pymwv.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyMpiOOaxfV60WyshUca4XEw",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/mauriciodev/pymwv/blob/master/pymwv.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from osgeo import ogr\n",
        "import math\n",
        "import matplotlib.pyplot as plt"
      ],
      "metadata": {
        "id": "JX59FoIezjAh"
      },
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "AHvEUvE_zdqy"
      },
      "outputs": [],
      "source": [
        "class MWVD:\n",
        "    def ApoloniusCircle(self, s1, s2, w1, w2, extent):\n",
        "        \"\"\"\n",
        "        Calculates the Apollonius circle for two sites (s1, s2) with weights (w1, w2).\n",
        "\n",
        "        Args:\n",
        "          s1: Tuple (x, y) representing the coordinates of the first site.\n",
        "          s2: Tuple (x, y) representing the coordinates of the second site.\n",
        "          w1: Weight of the first site.\n",
        "          w2: Weight of the second site.\n",
        "          extent: ogr.Geometry representing the extent of the area.\n",
        "\n",
        "        Returns:\n",
        "          ogr.Geometry: The Apollonius circle as an ogr.Geometry object.\n",
        "        \"\"\"\n",
        "        s1x, s1y = s1\n",
        "        s2x, s2y = s2\n",
        "\n",
        "        if w1 == w2:  # Regular Voronoi\n",
        "            mx = (s1x + s2x) / 2.\n",
        "            my = (s1y + s2y) / 2.\n",
        "            dx = s1x - s2x\n",
        "            dy = s1y - s2y\n",
        "            d = extent.GetBoundary().Length()\n",
        "            curve = ogr.Geometry(ogr.wkbLineString)\n",
        "\n",
        "            if dy != 0:\n",
        "                m = math.atan(-1. * (dx / dy))\n",
        "                curve.AddPoint_2D(-d * math.cos(m) + mx, -d * math.sin(m) + my)\n",
        "                curve.AddPoint_2D(d * math.cos(m) + mx, d * math.sin(m) + my)\n",
        "            else:\n",
        "                curve.AddPoint_2D(mx, -d + my)\n",
        "                curve.AddPoint_2D(mx, d + my)\n",
        "\n",
        "            shortCurve = curve.Intersection(extent)\n",
        "            diff = extent.GetBoundary().Difference(curve)\n",
        "            boundary = diff.GetGeometryRef(1)\n",
        "            endPoint = boundary.GetPoint(0)\n",
        "            boundary.AddPoint_2D(endPoint[0], endPoint[1])\n",
        "\n",
        "            ring = ogr.Geometry(ogr.wkbLinearRing)\n",
        "            for point in boundary.GetPoints():\n",
        "                ring.AddPoint_2D(point[0], point[1])\n",
        "            ring.AddGeometry(boundary)\n",
        "\n",
        "            domBoundary = ogr.Geometry(ogr.wkbPolygon)\n",
        "            domBoundary.AddGeometry(ring)\n",
        "\n",
        "        else:  # Weighted Voronoi\n",
        "            den=1./(w1*w1-w2*w2);\n",
        "            cx=(w1*w1*s2x-w2*w2*s1x)*den\n",
        "            cy=(w1*w1*s2y-w2*w2*s1y)*den\n",
        "            #print('Center:', cx, cy)\n",
        "            d= math.sqrt(((s1x-s2x)*(s1x-s2x) + (s1y-s2y)*(s1y-s2y)))\n",
        "            r=w1*w2*d*den\n",
        "            #print(\"Radius:\",r)\n",
        "            if (r<0): r=r*-1\n",
        "            #creating the circle boundary from 3 points\n",
        "            arc= ogr.Geometry(ogr.wkbCircularString)\n",
        "            arc.AddPoint_2D(cx+r,cy)\n",
        "            arc.AddPoint_2D(cx-r,cy)\n",
        "            arc.AddPoint_2D(cx+r,cy)\n",
        "            #creating the circle polygon\n",
        "            domBoundary=ogr.Geometry(ogr.wkbCurvePolygon)\n",
        "            domBoundary.AddGeometry(arc)\n",
        "\n",
        "        # Create ogr.Geometry from tuple for intersection check\n",
        "        s1_geom = ogr.Geometry(ogr.wkbPoint)\n",
        "        s1_geom.AddPoint_2D(*s1)\n",
        "\n",
        "        if s1_geom.Intersects(domBoundary):\n",
        "            return domBoundary.Intersection(extent)\n",
        "        else:\n",
        "            return extent.Difference(domBoundary)\n",
        "\n",
        "    def getMWVLayer(self, sites, extent):\n",
        "        \"\"\"\n",
