process_perf_profile_neurips_rebuttal.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import optperfprofpy\n",
    "\n",
    "%matplotlib inline\n",
    "%config InlineBackend.figure_format='retina'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "problems = pd.Series([1, 1, 2, 2, 3, 3], dtype=int, name='problem')\n",
    "methods = pd.Series(['A', 'B', 'A', 'B', 'A', 'B'], dtype=str, name='method')\n",
    "objective_vals = pd.Series([2, 20, 25, 5, 30, 4], dtype=float, name='obj')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "example_df = pd.DataFrame([problems, methods, objective_vals]).T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of problems per solver:  3\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 261,
       "width": 392
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    example_df, ['problem'], ['obj'], ['method'])\n",
    "optperfprofpy.draw_simple_pp(taus, solver_vals, solvers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def my_draw_simple_pp(taus, solver_vals, solvers, ax,\n",
    "        fname=\"performance_profiles.pdf\"):\n",
    "    \"\"\"Simple step plotter for performance profiles.\n",
    "    Parameters\n",
    "    ----------\n",
    "    taus : numpy.ndarray\n",
    "        x values of plot.\n",
    "    solver_vals : numpy.ndarray\n",
    "        y values of plot.\n",
    "    solvers : list\n",
    "        Labels of curves.\n",
    "    \"\"\"\n",
    "\n",
    "    # Add lines individually to support labels\n",
    "    for n, solver in enumerate(solvers):\n",
    "        ax.step(taus, solver_vals[n, :], alpha=0.5, label=solver, where=\"post\")\n",
    "\n",
    "    plt.legend(loc=4, frameon=False)\n",
    "    plt.xlim(1, taus.max())\n",
    "    #ax.set_xscale(\"log\")\n",
    "    ax.set_xlabel(r\"$\\tau$\")\n",
    "    ax.set_ylabel(r\"$\\rho_{s}(\\tau)$\")\n",
    "\n",
    "    plt.savefig(fname, bbox_inches='tight')\n",
    "    plt.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of problems per solver:  12\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 360x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 207,
       "width": 331
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "problems = pd.Series(\n",
    "        np.arange(12).tolist() + \\\n",
    "        np.arange(12).tolist() + \\\n",
    "        np.arange(12).tolist(),\n",
    "        # np.arange(12).tolist(), \n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        [r'GOGGLES'] * 12 + \\\n",
    "        [r'Snuba, $D=1$'] * 12 + \\\n",
    "        [r'Interactive WS'] * 12, \n",
    "        # [r'Supervised'] * 12, \n",
    "        dtype=str, name='method')\n",
    "goggles_perfs = [\n",
    "        0.3633, 0.4613, 0.2281,\n",
    "        0.4441, 0.4524, 0.2951,\n",
    "        0.2846, 0.3365, 0.2190,\n",
    "        0.2880, 0.3523, 0.3009]\n",
    "snuba_perfs = [\n",
    "        0.2726, 0.2793, 0.1826, \n",
    "        0.3715, 0.1414, 0.1309,\n",
    "        0.2834, 0.3311, 0.2409,\n",
    "        0.3244, 0.3135, 0.2499,]\n",
    "iws_perfs = [\n",
    "        0.2363, 0.2759, 0.1100,\n",
    "        0.3848, 0.4546, 0.1878,\n",
    "        0.3612, 0.3934, 0.1808,\n",
    "        0.2738, 0.3509, 0.2047,]\n",
    "# sup_perfs = [\n",
    "#         0.9, 0.8, 0.8,\n",
    "#         0.8, 0.6, 0.7,\n",
    "#         0.6, 0.7, 0.8,\n",
    "#         0.6, 0.6, 0.7,]\n",
    "objective_vals = pd.Series(\n",
    "        1.0 - np.array(goggles_perfs + snuba_perfs + iws_perfs), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "fig = plt.figure(figsize=(5, 3))\n",
    "ax = fig.add_subplot(111)\n",
    "my_draw_simple_pp(taus, solver_vals, solvers, ax,\n",
    "        fname=\"performance_profiles_search.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of problems per solver:  9\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 360x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 207,
       "width": 331
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "problems = pd.Series(\n",
    "        np.arange(3).tolist() + \\\n",
    "        np.arange(3).tolist() + \\\n",
    "        np.arange(3).tolist() + \\\n",
    "        np.arange(3).tolist() + \\\n",
    "        np.arange(3, 6).tolist() + \\\n",
    "        np.arange(3, 6).tolist() + \\\n",
    "        np.arange(3, 6).tolist() + \\\n",
    "        np.arange(3, 6).tolist() + \\\n",
    "        np.arange(6, 9).tolist() + \\\n",
    "        np.arange(6, 9).tolist() + \\\n",
    "        np.arange(6, 9).tolist() + \\\n",
    "        np.arange(6, 9).tolist(),\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        ([r'Raw features'] * 3 + \\\n",
    "        [r'PCA'] * 3 + \\\n",
    "        [r'ResNet-18'] * 3 + \\\n",
    "        [r'VAE'] * 3) * 3, \n",
    "        dtype=str, name='method')\n",
    "goggles_perfs = [\n",
    "        0.3633, 0.4613, 0.2281,\n",
    "        0.4441, 0.4524, 0.2951,\n",
    "        0.2846, 0.3365, 0.2190,\n",
    "        0.2880, 0.3523, 0.3009]\n",
    "snuba_perfs = [\n",
    "        0.2726, 0.2793, 0.1826, \n",
    "        0.3715, 0.1414, 0.1309,\n",
    "        0.2834, 0.3311, 0.2409,\n",
    "        0.3244, 0.3135, 0.2499,]\n",
    "iws_perfs = [\n",
    "        0.2363, 0.2759, 0.1100,\n",
    "        0.3848, 0.4546, 0.1878,\n",
    "        0.3612, 0.3934, 0.1808,\n",
    "        0.2738, 0.3509, 0.2047,]\n",
    "# sup_perfs = [\n",
    "#         0.9, 0.8, 0.8,\n",
    "#         0.8, 0.6, 0.7,\n",
    "#         0.6, 0.7, 0.8,\n",
    "#         0.6, 0.6, 0.7,]\n",
    "objective_vals = pd.Series(\n",
    "        1.0 - np.array(goggles_perfs + snuba_perfs + iws_perfs), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "fig = plt.figure(figsize=(5, 3))\n",
    "ax = fig.add_subplot(111)\n",
    "my_draw_simple_pp(taus, solver_vals, solvers, ax,\n",
    "        fname=\"performance_profiles_emb.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>problem</th>\n",
       "      <th>method</th>\n",
       "      <th>obj</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>Raw features</td>\n",
       "      <td>0.6367</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>Raw features</td>\n",
       "      <td>0.5387</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>Raw features</td>\n",
       "      <td>0.7719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>PCA</td>\n",
       "      <td>0.5559</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>PCA</td>\n",
       "      <td>0.5476</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2</td>\n",
       "      <td>PCA</td>\n",
       "      <td>0.7049</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0</td>\n",
       "      <td>ResNet-18</td>\n",
       "      <td>0.7154</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1</td>\n",
       "      <td>ResNet-18</td>\n",
       "      <td>0.6635</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>2</td>\n",
       "      <td>ResNet-18</td>\n",
       "      <td>0.781</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0</td>\n",
       "      <td>VAE</td>\n",
       "      <td>0.712</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1</td>\n",
       "      <td>VAE</td>\n",
       "      <td>0.6477</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>2</td>\n",
       "      <td>VAE</td>\n",
       "      <td>0.6991</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>3</td>\n",
       "      <td>Raw features</td>\n",
       "      <td>0.7274</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>4</td>\n",
       "      <td>Raw features</td>\n",
       "      <td>0.7207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>5</td>\n",
       "      <td>Raw features</td>\n",
       "      <td>0.8174</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>3</td>\n",
       "      <td>PCA</td>\n",
       "      <td>0.6285</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>4</td>\n",
       "      <td>PCA</td>\n",
       "      <td>0.8586</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>5</td>\n",
       "      <td>PCA</td>\n",
       "      <td>0.8691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>3</td>\n",
       "      <td>ResNet-18</td>\n",
       "      <td>0.7166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>4</td>\n",
       "      <td>ResNet-18</td>\n",
       "      <td>0.6689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>5</td>\n",
       "      <td>ResNet-18</td>\n",
       "      <td>0.7591</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>3</td>\n",
       "      <td>VAE</td>\n",
       "      <td>0.6756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>4</td>\n",
       "      <td>VAE</td>\n",
       "      <td>0.6865</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>5</td>\n",
       "      <td>VAE</td>\n",
       "      <td>0.7501</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>6</td>\n",
       "      <td>Raw features</td>\n",
       "      <td>0.7637</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>7</td>\n",
       "      <td>Raw features</td>\n",
       "      <td>0.7241</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>8</td>\n",
       "      <td>Raw features</td>\n",
       "      <td>0.89</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>6</td>\n",
       "      <td>PCA</td>\n",
       "      <td>0.6152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>7</td>\n",
       "      <td>PCA</td>\n",
       "      <td>0.5454</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>8</td>\n",
       "      <td>PCA</td>\n",
       "      <td>0.8122</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>6</td>\n",
       "      <td>ResNet-18</td>\n",
       "      <td>0.6388</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>7</td>\n",
       "      <td>ResNet-18</td>\n",
       "      <td>0.6066</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>8</td>\n",
       "      <td>ResNet-18</td>\n",
       "      <td>0.8192</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>6</td>\n",
       "      <td>VAE</td>\n",
       "      <td>0.7262</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>7</td>\n",
       "      <td>VAE</td>\n",
       "      <td>0.6491</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>8</td>\n",
       "      <td>VAE</td>\n",
       "      <td>0.7953</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   problem        method     obj\n",
       "0        0  Raw features  0.6367\n",
       "1        1  Raw features  0.5387\n",
       "2        2  Raw features  0.7719\n",
       "3        0           PCA  0.5559\n",
       "4        1           PCA  0.5476\n",
       "5        2           PCA  0.7049\n",
       "6        0     ResNet-18  0.7154\n",
       "7        1     ResNet-18  0.6635\n",
       "8        2     ResNet-18   0.781\n",
       "9        0           VAE   0.712\n",
       "10       1           VAE  0.6477\n",
       "11       2           VAE  0.6991\n",
       "12       3  Raw features  0.7274\n",
       "13       4  Raw features  0.7207\n",
       "14       5  Raw features  0.8174\n",
       "15       3           PCA  0.6285\n",
       "16       4           PCA  0.8586\n",
       "17       5           PCA  0.8691\n",
       "18       3     ResNet-18  0.7166\n",
       "19       4     ResNet-18  0.6689\n",
       "20       5     ResNet-18  0.7591\n",
       "21       3           VAE  0.6756\n",
       "22       4           VAE  0.6865\n",
       "23       5           VAE  0.7501\n",
       "24       6  Raw features  0.7637\n",
       "25       7  Raw features  0.7241\n",
       "26       8  Raw features    0.89\n",
       "27       6           PCA  0.6152\n",
       "28       7           PCA  0.5454\n",
       "29       8           PCA  0.8122\n",
       "30       6     ResNet-18  0.6388\n",
       "31       7     ResNet-18  0.6066\n",
       "32       8     ResNet-18  0.8192\n",
       "33       6           VAE  0.7262\n",
       "34       7           VAE  0.6491\n",
       "35       8           VAE  0.7953"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def my_draw_simple_pp(taus, solver_vals, solvers, ax,\n",
    "        fname=\"performance_profiles.pdf\", xlim=None, log=False, frameon=False, \n",
    "        colors=None, legend=True):\n",
    "    \"\"\"Simple step plotter for performance profiles.\n",
    "    Parameters\n",
    "    ----------\n",
    "    taus : numpy.ndarray\n",
    "        x values of plot.\n",
    "    solver_vals : numpy.ndarray\n",
    "        y values of plot.\n",
    "    solvers : list\n",
    "        Labels of curves.\n",
    "    \"\"\"\n",
    "\n",
    "    if colors is None:\n",
    "        colors = [f\"C{i}\" for i in range(10)]\n",
    "\n",
    "    # Add lines individually to support labels\n",
    "    for n, solver in enumerate(solvers):\n",
    "        ax.step(taus, \n",
    "            solver_vals[n, :], alpha=0.5, linewidth=3, \n",
    "            label=solver, color=colors[n], where='post')\n",
    "\n",
    "    if legend:\n",
    "        ax.legend(frameon=frameon, loc='center left', bbox_to_anchor=(1, 0.5))\n",
    "    if xlim == None:\n",
    "        ax.set_xlim(1, taus.max())\n",
    "    else:\n",
    "        ax.set_xlim(1, xlim)\n",
    "    if log:\n",
    "        ax.set_xscale(\"log\")\n",
    "    ax.set_xlabel(r\"$\\tau$\")\n",
    "    ax.set_ylabel(r\"$\\rho_{s}(\\tau)$\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "#axs[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of problems per solver:  16\n",
      "Number of problems per solver:  24\n",
      "Number of problems per solver:  12\n",
      "Number of problems per solver:  18\n",
      "Number of problems per solver:  12\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/hdd2/nick/github.com/Kaylee0501/FWRENCH/optperfprofpy.py:185: UserWarning: Solvers do not solve 100% of problems. Total amount of problems solved: 183.33333333333331\n",
      "  \"Total amount of problems solved: {}\".format(100 * solver_taus[:, 0].sum())\n",
      "/hdd2/nick/github.com/Kaylee0501/FWRENCH/optperfprofpy.py:185: UserWarning: Solvers do not solve 100% of problems. Total amount of problems solved: 122.22222222222223\n",
      "  \"Total amount of problems solved: {}\".format(100 * solver_taus[:, 0].sum())\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of problems per solver:  18\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/hdd2/nick/github.com/Kaylee0501/FWRENCH/optperfprofpy.py:185: UserWarning: Solvers do not solve 100% of problems. Total amount of problems solved: 116.66666666666667\n",
      "  \"Total amount of problems solved: {}\".format(100 * solver_taus[:, 0].sum())\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x306 with 6 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 300,
       "width": 884
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(2, 3, figsize=((3.5*3) + 2, 4.25))\n",
    "\n",
    "\n",
    "#specify order of items in legend\n",
    "# NOTE this is brittle and will need to change if more methods added\n",
    "order = [0, 3, 1, 2, 4, 5]\n",
    "\n",
    "\n",
    "\n",
    "ntotal_task = 10\n",
    "nrows = 4\n",
    "rowlist = [0, 1, 2, 3]\n",
    "tasklist = [0, 1, 2, 3] #[0, 1, 2, 3, 4, 5, 6]\n",
    "assert len(tasklist) <= 7\n",
    "ntask = len(tasklist)\n",
    "inds = np.array([[t + (ntotal_task * row) for t in tasklist] for row in rowlist]).reshape(-1).tolist()\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "goggles_perfs = [\n",
