gks1zKO6U3GmG97x/PLxpjuWtQTWA5jzB8bY4aNMR9dY78xxvwf3vH+gTHm/tWuI3CzlnFc7zTGvG6MyRljfnW167deMAYP7hhjjE/SMUnPSbog6U1Jn7PWHqoqs13SX0l6xlo7YYxptdYO16TCwA0s55i+rPw/kXSftfa/XL1aAjdnme/VfyjpXWvtHxhjdkn6O2vt5lrUF7iRZR7Tfy3pb621f2KMeUbSP7DW/v2aVBi4AWPME5JmJf2ptXb3VfZ/StI/kfQpSfsk/e/W2n2rW0vg5izjuG6VtEnSZyRNWGt/Z5WruC7Qggd30sOSTlhrT1lr85L+QtKPXlbmH0n6fWvthCQR7mCNW84xXe1zkv58VWoG3LrlHNdWUoO33ijp4irWD7hZyzmmd0n6jre+/yr7gTXDWvuKpPHrFPlRuT+SrbX2DUlJY0zH6tQOuDU3Oq6ttcPW2jclFVavVusPAQ/upC5J56vuX/C2VdshaYcx5nvGmDeMMc+vWu2Am7ecY1qS2/xf0hYt/oAA1qrlHNf/g6SfMcZckPR3cq8UA2vVco7p9yX9F976j0lKGGPSq1A3YCUs+/sJgI2FgAerzS9pu6Sn5LZ2+HfGmGRNawTcGT8t6UvW2lKtKwLcAZ+T9B+std1yuwD8mTGG7wyoZ78q6UljzLuSnpQ0IIn3awDAuuKvdQWwrgxI6qm63+1tq3ZB0gFrbUHSaWPMMbmBz5urU0XgpiznmF7w05J+ccVrBNy+5RzXX5T0vCRZa183xoQlNUuiWy3Wohse09bai/Ja8Bhj4pJ+3FrLoPioVzfz/QTABsLVONxJb0rabozZYowJyv3B+9XLyvyN3NY7MsY0y+2ydWo1KwnchOUc0zLG7JTUJOn1Va4fcCuWc1yfk/SsJBlj7pIUljSyqrUElu+Gx7QxprmqFdqvS/rjVa4jcCd9VdLPerNpPSJpylp7qdaVAlB7tODBHWOtLRpjfknSNyX5JP2xtfagMeZfSHrLWvtVb98njDGH5DaN/jVr7Vjtag1c2zKPacn9MfEXlmkJUQeWeVz/t3K70P5TuQMu/xzHN9aqZR7TT0n6V8YYK+kV0eISa5gx5s/lHrPN3lhovykpIEnW2n8jd2y0T0k6ISkj6R/UpqbA8t3ouDbGtEt6S+4kD2VjzK9I2mWtna5RlesS06QDAAAAAADUObpoAQAAAAAA1DkCHgAAAAAAgDpHwAMAAAAAAFDnCHgAAAAAAADqHAEPAAAAAABAnSPgAQAAAAAAqHMEPAAAAAAAAHXu/wdj8Lhvckj67gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1152x720 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(nrows=5, figsize=(16, 10))\n",
    "sns.distplot(y, ax=axes[0], bins=20)\n",
    "axes[0].set_title('y')\n",
    "sns.distplot(v, ax=axes[1], bins=20)\n",
    "axes[1].set_title('v')\n",
    "sns.distplot(n, ax=axes[2], bins=20)\n",
    "axes[2].set_title('n')\n",
    "sns.distplot(z, ax=axes[3], bins=20)\n",
    "axes[3].set_title('z')\n",
    "sns.distplot(p, ax=axes[4], bins=20)\n",
    "axes[4].set_title('p')\n",
    "for i in range(5):\n",
    "    axes[i].set_yticks([])\n",
    "fig.tight_layout()\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/tsalo/anaconda/envs/python3/lib/python3.6/site-packages/ipykernel/__main__.py:21: UserWarning: Matplotlib is currently using module://ipykernel.pylab.backend_inline, which is a non-GUI backend, so cannot show the figure.\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe4AAAOTCAYAAAB5NOBgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOzde3Bk2UHn+d9RVpaUUpVKJdmYdvbIbTBOHm6cCSWHveK1IrzJPDBqdmLA9oKCdUR7iZ0JFMvkooRgbdhhELQBze7MwNSGPTSgwSZoIQazTLfDbY3XLIupdqpGxu3uZqAp+tq4qyWlXpVKKaWzf0gqSp31kErn5slz9f1EVEh1lLr3Z1vWr8655+Y11loBAIAwdPgOAAAAjo7iBgAgIBQ3AAABobgBAAgIxQ0AQEAobgAAAnLGd4CjeM1rXmMfeugh3zFwRI1GQ5J05kwQP154tZ3NvY+pLr85gFPsmWeeecVa+9rbfS2I36wPPfSQrly54jsGjmhxcVGSNDAw4DkJ7svq83sfe9/sNwdwihlj/vpOX2OpHACAgAQx40ZYuru7fUcAgMSiuOFcJpPxHQEAEoulcji3s7OjnZ0d3zEAIJEobjhXrVZVrVZ9xwCARKK4AQAICMUNAEBAKG4AAAJCcQMAEBBuB4NzPT09viMAQGJR3HCuq4v3uAaAuLBUDucajcbNB40AANyiuOHcysqKVlZWfMcAgESiuAEACAjFDQBAQChuAAACQnEDCTS9MK2Hph5Sx8906KGphzS9MO07EgBHuB0Mzp07d853hFNtemFaj/7Bo7qxfUOS9Ncrf61H/+BRSdJ7H36vz2gAHDDWWt8Z7unSpUv2ypUrvmO0HfMzxncEAAlgP9D+PXDaGGOesdZeut3XWCoHACAgFDcAAAHhGnfA2nV5a3FxUZI0MDDgOcnp9Opr3JLUne7W5e+9fLRr3KvP733sfXNMCQGcBDNuIGHe+/B7dfl7L+sNF94gI6M3XHjD0UsbQNtjxg0k0Hsffi9FDSQUM24AAALCjBvOnT9/3ncEAEgsihvOnT171ncEAEgslsrh3NbWlra2tnzHAIBEorjh3NramtbW1nzHAIBEorgBAAgIxQ0AQEAobgAAAkJxAwAQEG4Hg3O9vb2+IwBAYlHccC6dTvuOAACJxVI5nKvX66rX675jAEAixT7jNsakJF2RFFlr/5Ex5o2SPippQNIzkn7IWsu7dSTI+vq6JKmzs9NzEuCw2Uqkx558Tl+q1vT6voxKxZxGC1nfsYBjacWM+8ckPXvL339B0q9Ya98kaVnS+1qQAcApN1uJVJ5ZUFStyUqKqjWVZxY0W4l8RwOOJdYZtzHmQUn/UNLPSfpfjDFG0oik9+y/5HFJH5T0q3HmANB+xsel+fnWna9yrVv1xlDT+Nh0hwqD8Z8/n5empuI/D5Iv7hn3lKT/VdLu/t8HJFWttY39v78k6bbrVMaYR40xV4wxV65fvx5zTABJV2/sHmscaFexzbiNMf9I0svW2meMMd913O+31l6WdFmSLl26ZB3HA+BZq2efw5NXFVVrTePZvozmJkZaGwY4gThn3MOS3mWMeVF7m9FGJP0rSX3GmIN/MDwoiQtMCXPhwgVduHDBdwzgkFIxp0w6dWgsk06pVMx5SgTcn9iK21pbttY+aK19SNIPSnraWvteSZ+S9I/3XzYm6ffjygA/zpw5ozNneIsAtJfRQlY///0PK9uXkdHeTPvnv/9hdpUjOD5+u/6EpI8aY/6FpIqkD3vIgBhtbm5Kkrq6ujwnAQ4bLWQpagSvJcVtrZ2TNLf/+V9Kelsrzgs/NjY2JFHcABAH3jkNAICAUNwAAASE4gYAICAUNwAAAeGeHTjX19fnOwIAJBbFDedSqdS9XwQAuC8slcO5Wq2mWq35rSUBACfHjBvO3bhxQ5KUyWQ8JwGA5GHGDQBAQChuAAACQnEDABAQihsAgICwOQ3OXbx40XcEnAKzlUiPPfmcvlSt6fV9GZWKOZ78hVOB4oZzHR0s5CBes5VI5ZkF1bZ3JElRtabyzIIkUd5IPIobzh3cDtbd3e05CXwZH5fm5+M7fuVat+qNoabxsekOFQbdny+fl6am3B8XuB9MjeAcb8CCuNUbu8caB5KEGTcA5+KenQ5PXlVUbf7HYbYvo7mJkXhPDnjGjBtAcErFnDLpw++Jn0mnVCrmPCUCWocZN4DgHGxAY1c5TiOKG0CQRgtZihqnEsUN5/r7+31HAIDEorjhnDHGdwQASCw2p8G5jY0NbWxs+I4BAIlEccO5zc1NbW5u+o4BAIlEcQMAEBCKGwCAgFDcAAAEhOIGACAg3A4G5wYGBnxHAIDEYsYNAEBAKG44t76+rvX1dd8xACCRKG44V6/XVa/XfccAgESiuAEACAib0zyarUQ8lhAAcCwUtyezlUjlmQXVtnckSVG1pvLMgiRR3gCAO6K472F8XJqfd3/cyrVu1RtDTeNj0x0qDLo9Vz4vTU25PSYAwA+K25N6Y/dY4yHhPm4AiA/FfQ9xzVSHJ68qqtaaxrN9Gc1NjMRzUgBA8NhV7kmpmFMmnTo0lkmnVCrmPCVyZ21tTWtra75jAEAiMeP25GADWhJ3lW9tbfmOAACJRXF7NFrIJqKoAQCtw1I5AAABobgBAAgIS+VwzhjjOwIAJBbFDef6+/t9RwCAxGKpHACAgFDccG51dVWrq6u+YwBAIrFUDue2t7d9RwCAxGLGDQBAQChuAAACQnEDABAQrnHDuY4O/j0IAHGhuOHcxYsXfUcAgMRiagQAQECYccO5lZUVSdKFCxe85pitRIl8bCqA043ihnONRsN3BM1WIpVnFlTb3pEkRdWayjMLkkR5AwgaxQ2vxsel+Xn3x61c61a9MdQ0PjbdocKgu/Pk89LUlLvjAcC9cI0biVRv7B5rHABCwYwbXsU1Wx2evKqoWmsaz/ZlNDcxEs9JAaAFmHHDuVQqpVQq5TVDqZhTJn04QyadUqmY85QIANxgxg3n+vr6fEe4uQGNXeUAkobiRmKNFrIUNYDEYakczlWrVVWrVd8xACCRmHHDuZ2dHd8RACCxmHEDABAQihsAgIBQ3AAABIRr3HDuzBl+rAAgLvyGhXO+nwoGAEnGUjkAAAGhuOHc8vKylpeXfccAgERiqRzO