        "        Calculates the Multiplicatively Weighted Voronoi Diagram for a set of sites.\n",
        "\n",
        "        Args:\n",
        "          sites: List of dictionaries, each with keys 'p' (tuple of x,y coordinates)\n",
        "                 and 'w' (weight).\n",
        "          extent: ogr.Geometry representing the extent of the area.\n",
        "\n",
        "        Returns:\n",
        "          list: A list of polygons, each represented as a list of (x, y) tuples.\n",
        "        \"\"\"\n",
        "        polygons = []\n",
        "        for site1 in sites:\n",
        "            dominance = extent\n",
        "            for site2 in sites:\n",
        "                if site1 != site2:\n",
        "                    twoSitesDominance = self.ApoloniusCircle(site1['p'], site2['p'], site1['w'], site2['w'], extent)\n",
        "                    dominance = dominance.Intersection(twoSitesDominance)\n",
        "\n",
        "            # Extract polygon boundary points\n",
        "            boundary = dominance.GetBoundary()\n",
        "            polygon = []\n",
        "            for i in range(boundary.GetPointCount()):\n",
        "                x, y, _ = boundary.GetPoint(i)\n",
        "                polygon.append((x, y))\n",
        "            polygons.append(polygon)\n",
        "        return polygons\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def create_extent(points):\n",
        "    \"\"\"\n",
        "    Creates an extent polygon from a list of points.\n",
        "\n",
        "    Args:\n",
        "      points: A list of (x, y) tuples representing the points.\n",
        "\n",
        "    Returns:\n",
        "      ogr.Geometry: An ogr.Geometry object representing the extent polygon.\n",
        "    \"\"\"\n",
        "    minx = min(p[0] for p in points)\n",
        "    miny = min(p[1] for p in points)\n",
        "    maxx = max(p[0] for p in points)\n",
        "    maxy = max(p[1] for p in points)\n",
        "\n",
        "    # Create a slightly larger extent for buffering\n",
        "    dx = (maxx - minx) / 10.\n",
        "    dy = (maxy - miny) / 10.\n",
        "    minx -= dx\n",
        "    miny -= dy\n",
        "    maxx += dx\n",
        "    maxy += dy\n",
        "\n",
        "    ring = ogr.Geometry(ogr.wkbLinearRing)\n",
        "    ring.AddPoint(minx, miny)\n",
        "    ring.AddPoint(minx, maxy)\n",
        "    ring.AddPoint(maxx, maxy)\n",
        "    ring.AddPoint(maxx, miny)\n",
        "    ring.AddPoint(minx, miny)\n",
        "\n",
        "    extent = ogr.Geometry(ogr.wkbPolygon)\n",
        "    extent.AddGeometry(ring)\n",
        "    return extent"
      ],
      "metadata": {
        "id": "jIlJGYhozxsc"
      },
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "points = [\n",
        "    (1, 1),\n",
        "    (3, 2),\n",
        "    (5, 1),\n",
        "    (4, 4),\n",
        "    (2, 5)\n",
        "]\n",
        "weights = [1.11, 1.12, 1.13, 1.1, 1.2]\n",
        "\n",
        "sites = [{'p': p, 'w': w} for p, w in zip(points, weights)]\n",
        "extent = create_extent(points)\n",
        "\n",
        "mwvd = MWVD()\n",
        "polygons = mwvd.getMWVLayer(sites, extent)\n",
        "\n",
        "# Plotting the polygons\n",
        "fig, ax = plt.subplots()\n",
        "print(polygons)\n",
        "for polygon in polygons:\n",
        "    xs, ys = zip(*polygon)\n",
        "    ax.plot(xs, ys, '-')\n",
        "\n",
        "# Plot the sites\n",
        "for i, point in enumerate(points):\n",
        "    ax.plot(point[0], point[1], 'ro')\n",
        "    ax.text(point[0], point[1]+0.1, f'{i+1}', fontsize=8, color='blue')\n",
        "\n",
        "ax.set_aspect('equal')\n",
        "\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 467
        },
        "id": "yNLnj1kyzzfJ",