    "        0.3778, 0.1697, 0.0897, 0.3778, 0.2710, 0.5284, 0.5283, 0.5880, 0.4732, 0.5516, \n",
    "        0.4563, 0.1989, 0.1074, 0.4561, 0.2688, 0.4863, 0.6728, 0.5880, 0.5466, 0.5628, \n",
    "        0.3258, 0.1768, 0.1192, 0.1894, 0.3127, 0.9353, 0.5312, 0.5200, 0.4553, 0.5152, \n",
    "        0.3309, 0.5436, 0.1884, 0.1405, 0.0000, 0.7505, 0.0000, 0.5280, 0.4779, 0.5000, \n",
    "]\n",
    "iws_perfs = [\n",
    "        0.4128, 0.1482, 0.1154, 0.3862, 0.2615, 0.5403, 0.4997, 0.4896, 0.6034, 0.5478,\n",
    "        0.3503, 0.1047, 0.1283, 0.4970, 0.2687, 0.5155, 0.4777, 0.5217, 0.5260, 0.5045,\n",
    "        0.0822, 0.2120, 0.1352, 0.5101, 0.4235, 0.7995, 0.7343, 0.7840, 0.5620, 0.5286,\n",
    "        0.6885, 0.7892, 0.1108, 0.3495, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
    "]\n",
    "snuba_perfs = [\n",
    "        0.2694, 0.1624, 0.0805, 0.2308, 0.2648, 0.5235, 0.7082, 0.7520, 0.6034, 0.5936,\n",
    "        0.2313, 0.0924, 0.1526, 0.2640, 0.2598, 0.8811, 0.4905, 0.6640, 0.7187, 0.5560,\n",
    "        0.4219, 0.1520, 0.1554, 0.1804, 0.2506, 0.9774, 0.6994, 0.7400, 0.5674, 0.5804,\n",
    "        0.6617, 0.7550, 0.3103, 0.2419, 0.0000, 0.8414, 0.0000, 0.6280, 0.5608, 0.5699,\n",
    "]\n",
    "snub_multi_perfs = [\n",
    "        0.2374, 0.1386, 0.0991, 0.2343, 0.2596, 0.5421, 0.7264, 0.7080, 0.6484, 0.6072,\n",
    "        0.5625, 0.1390, 0.2354, 0.5625, 0.2644, 0.8968, 0.5143, 0.6280, 0.7218, 0.5020,\n",
    "        0.3866, 0.2036, 0.1588, 0.2283, 0.3702, 0.9742, 0.7559, 0.7320, 0.5526, 0.5864,\n",
    "        0.3541, 0.6788, 0.2479, 0.2785, 0.0000, 0.7916, 0.0000, 0.7759, 0.5435, 0.5681,\n",
    "]\n",
    "sup_perfs = [\n",
    "        0.7295, 0.2315, 0.1496, 0.7295, 0.2583, 0.6316, 0.6517, 0.8920, 0.7871, 0.6472,\n",
    "        0.7306, 0.2262, 0.1483, 0.7305, 0.2548, 0.7863, 0.7488, 0.8880, 0.7868, 0.6464,\n",
    "        0.7966, 0.3335, 0.2544, 0.3485, 0.4700, 0.9900, 0.5329, 0.7000, 0.5682, 0.5856,\n",
    "        0.3579, 0.6857, 0.1011, 0.1009, 0.0000, 0.7658, 0.0000, 0.7000, 0.6347, 0.5636,\n",
    "]\n",
    "lp_perfs = [\n",
    "        0.7066, 0.1927, 0.1374, 0.7066, 0.2644, 0.5010, 0.5003, 0.8440, 0.5447, 0.5812,\n",
    "        0.7165, 0.1914, 0.1405, 0.7156, 0.2644, 0.5011, 0.5003, 0.8440, 0.5450, 0.5812,\n",
    "        0.7124, 0.1277, 0.1283, 0.2106, 0.2644, 0.9700, 0.5003, 0.7000, 0.5224, 0.5500,\n",
    "        0.1178, 0.5780, 0.1009, 0.1009, 0.0000, 0.5058, 0.0000, 0.6480, 0.5737, 0.5400,\n",
    "]\n",
    "\n",
    "goggles_perfs = np.array(goggles_perfs)[inds].tolist()\n",
    "iws_perfs = np.array(iws_perfs)[inds].tolist()\n",
    "snuba_perfs = np.array(snuba_perfs)[inds].tolist()\n",
    "snub_multi_perfs = np.array(snub_multi_perfs)[inds].tolist()\n",
    "sup_perfs = np.array(sup_perfs)[inds].tolist()\n",
    "lp_perfs = np.array(lp_perfs)[inds].tolist()\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(len(goggles_perfs)).tolist() + \\\n",
    "        np.arange(len(iws_perfs)).tolist() + \\\n",
    "        np.arange(len(snuba_perfs)).tolist() + \\\n",
    "        np.arange(len(snub_multi_perfs)).tolist() + \\\n",
    "        np.arange(len(sup_perfs)).tolist() + \\\n",
    "        np.arange(len(lp_perfs)).tolist(),\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        [r'GOGGLES'] * len(goggles_perfs) + \\\n",
    "        [r'Interactive WS'] * len(iws_perfs) + \\\n",
    "        [r'Snuba'] * len(snuba_perfs) + \\\n",
    "        [r'Snuba (multipolar)'] * len(snub_multi_perfs) + \\\n",
    "        [r'Few-Shot'] * len(sup_perfs) + \\\n",
    "        [r'Semi-Supervised'] * len(lp_perfs),\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1.0 - np.array(\n",
    "                goggles_perfs + \\\n",
    "                iws_perfs + \\\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs + \\\n",
    "                sup_perfs + \\\n",
    "                lp_perfs\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[0, 0], \n",
    "        xlim=None, log=True, legend=False, frameon=False, colors=[\n",
    "                f\"C{i}\" for i in np.argsort(order)])\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# TODO index into each of the above by task\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(0*ntask, 0*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(1*ntask, 1*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(2*ntask, 2*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(3*ntask, 3*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(4*ntask, 4*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(5*ntask, 5*ntask+ntask).tolist() * 4,\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        ([r'Raw features'] * ntask + \\\n",
    "        [r'PCA'] * ntask + \\\n",
    "        [r'ResNet-18'] * ntask + \\\n",
    "        [r'CLIP'] * ntask) * 6,\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1 - np.array(\n",
    "                goggles_perfs + \\\n",
    "                iws_perfs + \\\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs + \\\n",
    "                sup_perfs + \\\n",
    "                lp_perfs\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[1, 0], \n",
    "        xlim=None, log=True, legend=False,frameon=False, colors=[\n",
    "                f\"C{i}\" for i in [5, 6, 8, 9]])\n",
    "\n",
    "\n",
    "axs[0, 0].set_title('Image tasks')\n",
    "\n",
    "#plt.tight_layout()\n",
    "#plt.savefig(\"neurips/perf_all_image_tasks.pdf\")\n",
    "#plt.show()\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "ntotal_task = 10\n",
    "nrows = 4\n",
    "rowlist = [0, 1, 2, 3]\n",
    "tasklist = [4, 5, 6] #[0, 1, 2, 3, 4, 5, 6]\n",
    "assert len(tasklist) <= 7\n",
    "ntask = len(tasklist)\n",
    "inds = np.array([[t + (ntotal_task * row) for t in tasklist] for row in rowlist]).reshape(-1).tolist()\n",
    "\n",
    "\n",
    "goggles_perfs = [\n",
    "        0.3778, 0.1697, 0.0897, 0.3778, 0.2710, 0.5284, 0.5283, 0.5880, 0.4732, 0.5516, \n",
    "        0.4563, 0.1989, 0.1074, 0.4561, 0.2688, 0.4863, 0.6728, 0.5880, 0.5466, 0.5628, \n",
    "        0.3258, 0.1768, 0.1192, 0.1894, 0.3127, 0.9353, 0.5312, 0.5200, 0.4553, 0.5152, \n",
    "        0.3309, 0.5436, 0.1884, 0.1405, 0.0000, 0.7505, 0.0000, 0.5280, 0.4779, 0.5000, \n",
    "]\n",
    "iws_perfs = [\n",
    "        0.4128, 0.1482, 0.1154, 0.3862, 0.2615, 0.5403, 0.4997, 0.4896, 0.6034, 0.5478,\n",
    "        0.3503, 0.1047, 0.1283, 0.4970, 0.2687, 0.5155, 0.4777, 0.5217, 0.5260, 0.5045,\n",
    "        0.0822, 0.2120, 0.1352, 0.5101, 0.4235, 0.7995, 0.7343, 0.7840, 0.5620, 0.5286,\n",
    "        0.6885, 0.7892, 0.1108, 0.3495, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
    "]\n",
    "snuba_perfs = [\n",
    "        0.2694, 0.1624, 0.0805, 0.2308, 0.2648, 0.5235, 0.7082, 0.7520, 0.6034, 0.5936,\n",
    "        0.2313, 0.0924, 0.1526, 0.2640, 0.2598, 0.8811, 0.4905, 0.6640, 0.7187, 0.5560,\n",
    "        0.4219, 0.1520, 0.1554, 0.1804, 0.2506, 0.9774, 0.6994, 0.7400, 0.5674, 0.5804,\n",
    "        0.6617, 0.7550, 0.3103, 0.2419, 0.0000, 0.8414, 0.0000, 0.6280, 0.5608, 0.5699,\n",
    "]\n",
    "snub_multi_perfs = [\n",
    "        0.2374, 0.1386, 0.0991, 0.2343, 0.2596, 0.5421, 0.7264, 0.7080, 0.6484, 0.6072,\n",
    "        0.5625, 0.1390, 0.2354, 0.5625, 0.2644, 0.8968, 0.5143, 0.6280, 0.7218, 0.5020,\n",
    "        0.3866, 0.2036, 0.1588, 0.2283, 0.3702, 0.9742, 0.7559, 0.7320, 0.5526, 0.5864,\n",
    "        0.3541, 0.6788, 0.2479, 0.2785, 0.0000, 0.7916, 0.0000, 0.7759, 0.5435, 0.5681,\n",
    "]\n",
    "sup_perfs = [\n",
    "        0.7295, 0.2315, 0.1496, 0.7295, 0.2583, 0.6316, 0.6517, 0.8920, 0.7871, 0.6472,\n",
    "        0.7306, 0.2262, 0.1483, 0.7305, 0.2548, 0.7863, 0.7488, 0.8880, 0.7868, 0.6464,\n",
    "        0.7966, 0.3335, 0.2544, 0.3485, 0.4700, 0.9900, 0.5329, 0.7000, 0.5682, 0.5856,\n",
    "        0.3579, 0.6857, 0.1011, 0.1009, 0.0000, 0.7658, 0.0000, 0.7000, 0.6347, 0.5636,\n",
    "]\n",
    "lp_perfs = [\n",
    "        0.7066, 0.1927, 0.1374, 0.7066, 0.2644, 0.5010, 0.5003, 0.8440, 0.5447, 0.5812,\n",
    "        0.7165, 0.1914, 0.1405, 0.7156, 0.2644, 0.5011, 0.5003, 0.8440, 0.5450, 0.5812,\n",
    "        0.7124, 0.1277, 0.1283, 0.2106, 0.2644, 0.9700, 0.5003, 0.7000, 0.5224, 0.5500,\n",
    "        0.1178, 0.5780, 0.1009, 0.1009, 0.0000, 0.5058, 0.0000, 0.6480, 0.5737, 0.5400,\n",
    "]\n",
    "\n",
    "\n",
    "goggles_perfs = np.array(goggles_perfs)[inds].tolist()\n",
    "iws_perfs = np.array(iws_perfs)[inds].tolist()\n",
    "snuba_perfs = np.array(snuba_perfs)[inds].tolist()\n",
    "snub_multi_perfs = np.array(snub_multi_perfs)[inds].tolist()\n",
    "sup_perfs = np.array(sup_perfs)[inds].tolist()\n",
    "lp_perfs = np.array(lp_perfs)[inds].tolist()\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(len(goggles_perfs)).tolist() + \\\n",
    "        np.arange(len(iws_perfs)).tolist() + \\\n",
    "        np.arange(len(snuba_perfs)).tolist() + \\\n",
    "        np.arange(len(snub_multi_perfs)).tolist() + \\\n",
    "        np.arange(len(sup_perfs)).tolist() + \\\n",
    "        np.arange(len(lp_perfs)).tolist(),\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        [r'GOGGLES'] * len(goggles_perfs) + \\\n",
    "        [r'Interactive WS'] * len(iws_perfs) + \\\n",
    "        [r'Snuba'] * len(snuba_perfs) + \\\n",
    "        [r'Snuba (multipolar)'] * len(snub_multi_perfs) + \\\n",
    "        [r'Few-Shot'] * len(sup_perfs) + \\\n",
    "        [r'Semi-Supervised'] * len(lp_perfs),\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1.0 - np.array(\n",
    "                goggles_perfs + \\\n",
    "                iws_perfs + \\\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs + \\\n",
    "                sup_perfs + \\\n",
    "                lp_perfs\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[0, 1], \n",
    "        xlim=None, log=True, legend=False, frameon=False, colors=[\n",
    "                f\"C{i}\" for i in np.argsort(order)])\n",
    "\n",
    "\n",
    "\n",
    "# TODO index into each of the above by task\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(0*ntask, 0*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(1*ntask, 1*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(2*ntask, 2*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(3*ntask, 3*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(4*ntask, 4*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(5*ntask, 5*ntask+ntask).tolist() * 4,\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        ([r'Raw features'] * ntask + \\\n",
    "        [r'PCA'] * ntask + \\\n",
    "        [r'ResNet-18'] * ntask + \\\n",
    "        [r'CLIP'] * ntask) * 6,\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1 - np.array(\n",
    "                goggles_perfs + \\\n",
    "                iws_perfs + \\\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs + \\\n",
    "                sup_perfs + \\\n",
    "                lp_perfs\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[1, 1], \n",
    "        xlim=None, log=True, legend=False, frameon=False, colors=[\n",
    "                f\"C{i}\" for i in [5, 6, 8, 9]])\n",
    "\n",
    "\n",
    "axs[0, 1].set_title('Diverse tasks')\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "ntotal_task = 10\n",
    "nrows = 4\n",
    "rowlist = [0, 1, 2, 3]\n",
    "tasklist = [7, 8, 9] #[0, 1, 2, 3, 4, 5, 6]\n",
    "assert len(tasklist) <= 7\n",
    "ntask = len(tasklist)\n",
    "inds = np.array([[t + (ntotal_task * row) for t in tasklist] for row in rowlist]).reshape(-1).tolist()\n",
    "\n",
    "\n",
    "goggles_perfs = [\n",
    "        0.3778, 0.1697, 0.0897, 0.3778, 0.2710, 0.5284, 0.5283, 0.5880, 0.4732, 0.5516, \n",
    "        0.4563, 0.1989, 0.1074, 0.4561, 0.2688, 0.4863, 0.6728, 0.5880, 0.5466, 0.5628, \n",
    "        0.3258, 0.1768, 0.1192, 0.1894, 0.3127, 0.9353, 0.5312, 0.5200, 0.4553, 0.5152, \n",
    "        0.3309, 0.5436, 0.1884, 0.1405, 0.0000, 0.7505, 0.0000, 0.5280, 0.4779, 0.5000, \n",
    "]\n",
    "iws_perfs = [\n",
    "        0.4128, 0.1482, 0.1154, 0.3862, 0.2615, 0.5403, 0.4997, 0.4896, 0.6034, 0.5478,\n",
    "        0.3503, 0.1047, 0.1283, 0.4970, 0.2687, 0.5155, 0.4777, 0.5217, 0.5260, 0.5045,\n",
    "        0.0822, 0.2120, 0.1352, 0.5101, 0.4235, 0.7995, 0.7343, 0.7840, 0.5620, 0.5286,\n",
    "        0.6885, 0.7892, 0.1108, 0.3495, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
    "]\n",
    "snuba_perfs = [\n",
    "        0.2694, 0.1624, 0.0805, 0.2308, 0.2648, 0.5235, 0.7082, 0.7520, 0.6034, 0.5936,\n",
    "        0.2313, 0.0924, 0.1526, 0.2640, 0.2598, 0.8811, 0.4905, 0.6640, 0.7187, 0.5560,\n",
    "        0.4219, 0.1520, 0.1554, 0.1804, 0.2506, 0.9774, 0.6994, 0.7400, 0.5674, 0.5804,\n",
    "        0.6617, 0.7550, 0.3103, 0.2419, 0.0000, 0.8414, 0.0000, 0.6280, 0.5608, 0.5699,\n",
    "]\n",
    "snub_multi_perfs = [\n",
    "        0.2374, 0.1386, 0.0991, 0.2343, 0.2596, 0.5421, 0.7264, 0.7080, 0.6484, 0.6072,\n",
    "        0.5625, 0.1390, 0.2354, 0.5625, 0.2644, 0.8968, 0.5143, 0.6280, 0.7218, 0.5020,\n",
    "        0.3866, 0.2036, 0.1588, 0.2283, 0.3702, 0.9742, 0.7559, 0.7320, 0.5526, 0.5864,\n",
    "        0.3541, 0.6788, 0.2479, 0.2785, 0.0000, 0.7916, 0.0000, 0.7759, 0.5435, 0.5681,\n",
    "]\n",
    "sup_perfs = [\n",
    "        0.7295, 0.2315, 0.1496, 0.7295, 0.2583, 0.6316, 0.6517, 0.8920, 0.7871, 0.6472,\n",
    "        0.7306, 0.2262, 0.1483, 0.7305, 0.2548, 0.7863, 0.7488, 0.8880, 0.7868, 0.6464,\n",
    "        0.7966, 0.3335, 0.2544, 0.3485, 0.4700, 0.9900, 0.5329, 0.7000, 0.5682, 0.5856,\n",
    "        0.3579, 0.6857, 0.1011, 0.1009, 0.0000, 0.7658, 0.0000, 0.7000, 0.6347, 0.5636,\n",
    "]\n",
    "lp_perfs = [\n",
    "        0.7066, 0.1927, 0.1374, 0.7066, 0.2644, 0.5010, 0.5003, 0.8440, 0.5447, 0.5812,\n",
    "        0.7165, 0.1914, 0.1405, 0.7156, 0.2644, 0.5011, 0.5003, 0.8440, 0.5450, 0.5812,\n",
    "        0.7124, 0.1277, 0.1283, 0.2106, 0.2644, 0.9700, 0.5003, 0.7000, 0.5224, 0.5500,\n",
    "        0.1178, 0.5780, 0.1009, 0.1009, 0.0000, 0.5058, 0.0000, 0.6480, 0.5737, 0.5400,\n",
    "]\n",
    "\n",
    "goggles_perfs = np.array(goggles_perfs)[inds].tolist()\n",
    "iws_perfs = np.array(iws_perfs)[inds].tolist()\n",
    "snuba_perfs = np.array(snuba_perfs)[inds].tolist()\n",
    "snub_multi_perfs = np.array(snub_multi_perfs)[inds].tolist()\n",
    "sup_perfs = np.array(sup_perfs)[inds].tolist()\n",
    "lp_perfs = np.array(lp_perfs)[inds].tolist()\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(len(goggles_perfs)).tolist() + \\\n",
    "        np.arange(len(iws_perfs)).tolist() + \\\n",
    "        np.arange(len(snuba_perfs)).tolist() + \\\n",
    "        np.arange(len(snub_multi_perfs)).tolist() + \\\n",
    "        np.arange(len(sup_perfs)).tolist() + \\\n",
    "        np.arange(len(lp_perfs)).tolist(),\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        [r'GOGGLES'] * len(goggles_perfs) + \\\n",
    "        [r'Interactive WS'] * len(iws_perfs) + \\\n",
    "        [r'Snuba'] * len(snuba_perfs) + \\\n",
    "        [r'Snuba (multipolar)'] * len(snub_multi_perfs) + \\\n",
    "        [r'Few-Shot'] * len(sup_perfs) + \\\n",