7e7yBC4AiAszbgAAAkJxAwAQEJbKAaCNzVYinnKHQyhuOJdOp31HABJhthKpPLOg2vbe+/9H1ZrKMwuSRHmfYhQ3nOvt7fUdAYjd+Lg0Px/vOSrXulVvDDWNj013qDAYzznzeWlqKp5jww2ucQNAm6o3bn+Hxp3GcTow44ZzS0tLkqT+/n7PSYD4tGJWOjx5VVG11jSe7ctobmIk/gBoS8y44Zy1VtZa3zGA4JWKOWXSqUNjmXRKpWLOUyK0A2bcANCmDjagsasct6K4AaCNjRayFDUOYakcAICAMOOGc2fPnvUdAQASi+KGc+fPn/cdAQASi6VyAAACQnHDucXFRS0uLvqOAQCJRHEDABAQihsAgIBQ3AAABITiBgAgINwOBuc6Ozt9RwCAxKK44dy5c+d8RwCAxGKpHACAgMRW3MaYLmPMZ40xV40xf26M+Zn98V83xvyVMWZ+/08+rgzwg/u4ASA+cS6V1yWNWGvXjTFpSZ8xxvzR/tdK1trfjfHcQOLNViIe9wicQrEVt7XWSlrf/2t6/4+N63zAaTJbiVSeWVBte0eSFFVrKs8sSBLlDSRcrJvTjDEpSc9IepOkf2Ot/VNjzI9K+jljzP8m6ZOSJqy19ThzAO1ifFyanz/5cSrXulVvDDWNj013qDB4smM/eEH6hm+QfmryZMcBEI9YN6dZa3estXlJD0p6mzHmLZLKkr5e0pCkfkk/cbvvNcY8aoy5Yoy5cv369ThjAsGpN3aPNQ4gOVpyO5i1tmqM+ZSk77HWfmh/uG6M+feS/vkdvueypMuSdOnSJZbYA9LV1eU7QtuamnJznOHJq4qqtabxbF9GcxMjJzv46sm+HUC84txV/lpjTN/+5xlJ75T0RWPMA/tjRtKopM/HlQF+9PT0qKenx3eMRCsVc8qkU4fGMumUSsWcp0QAWiXOGfcDkh7fv87dIel3rLUfN8Y8bYx5rSQjaV7S/xRjBniwty9R2vu3GeJwsAGNXeXA6RPnrvL/Iqlwm/ETruOh3S0tLUmSBgYGPCdJttFClqIGTiHeOQ0AgIBQ3AAABITiBgAgIBQ3AAAB4bGecC6TyfiOAACJRXHDue7ubt8RACCxWCqHc7u7u9rd5a03ASAOFDecW15e1vLysu8YAJBIFDcAAAGhuAEACAjFDQBAQChuAAACwu1gcI7bwQAgPhQ3nOMNWID2M1uJeAxsQlDccG5nZ0eSlEqlPCcBIO2VdnlmQbXtvf9vRtWayjMLkkR5B4jihnPValUSz+MG7mV8XJqfj/88lWvdqjeGmsbHpjtUGIz//Pm8NDUV/3lOCzanAUDC1Ru3fyfDO42jvTHjBgBPWjULHZ68qqhaaxrP9mU0NzHSmhBwhhk3ACRcqZhTJn14z0kmnVKpmPOUCCfBjBsAEu5gAxq7ypOB4oZzPT09viMAeJXRQpaiTgiKG851dXX5jgAAicU1bjjXaDTUaDR8xwCARKK44dzKyopWVlZ8xwCARKK4AQAICMUNAEBAKG4AAAJCcQMAEBBuB4Nz586d8x0BABKL4oZznZ2dviMAQGKxVA7ntre3tb297TsGACQSxQ3nVldXtbq66jsGACQSxQ0AQEC4xg3cp9lKxNOWALQcxQ3ch9lKpPLMgmrbO5KkqFpTeWZBkihvALGiuBGc8XFpft5vhsq1btUbQ03jY9MdKgx6CHSLfF6amvKbAUB8KG44d/78ed8RYldv7B5rHABcobjh3NmzZ2M9fjvMJocnryqq1prGs30ZzU2MeEgE4LRgVzmc29ra0tbWlu8YsSoVc8qkU4fGMumUSsWcp0QATgtm3HBubW1NkjQwMOA5SXwONqCxqxxAq1HcwH0aLWQpagAtx1I5AAABobgBAAgIxQ0AQEC4xg3nent7fUcAgMSiuOFcOp32HQEAEoulcjhXr9dVr9d9xwCARGLGDefW19clSZ2dnZ6TAEDyMOMGACAgFDcAAAGhuAEACAjFDQBAQNicBucuXLjgOwIAJBbFDefOnOHHCgDiwm9YOPfEZ/9K/8cnX9C1lW0edwkAjlHccGq2EukX/mBem41dWXUpqtZUnlmQJMobAByguE+R8XFpfj7ec1Sudaun+2FJ0tLqxZvjY9MdKgzGc858XpqaiufYANBu2FUOp+qN3WONAwCOhxn3KdKKWenw5FWtrSxLks7arpvj2b6M5iZG4g8AAAnHjBtOlYo5dZ05/GOVSadUKuY8JQKAZGHGDadGC1nt7n6rfvmp57W2UmdXOQA4RnHDue//1kF9/7fGtBMNAE45lsrhXK1WU61W8x0DABKJGTecu3HjhiQpk8l4TgIAycOMGwCAgFDcAAAEhOIGACAgFDcAAAFhcxqcu3jx4r1fBAC4LxQ3nOvoYCEHAOLCb1g4d+PGjZu3hAEA3KK44RxvwAIA8aG4AQAICMUNAEBAKG4AAALCrnLAk9lKpMeefE5fqtZ4/CmAI6O44Vx/f7/vCG1vthKpPLOg2vaOJCmq1lSeWZAkyhvAXVHccM4Y4zvCiYyPS/Pz8Z6jcq1b9cZQ0/jYdIcKMT7KPJ+XpqbiOz6A+HGNG85tbGxoY2PDd4y2Vm/sHmscAA4w44Zzm5ubkqSenh7PSe5PK2akw5NXFVWb73XP9mU0NzESfwAAwWLGDXhQKuaUSacOjWXSKZWKOU+JAISCGTfgwcEGNHaVAzguihvwZLSQpagBHBtL5QAABIQZN5wbGBjwHQEAEosZNwAAAaG44dz6+rrW19d9xwCARKK44Vy9Xle9XvcdAwASieIGACAgFDcAAAGhuAEACAjFDQBAQLiPG85xHzcAxIcZNwAAAaG44dza2prW1tZ8xwCARIqtuI0xXcaYzxpjrhpj/twY8zP74280xvypMeYvjDEfM8acjSsD/Nja2tLW1pbvGACQSHHOuOuSRqy1b5WUl/Q9xpi3S/oFSb9irX2TpGVJ74sxAwA4N1uJNDz5tN448Ycannxas5XIdyScIrEVt91z8L6X6f0/VtKIpN/dH39c0mhcGQDAtdlKpPLMgqJqTVZSVK2pPLNAeaNlYt1VboxJSXpG0psk/RtJ/1VS1Vrb2H/JS5J4IDGAWIyPS/Pzbo9ZudatemOoaXxsukOFQTfnyOelqSk3x0LyxLo5zVq7Y63NS3pQ0tskff1Rv9cY86gx5oox5sr169djywj3jDEyxviOAcSi3tg91jjgWkvu47bWVo0xn5L0Dkl9xpgz+7PuByXddn3JWntZ0mVJunTpkm1FTrjR39/vOwIgKZ5Z6/DkVUXVWtN4ti+juYkR9ycEXiXOXeWvNcb07X+ekfROSc9K+pSkf7z/sjFJvx9XBgBwrVTMKZNOHRrLpFMqFXOeEuG0iXPG/YCkx/evc3dI+h1r7ceNMV+Q9FFjzL+QVJH04RgzwIPV1VVJUm9vr+ckgHujhb1tOY89+Zy+VK3p9X0ZlYq5m+NA3GIrbmvtf5FUuM34X2rvejcSant723cEIFajhSxFDW945zQAAAJCcQMAEBCKGwCAgPBYTzjX0cG/BwEgLhQ3nLt48aLvCACQWEyNAAAICMUN51ZWVrSysuI7BgAkEkvlcK7RaNz7RQCA+8KMGwCAgFDcAAAEhOIGACAgXOOGc6lU6t4vAgDcF4obzvX19fmOEITZSsQTpgAcG8UNeDBbiVSeWVBte0eSFFVrKs8sSBLlDeCuKG44V61WJSVn5j0+Ls3Puz1m5Vq36o2hpvGx6Q4VBt2eK5+XpqbcHhOAPxQ3nNvZ2fEdoe3VG7vHGgeAAxQ3cA9xzFaHJ68qqtaaxrN9Gc1NjLg/IYDE4HYwwINSMadM+vDu+0w6pVIx5ykRgFAw4wY8ONiAxq5yAMdFccO5M2f4sTqK0UKWogZwbPyGhXMXLlzwHQEAEotr3AAABITihnPLy8taXl72HQMAEomlcji3u8u9yAAQF2bcAAAEhOIGACAgFDcAAAHhGjecS6fTviMAQGJR3HCut7fXdwQASCyWygEACAjFDeeWlpa0tLTkOwYAJBJL5XDOWus7AgAkFjNuAAACwowbaHOzlYjHfwK4ieIG2thsJVJ5ZkG17R1JUlStqTyzIEmUN3BKUdxw7uzZs74jtJ3xcWl+/vjfV7nWrXpjqGl8bLpDhcHjHy+fl6amjv99ANoHxQ3nzp8/7ztCYtQbt39gy53GASQfxQ20wP3Ococnryqq1prGs30ZzU2MnDAVgBCxqxzOLS4uanFx0XeMRCgVc8qkU4fGMumUSsWcp0QAfGPGDbSxgw1o7CoHcIDiBtrcaCFLUQO4iaVyAAACQnEDABAQlsrhXGdnp+8IAJBYFDecO3funO8IAJBYLJUDABAQihvOcR83AMSH4gYAICAUNwAAAaG4AQAICMUNAEBAuB0MznV1dfmOAACJRXHDuZ6eHt8RACCxWCqHc9ZaWWt9xwCARGLGDeeWlpYkSQMDA7Gfa7YS8chLAKcKxY1gzVYilWcWVNvekSRF1ZrKMwuSRHkDSCyKG7EZH5fm5+M7fuVat+qNoabxsekOFQbdny+fl6am3B8XAI6Da9wIVr2xe6xxAEgCZtyITdyz0+HJq4qqtabxbF9GcxMj8Z4cADxhxg3nMpmMMplM7OcpFXPKpFOHz51OqVTMxX5uAPCFGTec6+7ubsl5DjagsascwGlCccO53d29a8wdHfEv6IwWshQ1gFOFpXI4t7y8rOXlZd8xACCRKG4AAAJCcQMAEBCKGwCAgFDcAAAEhF3lcK5Vt4MBwGlEccO5Vrz5CgCcViyVw7mdnR3t7Oz4jgEAiURxw7lqtapqteo7BgAkEsUNAEBAKG4AAAJCcQMAEBB2lQNo8okvfEUfnHuJp64BbYjihnM9PT2+I+AEPvGFr+iXP/Gcoo0HJElRtabyzIIkUd5AG6C44VxXV5fvCIk1Pi7Nz8d7jtWNLm3tPKy/feVNh8bHpjtUGIznnPm8NDUVz7GBpOEaN5xrNBpqNBq+Y+A+be3s3na83rj9OIDWYsYN51ZWViRJAwMDnpMkTytmpT/4oS/qK6t1rW+9cmg825fR3MRI/AEA3BUzbgCHvO/bvkZd6cO/GjLplErFnKdEAG7FjBvAIe/8xtdJkj44t82ucqANUdwAmrzzG1+nd779zb5jALgNlsoBAAgIM244d+7cOd8RACCxKG4419nZ6TsCACQWS+Vwbnt7W9vb275jAEAiUdxwbnV1Vaurq75jAEAiUdwAAASE4gYAICAUNwAAAaG4AQAICLeDwbnz58/7jgAAiUVxw7mzZ8/6jgAAicVSOZzb2trS1taW7xgAkEixFbcx5u8ZYz5ljPmCMebPjTE/tj/+QWNMZIyZ3//zD+LKAD/W1ta0trbmOwYAJFKcS+UNST9urf2cMea8pGeMMZ/Y/9qvWGs/FOO5AeDIZiuRHnvyOR5jiiDEVtzW2i9L+vL+52vGmGcl8f8EAG1lthKpPLOg2vaOJCmq1lSeWZAkyhttqSWb04wxD0kqSPpTScOS/qkx5oclXdHerHy5FTkAtLfxcWl+vrXnrFzrVr0x1DQ+Nt2hwmDrcuTz0tRU686HcMW+Oc0Yc07SE5LGrbWrkn5V0tdKymtvRv5Ld/i+R40xV4wxV65fvx53TACnVL2xe6xxwDdjrY3v4MakJX1c0pPW2l++zdcfkvRxa+1b7nacS5cu2StXrsSSEe4dPBksnU57ToL7svr83sfeN/vN0SLDk08rqtaaxrN9Gf3xxIiHRIBkjHnGWnvpdl+Lc1e5kfRhSc/eWtrGmAduedkjkj4fVwb4kU6nKW0Eo1TMKZNOHRrLpFMqFXOeEgF3F+c17mFJPyRpwRhzcNXqJyW92xiTl2QlvSjp/TFmgAf1el2S1NnZ6TkJcG8HG9DYVY5QxLmr/DOSzG2+9H/HdU60h/X1dUkUN8IxWshS1AgG75wGAEBAKG4AAAJCcQMAEBCKGwCAgPBYTzh34cIF3xEAILEobjh35gw/VgAQF5bK4dzm5qY2Nzd9xwCARGJqBOc2NjYkSV1dXZ6TAEDyMOMGACAgFDcAAAGhuAEACAjFDQBAQNicBuf6+vp8RwCAxKK44Vwqlbr3iwBHZisRj+TEqUJxw7larSZJymQynpMg6WYrkcozC6pt70iSompN5ZkFSaK8kVgUN5y7ceOGJIobe8bHpfn5eI5dudatemOoaXxsukOFwXjOmc9LU1PxHBs4CjanAQhWvbF7rHEgCZhxA4hVnLPT4cmriqq1pvFsX0ZzEyPxnRjwiBk3gGCVijll0oc3Q2bSKZWKOU+JgPgx4wYQrIMNaOwqx2lCccO5ixcv+o6AU2S0kKWocapQ3HCuo4MrMAAQF37DwrkbN27cvCUMAOAWxQ3narXazTdhAQC4RXEDABAQihsAgIBQ3AAABITiBgAgINwOBuf6+/t9RwCAxKK44ZwxxncEAEgslsrh3MbGhjY2NnzHAIBEorjh3ObmpjY3N33HAIBEorgBAAgIxQ0AQEDYnHYPs5WIRwYCANoGxX0Xs5VI5ZkF1bZ3JElRtabyzIIkUd4AAC9OXXGPj0vz80d7beVat+qNoabxsekOFQbv/f35vDQ1dcyACTAwMOA7AgAkFte476Le2D3WOAAAcTt1M+7jzICHJ68qqjY/njLbl9HcxIjDVMmyvr4uSTp37pznJACQPMy476JUzCmTTh0ay6RTKhVznhKFoV6vq16v+44BAIl06mbcx3GwAY1d5QCAdkFx38NoIUtRAwDaBkvlAAAEhOIGACAgLJXDOe7jBoD4MOMGACAgFDecW1tb09ramu8YAJBILJXDua2tLd8RACCxmHEDABAQihsAgIBQ3AAABIRr3HDOGOM7AgAkFsUN5/r7+31HAIDEYqkcAICAMOOGc6urq5Kk3t5ez0mAO5utRDz5D0GiuOHc9va27wjAXc1WIpVnFlTb3pEkRdWayjMLkkR5o+1R3ACcGx+X5ud9p7izyrVu1RtDTeNj0x0qDHoIdAT5vDQ15TsF2gHXuAGcOvXG7rHGgXbCjBuAc+0+MxyevKqoWmsaz/ZlNDcx4iERcHTMuOFcR0eHOjr40UL7KhVzyqRTh8Yy6ZRKxZynRMDRMeOGcxcvXvQdAbirgw1o7CpHiChuAKfSaCFLUSNIrGfCuZWVFa2srPiOAQCJxIwbzjUaDd8RACCxmHEDABCQIxW3MeZ7jTGUPAAAnh21jH9A0gvGmF80xnx9nIEAAMCdHam4rbX/g6SCpP8q6deNMX9ijHnUGHM+1nQIUiqVUiqVuvcLAQDHduTlb2vtqqTflfRRSQ9IekTS54wx/yymbAhUX1+f+vr6fMcAgEQ66jXudxljfk/SnKS0pLdZa/++pLdK+vH44gEAgFsd9Xaw/17Sr1hrP33roLX2hjHmfe5jIWTValWSmHUDQAyOVNzW2rG7fO2T7uIgCXZ2dnxHAIDEumtxG2PWJNk7fd1a2+s8EQAAuKO7Fre19rwkGWP+d0lflvSbkoyk92pvgxoAAGiho+4qf5e19t9aa9estavW2l+V9H1xBgMAAM2OujltwxjzXu3dCmYlvVvSRmypELQzZ3gLfNzZbCXicZrACRz1N+x7JP2r/T9W0h/vjwFNLly44DsC2tRsJVJ5ZkG17b0NjFG1pvLMgiRR3sARHXVX+YtiaRw4NX7uX0qf+P/cH7dyrVv1xlDT+Nh0hwqD7s6Tz0tTU+6OB7STIxW3Mebf6za7y621/6PzRAje8vKyJOnixYuek6Dd1Bu7xxoH0OyoS+Ufv+XzLu293emX3MdBEuzu8ks4dD/1k9JPxXCz5/DkVUXVWtN4ti+juYkR9ycEEuioS+VP3Pp3Y8xvS/pMLIkAJFapmDt0jVuSMumUSsWcx1RAWO53++/XSfoql0EAJN/BBjR2lQP376jXuF/9Dmp/K+knYkkEINFGC1JbT4gAACAASURBVFmKGjiBoy6V89xtHFk6nfYdAQAS66iP9Wx6kMjtxgBJ6u3tVW8vb2MPAHG410NGuiR1S3qNMeai9t6nXJJ6JbHWBQBAi91rqfz9ksYlvV7SM7eMr0n613GFQtiWlpYkSf39/Z6TAEDy3Gup/P+V9N9I+ufW2q+R9DOSPi/pP0v6DzFnQ6CstbL2jk+DBQCcwL2K+99Jqltr/09jzHdI+nlJj0takXQ57nAAAOCwey2Vp6y1S/uf/4Cky/tvxvKEMWY+3mgAAODV7jXjThljDsr9uyU9fcvXeHYjAAAtdq/y/W1J/9kY84qkmqT/R5KMMW/S3nI50OTs2bO+IwBAYt21uK21P7d/v/YDkp6yf7fjqEPSP4s7HMJ0/jzv1wMAcbnncre1tumpvNba5+OJAwAA7uZI75wGHMfi4qIWFxd9xwCARIqtuI0xf88Y8yljzBeMMX9ujPmx/fF+Y8wnjDEv7H+8GFcGAACSJs4Zd0PSj1trv1HS2yX9z8aYb5Q0IemT1tqvk/TJ/b8DOKbZSqThyaf1xok/1PDk05qtRL4jAWiB2IrbWvtla+3n9j9fk/Ss9t7f/Pu09yYu2v84GlcGIKlmK5HKMwuKqjVZSVG1pvLMAuUNnAItuRfbGPOQpIKkP5X0Omvtl/e/9LeSXteKDIBL4+PSvMe3IKpc61a9MdQ0PjbdocLgyY794IW9jy8d44bPfF6amjrZeQEcTeyb04wx5yQ9IWncWrt669f2by+77ZtaG2MeNcZcMcZcuX79etwx4VBnZ6c6Ozt9x0i0emP3WOMAkiPWGbcxJq290p621s7sD3/FGPOAtfbLxpgHJL18u++11l7W/vuhX7p0iSdWBOTcuXO+I8TO9+xyePKqomqtaTzbl9HcxMjJDn7wz2seqQ60pTh3lRtJH5b0rLX2l2/50n+UNLb/+Zik348rA5BUpWJOmXTq0FgmnVKpmPOUCECrxDnjHpb0Q5IWbnkgyU9KmpT0O8aY90n6a0n/JMYM8ODgHu6BgQHPSZJrtJCVJD325HP6UrWm1/dlVCrmbo4DSK7Yitta+xlJ5g5f/u64zgucFqOFLEUNnEK8cxoAAAGhuAEACAjFDQBAQFryBiw4Xbq6unxHAIDEorjhXE9Pj+8IAJBYLJXDOWut9t4UDwDgGsUN55aWlrS0tOQ7BgAkEsUNAEBAKG4AAAJCcQMAEBCKGwCAgHA7GJzLZDK+IwBAYlHccK67u9t3BBzRbCVqfsLY1/pOBeBuWCqHc7u7u9rd3fUdA/cwW4lUnllQVK3JSoqqNZVnFvSJL3zFdzQAd8GMG84tLy9L4nncroyPS/Pz937dcVWudaveGGoa/4n/9Jd6/PXSSytuz5fPS1NTbo8JnEbMuIFTqt64/arI1g6rJUA7Y8YNtLm4ZqnDk1cVVWtN4295zSv6rUffIfXGc14AJ8OMGzilSsWcMunUobFMOqX3fdvXeEoE4CiYcQOn1GghK0lNu8rf+bUbnpMBuBuKG85xO1g4RgvZmwV+0+rzfsIAOBKKG87xBiwAEB+uccO5nZ0d7ezs+I4BAIlEccO5arWqarXqOwYAJBLFDQBAQChuAAACQnEDABAQihsAgIBwOxic6+np8R0BABKL4oZzXV1dviMAQGKxVA7nGo2GGo2G7xgAkEgUN5xbWVnRyorjhzkDACRR3AAABIXiBgAgIGxOA3DTbCXSRz/5J3p5ta5690sqFXPNTw8D4BXFDUDSXmmXZxb01aYuSYqqNZVnFiSJ8gbaCMUN586dO+c7QvDGx6X5+daes3KtW/XGkFKvuShJ+ttX3iRJGpvuUGGwdTnyeWlqqnXnA0JDccO5zs5O3xFwH+qN3WONA/CD4oZz29vbkqR0Ou05Sbh8zDiHJ68qqtb02rORJGl96xVJUrYvo7mJkdYHAnBb7CqHc6urq1pdXfUdA8dUKuaUSacOjWXSKZWKOU+JANwOM24Akv5uA9pHP/kVvbxaV7Yvw65yoA1R3ABuGi1kNfq179j7S++b/YYBcFsslQMAEBCKGwCAgLBUDufOnz/vOwIAJBbFDefOnj3rOwIAJBZL5XBua2tLW1tbvmMAQCJR3HBubW1Na2trvmMAQCJR3AAABITiBgAgIBQ3AAABobgBAAgIt4PBud7eXt8RACCxKG44x+M8ASA+LJXDuXq9rnq97jsGACQSM244t76+Lknq7Oz0nARJN1uJ9NiTz+lL1Zpez2NIcUpQ3ACCNFuJVJ5ZUG17R5IUVWsqzyxIEuWNRKO4ATgzPi7Nz7fmXJVr3ao3hprGx6Y7VBiM//z5vDQ1Ff95gFfjGjeAINUbu8caB5KCGTcAZ1o5Ax2evKqoWmsaz/ZlNDcx0rogQIsx44ZzFy5c0IULF3zHQMKVijll0qlDY5l0SqVizlMioDWYccO5M2f4sUL8Djagsascpw2/YeHc5uamJKmrq8tzEiTdaCFLUePUobjh3MbGhiSKGwDiwDVuAAACQnEDABAQihsAgIBQ3AAABITNaXCur6/PdwQASCyKG86lUql7vwgAcF9YKodztVpNtVrzW1ECAE6OGTecu3HjhiQpk8l4TgIAycOMGwCAgFDcAAAEhOIGACAgXOMGgDYyW4l44hnuiuKGcxcvXvQdAQjSbCVSeWZBte0dSVJUrak8syBJlDduorjhXEcHV2CQLOPj0vx8/OepXOtWvTHUND423aHCYPznz+elqan4z4OT4TcsnLtx48bNW8IAHF29sXuscZxOzLjh3MGbr3R3d3tOArjRqlno8ORVRdXmNy/K9mU0NzHSmhBoe8y4AaBNlIo5ZdKH3zI4k06pVMx5SoR2xIwbANrEwQY0dpXjbihuAGgjo4UsRY27YqkcAICAMOOGc/39/b4jAEBiUdxwzhjjOwIAJBZL5XBuY2NDGxsbvmMAQCJR3HBuc3NTm5ubvmMAQCJR3AAABITiBgAgIBQ3AAABobgBAAgIt4PBuYGBAd8RACCxmHEDABAQihvOra+va3193XcMAEik2IrbGPMRY8zLxpjP3zL2QWNMZIyZ3//zD+I6P/yp1+uq1+u+YwBAIsV5jfvXJf1rSb/xqvFfsdZ+KMbzAvdlthLxOEUAbS+24rbWftoY81Bcxwdcmq1EKs8sqLa9I0mKqjWVZxYkifIG0FZ87Cr/p8aYH5Z0RdKPW2uXPWRAGxkfl+bn/WaoXOtWvTHUND423aHCoIdA+/J5aWrK3/kBtJ9Wb077VUlfKykv6cuSfulOLzTGPGqMuWKMuXL9+vVW5cMpVW/sHmscAHxp6YzbWvuVg8+NMf+XpI/f5bWXJV2WpEuXLtn408GV497H3Q4zyuHJq4qqtabxbF9GcxMjHhIBwO21dMZtjHnglr8+Iunzd3ot0EqlYk6ZdOrQWCadUqmY85QIAG4vthm3Mea3JX2XpNcYY16S9AFJ32WMyUuykl6U9P64zg9/1tbWJEnnz5/3nOToDjagsascQLuLc1f5u28z/OG4zof2sbW15TvCfRktZClqAG2Pd04DACAgFDcAAAGhuAEACAiP9YRzxhjfEQAgsShuONff3+87AgAkFkvlAAAEhOKGc6urq1pdXfUdAwASiaVyOLe9ve07AgAkFjNuAAACQnEDABAQihsAgIBwjRvOdXTw70EAiAvFDecuXrzoOwIAJBbFDQBHMFuJeOwr2gLFDedWVlYkSRcuXPCcBHBjthKpPLOg2vaOJCmq1lSeWZAkyhstR3HDuUaj4TsC0GR8XJqfv7/vrVzrVr0x1DQ+Nt2hwuDxj5fPS1NT95cFYBcRANxDvbF7rHEgTsy4AZwKJ5nhDk9eVVStNY1n+zKamxg5QSrg+JhxA8A9lIo5ZdKpQ2OZdEqlYs5TIpxmzLjhXCqVuveLgIAcbEBjVznaAcUN5/r6+nxHAJwbLWQparQFlsoBAAgIxQ3nqtWqqtWq7xgAkEgslcO5nZ0d3xEAILGYcQMAEBCKGwCAgFDcAAAEhGvccO7MGX6sACAu/IaFczwVDADiw1I5AAABobjh3PLyspaXl33HAIBEYqkczu3u8qhDAIgLM24AAAJCcQMAEBCWygEEa7YS8ahNnDoUN5xLp9O+I+AUmK1EKs8sqLa99974UbWm8syCJFHeSDSKG8719vb6joA2Mz4uzc+7PWblWrfqjaGm8bHpDhUG3Z0nn5emptwdDzgprnEDCFK9cfu7F+40DiQFM244t7S0JEnq7+/3nATtIo4Z6/DkVUXVWtN4ti+juYkR9ycE2gQzbjhnrZW11ncMJFypmFMmnTo0lkmnVCrmPCUCWoMZN4AgHWxAY1c5ThuKG0CwRgtZihqnDkvlAAAEhBk3nDt79qzvCACQWBQ3nDt//rzvCACQWCyVAwAQEIobzi0uLmpxcdF3DABIJIobAICAUNwAAASE4gYAICAUNwAAAeF2MDjX2dnpOwIAJBbFDefOnTvnOwIAJBZL5QAABIQZN5w7uId7YGDAcxLgeGYrEU8bQ9ujuAFAe6VdnllQbXtHkhRVayrPLEgS5Y22QnEDSITxcWl+/v6/v3KtW/XGUNP42HSHCoP3d8x8Xpqauv9MwO1wjRsAJNUbu8caB3xhxg0gEU46sx2evKqoWmsaz/ZlNDcxcrKDAw4x44ZzXV1d6urq8h0DOJZSMadMOnVoLJNOqVTMeUoE3B4zbjjX09PjOwJwbAcb0NhVjnZHccM5a60kyRjjOQlwPKOFLEWNtsdSOZxbWlrS0tKS7xgAkEgUNwAAAaG4AQAICMUNAEBAKG4AAALCrnI4l8lkfEcAgMSiuOFcd3e37wgAkFgslcO53d1d7e7y/s4AEAeKG84tLy9reXnZdwwASCSKGwCAgFDcAAAEhOIGACAgFDcAAAHhdjA41063g81WIh7TCCBRKG441y5vwDJbiVSeWVBte0eSFFVrKs8sSBLlDSBYFDec29nZK8pUKnXP146PS/Pz8eSoXOtWvTHUND423aHCoPvz5fPS1JT74wLArbjGDeeq1aqq1arvGKo3bv8mMHcaB4AQMOOGV3HOUIcnryqq1prGs30ZzU2MxHdiAIgRM24kVqmYUyZ9eLk+k06pVMx5SgQAJ8eMG4l1sAGNXeUAkoTiRqKNFrIUNYBEobjhXE9Pj+8IAJBYFDec6+rq8h0BABKLzWlwrtFoqNFo+I4BAIlEccO5lZUVrays+I4BAIlEcQMAEBCKGwCAgFDcAAAEhOIGACAg3A4G586dO+c7AgAkFsUN5zo7O31HAIDEYqkczm1vb2t7e9t3DABIJIobzq2urmp1ddV3DABIpNiK2xjzEWPMy8aYz98y1m+M+YQx5oX9jxfjOj8AAEkU54z71yV9z6vGJiR90lr7dZI+uf93AEis2Uqk4cmn9caJP9Tw5NOarUS+IyFwsRW3tfbTkpZeNfx9kh7f//xxSaNxnR8AfJutRCrPLCiq1mQlRdWayjMLlDdOpNW7yl9nrf3y/ud/K+l1LT4/gFNifFyan/eboXKtW/XGUNP42HSHCoMeAu3L56WpKX/nx8l4ux3MWmuNMfZOXzfGPCrpUUkaHPT4E45jO3/+vO8IQFuoN3aPNQ4cRauL+yvGmAestV82xjwg6eU7vdBae1nSZUm6dOnSHQse7efs2bO+IwBtMaMcnryqqFprGs/2ZTQ3MeIhEZKg1beD/UdJY/ufj0n6/RafHy2wtbWlra0t3zEA70rFnDLp1KGxTDqlUjHnKRGSILYZtzHmtyV9l6TXGGNekvQBSZOSfscY8z5Jfy3pn8R1fviztrYmSRoYGPCcBPBrtJCVJD325HP6UrWm1/dlVCrmbo4D9yO24rbWvvsOX/ruuM4JAO1mtJClqOEU75wGAEBAKG4AAAJCcQMAEBAe6wnnent7fUcAgMSiuOFcOp32HQEAEoulcjhXr9dVr9d9xwCARGLGDefW19clSZ2dnZ6TAEDyMOMGACAgFDcAAAGhuAEACAjFDQBAQNicBucuXLjgOwIAJBbFDefOnOHHql3MViKeTAUkDL9h4dzm5qYkqaury3OS0222Eqk8s6Da9o4kKarWVJ5ZkCTKGwgYxQ3nNjY2JCWruMfHpfl53ymOp3KtW/XGUNP42HSHCoN3/r4H9690vLQSU7BjyuelqSnfKYD2weY0IKHqjd1jjQMIAzNu4AhCnPENT15VVK01jWf7MpqbGLnzN67uf+RZMUBbYsYNJFSpmFMmnTo0lkmnVCrmPCUC4AIzbiChDjagsascSBaKG8719fX5joB9o4UsRQ0kDMUN51Kp1L1fBAC4L1zjhnO1Wk21WvOmKADAyTHjhnM3btyQJGUyGc9JACB5mHEDABAQihsAgIBQ3AAABITiBgAgIGxOg3MXL170HQEAEovihnMdHSzkAEBc+A0L527cuHHzljAAgFsUN5zjDVgAID4UNwAAAaG4AQAICJvT0NZmKxGPpQSAW1DcaFuzlUjlmQXVtnckSVG1pvLMgiRR3gBOLYobxzI+Ls3P3+tV/U7OVbnWrXpjqGl8bLpDhUEnp7ijfF6amor3HABwPyhuxMA4OUq9sXuscQA4DShuHMtRZqEbGxuSpJ6enhOda3jyqqJq821l2b6M5iZGTnRsAAgVu8rh3ObmpjY3N098nFIxp0w6dWgsk06pVMyd+NgAECpm3GhbBxvQ2FUOAH+H4kZbGy1kKWoAuAVL5QAABITiBgAgICyVw7mBgQHfEQAgsZhxAwAQEIobzq2vr2t9fd13DABIJIobztXrddXrdd8xACCRKG4AAAJCcQMAEBCKGwCAgFDcAAAEhPu44Rz3cQNAfJhxAwAQEIobzq2trWltbc13DABIJJbK4dzW1lZLzzdbiXj0J4BTg+JG0GYrkcozC6pt70iSompN5ZkFSaK8ASQSxY1Dxsel+fmTHSOT2ftYq508z71UrnWr3hhqGh+b7lBhMP7zu5DPS1NTvlMACAXXuBG0emP3WOMAEDpm3DjExcxvaclIkvr7T36sexmevKqo2jy1z/ZlNDcxEn8AAGgxZtxwrr+/X/2taG1JpWJOmXTq0FgmnVKpmGvJ+QGg1ZhxI2gHG9DYVQ7gtKC44dzq6qokqbe3tyXnGy1kKWoApwbFDee2t7d9RwCAxOIaNwAAAaG4AQAICMUNAEBAuMYN5zo6+PcgAMSF4oZzFy9e9B0BABKLqREAAAGhuOHcysqKVlZWfMcAgERiqRzONRoN3xEAILGYcQMAEBCKGwCAgFDcAAAEhGvccC6VSt37RQCA+0Jxw7m+vj7fEQDgkNlKlJjH/1LcAIBEm61EKs8sqLa9I0mKqjWVZxYkKcjyprjhXLValcTMG8D9Gx+X5ufdHKtyrVv1xlDT+Nh0hwqDbs6Rz0tTU26OdS8UN5zb2dnxHQEAbqo3do813u4obgBA23E5ex2evKqoWmsaz/ZlNDcx4u5ELcLtYACARCsVc8qkD9/tkkmnVCrmPCU6GWbcAIBEO9iAxq5y4A7OnOHHCkB7GS1kgy3qV+M3LJy7cOGC7wgAkFhc4wYAICAUN5xbXl7W8vKy7xgAkEgslcO53d0w740EgBAw4wYAICAUNwAAAaG4AQAICNe44Vw6nfYdAQASi+KGc729vb4jAEBisVQOAEBAKG44t7S0pKWlJd8xACCRvCyVG2NelLQmaUdSw1p7yUcOxMNa6zsCACSWz2vc/6219hWP5weCNVuJEvOkIwDHw+Y0IDCzlUjlmQXVtnckSVG1pvLMgiRR3sAp4Ku4raSnjDFW0r+z1l72lAOI1fi4ND/v9piVa92qN4aaxsemO1QYPPnxH9x/uNtr3ihNTZ38eADc8lXc32atjYwxXyXpE8aYL1prP33rC4wxj0p6VJIGBx38NkLLnD171neERKs3bv9e8HcaB5AsxvdGImPMByWtW2s/dKfXXLp0yV65cqV1oYA2Njz5tKJqrWk825fRH0+MnPwEq8/vfex988mPBeC+GGOeudPG7ZbfDmaM6THGnD/4XNJ/J+nzrc4BhKpUzCmTTh0ay6RTKhVznhIBaCUfS+Wvk/R7xpiD8/8Ha+1/8pADMVlcXJQkDQwMeE6STAcb0NhVDpxOLS9ua+1fSnprq88LJMloIUtRA6cU75wGAEBAKG4AAAJCcQMAEBDeOQ3OdXZ2+o4AAIlFccO5c+fO+Y4AAInFUjkAAAGhuOHc4uLizXu5AQBuUdwAAASE4gYAICAUNwAAAaG4AQAICLeDwbmuri7fEQAgsShuONfT0+M7AgAkFsUN56y1kqT9R7ciELOVSI89+Zw6b/yFvqq3Uz/43T08gQxoQ1zjhnNLS0taWlryHQPHMFuJVJ5ZUFStyUr6ympd5ZkFzVYi39EAvAozbqCFxsel+XnfKZpVrnWr3hiSJJ17zV9Ikv7qlTdpbLpDhUGfyZrl89LUlO8UgD/MuAGo3tg91jgAf5hxAy3UrjPF4cmriqo1SdJrz+4tj69vvaJsX0ZzEyM+owF4FWbcAFQq5pRJpw6NZdIplYo5T4kA3AkzbjiXyWR8R8AxHewef+zJ52RuSF/V26kf++6H2VUOtCGKG851d3f7joD7MFrI7hX16oN7A72UNtCOWCqHc7u7u9rdZVMTAMSB4oZzy8vLWl5e9h0DABKJ4gYAICAUNwAAAaG4AQAICMUNAEBAuB0MznE7GADEh+KGc7wBCwDEh6VyOLezs6OdnR3fMQAgkShuOFetVlWtVn3HAIBEYqkcwKmzvb2tl156SZubm76j4JTr6urSgw8+qHQ6feTvobgBnDovvfSSzp8/r4ceekjGGN9xcEpZa7W4uKiXXnpJb3zjG4/8fSyVAzh1Njc3NTAwQGnDK2OMBgYGjr3yw4wbQGxmK5Eee/I5fala0+v7MioVc23zqFBKG+3gfn4OmXHDuZ6eHvX09PiOAc9mK5HKMwuKqjVZSVG1pvLMgmYrke9o3v3e7/2e8vn8oT8dHR36oz/6I9/REABm3HCuq6vLdwScwM/9S+nZZ6WXVk52nMq1btUbQ03jY9MdKgye7Nj5vDQ1dbJj+PTII4/okUceufn3y5cva3p6WsVi8Z7fa62VtVYdHcy7Tiv+l4dzjUZDjUbDdwx4Vm/c/pnsdxo/rZ5//nn97M/+rH7zN39THR0deuyxxzQ0NKRv/uZv1gc+8AFJ0osvvqhcLqcf/uEf1lve8hb9zd/8jUqlkt7ylrfo4Ycf1sc+9jHP/ynQSsy44dzKyt5UbWBgwHMS3I+f+sn9T3pPdpzhyauKqrWm8WxfRnMTIyc7uGOLi4tNY11dXerp6ZG1VktLS01fz2Qy6u7u1u7ubtPz54/6s7+9va33vOc9+qVf+iUNDg7qqaee0gsvvKDPfvazstbqXe96lz796U9rcHBQL7zwgh5//HG9/e1v1xNPPKH5+XldvXpVr7zyioaGhvQd3/EdeuCBB+7vvwAEhRk3gFiUijll0qlDY5l0SqVizlOi9vPTP/3T+qZv+ib9wA/8gCTpqaee0lNPPaVCoaBv+ZZv0Re/+EW98MILkqQ3vOENevvb3y5J+sxnPqN3v/vdSqVSet3rXqfv/M7v1J/92Z95+8+B1mLGDSAWB7vH23VX+a3uNkM+uGXnTjo6Ou5rdWlubk5PPPGEPve5z90cs9aqXC7r/e9//6HXvvjii2z4xE3MuAHEZrSQ1R9PjOivJv+h/nhipC1L24fl5WX9yI/8iH7jN35D58+fvzleLBb1kY98ROvr65KkKIr08ssvN33/t3/7t+tjH/uYdnZ2dP36dX3605/W2972tpblh1/MuAGgxX7t135NL7/8sn70R3/00Hi5XNZ73vMeveMd75AknTt3Tr/1W7+lVOrwJYdHHnlEf/Inf6K3vvWtMsboF3/xF/XVX/3VLcsPv4y11neGe7p06ZK9cuWK7xg4onq9Lknq7Oz0nAT3ZfX5vY+9b/abI0bPPvusvuEbvsF3DEDS7X8ejTHPWGsv3e71zLjhHIUNAPHhGjec2/7/27v7+J7r/fHjjxcmQ3OR6sRirmKzi8/GXKR1zMUQrRyE20RqB6XLc+gkSXXo4vCLxrEo4mSVkp+ro8LiWCFnMRJLCE1OsRk2o82e3z8+n737zC4yPvPZZ3veb7fd7P16X73e77fPnp/X+/16v565ueTm5rq7GkopVSlp4FYud+bMGc6cOePuaiilVKWkgVsppZTyIBq4lVJKKQ+igVsppZTyIBq4lVLKDYwxDB8+3JrOy8vjxhtvpH///m6pT2pqKjabjdDQUA4ePFho3unTpxkxYgStWrWiZcuWjBgxwspJsGnTJrfVuTR+fn6cPHmy2PKgoCArnerjjz8OFD3+uLg4/P39iYmJKfO+Z82axblz5676GEqigVu53PXXX19oNCilVFF16tRhz5495OTYE7GsX7+eJk3cN7LcihUrGDRoEDt37qRly5aF5j300EO0aNGCAwcOcPDgQZo3b05sbGy51aW8swtu3LiRlJQUUlJSiIuLA4oe/9y5c1m/fj0JCQll3r4GbuVxatasSc2aNd1dDaUqvLvuuot///vfALz//vsMGzbMmpednc2DDz5Ix44dCQ0NZeXKlYB93PKIiAjCwsIICwtjy5YtgL3l261bNwYNGkTbtm2JiYmhuAG2UlJS6Ny5M8HBwQwYMIBTp06xdu1aZs2aRXx8PJGRkYWWP3DgAF9//TWTJ0+2yp5//nmSk5OtlvmZM2fo168fbdq0YezYseTn53Px4kUeeOABK/XozJkzATh48CB9+vShffv2REREkJqaCsADDzzA2LFj6dSpE08//TR+fn5kZmZa+2zdujU///wzJ06cYODAgYSHhxMeHs6XX34J2DO8RUVF0a5dO2JjY4s99pJcevxjx47l0KFD9O3bl5kzZ5Z4LS5evMj48eMJDAwkODiY2bNnQLVUmwAAGqtJREFUExcXx08//URkZCSRkZElnoerUpCUvSL/tG/fXpTnuHDhgly4cMHd1VBX6vR39p9KbO/evb9NZB397Zhd9ZN19HfrUKdOHdm1a5cMHDhQcnJyJCQkRDZu3Cj9+vUTEZGJEyfKu+++KyIip06dktatW0tWVpZkZ2dLTk6OiIjs379fCv4+bty4UXx8fOTHH3+UixcvSufOnSUpKanIfoOCgmTTpk0iIjJ58mR54oknRERkypQpMn369CLLr1y5Uu69994i5ffee6+sXLlSNm7cKNddd50cPHhQ8vLypGfPnvLRRx9JcnKy9OzZ01r+1KlTIiLSvXt32b9/v4iIbNu2TSIjI0VEZOTIkdKvXz/Jy8sTEZHHH39cFi5caC3Xo0cPEREZNmyYdVxHjhyRtm3biojIY489Ji+++KKIiKxZs0YAOXHiRJF6N2vWTAIDAyUkJERCQkLk9ddfL/b4mzVrZq1f0rWYO3euDBw4UHJzc0VEJD09vci6JZ0HZ4X+PzoAyVJCTNSR05TLnT17FtB83JXVip3HPCLjlycIDg7m8OHDvP/++9x1112F5q1bt45Vq1YxY8YMAM6fP8/Ro0dp3Lgxjz76KCkpKVSvXp39+/db63Ts2BFfX18AbDYbhw8f5o477rDmnz59mszMTP74xz8CMHLkSAYPHnzVx9GxY0datGgBwLBhw/jiiy/o0aMHhw4d4rHHHqNfv35ERUWRlZXFli1bCu2zYIhkgMGDB1vjsg8ZMoSXXnqJUaNG8cEHH1ipTzds2MDevXutdc6cOUNWVhabN29m+fLlAPTr148GDRqUWN+NGzfSqFGjyz6+kq7Fhg0bGDt2LDVq2ENpw4YNi6zbokWLIufhamngVkpdthU7jzFx+Tfk5F4E4FhmDhOXfwPgucG7zq1u3X10dDTjx49n06ZNpKenW+Uiwscff0ybNoXzl7/wwgvcfPPN7Nq1i/z8fGrVqmXNcx5uuHr16i55VhwQEEBKSgr5+flUq2Z/upqfn09KSgoBAQGkpaVhjCm0jjGGBg0asGvXLj777DPefPNNPvzwQ2bNmkX9+vVJSUkpdl/OqUu7dOnCgQMHOHHiBCtWrOC5556z9r1t27ZCx13eSroWl6O487Bw4cKrqo8GbqWqiCefhBL+XhbiW8/+b9rpovN2Hq3NhbzwIuUjE6oR2vTK62azwaxZV76+J3vwwQepX78+QUFBbNq0ySrv3bs3s2fPZvbs2Rhj2LlzJ6GhoZw+fRpfX1+qVavG4sWLuXjx4mXvq169ejRo0ICkpCQiIiJ49913rdZ3SVq1akVoaChTp07l+eefB2Dq1KmEhYXRqlUr0tLS2L59Oz/88APNmjVj6dKljB49mpMnT1KzZk0GDhxImzZtGD58OD4+PjRv3pyPPvqIwYMHIyLs3r2bkJCQIvs1xjBgwAD+8pe/4O/vb93Bi4qKYvbs2UyYMAGwP7O32WzceeedvPfeezz33HN88sknnDp16rLPy+8p6Vr06tWLefPmERkZSY0aNcjIyKBhw4Zcf/31nD17lkaNGhV7Hq6Wdk5TSl22C3n5ZSpXv8/X19d6JcnZ5MmTyc3NJTg4mHbt2lmdwx555BEWL15MSEgIqamphVqpl2Px4sVMmDCB4OBgUlJSrGBcmgULFrB//35atmxJy5Yt2b9/PwsWLLDmh4eH8+ijj+Lv70/z5s0ZMGAAx44do1u3bthsNoYPH84rr7wCQEJCAgsWLCAkJIR27dpZHb2KM2TIEJYsWWLdJgeIi4sjOTmZ4OBgAgICePPNNwGYMmUKmzdvpl27dixfvpymTUv+JhkZGWm9DjZixIjfPf6SrkVsbCxNmzYlODiYkJAQ3nvvPQBGjx5Nnz59iIyMLPE8XA1N66lcruB2nz7j9lClpPXs+urnHMvMKVLepL43Xz7Tvbxr5jKa1lNVJGVN66ktbuVyPj4++Pj4uLsaqhxM6N0Gb6/qhcq8vaozoXfZn/0ppa6MPuNWLufl5eXuKqhyUtABTXuVK+U+GriVyxW83uHcw1VVHveGNtFArZQbaeBWLpeVlQVo4FZKqfKgz7iVUkopD6KBWymllPIgGriVUsoNqlevjs1mIyQkpFCykJJkZmYyd+7ca1S70q1Zs4bQ0FBCQkIICAhg3rx5gD3DlvNwpJerW7dulOWV36VLl1rvVP/tb3+zyhctWsSNN95ovaP99ttvA/Ddd9/Rvn17goOD2bp1K2DPQNazZ89Ss3jNmDGDtm3bYrPZCA8P51//+tcV1dfVNHArpZQbeHt7k5KSwq5du3jllVeYOHFiqctXlMCdm5vL6NGjWb16Nbt27WLnzp1069YNuPLAXRbp6elMmDCBxMREvv32W/73v/+RmJhozR8yZIiVsrMg9ei8efN44403WLt2rTXeeHx8PMOHD6d27drF7ufNN99k/fr1bN++nZSUFBITE8uUcaw8aeBWLlevXj3q1avn7moo5THOnDlTKCnG9OnTCQ8PJzg4mClTpgDwzDPPcPDgQWw2GxMmTGDcuHGsWrUKgAEDBvDggw8CsHDhQiZNmgTAkiVL6NixIzabjTFjxljDo65bt44uXboQFhbG4MGDrQ6lfn5+TJkyhbCwMIKCgqyUm87Onj1LXl6eNcDSddddR5s2bdiyZQurVq1iwoQJ2Gw2Dh48WKhlevLkSfz8/ADIyclh6NCh+Pv7M2DAACsn+cKFC3nyySetfb311ls89dRThfZ/6NAhWrduzY033ghAz549+fjjj0s9v15eXpw7d45z587h5eVFZmYmq1evLnXUtJdffpn4+HhrTAofHx9GjhxZ6n6uFe1VrlyuIFOOUp7AvGh+f6ErJFNKbqHl5ORgs9k4f/48x48f5/PPPwfsQfX7779n+/btiAjR0dFs3ryZV199lT179lgJOj744AOSkpKIjo7m2LFjHD9+HICkpCSGDh3Kvn37WLp0KV9++SVeXl488sgjJCQkcNdddzF16lQ2bNhAnTp1eO2113j99detoU8bNWrEjh07mDt3LjNmzLBuNxdo2LAh0dHRNGvWjB49etC/f3+GDRvG7bffTnR0NP3792fQoEGlnpf4+Hhq167Nvn372L17N2FhYQDcd999TJs2jenTp+Pl5cU777xj3YYv0KpVK7777jsOHz6Mr68vK1as4Ndff7Xmf/zxx2zevJnbbruNmTNncuuttzJu3DhGjBjBhQsXmDdvHn//+9959tlnraQplzpz5gxnz561Mp5VNNriVi53/vx5zp8/7+5qKFWhFdwqT01N5dNPP2XEiBGICOvWrWPdunWEhoYSFhZGamoq33//fZH1IyIiSEpKYu/evQQEBHDzzTdz/Phxtm7dyu23305iYiJff/014eHh2Gw2EhMTOXToENu2bWPv3r107doVm83G4sWLOXLkiLXdP/3pTwC0b9+ew4cPF1v3t99+m8TERDp27MiMGTOs1v7l2rx5s5VsIzg4mODgYADq1q1L9+7dWbNmDampqeTm5hIUFFRo3QYNGhAfH8+QIUOIiIjAz8/PSgV69913c/jwYXbv3k2vXr2sFnLTpk3ZtGkTW7dupXbt2qSlpeHv78/999/PkCFDCqVG9QTaNFIul52dDXBN0+4p5cm6dOnCyZMnOXHiBCLCxIkTGTNmTKFlLg2iTZo0ITMzk08//ZQ777yTjIwMPvzwQ+rWrcv111+PiDBy5MgiSS1Wr15Nr169eP/994utS8H4C85pQXv37s3PP/9Mhw4drBZ4UFAQQUFB3H///TRv3pxFixYV2VaNGjXIz7cnoLncL/OxsbG8/PLLtG3bllGjRhW7zN13383dd98NwPz5863A7ZwfITY2lqeffrrIupMmTWLq1KnExcURGxuLn58fzz77LAkJCdYyPj4+1K1bl0OHDlXIVrcGbqVUlVba7exrJTU1lYsXL3LDDTfQu3dvJk+eTExMDHXr1uXYsWN4eXlZqSKdde7cmVmzZvH555+Tnp7OoEGDrNvUPXr04J577uGpp57ipptuIiMjg7Nnz9K5c2fGjRvHgQMHaNWqFdnZ2Rw7dozbbiuaVKbAZ599Zv2elZVFcnKy1SEtJSWFZs2aARSpo5+fH19//TUdO3Zk2bJlVnlBCs7u3buzZ88edu/ebc3r1KkTP/74Izt27ChU7uyXX37hpptu4tSpU8ydO5cPP/wQgOPHj3PLLbcAsGrVqiKJO/7zn//QuHFjWrduzblz56hWrRrVqlUrtmf5xIkTGTduHEuXLsXHx4esrCyWL19+WdnEypsGbqWUcoOCZ9wAIsLixYupXr06UVFR7Nu3jy5dugD228dLliyhZcuWdO3alcDAQPr27cv06dOJiIhg3bp1tGrVimbNmpGRkUFERAQAAQEBTJ06laioKPLz8/Hy8uKf//wnnTt3ZtGiRQwbNswannjq1KmlBm5nIsI//vEPxowZg7e3N3Xq1LFa20OHDuXPf/4zcXFxLFu2jPHjx3Pfffcxf/58+vXrZ23j4YcfZtSoUfj7++Pv70/79u0L7eO+++4jJSWlUIc9Z0888QS7du0C4Pnnn7fqHhcXx6pVq6hRowYNGzYsdBdARJg6dSpLly4F7Kk3Y2JiyMvLIz4+vsg+Hn74YbKysggPD8fLywsvLy/++te/XtY5Km+a1lO5nKb19HClpPWsLDStZ8XWv39/nnrqKXr06OHuqlwTmtZTKaWUR8rMzOS2227D29u7ygTtK6G3ypXL1a9f391VqJJW7Dym6TaVR6tfv77H9fB2Bw3cyuUKeniqa2fFzmNMXP4NObn2ATaOZeYwcfk3ABq8lapkNHArlysYBcnb29vNNan4nnwSHONpXJWdR2tzIS+8SPnIhGqENi3btnwdg96lnf6tzGaDWbOuooJKKZfRwK1cruDVCg3c186FvPwylSulPJcGbqXcyFWt2K6v7uJYZk6R8ib1vdn0TPeybeyM41+fq6+XUsr1tFe5UpXAhN5t8PYq3LfA26s6E3q3cVONVGnS09Ot1JN/+MMfaNKkiTXtPO721ejWrRtt2rSxtlswMMuJEyfo1KkToaGhJCUl8dFHH+Hv709kZGSZ97Fo0SJ++uknl9RXXT5tcStVCRR0QNNe5Z7hhhtusJKFvPDCC9StW5fx48db8/Py8lySrCchIYEOHQq/CpyYmEhQUJA1dGmfPn146623uOOOO8q8/UWLFhEYGEjjxo2vuq7q8mngVqqSuDe0iQbqcpLwTQKTEidx9PRRmtZryrQe04gJinHpPh544AFq1arFzp076dq1qzVedkFADwwMZM2aNfj5+bFkyRLi4uL49ddf6dSpE3Pnzr2stzlSUlJ4+umnycnJITk5mQEDBvDFF1/w0EMPER0dzauvvsozzzzDpk2buHDhAuPGjbPGTH/ttddYsmQJ1apVo2/fvnTo0IHk5GRiYmLw9vZm69at2q/lGtHArVyupGEKlfJECd8kMHr1aM7l2jtdHjl9hNGrRwO4PHinpaWxZcsWqlevzgsvvFDsMiWl6yxuDO2CoArQq1cvpk+fzksvvURycjJz5swBYOPGjcyYMYMOHTowf/586tWrx3//+18uXLhA165diYqKIjU1lZUrV/LVV19Ru3ZtMjIyaNiwIXPmzLHWVdeOBm7lciXluFXKE01KnGQF7QLncs8xKXGSywP34MGDf7fl7JyuE+yvX950003FLlvcrfLSrFu3jt27d1sJQU6fPs3333/Phg0bGDVqFLVr1wbsObmV+2jgVi5X8DpYwYdcKU929PTRMpVfjTp16li/O6fEhN/SYpaUrtMVRITZs2fTu3fvQuXO2cGU+2nTSLlcTk6ONQiLUp6uab3iR7ApqdxV/Pz82LFjBwA7duzghx9+AOzpOpctW8Yvv/wCQEZGBkeOHHHJPnv37k18fDy5ubkA7N+/n+zsbHr16sU777xjfSnPyMgAiqbxVNeGBm6llCrFtB7TqO1V+O5Rba/aTOsxrVz3O3DgQDIyMmjXrh1z5syxUlc6p+sMDg6mV69eHD9+vNhtxMTEWK+D9ezZ83f3GRsbS0BAAGFhYQQGBjJmzBjy8vLo06cP0dHRdOjQAZvNxowZMwB7h7qxY8dis9n0y/o1pGk9lctpWk8Pp2k9i7gWvcpV1VXWtJ76jFsppX5HTFCMBmpVYeitcqWUUsqDaItbuZy+KqKUUuVHW9zK5YwxGGPcXQ2lSuUJ/XtU5Xcl/w81cCuXy87OJjs7293VUKpEtWrVIj09XYO3cisRIT09nVq1apVpPb1VrlyuYKAI58EklKpIfH19SUtL48SJE+6uiqriatWqha+vb5nWcUvgNsb0Ad4AqgNvi8ir7qiHUqpq8vLyonnz5u6uhlJX5JrfKjfGVAf+CfQFAoBhxpiAa10PpZRSyhO54xl3R+CAiBwSkV+BD4B73FAPpZRSyuO4I3A3AX50mk5zlCmllFLqd1TYzmnGmNHAaMdkljHmu8tctRFwsnxqpcpAr0PFoNehYtDrUDF40nVoVtIMdwTuY8CtTtO+jrJCRGQ+ML+sGzfGJJc0vqu6dvQ6VAx6HSoGvQ4VQ2W5Du64Vf5foLUxprkxpiYwFFjlhnoopZRSHueat7hFJM8Y8yjwGfbXwRaKyLfXuh5KKaWUJ3LLM24RWQusLafNl/n2uioXeh0qBr0OFYNeh4qhUlwHj8jHrZRSSik7HatcKaWU8iCVJnAbY6YbY1KNMbuNMf/fGFPfad5EY8wBY8x3xpje7qxnVWCM6eM41weMMc+4uz5VgTHmVmPMRmPMXmPMt8aYJxzlDY0x640x3zv+beDuulYFxpjqxpidxpg1junmxpivHJ+JpY6OuaocGWPqG2OWOeLCPmNMl8ryeag0gRtYDwSKSDCwH5gI4BhOdSjQDugDzHUMu6rKgQ5p6zZ5wF9FJADoDIxznPdngEQRaQ0kOqZV+XsC2Oc0/RowU0RaAaeAh9xSq6rlDeBTEWkLhGC/HpXi81BpAreIrBORPMfkNuzvh4N9ONUPROSCiPwAHMA+7KoqHzqkrRuIyHER2eH4/Sz2P1JNsJ/7xY7FFgP3uqeGVYcxxhfoB7ztmDZAd2CZYxG9DuXMGFMPuBNYACAiv4pIJpXk81BpAvclHgQ+cfyuQ6xeW3q+3cwY4weEAl8BN4vIcces/wE3u6laVcks4Gkg3zF9A5Dp1LDQz0T5aw6cAN5xPLJ42xhTh0ryefCowG2M2WCM2VPMzz1Oy0zCftswwX01Vco9jDF1gY+BJ0XkjPM8sb9Coq+RlCNjTH/gFxH52t11qeJqAGFAvIiEAtlcclvckz8PFXas8uKISM/S5htjHgD6Az3kt/fcLmuIVeUyer7dxBjjhT1oJ4jIckfxz8aYW0TkuDHmFuAX99WwSugKRBtj7gJqAT7Yn7XWN8bUcLS69TNR/tKANBH5yjG9DHvgrhSfB49qcZfGGNMH++2paBE55zRrFTDUGHOdMaY50BrY7o46VhE6pK0bOJ6jLgD2icjrTrNWASMdv48EVl7rulUlIjJRRHxFxA/7//3PRSQG2AgMciym16Gcicj/gB+NMW0cRT2AvVSSz0OlGYDFGHMAuA5IdxRtE5GxjnmTsD/3zsN+C/GT4reiXMHR2pjFb0PaTnNzlSo9Y8wdQBLwDb89W30W+3PuD4GmwBHgPhHJcEslqxhjTDdgvIj0N8a0wN5RsyGwExguIhfcWb/Kzhhjw95BsCZwCBiFvbHq8Z+HShO4lVJKqaqg0twqV0oppaoCDdxKKaWUB9HArZRSSnkQDdxKKaWUB9HArZRSSnkQDdxKVTDGGDHGLHGarmGMOVGQaaqc9z3LGHOn4/fuxpgdjtEJFxtjajjKuxljThtjUhw/zzvKbzTGfOFY/l6nba40xjQuZZ8jHOt84xiecryjfJExZpDj9w+MMa3L89iV8hQauJWqeLKBQGOMt2O6F9dgpC1jzA1AZxHZbIyphj0Jw1ARCcT+zutIp8WTRMTm+HnJUTYMeBN7opknHdu8G9gpIj+VsM++jmWjRCQIe2az08UsGo99gCWlqjwN3EpVTGuxZ5gCe0B8v2CGMaaOMWahMWa7o4V6j6PczxiT5Ggl7zDG3O4o72aM2eSUmzjBMdLapQYCnzp+vwH4VUT2O6bXO+aXJheojX0gpIuOFvqTwD9KWWci9kFKfgJwZPF7q5jlkoCeBa1+paoyDdxKVUwfYB+qtxYQjH0EtAKTsA+l2RGIBKY7Mh/9AvQSkTBgCBDntE4o9iAaALTAPqb2pboCBckxTgI1jDEdHNODKDwGfRdjzC5jzCfGmHaOsvewp01cD7wMPAK8e8kQxJcKdNpniUQkH3tK3pDfW1apyk6/vSpVAYnIbkd6zmHYW9/OorAnshjvmK6FfQjHn4A5jqEeLwK3Oa2zXUTSAIwxKYAf8MUl270FeypERESMMUOBmcaY64B1jm0C7ACaiUiWY3jbFUBrETmN4y6BMaYB9qQOA4wxbwENgP8nIluv7IwA9i8mjbmMQK9UZaaBW6mKaxUwA+iG/dZ1AQMMFJHvnBc2xrwA/Iy9VVoNOO8023lc7IsU/9nPwf4lAABHkI1wbDsKxxcB53ShIrLWGDPXGNNIRE46bWsyMA37F48vsGdnWg70vmSf3wLtgc+Lqc+lajnqqFSVprfKlaq4FgIvisg3l5R/BjxW8JzaGBPqKK8HHHfcVr4fe5KXstgHtCqYMMbc5Pj3OuBv2DueYYz5g9O+O2L/O5LutF5rwFdENmF/5p2PPe9xQWc7Z69gv9X/B8e6NY0xsSXU7zZgTxmPSalKRwO3UhWUiKSJSFwxs/4OeAG7jTHfOqYB5gIjjTG7gLbYe6eXxb+xt+4LTDDG7AN2A6tFpKBVPAjY49hPHPae587ZiqZhfw4P9k51D2NP9/pGMce4FpgDbHAcyw7sOawLMcbcDOQ40jUqVaVpdjCllMUY8wXQX0Qy3V0XZ8aYp4AzIrLA3XVRyt20xa2UcvZX7B3dKppM7O+VK1XlaYtbKaWU8iDa4lZKKaU8iAZupZRSyoNo4FZKKaU8iAZupZRSyoNo4FZKKaU8iAZupZRSyoP8H/HqLF1dePrXAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 576x1152 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(8, 16))\n",
    "study_ticks = np.arange(n_studies)\n",
    "\n",
    "# Get 95% CI for individual studies\n",
    "lower_bounds, upper_bounds = var_to_ci(y, v, n)\n",
    "ax.scatter(y, study_ticks+1)\n",
    "for study in study_ticks:\n",
    "    ax.plot((lower_bounds[study], upper_bounds[study]), (study+1, study+1), color='blue')\n",
    "ax.axvline(0, color='gray', alpha=0.2, linestyle='--', label='Zero')\n",
    "ax.axvline(np.mean(y), color='orange', alpha=0.2, label='Mean of Observed Effects')\n",
    "\n",
    "# Get 95% CI for true effect\n",
    "lower_bound, upper_bound = var_to_ci(true_effect, between_study_var, 1)\n",
    "ax.scatter((true_effect,), (n_studies+1,), color='green', label='True Effect')\n",
    "ax.plot((lower_bound, upper_bound), (n_studies+1, n_studies+1),\n",
    "        color='green', linewidth=3, label='Between-Study 95% CI')\n",
    "ax.set_ylim((0, n_studies+2))\n",
    "ax.set_xlabel('Mean (95% CI)')\n",
    "ax.set_ylabel('Study')\n",
    "ax.legend()\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create a Dataset object containing the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "dset = pymare.Dataset(y=y, X=None, v=v, n=n, add_intercept=True)\n",
    "\n",
    "# Here is a dictionary to house results across models\n",
    "results = {}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fit a model\n",
    "\n",
    "When you have `z` or `p`:\n",
    "- Stouffer\n",
    "- Fisher\n",
    "\n",
    "When you have `y` and `v` and don't want to estimate between-study variance:\n",
    "- WeightedLeastSquares\n",
    "\n",
    "When you have `y` and `v` and want to estimate between-study variance:\n",
    "- DerSimonianLaird\n",
    "- Hedges\n",
    "- VarianceBasedLikelihoodEstimator\n",
    "\n",
    "When you have `y` and `n` and want to estimate between-study variance:\n",
    "- SampleSizeBasedLikelihoodEstimator\n",
    "\n",
    "When you have `y` and `v` and want a hierarchical model:\n",
    "- StanMetaRegression"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### First, we have \"combination models\", which combine p and/or z values\n",
    "The two combination models in PyMARE are Stouffer's and Fisher's Tests.\n",
    "\n",
    "Notice that these models don't use `Dataset` objects."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "z: [1.00546702]\n",
      "p: [0.31467202]\n"
     ]
    }
   ],
   "source": [
    "stouff = pymare.estimators.Stouffers(input='z')\n",
    "stouff.fit(y=z[:, None])\n",
    "stouff_summary = stouff.summary()\n",
    "print('z: {}'.format(stouff_summary.z))\n",
    "print('p: {}'.format(stouff_summary.p))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "z: [1.00546702]\n",
      "p: [0.31467202]\n"
     ]
    }
   ],
   "source": [
    "fisher = pymare.estimators.Stouffers(input='z')\n",
    "fisher.fit(y=z[:, None])\n",
    "fisher_summary = fisher.summary()\n",
    "print('z: {}'.format(fisher_summary.z))\n",
    "print('p: {}'.format(fisher_summary.p))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Now we have fixed effects model\n",
    "\n",
    "This estimator does not attempt to estimate between-study variance. Instead, it takes `tau^2` as an argument."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "        name   estimate        se   z-score       p-value  ci_0.025   ci_0.975\n",
      "0  intercept  20.326462  2.967198  6.850389  7.364998e-12  14.51086  26.142063\n"
     ]
    }
   ],
   "source": [
    "wls = pymare.estimators.WeightedLeastSquares()\n",
    "wls.fit(dset)\n",
    "wls_summary = wls.summary()\n",
    "results['Weighted Least Squares'] = wls_summary.to_df()\n",
    "print(wls_summary.to_df())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Methods that estimate between-study variance\n",
    "\n",
    "The `DerSimonianLaird`, `Hedges`, and `VarianceBasedLikelihoodEstimator` estimators all estimate between-study variance from the data, and use `y` and `v`.\n",
    "\n",
    "`DerSimonianLaird` and `Hedges` use relatively simple methods for estimating between-study variance, while `VarianceBasedLikelihoodEstimator` can use either maximum-likelihood (ML) or restricted maximum-likelihood (REML) to iteratively estimate it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "        name   estimate        se   z-score       p-value  ci_0.025   ci_0.975\n",
      "0  intercept  20.326462  2.967198  6.850389  7.364998e-12  14.51086  26.142063\n"
     ]
    }
   ],
   "source": [
    "dsl = pymare.estimators.DerSimonianLaird()\n",
    "dsl.fit(dset)\n",
    "dsl_summary = dsl.summary()\n",
    "results['DerSimonian-Laird'] = dsl_summary.to_df()\n",
    "print(dsl_summary.to_df())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "        name   estimate        se     z-score  p-value   ci_0.025   ci_0.975\n",
      "0  intercept  20.326462  0.158114  128.555832      0.0  20.016564  20.636359\n"
     ]
    }
   ],
   "source": [
    "hedge = pymare.estimators.Hedges()\n",
    "hedge.fit(dset)\n",
    "hedge_summary = hedge.summary()\n",
    "results['Hedges'] = hedge_summary.to_df()\n",
    "print(hedge_summary.to_df())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "        name   estimate        se   z-score       p-value   ci_0.025  \\\n",
      "0  intercept  20.606514  3.077402  6.696075  2.140910e-11  14.574917   \n",
      "\n",
      "    ci_0.975  \n",
      "0  26.638111  \n"
     ]
    }
   ],
   "source": [
    "vb_ml = pymare.estimators.VarianceBasedLikelihoodEstimator(method='ML')\n",
    "vb_ml.fit(dset)\n",
    "vb_ml_summary = vb_ml.summary()\n",
    "results['Variance-Based with ML'] = vb_ml_summary.to_df()\n",
    "print(vb_ml_summary.to_df())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "        name   estimate        se   z-score       p-value   ci_0.025  \\\n",
      "0  intercept  20.660719  3.101268  6.662023  2.700840e-11  14.582345   \n",
      "\n",
      "    ci_0.975  \n",
      "0  26.739093  \n"
     ]
    }
   ],
   "source": [
    "vb_reml = pymare.estimators.VarianceBasedLikelihoodEstimator(method='REML')\n",
    "vb_reml.fit(dset)\n",
    "vb_reml_summary = vb_reml.summary()\n",
    "results['Variance-Based with REML'] = vb_reml_summary.to_df()\n",
    "print(vb_reml_summary.to_df())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The `SampleSizeBasedLikelihoodEstimator` estimates between-study variance using `y` and `n`, but assumes within-study variance is homogenous across studies."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "        name  estimate        se   z-score  p-value   ci_0.025   ci_0.975\n",
      "0  intercept  22.91788  2.450766  9.351311      0.0  18.114466  27.721294\n"
     ]
    }
   ],
   "source": [
    "sb_ml = pymare.estimators.SampleSizeBasedLikelihoodEstimator(method='ML')\n",
    "sb_ml.fit(dset)\n",
    "sb_ml_summary = sb_ml.summary()\n",
    "results['Sample Size-Based with ML'] = sb_ml_summary.to_df()\n",
    "print(sb_ml_summary.to_df())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "        name   estimate        se   z-score  p-value   ci_0.025  ci_0.975\n",
      "0  intercept  22.917806  2.481961  9.233751      0.0  18.053253  27.78236\n"
     ]
    }
   ],
   "source": [
    "sb_reml = pymare.estimators.SampleSizeBasedLikelihoodEstimator(method='REML')\n",
    "sb_reml.fit(dset)\n",
    "sb_reml_summary = sb_reml.summary()\n",
    "results['Sample Size-Based with REML'] = sb_reml_summary.to_df()\n",
    "print(sb_reml_summary.to_df())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### What about the Stan estimator?\n",
    "We're going to skip this one here because of how computationally intensive it is."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Let's check out our results!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/tsalo/anaconda/envs/python3/lib/python3.6/site-packages/ipykernel/__main__.py:19: UserWarning: Matplotlib is currently using module://ipykernel.pylab.backend_inline, which is a non-GUI backend, so cannot show the figure.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(8, 8))\n",
    "\n",
    "for i, (estimator_name, summary_df) in enumerate(results.items()):\n",
    "    ax.scatter((summary_df.loc[0, 'estimate'],), (i+1,), label=estimator_name)\n",
    "    ax.plot((summary_df.loc[0, 'ci_0.025'],  summary_df.loc[0, 'ci_0.975']), \n",
    "            (i+1, i+1), \n",
    "            linewidth=3)\n",
    "\n",
    "# Get 95% CI for true effect\n",
    "lower_bound, upper_bound = var_to_ci(true_effect, between_study_var, 1)\n",
    "ax.scatter((true_effect,), (i+2,), label='True Effect')\n",
    "ax.plot((lower_bound, upper_bound), (i+2, i+2),\n",
    "        linewidth=3, label='Between-Study 95% CI')\n",
    "ax.set_ylim((0, i+3))\n",
    "ax.set_yticklabels([None] + list(results.keys()) + ['True Effect'])\n",
    "\n",
    "ax.set_xlabel('Mean (95% CI)')\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:python3]",
   "language": "python",
   "name": "conda-env-python3-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}