        "outputId": "2f29ecb3-ff61-4c27-fd08-1180df4246d3"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[(0.6, 0.6), (0.6, 3.1126428394299883), (1.2186356643521516, 2.9714432860165814), (2.3606625608392626, 0.6), (0.6, 0.6)], [(2.3606625608392626, 0.6), (1.223501903483414, 2.9613384363396777), (1.6384340130675807, 3.1353105168661837), (2.6506182995218257, 3.4797128604182825), (4.445320347487413, 2.5769423775420512), (3.49327507561819, 0.6), (2.3606625608392626, 0.6)], [(4.411918942123078, 2.5075836446274207), (4.8394978518656435, 2.63678471508706), (5.4, 2.8493551409819595), (5.4, 0.6), (3.49327507561819, 0.6), (4.411918942123078, 2.5075836446274207)], [(4.8394978518656435, 2.63678471508706), (4.518925291347588, 2.539917640881138), (2.607735405128111, 3.5012838010225065), (2.799848721585431, 3.962670994610289), (3.0716146398498942, 4.534192115847554), (3.371225291177893, 5.091620643966734), (3.561300457774457, 5.4), (5.4, 5.4), (5.4, 2.8493551409819595), (4.8394978518656435, 2.63678471508706)], [(1.6384340130675807, 3.1353105168661837), (1.2373920319681704, 2.9671622675049623), (0.6, 3.1126428394299883), (0.6, 5.4), (3.561300457774457, 5.4), (3.371225291177893, 5.091620643966734), (3.071614639849872, 4.5341921158475085), (2.799848721585408, 3.9626709946102663), (2.5901927632898905, 3.459152675314469), (1.6384340130675807, 3.1353105168661837)]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZgAAAGdCAYAAAAv9mXmAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAABJTElEQVR4nO3dd3hUZd7G8e+ZSTJJSDLpkJBAIPRACJ3QDCAgIIJY1oa4CrourrAuFty1vRZwXQu7uojogg1RkaKsVKUJ0gm9hZZOSO+TZGbePwZQVkQCmXmm/D7XNVcyJ4fMnWEyd057Hs1qtVoRQgghGphOdQAhhBDuSQpGCCGEXUjBCCGEsAspGCGEEHYhBSOEEMIupGCEEELYhRSMEEIIu5CCEUIIYRdejn5Ai8VCdnY2gYGBaJrm6IcXQghxDaxWK2VlZURHR6PTXX4bxeEFk52dTWxsrKMfVgghRAPKyMggJibmsus4vGACAwMBW7igoCBHP7wQQohrUFpaSmxs7IX38stxeMGc3y0WFBQkBSOEEC7qSg5xyEF+IYQQdiEFI4QQwi6kYIQQQtiFFIwQQgi7kIIRQghhF1IwQggh7EIKRgghhF1IwQghhLALKRghhBB2IQUjhBDCLqRghBBC2IUUjBBCCLtw+GCX16o0v4oDG7Ooq7WojiKES7HWmTEd3E9zr0yCfKpUxxGK6I1GQsePRx8QYPfHcrmC2bniNAd/yFYdQwgXFUbJmRMkHPpYdRChkFdkJCG33Wb/x7H7IzSwWpMZgJh2ITSOk+H+hbgSBcfOcOp4NQCt2vsTNuAhxYmECmXfraEm7TjWykqHPJ7LFcx5cZ3C6TxYZsYU4rdY6sx89odt4NWYaF02Sa/8SXUkoUhtVhY1accd9nhykF8IN7fzn8so9mqM3lzNwD+nqI4jPIgUjBBurOxULrsO2nZUdG5VTXDry8+hLkRDkoIRwo2tfX01dV5+BNXm0XPKKNVxhIeRghHCTaUt3kxGbVOwWrjujtboDd6qIwkPIwUjhBuqrajmh2VZALRslEuzId0UJxKeSApGCDf042tLqfAOw6e2jOueHKE6jvBQUjBCuJn8Pcc5kGUEoEdPA/6NQxUnEp5KCkYIN2KxWPj+3z9i0fsQbskhceJQ1ZGEB5OCEcKNHJi7mrNaNJqljoEP9UCnk19xoY68+oRwE1X5JWzdbBvEsn3jIiK7tVGcSHg6KRgh3MSGGcsweQfhX1tI3ydvUh1HCCkYce3i4qBtW0hKst0+/1xxIA+UsTaVtLLGAPS9oTE+AX6KEwnhwoNdCufy+ee2chGOZ66tY/2nh8Erkqb6LNrcNk51JCEA2YIRwuXtmPkNJV6R6OuqGfjYINVxhLhACkY0iHvvhU6d4IEH4OxZ1Wk8R+mJbHYfNgCQ1NaEMb6p4kRC/EQKRlyzDRtg717YtQvCw2H8eNWJPMfa17/D7OWLsTaPHjKYpXAycgxGXLNmzWwfvb1hyhRoI2fHOsSxhRvJNJ8bzPLudui95ddZOBfZghHXpKICiot/uv/ZZ9Cli7I4HqOmvIoflp8BID7wDLGDktQGEuIS5E8ecU3OnIFbbgGzGaxWaNkSPvpIdSr3t/m1r6n0jsBQW8aAJ2QwS+GcpGDENWnZEnbvVp3Cs5SeyuFwdiDooXtPA/6RIaojCXFJUjCi/sxm2LgRcnIgKgr69we9XnUqj7H5X99j1kdhrMsjceLtquMI8avqdQzm+eefR9O0i27t2rWzVzbhjBYtsl26P3Ag3HWX7WNcnG25sLv8Pcc5URYJQO+RzWQwS+HU6r0Fk5CQwJo1a376Bl6yEeQxFi2CW2+1HWz5uaws2/KFC2HsWDXZPMSm937EqosmwpJDq5vvVh1HiMuqdzt4eXnRpEkTe2QRzsxshsmTf1kuYFumabZzlEePlt1ldpKxNpVMczQAfe/ppDiNEL+t3tvXx44dIzo6mpYtW3L33XeTnp5+2fVNJhOlpaUX3YQL2rgRMjN//etWK2Rk2NYTdvHjZwcAiPHKpumARMVphPht9SqYXr16MW/ePFasWMGsWbM4efIk/fv3p6ys7Ff/zfTp0zEajRdusbGx1xxaKJCT07DriXo5tnAjZ3VRaJY6+j7YR3UcIa5IvQpm+PDh3HbbbSQmJjJs2DC+/fZbiouL+eKLL37130ybNo2SkpILt4yMjGsOLRSIimrY9cQVs9SZ2bI8C4CWQWcJT2ypOJEQV+aajtAHBwfTpk0b0tLSfnUdg8GAwWC4locRzqB/f4iJsR3Qv9RxGE2zfb1/f8dnc3P7562h1DsSvbmavn8arDqOEFfsms5xLC8v5/jx40TJX63uT6+HmTNtn2vaxV87f/+tt+QAfwOrqzax80fbLuh2UWUENpcTbITrqFfBTJ06lfXr13Pq1Ck2b97MzTffjF6v584777RXPuFMxo61nYrc9H+GhI+JkVOU7WTXO8up9A7Fp7ac5CnDVccRol7qtYssMzOTO++8k4KCAiIiIujXrx9btmwhIiLCXvmEsxk71nYqslzJb3emwlL2HNTAGzq1M2MIDVIdSYh6qVfBLFiwwF45hCvR6yElRXUKt/fjzOXUeEfgV1tE90kjVccRot5knAkhnFBZep5tQEugW69GePn7Kk4kRP1JwQjhhDb/azVmvS9BtXl0un+I6jhCXBUpGCGcTP6+kxwvsR3X7HVDU3RecnxLuCYpGCGczKb3fsCq8yLckkOb2+S6IuG6pGCEcCJZG/eRWWO7rqzPHQmK0whxbaRghHAimz/ZA5qOpvosYgclqY4jxDWRghHCSaQt2UyeFo1mMdN3QrLqOEJcMykYIZyAxWJhy7LTALQIOENEl1aKEwlx7aRghHACB+Z9R4lXY/RmE30fGaQ6jhANQgpGCMXMplp2bCoBoG3jEoJaRitOJETDkIIRQrFd//6WSu9QvOsq6C0DWgo3IgUjhEKmknL27Ld93ql1LX7hRrWBhGhAUjBCKLT1reWYvAPxqy2m+yMjVMcRokFJwQihSHnmWQ5mNAKgaw9fvBvJgJbCvUjBCKHI5n+uxuzlS2DtWRInDFMdR4gGJwUjhAKFB06RVhwOQK+hTWRAS+GWpGCEUGDTu7YBLcPMObSWAS2Fm5KCEcLBsjcdIL2mCQB9bm+HTie/hsI9yStbCAfb/PFu0HRE67JpNqSb6jhC2I0UjBAOdGLZFs4QjWY10/eBnqrjCGFXUjBCOIjFYuHHJScBiPM/Q2S3NooTCWFfUjBCOMihj7+n2KsxOnMNfR8ZqDqOEHYnBSOEA5hNtWzfUARAm4gijPFNFScSwv6kYIRwgN3vLqfCOwzvukqSp9ygOo4QDiEFI4Sd1ZRWkLrXAkDHeBP+kSGKEwnhGFIwQtjZtn8ux+QdhG9tCd3/JANaCs8hBSOEHdWUVnDopA8ASUne+AT4KU4khONIwQhhR7tmr6LGOwDf2mI6TxyiOo4QDiUFI