    "        [r'Semi-Supervised'] * len(lp_perfs),\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1.0 - np.array(\n",
    "                goggles_perfs + \\\n",
    "                iws_perfs + \\\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs + \\\n",
    "                sup_perfs + \\\n",
    "                lp_perfs\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[0, 2], \n",
    "        xlim=None, log=True, frameon=False, colors=[\n",
    "                f\"C{i}\" for i in np.argsort(order)])\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# TODO index into each of the above by task\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(0*ntask, 0*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(1*ntask, 1*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(2*ntask, 2*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(3*ntask, 3*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(4*ntask, 4*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(5*ntask, 5*ntask+ntask).tolist() * 4,\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        ([r'Raw features'] * ntask + \\\n",
    "        [r'PCA'] * ntask + \\\n",
    "        [r'ResNet-18'] * ntask + \\\n",
    "        [r'CLIP'] * ntask) * 6,\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1 - np.array(\n",
    "                goggles_perfs + \\\n",
    "                iws_perfs + \\\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs + \\\n",
    "                sup_perfs + \\\n",
    "                lp_perfs\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[1, 2], \n",
    "        xlim=None, log=True, frameon=False, colors=[\n",
    "                f\"C{i}\" for i in [5, 6, 8, 9]])\n",
    "\n",
    "\n",
    "axs[0, 2].set_title('Text tasks')\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "#get handles and labels\n",
    "handles, labels = axs[0, 2].get_legend_handles_labels()\n",
    "\n",
    "#add legend to plot\n",
    "axs[0, 2].legend(\n",
    "        [handles[idx] for idx in order],[labels[idx] for idx in order],\n",
    "        frameon=False, loc='center left', bbox_to_anchor=(1, 0.5))\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "plt.suptitle('Accuracy performance profiles across task types')\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"neurips/perf_all_tasks.pdf\")\n",
    "plt.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of problems per solver:  16\n",
      "Number of problems per solver:  12\n",
      "Number of problems per solver:  12\n",
      "Number of problems per solver:  9\n",
      "Number of problems per solver:  12\n",
      "Number of problems per solver:  9\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/hdd2/nick/github.com/Kaylee0501/FWRENCH/optperfprofpy.py:185: UserWarning: Solvers do not solve 100% of problems. Total amount of problems solved: 175.0\n",
      "  \"Total amount of problems solved: {}\".format(100 * solver_taus[:, 0].sum())\n",
      "/hdd2/nick/github.com/Kaylee0501/FWRENCH/optperfprofpy.py:185: UserWarning: Solvers do not solve 100% of problems. Total amount of problems solved: 188.88888888888889\n",
      "  \"Total amount of problems solved: {}\".format(100 * solver_taus[:, 0].sum())\n",
      "/hdd2/nick/github.com/Kaylee0501/FWRENCH/optperfprofpy.py:185: UserWarning: Solvers do not solve 100% of problems. Total amount of problems solved: 175.0\n",
      "  \"Total amount of problems solved: {}\".format(100 * solver_taus[:, 0].sum())\n",
      "/hdd2/nick/github.com/Kaylee0501/FWRENCH/optperfprofpy.py:185: UserWarning: Solvers do not solve 100% of problems. Total amount of problems solved: 222.2222222222222\n",
      "  \"Total amount of problems solved: {}\".format(100 * solver_taus[:, 0].sum())\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x270 with 6 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 264,
       "width": 884
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(2, 3, figsize=((3.5*3) + 2, 3.75))\n",
    "\n",
    "\n",
    "\n",
    "ntotal_task = 10\n",
    "nrows = 4\n",
    "rowlist = [0, 1, 2, 3]\n",
    "tasklist = [0, 1, 2, 3] #[0, 1, 2, 3, 4, 5, 6]\n",
    "assert len(tasklist) <= 10\n",
    "ntask = len(tasklist)\n",
    "inds = np.array([[t + (ntotal_task * row) for t in tasklist] for row in rowlist]).reshape(-1).tolist()\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "iws_perfs = [\n",
    "        0.3617, 0.4616, 0.3666, 0.3656, 0.5125, 0.8163, 0.9999, 0.9640, 0.9999, 0.7908,\n",
    "        0.5107, 0.2731, 0.2696, 0.4797, 0.5448, 0.9874, 0.8873, 0.9200, 0.9616, 0.9816,\n",
    "        0.7422, 0.3580, 0.0917, 0.1631, 0.6109, 0.9999, 0.6142, 0.9999, 0.6368, 0.8596,\n",
    "        0.2758, 0.7239, 0.0993, 0.1757, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
    "]\n",
    "snuba_perfs = [\n",
    "        0.6842, 0.6409, 0.8387, 0.8105, 0.9090, 0.5379, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.9608, 0.7183, 0.9557, 0.9716, 0.9956, 0.9999, 0.8072, 0.9999, 0.9971, 0.9999,\n",
    "        0.8761, 0.9772, 0.9230, 0.9320, 0.8998, 0.9999, 0.5912, 0.9999, 0.9884, 0.9999,\n",
    "        0.7667, 0.7262, 0.1521, 0.1323, 0.0000, 0.4116, 0.0000, 0.9999, 0.4781, 0.2716,\n",
    "]\n",
    "snub_multi_perfs = [\n",
    "        0.7422, 0.8129, 0.4470, 0.5407, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.3472, 0.5807, 0.4763, 0.3472, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.6185, 0.6337, 0.8232, 0.7193, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.5027, 0.3372, 0.2033, 0.7235, 0.0000, 0.9999, 0.0000, 0.7320, 0.5926, 0.2816,\n",
    "]\n",
    "\n",
    "iws_perfs = np.array(iws_perfs)[inds].tolist()\n",
    "snuba_perfs = np.array(snuba_perfs)[inds].tolist()\n",
    "snub_multi_perfs = np.array(snub_multi_perfs)[inds].tolist()\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(len(iws_perfs)).tolist() + \\\n",
    "        np.arange(len(snuba_perfs)).tolist() + \\\n",
    "        np.arange(len(snub_multi_perfs)).tolist(),\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        [r'Interactive WS'] * len(iws_perfs) + \\\n",
    "        [r'Snuba'] * len(snuba_perfs) + \\\n",
    "        [r'Snuba (multipolar)'] * len(snub_multi_perfs),\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1.0 - np.array(\n",
    "                iws_perfs + \\\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs,\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[0, 0], \n",
    "        xlim=None, log=True, legend=False, frameon=False)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# TODO index into each of the above by task\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(0*ntask, 0*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(1*ntask, 1*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(2*ntask, 2*ntask+ntask).tolist() * 4,\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        ([r'Raw features'] * ntask + \\\n",
    "        [r'PCA'] * ntask + \\\n",
    "        [r'ResNet-18'] * ntask + \\\n",
    "        [r'CLIP'] * ntask) * 3,\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1 - np.array(\n",
    "                iws_perfs + \\\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs,\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[1, 0], \n",
    "        xlim=None, log=True, legend=False,frameon=False, colors=[\n",
    "                f\"C{i}\" for i in [5, 6, 8, 9]])\n",
    "\n",
    "\n",
    "axs[0, 0].set_title('Image tasks')\n",
    "\n",
    "#plt.tight_layout()\n",
    "#plt.savefig(\"neurips/perf_all_image_tasks.pdf\")\n",
    "#plt.show()\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "ntotal_task = 10\n",
    "nrows = 4\n",
    "rowlist = [0, 1, 2, 3]\n",
    "tasklist = [4, 5, 6] #[0, 1, 2, 3, 4, 5, 6]\n",
    "assert len(tasklist) <= 10\n",
    "ntask = len(tasklist)\n",
    "inds = np.array([[t + (ntotal_task * row) for t in tasklist] for row in rowlist]).reshape(-1).tolist()\n",
    "\n",
    "\n",
    "iws_perfs = [\n",
    "        0.3617, 0.4616, 0.3666, 0.3656, 0.5125, 0.8163, 0.9999, 0.9640, 0.9999, 0.7908,\n",
    "        0.5107, 0.2731, 0.2696, 0.4797, 0.5448, 0.9874, 0.8873, 0.9200, 0.9616, 0.9816,\n",
    "        0.7422, 0.3580, 0.0917, 0.1631, 0.6109, 0.9999, 0.6142, 0.9999, 0.6368, 0.8596,\n",
    "        0.2758, 0.7239, 0.0993, 0.1757, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
    "]\n",
    "snuba_perfs = [\n",
    "        0.6842, 0.6409, 0.8387, 0.8105, 0.9090, 0.5379, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.9608, 0.7183, 0.9557, 0.9716, 0.9956, 0.9999, 0.8072, 0.9999, 0.9971, 0.9999,\n",
    "        0.8761, 0.9772, 0.9230, 0.9320, 0.8998, 0.9999, 0.5912, 0.9999, 0.9884, 0.9999,\n",
    "        0.7667, 0.7262, 0.1521, 0.1323, 0.0000, 0.4116, 0.0000, 0.9999, 0.4781, 0.2716,\n",
    "]\n",
    "snub_multi_perfs = [\n",
    "        0.7422, 0.8129, 0.4470, 0.5407, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.3472, 0.5807, 0.4763, 0.3472, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.6185, 0.6337, 0.8232, 0.7193, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.5027, 0.3372, 0.2033, 0.7235, 0.0000, 0.9999, 0.0000, 0.7320, 0.5926, 0.2816,\n",
    "]\n",
    "\n",
    "\n",
    "iws_perfs = np.array(iws_perfs)[inds].tolist()\n",
    "snuba_perfs = np.array(snuba_perfs)[inds].tolist()\n",
    "snub_multi_perfs = np.array(snub_multi_perfs)[inds].tolist()\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(len(iws_perfs)).tolist() + \\\n",
    "        np.arange(len(snuba_perfs)).tolist() + \\\n",
    "        np.arange(len(snub_multi_perfs)).tolist(),\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        [r'Interactive WS'] * len(iws_perfs) + \\\n",
    "        [r'Snuba'] * len(snuba_perfs) + \\\n",
    "        [r'Snuba (multipolar)'] * len(snub_multi_perfs),\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1.0 - np.array(\n",
    "                iws_perfs + \\\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs,\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[0, 1], \n",
    "        xlim=None, log=True, legend=False, frameon=False)\n",
    "\n",
    "\n",
    "\n",
    "# TODO index into each of the above by task\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(0*ntask, 0*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(1*ntask, 1*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(2*ntask, 2*ntask+ntask).tolist() * 4,\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        ([r'Raw features'] * ntask + \\\n",
    "        [r'PCA'] * ntask + \\\n",
    "        [r'ResNet-18'] * ntask + \\\n",
    "        [r'CLIP'] * ntask) * 3,\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1 - np.array(\n",
    "                iws_perfs + \\\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs,\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[1, 1], \n",
    "        xlim=None, log=True, legend=False, frameon=False, colors=[\n",
    "                f\"C{i}\" for i in [5, 6, 8, 9]])\n",
    "\n",
    "\n",
    "axs[0, 1].set_title('Diverse tasks')\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "ntotal_task = 10\n",
    "nrows = 4\n",
    "rowlist = [0, 1, 2, 3]\n",
    "tasklist = [7, 8, 9] #[0, 1, 2, 3, 4, 5, 6]\n",
    "assert len(tasklist) <= 10\n",
    "ntask = len(tasklist)\n",
    "inds = np.array([[t + (ntotal_task * row) for t in tasklist] for row in rowlist]).reshape(-1).tolist()\n",
    "\n",
    "\n",
    "\n",
    "iws_perfs = [\n",
    "        0.3617, 0.4616, 0.3666, 0.3656, 0.5125, 0.8163, 0.9999, 0.9640, 0.9999, 0.7908,\n",
    "        0.5107, 0.2731, 0.2696, 0.4797, 0.5448, 0.9874, 0.8873, 0.9200, 0.9616, 0.9816,\n",
    "        0.7422, 0.3580, 0.0917, 0.1631, 0.6109, 0.9999, 0.6142, 0.9999, 0.6368, 0.8596,\n",
    "        0.2758, 0.7239, 0.0993, 0.1757, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
    "]\n",
    "snuba_perfs = [\n",
    "        0.6842, 0.6409, 0.8387, 0.8105, 0.9090, 0.5379, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.9608, 0.7183, 0.9557, 0.9716, 0.9956, 0.9999, 0.8072, 0.9999, 0.9971, 0.9999,\n",
    "        0.8761, 0.9772, 0.9230, 0.9320, 0.8998, 0.9999, 0.5912, 0.9999, 0.9884, 0.9999,\n",
    "        0.7667, 0.7262, 0.1521, 0.1323, 0.0000, 0.4116, 0.0000, 0.9999, 0.4781, 0.2716,\n",
    "]\n",
    "snub_multi_perfs = [\n",
    "        0.7422, 0.8129, 0.4470, 0.5407, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.3472, 0.5807, 0.4763, 0.3472, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.6185, 0.6337, 0.8232, 0.7193, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.5027, 0.3372, 0.2033, 0.7235, 0.0000, 0.9999, 0.0000, 0.7320, 0.5926, 0.2816,\n",
    "]\n",
    "\n",
    "\n",
    "iws_perfs = np.array(iws_perfs)[inds].tolist()\n",
    "snuba_perfs = np.array(snuba_perfs)[inds].tolist()\n",
    "snub_multi_perfs = np.array(snub_multi_perfs)[inds].tolist()\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(len(iws_perfs)).tolist() + \\\n",
    "        np.arange(len(snuba_perfs)).tolist() + \\\n",
    "        np.arange(len(snub_multi_perfs)).tolist(),\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        [r'Interactive WS'] * len(iws_perfs) + \\\n",
    "        [r'Snuba'] * len(snuba_perfs) + \\\n",
    "        [r'Snuba (multipolar)'] * len(snub_multi_perfs),\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1.0 - np.array(\n",
    "                iws_perfs + \\\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs,\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[0, 2], \n",
    "        xlim=None, log=True, frameon=False)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# TODO index into each of the above by task\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(0*ntask, 0*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(1*ntask, 1*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(2*ntask, 2*ntask+ntask).tolist() * 4,\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        ([r'Raw features'] * ntask + \\\n",
    "        [r'PCA'] * ntask + \\\n",
    "        [r'ResNet-18'] * ntask + \\\n",
    "        [r'CLIP'] * ntask) * 3,\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1 - np.array(\n",
    "                iws_perfs + \\\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs,\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[1, 2], \n",
    "        xlim=None, log=True, frameon=False, colors=[\n",
    "                f\"C{i}\" for i in [5, 6, 8, 9]])\n",
    "\n",
    "\n",
    "axs[0, 2].set_title('Text tasks')\n",
    "\n",
    "plt.suptitle('Coverage performance profiles across task types')\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"neurips/coverage_all_tasks.pdf\")\n",
    "plt.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Coverage analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of problems per solver:  16\n",
      "Number of problems per solver:  12\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 378x252 with 2 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 247,
       "width": 362
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(2, 1, figsize=(5.25, 3.5))\n",
    "\n",
    "\n",
    "\n",
    "ntotal_task = 7\n",
    "nrows = 4\n",
    "rowlist = [0, 1, 2, 3]\n",
    "tasklist = [0, 1, 2, 3] #[0, 1, 2, 3, 4, 5, 6]\n",
    "assert len(tasklist) <= 7\n",
    "ntask = len(tasklist)\n",
    "inds = np.array([[t + (ntotal_task * row) for t in tasklist] for row in rowlist]).reshape(-1).tolist()\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "iws_perfs = [\n",