4Sd1FWbOHDY9nnH9hpevga1gYRwMCkYIexk7/trqPY24lNbRpcHh6qOI4TDScEIYQeWOjN7d1cC0D6uBp+gRooTCeF4UjBC2MHBj7+nwjsMr7oquj0sWy/CM0nBCNHALBYLu38oAKB14zL8wo2KEwmhhhSMEA3s2MJNlHpHojPX0OPhwarjCKGMFIwQDWz36nQAWgYXEBgToTiNEOpIwQjRgE5+u50CfRSapY6eD8hUyMKzScEI0YB2fn0EgGZ+eYS0a6Y4jRBqScEI0UCyf9jPGaLBaqHX+F6q4wihnBSMEA1k+2epAER75RDRtbXaMEI4ASkYIRrA2d1pZNY1AaDn7zorTiOEc5CCEaIBbPtwC2g6Iq3ZNB2QqDqOEE5BCkaIa1R8LJPTlZEAdL9Jdo0JcZ4UjBDXaNucDVh1XoSac2kxUg7uC3GeFIwQ16A88yzHi0IB6Hp9U8VphHAuUjBCXIPt736HRe9DUG0erW+TCyuF+DkpGCGuUlV+KUdzAwFI6hOCTie/TkL8nPxGCHGVdr27ijovP/xrC+lwrwxqKcT/koIR4irUlFdx8KQ3AImdfdF7eylOJITzkYIR4iqkzl5JjXcgvrUldJ5wveo4QjglKRgh6slsqmX/QTMAHdpY8fL3VZzINc2dC5oGS5aoTiLsRQpGiHraN3c1Vd4heNdV0PUPMh3y1Th1CubMgd69VScR9iQFI0Q9WOrM7NleDkC72CoMxgDFiVyPxQITJsC//gUGg+o0wp6kYISoh0Pz11HuHY7eXE33P8ixl6vxxhvQty9066Y6ibA3OfVFiCtksVjYvSEPvBrTOrwE/8ahqiO5nP374auvYMMG1UmEI0jBCHGFTiz9kRKvxugstfR4KEV1HJe0caPt+Evrc2OC5ubCgw9CTg48/LDSaMIOZBeZEFdo5/JTALQIzCcoLkptGBf18MO2Mjl1ynbr3Rvee0/KxV1JwQhxBdJX7yRfF4VmMdPz/r6q4wjhEmQXmRBXYPuiQ0A0sYYzhCYMUR3HbaxbpzqBsCcpGCF+Q+6PB8m1RoPVQo9x3VXHcT1ms+3gS04OREVB//6g16tOJRzgmnaRzZgxA03TmDJlSgPFEcL5bJ2/E4AoXS5NerZTnMbFLFoEcXEwcCDcdZftY1ycbblwe1ddMNu3b2f27NkkJsr848J95e87SVZNEwB63JqgOI2LWbQIbr0VMjMvXp6VZVsuJeP2rqpgysvLufvuu5kzZw4hISENnUkIp7Ft7iasmp4ISw6xg7uojuM6zGaYPBms1l9+7fyyKVNs6wm3dVUFM2nSJEaOHMn11//2lcwmk4nS0tKLbkK4gpLjWZwujwCg28iWitO4mI0bf7nl8nNWK2Rk2NYTbqveBbNgwQJ27drF9OnTr2j96dOnYzQaL9xiY2PrHVIIFbbNWY9F501w3RlajOqlOo5ryclp2PWES6pXwWRkZDB58mQ+/fRTfH2vbIjyadOmUVJScuGWkZFxVUGFcKTKnAKOFwQD0HVgY5kOub6irvBC1CtdT7ikep2mvHPnTvLy8ujateuFZWazmQ0bNvD2229jMpnQ/8/phwaDAYMMmSpczPbZazDrIwisPUvbO29THcf19O8PMTG2A/qXOg6jabav9+/v+GzCYepVMIMHD2bfvn0XLfv9739Pu3btePLJJ39RLkK4orpqE8cyfMEbEnsEytbL1dDrYeZM29limnZxyWia7eNbb8n1MG6uXgUTGBhIx44dL1rWqFEjwsLCfrFcCFd16NN1mLwD8akto+O9w1THcV1jx8LChbazyX5+wD8mxlYuY8cqiyYcQ67kF+J/7N9SAPomxDepkumQr9XYsTB6tFzJ76GuuWDWyWBCwo3k/niQQn0TsFroMq6P6jjuQa+HlBTVKYQCsnNZiJ/Z/eUuwDYsTEi7ZorTCOHapGCEOKcqv5TTZbZZKjsNiVMbRgg3IAUjxDl7532HWe9Lo9oC4sfI7jEhrpUUjBCAxWLh8KFaANq20cupyUI0APktEgI4+c1Wyr3D0Zlr6Hxfiuo4QrgFKRghgL0rjwPQrFE+/o1DFacRwj1IwQiPV3wsi5y6xgAk3dJZcRoh3IcUjPB4qR//gFWnJ6Qul6b9O6mOI4TbkIIRHq2u2kRalm0w1g49ZNeYEA1JCkZ4tEOfrsfkHYRPbTkJ96SojiOEW5GCER7twJazAMQ3qcC7kYw7JkRDkoIRHit3yyEK9FHnxh3rqzqOEG5HCkZ4rN1f2MYdayLjjglhF1IwwiPZxh0LASBxcHPFaYRwT1IwwiPt+/B7zHpf/GsLZdwxIexECkZ4HIvFwqGDNQC0a61D5yWTXwlhD1IwwuOcWvazccd+n6I6jhBuSwpGeJw9K9MAGXdMCHuTghEepeR4Fjm1TQBIGpuoOI0Q7k0KRniU3R/+bNyxAVIwQtiTFIzwGBeNO9Y9RHEaIdyfFIzwGIfmb/hp3LFxA1XHEcLtScEIj3HgxzwAWjaWcceEcAQpGOERcrcdvjDuWNdxcmGlEI4gBSM8QurnOwFoouUS0l6GhhHCEaRghNurLijlVKntoH4nGXdMCIeRghFub++8n8Yda3Wz7B4TwlGkYIRb+/m4Y21bIuOOCeFAUjDCrWUfKTg37piJoGX/4uw/38JcVKA6lhAeQQpGuLUDm84AEF2+D31RHvn/nk3adX3Jf+JuzPnZitMJ4d6kYIRbqikp5siXSzmxMxeAPnGfEp1ciE9QLZYajbNf7yJt0CDyn7wbc2Gu4rRCuCcv1QGEaCi15aWcXrOetJ15nDrbFDOBAER77ycixg9u+AtBbUdRuvgz8j9cSE0xnF26i8IVKYSO6EHI46+hD22i9ocQwo1IwQiXVldZSfr360jbnsPJvCjqrI2AFgAYfc7SOt5E59GDIe5RADTA+PD/ETThb5S+93/kf7zIVjSLd1D4bQqhI3sS+vg/0IVEKvuZhHAXUjDC5Zirq8lYt4G0rZmczI2kxurP+VIJ9C6gdcsKWl2XSHhSCpru0nuBNW8fjJNeImjiM+eKZjE1JXB20XYKvx1A2MjehDz+D3TB4Q78yYRwL1IwwiVYamrJ2rCBY1tOcSI7ApMlAIgDIMCriPi4UloP6EBk918vlUvRfAwYH3mZoInPUDL7BfI/XUptCeR9tZWC//YjbFQyIX95TYpGiKsgBSOclqWujpxNP3Bs0wlOZIZSZQni/JaKv76E+GbFtO7Xhia9xqBd4/UtmsGX4EenY3zoOUrefd5WNKUaeV9uoeCbfoTd1JeQv/wdnTGsAX4yITyDFIxwKtY6M7lbt3Dsh6McTw+m0mzk/JaKr66M+JgCWvWJJ7rfKHReDf/y1Qy+BE+eYSuaWc+SP38ZtWUaeV9spuDrvoSN7kfIY69K0QhxBaRghHJWi4W87ds5tvEgx08FUl4XCtjGDDPoKmgZnUer3i2IGTACnY+3QzJpvn4E//k1jA+/QMm/nyX/s//aiubzTRR83ZfwMQMIfuxVdIEycZkQv0YKRihhtVjI37ObtHX7SDvhT2ltOOdLxVuromWTM7TqFUNsyhD0vurmbtF8/Ql+7B8Y//h/FL/zNwo+W05tucaZzzaSvySZ8JuvI/jPM6RohLgEKRjhUAX795L2/W7S0gwU10QCzQDw0qqJi8yldfcomg0eiJe/v9qg/0Pz9SfkL28Q/McXKX77b+R/voK6co0z8zdQsCSZsLEDCZ4yA12AUXVUIZyGFIywu+LDhzj23Q7SjnpRaGoMxAKgp4bmEdm06hZJ3ODr8A4MVBv0Cmh+jQh5/E2CH3mJ4n8+Tf6Xq2xF88k6Chb3ImzsIFvRNApSHVUI5aRghF2UHj9G2uqtHDuikV8VBTQFQEcdzUKzaNU1lBZDrsPHGKw059XS/BoR8uRMjH8qp+RfT5P/xWrqKjTOfLyWgkW9CL91MMbJ09H5O39pCmEvUjCiwZSdPsnx1T9y7KCZvMqmQDQAGmZigzNp1dlIi6ED8A0bqjZoA9L5BxDy5D8x/qmM4pnTKFj4HXUVkPvhd+Qv7En4bdcTPHk6ml+A6qhCOJwUjLgmFVmZHF/1A2kHasgpjwFsY3lpmIkOyqJ1YiNaDr0Ov8ghaoPamc4/kNBpbxP8aCnFbz1FwVffU1ehkTtvDfkL1xB++1CC//QKml8j1VGFcBgpGFFvVWdyOb5yA2n7q8gqbQqcH7fLQlRgFq0TfGk5tB+Not27VC5F1yiI0L/+m+DJJbaiWbSWunKN3P+souCLVYT9bhjBj7wsRSM8ghSMuCLV+XmcWLWBtL1lZBbHYOWnoVMaN8qkdQdv4of2ISD2eoUpnYcuwEjo32YRPLmY4jefJH/xemrLNXI/WEnBFysJv2M4xkkvofk619lyQjQkKRjxq2qKizi5aj3HUovIKGyKhVAgFIAIv2xatddodX1vgloOUhvUiekCgwl9djbBfy6i+I0nyV+ygdoyjZw5y8lfsJzwO0di/OP/SdEItyQFIy5SW1bKqdXrOLYrn/T8aMwEAbZTbsN8c2nVxkyrwT0IbiulUh+6wBBCn3uP4CkFFL35JAVLfrAVzXv/JX/Bfwm/axTGh/8PzaDuolIhGpoUjKCuopzT360nbUcup85GU2cNAGxnPYUY8mjVqoZWg7oSmiClcq10xjDCnn+fkD8XUPT6ExR8vYnaUo2cd78hf/43hN99E8Y/vCBFI9yCFIyHMldVkb52A2nbsjh5pjG1Vj/Oj1Qc5J1Pq/gqWqckEpZYv+HvxZXRGcMI+78PCHksn6LXH6fgmx9tRTPrawrmf0P4PaMJevA5KRrh0qRgPIi5xkTmuo2kbU3nRPb5ibriAAjwKqRVizJaD0ggopuUiqPogsMJe3EuIX85S9Frj1OwbAs1JZD9zhLyP11K+D1jCHroeTRvH9VRhag3KRg3Z6mrI2vjRtI2n+REVhjVlkDOl4q/vphWzUto3b8djXuORdNLqaiiC44g7OV5hEzNo/C1xyn871ZqiiH77cXkf7KE8HtvIWjiM1I0wqVIwbghc20dh5ZtpvBIGmkZIVT9bE4VP10p8bFFtOrbiqg+o9Fd40RdomHpQiIJf+VDQqbmUvT3xylcvt1WNP9cSP7HXxF+760ETfibFI1wCVIwbqT6bB475q9lz6GIc0viADDoyolvmk+r5BY07TfSYXOqiKunD21C+IyPCXk8h6K/P07B8h3UFEH2zC/J/+grIu67jcAH/oZmh0nXhGgoLvfqrKmzAHA8vYTWVU3w95M3y7zt29i3fB/HsptiJuLCcqN/Gf1v8CcmZRh6H4PChOJq6cOiCH/1E0KmZtt2na3YSU2Rhaw3P8fnwy+JuO92Au//qxSNcEou96o8crYMXyBnax5ztp6h3F+HX5Q/zduG0KtHFDFRnjF6bV1lJceXrWTvlvJzA0vazgCL8Msmtm0gXe8chMHoGc+FJ9BHRBPx908JfTyLwr8/TuGKXdQUWsh6YwGGD78g/Pd3EHjfNCka4VRc7tWYG+VDzZkKYup0+Fs1giutcLyC3OMVLP02kzJvIMyHxi2NJHaOJCkhAr2X+xy8Lj15nANLNnLwWAjVFiNgREctraKy6DSsA4173iVngLkxfURTIl6bT+jjGRS++jiFq1IxFVjI+sd8DHM/J/z+Owgc/5QUjXAKLvcqrGmkZ2mjGv46oh19Ggeze1cuWcdLsORVE1QDgbVAbg2VuWfZsvks6zUrVUFeBMU0ok1COL26RxEc5Fq7i6wWC5lr17Hv+1OcKojFem4WyACvQhISqugw5nr8o4YpTikcSR8ZS8TrCwg9k24rmtV7MBWYyXrtUwxzFxD+wN0E3vsEml5O4hDquFzBnKfT6UhoG0ZC27ALy/IKKtm6I4cTBwspz66gUbkZg1XDUGKGklLSDpRy9IvjlPlqeDf2I6ZVMN17RBHfLAidE/7Vbyoq5PDiVexP5dz0wnEAxBjT6dQ/kriho+WAvYfTN25GxBufE5p7moJXp1K0Zh+mfDNZr36E4T+fEvHAPQTc+4Rs1Qol6lUws2bNYtasWZw6dQqAhIQEnn32WYYPH26PbPUWGebPqGHxMCwegJqaOnbuPcv+PXnknypFX1hLI7OGsRo4XUXh6SpWfZdDhd5KXagP4c0DSUiMoHtSYww+6rq3YO8e9i3byZGMJtRZbUPhe2tVtIs7Q8cbe8qQLeIX9E2aE/nml4TlnDpXNPsxnTWTOeNDDB98QsTEewm4Z6oUjXCoer2LxsTEMGPGDFq3bo3VauXDDz9k9OjR7N69m4SEBHtlvGo+Pl4kd48iuXvUhWXHTxezY2cuGUeLqDlTTWCVhUZmDc7WUnu2kNQdhWznMBUBehpF+9OifSi9u0fROMK+83eYa0yc/HYV+zYVkl0Wy/mtlRDDGRK76WgzepjLTi8sHEcfFUfkWwsJzT5J4YypFH1/wFY0r8zF9/2PCX9wPAF3PSZFIxyiXgUzatSoi+6//PLLzJo1iy1btjhlwVxKfPNg4psHX7hfUmZi285cjh7IpySjHN+SOgxWjeByCxwtJ/NoOQuXplPqA7oIA1HxRpK6NKZj27AG2a1WkZXBgcVrOXgogApzMNAIDTMtIzLodH080f1/J28Got68olsQ+c+vCM08bjtGs/Yg1Xl1ZL70Ab5zPiT8wd8TcOcUeW0Ju7rq/UBms5kvv/ySiooKkpOTf3U9k8mEyWS6cL+0tPRqH9IujIEGhqQ0Z0hKcwDMdRb2Hc4ndfcZck+UQr6JwFoIqgGyTJRl5bFxQx6rdVZMRi+CmwXSNiGcXt2aENDoyq6utlos5GzezL5VRzmRF4OFGAD89CUktC0jYfQAApp73myQouF5xcQT+a9FhGamUThjKoVrD1N9po7MF+fgO2ceEQ/dT6PfPSpFI+yi3gWzb98+kpOTqa6uJiAggMWLF9OhQ4dfXX/69Om88MIL1xTSkfReOpI6RpLUMfLCsuwzFWzbns2Jw0VU51TSqMKMr0XDt8iMtaiYw3uKOTD/GGV+OgxRfjRvE0LP7lE0jwm66HvXlpVydOlK9u2opaC6Ced3g0UFZNIxOZj4kcPR+8rouaLhecW0IvLtJYRmHqNg+lSK1h2hOreWjBdm4/vef4h4aAKNbn9EikY0KM1qtVrr8w9qampIT0+npKSEhQsX8v7777N+/fpfLZlLbcHExsZSUlJCUFDQJf/N5UxesJulqdk8c2MHHujXot7/viFUVdexfXcuB/edpfB0GT5FtfhZtF+sV+4FllAf+nbyRnfyIIdORpwbwRi8NBNtYnLoOCKJiC5dHf0jCA9Xl36UghlTKVp/FKvZ9tr1i/Im/OGJNLp1khSNm8qa+jily5bReNpThI4ff1Xfo7S0FKPReEXv4fXegvHx8aFVq1YAdOvWje3btzNz5kxmz559yfUNBgMGg2tdd/Jb/Hy9GJAcw4Bk264ti8XC0RPF7NqZS2ZaMea8agJNVgLqNMirYd93NYBtF1yQdz6dOptpd/P1+IZFXOZRhLAfr2ZtaPzvrwk7fZiC6Y9TtPEYVTm1ZDz7b/zefZ/wPz5Io7EPS9GIa3LN5+JaLJaLtlA8kU6no12rUNq1Cr2wrKC4iq3bczCs+BcVtTHg403noR1oNugWNBnBWDgJr+btaPzuN4SdPEjBjCco+iGNquwaMv72Nn6z5hAx6Q/4j3lQikZclXq9aqZNm8aGDRs4deoU+/btY9q0aaxbt467777bXvlcVliwHyOGtMS/o4GbQv+PxKA5xA4eKOUinJJXiw40nr2M+GVfETKgJZrOSlWWifSnZ3J6aFcqFs3GarGojilcTL0KJi8vj3vvvZe2bdsyePBgtm/fzsqVKxkyRM54+jUdxkylzOpHnCWdPd99pjqOEJfl3SKBJu/9l/hvviSkXwtb0WSaSH/6LdKHdaNi6fuqIwoXUq+C+eCDDzh16hQmk4m8vDzWrFkj5fIbjCHh7I/5HQABW9+UvwKFS/CO70ST978l/uvPCekXh6azUplRTfqTr3N6aBcqvv5AdUThAmTHqgO0Hf0ElVYDrc1p7NuwWHUcIa6Yd6vONHl/OfFLFhDSp5mtaNKrSX/iH5we2oXKZfNURxROTArGAUIjm7K38c0AeG16Q3EaIerPu00STf6zkvgl8wlOjoVzRXN66qucvqErlf/9UHVE4YSkYByk5ehp1Fi96FC7n4M/LlcdR4ir4t2mK1FzV9Hqq08I7h1jK5pTVZz+ywzSb+hG5fJPVEcUTkQKxkEim8axO3wkAHXrX1OcRohr492+O1HzVtPqq48J7tUUNCsVpyo5/eeXSR/RnaqV81VHFD9TXpLP9+8+Q/rm1QBklWc75HGlYBwo9sanqbPqSKzeydFd61XHEeKaebfvQdSHa4hf+CHBPaNtRXOiglOTX7QVzeoFqiN6LIvFQuqq+Sx7YDjH+vUn6q2FBBeaqNNBWqRjTjaSgnGg6Bbt2B1sO+uuYs2ritMI0XB8EnoR9dF3xH85F2P3qJ+K5k8vkD6yB1VrvlAd0WNkn9jH8ucn8EP/JAyPvkj8plP41sLZMC++HRrCIw/rKU9o5pAsUjAOFjniaSxWjS6Vmzh5YKvqOEI0KJ+OyUR/8j3xn/8HY7dzRXO8nFOPPEfGqJ5Uff+l6ohuqaqylA3zprNiTB+KRtxO3IJNRBTUUuUDx/vFYfrXs/TbuIe8W/tTGPTLcRPtxWWnTHZVzdsmsSuwP13LN1Cw4lVaJCxSHUmIBueT2IfoT78nfM9G8l99lpLdOZQfK6P8j88S0Po1wv/yJH4pt6iO6dIsFguHNn/Dyfkf0GRzGhHVP41bnB4fiM9NN9DjjkfpagxXllEKRoGgodNg0Qa6lH5PRto+Ylt1Uh1JCLvw6dyf6PlrCU/dYCua1Fxb0fzhbwS0+Tvhjz2FX8rNqmO6lLNZaez66E18Vm6iSa6J+HPLC416SgZ3IeGeRxjWoZfSjOfJLjIFWiX2YY9fL/SalZxlr6iOI4Td+SQNIPqzdbT8dDZBSZGgWSk/WsqpPzxNxujeVG9cqjqiU6sxVbJ5wVt8e/sAcoeMotmH39Mk10SNFxzv2ZSyvz9Grx92MeKVj2nuJOUCsgWjjGHQE/DfW+hStJLc9GM0adZadSQh7M7Q9TqaLlhP+K615L/6HKV78ig/UkL5xKcIbDed8Kl/xbffqN/+Rh7i2I7vOPrpu4RvOEBIhZWQc8uzmvnDiEF0v2cKncObKs14ObIFo0i7HtdzwKcz3pqZ019PVx1HCIcydB1I08830PKTfxOUGA5YKTtcwskJT5B5czLVm/6rOqIyxfmZrH5zKmuGdKPunkdouXw/QRVWSgJ0nBjRCa9P3+b6VTu5fsprBDtxuYBswShlHTAV1oyj89mvyc9NJ7yJY04dFMJZGLoPoukXgwjftpr8116gdF8+ZYeKKXtgKoEdXiZ86t/w7TNCdUy7q6utYde38zi78HNidmUTYz63XAcZiZEEj72F7qMn4GPwVxu0nmQLRqGEPjdyxKsdvlotaUtmqI4jhDKGnkNo+uUPtPxwJoEdQwErZQeLOHn/X8i8pS+mrStVR7SL0we28O1T97Ctb1cCn3yTltuz8TFDbhMD6fcNpsnqbxixYD19bn/U5coFZAtGKU2noyr5z7DxITrlfEVJwXMYwxqrjiWEMoZew4hZOIzqH1eQ/48XKTtQaLuNn0xQx3DCn3gOQ0/XniKkvCSfbfPfou6bVcSeKKPFueUVvhq5fVvT4q4HuC75RnRuMIuo6/8ELq7zwNs5rm9BI62ag0tkjDIhAHyTbyDmq020+M/rBHYIATRK9xdw4t4/kXVbP0zb16iOWC8Wi4Xdqz5l2QM32IZtmfkVsSfKsGhwqkMIZ5++l4TNW7jxnaUk9L3JLcoFZAtGOU2no7jbZNg2hYSM+ZSVPE2gMVR1LCGcgm+fEcQsGkH1D9+Q//orlB0qpnRfAaXjHiGocwThj7+Aofsg1TF/Vfbxvez5eCYBq7YRXlh34ZqVs2HeVAztRed7HyWhhfteBycF4wSSho4jfcdrNLNk8ePSN0m+90XVkYRwKr79RhHTbxTVG7/m7OuvUH64hNI9+ZSO+yNBiZGEP/kihq7XqY4JQGXRWfa/+CSFJw4Te7iIuHPLq3wgu2ccMb+7l36Df+c2WymXIwXjBPReXpxJ/CPNUv9KmxMfUl35JL7+AapjCeF0fPvfRGz/m6jesISzr8+g/EgJpXvOUnr3QxiTGhP+5Iv4JA1weK68zKPs//pDzCvXEXOkkEAg8NzXnGXYFhWkYJxE0oiJZO+ZSbQ1jy1L/0nvO59WHUkIp+U7YAyxA8ZQtW4x+W/MoPxoKSW78yi580GMSU0If+pFfDr3t9vjWywWju1cw4n/foHXj6lEn64g6n/WKW4bRfT06U4zbIsKUjBOwtvHQEb7B4k++BItjrxPjekxfAy+qmMJ4dT8Um4mNuVmqr5fSP6bf6f8WBklu89QcsdEjF2jCH/yJXwS+zbIY9VUVZK6+lPyVn9L8I5jhBWZL+z+AsiK8aMmOZEWycNoO/Q2dF7y9irPgBPpPOqPnD34Do0pYNuyWfS85c+qIwnhEvwG3UrsoFupWvMFZ996jYq0ckp25lLyuwcwdou2FU2nPvX+voW5p9n7zVwq1q4nan8ugTU/7fqq8YKsduH4DOhDx9Hjad+8Q8P+UG5ACsaJ+Po1IrX1/UQce53o/e9Sd9MkvLx9VMcSwmX4XX87za6/narVCzj71utUHC+nZEcOJbffj3FwL8KfegWfmMsPr3J870aOffMpuk27aHqyjMY/jYJPSYCO/K5xhA4eSucR4+gcKGd8Xo4UjJNJHD2Zon+8R4w1l50r/kO3UX9QHUkIl+M35A6aDbmDqpWfcfbVZ6jI9qZkzTZK1t1A8M03E/bQQxeKprammr3ff0nOyqUEbTtKREEtzX/2vXKiDFT16kiz4bfQo98o9Hp527xS8kw5Gf8AI3vi7iH51CzCdr+NZcREdHq96lhCuCS/Vk1oNuAslUUB5BcPomLrDoq//JLixYspG9yNvOJMmuzJplG19cI1KnU6yGoTgq5/LzrcdC/tW3dR+jO4MikYJ9Rh9FTK3ppHnCWDXWvm03XYONWRhHBNG/4BgP8N96D1e4TDqz7B78OlND9aQuDKrReOp5T7aZxJisU4eDCdR46nU4gM2dQQpGCckDEknB9j7iA5ay6B297COuRuNA+4KEuIhnTi0GKO5u/icGgo66r2c3zRuVGZb4F2GXpuOuRPaFQLmt4whi7X3SLHO+1ACsZJtR39BJXvzKe1OY296xeROPBW1ZGEcBlmi5nR256FyHMXNpalo9f0dG3clZSYFFJuTqFZkEyPYW9SME4qNDKaLU1upveZBXhvfgOkYIS4YoeP/jQFc6fg1tzTaQJ9m/bFaDAqTOV5ZL+LE4sfPY0aqxftaw9w8MflquMI4TLe3/4GADd4hTF/9CJGtBwh5aKAFIwTi4iOY3f4jQCY1/1dcRohXEPa8dWssZQA8Ifef1WcxrNJwTi52BunUWfV0cm0i6O71quOI4TTe3/LKwBcrzMSH+/ak5O5OikYJxfdoh27g22/JBVrZFplIS4nI2MTy2vPAjChx2OK0wgpGBcQOeJpLFaNLpWbOXlgq+o4QjitD354AYum0Y9GJLQbqzqOx5OCcQHN2yaRGmib46JwhWzFCHEpubmpLDVlA/Bgl0cUpxEgBeMyjEOfAiCpdC0ZafsUpxHC+czb8Ax1mkZ3q4EuifeojiOQgnEZ8Yl9SPXrjV6zkrPsFdVxhHAqBflH+ariJAATOz6gOI04TwrGhfgOegKALkUryU0/pjiNEM7j4/VPU63T6GTxIrnrQ6rjiHOkYFxIux6D2W9Iwlszc/pr2YoRAqCkJJ0FpYcBmNhOxu1zJvI/4Wr6/wWApLPfkJ+brjiMEOrNXzuNCp1Ga4uO63pOUR1H/IwUjItJ6HMjh73aY9BqSVsiZ5QJz1ZZnsenRXsAeDB+LDqZDMypSMG4GE2nw9RnCgCJOQspzs9VG0gIhb5Y9zQlOo04MwzpM011HPE/pGBcUGLK7RzXt8RfM3Fo6Wuq4wihRHVVEfPObgHg/ubD0XvJfC7ORgrGBWk6HcXdHgUgIeMzykoKFScSwvEWr3+GAp1GlNnKjQOeVx1HXIIUjItKGjqO07oYgqhg/9I3VccRwqFqayuZm70OgPujB+Lt7a82kLgkKRgXpffyIi/xjwC0PTGPqooyxYmEcJxlG54nR68RbrZyc8pLquOIXyEF48KSRkwgW4sklFL2fP1P1XGEcAhzXQ0fnLZNwDc+sjcGX5lIzFlJwbgwbx8DGe1tVy23OPIBNdVVihMJYX+rNr3CaT0YLVZuT5ELjp2ZFIyL6zzqj+QRSmMKSF02S3UcIezKYq7jvROLAbgnNAn/gEjFicTlSMG4OF8/f060/j0ATQ+8S11tjeJEAqC6GsaMgTZtoHNnGDIE0tJUp3J967a+SZrOQiOLlTtl68XpScG4gcTRkykiiKbWM6Qu/4/qOOKcBx+EI0dgzx4YPRomTFCdyLVZLRbmHJkPwB1B7TAamylOJH6LFIwb8A8wcjjONv9FeOrbWMxmxYmEry+MGAGaZrvfuzecOqU0ksv7cde77NfV4WuxMi5luuo44gpIwbiJhDFTKcWfOEsGe9Z8ojqO+B8zZ9q2YsTVm7PftnV+a6MWhIW1VpxGXAkpGDcRFBzGgaa/AyBw20ysFoviROK8V16xHX+ZLn90X7Vdez5ih2bCy2pl/IAXVccRV0gKxo20G/MklVYDrczH2bd+keo4AvjHP2DRIli+HPzlYvOr9l7qOwCM9m1KkyZJasOIKyYF40ZCIqLY22QsAN6bX5etGMXeeAM++wxWr4bgYNVpXNeBw4vYRCV6q5UH+j6vOo6oBykYNxM/+ilqrF60rz3IwS0rVMfxWJmZ8Je/QHExDBwISUnQq5fqVK5pzvY3ABjuHUFsbLLiNKI+ZHYeNxMRHcfW8BvpVbAEy/rXoM8I1ZE8UkwMWK2qU7i+tOOr+M5SAsCE3k8rTiPqS7Zg3