    "        0.3617, 0.4616, 0.3666, 0.3656, 0.5125, 0.8163, 0.9999,\n",
    "        0.5107, 0.2731, 0.2696, 0.4797, 0.5448, 0.9874, 0.8873,\n",
    "        0.7422, 0.3580, 0.0917, 0.1631, 0.6109, 0.9999, 0.6142,\n",
    "        0.2758, 0.7239, 0.0993, 0.1757, 0.0000, 0.0000, 0.0000,\n",
    "]\n",
    "snuba_perfs = [\n",
    "        0.6842, 0.6409, 0.8387, 0.8105, 0.9090, 0.5379, 0.9999,\n",
    "        0.9608, 0.7183, 0.9557, 0.9716, 0.9956, 0.9999, 0.8072,\n",
    "        0.8761, 0.9772, 0.9230, 0.9320, 0.8998, 0.9999, 0.5912 ,\n",
    "        0.7667, 0.7262, 0.1521, 0.1323, 0.0000, 0.4116, 0.0000,\n",
    "        ]\n",
    "snub_multi_perfs = [\n",
    "        0.7422, 0.8129, 0.4470, 0.5407, 0.9999, 0.9999, 0.9999,\n",
    "        0.3472, 0.5807, 0.4763, 0.3472, 0.9999, 0.9999, 0.9999,\n",
    "        0.6185, 0.6337, 0.8232, 0.7193, 0.9999, 0.9999, 0.9999,\n",
    "        0.5027, 0.3372, 0.2033, 0.7235, 0.0000, 0.9999, 0.0000,\n",
    "]\n",
    "\n",
    "\n",
    "iws_perfs = np.array(iws_perfs)[inds].tolist()\n",
    "snuba_perfs = np.array(snuba_perfs)[inds].tolist()\n",
    "snub_multi_perfs = np.array(snub_multi_perfs)[inds].tolist()\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(len(iws_perfs)).tolist() + \\\n",
    "        np.arange(len(snuba_perfs)).tolist() + \\\n",
    "        np.arange(len(snub_multi_perfs)).tolist(),\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        [r'Interactive WS'] * len(iws_perfs) + \\\n",
    "        [r'Snuba'] * len(snuba_perfs) + \\\n",
    "        [r'Snuba (multipolar)'] * len(snub_multi_perfs),\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1.0 - np.array(\n",
    "                iws_perfs + \\\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs,\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[0], \n",
    "        xlim=None, log=True, frameon=False)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# TODO index into each of the above by task\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(0*ntask, 0*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(1*ntask, 1*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(2*ntask, 2*ntask+ntask).tolist() * 4,\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        ([r'Raw features'] * ntask + \\\n",
    "        [r'PCA'] * ntask + \\\n",
    "        [r'ResNet-18'] * ntask + \\\n",
    "        [r'CLIP'] * ntask) * 3,\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1 - np.array(\n",
    "                iws_perfs + \\\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs,\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[1], \n",
    "        xlim=None, log=True, frameon=False, colors=[\n",
    "                f\"C{i}\" for i in [5, 6, 8, 9]])\n",
    "\n",
    "\n",
    "plt.suptitle('Coverage on image tasks')\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"neurips/perf_coverage_all_image_tasks.pdf\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of problems per solver:  12\n",
      "Number of problems per solver:  9\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/hdd2/nick/github.com/Kaylee0501/FWRENCH/optperfprofpy.py:185: UserWarning: Solvers do not solve 100% of problems. Total amount of problems solved: 175.0\n",
      "  \"Total amount of problems solved: {}\".format(100 * solver_taus[:, 0].sum())\n",
      "/hdd2/nick/github.com/Kaylee0501/FWRENCH/optperfprofpy.py:185: UserWarning: Solvers do not solve 100% of problems. Total amount of problems solved: 188.88888888888889\n",
      "  \"Total amount of problems solved: {}\".format(100 * solver_taus[:, 0].sum())\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 378x252 with 2 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 247,
       "width": 369
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(2, 1, figsize=(5.25, 3.5))\n",
    "\n",
    "\n",
    "\n",
    "ntotal_task = 7\n",
    "nrows = 4\n",
    "rowlist = [0, 1, 2, 3]\n",
    "tasklist = [4, 5, 6] #[0, 1, 2, 3, 4, 5, 6]\n",
    "assert len(tasklist) <= 7\n",
    "ntask = len(tasklist)\n",
    "inds = np.array([[t + (ntotal_task * row) for t in tasklist] for row in rowlist]).reshape(-1).tolist()\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "iws_perfs = [\n",
    "        0.3617, 0.4616, 0.3666, 0.3656, 0.5125, 0.8163, 0.9999,\n",
    "        0.5107, 0.2731, 0.2696, 0.4797, 0.5448, 0.9874, 0.8873,\n",
    "        0.7422, 0.3580, 0.0917, 0.1631, 0.6109, 0.9999, 0.6142,\n",
    "        0.2758, 0.7239, 0.0993, 0.1757, 0.0000, 0.0000, 0.0000,\n",
    "]\n",
    "snuba_perfs = [\n",
    "        0.6842, 0.6409, 0.8387, 0.8105, 0.9090, 0.5379, 0.9999,\n",
    "        0.9608, 0.7183, 0.9557, 0.9716, 0.9956, 0.9999, 0.8072,\n",
    "        0.8761, 0.9772, 0.9230, 0.9320, 0.8998, 0.9999, 0.5912 ,\n",
    "        0.7667, 0.7262, 0.1521, 0.1323, 0.0000, 0.4116, 0.0000,\n",
    "        ]\n",
    "snub_multi_perfs = [\n",
    "        0.7422, 0.8129, 0.4470, 0.5407, 0.9999, 0.9999, 0.9999,\n",
    "        0.3472, 0.5807, 0.4763, 0.3472, 0.9999, 0.9999, 0.9999,\n",
    "        0.6185, 0.6337, 0.8232, 0.7193, 0.9999, 0.9999, 0.9999,\n",
    "        0.5027, 0.3372, 0.2033, 0.7235, 0.0000, 0.9999, 0.0000,\n",
    "]\n",
    "\n",
    "\n",
    "iws_perfs = np.array(iws_perfs)[inds].tolist()\n",
    "snuba_perfs = np.array(snuba_perfs)[inds].tolist()\n",
    "snub_multi_perfs = np.array(snub_multi_perfs)[inds].tolist()\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(len(iws_perfs)).tolist() + \\\n",
    "        np.arange(len(snuba_perfs)).tolist() + \\\n",
    "        np.arange(len(snub_multi_perfs)).tolist(),\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        [r'Interactive WS'] * len(iws_perfs) + \\\n",
    "        [r'Snuba'] * len(snuba_perfs) + \\\n",
    "        [r'Snuba (multipolar)'] * len(snub_multi_perfs),\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1.0 - np.array(\n",
    "                iws_perfs + \\\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs,\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[0], \n",
    "        xlim=None, log=True, frameon=False)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# TODO index into each of the above by task\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(0*ntask, 0*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(1*ntask, 1*ntask+ntask).tolist() * 4 + \\\n",
    "        np.arange(2*ntask, 2*ntask+ntask).tolist() * 4,\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        ([r'Raw features'] * ntask + \\\n",
    "        [r'PCA'] * ntask + \\\n",
    "        [r'ResNet-18'] * ntask + \\\n",
    "        [r'CLIP'] * ntask) * 3,\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1 - np.array(\n",
    "                iws_perfs + \\\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs,\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[1], \n",
    "        xlim=None, log=True, frameon=False, colors=[\n",
    "                f\"C{i}\" for i in [5, 6, 8, 9]])\n",
    "\n",
    "\n",
    "plt.suptitle('Coverage on non-image tasks')\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"neurips/perf_coverage_all_non_image_tasks.pdf\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Polarity analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of problems per solver:  16\n",
      "Number of problems per solver:  16\n",
      "Number of problems per solver:  12\n",
      "Number of problems per solver:  12\n",
      "Number of problems per solver:  12\n",
      "Number of problems per solver:  12\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/hdd2/nick/github.com/Kaylee0501/FWRENCH/optperfprofpy.py:185: UserWarning: Solvers do not solve 100% of problems. Total amount of problems solved: 116.66666666666667\n",
      "  \"Total amount of problems solved: {}\".format(100 * solver_taus[:, 0].sum())\n",
      "/hdd2/nick/github.com/Kaylee0501/FWRENCH/optperfprofpy.py:185: UserWarning: Solvers do not solve 100% of problems. Total amount of problems solved: 141.66666666666669\n",
      "  \"Total amount of problems solved: {}\".format(100 * solver_taus[:, 0].sum())\n",
      "/hdd2/nick/github.com/Kaylee0501/FWRENCH/optperfprofpy.py:185: UserWarning: Solvers do not solve 100% of problems. Total amount of problems solved: 158.33333333333331\n",
      "  \"Total amount of problems solved: {}\".format(100 * solver_taus[:, 0].sum())\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABvYAAAI0CAYAAADGGB4dAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAABYlAAAWJQFJUiTwAADPiUlEQVR4nOzdd7gkRbn48e8LS1hwXZIiV9QVBMGIgjkAcgnqVbmGn1kxe9VrTqhXwXDNOVwza06omBUDIALmjCIILIqSYQFhye/vj6rm9M7OzJk5Z86ZnXO+n+fpp2e6u7qrY3V3dVVFZiJJkiRJkiRJkiRp/bbBuCMgSZIkSZIkSZIkaXpm7EmSJEmSJEmSJEkTwIw9SZIkSZIkSZIkaQKYsSdJkiRJkiRJkiRNADP2JEmSJEmSJEmSpAlgxp4kSZIkSZIkSZI0AczYkyRJkiRJkiRJkiaAGXuSJEmSJEmSJEnSBDBjT5IkSZIkSZIkSZoAZuxJkiRJkiRJkiRJE8CMPUmSJEmSJEmSJGkCmLEnSZIkSZIkSZIkTQAz9iRJkiRJkiRJkqQJYMaeJElar0TE0RGREXHQuOMiDSqK50TEbyPi8noMZ0SsGHfctDhExLKIeEdEnBoRV9Xjb1Udd1D9f3SXcF5ztV6IiFX1WNxr3HHR+mMxHRcRsVf72r3YRcQhdXusHMOy3ReSpPXaknFHQJI0eSLiQOCr9e8PMnPfMUZHktYHrwBeX39fAZxTf187nuhoEfoK8O/19yXAhcB544uO1mc1k2Qv4LeZecQ44zKXImI34EBgVWauHGtkNO8iYgnwOOBRwB2BrYHLgLOB04BjgR9l5s/HFklNnPosvBtwdGYePdbISJIWLTP2JEkz8cTW7/tFxE0z8x9ji40kjd/zav+FwLsyM8cZGS0uEXFbSqbe1cB9M/OnHZNcDPwF+Nt8x03rrb2A1wCfAI4Ya0zm1m6U9TwGWDnWmGheRcSNgG8De7QGXwEEcGtgF+ABlOvjFvMdP633Lqekm92ecQ9k6nn46HmKjyRJa7EqTknSUCJiG+CBlK9dP0tJSx4/1khJ0hhFxI2BG9W/HzFTT2Nw29r/fZdMPTLzq5m5S2Y+YZ7jJUnj8mlKpt6lwEuB7TJzaWZuASwH9gU+AKweVwS1/srMn9d0c59xx0WSpG7M2JMkDevRwEbA14EP1WFP7D25JC14S5sfmfmvcUZEi1ZzDHr8SVr0ImIXYL/698mZ+dbMPLsZn5mXZuYPMvPZlJJ7kiRJE8WMPUnSsJpMvM9Q2qX4G7BLRNx1uoARsWtEfDAiTo6IyyNidUT8ISLeExG79wizdUQcGhG/qtNfXsN/vrZv0J522gbWI2JlneaQjuFrNZAeEfePiO9ExLkRcV1EPL817X0j4t0R8bOI+GdEXFWn+25EPHyA7TDQOkXEx2ucDp9mfofW6Y6fbtmtMBvWdfxQjcc5dT3+GRFfjYj79Ql7dF3eQRGxtG73v0TEmrodPh8RO02z/AMi4kcRcXFEXBIRP42IkZT8jIibRcTbI+KPEXFp7f4UER+LiL17hNmxbovTIuKKiLgoIn4cEU+NiA07pt20xjkj4j+mictJdbrndRl3g4h4RUT8om6HKyLilHo+3KzH/NrbfouIeHNdxuURsbo13TYR8ayI+Fodf2lEXFa3wzsi4t+mifeWEfHOiFgVEVdGxN8j4qN12651rvQIf7t6/J5e12t1RBwXEc+MiI36LbvH/K4/tyNig4h4QUT8rq7TBRHx9ZjmGlTDPT4ivh8R57WO9y9ExN0GXO5zIuLndX0yIp4fEQmsaoXJVndIx/xuWOf5u4j4V+1+H+UcXj7DOOzWsdwVEXHriPhMRJxVj43fROv8iuLpEfHLemxcGOW8vXmPOCyrx9wXo5xXq6Oc73+NiA9Hn/O9I143j4iPRMSZ9bg6PSLeFhE37L3nZpx23Cgi3lin+1c9Vv4YEW+IiK36La/H/DrTiAdFxFFRrhX/iogTIuIxA26HXSPiE1HOq6sj4oiOae8UEZ+u46+MiPMj4nsR8bAu8z2kHoMr66A9O47Bvep0B9X/R89g3Yc+d2q4PSPi8Lq/r4pynTslIo6IiGdExFDPojFPaW+ddq3tFRGPjYhjolxvMtZOqzeIiKfU8RdGueadXs+NW/WJzy0j4v9qHNbU+JwR5Tp/cJQaEuhYzkH1uLugHjvnRcSJUa63Bwy4HVfUY+Y1ddATO46ZjIgVrel3johXR0mz29f0n0bEiyJiadcFlbB3jIhPxlRacmmUdPa7Ua6fmw0S5zqvLaOcZxnlGrrtAGESOKz+7Tw3Msp5fd/6+8qI2LrPvHaIcj+YEXHr1vB2urxllLSzuZc4sx4H200Tz5ncD4zkeGjNb5OIeETdX7+Lct25oh6Tn4ke19oadlVre24V5T7j9LpN/xHluj/dNnhsPab+Vc+jH0XEA4dZhw63b/3+Zr8JM/OKLvFZ65kiIp4Y5dpzaZR7wKMiomsb451he0zT9Xmky3TDpjXbR8SL6zl2SpTryiVR7gMOjYgt+i1vkPhGuQ8+NMr9ZXPv/7mI2HmaeWwb5f68uW+9OMr9zIsiYpMZxGlG6xoDPvN1TtcextTz8Gui47pSp3t1/f/LadbhSXW6v8eQaaIkSWSmnZ2dnZ3dQB2lqq8Ezgc2qsPeVIe9f5qw/w1cU6dNSqmCi1r/j+4S5j51Wc00VwIXtOfTMf0hdfjKPvFYWac5pGP4XnX4KuBF9fd1NY7XAM+v092gFZ8ELqG0zdEe9qE+yx94nYB7tqbZusf8NqhxTuCpQ+zL23XE+eK6T9rDDu4R9ug6/rnAr+vvKyhtUTRhLwB27BH+Ja3pmm18bf3/9tb8D5rBMfqwjnisAS6sy0lgVZcw/1Gna8KsBq5q/f8+sHlHmE/UcZ/tE5c712muAW7SMW7X1n5LSrtY7e1/IXCvPtv+JcCprW1/CbC6Nd3bOua91jEGnAvcoUe8twdOb017OaUaqybcU3ttyxr+Oa39mTVse9lHAZsNuV8PqWE/AXyltV6rW/O9Bnhkj/DL6n5sH3ft8/Za4DnTLPeI1nKaa9c9gbOB81rzOrvVvbg1r1t17PPLatf8PwPYaQZx2K1O18zn/9XjoTmWr2uNexGlbaHP1v9XsfZx9ze6XGvqPm1v5wso16X29fzfe2z7ZpqH1HDNdfPq1rhfUNOULuFnknbcu7Ws5hraPsf/Btx6yGNwL6bSiOe3jqP29SuB902zHR7f2u+X1Hgd0Zru6R3zu6hj/T8FbNia/sX1WGuO56tY+xi8Z53uoD7b62h6XHOZ+bnz9NY0zfHemcZsOsT2n7e0t3N7Ae9preuFtX9gnW4z4Hut+V7F2telNcBDeqQPl3SEax/XCRzQEeYzHeNXs/Z5+NMBt+XN6rHxr1Ycz+7obtaa/pcd63MBa19XfgEs67KcB7B2WnpFl322S0eYVXX4Xh3DbwL8vo47AdhywHXtd260z4+/1Gn+u8+8Xlen+UmP8+dFwF+ZSjfbx/u5wK495jvT+4GRHA+t+f1HK+x1dblrOuL1+B5hm/g/rvX7srrPm/Cn99pvwPta011LOReaY+y5vY6LadbnEa15dr0fnSb8ITXsSuCjTKV/nde/h/UL22f+K+s0h3QM34vZpTWHt6ZprnPtcH8Ftp/B9mji+0bKOdjMv709LqO08dot/F1ZO11u0r/m/2+BGw+zLWe6rgz+zHf9dK2wzX1fE/d/0XFdqdNt34rL7fts12PrNK8fdp/Y2dnZ2dmNPQJ2dnZ2dpPTAW+pDx8faA27fR12AbBxj3Dth+sv0Xq5AWwFPBZ4e0eYHVsPi78B9qa+zKRUObYv8OWOMD0f/lrTNA+mh3QMbx7e1tSHuvcD29ZxmzYPhpSXeF+iNJq+VSv8FsCzmcoAeUSXZc9knU6s0z+3x/r8e+vBcp0Xa322w87AxyjVFN2wNfzGwKvqNrgOuFuXsEfXZV5EeVGzP7AhJZPxPsDf6/gvdgl7b6Ze1nyKmuFVt9+bmXo5lQyZsUd52G4yC34E3AWIOm5Z3Wcf77JPmhdoR1Nf9gObUF5MNy+lPtoRbv/Wdu+aSQW8tU7z/Y7hy5nKOPsicIfWcbADUy/rzga26LHtL6VkThwAbFDH3ao13XOBgynn55I6bENgd+C7dR5/bLZPj2WcTWlPc4PW9j2JqZfPq7qEPZCpFzYvAbapwzeu2+xkpnkB32NbHtI6Nq4BXgAsbe3DI5l6mbrOCzzgq3X8ryjH/KZ1+JbAKykvfK+l4+Vpa7mX1mPhv5r9TTlXblh/r6BL5kBrPhsDv6vT/I1yrkft9qFk6jX7ZJMZxqG5xq4GvgHcsg6/IfB/TL10e12d1+NqvIJyXp5Vp3lLl/g/Cng95ZzauA4LSvVln2bqxfXmXcI28boI+CFwu9Y59mSmzrFndQk7k7TjFkwdox+gZKhuULvbMZUJcyKtDLIBjsG9WtvwKkpGa5NGbMnamemP6bMdLqWcY7drbccdW+dY8yLwS0ylOzegHKfNtfNVXeZ/UB13dI/49xxP/4y9oc8dSjrZpIUfY+1Moq0o163P0uOeoUf85zvtbbbXpXW7v5p6PaacUzeuvz9Yp7sCeAb1/KWksUe1jpmdO+b/ozrup8CdOtZzD+CdwD1aw+/LVMbC86npfT1+tqOUHnnbDK+rK6eZ7v3AU4BbtIZtAjyIqQyxdT7uAk6r477RXv+6/e4DfBhY0RFmFR0ZOJRz+pQ6/Ad0uc5ME/9mX3Y9N+o0L63T/LrH+A0o1+6kVOvY7fxZDZxDySBr0s09W9vhj3R8wMAM7wfm6HjYC3h33TebtYbfvB6PSblHvnmXsM1+u4hyjt2jDl8CPJipa3K39OWxTF0f38rUebYt5Tp7FVMfQ+w1xPrs0Jrvd4EbzfD8uKiu9zOZSntvCRxTx/+Tep81zLnF9M8jM01rXkf5IGan1nG4UT0Wf17DfWuYbdER39U1bo9n6iPP3ShpRHOsbtkRdsu6nZKSQX+XOnxD4OGUTOSk4355um0503Vl8Ge+ZrpVg+6/jmm+Xad5R4/xOzGVqTh05rOdnZ2dnd3YI2BnZ2dnNxldffhqHsru3TGu+Yq621erGwFn1vE9Szd1CffFGuYvDJhhNaIH6aHi2WX+j6/zOGpE6/SCGuY3PcZ/drp1nuF6/E+d72Fdxh3NVCbKrbqMfxhTLzo37hj3Q6Yy3rplKjVfRSfDZ+z9rIY7hh6lf7qE+VgN81e6ZNAxVerkOtbOONuQ8gIvgUd3CRdMvQR8Use41093nAHfqdO8uGN4s+2vomYMzGDfbsJUhvGeHeP2bq1vtxICK5gqEbmqY9yGTL3c27/HsnekvBC6GthuiDg353YCr+wyflNKpmOybiZsk/l9ErC8x/xfXqf5Zp/lPr1P/FY00/UY31wXuu43SmnopmRL50vjQePQTHMy675k3ICpF+MJPKFPHE8b8ngKpkp0PbFPvNbJtKzj31vH/6hj+EzTjiaj8Y09xrczWR8+xHz3aq3LkXS/fq2s40/pHN8Keyo1U7pL+Ob6+BO6ZDoC/8tUZtMNO8YdxIgz9mZ67lBKZiTlw4eBM09n0zH6tLfZXgn8b49pVjCVEfuMLuM3Y6oE1yc7xjXX0XU+numxrCbj6Tsj3GaHMMv7B0oGx9WU63o7M+jGre237RDzW0UrA4fy8UDzsdARdLmGDLEvu54brfg21+A7dhm/X+vcu0HHuOb8uY6O++M6/tZMlaR7XMe4Gd0PzMXxMMB2bO6XXtNnv51N91LfTamo0zqGR+scWec4ZO305frjYog4f6IV9kpKxvDrKSXI+2b0sXba+9gu4/+ttV/v2yNsz3OLwZ5Hhk5rplmnrSgf4VxHR6b6AGFX0n97bMNUqehXdYxrnikuoqMGizp+v9a87zfsthx2XRnwmY/ZZ+z9Z53mXLo8lzCVph89zLrZ2dnZ2dk1nXU4S5IGtS/lK+AzgOM6xn2m9p/YJdw+wE0pL79eMsiCIuIGlIchgFdn5qVDx3Z23jqLsN+o/btHq222WazTJykvm3aLiDu1R9S2I5p5fnzGMe6uWY979Znm8Mz8a5fhX6c8qG5CKS0DQJR2rZo27t6cmdkl7P/OIK5ExC6Ul8kAL83MqwcIE5RMSIB3ZublXSb7KPAPysul69twysxrKS+LAR7dJdy9KdWdXUGpOrKtOU/e3id6n639ru23UF7m/bFP+J4y80rKizJYd/8+tPaPy8zO85zMXAV8vses96KUrPhjZn6vx7JPpZRQWVKnH9blwLu6zPcKprbnw+q+bTTb+yOZeXGP+TbXsL2jo03F6gJmd441x87Xuu23zDyRUqUUlKo0uxk0Dm/LzGs65n8dJTMdSmbZp7uE+2Ht3zIiNh9gOc28E/hW/dvvevGOeux1OqL2b9cxfCZpx2aUUn7XAe/oEd+rmNrWvc6v6byxx/XrDbV/K+COPcK+LzPXdA7suD6+sV5jOr2Zck25AaWaw7k203PnktrfCOjZZtmIjTrtbVxLj2OpzncDSmbGRztH1jTlLfXvQzuuLc026tvuWJfpb7w+tcOUmadTPhTZjFJqp/EvynkIg6/jWiLizpRq6ranHGcP73ENmbXMPJepY+jJXSZ5Uu1/KTP/1WM2x2bmT7rM+y9MXXM624Kc6f3AOI6HQe4LP5yZF3QZfkTtd6Yvu1E++IFSxeNa6nV2RveF1dMo5+9VlI869qGUND4COLe27/bYjnuGTn9jah+04/ZPSqkwWDf9GpXZpDXryMwLgeMp97T3nGGczqD79jgf+FD923mcN/8/mplndwl7JKV6T+h9DzSUIdZ1Ns980/kG5SPAG1FK8l6vnrdPqH9H/QwnSVok1puHAknSeu+g2v9cl4fMz1Eycu4fETfqGHf32v9dZv5jwGXtQXnxn5Tqc+bTGkppjp4iYklEPKU21n5WRFzZajD9ojrZppSqZxozWqf6guSI+vdJHaMfXZdzSmb+eNB5ttZjaUS8ICKOrg3GX91aj9/Uyf6tzyx+0SPOV1O+ToW1t8GdKA/Y11FKpHQLexrl6/xhNcfZhZn5swHD7ECpBgtKlWnd4nMd5Wt8KG0itTUvNvavL+XbHlP7326/EI+Im1FeUgJ8OyLO7tZRqsOCkjnYzQk9hl8vInaJiPdFxO8j4pKIuK61f59XJ+vcv03mcdf9Ux3bY3jz4mSnXutV162Zrte69fPLzLysx7hjan8LSimSzni9qk+cmmN5M7pnRPyyM7NsSM2x0/U4q5qMt87jbNg4/KHH8Oac/FM9rjud0/q9RefIiNg+It4cEb+KiNURcW3reHpnnWzo6wUl4xzWvlbAzNKO3ZmqXvQPffb3i+v0MzkGr2bdj1sAyMxTKFWaQu/92Ovcba6PydSx3Dn/iynVnfWb/yjN9Nw5pXYbAyfUdGaXaV6eT2s+096Wv9YX1t00++DYHhmxMHVeb04pudX4du1/MiLeFBF3j4iN+sTjh5TMiTsDR0fE4yKi3/k2UhGxb0R8LiJOjYjLm+1et32TsXB9fGqmZnMcfy8iXhURu/X4cKKb+1Cul9tQqhJ+/CyvwYNoMmcfGxEbNwMjYktKFbBQSq31cnSfcc22uP68neX9wJwcDxGxVUT8T0QcHxEXRMQ1rf381TrZbK7zsHb60myPc2oGaDfHU6pLHFpmXpWZL6Jsu2dSnlea0utQqpf+NPCFPhmkv+yRuQa9069RmHFaExF3jYiPR8RJEfGvjvP1IXWymR4vx/TZHs1xfrvmHKr9JuNzNvdAXc1yXad95puNes36RP3b+Qy3P+XjpUuYyviXJGkoS8YdAUnS+i8iljP1cNTtK82/RcSxlDY/HsPUSwgobWRA+eJ1UE2Yi/uUEpgrF/R46Q1c//X/91j76881wHlMfZ3exH9zSrU07WEzWaePUr5gfUxEvLiWOIGpr8oPG3J+RMR2lJdQO7cGX0Z5OXodpVrFbSjr0Eu/kg9X1H77RWWT6Xtxn8wZKC9Khn3hPpPjrJ0J3S/j4Mwu05OZJ0TE6ZRMpIcBH4Hy8pmpr5M7z5d2yYUbDxDHzXoMP69foIh4FKW0Z7P9r6O0MdWUdrgBZd927t9tav8sevtnj+HNum3C1P7op9e69dNvP7XH3YjSrlE7XlsMuIxu8eq7vQfQHDuDHGdbR0R0eXE2aBx67btr+43PzGtb+S5rZTBExJ7ANynHTeNips7zpZR2s2ZyvWjm0flcMpNzutnXwdwdg+e3rsHd/KPGo/Mjl0av/di+PvYqEQQ9rkdzZEbnTj2WHkP5KGUHSomZdwAXRsSPKO2rfqPPy+F1jCnthf7n3TDndXt6KKVQb01Zn5fV7oqIOIHSluDKdsnOzDwlIv4LeB8l0+s+ABGxipJh+eHM/A1zICLeQ2nHqnE1pU2spmT8VpRrRuf5/1TKdWNXSltYrwP+FRE/pmSwfL5PZt1ra/9HmfmsWa/EYL5H+bDoZpT2A79chz+GkmH8l24l2VsGSZ/ax8CM7wfm4niIiNtQMlfa185LKedZUjLqt2QG1/nMvKJH+tJsj173FWTmlRFxPnCTfvHvp5bI/FDtiIhtKfv41ZT9/QhKJtq7uwQf9l53VGaU1kTEiyklhZsNfi3l3r6Z13LK8Txwyfwuy51u3IaUY+UcyvVhg47x3Qydto1gXfs+843IRylV594/IrbNzOYjquYZ7vM9agyRJGlaltiTJA3ikZQHI4Dft7+GbH0Ved86vlt1nJOk11f3jf+hvIg7n7Ku22bmZpl548y8CeXry8asSie0/AA4nVIa4sEAEXE7SkmEa5n6GnQY76Jk6p1GyZTaKjNv0FqPu/cLvABtOv0kXX2u9h/TGrYvJYPsEqaqKGy07722zMyYplvRY7k9j9MopWY/QnnR9AXKcbJpZm6ZmTep+7cpYTWqYxSm1u1rA6xXZOYhI1z2IPH6zwHjtarLPKa7LgxqpsfZKOMwlFqK6NOUTL0fUK71SzNzi9bx9MJm8nHEsaXZ1xcPuK/3GkMcp9uPm8xLLAYz43MnM38J7AQ8jvKRwWmUF7wPB74GfGuI0lswnrQXBjvvhj6va2n8e1PSi/dQSslvTKmO9QPAHyNi+44wH6d8SPJ8yja8gNLO3zOBX0XEK4aNx3Qi4v6UTL1rKW1d3YrSzt3WrfO/KSW/1navJfDvQKmy9MPAn5mqRvZTwM9qhm03X6j9+9UMrDlXX/A3VeK1S9c0v4f+iGoas7ofmIPj4TBKpt6vgQMo7VHeMDO3rfv5EXW6cV/nZy0zz8nMj1JKh3VmtEysiLgtpcrmoGT63pZyvm7VOl+b0mHj2I+zuQday4jWdc7vq2rpymMoHy89HiAirn+ew2o4JUmzYMaeJGkQw2TW3Skibt/63zww32KIeTRhltfSgoNqvvzu9+A4zPy6aV5s/HdmfrJ+BdzWq5TITNepaWOk82VT8wLie1na+RhYrRanKYH52Mz8SmZe1DHZIKVdhtWUfFgepS2sXmZSPVCzfW8+g/hMF655udqt5EZTIu++rWqwmjb3vpKl7be2dnWHw8R1GPenvDz9E/CYzPxVrtvmYK/925Ry6dcmUq9xM9kHw+p3bLTHtffVfMRrOk18BjnOLhimJNM8uAclbhcCD8nMY7sc13NxvZhN2nHDYa+zQ9imXU1fF81xOGwpz2b6pbFuldZt/a5Hozarcycz12TmZzLziZm5I6X03hupVXdTMiAGNe9p7wCGOa/b0wMlbc/MH2Tm8zLzzpQPQp5BOdd2YOoDjHaYczLz3Zl5IKVky10pVSQG8LqIuMNMV6aHZrt/NDMPzcxTu1yfep7/mXlNZh6Rmc/IzNtQ0o+XUEo63Rl4TY+gH2Tqg4H3R8R8Zbp8nFIC9ICI2K5uz90pGQCfnCbsIOlTt7QJZn6OjeR4iIib17DXAg/OzO/luiWH5/K+sOe2q9fbbXqNn40s1ex+rf7dud+0QxrF88hM0pqHUd7xfS8z/zsz/5TrVhM82/04yHHelJyDcj1rSsXN9F67m/lY11FpqvltnuEeS/mQ48QcvPkASZLWYcaeJKmviNiJqaqvdqNUrdKr+0adrp0R+NPav0NEtL+o7+eXlIfioLz8G9Tq2t++28javs/uQ8yvm2bevao4+vcew2e6To3DKA/K+0fELSilIGBmX3puw1SpkGHXYzZ+Q3mhuwGlpMI6IuKWzOwFV3OcbRURg5Y2PI2pY2bvHvHZANir/v115/jMPBH4PWWdHhURmzLVFk+3amtPZ+pl3kyOg0E0x+jvs0sVQ/U8uF+PsM3x0HX/VPfpMbxpO2yYc31Ye/TJFN6z9ldTSrh2xmuutvcgmmOn63FWNftkneNszJrj6eTsXV3UXFwvZpt2HDAHcYJSEvYe3UZExK2YerE57H5sro/Q+3q0nKk0bD6Ok5GeO5l5ema+gqnSWHv2m77DuNLefpp9cLc+16XmvL4M6NWGGACZeVFmfhhoSlr13T41Y/AXlMy3M+mTtvbQpA/9Su703e71fuRWgy4wM8/OzLdRag2APuuYme8EXl7j95GIeFyvaacxyHo2y/wb8H1KVYJPYOpF/Hcys18V1dB/fzXjrj9vR30/MMvj4fpMlezdpulcXOeb7bFtRPTKWLsnc9uETFM1fL9qL4e1uvZn8zwyk7RmuvN1c2ZfI8cgx/kfs1YjWvt/rMNHeQ80H+vaz8DXFUrJwdXAbSLibsxdKWBJ0iJjxp4kaTpPqP3fZebvMnN1r47SLgzAY1tVbP2Q0qbChsBbB1lg/Ur4q/XvoRGxbMC4/qH271LbkOv0WIZvu61T00bP7TtH1CqlXtkt0CzWqQn/D+A7lO34GcqX2ecBXx9mPtWlTL1A7rYe27F2ezojkZkXUtpvAXhptBpcaXn5DOd9EvDz+vcttfrA6cIk8JX693k9Xsw+lVLFWzJ1fHdqMvAeTWmzZRlwNlPr2mll7b+4X4ZFFFv0Gt9Hc4zersc2fhqwY4+wzTF6r4hY54VS/ar/UT3C/pDSPtG053pEbNlvfB+bA8/rMr9NmCrdcXhHiZKVtb9/RPTN7JlFvKbTVAd1/4i4U5fl3papdhm/OEdxmKnmeNqpZlyvJSL2o//LupmaSdpxKVPtYr2233U2Ipb0qQZwOgf3OLcOrv1TMvO3w8ywXh+Pqn9fVj8q6PQySgmQfwHfHmb+M7Sy9oc6d6YpZQKlzS4YrtrRsaS90/gK5eXu1sDTu8RrM0rpNCgluK+twzeI0hZrL+tsn37btM63KZU9zDa9pPa36DNNz+1e/S9dXmxHxEY9zpHGQMdAZr6ZUqpvA2BlRPy/ftP3MMh6tn2k9p9MuW8E+NgA4faMiHt2DqwfyDXX9877iJW1P9T9wBwcD81+3jYi1mnvr9bE8ZjO4SPwW+Cv9ffLuiw3mOF9YUTcMiJ63es002zG1MdYv53JcnoY1fPIsGnNdOfrKyn3qLOxIiIe3TkwIrZi6jrYeZw390AHddse9T6iuecc9B5oPta1n4GvK7WWg0/Xv2+nfCh7NaVKYkmSZsyMPUlST/Vh8vH171f6TVt9g/KgchNgf4BaBeCL6vhHR8QXI2KX1jK2ioinRcR7Oub1CkoG1M7AjyNi7+ZFZ0QsjYgHRkTni83jgH9Sqjf5XC39RURsFhHPoLys6axycljfr/13RMSezQN3RNyF8iJ66z5hZ7JObU1VLveq/U93qWJxWvXld1Ma5uMRsVuNwwYRsQ+lLYi5anvjEEom2T6Ul3Tb1mUvj4j/pbwUuLh38L5eSCnVeB/guxGxRzMiIpZFxKMi4jMdYf6X8rX2v1HafLp1nX6TiHgape0jgI9l5qk9lvu5uk57MPWy5YtdqgRqvIlSWnAb4PiI+H8RsbQV15tHxNMpXy0fOMB6d/pBjc/tgPc0LwMj4oYR8RLg/ZS2eLo5CjiWsv+/HBH3bx3jdwe+S4+v2uux+Jy67EdHxBHNsVXDbxQRe0TEW1i7RN0wLqZUL/a8ZptFxA6UqrR2pVTv9qaOeH2Xcv0K4KsR8ZJoVXVYr0EHRsTXgXfMMF7T+QKlZCfAERHx763tug8lk2Yj4ERKxv365Djgcsq17ZPNS7l6zXoyJSOt1/E0Y7NIO15OqfprZ8r5dUDUjP76cnyniHghcBLlnB3W5ZTr18eaF+ARsUVEvJmpKpIPmcF8obQjdx2lisLPR21jLSJuEKW9rOYF95sy85Ie8xiZWZw7D4iIE+r+uUVr+s3qdbXJLPneENEZZ9rbVWaeQWk7DuBNEfH0KB8ZEKX00bcopdkuB17fCnpD4K8R8cqIuH3UD6FaafAb6nTt7fO/EXF43d5bNQMjYtt6DtyScu39PoM7sfbvHSXzqZtmfs+IiCdHzVCq6dQnKB+0dLuvui2lncDnR8TOrf21UUQ8jKkPMaY9BjLztZRtsiHwmYg4cIB1a2vWsyktM52vA+dSjpcb1d/fHCDcJcBXIuIBrfW9D+WjrE1qPDozLWZ6PzDq4+HPlFJ+AXwhSomwZn89tM6ns2rOWasf4RxS/z45It7cumfZllIrxf0o59Cwbgv8JSK+Urfr9RlKEbF5RDyIcr9zyzr43TNbi65G8Twyk7Sm2d8PjIiDo36wFhE3ioi3Uu5RZ5teX0wpQfvYqB8oRKny9XtMnS8f6AjzPuAsYCmt+/OI2LBeDz5fp/tBZvb6KK7TfKxrP8115YDonnnbqfMZ7ptdqpSWJGk4mWlnZ2dnZ9e1o5TCyNrddsAw363Tf6FjeJPp0szvUspDbfP/6B7Lb09zBaUNsGuaYV3C/GfHci6mZDYm5YvrlfX3IR3h9qrDV02zfjtQSso1819DedmRlIfw/VrjVoxinVphl1BeFDRhbzeLfXu3Gt9mXv9q/b+A0gZfr218dB13UJ/5r6rT7NVl3Etay72O8hK+Wf+3DzL/Pst9VN2mzfwvr+tzXa/9Syllt6YV5iJK5lXz/wfA5tMs99jW9AncbZrpb0VpA6+Z/pp6HFzeMZ8nDrvt63Tv6JjPRUydF9+lvGROYGWXsDcHzujYhk0pz7OBp9Tff+mx7CcBV3bZB9e04zTkfj2khvsEJaMh6z5qn0vXAI/qEX5zSqmd9nF3EeVFbHs7HdZjuetsp47pVky3XnWfr2ot67LaNf/PAHbus+7TxaHndWfQ+fSaB/Dcju20mqnr6m8oJXx7Xceni1ffbcfM0o67UEr7NdNcRTm/2sdlAnsOcQzuVcOsAp7fOo4u7Ijf+2ayf1rTPaM1v87rY1K++t+wS7iDem2P6cbT57rCDM4dSgZEe9zlTLWz1Az7FrBkiO0/r2nvdNuzNd1mwJEdx1rnch7SEWaLju1zFeteI08Ftm+FeVdHmIu77INXDLo96zw3opSWavbruZTje1WzbErGxAmtZVzTsX7/0+34oZRIacftirqO7XPlF8ANO+K0qo7bq0t831rHXQk8cMh1Paa13Ata63n3HtO/tTX926aZd7P+L2ptz3a6mXXb3qZH+KHvB+boeOi8f76EqWvmGZQq4JPu91E991trmn7n5/s61r99vXjuIPPvMs/9O7ZHs19Wdwy7ptu2YrA0c2Wd5pABtufQzyPMLK35cmuaJlyzLT/aL87TbM8m3BspHwc25/XFreVdBty3R/i71ri0j6/2vffvgBsPsx9muq4M/szXczpKZvwFdfy1lIzLVf3mSbnmNfH9j2G2v52dnZ2dXbfOEnuSpH6eWPsnZ2lLbBBNNWgPjla1QZn5DuBOlPYEVlFeKCWlFMu7gRd0zigzjwJuDbyZ0j7DNZRqyE6llJJ6cJcwX6W84DuK8lJlQ0r1Ok/JzKcMuA49ZeZplIfTT1Ne1GxIeUnwGeAumXnkNOGHXqdW2GuYasfwF5n5x17TDrAeP6NUe3ME5SXdRnV9PkR5Ife7mc57gGW/ldKezFGUF7NLKO0gPSEzX9Qv7ADz/jyl5Nb7gJPr4CWU0jkfZapq2XaYb1Cq8vkI5djcjPLy5yeUEoT7Z+ZlneE6tNvTO7Vu337x/CvlfHgWZTtcBCynHA+/p5QCeSBTVfcMJTNfWOP+G8qLuQ3r7+fX+V7TJ+zfKCWG3gP8jalj/COUNmGaL6BX9wh/GOUYfxfli+ZrKSVULqC8AH1NHT+jVaO0H/RCSgmDjSnb7pvAPev+7xanyzLzP4H/oGQM/pOyn5sX21+kZEiOvAraVhz+CtwReC1T7c1Qf78OuENmntwt7Lhl5nuAhzJVeq85p15Daf/o0jlc9kzSjl8Au1Cqdjuecp3Zosb9l5Rje8/MPGaGcXoX5Vp9DKUGlCsoLzofl5nPmck8W/P+ECVj8rOUF4U3oLw4/T7wiMx8XPYuDTxyMzx3fkQp7f8JSpV0l1OqRLugrscTgAfVNG3QeIwt7Z1mvpdT0rOnUj7wuJyyfc6gpDm3z8yvdQS7hLI930WpQvo8yva5jPLi95XAbpl5ZivMOykZHF+jpG1BKQX2d0qJ4Ptm5v8OGferKSWCPkXJCN8SuEXtltRprqK0rdaULLuOsu2+T9mHr+sx+z9Tqp/8ICXtWU1JBy6mpK3/Ddwrhyh5mpkvoZzzG1NKlO83+NryUEopotMp51SznutUL1y1a6kYtC3jCyjH6LsobedtTDlfPkLZn3/qFmiG9wNzcTx8lVI67vuUa/pGlOP4bTV+Z/YOPTv1uvk44GeUe5agXF//o6Y/M5nn9yjn/Isp97pNlZ83oByPv6bsqzsOu60GXP6sn0dmmNY8klK6+8+UjMSgpN1PzMynznB12q6kZHi9lnJ8bEy5hn0euHNm/rjHuvwcuA3l2D2ZcnxdQ0mTX0L5IG7YEmxzva49Zeb5lA9GvkJZ/xsxdV3ppbmunEUpyStJ0qxEZo47DpIkaUARcTKwE/BfmfnBccdHi09EvA54FfCJzDxonpZ5CCUTad6WKbVFxF6UF7RnZOaKsUZG0pyKiFdSSrb/LDPvPs20RwN7Ak/KzJVzHztp/kXESsoHn4dm5iHjjc1kiojvUz6UeHNmzqjtSEmS2iyxJ0nShKht7+xE+ar/s9NMLo1cbcun+dJ8mLacJEla79U2D5vSPh/uN60kDaK2W7kPU9WESpI0a2bsSZI0ASJiG0qbLwAfH6b6KmkYEXG3iHhvROwREZvWYUsi4n6UEkvbUapE/HKf2UiSNFEiYgPg1ZS2P8+hVNMqSTMWETcA3kupKvSbtfpdSZJmbcm4IyBJknqLiLcB/w+4CaU9ivMp1UNJc2UZ8JzaEREXAZtT2lEBuBB4ZGZeMZ7oSZI0OhFxd0obYVtS2gIEeEVmrhlfrCRNsoh4PqVt65tQ2sC8gtImoCRJI2GJPUmS1m/bADcD1gBHAvebQePy0jB+S2lD7xjg78BmwNXAicDbgdtl5s/HFjtJkkZrU+AWwFLgJODpmfnx8UZJ0oTbgnJduRY4HtgvM/801hhJkhaUyMxxx0GSJEmSJEmSJEnSNCyxJ0mSJEmSJEmSJE0AM/YkSZIkSZIkSZKkCWDGniRJkiRJkiRJkjQBzNiTJEmSJEmSJEmSJoAZe5IkSZIkSZIkSdIEMGNPkiRJkiRJkiRJmgBm7EmSJEmSJEmSJEkTwIw9SZIkSZIkSZIkaQKYsSdJkiRJkiRJkiRNADP2JEmSJEmSJEmSpAlgxp4kSZIkSZIkSZI0AczYkyRJkiRJkiRJkiaAGXuSJEmSJEmSJEnSBDBjT5IkSZIkSZIkSZoAZuxJkiRJkiRJkiRJE8CMPUmSJEmSJEmSJGkCmLEnSZIkSZIkSZIkTQAz9iRJkiRJkiRJkqQJYMaeJEmSJEmSJEmSNAHM2JMkSZIkSZIkSZImgBl7kiRJkiRJkiRJ0gQwY0+SJEmSJEmSJEmaAGbsSZIkSZIkSZIkSRPAjD1JkiRJkiRJkiRpApixJ0mSJEmSJEmSJE0AM/YkSZIkSZIkSZKkCWDGniRJkiRJkiRJkjQBzNiTJEmSJEmSJEmSJoAZe5IkSZIkSZIkSdIEMGNPkiRJkiRJkiRJmgBm7EmSJEmSJEmSJEkTwIw9SZIkSZIkSZIkaQKYsSdJkiRJkiRJkiRNADP2JEmSJEmSJEmSpAlgxp6kBSUiVkVERsRe446LRqvu14yIFeOOiyRpbkXEiua632XcyjrukDFETXPAfSpJ0vzzGVuSJpcZexqLiDiwdQPx/XHHR9La6jl6iBmk0uJmei0tPDXT9JCIeP644yJpMnl/oLkSEc+vadSKccelLSIOqvHabRbz8BlbkjQyS8YdAS1aT2z9vl9E3DQz/zG22GghORW4Arh83BGZcAcydZ4ePb5oSBoz02utj84C/gKcP+6ITKgVwGuAM4B3jTUmkiaV9weaK88HbkF5Bl01zoh0OAjYkxKn385wHgfiM7YkaUQssad5FxHbAA8ELgM+SzkOHz/WSGnByMx9MnOXzPz5uOMiSZPM9Frrq8w8uKb17xt3XCRpsfH+QJIkafzM2NM4PBrYCPg68KE67Im9J5ckSWNgei1Jkjp5fyBJkjRmZuxpHJqb/s8AxwJ/A3aJiLtOFzAido2ID0bEyRFxeUSsjog/RMR7ImL3HmG2johDI+JXdfrLa/jPR8SBHdMeUtsJWNknDivrNId0DN+rDl9V/98/Ir4TEedGxHXtdkwi4r4R8e6I+FlE/DMirqrTfTciHj7AdhhonSLi4zVOh08zv0PrdMdPt+wuYXeMiA9FxGkRcUVEXBQRP46Ip0bEhj3CHF2Xd1BELK3b/S8RsaZuh89HxE7DxqXOe1Wd914dww+qw4+u/x8dEcdHxCURcV5EfDUidm1Nv11EvLfO74qI+GtEvLzPOu0cEa+OiB9FxOk1zOqI+GlEvCgilk4T79tExBfq+q+JiJPqftl0wOPyQRHxtYg4u3U8fSMi9h98600dx0ydp6+JqfYzso5rT3/niHhTRPwkIv4WEVdGxAV1H/c8BmrYPSPi8Ig4s8b54og4JSKOiIhnRMTAaVRE3LyeAxkRP4iIzVvjbhkR/1fHr6nnyxk1jgdH+epY0rpMr02v5yy9rvPfNCL+J0qad0VEnFXneZtpwq2zbyPisXXY2dOkPfeo013d7fofEber+6Odlh8XEc+MiI26TL8iWuljRNw9Stp2VkRcGxHvak07o/SoLuO9ddtfHhGX1mPqZdFK7wZRj/uj6t9bREcaHxEHtabdJiKeFeX+4qS63Msi4k8R8Y6I+Lc+y7lxRLw1Iv5Yw1wREX+Pcu/12oi4xRBx3iAiPlDjd1FE3KM1buOIeF6d7+q6X8+JiN9FxPvb00oaGe8PvD8Y+f1Bs+8o1XACHNWRPh3dJcwNIuIVEfGLKM+SV0R5nnxPRNysY9oNIuLYOq8fR5dnzbpf/lmneU8ddlCN1551ssM64rVqgHXzGVuSNHqZaWc3bx1wWyApbaJsVIe9qQ57/zRh/xu4pk6bwL+Ai1r/j+4S5j51Wc00VwIXtOfTMf0hdfjKPvFYWac5pGP4XnX4KuBF9fd1NY7XAM+v092gFZ8ELgEu7hj2oT7LH3idgHu2ptm6x/w2qHFO4KlD7s//ANa04rIauKr1//vA5l3CHV3HPxf4df3dtIvXhL0A2HEGx1izLnt1DD+oOU6AN9ffV9ft3yzzfGBnYCfg76390z7uuh6nwC9b06yp8b+uNewXwLIeYf+9YzteXPdZAicAb+x1XFK+lv10x/HTeTy9eYjtd0/g7FZ8/lX/X991TN8+Fi9j7XMygW8BS7os5+kd011Wl9UetmlHmGb4io7ht6a8UEjgCGCT1rg7d+zjq7rE8YBRXufs7BZCh+k1mF7D3KbXNwB+2rF9mu37L0qJkHX2fa99C2xOSUsS2K/Pct9Tp/lWl3HPAa5txelS1j6WjwI26wizojX+kZR7i/Y2fledbkbpEfDQjn13Wce++z2w7RDb/RfAhTXstXSk8cAjW9O+rbWcqzuPX+Bc4A5dlnEL4J+t6a6py2zfFz1zwPN1CSXzIIFzgDt2jDu6Nc/2edwM+/ywx6adnV3vDu8PwPsDmIP7A+DFlHSoSYcvZO306Ssd0+/aWu8mnWo/T14I3KsjzC2ZSotf1iUOX6rj/gQsrcMeWZffbJeLO+L1iwHWzWdsOzs7O7uRd2OPgN3i6oC31ET+A61ht2/d+G3cI9wjWjcIXwJ2bY3bCngs8PaOMDsydYP9G2BvYMM6bimwL/DljjCHMPsHgTWUm/L3U1+0AJsC29ffm9V1OBDYqhV+C+DZlJdICTyiy7Jnsk4n1umf22N9/p2pm8uuGU89wu3Yukk8Grh1Hb5JvaG8oo77aJewR9dxFwGnA/sDG1IeSu7DVKbaF2dwjK2qYffqGH4Qaz+sPI/6cq4egyfV8V8BfgYcT315VPfZK5l6uLtdl+W+H3gKcIvWsE2ABwF/ocfDLrANUzfuP2vmTcmwe0w9Hpqb5HWOS+CdddwplPNk8zp8GfBfTN1wP3rI7biSLsd5l+k+CzwKuElr2ObA44Cz6jxe0hFmM6aO848BN+s4nw+o8924I9w6Dx3AnSgvFpOSwbmkI8yP6rifAnfqiMMedfvdY9jjzM5uoXeYXjfXCdPruUuvP1LDXk5Jo5sXxHegZD6tbo6lIfbt5+rww3osc0PKC7QEHtsx7kCmXtC+BNimDt+4rvfJdHlZy9oZe5cCh1PTKUrGU/N76PQIuAvlnuVq4PXATVvrcY+6nRL43pDbfq8abtU00z0XOJhy7i9pLXt34Lt1Hn8EoiPcx5m6N7kPsEHrmLsd8DrgwOn2KeV8/Hod/jdg544wT2DqxeXjqC8raxxvTjlPDx722LSzs+vd4f1Bc932/mCen+c7pllel5vAFyn3Ds123IGpD0LOBrboCHsQU5mlu7WGN2nKVcDufdb5oFmcP12PvS7T+YxtZ2dnZzdtN/YI2C2ert7oNV/v3rtj3O/r8Id1CbcRcGYd/9khlvfFGuYvg97gMpoHgaHi2WX+j6/zOGpE6/SCGuY3PcZ/drp17hHuYzXcX+n4er2Ob74Wuw64Vce45qb48s5xdfzD6vgrOm88B4jXKvpn7CXwmi7h7tMafyEdDwB1mh/W8a8eMk63pLyUu6xzWwGH1nme02OZ/68Vr5Ud43aq2/dcWjfuHdM8qob945Bx7nqcDzmPZpue3jH8rkw9fG44xPzWeugA7sXUi98P0PFisU7TfDV6t5muh53dYuswvR403qbXM0yvKSW6mi/yD+oyfiumXijlEPv2QXX4ajq+SK/jm5efl9EqgVCP+VV13P494rxjDXc1sF1r+IrW8fQTaiZWl/BDp0d1fgk8o8f4rVrn6h5DzLc5B1bN4vjfhKmXzXt2jPtTHf7IIea31j6lfKB0VB12MnDzLmE+UMf/30zXw87ObvAO7w8Gjbf3B3PwPN8xzeun20/Ad+o0L+4y7nCmPk7ZlPIxSJPh+soe82vW+aBZHBtdj70h5+Eztp2dnZ0dmWkbe5pX+wLbAWcAx3WM+0ztP7FLuH2Am1JeAL1kkAVFxA2A/6x/X52Zlw4d29l56yzCfqP2796uO30W6/RJyldnu0XEndojImKL1jw/PmgEIyIoN+sA78zMy7tM9lHgH0AAvdoZODwz/9plePN19ibArQaN14CuAt7RZfhxlAcPKC+IVneZ5oe1f7thFpiZp1Nefm0G7NYx+qG1/+Fuy8zMLwKn9Zj1Eyjb9wuZ+fce0xxO+RrxthGx3TDxnq3MPJbyULAi1m6H55La3wjYeibzjtJ24JGUrzXfnJnPyszsMmmzrHldd2nCmV4PxvR65un1Qylf9f+Tst5rycwLgf8bYn6N71I+zlkOPKDL+EfX/tcz87LW8L0omY1/zMzvdZtxZp5K+TJ9SZ2+m7dn5nU9xg2VHkXEjky9XPtYjzhdSHlxCeW8nTeZeSWlijYo8WybVdobEVtT7rn2omQW3Ccz/9ZlUtN4aX55fzAY7w/m9nkepo6zt/eZ5rO13y19fAal5NttKaVQPwHckFJrz5tGFMc54TO2JKlhxp7m00G1/7kuNwdN1Un3j4gbdYy7e+3/LjP/MeCy9qC8eEnKS575tAb4Xb8JImJJRDwlSuPaZ9XGkJtGky+qk20KbNkKNqN1yswLKHWiAzypY/Sj63JOycwfDzpPSvUWy+vvo3os9zrKV21Q6mDv5hc9wl5N+VIf1t4Go7Cq20NUje/59e8fe4Q9p1+cImLfiPhcRJxaG49uN4Z9xzrZv7Wm3wS4Tf37kz5x7jXunrX/xIg4u1tH+Tp2ozrdzbrPZnYi4hG1Me6/1Yaz2+u9RZ2s/dBxSu02Bk6IiBdExC71AXMQj6A8LG5GqWLr5X2m/Xbtf7I2QH73iNioz/SSTK+vZ3p9vVGn181yju2TEXbMEPNrx+fw+vcx7XE1zW0+pvksa2vS0516pac1TW2m65WentAnesOmR82ybgCc2SdOj5wmTrNS0+f3RcTvI+KSiLiudQ48r072bx3BmnV9c0S8PyL2joilAy7y3yj7/i6UjNS9MvOcHtM2mZoPiYivR8RDa6agpLlxUO17f+D9QWPen+cj4mbA9vXvt/ukj++u06yTPtZt+iTKvvhvyock/wIen5nXjjK+M+UztiRpOmbsaV5ExHLgIfVv58sU6le4x1JudB/TMXrb2u/2pW4vTZiLM/PiIcKNwgV9XlI1X+odQ/kCbn/gJpSvF8+jZBy1X15s3vo9m3X6aO0/JiI2bg1/cu0fNuT82g9r/R7OzuwyfVu/rxSb0nOjvkE8q8+4a6eZphm/Tpwi4j2Ur9seRXlQWkIpNdDs06vrpO19uiVT1+F+8fpnj+HNF3LLKMdHr65ZxmZ9ljG0+kD7FUqVMg+hPDQFJYO0We/mXLh+vevD0mMox84OlBKUfwbOj4gvRcSDp3kAeQvlgeXjmTndF5UvoXx5uQx4GeWl6yUR8aOI+K8hXjRKi4Lp9RTT67WMOr1ultMrfYP+8e2nOW4fGBHLWsPvT3kRdiHrvlBt0tNN6J+eblqn65WentcnXsOmR02clkwTp+bYG2kaDxARj6KUmHs2pQ2tzSlVlTXHf1PqcfOOoG+mvBzcGHgWpS2eSyLi+Ih4SZQSJr08jVKC4iLggMy8qNeEmXkM8GpKW1gPAr5MuZf4c0S8LSJ2GmJ1JfXh/cEU7w/WMo7n+XYpsRvTO31sMhS7po+1hP7nW4Nelpm9asqZNz5jS5IGZcae5ssjmXoZ8vv210atr47uW8d3q75jkkz3hdf/UL7CPp+yrttm5maZeePMvAmlmpLGoF9XTecHlMaltwYeDBARt6N8NXgtpeqJmdp0+kkWtoi4P+VLv2sp7TrcCtgkM7fOzJvU/fqzZvIRLrq5hr8gM2OA7ugRLhvKy7f/pNSx/1xKO3+bZuaNWuvdvLRda70z85eUNgIfR6le5jRKW0EPB74GfCtaVdd0aB7AHh8RD+oXwfo15r0pVbC8h9JI/caUhuo/APwxIrbvPQdp0TG9nmJ6PZl+THkRuSlTJfRgqhrOw2tJgrYmPf3agOnpId0W3O8r/xmkR02cfjdgnA6aZrsMpZa4+QjlhewXKMfgppm5ZSuNf2czece6XpmZDwHuQXlR+FNKqYjm/8kRcUe6+zblRfGWwAciou/zama+DtgZOBj4HqV6sF2AFwF/iognDLXiknrx/mCK9wfj1U4XthwgfVzRbSZRqrHcvzXo3nMZ6SH4jC1JGogZe5ovw9zc3ykibt/633zxdosh5tGEWV6/LhzUNbXf7+Z2mPl184ja/+/M/GRmntsxftvOANVM14laVUpT535TfUfzdd/3MrPfF/PdtL9Iv3mf6ZqbuX5fsC8EzT79aGYempmndqmeptt+vYipr+361U/fa1xzTPTbB3OpWe/XZeZ7M/PM9sj60LBNr8CZuSYzP5OZT8zMHSlfFr6RWo0P8MweQQ+mVK2yEfCliDigXySz+EFmPi8z71zj9AxKqY0dmHoxKcn0us30eu40y+mswrGt37ie6jZsXk49Gq4vXdG8pFqnpAnzlJ4OmR41cZqTKjYHcH9KNaB/Ah6Tmb/qkiHa6xwAIDN/mpkvy8x7UDLqHk0psXMjpkqfdPoFpX3EyyglDz42XTVimXl6Zr4pMw+gvMDcm5LBu4SSOXjjfuElDcT7gyneH4xXu0TkjNLtmq4cRkkz/kI5bh4dEY/sG3B++IwtSRqIGXuac7UanKadkN0oD/a9uqah6faDw09r/w4R0f76rZ9fUm7OgnLzMqjVtd/166J6A7j7EPPrppn3b3qM//cew2e6To3DKF/z7R8Rt6B8xQVDNLLdchpT22rvbhPUL6z3qn9/PYNlTJK++7Ru73UaDc/MKykvzKD/F4L36TG8acun7033DDUZjv1epk13LN+LIb4ArS/mXkEpGQCwZ59pnw/8H6XatK9GxD5DLOeizPww8IrpliMtJqbX6zC9njvNcu7dJ9NmNtfmJvNun5qp8xBgKaUkX7c2iJr0dJhjd9amSY+aOG0VEXcb8aKHSeN/n12qpKv77X6DLjAzL8vMzwNPr4N2j4jOKjybaX9CyYhdQ2nT64ODthGUmddmqaHgPyjVoG9OKdEiaYa8P1iH9wdzq28alZmnM5W5N5PtCPAcYD9KqbiHAK+vw/+vzzE6SNo5HZ+xJUkjY8ae5kNTBc7vMvN3mbm6Vwd8qU772FYVAT+k1BO+IfDWQRaYmf8Cvlr/HtrRxko/f6j9u0REt1JSj2X2X0439enfvnNE/aL8ld0CzWKdmvD/AL5D2Y6foXwtfR6lDZSh1C8Gv1L/Pi8iutVb/1RKNSTJ1H5dqHru0+p/6X3z3uzTp3X7cjMiHkb54q2bT1K2764R8Yx+EYyIYRstv6T2t+gzTb9jeQlTD0id4zbuNrxlTe1vMs10zwY+Rnmw+XpErJUBGhEb1HjMdjnSYmF6vTbT67nzFcrLrZsy9WLyejXN6vVF+bQy8zfASZQSW49gqr2nz3cpUQ/l2P07Axy7M0hPZ5QeZeZJTL0Mf0tE9GyjKCKWRsQwaVmTxvcrMdIc/7frkan2NGDHHvHpl8436xqUaru6ysyjgAOBKymZge8ecjlXMVWdnum8NDveH6zN+4O5Nchz6Mraf3G/zOIotugYtiulLViAl2bmX4A3AD+nZE4f1iPdGyRe0/EZW5I0MmbsaU7VG6LH179f6Tdt9Q3K17U3odZ3Xqv9eVEd/+iI+GJE7NJaxlYR8bSIeE/HvF5BaaNjZ+DHEbF3/eqseQHywIj4dkeY4yj1lW8MfC4iblmn36xmnHyEUn3ibHy/9t8REXs2N40RcRfKQ8/WfcLOZJ3ammqP7lX7n+5SrdKg/pdSTdK/Uepqv3WNxyYR8TRKXesAH8vMU2e4jEnR7NNnRMSTm5vqiLh5RHyCUvVUr+PmvXXctsB3IuK2NeySiHgU5cvM1d0CZuafmKri4gMR8cZo1WUfEcsiYr+I+DTDP4ydWPsH9Hgohqn1/p+IeEjz8F7Pz28Ad6UcI50eEBEn1PP2+ip56nn2NMoDN5S2cnqqD6RPBz5FaRT92xFxj9YkNwT+GhGvjIjbt+K3Qf368A2DLEdaDEyvuzK9niOZeQZTJQw+GBFPaDKuolTf9l1m3+ZPU2rvGZQ2YNrDOuNzNeXr/aQcu0dExG7N+IjYKCL2iIi3UNo4GtZM06PnUjK27gv8MCLu3TqONqzzejWl5EW/Kr07nUI5f5dH+YComx9QtsftgPc0L0Yj4oYR8RLg/cAFPcL+MSL+NyLu0ronioi4K+W+B+AXmdn3HM3MI4GHUTLp/jsi3tYxyScj4rCI2D9aL8gjYgWlvalNKS8Yj+23HEm9eX/QlfcHc6t5Dn10RPS6F3gTJe3bBjg+Iv5fRCxtRtbn8KdTShoe2Bq+EeXZcSmlCtP3A2TmNZTj/HLKPcN/94nXQ2PIqlS7zMNnbEnS7GWmnd2cdZRqHbJ2tx0wzHfr9F/oGP5Cype3zfwupdyUN/+P7rH89jRXUBq5vqYZ1iXMf3Ys52LKw0lSvlxaWX8f0hFurzp81TTrtwPly7pm/muAf9Xfl1OqhGjGrRjFOrXCLqE86DRhbzfL/dtUk9TM7yLKy5fm/w+AzbuEO7qOP6jPvFfVafYaMk5dw1Gqcup6nAy6zF7zoDw4ntBa72s69tH/9FtnykPvFa3pV7f+/4SpOvE/1CXshpQGqrPVXVzncV1r2FFDbsdtKC/sknI+nFW3z6rWNFsBf20t46q67GYbHNRtm1IertrxvZxSF387vt8ClnTEqet5UbfB51rbbo86fIuO5VxV1+ma1rBTge1ncx7Y2S2EDtPrbutnej236fUNKCXS2ttndf39L8pHMb32fdd92zHNjh1pwJ8HiNOTKBlp7fSpM93IjjArBtifW3TEZeD0iFLN2GpY5zi6qmOetxhy+3+iFXZ13Y+rgIe3pnlHxzIuYuqc+y6l1EACKzvm3Y7vNXVd2/E9D7jDoPuUct/QnNtvaA0/ojXP62r8LutY9uNnc+7Y2S32Du8Puq2f9wdze39wv9byr6SUqF9FKXXfnu5WlGYt2tf88+s+aKddT2yFeUMddgHwb12W/azWfty1Y9wuTN0jXE0phboK+MkQ6+Yztp2dnZ3dyDpL7GmuPbH2T87ME/tOOeXLtf/gaFWbkJnvAO5EKcG0itKobwK/p1TP84LOGWWpxufWlKoW/ki52diUcqPxOeDBXcJ8lXIzfhTlYWND4LfAUzLzKQOuQ0+ZeRrlK6tPA+fW+a+mVKdxlyxfJ/cLP/Q6tcJew1S7B7/IzD/Ocl2+Qaki4iOUfbIZ5QbyJ5SvvPbPzG5fky0omXkVpS2F5svB6yj75fvAgzLzddOE/x6l/ZfDKTfFm1BKBLwG2IfyRSF0KbmXpS2ZZ1Ha6Ps0cEYNvynwN0rVLM8BHj7kOp1Peej8CuXB9UaUBu9v0ZrmQuDulHr4m0a911BetO2ZmSt7zP5HlC8iP0GpLudyYFld9+9Tqvt5UD1eB4nrtXV+X6FUK3ZkLWlxCaWNnXdRqlY5ry7nMuAXlGpydsuOBsmlRcr0et35m17PoSxVku0FvBo4uQ6+gtIGzF2ZamNupvM/lXLtb3QtrdcR5jDKPnsX5av6aylfpl9AeYn5mjp+WDNOjzLzO5SSHa+nlDy4kvJS7RLgeMq9x+5ZSkEO45mUD4dOotw3NGn8DVrLfiHl+PhNXe6G9ffzgQdSjuluHlLn3ZScuQHlxd/va3xvm5m/HzSimXkEJaP3WuAVEfGaOurlwEspmQinUT602pByjh0G3DkzPzXociR15f3BuvP3/mAOZeaPKJmzx1CeLW9KSZ9u0jHdXynH07Mo+/oiyrPgNZRj6sOUtOrTABFxT+BlNfgzM/OfXZb9AUpJs6XAp6JVDXaWKrL3paQ5F9f43IIe7Tn2WDefsSVJIxOZOe44SJpHEXEysBPwX5n5wXHHR9OLiGMpGXdP6nMjL0laQEyvJUlSJ+8PJEkSmLEnLSq1zvMfUOvSz8xLpgmiMav12R9PKQW4IjP/PuYoSZLmmOm1JEnq5P2BJElqLBl3BCTNj4jYBnhr/ftxHwLWH7Vh720o1Y+tysxrI+IGwEOBd9bJvmimniQtfKbXkiSpk/cHkiSpzRJ70gIXEW8D/h+lDviNKA1K3zYzzx1rxHS9iHg9pS56KO3HXExpP6dpB/W3wL61Tn5J0gJkei1Jkjp5fyBJkrrZYPpJJE24bYCbURpcPhK4nw8B653PA+8AfkV5UFtGaZT6eOCFwD3N1JOkBc/0WpIkdfL+QJIkrcMSe5IkSZIkSZIkSdIEsMSeJEmSJEmSJEmSNAHM2JMkSZIkSZIkSZImgBl7kiRJkiRJkiRJ0gQwY0+SJEmSJEmSJEmaAEvGHQH1FhGnAzcEVo05KpKk9dsK4JLMvOW4I6K1mZZLkga0AtPy9ZJpuSRpQCswLZc0T8zYW7/dcOnSpVvtuuuuW407IpKk9def//xn1qxZM+5oqDvTcknStEzL12um5ZKkaZmWS5pPE5+xFxEPB/YEdgPuCCwDPpOZj5vBvLYHXgscAGwNnAUcARyamRf1CHMb4BBgL8pXfGcAnwfelJmzvZqv2nXXXbf61a9+NcvZSJIWst13351f//rXq8Ydj5kyLZckLXam5WvNy7RckjRxJj0tlzRZJj5jD3gV5cHhX8CZwC4zmUlE7AgcD9wY+BpwEnBX4HnAARFxr8y8oCPM3YAfARsBhwN/B+4HvBrYJyL2ycwrZxIfSZIWEdNySZImm2m5JEmSNE8WQsbeCygPDn+lfCF41Azn8wHKw8NzM/O9zcCIeEddxhuAZ7aGbwgcBmwGPCQzv16HbwB8EXhYDfemGcZHktRDZnLCaRdw0lmXcvGaq8cdnbG5+98+DMBl55855pjMmmm5JM0R08z1m2n5OkzLNbxMWHUsnHMirFk97thIWmROOK18b7IA0nJJE2SDcUdgtjLzqMw8JTNzpvOoXwXuR2kM+/0do18DXAY8PiI2bw3fE9gV+HHz8FDjcx3w0vr3mRERM42XJKm7E067gJ+ddqEvKBcI03JJmjummZoPpuUaq1XHwqrjzNSTJEmLxkIosTcKe9f+kfUB4HqZeWlEHEd5wLg78MM66n61/93OmWXmaRFxMrAzsANwar+FR0SvyvpnVH2JJC10J5116bijoPWPabkkdWGaqQliWq6ZOefEccdAkiRpXk18ib0RuXXtn9xj/Cm1v/Msw0iSRsBSB+rCtFySujDN1AQxLdfMWFJPkiQtMpbYK5bX/sU9xjfDt5hlmK4yc/duw+sXg3eeLrwkSTItlyRpwpmWS5IkSQOwxJ4kSZIkSZIkSZI0ASyxVzRf8S3vMb4ZvnqWYSRJc+gF+y6yWpaO2hqAzTcxOce0XJKGsujSzPWVaXmbablGa++Dxx0DSYvAT68ptUFftvF3gXPHGxlJi4Yl9oq/1H6vp9udar9db/9MwkiSpLlhWi5J0mQzLZckSZIGYMZecVTt7xcRa22TiFgG3Au4HPhpa9SPav+AzplFxA6UB4szgNNGHltJktTJtFySpMlmWi5JkiQNYFFl7EXERhGxS0Ts2B6emacCRwIrgGd3BDsU2Bz4VGZe1hp+DPBn4L4R8eDWMjYA3lz/fjAzc7RrIUnS4mVaLknSZDMtlyRJkmZn4ivyj4gDgQPr35vU/j0iYmX9fX5mvrj+vinlpv8MysNC27OA44H3RMQ+dbq7AXtTqu14ZXvizLw2Ip5E+ULw8Ig4HPgbsA+wB3Ac8M5Zr6AkSQucabkkSZPNtFySJEmaPxOfsQfsBjyxY9gOtYPysPBippGZp0bEHsBrKdV4PAA4C3g3cGhmXtQlzM8i4i6Urwf3A5bV5b0WeFNmXjmTFZIkaZHZDdNySZIm2W6YlkuSJEnzYuIz9jLzEOCQAaddBUSf8X8HnjTk8v8EPGKYMNKsZMKqY+GcE2HN6nHHRotQkpx50RrOv/RKrrjmurHE4e69Rhy19XxGQyNiWi5J0mQzLZckSZLmz8Rn7EmLzqpjYdVx446FFrEzL1rDmRetGXc0JEmSJEmSJGnR2WDcEZA0pHNOHHcMtMidf6m1GUmSJEmSJEnSOJixJ00aq9/UmI2r+s3pbLrEJE2SJEmSJEnSwuZbUEnSgrDNsk3GHQVJkiRJkiRJmlO2sSdJmmibLtmAbZZtwvZbLh13VCRJkiRJkiRpTpmxJy0Uex887hhokfjpNSd3Hf6CfXee55joessOB/457lhIWogyYdWxpY3fRVIdeJKcedEazr/0yvW2+ulh3L3XiKO2ns9oSFqIOtKI9e362eu5RZIkadKZsSdJkiSpu1XHwqrjxh2LeXXmRWs486I1446GJK3/OtIIr5+SJEnzwzb2JEmSJHV3zonjjsG8O//SK8cdBUmaDB1phNdPSZKk+WHGniRJkqTuFkn1m23rQ/Vxc23TJT4GShqBjjRifbp+Xrlk2bijIEmSNGd8opMkSZKkRWSbZZuMOwqSNKfO22yncUdBkiRpztjGniRJkiQtApsu2YBtlm3C9lsuHXdUJGlOXLlkGedtthNnLt993FGRJEmaM2bsSZIkSRrO3gePOwZz5qfXnNx1+Av23XmeY6JFZdnhwD/HHQtpJH5686ev9d/rp6TF4NM33IQzxx0JSYuGGXsaXiasOrY0lN1Rp36SnHnRGs6/9Mr1qn79xaDXSyhJkiRJkiRJkrQwmLGn4a06FlYd13XUmRet4cyL1sxzhCRJkiRJkiRJkha+DcYdAU2gc07sOer8S6+cx4hIkiRJkiRJkiQtHmbsaXgd1W+2Wf3meFy5ZNm4o6BFbvnSjcYdBUmSJEmSJEla8MzYkxaA8zbbadxR0CK3y3ZmLkuSJEmSJEnSXLONPWmCXblkGedtthNnLt993FHRIrV86Ubsst0y7rHD1uOOiiRJkiRJkiQteGbsaXT2PpifXnNy11Ev2HfneY6MJEmSJEmSJEnSwmJVnJIkSZIkSZIkSdIEMGNPkiRJkiRJkiRJmgBm7EmSJEmSJEmSJEkTwDb2JEmSxiETVh0L55wIa1aPOzbSyGQmJ5x2ASeddSkXr7l63NGRpPXGRF0fM9n+4l9xo8tPYZNrLh13bCRJktRixp4kSdI4rDoWVh037lhII3fCaRfws9MuHHc0JGm9M0nXx+0v/hXbX/LrcUdDkiRJXZixt9hlkqf/mFV/+jmrLziPK665bsaz+uk1J48wYpIkLXDnnDjuGEhz4qSzLNkhSd1M0vXxRpefMu4oSJIkqQcz9ha7Vcdy5m9/wNkXrRl3TCRJWlysflOTaukWfUev99XLzcDypRuNOwqSFoBJuj7OtPrNK5csW+u/109JkqTR22DcEdCYnXMi51965bhjIUmSpEmx7W3HHYN5t8t2y6afSJLEeZvttNZ/r5+SJEmjZ4m9xW7N6llVv9no/CqvzS/0JEmSFoClW5RMvRX3GXdM5s3ypRuxy3bLuMcOW487KpK0XrtyyTLO22wnzly+O+D1U5IkaS6ZsaeR6Pwqr80v9CRJGsLeB487BtKceMG+O487CpK0Xlovr49H9ciQ8z5FkrSARcQqgMxcMd6YSP2ZsadZ6fwqr80v9CRJkiRJkiRJnSJiQ+DJwOOA2wPLgIuAs4GfA1/PzK+PL4bS+mukGXsRsS+wL3Bf4ObANsAa4Fzgt8CPKCfkP0a5XI3eT2/+9PXzq0FJ0pwyLZckafKZnkuSpPVZzdT7JnAAsBr4FnAmsDFwW+AxwC6AGXtSF7PO2IuIzYDnAs+gPDBEHXUF5aFhKbADsCPwMODdEfEN4G2ZecJsly9JkmbHtFySpMlnei5JkibIoymZer8D9szMi9sj633N3cYRMWkSzCpjLyKeDLwO2A44CTgUOA74RWZe0pougFsDdwf2Bx4CHBgRhwMvycy/zSYekiRpZkzLJUmafKbnkiRNthUv/9Yh447DTKx60wMPmWHQe9b+ys5MPYDMvBw4qvkfEQcBhwFPAs4AXgPsDiRwLPDizPxzex4RcTQl0zDo0J5fZq7sMn458AbgP4GtgdOADwLvzczsMq8HAXei3ItdDfwB+L/M/HS/jSDN1AazDP9R4GfA3TLzNpn52sz8YfvBASCLkzJzZWY+GrgJ8Hzg3sBBs4wDEbF9RHw8Iv4ZEVdGxKqIeFdEbDlg+L0iIgfobtYRrt+0P53tekmSNA/Wi7RckiTNium5JEmaJBfU/rBtQf0HcCRwCSWj7VjgAcAxEbHNiOK2MfADykdQnwc+AmwBvBt4X5fp/w+4BfBj4F01zC2AT0XE60YUJ2kts62Kc4/M/PWwgerDxXsj4iPAitlEICJ2BI4Hbgx8jfJ14l2B5wEHRMS9MvOCPrMAWEX5orGb2wMPBf6YmX/vMv4MYGWX4WdOG3lJksZv7Gm5JEmaNdNzSZI0Sb4CvAx4ZkQsA74K/Cozz5gm3IHA/pn5w2ZARLwReDnwZOAtI4jbdpQSerfLzCvrMl4D/AJ4VkR8ITN/3Jr+dpl5ansGEbEx8B3g5RHxQds11qjNKmOv88EhIjbOzKuGCH8FJSNuNj5AydR7bma+txWXdwAvoBSZfeY08VgFHNJtXER8rv78SI/gqzKza1hJktZ360laLkmSZsH0XJIkTZLM/E1EPI5SCu5xtSMiLqSUfPt4Zn6jS9DPtzP1qg9TMvbuOsIoHtxk6tX4XlhL3zXVgf64Ne7UzsCZeVVEvB+4H7AP8MkRxk2adYm9ThdHxAsy84Mjnm9XtbTefpQSd+/vGP0a4OnA4yPiRZl52Qzmvw2lHt01TOrJlwmrjoVzToQ1q9cdTa4bRpK0mM1rWi5JkuaE6bkkSVqvZeYXI+KrwN6UasHvVPsHUtoA/iRwUEebdr/sMqumlr2BmuUawDWUGgI7HV37d2oPjIibU0of7gPcHFjaEe6mI4qXdL3ZtrHXaRNK6bn5snftH5mZ17VHZOallMbCN6M0DD4TT6Ss05cyc3WPabaIiCdHxCsi4tkRMdNlzY1Vx8Kq47pm6gGcedGa+Y2PJGl9N99pOWB7uZIkjdi8p+em5ZIkaViZeXVmHpmZr87MBwHbAI8ELgOeADykI8jqLvO4pv7ccETROj8zr+0y/OzaX94MiIgdgF9Tagw8m9Lu8espzX59ok62yYjiJV1v1CX25tuta//kHuNPoZTo2xnoLKI7iKfV/of6THNH4GPtARHxO+DxmfmHQRYSEb/qMWqXQcL3dc6JfUeff+mVfcdLkjTXbC9XkqTJZlouSdLsrHrTAw8ZdxzWBzVD7YsRcXvgVZSqLI+Y4eyuA4iIJa3Mv8YWfcJtExEbdsncu0ntX9wa9kJga+BJmbmyPXFEPJpScEgauXnP2IuIRwG7Z+ZLRjC7Jnf84h7jm+FbDDvjiNiTknH4x8zsVvQW4B3AlykZi1dQMuJeBjwc+FFE7Db2hjF7lNRrXHHNdesMu3LJMpYv3WiOIiRJmnQjTsthobeXe+nZcNQb52z2i0FmcsJpF3DSWZdy8Zqrxx0dSVoQRpyeL+y0vIdZpU+ZbH/xr7jR5afAURvPTQQlSZpcl9Z+zGIeF9X+zYDTO8bt0SfcEuCewLEdw/eq/d+0ht2q9r/cZT57Th9FaWZGXRUnwH9ExAsj4n4RsXWX8Xeh3Niv755e+x/uNUFmvigzj8/M8zPzX5n5y8x8BOVE3gZ48SALyszdu3WMqfHy8zbbiV22WzaORUuS1g/zlpYP0F7uZZT2cjef4fwnv71cccJpF/Cz0y40U0+ShjMv6fliTstnkz5tf/Gv2P6SX7PJNZdOP7EkSQtMRDw6IvaNiHXyJyLiJkzVpPfjWSzm57X/tPbAiNgHePQ0Yd8YEddXoRkRW1FKEAIc1ppuVe3v1bGM/YGnDhddaXBzUWJvj9olQET8E/gd8EfgEkrduOeMaFlNibzlPcY3w1cPM9N6oj6M8uDwqRnE64M1/H1nEHZsrlyyjPM224kzl+/Ow3fo9twnSVok5jMt79tebkQcR3lZeHdmVq12017uJ6drL5dSrcbFwK8yc6g2eea0Wm1x0lm+9JSkGZiv9HzRpuWzSZ9udPkpMw4rSdICcDdKld1nR8RPmCpRd0vggcBSSvXeh89iGYcBLwEOjog7An+iNNl1f+CrlPf33ZxFuff4Y0R8HdiIUkPfdsAHMrOd2fgB4EnAlyLicOCfwO2AA4AvUtoLlEZuLjL2VgLHAbtR2p+7A/CA2jVeOaJl/aX2d+4xfqfa79UGXy/Ng8Mn+jw49HNe7c/oi8R5sffBAPz0mu6bJmI2pZwlSRNuJfOXli+I9nJHZukW87q4SWFJvclnNe/SWKxkftLzRZuWzyZ9akrqbbpkLipSmiPep0iSRuftlHuEf6fco+wPbApcABwNfBb4bGbmTBeQmefW5rbeSimAsyfwS2BfSgZir4y9q2q8/hd4FKVmvtOANwHvbU+Ymb+PiL2B11MyJJdQPqR6KKWwkRl7mhNzkbG3KjM7b6hvSfnK7YbAyZn5m64hh3dU7e8XERu0vw6MiGXAvYDLgaG+1GPqwaFnNZzTuHvtnzbD8JIkjdN8puULor3cWoV2tzj8CrjzwJHe9rYDTypNEqt5l8ZivtJz0/JZ2GbZJtNPtL7wPkWSNCKZ+XdKFd6d1Xj3mn4l5aOlXuO7llLJzBNZ+6OmxjHd5peZK1p/n1276eJ2PHC/HqMtPaM5MRcZe+vIzNNZt4HKUcz31Ig4kvL137NZO8f8UEqJuQ9l5mXNwIjYpYbt2n5dRNwH2JX+Dw5ExB2AP2fm1V2Gv6H+/fTQKyVJ0nportLyOTZQe7kdg34JPKJWofEwSnu5c9828NItysuyFfeZ80VJ82n50o3YZbtl3MNq3qX1wgSm55OTlg9p0yUbsM2yTdh+y6Xjjsr0vE+RJElar4w6Y2/GRWNn4VnA8cB7asOXf6bU0bs35Yu9zqpF/lz7vXLLp31wqF4IPCgijgX+DlxJ+TLwAGBD4CPA5wZfDUmS1gvznZYv/PZyl93k+iqwNVov2LdXbeyStOjNZ3q+8NPyIQ2UPh21Zffh3jNIkiRpGqPO2LsxcIMRz7OvWmpvD+C1lEy1B1AauHw3cGhmXjTovCJiS0p1HYM8OBxBqb7kDpSitk0dwN8BPpKZXx9uTSRJWi/Md1pue7mSJI3efKbnpuWSJEnSPBppxl5mXkDJ3JpXtU7eJw04bc96bWsm4ED1YGTmEZTMPUmSFowxpOW2lytJ0ojNc3puWi5JkiTNo3lpY0/zIBNWHQvnnAhrVveflOSEU8/npLMunZ+4SZLUg+3lSpI02UzLJUmSpPk1q4y9iFiamWvGPQ9RMvVWHTfQpGdetIafXXPhHEdIkjQJ1pO03PZyJUmahfUgPTctlyRJkubJBrMMf3pEPC8iNhk2YETcMSK+Brx4lnEQlJJ6Azr/0ivnMCKSpAkz9rQ8M08F9gBWUl4CvgjYkdJe7t1rdWKDxmnY9nKPAW5HacfnucDulPZyH5KZT8/MHGZdJEkak7Gm56blkiRJ0vyZbVWc3wPeAbwmIr4AfBH4aa+v/CJiB2B/4AnAXSlf1L11lnEQTFv9ZtvF07QdvnzpRrOMjCRpgqwXabnt5UqSNCtjT89NyyVJkqT5MauMvcx8YkS8j1J3/dNrd21E/Bk4C7gI2BTYGrg1sA2lqo1zKVVxvDMzLT42z87bbKe+43fZbtk8xUSSNG6m5ZIkTT7Tc0mSJGnxmG2JPTLzF8B+EbET8BRgH2A34PYdk54HfAX4MvDlzsatNQ+WbgHb3pYzr9626+jlSzdil+2WcY8dtp7feEmSxsq0XJKkyWd6LkmSJC0Os87Ya2TmKcDLASJiM+CmlK8B1wDnZuZZo1qWhrD3wesO++vJXSd98r1vOceRkSStz0zLJUmafKbnkiRJ0sI2soy9tsy8HDildpIkacKYlkuSNPlMzyVJkqSFZ4NxR0CSJEmSJEmSJEnS9OakxF6niPgS8P3M/HD9f2vgDsDRmXnefMRBkiTNnGm5BpGZnHDaBZx01qVcvMYmmyRpfWN6PhpNejeDgLDqWDjnxNFHSpIkjURErALIzBVjjMPGwInAKZn5gHHFo1NErABOBz6RmQcNEe5oYM/MjLmJWc/l7gUcBRyamYfM4XJuCvwVeHtmvmqultM2XyX27gv8FiAitgZ+BnwUODEiOhvyliRJ6x/Tck3rhNMu4GenXWimniStv0zPR6BJ74a26lhYdRysWT3yOEmSNGkiYsOIeFpEHBMRF0bE1RFxbkT8PiI+GhEPHnccx+i5wK2Aeckkmq2IWBkRWTP+Fp3M/AfwQeCFEXGz+VjmfGXsLQOaBrofRsnV3Qr4CPCGeYqDJEmaOdNyTeuksy4ddxQkSf2Zno/AjNM7S+pJkgSUTD3gm8CHKbUHfBt4O/Bpyr3KY4CXji2CYxQRmwOvpNSy8Otxx2dEngDsOu5IzLG3AhsD/zMfC5uXqjiBvwE7An8HHg58KjOvjYiVwHHzFIfJdOnZcNQbRzKrGVcXIkmSabkGMJ8l9ZYv3WjeliVJC4jp+Qj0S+/6pk/TldRbusWM4iNJ0gR6NHAA8DtKFY0Xt0dGxGbA3cYRsfXAY4AtgJXjjcboZObfxh2HuZaZ/4yI7wOPiYiXdB7TozZfGXsfB94XEd8B9gae2Vr+ZvMUh0VvxtWFSJJkWq71zC7bLRt3FCRpEpmez7FZpU/b3nZ0EZEkTZZDlh8y7ijMyCEXHzLDkPes/ZXdMkAy83JK22jXi4iDgMOAJwFnAK8BdgcSOBZ4cWb+uSPM0fRo2609v8xc2WX8ckqNBv8JbA2cRqlu8b2ZmV3m9SDgTsB2wNXAH4D/y8xP99oIPTwFuAo4okucDqGs9951OS+mlIRbDXweODgzr4yI+wGvBu4MXEspHfn8zLygY34JHJOZe3VZ1krgicAtM3NVr8jWeTROj7h+U5/RtFPYbT+0278Dvge8DrgLpZbJ44FXZuYvuyxvOfBy4KHALYA1wM+Bt2bmD3rFs2Meu1NKEe4F3IxyH/x34OvA6zPzoo7pD2Lq2Du7Lv9OwA07jq3PUzKsHwV8aJC4zNS8ZOxl5lvqDt2fcoKdVkfdlXISah5YPZYkaaZMy7W+WL50I3bZbhn32GHrcUdFkiaO6fncutsOW80sfVq6RcnUW3GfkcdJkqT1VJPBtPMMwv4H8BDgO5SMttsADwDuEhG3yczzRxC/jYEfUErOfb7+fxjwbuDWwLM7pv8/4ETgx5SqRLeucfpURNw6MweqnrFmWu0B/KJmbvby38D9KZl/RwP7AS8AtoqIr9U4f4tS1ek9gccB29Qwo3YocCBwR8r2WV2Hr+4++TruBhxM2d7vp7Qt+FDgvhGxX2Ye20wYEVtQapm4DfAL4F2U9fp/wJER8V+ZOUiG2tMoGbbH1OVuQMkkfiFw/4i4W2Z2y0x5OCXjrjn2btExvqkBY18mNWOvNhJ4bWb+E8oDBPCWjsm2pRxkmgsd1XjMuLoQSdKiZFquUXnBvjN5VpMkjYLp+fy5547bzCzg3f9rtBGRJGn99xXgZcAzI2IZ8FXgV5k5yIdGBwL7Z+YPmwER8UZKKaons+59zkxsRymhd7vMvLIu4zWUzKRnRcQXMvPHrelvl5mntmcQERtTMoBeHhEfzMx/DLDcewAbAuuUVOvw78DuTQnFiNgE+DXweErJwf0y85g6bgNKibgDImK3zPztAPEYWGYeEhErKBl77+pXuq+HA4D/zsz3NQMi4iGUTMuP14zR6+qoN1My9T4MPLMpORkRb6Zss/dExPcGiMMbgWdn5rXtgRHxFOCjwLPqsjo9AHhAZn6320wz868RsRq47zTLn7UNRj3DiHhlRFwArAL+HhGXRMQREfGAzmkz8y2Z+bpRx0HVENV4WJ2VJKlhWi5J0uQzPZckSeurzPwNpRTZObX/ZWBVRFwQEV+NiAf1Cf75dqZe9eHav+sIo3lwk6lX43whpbpIKFUy0hq3VqZeHXYVpQTaEmCfAZd589o/a5rp3tOudrTG8wuU/J5vNZl6ddx1QFMd6B0HjMd8+ivwgfaAzPwapTTdrYD7wPUZpY8D/kXZN9ma/hTgPZSSlU+YboGZeUZnpl71ceASSu0W3XytV6Zey9nAjSJi0+niMRsjzdiLiFdSDu4tgb9Qcon/BTwY+EZEfLMWJ9VcWroFrLjXwNV4zLi6EEnSgmNaLknS5DM9lyRJ67vM/CIlI2t/yn3LNyn5FQcCX4+IT0SrwbaWbqXZ/l77W44oetdQ2nnrdHTt36k9MCJuHhHvj4iTIuLyiMja9tyX6yQ3HXC5zUv6i/pO1X0b/LP2f9VlXFNacPsB4zGfjm2VyGs7uvabbX1rSlt4v6uZrJ1+1DF9TxGxUUQ8JyJ+EhEXRsS1dX9dB9yQ3vvr59PNG2jiNsOqHAYz6qo4n0Y56PZpF+mMiLsAL6LUdfq9iLhvzbHWdJbdBPY+eE4XMePqQiRJC5FpuSRJk8/0XJIkrfcy82rgyNoRERtS2rL7OKXk1VcpVTK2re4yn2tqHuCGI4ra+T1KdJ1d+9d/IBURO1AyfLYEjqWsy8XAtcAK4InAJgMud03tT1fa6+Iuw64ZYNz62B7XOT2Gd27rpt+rNGMzfIsBlvkFSht7pwFfq8tqSmc+n9776+wew9uW1v6avlPN0qgz9m4KfLizntbM/AXwqIg4FngvpRHCN4142ZIkafZMyyVJmnym55IkTZJDLj5k3FFYH9TMtC9GxO2BVwH3Y92MvWFcBxARSzLzmo5xW/QJt01EbNglc+8mtd/OPHshpaTdkzJzZXviiHg0JWNvUOfW/nxVr5f0ziPaYp7isG2P4Z3b+uKO4Z2265iuq4jYg5Kp9wPg/u3jorZH+NI+wbPPuMbWlIzUbqUKR2bUbexdDlzaa2Rmvh84jgHqOZUkSWNhWi5J0uQzPZckSZOsuY/pVhXnMJoqLW/WZdwefcItAe7ZZfhetf+b1rBb1f6XWdee/SLXxe9rf5chw83URXTZNrXk5G5DzKfJAJ1Jicl71wy1TnvVfrOt/0K5x71jRGzRZfq9a//X0yyv2V9f75LZe1emStwNLSJuQPnA7vftNgDnwqgz9n4D7DfNNMcCtxzxciVJ0miYlkuSNPlMzyVJ0norIh4dEft2y9CJiJtQqhUH+PEsF9W0ifa09sCI2Ad49DRh3xgR11fJGBFbUUoRAhzWmm5V7e/VsYz9gacOF11OBM4D7j5kuJn6OXDziOi8b3wVcIsh5nNB7d98BnHYCXhWe0BEPISSKfpXyj0rtfr4zwDLKG0ytqffEXgucDXwqWmWt6r29+qYx42B988g/m13oWRuHjXL+Uxr1FVxvgo4JiL+NzNf0WOaGwHnj3i5kiRpNEzLJUmafKbnkiRpfXY34HnA2RHxE+D0OvyWwAMppaa+Bhw+y+UcBrwEODgi7gj8CdgZuD+l/b6H9Qh3FqWdtT9GxNcpbdM9nFLd4wcys53h+AHgScCXIuJw4J/A7YADgC8Cjxw0spmZEfFV4OkRcdvMPHHgNZ2ZtwH7A1+LiC9Qqo+8J2U/HE1H5lcfP6Rs549ExJcpJS5XZ+b7Bgj7XeDtEXF/4HeUEnUPBa4AnpyZ17WmfTlwH+A5te3oo4BtKO1HLwOek5mn098vKDVXPDQijgd+QqkO9P6UUoH/HCDOvTQZpN1Kb47USEvsZeZPKPXzvzwijo2I/4yI6xt6rDvnMcAnR7lcSZI0GqblkiRNPtNzSZK0nns78Bzgp8AdgGcCzwfuTclQejzw0NlWZ5iZ51JKfn0HuC/wX8ByYF/gm32CXgX8O3Ak8CjgGZS2255X491exu8p1UAeT8mU/C/ghpTMqQ/OINofqP05rzI9M38IHEgpKfgoSnuAqyhVUp4xxHy+B7yIUmLu+ZQSdS8eMPjPKBmIm1C27f2BHwH3zcxjO5ZzIXAP4C2UtuxeCDyCUvLwgMz8ANOo7SY+GPg/4N8oJf3uDXyUksl59YDxXkstffo44HeZecJM5jGMUZfYIzNfGRHnA6+l5KhfV/9vTGlw8Rt1nGYhMznhtAs46axLuXjNjI41SZK6Mi2XJGnymZ5LkqT1VWb+nVLt4cBVH2bmSmBln/Fd2+Orpd4e0GXUMd3ml5krWn+fXbvp4nY8cL8eo4dqJzAzfxcRRwJPiIhDMnNNa9whwCE9wq2kx/bJzKN7xSMzvw58vcuog2rXnnZVn/m8A3hHj3F7dRveGn8CJSN1Wpm5GnhZ7aab9mi6xLdmED5rnQDFii7Tr6TPsVc9ENgeOHi6eI3CqNvYAyAz30kp0vpS4ATKQ8OWlI34IGB1RPwiIj4UEc+MiLvORTwWshNOu4CfnXahmXqSpDlhWi5J0uQzPZckSZpIL6ZUm94r80nrkYgI4FDgl5R2AOfcyEvsNTLzLEqR2rdHxIbAbYHdW90dah8gKY0KakAnnXXpuKMgSVrgTMsF1hIgSZPO9Hz2zr3kSt75/ZNnPoNMWHXs9NNJkiQBmfmHiHgypd04rf9uQin1eMRsq48d1Jxl7LXVekt/X7vD4Po6R28D7AHceT7isZCM6sXa8qUbjWQ+kqSFzbR88WpqCZAkTT7T8zFZdSysOm7csZAkSRMkM20LeULUD+kOmc9lzkvGXjeZeR3wx9qtHFc8FrtdtjPTX5I0M6bli4O1BEjSwmZ6Pg/OOXHcMZAkSVqv9Gv3T9MbW8aexmv50o3YZbtl3GOHrccdFUmStB6bbS0B1g4gSVoM+qZ3a1b3Hrd0i1FHRZIkSQucGXsLzAv23XncUZAkSbqetQNIkhaDGad32952tBGRJEnSgmfGniRJkkbO2gEkSYvBrNK7FfeCFfcZfaQkSZK0oJmxJ0mSpKFZS4AkabG48Q03mZt075b3Hf08JUmStOBtMO4ISJIkSZIkSZIkSZregsjYi4jtI+LjEfHPiLgyIlZFxLsiYssh5nF0RGSfbtMe4W4TEV+MiHMj4oqI+EtEHBoRS0e3hpIkSZIkSZIkSVrsJr4qzojYETgeuDHwNeAk4K7A84ADIuJemXnBELM8tMfwa7os+27Aj4CNgMOBvwP3A14N7BMR+2TmlUMsW5IkSZIkSZIkSepq4jP2gA9QMvWem5nvbQZGxDuAFwBvAJ456Mwy85BBpouIDYHDgM2Ah2Tm1+vwDYAvAg+ry3/ToMuWJEmSJEmSJEmSepnojL1aWm8/YBXw/o7RrwGeDjw+Il6UmZeNePF7ArsCP24y9QAy87qIeCklY++ZEfHmzMyZLuSyK6/h4z85nYvXXD37GEuSJEmSJEmSJGliTXobe3vX/pGZeV17RGZeChxHKVF390FnGBGPjIiXR8QLI+L+EbFJj0nvV/vf7RyRmacBJwO3AHYYdNndXHbltWbqSZIkSZIkSZIkabJL7AG3rv2Te4w/hVKib2fghwPO8/Md/8+NiGdn5uEzWPbOtTu13wIj4lc9Ru3SL5wkSZIkSZIkSZIWj0kvsbe89i/uMb4ZvsUA8/oa8CBge2ApJVPtjTXsFyLigDlc9kgsX7rRfC1KkiRJkiRJkiRJ82zSM/ZGJjPfmZnfzMx/ZOYVmfmXzHwF8CLKdnrjHC57924dcNIw89llu2VzFENJkuZWRGwfER+PiH9GxJURsSoi3hURWw4xj6MjIvt0m/YId5uI+GJEnBsRV0TEXyLi0IhYOro1lCRpYTMtlyRJkubHpFfF2ZSKW95jfDN89SyW8VHgncBuEbGstt03X8seyPKlG7HLdsu4xw5bz/WiJEkauYjYETgeuDGlBP1JwF2B5wEHRMS9MvOCIWZ5aI/h13RZ9t2AHwEbAYcDf6e0o/tqYJ+I2Cczrxxi2ZIkLTqm5ZIkSdL8mfSMvb/U/s49xu9U+73awZtWZl4REZcCWwKbA03G3pwvu5cX7NtrkZIkTaQPUF4EPjcz39sMjIh3AC8A3gA8c9CZZeYhg0wXERsChwGbAQ/JzK/X4RsAXwQeVpf/pkGXLUnSImVaLkmSJM2TSc/YO6r294uIDTLzumZERCwD7gVcDvx0pguIiFtTMvUuBc5vjfoR8ErgADqq6YyIHSgZfmcAp8102ZIkLXT1C//9gFXA+ztGvwZ4OvD4iHhRZl424sXvCewK/Lh5EQiQmddFxEspLwOfGRFvzswc5YIzkxNOu4CTzrqUi9dcPcpZS5I0rxZNWn76j+GcE2HN6lnNRpIkSZqtiW5jLzNPBY4EVgDP7hh9KKWE3afaDw8RsUtE7NKeMCJuGRFbdc4/Im5E+foP4POZ2a724xjgz8B9I+LBrTAbAG+ufz846heBkiQtMHvX/pHtD3QAavXXx1G+wr/7oDOMiEdGxMsj4oURcf+I2KTHpPer/e92jsjM0yil7m8B7DDosgd1wmkX8LPTLjRTT5K0ECz8tPyqf8Gq48zUkyRJ0nph0kvsATyLUpf/eyJiH0pm290oDxcnU0rVtf259qM1bE/ggxHxE0oJuwuBmwMPoLSV90vgpe2ZZOa1EfEkSsm9wyPicOBvwD7AHpSHl3eOaB0lSVqobl37vaquPoVSCmBn4IcDzvPzHf/PjYhnZ+bhM1j2zrU7td8CI+JXPUbt0m3gSWdd2m2wJEmTaOGn5Vev6RdUkiRJmlcTXWIPri+1twewkpKh9yJgR+DdwN0HbKD7V5QHh20pVXW8iFLF5h+A5wL3yszVXZb9M+AulMbB96PU3b8ceC2wrw10S5I0reW1f3GP8c3wLQaY19eABwHbA0spmWpvrGG/EBEHzOGyhzLpJfWWL91o3FGQJK0/Fn5aft21swre1dItRj9PSZIkLQoLocQemfl34EkDThtdhv0BOGiGy/4T8IiZhJUkSaOTmZ0l5f8CvCIi/gm8l/JicJ2quka07N27Da9f/995LpY5Trtst2zcUZAkLUCLKi3f9rYjn6UkSZIWh4kvsSdJkiZa8yX98h7jm+GrZ7GMjwLXALtFRDtHaj6WvaAsX7oRd9thK+6xw9bjjookaf1hWj6MpVvAinvBivuMOyaSJEmaUAuixJ4kSZpYf6n9nXuM36n2e7WdM63MvCIiLgW2BDYHmgbu5nzZw3rBvr2iIknSemvxpuV7Hzwns5UkSZL6scSeJEkap6Nqf7+IWOu+pH6Rfy/gcuCnM11ARNya8iLwUuD81qgf1X5nez1ExA6Ul4RnAKfNdNmSJC0CpuWSJEnSPDJjT5IkjU1mngocCawAnt0x+lDKV/mfyszLmoERsUtE7NKeMCJuGRFbdc4/Im4EHFb/fj4zr2mNPgb4M3DfiHhwK8wGwJvr3w9mZs5k3SRJWgxMyyVJkqT5ZVWckiRp3J4FHA+8JyL2obyguxuwN6XqrFd2TP/n2o/WsD2BD0bETyhf5V8I3Bx4AKV9nV8CL23PJDOvjYgnUb72PzwiDgf+BuwD7AEcB7xzROsoSdJCZlouSZIkzRMz9iRJ0lhl5qkRsQfwWkpVWg8AzgLeDRyamRcNMJtfAZ8HdgfuBNyQUl3XH4AvAh/KzKu6LPtnEXEXSomC/YBllCq7Xgu8KTOvnOXqSZK04JmWS5IkSfPHjD1JkjR2mfl34EkDThtdhv0BOGiGy/4T8IiZhJUkSYVpuSRJkjQ/bGNPkiRJkiRJkiRJmgBm7EmSJEmSJEmSJEkTwIw9SZIkSZIkSZIkaQKYsSdJkiRJkiRJkiRNADP2JEmSJEmSJEmSpAmwZNwRkCRJWsjOveRK3vn9k8cdDUmSJEmSJC0AltiTJEmSJEmSJEmSJoAZe5IkSZIkSZIkSdIEMGNvwixfutG4oyBJkuaAabwkSRNk6RbjjoEkSZIWKTP2Jswu2y0bdxQkSdIcMI2XJGmCbHvbccdAkiRJi9SScUdAg1m+dCN22W4Z99hh63FHRZIkjZBpvCRJE2TpFiVTb8V9xh0TSZIkLVJm7K3nbnzDTXjBvjuPOxqSJGmGTMslSZpwy24Cex887lhIkiRJgFVxSpIkSZIkSZIkSRMhMnPccVAPEXHB0qVLt9p1113HHRVJ0nrsz3/+M2vWrLkwM63LcT1jWi5JGoRp+frLtFySNAjTcknzyYy99VhEXAlsCPxu3HFZgHap/ZPGGouFyW07d9y2c2fSt+0K4JLMvOW4I6K1mZYvWpN+TdHMuN8Xp1Ht9xWYlq+XTMsXPK/dC5v7d2Fb3/bvCkzLJc0T29hbv/0RIDN3H3dEFpqI+BW4beeC23buuG3njttWc8i0fBHymrI4ud8XJ/f7omBavoB5Di9s7t+Fzf0raTGzjT1JkiRJkiRJkiRpApixJ0mSJEmSJEmSJE0AM/YkSZIkSZIkSZKkCWDGniRJkiRJkiRJkjQBzNiTJEmSJEmSJEmSJkBk5rjjIEmSJEmSJEmSJGkaltiTJEmSJEmSJEmSJoAZe5IkSZIkSZIkSdIEMGNPkiRJkiRJkiRJmgBm7EmSJEmSJEmSJEkTwIw9SZIkSZIkSZIkaQKYsSdJkiRJkiRJkiRNADP2JEmSJEmSJEmSpAlgxp4kSZIkSZIkSZI0AczYmycR8fCIeG9EHBsRl0RERsSnZziv7SPi4xHxz4i4MiJWRcS7ImLLUcd7Uoxi+0bE1hHx1Ij4akT8NSLWRMTFEfGTiHhKRCzK82WUx27HfB9X55UR8dRRxHXSjHrbRsQ+9fg9u14b/hkR34uIB4wy3pNgxNfcB0bEkRFxZr0unBYRX4qIe4w63lq/mZYvXqPa93OVpmpueH+5OI3wfH9zRPwwIv5e9/uFEfGbiHhNRGw9F3HXYEzPFy6v2wub7yYWNt+PSNJwlow7AovIq4A7Av8CzgR2mclMImJH4HjgxsDXgJOAuwLPAw6IiHtl5gUjifFkGcX2fQTwf8BZwFHA34BtgYcCHwXuHxGPyMwcSYwnx0iO3baIuBnwvjrPG8x2fhNsZNs2It4CvKTO5+vA+cCNgN2BvYBvzzKuk2ZU19w3Ay8FLgCOoGzXWwEPAR4WEU/ITF/KLx6m5YvXqK7XI09TNae8v1ycRnWevgD4NfB94Fxgc+DuwCHA0yPi7pn591nHVjNher5wed1e2Hw3sbD5fkSShmDG3vx5ASVB+SuwJ+UGcSY+QHlweG5mvrcZGBHvqMt4A/DM2UV1Io1i+54MPBj4VmZe1wyMiFcAPwceRrmZ//KsYztZRnXsAhARARxGySj5CvDi2UZwgo1k20bE0yg3rZ8Anp6ZV3WM32iW8ZxEs962EXETyvF5DnCHzDy3NW5v4EfAawEz9hYP0/LFa1T7fqRpquac95eL06jO0xtm5hWdAyPiDcArgIOBZ800kpoV0/OFy+v2wua7iYXN9yOSNASrEJgnmXlUZp4ym6+66heB+wGrgPd3jH4NcBnw+IjYfMYRnVCj2L6Z+aPM/Eb75r0OPxv4YP271yyiOZFGsW07PBe4H/AkyjG7aI3ourAJ5aXB3+hy01qXc/UsojmRRnTc3oKSTv6snanXzB+4lPLVnxYJ0/LFa1Rp4RykqZpD3l8uTiM839fJ1Ku+WPs7zWb+mjnT84XL6/bC5ruJhc33I5I0HDP2JsvetX9kl5vMS4HjgM0oVbxotJqE/5qxxmLCRcSuwJuAd2fmj8cdnwViX0rm0leA66K0B/eyiHhe2AbcbJ0CXAXcNSK2aY+IiPsCy4AfjCNimmim5ZIa3l8uTg+q/d+PNRaaLdPzxcnr9gLhu4kFy/cjkhYNq+KcLLeu/ZN7jD+F8tXgzsAP5yVGi0BELAGeUP9+d5xxmWR1O36K8uXUK8YcnYXkLrV/BfAb4HbtkRHxY+DhmXnefEds0mXmhRHxMuAdwJ8i4ghKNS07Uqrn+T7wjPHFUBPKtFyS95eLSES8mNJu03JgD+DelEy9N40zXpo10/NFxuv2wuG7iQXN9yOSFg0z9ibL8tq/uMf4ZvgWcx+VReVNlJuBb2fm98YdmQn2auBOwL0zc824I7OA3Lj2XwL8CbgP8FvglsDbKC8UvoTVxcxIZr4rIlYBHwee1hr1V2BlZxWd0gBMyyWB95eLyYuBbVv/vwsc5EvFiWd6vvh43V44fDexcPl+RNKiYVWcUh8R8VzgRcBJwOPHHJ2JFRF3o3wJ9/bMPGHc8Vlgmuv4NcCDM/MnmfmvzPwD8J+Uxqf3tNqJmYmIlwKHAyspJfU2B3YHTgM+ExFvGV/sJEmTyPvLxSUzb5KZAdwEeCiwA/CbiLjzeGMmaVBetxcO300seL4fkbRomLE3WZqv/pb3GN8MXz33UVn4IuI5wLspX/nsnZkXjjlKE6lWc/FJSjU1/zPm6CxEq2v/N5m5qj0iMy8Hmq9J7zqPcVoQImIv4M3A1zPzhZl5WmZenpm/pjwU/AN4UUTsMMZoavKYlkuLmPeXi1dmnpOZX6WUFtiacn+syWV6vkh43V44fDexKKyufd+PSFrwzNibLH+p/Z17jN+p9nvV868BRcTzgfcCf6TcvJ893hhNtBtQjtldgSsiIpsOeE2d5iN12LvGFckJ1lwXVvcYf1HtL537qCw4/1H7R3WOqA8FP6eko3eaz0hp4pmWS4uU95cCyMwzKBkEt42IbcYdH82Y6fki4HV7wfHdxMLn+xFJi4Zt7E2W5uXyfhGxQWZe14yIiGXAvYDLgZ+OI3ILRUS8jFJ//m+BfTPz/PHGaOJdCXysx7g7UzJFfkK5AbMqjOH9EEjgNp3XhappLPr0+Y3WgrBJ7d+ox/hm+FXzEBctHKbl0iLk/aU6/FvtXzvWWGg2TM8XOK/bC5LvJhY+349IWjQssbceioiNImKXiNixPTwzTwWOBFYAz+4Idiil7adPZeZl8xLRCdVr+9Zx/0O5ef8VsI8378Pptm0zc01mPrVbB3y9TvaJOuwL44n5+q/PdeEM4BvAzYHndYTZD9if8rXad+cpqhOnzzXh2Np/ekTctCPM/SkvbK4Ajp+HaGrCmJYvXv3uM7RweX+5OPXa7xGxc0SsU0VjRGwQEW8Abgwcn5kXdU6j9Yvp+cLldXth893Ewub7EUmyxN68iYgDgQPr35vU/j0iYmX9fX5mvrj+vinwZ+AMyoNC27MoL5HfExH71OnuBuxNqebjlaOP/fpvFNs3Ip4IvJby5eyxwHMjonNRqzJzZefAhWyEx646jHDbPpvydeE7IuKBwG+AW9Z5Xws8NTMvZhEZ0bY9HPgB8O/AnyPiq8DZlKpb/gMI4OWZecGcrITWO6bli9eo9v2Q89GYeX+5OI3ofH8A8MaI+AmlVMAFwLbAnsAOlPuJp83JCmhapucLl9fthc13Ewub70ckaThm7M2f3YAndgzboXZQEqNpX+Rk5qkRsQflRvMAykPjWZTGnA9dxF997sbst+8ta39D4Pk9pjkGWDl07Cbbbozg2FVXuzGa68KZEbE78GrgwcB9gUsoX6q9MTN/PqoIT5DdmOW2zczrIuIBlAeDRwH/CWwGXAh8G3hPZh45wjhr/bcbpuWL1W6MJi0c1Xw0P3bD+8vFaDdmv99/ANwKuDfl5eIWwGWUzJ5PUe4hLhxNdDUDu2F6vlDthtfthWw3vI9ayHbD9yOSNLDIzHHHQZIkSZIkSZIkSdI0bGNPkiRJkiRJkiRJmgBm7EmSJEmSJEmSJEkTwIw9SZIkSZIkSZIkaQKYsSdJkiRJkiRJkiRNADP2JEmSJEmSJEmSpAlgxp4kSZIkSZIkSZI0AczYkyRJkiRJkiRJkiaAGXuSJEmSJEmSJEnSBDBjT5IkSZIkSZIkSZoAZuxJkiRJkiRJkiRJE8CMPUmSJEmSJEmSJGkCmLEnSZIkSZIkSZIkTQAz9iRJkiRJkiRJkqQJYMaeJEmSJEmSJEmSNAHM2JMkSZIkSZIkSZImwJJxR0CS+omIbYBzgZhm0quA5Zl5xdzHSpIkDcq0XJKkyWZaLknS+sWMPUnruxsAr239XwE8EfgV8M3W8PN8eJAkab1kWi5J0mQzLZckaT0SmTnuOEjSwCLiqcBHgJdk5tvGHR9JkjQc03JJkiababkkSeNlG3uSJs1utf/bMcZBkiTN3G61/9sxxkGSJM3cbrX/2zHGQZKkRcuMPUmT5o61/7uxxkKSJM2UabkkSZPNtFySpDGyKk5JEyMiAlgN/Cszbzrm6EiSpCGZlkuSNNlMyyVJGj9L7EmaJLcEbohfBUqSNKlMyyVJmmym5ZIkjZkZe5ImyS61/8exxkKSJM2UabkkSZPNtFySpDEzY0/SJLlh7V8y1lhIkqSZMi2XJGmymZZLkjRmS8YdAUkawsm1//yI2Ar4RWZ+bpwRkiRJQzEtlyRpspmWS5I0ZpbYkzQxMvPXwKuAq4HnA3uMNUKSJGkopuWSJE0203JJksYvMnPccZAkSZIkSZIkSZI0DUvsSZIkSZIkSZIkSRPAjD1JkiRJkiRJkiRpApixJ0mSJEmSJEmSJE0AM/YkSZIkSZIkSZKkCWDGniRJkiRJkiRJkjQBzNiTJEmSJEmSJEmSJoAZe5IkSZIkSZIkSdIEMGNPkiRJkiRJkiRJmgBm7EmSJEmSJEmSJEkTwIw9SZIkSZIkSZIkaQKYsSdJkiRJkiRJkiRNADP2JEmSJEmSJOn/t3fv8XLU5eHHPw8Q4AAh4SKQihihhFC8oKBIESVQEG1tUbFeKhLbn1alVfFWUVsB661VES/UWzWtWtFiC9p6wQpohEAV70gAiUdBQtBAQggh3J7fHzPbs9nsnsuePTs7ez7v12tec85cn5nv7j4z8535jiRJNWDFniRJkiRJkiRJklQDVuxJkiRJkiRJkiRJNWDFniRJkiRJkiRJklQDVuxJkiRJkiRJkiRJNWDFniRJkiRJkiRJklQDVuxJkiRJkiRJkiRJNWDFniRJkiRJkiRJklQDVuxJkiRJkiRJkiRJNWDFniRJkiRJkiRJklQDVuxJkiRJkiRJkiRJNWDFniRJkiRJkiRJklQDVuxJkiRJkiRJkiRJNWDFniRJkiRJkiRJklQDVuxJkiRJkiRJkiRJNWDFniRJkiRJkiRJklQDVuxJkiRJkiRJkiRJNWDFnqShFBFnRkRGxLKqY1FvRcSysmzPrDoWSVJ/RMRl5W//0pbhx5TDR6uJTL1mmUqS1F+eY0tS/WxXdQCaHSJiJ+BU4OnAY4A9gQRuA64GLgS+mJmbqopR0tYiYiGwFFiXme+vNBhJfWHOlmaXiHg1MB9YlpmjlQYjaSB5bKB+i4hjgGOAH2bmhVXG0iwiDgVOAkYzc1mXy1iI59iSpGnyiT3NuIh4BnAjcB7wR8DDgAeBB4CFwLOBTwM/j4hjKwpTw+e3wHXA6qoDqbmFwFuBV1cbhqR+MGerhu6myPc3Vh1Ijb2aItcvrDYMSYPIYwNV5BiK3HRStWFs5VCKuJZOYxkL8RxbkjRNVuxpRpXNJV0I7ENx0eUUYM/M3CUzd6W4O/hk4DLgd4AnVxGnhk9mfigzF2fmGVXHIkl1YM5WHWXm/5b5/riqY5GkYeOxgSRJ0mCyKU7NmIh4DPARigrkrwAntzbNkZnrgS8CX4yI5wL79j1QSZJmOXO2JElq5rGBJEnS4PKJPc2kvwd2AH4NvGCi9vYz8/PA+1qHR8QOEfGaiLgqItZHxKaIuC4i3hcR+7SZ/uPlS38vGG99EXFGOd0POox/RkRcFBG3RsS9EXFbRHw5Ip7aYfql5fIuK///s4j4VkSsLYefVA7fNiKeFhEfjYirI2JNufxbIuI/J9N8SUScWu6PjRFxe0RcGhF/VI4bLdd3TId5d4mIN0XEd8v9eU9E3BARH4iIh0207nFimlI5lfO07rNnlNuyLiLuiogrI+L5XcZzZrnsZW3GZdktjIiDIuKzEbE6Iu6OiB9ExClN00ZEvDQivhcRG8r9fX5E7NdhvXPL7fpCRPy03JZNEfHziPhYRBw4QdwjZezXlWWzulzfI8t4MyJynPkXRsQHy/nvLmO+OiL+JiJ2nsIuJCJGgUvLfx/etN8a3dKmafeMiFeU35mV5Xo3RsTPys/A74yznr0i4h/L/bWx3O6bIuKKiDg7Ih4+hZi3iYjzyvjuiIgjm8ZtHxGvKpe7LiLuK79/P4qIDzdPK81C5mxzdmU5u2kdJ0bEJWVMd5bLPGWCeY4p4xptGrZjOX829vU4868sp3tVm3FT3v8RcVm5vKURMT8i3l2u4+6IWNc0XVc5qZzvryJiefl52hwRv4yIT0bEweNta5tlnRnFMUUjz14aW+b5y5qm7fq7EEVuXlp+XtaW2/qbiLimjPvEKcb9B+VnLiPiXS3jnhIRF0TEzWV868syuzAi/jIiPP+VJs9jA48N+n5sEOU5L0VTlQCnxtbnoQvbzDfp8o7i/D7LfffIDnF8tJzmVxExvxyWwKfKSZ7SJq5jJrF9o3iOLUnqhcy0s+t5BzyUot39BN4wjeU8BPh+uZwE7gHubPr/duCJLfMsKcdtAnYdZ9k/Lqd7fcvwOcBnmtaRwPqW/9/dZnlLy3GXAR8o/36gjPEB4KRyuke2WfZdLcPOGCfujzdN9wBwR9O+fhUwWv59TJt5D24an8B9Leu+HTiqH+XUZp/9bdM2rWvZH6/uIqYzy3mXtRnXWO6fNsW5rmk/JvBaIIB/K/+/t2Vf/QrYo82y/6ppmvuBtcDmpmF3AX/QIeZ5wPeapt3c9Nm7C3hBY1yH+Z9F8blvzL+xjLvx/4+BvaewD79bll2jXG5t6Z7bNO17Wj5Xa8vtbwy7DXh0m3U8HLilZZ/d3lIWL2uZZ1k5/MyW4dsBny3HrQEe0zLusqZlPkjx3WmO8fzp/O7Z2dW1w5xtzq44Z5fLf32b3+gHyv/fy9hv+NKW+Y4ph4+2DP+Xcvi/jbPOxzGWe/bpxf5vivP1FO+kat7H68ppuspJwALghy2fqeay2wQ8awr7/HUU+byxn29nyzz/H03Tdv1dYCw3N7p1bHlsdOVkyrQc98xyfybwxpZxL21Zz8Y2Me7Y7W+cnd1s6vDYwGODio4NKN7heGvTdm1i6/PQh/WgvL9cjvshsH3LuD9iLD8f2zT81qZl39smrt+fxPZ5jm1nZ2dn15Ou8gDshrMD/qwpkS+exnK+2nQg+Rxg23L44YwdyN9K0c5/Y55tgJvLcad2WO4hTQce+7aMO6ccd0O5zp3L4XOBlzN2gPv8lvmWlsM3lMv9O2B+OW5XYK/y70XAPwMn0HSiAuwFvKU8CHoQOKJN3C9u2q/vAOY1zfsJioPLjbQ5EaCoNPpFOe4LwKOb9uf+TQdrtzbinslyatln68rtfkvTPtsb+HfGDuZ3n2JMZ5bzLmszLpvW+2XgEU3l9E+MXQx6W1meLwS2p6joexKwupzmH9os+3kUd7c+nvIEoZxvMWMnHLc1Plct8zYuQt5VrnNO0+d1BcVBcgLZZt7Hl+V/X7n+h5bDtwWOpDiBSODrU9yPx9Dh4lrLdK8EzgAeBWzXtO7DgK+Vy/gpEC3zfZKx79vRwDbl8B0oTprfRnkS3TTPMlpOOoAdgS+Vw38FLGqZ50VN5fpCygt7ZYz7Aacxzgm4nd0wd5izzdkTlFPLPltH73P2kxi72PRpyko2inc3vbtpvcnkK/aeylhO3anDev+xnOYbvdr/jF3g2kCRj05kLLf9btmfck6iuHD5v+V8/0OR2xvHCQsY+y5sBA6Y4v4fbfc5bJmmq+8Cxfu2shz/amBuOTzKuE8F3jPJMn1R03pe3jJup3KfZxln80Xf3cty+DdaLt7a2dm17/DYwGODCcqpZZ+to4/n8z0q770oKqqSpvN6ikrOW8vh722zvsY2XzaN78UxeI5tZ2dnZzfNrvIA7Iazo6hYSIo7vaLLZRzN2EHvU9uM35uxO53Obhn3XsapxGiK71stww+kOAi/jaYLAi3TPK9xANUyfGlTvO+Yxr5r3On2qZbhwdiB/Mc6zPvfTTEc02Gbx7tzvXFA/7o+lVPzPntzm3lHyrJI4EVT3I9nMnHF3vWUB8hN47YpD4Cz03opXhqfwKopxhTAN2hzkkpxMta4qPmCNvPOo+muuzbjv1OO+8sO6969af7DpxDzMUzipGOCZewAXFMu5ykt435WDn/uFJa3jKaTDoqTtkubynS/NvOcV47/p263w85uWDvM2ebsyZVT8z7rdc7+ZjnfJe0+gxQXOxvrXtoyrm2eorio1Lhg9/w2ywyKi1QJvLhX+5+xir17gUd2mHfKOQn4f+U836as0GszzUfKaT40xf0/2u5z2KPvwhvK4V+dwrK2KlPgr8vv+33AC9vM8wTGKnK37XY77Ozsig6PDTw2mFw5Ne+zvp3P96K8y3HPKMc9QHmeCvxnOewnwA5t5mls82XT+Ixslee6WIbn2HZ2dnazvPMdA5ope5T9OzIzu1zGyWX/e5n59daRmbmG4gIGFE0qNvu3sn9cROzVZtnPa5mu4UUUB9yfz8ybOsR1AUXzQYdExII24x+gzbsFpuDLZf+oluGPAxaWf/9Dh3nfPc5yTy377x1nmsb+OH6caVpNp5wa7gHe32beTUBjmW3bvp+m92Tm/S3rfJDiwiIUd4p+ps183yz7j4gpvLeu/C78d/lva/k+k+KzdxPwuTbzrmdsP24hIg4ol7eO4u7Rduu+neJED6ZWvtOWmZspKjRh6+2+s+y3+y5NKCL2oCiPYyjuJj06M3/VZtJprUcacubs7pmzx+btKmdHxO4Uza5B0VRWu8/gOya7vKZ4HqB4ogGg3ft9nkTR3Nc9wH+0jOvF/v9qZv60w7huclIjpnMz874O03x2gphmUqfvQmNb94ou328XEX9L0SzevcDJmdnu2KyxnjmM/aZJ6p7HBt3z2GBs3pk+n59WeWfml4GPUdzc+68RcTpwEkW+eWF5HjuQPMeWJFmxp0H2uLJ/6TjTNCpgFjVXsGTm1cB1FHdrb3HwGRFHAAdQ3PH77y3L+/2yf2r50uWtOorKnjnldO1eTv3zzPzteBsWESMRcXpEXFa+1Pm+xsuSgR+Uk7W+CPmxZf/WzPx5h0VfWW5X6/oeBuxb/vuVcbbt3HG2q5Ouy6nJzzJzY4d5f132d5tCTJP1kw7Dbyv7Pysr+lqtafp7fuvIiNg3It4dxcvU10XEA03le045WafyvXyck+flHYY3Pre7ADePU77PLafr+qXq44mIxRHxoYj4cUTcGREPNm33q8rJWrf7K2X/3eXLtZdExMgkV/k7wLcomiG9kuKu1jUdpm1Uav5JRHwpIp5VnrBI6g1z9pbM2ZP3WIqLcg9SPH2+lcxcRXHjy1Q1LnA+taxAbPaCsv+V8uYZoKf7f8U4cU0pJ0XEdhRPpAF8dJyYGhWUM5Xnu/kufJPiAunjgMsi4oUR0TrNOKuM9wFnUzTz9YeZeVGHaW8ou+2BFWWciyMiprKNknrKY4MteWzQe70o79cAP6doOrJRofu3mfmjGYh3yjzHliR1sl3VAWhorS37u0VEdHmX30PK/q/Hmebmsh/AnhQn/Q2fo2i+4fnAh5qGN+7a/nr5FFOzxp1Gc8tuIju1Gfab8WaI4i6xyyja5m/YyNhLs7el2JbWA+Y9y/7qTsvOzHsjYi2wT8uo5juo2t3x2KrddnUy3XKC4j0GndxT9ueMM023Ou3LB8Ybn5kPNF0n2iKuiHgK8F8UlWwN6xnbjhGKdzRMuXwpmtJsp1G+21E0lTKRqZTvpETE84B/ZWx/PEix3Y27HHeh2ObW7X43xTsC/hh4RdndHxHfpWgG5eOZua7Dal9S9u8ATmy+MNsqM78VEX9H8a6MZ5QdEbGS4inKj2bmDZPaWGn4mLM7MGf3JWc3Ylo/zkVBKGKeUoVVZq6IiF8AjwCeDXwc/q+irPGEQuvTHr3a/x0/W13kpN0pKqxgck+jTfbi3aR1+13IzBsi4uUU3+ujy46IGKV4P9DHMvMHtLcfcHr598sz85sdpmscm70AuJCiefP3ld3tEXEJxbsbvzyNJ4+k2cZjgw48Nhio8/lpl3dmboyI0xh7svBK4D29CW96PMeWJI3HJ/Y0U64t+zsAB01zWTt2OV/jQs2REbEQIIomgP60ZXyzxnfi9MyMSXSXtVnGA22GNXs/xUnAKoqLTLtn5i6ZuVdm7gM8cdJbOHnN3/XdJrFdC7tYR7flNDQiYg5F0527AP8DPBkYycz5mblPWb6vaUzew1U3yvdHk/zcLu3huomIh1BcLJ0DfJ7iJes7ZuZuTdvdeFJxi+3OzM2Z+SfAkRRN0lxJ0U5/4//rI+IxHVb9FYqTyN2A82KCJr4y820U370zKE7c7gQWA68FfhYRL5rShkvDw5zd2fsxZ9ddo3nrFzQNO57iwuSdjDWR3dCr/T/uZ2uKOak5psdO5vM+3rq79H66/C5k5icpKldfDVxEUWGwEHgZcHVEvKnDrLdSPDUA8M4omh7vKDO/R/G+pRdSXAhdRVEpenK53v+OiG0n2lBJgMcG43k/HhsMil6UN8CfN/19IJO7WXZGeY4tSZqIFXuaKd+iOHCA4i6hbjTulNtvnGkazVEksEVzGeWdQd+jOMhptMF/DMVdXRspTvBbNZoYGG+dXYuI7YE/Kf/9s8z8j8y8o2WyTgeRje3r2H55ufx2d3I3N53Q622bVjkNmSMptvV24E8yc3lm3tMyTdflO864RvnOSNNbk/A0isrMnwEvyMyrc+v3/4x7cpSZV2bm32TmkRQnEc8HfkVxB+knOsz2XeDpFN/nFwD/PFGTW5n5i8x8V2aeSHGxbwnwbYqnHc+L9u/wkIadObsNc3bfcnYjpnkRMd4TBpNtvrFV48Lvk5uagGw87fEfbfL0TO7/LUwhJ61l7ELzjMbUzjS/C0DxjqbMPDczT6LI7U+geGoggLdFxKPbzLYZ+CPgcuChwCUR8fAJ1rMpMz+bmadm5gEUT++9k+Lz/DSKykRJE/PYoA2PDQbufH7a5R0Rf0bxyor7KZp/3QP45PRDmzbPsSVJ47JiTzMiM29mrF3vv46IXSczX8sBw/fL/lPGOZA4tuxfn+2bb2pczGncpd24kHNRZt7dZvrG+1BOnEy8XdiT4q5HGGt7v9UfdBjemH6fce5YPoI2TVxk5i8YO+h92iTinIpelNOwaJzwXN/h8wUTl+9R4+zHozsMb3xud4/inRO91HjH4HgH843t/nG2eSdhuT3Htg7vJDM3Zub5wEvLQYd1eJcDmfkdiiY/NgFLgY9MdOLRNO8D5V2bf0TxLoudKe6ElGYVc3ZH5uz+5OwfUFwo3AZ4UrsJIuIRdHnRLjOvAX5cLv95EbEjcFI5equnPWZ4/3c0Xk4qL+R9bwZjmijXT+e7sJUsfBd4DkXzbh3LPjPvorjA+L8Un4FLImLfdtN2mP8XmfkmiqcdAJ4y2Xml2cxjg448Nujf+fxkzkOnVd5RvLuw0czr2RTHB5uAEyPiFdOIayKeY0uSps2KPc2kt1Dcabsv8G/lhZSOIuJPGWumEOCCsn8IY3fFNU+/N2N33X6hw2LPpzhoelREPJaiqQxo32wHFM32JHBwRPzlBPF28/LnDYzd+fioNstcAPx1h3l/APyy/Pt1HaZ5wzjrXtaYNyIe2mmiKMwfZzmtelFOw6LR/vyB7T7vEXECxd1r7VxI8dl4GC0viC/n3ZUOd5ln5kqK5jUA/qFsErStKF70vkOn8W3cWfbnjTNNY7sf2eGA/yUUL7hvF8/27YaXNjUmY+zdQlvJzEspTsI2U5yonNs6zQTruZexJyGmsm+kYWLO3po5uw+yeD/SJeW/b+iQR944zdU0PkPPp7hQNZeimcdLOky/rOz3ev835usmJzViWjpO81mN5U/1897I9fM7jO/6uzDetmbmAxQX/WCc/JuZdwJPpbj4vD9F5d4WT7xMsE9h7JjCPC9NnscGW/PYoH8myk0wjfIujzeWlcu/EnhHeV79N+Uk/xgRze9RnEpcE/EcW5I0fZlpZzdjHfAXFAfiSdFO/wsp2qBvjJ8HPAu4tJzmzJb5v1oOv53i/RjblsMPA35UjrsV2GOcGL5ZTnc1Y01HzBln+veW0z1A0XTPvk3j5gInULxH7X9a5ltaznfZBPvkinK6HwOHlsO