FDsqKeps+roZNrF0V3rVMfxHGYzrFtn2y+2bp3tvrgm72+xnRkxRGckPn6I4jSivupVMNOnT6dHjx4EBgYSGRnJmDFjOHLkiL2yiasUHdeW3cFDAahY86riNB5i0SKIi7PtD7vrLtvHuDjbcnFVMjI2sbz2LAATez6uOI24GvUqmPXr1zNp0iS2bNnC6tWrqa2tZejQoVRUVNgrn7hKjUdOw2LV6FK5mZMHtqqO494WLYJbb7UdePm5rCzbcimZq/LBDy9g0TT60Yj2beUiIldUr4JZsWIF9913HwkJCXTu3Jl58+aRnp7Ozp077ZVPXKVmbZLYHXgdAIUrZihO48bMZpg8+dIHXM4vmzJFdpfVU27ObpaasgF4sMsjitOIq3VNx2BKSmwH30JDQ391HZPJRGlp6UU34RjGYU8B0KV0LRnH9ihO46Y2bvzllsvPWa2QkWFbT1yxeRufpU7T6GE10CXxHtVxxFW66oKxWCxMmTKFvn370rFjx19db/r06RiNxgu32NjYq31IUU+tOiWT6tcbnWYl57+yFWMXOTkNu54gP/8wCytOAjCxk4wQ6squumAmTZrE/v37WbBgwWXXmzZtGiUlJRduGRkZV/uQ4ir4Dn4CgC5FK8k5LSdkNLioqIZdT/Dx+r9i0ml0snjRu8uDquOIa3BVBfPII4+wbNky1q5dS0xMzGXXNRgMBAUFXXQTjtOu+2D2G5Lw1sykfyNbMQ2uf3/bRS/nh03+X5oGsbG29cRvKilJ5/NS2x9CE9vdjaaTKylcWb3+96xWK4888giLFy/m+++/p0WLFvbKJRrSgKkAJJ39hvzs04rDuBm93jZUMvyyZM7ff+st23riN81fO40KnUYbi47rev1ZdRxxjepVMJMmTeKTTz5h/vz5BAYGkpubS25uLlVVMhe8M0tIHslhr/YYtFrSvpbrYhrc2LGwcCE0bXrx8pgY2/KxY9XkcjEV5bl8UmQ7GWVi/Fh0OillV1evgpk1axYlJSWkpKQQFRV14fb555/bK59oAJpOh6nPFAAScxZSnJ+rNpA7GjvWNqPY2rUwf77t48mTUi718MW6v1Kq04gzw5A+01THEQ2gXmORWWVwJZeVmHI7xze9Srz5BHuX/J3eE95QHcn96PWQkqI6hUuqririw7ytoNd4oPkI9F4+qiOJBiBH0DyEptNR0v1RADpkLqCspFBxIiF+smj9MxToNaLNMHLAc6rjiAYiBeNBkobey2ldDEFUsH/J66rjCAFAramCudnrAPh90xS8vWVmNnchBeNBdHo9eYmTAGh78iOqKsoUJxIClm18nly9RrjZys3XvaQ6jmhAUjAeJmnkBLK1xoRSyt6vZ6qOIzycua6G99NtE+Pd1zgZg69RcSLRkKRgPIy3tw8ZHWxXR7c48gGm6krFiYQnW7npZdL1YLRYue26l1XHEQ1MCsYDJY36I3mEEkkhe5a9qzqO8FAWcx1zTiwB4J7QLvgHRKoNJBqcFIwHMvj6c7z1/QA0PfAudbU1ihMJT7Ru6xuk6SwEWKzclfKK6jjCDqRgPFTn0Y9SRBBNrWdIXf6B6jjCw1gtFuYc+QyAO4LaEWSUUdbdkRSMh/IPMHIkbhwA4anvYJEJsYQD/bhzFvt1dfharIwbKIOwuispGA/WYcxfKMWfOEsGe9Z8ojqO8CDvHZgLwK0BLQkNbaU4jbAXKRgPFhQcxoGYOwAI3DYTq8WiOJHwBDv3fMROzYS31cr4AS+qjiPsSArGw7Ub/QSVVgOtzMfZu/4r1XGEB5iT+g4Ao32b0qRxZ8VphD1JwXi4kIgo9jaxjfjrs/kN2YoRdnXg0FdsohK91cr9fZ9XHUfYmRSMIH70U5is3rSvPcjBLStUxxFubM4O2yjew70jiY1NVpxG2JsUjCAiOo7U8JEAWNb/XXEa4a7Sjq/iO0spmtXKhOSnVccRDiAFIwCIHfU0dVYdnUy7Obprneo4wg3N2WK7mPJ6fQjxLa9XnEY4ghSMACA6ri27gocBULFGplUWDSs9/QdW1OYDMLHnVMVphKNIwYgLmox8CotVo0vlZk7s36o6jnAj/9n0f1g0jf5aI9q3Ha06jnAQKRhxQbM2SewOvA6AopXTFacR7iI3ZzdLTdkAPNjlT4rTCEeSghEXCR42DYAupevIOLZHcRrhDuZufIY6TaOH1UBSp7tVxxEOJAUjLhLfqTepfsnoNCs5/5UxosS1yc8/zFcVpwCY2GmC2jDC4aRgxC/4Dn4CgC5FK8k5fURxGuHKPl7/V0w6jUSLF727PKg6jnAwKRjxC+26D2K/IQlvzUz6N7IVI65OSfEpFpTa/kCZ2H4cmk7ebjyN/I+LSxvwOABJZ78hP/u04jDCFc1f9zSVOo02Fh3X9ZyiOo5QQApGXFJC8ggOe7fHoNWS9rVsxYj6qSjP5ZOivQBMjL9Ftl48lPyvi0vSdDpMyY8BkJjzFcX5uYoTCVfy+bqnKdVpxJk1hvR5SnUcoYgUjPhViSm3kqaPx18zcXiJjFEmrkx1VREf5W0D4IG4Eei9fBQnEqpIwYhfpel0lPZ4FIAOmQsoKy5QnEi4gkXr/0aBXiPaDCP7P6s6jlBICkZcVtKQcZzWxRJEBQeWvq46jnBytaYK5mavB+D+pgPx9vZXnEioJAUjLkun15PX+Y8AtDn5MVUVZYoTCWf2zcbnydVrRJitjLlOpkP2dFIw4jd1GTGBbK0xoZSyZ+lM1XGEk6qrreaDdNuEdeMbJ2PwNSpOJFSTghG/ycvbh4wODwHQ8ugHmKorFScSzmjV5umk68FosXJbyiuq4wgnIAUjrkjSqIfJI5RICkn9ZpbqOMLJWMx1zDmxBIBxoV3wbxShNpBwClIw4ooYfP050eZ+AGIOzqautkZxIuFM1m59nTSdhQCLlTtl60WcIwUjrljiTY9SSBBNrWdIXf6B6jjCSVgtFuYc+QyAO4ztCTLGKk4knIUUjLhi/gFGjrQYB0BE6jtYzGbFiYQz+HHnLA7ozPharIxLkYnqxE+kYES9JIz+C6X409ySQerqj1XHEU5g9oG5ANwa0JLQ0FaK0whnIgUj6iUoOIwDMXfYPt/+T6wWi+JEQqWdez5kl2bC22rlvgEvqY4jnIwUjKi3dqOfoMJqoJX5OHvXL1QdRyg0J/XfAIz2bUrjxomK0whnIwUj6i0kIop9TcYCYNj8pmzFeKgDh75iE5XorVbu7/eC6jjCCUnBiKvSavQ0TFZv2tUe5OCPy1XHEQq8t+MNAEb4RBIb01txGuGMpGDEVQmPbs7uiBsBsGx4TXEa4WjH0lbwvaUUzWplQvLfVMcRTkoKRly1ZjdOo9aqp5NpN0d2fK86jnCg97faZjm9Xh9CyxaDFKcRzkoKRly16Li2pIYMBaDqe5mQzFOkp//Aitp8ACb2nKo4jXBmUjDimjQe8RQWq0ZS5Y+c2L9FdRzhAB9segGLptFfa0T7tqNVxxFOTApGXJNmbZLYFZQCQNEKuYrb3eXm7OZrUw4AD3Z9VHEa4eykYMQ1Cx02DYAuZetJP5qqNoywq7kbn6FO0+hpNZDU8S7VcYSTk4IR16xlx16k+iej06yc+XaG6jjCTvLzD/NVxSkAJnZ6UG0Y4RKkYESD8Bv0BABJRavIOX1EcRphDx+tfxqTTiPR4k2vLhNUxxEuQAqmHh59FOLiQNMgNVV1GufStvsg9hm64K2ZSf9ajsW4m5LiU3xeehSAie3vQdPJW0dDGzoUEhMhKQn694fdu1UnunbyKqmHW2+FH36A5s1VJ3FOugGPA5CUv4z87NOK04iG9Om6p6nUabS16Liu5xTVcdzSF1/A3r22P14fewzuu091omsnBVMPAwZATIzqFM6rQ/JwDnt3wKDVkrZUtmLcRUV5Lp8W7QVgQqtbZevFToKDf/q8pMS2p8TVyStFNBhNp8OU/BgAibmLKD6boziRaAifr32aUp1GnFljSPKTquO4tXvvhdhYeOYZ+NgNpluSghENKjHlFtL08fhrJg4vlav7XV11VREf5m8DYELcCPRePooTubePPoKMDHjpJXjSDbpcCkY0KE2no7SH7QK8DpkLKC0uUJxIXIuv1v2NQp1GUzOM6P+s6jgeY/x4WLsWClz810cKRjS4pCHjOK2LJYhKDix9XXUccZVqTRXMzVkPwO+bDsTb219xIvdVXAzZ2T/dX7IEwsIgNFRVooYhBVMPDz1kO8ifmQnDhkErmX78knR6PXmdJwHQ7uTHVJaXKE4krsY3G5/jjF4jwmxlzHUvqo7j1kpKYMwY6NQJOneGt9+GZctc/0C/l+oArmT2bNUJXEeXEQ+QlfomTa1n2PL1P+l91zOqI4l6qKut5v30laCH8Y37YPA1qo7k1po3h23bVKdoeLIFczlmM6xbB599ZvtoNqtO5DK8vH3I7PAQAC2P/gdTdaXiRKI+Vm56hQw9BFus3DbwFdVx3JMHvL/Uu2A2bNjAqFGjiI6ORtM0lixZYodYTmDRIttl+wMHwl132T7GxdmWiyuSNOph8gglkkL2fPNv1XHEFbKY63j/xBIA7gntgr9/uNpA7shD3l/qXTAVFRV07tyZd955xx55nMOiRbbL9jMzL16elWVb7mYvAnsx+Ppzos0DAMQcfI+62hrFicSVWLv1ddL0VgIsVu5Mka2XBudB7y/1Lpjhw4fz0ksvcfPNN9sjj3pmM0yeDFbrL792ftmUKW65OWsPiaMfpZAgoq1n2P3t+6rjiN9gtVh478hnANxp7ECQMVZxIjfjYe8vdj8GYzKZKC0tvejm1DZu/OVfFj9ntdquhNq40XGZXJh/oyCOtBgHQOSed7C4yS+Ou9q8898c1Jnxs1i5J0WG+2lwHvb+YveCmT59Okaj8cItNtbJ/yLKucLhTa50PUHCmKmU0ojmlkz2rPpIdRxxGe8dmAvALQEtCQ2NV5zGDXnY+4vdC2batGmUlJRcuGVkZNj7Ia9NVFTDricIMoZyMOYOAAJ3/AurxaI4kbiUHanz2KXV4G21ct+Al1XHcU8e9v5i94IxGAwEBQVddHNq/fvbrqb8tSucNM02Gl3//o7N5eLajXmCSquBVubj7F23UHUccQlz9tjO9BvjG0Pjxp0Up3FTHvb+ItfB/C+9HmbOtH3+vy+C8/ffesu2nrhiweFN2BN1CwCGH9+QrRgns//gl2ymCr3Vyv39nlcdx3152PtLvQumvLyc1NRUUs9N6Xjy5ElSU1NJT09v6GzqjB0LCxdC06YXL4+JsS0fO1ZNLhfX+qanMFm9aVd7iIM/fqs6jviZOTvfAmCETyQxMb3VhnF3HvT+Uu+hYnbs2MHAgQMv3H/sMdv8H+PHj2fevHkNFky5sWNh9Gjb2Rw5ObZ9ov37u81fFiqERzdna8QoeuUvwrrhNeh7o+pIAjiatoLvLaVoVisTkv+mOo5n8JD3l3oXTEpKCtZLncPtjvR6SElRncKtNBv1FLX/WUpHUypHdnxH2+6DVUfyeO9vsZ2OfL0+hJYtBilO40E84P1FjsEIh4pq3pbdIcMAqP5OJiRTLT39B1bW2SYdmdjzccVphLuRghEOFzXyKcxWjc5VWzi+b4vqOB7tg00vYNE0BmgBtG97k+o4ws1IwQiHi23dmdSgFACKV8rV4qrkZO/ka5Ptgr6JXf+kOI1wR1IwQomQYdMA6FK2nvSjqWrDeKi5G5+lTtPoiS9JHe9SHUe4ISkYoUTLjr3Y7Z+MTrNy5lvZinG0/PzDLKo8DcDEjhMVpxHuSgpGKOM/+EkAuhStIvvUEcVpPMtH657GpNNItHjTq8sE1XGEm5KCEcq07TaQfYYueGkWMr6ReUccpaT4FJ+XHQXgwQ7j0HTyNiDsQ15ZQinddU8A0CV/GfnZp9SG8RCfrp1GpU6jrUXHgB6TVccRbkwKRijVofcNHPLugI9WR9rSGarjuL3y8lw+Ld4HwMRWt8rWi7AreXUJpTSdjpo+tuGGEnMXUXTWPebBcFafr51GqU4jzqxxffKTquMINycFI5RLvO4W0vTx+GsmDi+Vq/vtpbqqiI/ytwMwIW4kei8fxYmEu5OCEcppOh2l544FJGQuoLS4QHEi9/TVur9SqNNoaoYR/Z9RHUd4ACkY4RSShtzDKV0sQVRycMnrquO4nVpTBXNzNgBwf8wgvL39FScSnkAKRjgFnV5PftIkANqe+pjK8hLFidzL1xue5YxeI8JsZfR1L6qOIzyEFIxwGknDHyBLa0wIpexdOlN1HLdRV1vNBxmrABjfpA8Gg5NPWy7chhSMcBpe3j5kJvwBgJbH5mKqrlScyD2s3PQKGXoItli5LUUuaBWOIwUjnErSjX8gj1AiKST1m3+rjuPyLOY65pxYAsC4sK74+4erDSQ8ihSMcCoGX39OtHkAgNiDs6mrMSlO5NrWbvkHx/VWAixW7pCtF+FgUjDC6XQePZlCgoi25pG6/H3VcVyW1WLhvaMLALjT2IGgoBjFiYSnkYIRTsevUSBHWtwLQMSef2MxmxUnck2bdr7DQZ0ZP4uVe1JkSgTheF6qA9TX0tRsAF5cdpDMIjkI7K5qDSNJ5m2aWzKp+kd7/BLHqo7kcuZkLwcvCPVuxJy0harjCCdwsOCgQx/P5Qrm5+ZuOqU6grCjRw3BRGrF+FWdga2zVMdxKZWaxq64WACyzJV8cugTxYmEMwnwDnDI47hcwQzv2ITl+3Pp1SKU7nEhquMIO1pf9Dy3HZ5CdUxffFv0Vh3HpfgD/6jM4Ih/EIQ0Vx1HOBGjwciwuGEOeSzNarVaHfJI55SWlmI0GikpKSEoSC74EkIIV1Kf93A5yC+EEMIupGCEEELYhRSMEEIIu5CCEUIIYRdSMEIIIexCCkYIIYRdSMEIIYSwCykYIYQQdiEFI4QQwi6kYIQQQtiFFIwQQgi7kIIRQghhF1IwQggh7MLhw/WfH7y5tLTU0Q8thBDiGp1/776SgfgdXjBlZWUAxMbGOvqhhRBCNJCysjKMRuNl13H4fDAWi4Xs7GwCAwPRNM2RD31VSktLiY2NJSMjQ+av+Rl5Xi5NnpdLk+fl0lzxebFarZSVlREdHY1Od/mjLA7fgtHpdMTExDj6Ya9ZUFCQy7wAHEmel0uT5+XS5Hm5NFd7Xn5ry+U8OcgvhBDCLqRghBBC2IUUzG8wGAw899xzGAwG1VGcijwvlybPy6XJ83Jp7v68OPwgvxBCCM8gWzBCCCHsQgpGCCGEXUjBCCGEsAspGCGEEHYhBXMZGzZsYNSoUURHR6NpGkuWLFEdSbnp06fTo0cPAgMDiYyMZMyYMRw5ckR1LOVmzZpFYmLihQvmkpOTWb58uepYTmfGjBlomsaUKVNUR1Hq+eefR9O0i27t2rVTHavBScFcRkVFBZ07d+add95RHcVprF+/nkmTJrFlyxZWr15NbW0tQ4cOpaKiQnU0pWJiYpgxYwY7d+5kx44dDBo0iNGjR3PgwAHV0ZzG9u3bmT17NomJiaqjOIWEhARycnIu3H744QfVkRqcw4eKcSXDhw9n+PDhqmM4lRUrVlx0f968eURGRrJz504GDBigKJV6o0aNuuj+yy+/zKxZs9iyZQsJCQmKUjmP8vJy7r77bubMmcNLL72kOo5T8PLyokmTJqpj2JVswYhrUlJSAkBoaKjiJM7DbDazYMECKioqSE5OVh3HKUyaNImRI0dy/fXXq47iNI4dO0Z0dDQtW7bk7rvvJj09XXWkBidbMOKqWSwWpkyZQt++fenYsaPqOMrt27eP5ORkqqurCQgIYPHixXTo0EF1LOUWLFjArl272L59u+ooTqNXr17MmzePtm3bkpOTwwsvvED//v3Zv38/gYGBquM1GCkYcdUmTZrE/v373XLf8dVo27YtqamplJSUsHDhQsaPH8/69es9umQyMjKYPHkyq1evxtfXV3Ucp/HzXe+JiYn06tWL5s2b88UXX/DAAw8oTNawpGDEVXnkkUdYtmwZGzZscMnpF+zBx8eHVq1aAdCtWze2b9/OzJkzmT17tuJk6uzcuZO8vDy6du16YZnZbGbDhg28/fbbmEwm9Hq9woTOITg4mDZt2pCWlqY6SoOSghH1YrVa+dOf/sTixYtZt24dLVq0UB3JaVksFkwmk+oYSg0ePJh9+/ZdtOz3v/897dq148knn5RyOae8vJzjx48zbtw41VEalBTMZZSXl1/0F8XJkydJTU0lNDSUZs2aKUymzqRJk5g/fz5Lly4lMDCQ3NxcwDYBkZ+fn+J06kybNo3hw4fTrFkzysrKmD9/PuvWrWPlypWqoykVGBj4i+NzjRo1IiwszKOP202dOpVRo0bRvHlzsrOzee6559Dr9dx5552qozUoKZjL2LFjBwMHDrxw/7HHHgNg/PjxzJs3T1EqtWbNmgVASkrKRcvnzp3Lfffd5/hATiIvL497772XnJwcjEYjiYmJrFy5kiFDhqiOJpxQZmYmd955JwUFBURERNCvXz+2bNlCRESE6mgNSobrF0IIYRdyHYwQQgi7kIIRQghhF1IwQggh7EIKRgghhF1IwQghhLALKRghhBB2IQUjhBDCLqRghBBC2IUUjBBCCLuQghFCCGEXUjBCCCHsQgpGCCGEXfw/ekpTAM5jt/cAAAAASUVORK5CYII=\n"
          },
          "metadata": {}
        }
      ]
    }
  ]
}