2AY4Dri/jy+Lr2XZ/Ztm9Ddi1HP4Q4GMUB08by/HHtMw7H7ixHPdLisqjkabx+1EcsP0AWDrFcu6qnCazzyhemJ7AsinG1HG+pn24sNt1tltGuY8b+/8LwIJy+AhFu/0bmsp3q20GPl2O20BxV+p25fDfo2iG6o5xPhuPp3gxeVI0e/EkYJty3LYUJ55/R/Gy9rbb3WE7dy4/Vwk8u8M0BzH2Pf8gML8cvivweoqLdo3tXtYy7/XAO8r4ty+HBUUTXY3Pzv+2zLOM9r8Xf0hx4pHAe1rGnQ98iuLC4Nym4QvLcQncDew5lc+Znd0wdZiz2y3fnD2FfUb3OftJTZ+9fwH2bvrMvaMcvq7sL22Z95hy+Og4y9+vafnfL/vnjjN91/sfuKxdnC3TTDknUTzBsaIct5bigt6uTeP3Af6Movm8Mzutu0M8n2Xs2GXHXn4XgPdQXDQ+iS1/T/YGPlDO8yDwyInKlKJ5r8Zn9GfAQ5rGnVTun5cAD28avlM5rHF88Iqp7Bs7u9ne4bFBu+V7bDCFfUb3xwbHM5bzDpyB8n5NOd9dwO82DQ/gG+W4qyjPyZvGH1iOuw84osvvlefYdnZ2dnbT7ioPwG74O4oT7TWMHcAmRcXFnS3DRoEnt8z7EMaaaEqKu4ua57sdOHKC9f9Fy3r+aYLptwXOa5lnPcUFpQebhl3aMt9SJncicER5cNNYzl1N/6+luEuu04lAULT33pj3/nIfPFh2f0VxkJ/t9gvwuxQXQprn/21LPAmcOsUy7qqcJrPPqFHFXjn8lS37ch3FQXeW++ivO20zxcla8368h7ELmRsonjRIYHOHmJ7WNH1j/t8ydtLQ6B4+xX35Ly3bM1p2JzdN876WddxBcXKVwNeAv2+3T1vivZ/iO9Ac72+AR7fMs4w2Jx1NvzeN/f32puEXNi3zwTK+jS3rPmUq+8XObhg7zNmtyzdnT2Gf0WXOLud9fcvv9O3lNifFRbvLyr+Xtsx3TOMzOcHyl7fst3EvxnW7/zvF2TLNhS3bOqmcBOwFfKdpugfKz+FdLTG9dYr7/timeTcDN1F8x8+f7ncBeH9LbOvZ+vfkTZMt0/Lze005/keUF5kpfrual3k3Y9+3xrD/puUCrZ2d3cQdHhu0Lt9jgynsM7o/n58D/JyxXHkbY+ehzZV3Uy5v4JGM3RT7l23W/dBym5P255zfalru2qa4njiF7fMc287Ozs5uWp1NcWrGZeaFFM3mnEbRTv/NFC/R3Y7iwOUCiqeTDsrMb7fM+xvgSIqmKr5HcTCxPXADxYWCQzJzBeP7IsVFioZOzXY01vlAZr6C4u7xz1AcWO8A7EjxouEvURxwnzzBejst/yqKbbqQ4sBnDsVB6keBQykuUnSaNylObP6c4qXGmylODi4D/jAzP0RxBxcUB3Ot8/8ceCzwCoq7Ku+guMvyfoo7Dj9GcUfWZ6a4Tb0op6GQmR+guGv1cooTrO2AlcBbgd+nOAnuNO864CiKuzd/TlG29wCfo7i77tpy0nUd5v8qsIjiAP/7FJ+P+RQnZVcA7wIOy8xftpt/HC+juPtxJcV34eFlt0vTul/D2B2imylOsH4AvJriM3V/h2X/Sbnsy4FbymXeS/F5fBfFZ+fHkw20/L15PsUJz5si4q3lqDdSNG3zNWAVxedzW4q7Xj8FPC4zPz3Z9UjDypy91fLN2X2Smf9IcYPKpRQXSbcr43tRZr62B6to/izdWJbtePHMyP4vdZWTMvM2infE/RnF9/M3FE8jQJGj/5XiCY53TSWYzLwEeCbFhcpNFBc0H07xFGBjmm6/C+dQ3PR0EcUTBEHxHb2J4r13T87Md0wh1t9QPBVzHfBo4Btlk3OXAKdQXCj9CcUx2FyKi5nfAF4EPCMzOx2PSOrAY4Otlu+xQR9k8X7Z4yhatfk1sBtj56HbNU03pfIum4/8TDnNf2fmR9us+9cUn3eAN0fEE1omeRZFZeIvKM5fG3GN21xtC8+xJUnTEsVxhaRhUL6E++cUB21zM/PeikNSD0XEXwCfAL6VmcdUHI4kaRrM2ZIkqZnHBpIkabJ8Yk8aLo2XbX/bk4DhUt5Z+Kry329UGYskqSfM2ZIkqZnHBpIkaVKs2JNqJiI+FREnR8QeTcMeERHnUTTTAMW7aFQzEbFfWb5HR8TO5bBtyqY/vg48iuJ9AZ+oMk5J0uSYsyVJUjOPDSRJUi/YFKdUMxFxM8W7T6B4MfGDjL1fBeDvM/Nv+x6Ypi0ifpfiPQYN6yja6W+01X8P8JzM/K8+hyZJ6oI5W5IkNfPYQJIk9YIVe1LNRMTzKV6G/Fhgb2An4DfACuC8zLykwvA0DRGxE8VLtE8AFgMPoXiZ+s3AJcB7M/OGzkuQJA0Sc7YkSWrmsYEkSeoFK/YkSZIkSZIkSZKkGvAde5IkSZIkSZIkSVINWLEnSZIkSZIkSZIk1YAVe5IkSZIkSZIkSVINWLEnSZIkSZIkSZIk1cB2VQegziLiF8CuwGjFoUiSBttC4M7MfETVgWhL5nJJ0iQtxFw+kMzlkqRJWoi5XFKfzLqKvYg4GXgKcCjwGGAu8NnMfGEXy9oXOBs4EdgDWA1cCJyVmXf0INxdR0ZGdj/44IN378GyJElD6tprr2XTpk1Vh9G1qnNzRPwecCZwDMWFu18C5wPvyszp7lhzuSRpQnXP5d2oOv9PgblckjSh2ZjLJVVn1lXsAW+hOGm4C7gZWNzNQiLiAOAKYC/gImAl8ATgVcCJEXFUZq6dZqyjBx988O5XX331NBcjSRpmhx12GN///vdHq45jGirLzRFxBHAJMAe4ALgJOBb4O+C4iDguMzd3E0/JXC5JmtAQ5PJu1OXc3FwuSZrQLM3lkioyG9+xdzqwiOKO/JdPYznnUZw4vDIzT8rMN2bmscA5wEHA26cdqSRJs0MluTkitgU+BewEnJyZL8jMvwGOAL4IHFXGJkmSes9zc0mSJKkLs+6Jvcy8tPF3RHS1jPKOwBMo2tj/cMvotwIvBU6JiNdm5sbuIpW6l5msWLWWlas3sH7TfVWHI2mGPPFXHwNg429vrjiS6akwNz8FOBj4dmZ+qSmeByPiDcCzgZdFxLszM7sKDGDDrXDpO7ueXZI0vFasKh4kq3su74bn5pJmgtdD1G/Dcl4uqV5m4xN7vbCk7F+cmQ82j8jMDcDlFHf/P7HfgUlQXCC4atXtHsRKmk26yc3Hlv2vtS4sM1cB1wMPB/bvebSSJKkXPDeXtAWvh0iSZoNZ98RejxxU9q/vMP4GirsGFwHfnGhhEdGpsf6u3jEgrVy9oeoQJKnfusnNk5lnUdndON7KzeWSJFWiZ+fm5nJpOHg9RJI0G/jEXnfmlf31HcY3hs+f+VCkrXlnmqRZqJvcbD6XJKnezOWStuD1EEnSbOATewMgMw9rN7y8Y/BxfQ5HkiRNkblckqR6M5dLkiSpLnxirzuNu/7mdRjfGL5u5kORJEl0l5vN55Ik1Zu5XJIkSbOOT+x157qyv6jD+APLfqd2/qVKnH58p4+spFq6dA8Adt7BdE53ubk/+XzuPrDkjGktQpI0nK68v0gxG7f/GnBbtcHUk+fmkibF6yGaMZ6XS6qAT+x159Kyf0JEbLEPI2IucBRwN3BlvwOTJGmW6iY3X1L2T2xdWETsT3GR8JfAqp5HK0mSesFzc0mSJM06VuyNIyLmRMTiiDigeXhm3ghcDCwETmuZ7SxgZ+DTmbmxL4FKkjRL9Dg3fwu4FnhyRPxx0zq2Ad5d/vuRzMzeboUkSZoKz80lSZKkMbPuGeGIOAk4qfx3n7J/ZEQsK//+bWa+rvz7oRQX/H5JcaLQ7BXAFcAHIuK4crojgCUUzXy8uffRS5I0fKrKzZn5QES8mOLJvQsi4gLgV8BxwOHA5cA5095ASZK0Fc/NJUmSpO7Muoo94FDg1JZh+5cdFCcKr2MCmXljRBwOnE3RhNfTgdXAucBZmXlHrwJWdzKTFavWsnL1BtZvuq/qcCTVTSaMLoc118CmdVVHM+wOpaLcnJlXRcTjKe7qPwGYW67vbOBdmbm5mw2SJEkTOhTPzSVJkqQpm3UVe5l5JnDmJKcdBWKc8TcBL+5FXOq9FavWctWq26sOQ1JdjS6H0curjmJWqDo3Z+bPgOdMZR5JkjQ9Ved/SZIkqa58x56G1srVG6oOQVKdrbmm6ggkSZIkSZIkaQtW7Glo2fzmluaNzKk6BKlebH5TkiRJkiRJ0oCxYk+aJRYvmFt1CJIkSZIkSZIkaRpm3Tv2pNlm3sgcFi+Yy5H771F1KJIkSZIkSZIkaRqs2NOsc/rxi6oOQVKdLTmj6gi2NvcC4Jaqo5AkSZIkSZI0w6zYkyRJkiSpRzKTFavWsnL1Bt/7LUmSJKnnrNiTJEmSJKlHVqxay1Wrbq86DEmSJElDapuqA5AkSZIkaVisXL2h6hAkSZIkDTGf2JvFbCJGkppkwuhyWHMNbFpXdTSSJKmmPLeSJEmSNJOs2JvFbCJGkpqMLofRy6uOQpIkSZIkSZI6sinOWcwmYiSpyZprqo5AkiRJkiRJksblE3uz2GxsImbeyJyqQ5A0qCbT/ObI/JmOQpIkDanTj1/Ucdxndt2Bm/sYiyRJkqT68ok9zSqLF8ytOgRJdbb3IVVHIEmSJEmSJGkW84k9zQrzRuaweMFcjtx/j6pDkVRHI/OLSr2FR1cdiSRJkiRJkqRZzIo9bWW8JmIkadZZckbVEUiSJEmSJEkSYFOckiRJkiRJkiRJUi1YsSdJkiRJkiRJkiTVgBV7kiRJkiRJkiRJUg34jr0aykxWrFrLytUbWL/pvqrDkaTJy4TR5bDmGti0rupoJEmSpIltuBUufWfVUUhbSJKb79jEbzds5p77H6w6nIHxxE4jLt2jn2FIkjSjrNiroRWr1nLVqturDkOSpm50OYxeXnUUkiRJklRrN9+xiZvv2FR1GJIkqQI2xVlDK1dvqDoESerOmmuqjkCSJEmSau+3GzZXHYIkSaqIFXs1NJPNb84bmTNjy5ak2jW/OTK/6ggkSZIkaSs2vzl5O27n5U9J0nAxs2kLixfMrToESRocex9SdQSSJEmSpGnYc+4OVYcgSVJP+Y49AcWTeosXzOXI/X2ZsCQxMr+o1Ft4dNWRSJKkAZWZrFi1lpWrN8xoqyqSpO7suN027Dl3B/bdbaTqUCRJ6ikr9obI6ccvqjoESerOkjOqjkCSJGlKVqxay1Wrbq86DPXD3H08XtXAufL+69sO99qQVJG5FwC3VB2FpFnCpjglSZIkSZqilas3VB2CJEmSpFnIij1JkiRJkqZoKs1vzhuZM4ORSJIkSZpNrNiTJEmSJGkGLV4wt+oQJEmSJA0J37E34G67czPnfKN9u+nSwMuE0eWw5hrYtK7qaCRJkqS+mjcyh8UL5nLk/ntUHYokSZKkIWHFnqSZM7ocRi+vOgpJkiSpb04/flHVIUiSJEkaYjbFKWnmrLmm6ggkSZIkSZIkSRoalT+xFxHHA8cDTwb2A/YENgG3AT8ELgG+lJm/rirGOvBl7BpINr+pyRiZX3UEamJeliRp9jH/S5IkSfVRScVeROwEvBL4S4qThihH3UNx4jAC7A8cADwbODcivgy8JzNX9D/iwefL2CXV1t6HVB3BrGdeliRp9jH/S5IkSfXU96Y4I+LPgRuAd1DcAXgWxZ2B8zNzp8zcNzP3oKh0/D3gz4EvAk8DvhMRn4+I/fod96CaNzKHI/bf3ZexS6qfkfmw8ChYeHTVkcxq5mVJkmYf878kSZJUX1U8sfcJ4ELgnZn53U4TZWYCK8tuWUTsCpwKvBFYCpw945EOgL123cGXr2v4LDmj6ggkjTEvS5I0+5j/JUmSpJqqomLv8Mz8/lRnysw7gQ9GxMeBhT2PSpKk2cm8LEnS7GP+lyRJkmqq701xtp48RMT2U5z/nsxc2duoJEmanczLkiTNPuZ/SZIkqb76XrHXxvqIeFnVQUiSJKCivBwR+0bEJyPilojYHBGjEfH+iNhtkvMfExE5ie5hLfONN+2VM7O1kiQNHM/LJUmSpJqooinOVjsAe1UdhCRJAirIyxFxAHBFud6LKN7j8wTgVcCJEXFUZq6dYDGjwFkdxj0KeBbw08y8qc34XwLL2gy/ecLgJUkaDp6XS5IkSTUxCBV7kqqUCaPLYc01sGld1dFImp3Oo7iY+MrM/GBjYES8DzgdeDsw7lMEmTkKnNluXER8rvzz4x1mH83MtvNKkiRJkiRJg6QWFXsR8TzgsMx8fdWxSENndDmMXl51FJJqpJd5uXxa7wSKJ+4+3DL6rcBLgVMi4rWZubGL5e8JPBPYBPzr9KKVJA2CzGTFqrWsXL2B9ZvuqzqcWcPzckmSJGkwDMI79gD+KCJeExHHRsQebcY/nuKOfUm9tuaaqiOQNHj6mZeXlP2LM/PB5hGZuQG4HNgJeGKXyz+Vonmxf8/MdR2mmR8Rfx4Rb4qI0yKi23VJkvpgxaq1XLXqdiv1es/zckmSJKkGBuWJvcPLLgEi4hbgR8BPgTuBFwFrKotOGmb9bn5zZH5/1yepG/3MyweV/es7jL+B4om+RcA3u1j+S8r+R8eZ5jHAPzcPiIgfAadk5k8ms5KIuLrDqMWTmV+SNHkrV2+oOoRh5Xm5JEmSVAODUrG3jOKO/EMpLq49Gnh62TW8ue9RSeq9vQ+pOgJJE1tG//LyvLK/vsP4xvD5U11wRDyFouLwp5l5RYfJ3gd8kaJi8R6Kiri/AU4GLomIQzPz11NdtyRp5gzyk3rzRuZUHcJ0LMPzckmSJGngDUrF3mhmtt4p/wiKi2u7Atdn5g8qiUxSb4zMLyr1Fh5ddSSSJjYsefmlZf9jnSbIzNe2DPoe8JyIuAB4NvA6JtHsWGYe1m54+STf4yYVrSSp9hYvmFt1CNMxLPlfkiRJGmqDUrG3lcz8BfCLquOQZq0lZ1QdgaQBMoN5ufFE3rwO4xvD101loRGxO0XF3Cbg013E9ZFy/id3Ma8kaZaZNzKHxQvmcuT+7V5NV1+el0uSJEmDZxAq9rLqACRJ0v/pd16+ruwv6jD+wLLf6R18nZwK7AD8S2au6yKu35T9nbuYV5JUgdOP75RKNAmel0uSJEk1MQgVe3sBu1QdhCRJAvqfly8t+ydExDaZ+WBjRETMBY4C7gaunOJyX1L2OzbDOYEnlv1VXc4vSVKdeF4uSZIk1cQ2VQeQmWsz85dVxyFJkvqflzPzRuBiYCFwWsvosyiemPt0Zm5sDIyIxRGxuNMyI+Jo4GDgp5l5xTjTPToi5rQbDry9/Pczk9wUSZJqy/NySZIkqT4G4Yk9Sb2QCaPLYc01sGld1dFI0lS8ArgC+EBEHAdcCxwBLKFogvPNLdNfW/ajw/JeWvYnelrvNcAzImI5cBOwGVgMnAhsC3wc+NzkN0OSJM1mmcmKVWtZuXoD6zfdV3U4kiRJGlJ9r9iLiJHM3FT1MqShM7ocRi+vOgpJNTMIeTkzb4yIw4GzKSrVng6sBs4FzsrMO6YQy27AycAm4NMTTH4hsCvwaOBYYEdgLfBV4OOZ+aWpbYkkSfUwCPl/GK1YtZarVt1edRiSJEkaclU0xfmLiHhVROww1Rkj4jERcRHwuhmIS6q3NddUHYGkehqIvJyZN2XmizNzQWZun5kPz8xXt6vUy8zIzLZP62XmHZk5kpk7Zea6CdZ5YWY+KzN/NzN3Lde7IDOfYaWeJGnIDUT+HzYrV2+oOgRJkiTNAlVU7H0deB+wOiL+KSKWRMRIp4kjYv+IeHlErAC+DzwGuLRPsUr10cvmN0fm925ZkgadeVmSpNnH/D8DbH5TVZs3stXroyVJ0hDqe1OcmXlqRHwIeDvFO3BeCjwQEddSNLt1B0VTWHsABwF7UrxD5zaKd+yck5mb+x23NKvsfUjVEUjqE/OyJEmzj/lfGk6LF8ytOgRJktQHfa/YA8jM7wInRMSBwF8AxwGHAo9qmfQ3wH8AXwS+mJne/ibNpJH5RaXewqOrjkRSH5mXJUmafcz/0vCYNzKHxQvmcuT+e1QdiiRJ6oNKKvYaMvMG4I0AEbET8FCKOwI3Abdl5uoKw5OGw5Izqo5AUk2YlyVJmn3M/zPv9OMXVR2CJEmShkilFXvNMvNu4IaykyRJFTIvS5I0+5j/JUmSpME3MBV7kiRJkqT6yExWrFrLytUbWL/J1hklSZIkqR+2qTqATiLi3yPipU3/HxQRz4mIh1QZlyRJs5F5WZLUasWqtVy16nYr9YbYTOf/iNg3Ij4ZEbdExOaIGI2I90fEblNczpMi4qJy/nsi4lcR8ZWIOLEXcUqSJEmDZJCf2Hsy8I8AEbEHcBUQwOaIOC4zf1JlcFJHmTC6HNZcA5vWVR2NJPWKeVmStIWVqzdUHYJm3ozl/4g4ALgC2Au4CFgJPAF4FXBiRByVmWsnsZyXA+cBG4H/BG4G9gWeBTwtIt6SmW/vNk5JkiRp0AzsE3vAXKDxku5nA78Adgc+DnhQrsE1uhxGL7dST9KwMS9LkrYwSE/qzRuZU3UIw2om8/95FJV6r8zMkzLzjZl5LHAOcNBklh8Rc4B3AvcAh2XmKZl5RmaeAhwObAbeHBE7TDNWSZIkaWAMcsXer4ADyr9PBj6dmQ8Ay4AnVhWUNKE111QdgSTNBPOyJGlgLV4wt+oQhtWM5P/yab0TgFHgwy2j30rx9N0pEbHzBIvaHZgHXJ+Z1zWPyMxrgeuBEWCXbmOVJEmSBs0gV+x9EvhQRPwjsAS4sBy+HbBTVUFJExqkJ/VG5lcdgaThYV6WJA2ceSNzOGL/3Tly/z2qDmVYzVT+X1L2L87MB5tHZOYG4PJy+RNVHt4G/AZYFBEHNo+IiEXAgcAPJ9OkpyRJklQXA/uOvcz8h4gAeCrwusxcVY56AvDLygKT6mTvQ6qOQNKQMC9Lkibr9OMXVR2CemQG8/9BZf/6DuNvoHiibxHwzXHiy4g4DfgMcHVE/CdwC/BQ4JnANcDzJhNQRFzdYdTiycwvSZIk9ctAVexFxMOABzLzFihOIoB/aJlsb+D8fscm1crI/KJSb+HRVUciqcbMy5IkzT59yv/zyv76DuMbw+dPtKDM/PeIuAX4HPCiplFrgE8Bq9rOKEmSJNXUQFTsRcSbgddQHrRHxEbgEuBjmfmV5mnLkwqpfpacUXUEkjQp5mVJkmafuub/iHgh8HHgP4C3UTxJ+HDgb4EPAU8B/nSi5WTmYR2WfzXwuF7FK0mSJE1X5e/YK08e3gbsBlwHfB+4C/hj4MsR8V8RMW+cRXS73n0j4pMRcUtEbI6I0Yh4f0TsNoVlXBYROU63Y6/jliRpJlWVlyVJUnUqyP+NJ/I6LbMxfN14Cynfo/dJiiY3T8nMlZm5KTNXAqcAVwPPiYhjphuwJEmSNCgG4Ym9lwB3AMdl5g8bAyPi8cBrKe6s+3pEPDkz7+3FCiPiAOAKYC/gImAlxTsCXgWcGBFHTfHl2md1GH7/tAKVJKn/+p6XJUlS5fqd/68r+51eyHhg2e/0Dr6GE4A5wLcy88HmEZn5YER8Gzis7C7rLlRJkiRpsAxCxd5DKZr2+GHzwMz8LvC8iFgOfJCiSZB39Wid51FU6r0yMz/YGBgR7wNOB94OvGyyC8vMM3sUlyRJVasiL0uSpGr1O/9fWvZPiIhtmivlImIucBRwN3DlBMvZoew/pMP4xnBvRpIkSdLQqLwpToqD9Q2dRmbmh4HL2fIl2F0rn9Y7ARgFPtwy+q3ARuCUiNi5F+uTJKlm+pqXJUnSQOhr/s/MG4GLgYXAaS2jzwJ2Bj6dmRsbAyNicUQsbpl2edk/OSIe3TwiIg4FTgaS4l2BkiRJ0lAYhIq9H1BUtI1nOfCIHq1vSdm/uE1THRsoTlZ2Ap442QVGxHMj4o0R8ZqIeFpE7DDxXJIkDaR+52VJklS9KvL/K4DbgA9ExIUR8c6IuISiFZ3rgTe3TH9t2f2fzPxf4FPACPDdiDg/It4dEZ8HrgJ2BM7NzGt6GLckSZJUqUFoivMtwLci4h2Z+aYO0zwE+G2P1ndQ2e/UVv8NFCc0i4BvTnKZ57f8f1tEnJaZF0xm5oi4usOo1rsRJUmaaf3Oy5IkqXp9z/+ZeWNEHA6cDZwIPB1YDZwLnJWZd0xyUX8BfBtYCjwVmAvcCXwH+Hhmtp6vS5IkSbVW+RN7mfkdijb63xgRyyPimRGxY2N8RDwNeAHwrz1a5byyv77D+Mbw+ZNY1kXAM4B9Ke4QXAy8s5z38xFxYtdRSpJUgQrysiRJqlhV+T8zb8rMF2fmgszcPjMfnpmvblepl5mRmdFmeGbmssw8JjN3y8ztMnP3zDzOSj1JkiQNo0F4Yo/MfHNE/JbiTr0LgAfL/7enqCT7cjluoGTmOS2DrgPeFBG3ULxY/J3A1yaxnMPaDS+f5HvcdOOUJGkq6pqXJUlS98z/kiRJUj1U/sReQ1lJtgh4A7CC4sRhNyAonopbFxHfjYiPRsTLIuIJXa6q8UTevA7jG8PXdbl8gE8A9wOHRsTcaSxHkqRK9DEvS5KkAWH+lyRJkgbfQDyx15CZq4H3Au+NiG2BQ4DDmrpHl32ABLbtYjXXlf1FHcYfWPY7vYNvQpl5T0RsoDgB2hnY0O2yJEmqSp/ysiSpJjKTFavWsnL1BtZvuq/qcDRDzP+SJEnSYBuoir1mmfkA8OOy+xRARGwD/B5wON03UXlp2T8hIrbJzAcbI8qn644C7gau7HL5RMRBFJV6G+jhy8UlSarKDOZlSVJNrFi1lqtW3V51GOoj878kSZI0eAa2Yq+dshLup2W3rMtl3BgRFwMnAKdRvAuv4SyKJ+w+mpkbGwMjYnE578qmYY8A1mfmFme2EfEQyhMe4PzMvL+bOCVJGnS9yMuSpPpYudqGSGT+lyRJkqpWq4q9HnoFcAXwgYg4DrgWOAJYQtEE55tbpr+27EfTsKcAH4mI7wCrgNuB/YCnU7yn73sU7yWQJEmSpNqbTPOb80bm9CESSZIkSZq9ZmXFXvnU3uHA2cCJFJVxq4FzgbMy845JLOZq4HyKdws8FtiVounNnwBfoHjq794ZCF+DIBNGl8Oaa2DTuqqjkSRJkgbC4gVzqw5B6rm1d93LOd+4vuowJEmSJGCWVuwBZOZNwIsnOW20GfYTYGmPw1JdjC6H0curjkKSJEkaCPNG5rB4wVyO3H+PqkOReu6BB7PqECRJkqT/M2sr9qRpWXNN1RFIkiRJA+H04xdVHYIkSZIkzRrbVB2AVEtTbX5zZP5MRCFJkiRJGmC+d1KSJEm9ZsWe1A97H1J1BJIkSZKkPvO9k5IkSeo1m+KUZtLI/KJSb+HRVUciSZIkSeoT3zspSZKkmWLFntRLS86oOgJJkiRJ0gzyvZKSJEmqkk1xSpIkSZIkSZIkSTVgxZ4kSZIkSZIkSZJUA1bsSZIkSZIkSZIkSTXgO/akVpkwuhzWXAOb1lUdjSRJkiRJkiRJEuATe9LWRpfD6OVW6klSH0XEvhHxyYi4JSI2R8RoRLw/InabwjIui4gcp9uxw3y/FxFfiIjbIuKeiLguIs6KiJHebaEkSZIkSZI0fT6xJ7Vac03VEUjSrBIRBwBXAHsBFwErgScArwJOjIijMnPtFBZ5Vofh97dZ9xHAJcAc4ALgJuBY4O+A4yLiuMzcPIV1S5IkSZIkSTPGij2pVbdP6o3M72UUkjSbnEdRqffKzPxgY2BEvA84HXg78LLJLiwzz5zMdBGxLfApYCfgTzLzS+XwbYAvAM8u1/+uya5bqqvMZMWqtaxcvYH1m+6rOhxJkiRJktSBTXFKvbL3IVVHIEm1Uz6tdwIwCny4ZfRbgY3AKRGx8wys/inAwcC3G5V6AJn5IPCG8t+XRUTMwLqlgbJi1VquWnW7lXqSJEmSJA04n9iTpmtkflGpt/DoqiORpDpaUvYvLivU/k9mboiIyykq/p4IfHMyC4yI5wKPAO4FrgUu6dCc5rFl/2utIzJzVURcDywC9gdunGCdV3cYtXgyMUtVW7l6Q9UhSJIkSZKkSbBiT5qsJWdUHYEkDaODyv71HcbfQFGxt4hJVuwB57f8f1tEnJaZF3Sx7kVlN27FnlR3Pqmnbs0bmVN1CJIkSZI0q9gUpyRJqtK8sr++w/jG8PmTWNZFwDOAfYERiqfl3lnO+/mIOHGm1p2Zh7XrgJWTiFuSamvxgrlVhyBJkiRJs4pP7EmSpKGQmee0DLoOeFNE3AJ8kKKSb6tmNyVJUzdvZA6LF8zlyP33qDoUSZIkSZpVrNiTJElVajwVN6/D+MbwddNYxyeAc4BDI2JuZjZeJtaPdUu1dvrxi6oOQZIkSZIkNbFiTzMnE0aXw5prYNO6qqORJA2m68p+p9qDA8t+p/fgTSgz74mIDcBuwM5Ao2JvxtctSZIkSZIk9ZLv2NPMGV0Oo5dbqSdJGs+lZf+EiNjiuCQi5gJHAXcDV3a7gog4iKJSbwPw26ZRl5T91nfvERH7U1T4/RJY1e26JUmSJEmSpF6yYk8zZ801VUcgSRpwmXkjcDGwEDitZfRZFE/YfTozNzYGRsTiiFjcPGFEPCIidm9dfkQ8BPhU+e/5mXl/0+hvAdcCT46IP26aZxvg3eW/H8nM7GbbJEmSJEmSpF6zKU7NnGF6Um9kftURSNIwewVwBfCBiDiOorLtCGAJRTOYb26Z/tqyH03DngJ8JCK+Q/GE3e3AfsDTKd6V9z3gDc0LycwHIuLFFE/uXRARFwC/Ao4DDgcup3g3nyRJkiRJkjQQrNiTJmPvQ6qOQJKGVmbeGBGHA2dTNIv5dGA1cC5wVmbeMYnFXA2cDxwGPBbYlaLpzZ8AXwA+mpn3tln3VRHxeIqnA08A5lI0v3k28K7M3DzNzZMkSZIkSZJ6xoo9aTwj84tKvYVHVx2JJA21zLwJePEkp402w34CLO1y3T8DntPNvJIkSZIkSVI/WbGn/ltyRtURSJIkSZIkSZIk1c42VQcgSZIkSZIkSZIkaWJW7EmSJEmSJEmSJEk1YMWeJEmSJEmSJEmSVANW7EmSJEmSJEmSJEk1sF3VAWgCG26FS99ZdRSSJKlLt925mXO+cX3VYUiSJEmSJGkI+MSeJEmSJEmSJEmSVANW7EmSJEmSJEmSJEk1YMWe+mtkftURSJIkaRLmjcypOgRJGjj+NkqSJKlqVuypv/Y+pOoIJEmSNAmLF8ytOgRJGjj+NkqSJKlq21UdgGaJkflFpd7Co6uORJIkSeOYNzKHxQvmcuT+e1QdiiQNDH8bJUmSNCis2Bt0c/eBJWdUHYUkSerSXrvuwOnHL6o6DEmS1KW9dt2BP3/SI6oOQ5IkSQJsilOSJEmSJEmSJEmqhcjMqmNQBxGxdmRkZPeDDz646lAkSQPs2muvZdOmTbdnpm1DDRhzuSRpMszlg8tcLkmaDHO5pH6yYm+ARcRmYFvgR1XHMqS2AfYG1gAPVhzLdA3atvQ7nplcX6+X3YvldbuMbudbXPZXTmEeTU6vPl8LgTsz0zaiBswA5PJefca6Wc5U5plo2vHGdxrXbni7YVX/xllGllGv5rGMBruMJppuIebygTQAubzZIJ2b9PL3aTLTTOV3arzhVf9eNczEOWw/zhNnsizHGzeIuaeZ302/m2Aul9RHVuwNsIi4GiAzD6s6lmEUEb8D/Bp4aGbeUnU80zFo29LveGZyfb1edi+W1+0ypjGfv0UzZNC+u+q9qr8/vfqMdbOcqcwz0bTjje80rt3wDsMsI8toXJaRZdSLecz59VX157slloE5N+nl79NkppnK79QEwweiPGfiN6Ef54kzWZbjjRvE3NPM76bfTUnqN9+xJ0mSJEmSJEmSJNWAFXuSJEmSJEmSJElSDVixp9lsA3BW2a+7QduWfsczk+vr9bJ7sbxulzFonxNZJpp5vfqMdbOcqcwz0bTjje80rt3wQfzOWUaWUa/msYwGu4wGcb+pfgbp3KSXv0+TmWYqv1NTja8KMxFfP84TZ7Isxxs328rT72Z1Bj0+SQJ8x95AG5T2pSXNbv4WSd3z+zP4LKPBZxkNPstIw8zP93CxPIeHZTlcLE9Jmhqf2JMkSZIkSZIkSZJqwCf2JEmSJEmSJEmSpBrwiT1JkiRJkiRJkiSpBqzYkyRJkiRJkiRJkmrAij1JkiRJkiRJkiSpBqzYkyRJkiRJkiRJkmrAij1JkiRJkiRJkiSpBqzYkyRJkiRJkiRJkmrAij1JkiRJkiRJkiSpBqzYGxIR8YqI+EVE3BMRV0fE0VXHJGl2iYgnR8SXIuLXEZERsbTqmKQ6MZcPPn/nBl9EnBER342IOyPiNxHx5Yh4ZNVxqRARp0XEj8vyuTMiVkTEH1Ydl9Qr5vLhYc4fLh4fDA+PJSSpYMXeEIiI5wLnAu8AHgtcAXw1IvarNDBJs80uwE+BVwGbKo5FqhVzeW34Ozf4jgHOA34fOBa4H/ifiNi9yqD0f24G/gZ4HHA4cAlwYUQ8utKopB4wlw8dc/5wOQaPD4aFxxKSBERmVh2DpikirgJ+nJkvaRp2A3BBZp5RXWSSZquIuAv4q8xcVnUsUh2Yy+vH37l6iIhdgPXASZn55arj0dYi4nbgjMz8aNWxSNNhLh9e5vzh4/HBcPFYQtJs5BN7fRARJ0fEByNiefmYeEbEZyaYZ9+I+GRE3BIRmyNiNCLeHxG7tUy3PXAYcHHLIi6muBNJkoCZ/S2Shp25vB78nRt8FZTRXIpznjt6sgFDrp/lExHbRsTzKJ6KuaKX2yG1Yy4fLub84eLxwfDwWEKS+mO7qgOYJd4CPAa4i+KR8cXjTRwRB1AkpL2Ai4CVwBMomoA4MSKOysy15eR7AtsCa1oWswb4g15tgKShMJO/RdKwM5fXg79zg6/fZXQu8ENgxbQjnx1mvHwi4lEU5bFjuZ5nZuZPerwdUjvm8uFizh8uHh8MD48lJKkPfGKvP04HFgG7Ai+fxPTnUSS0V2bmSZn5xsw8FjgHOAh4+4xFKmmY+Vskdc/vTz1YToOvb2UUEe8DngQ8OzMfmHbks0M/yuc64FDgCOCfgH+JiEf2IHZpIuaI4WJ5DhePD4aHxxKS1AdW7PVBZl6amTfkJF5oWN6pcgIwCny4ZfRbgY3AKRGxcznst8ADwN4t0+4N3DqduCUNlxn+LZKGmrm8HvydG3z9KqOIOAd4PnBsZq6aduCzRD/KJzPvzcyfZ+bV5XvHfkhxEVCaUeby4WLOHy4eHwwPjyUkqT+s2Bs8S8r+xZn5YPOIzNwAXA7sBDyxHHYvcDVwfMtyjsf2pSV1b0q/RZK2YC6vB3/nBl9XZRQR5zJ20W5lPwKdpXr1HdoG2KH34UnTYi4fLub84eLxwfDwWEKSumTF3uA5qOxf32H8DWV/UdOw9wFLI+L/RcTB5cHK7wAfmaEYJQ2/Kf8WRcQuEXFoRBxKkV/2K//fb+bClAaSubwe/J0bfN2U0YeBFwMvAO6IiH3KbpeZC3PW6qZ83hURR0fEwoh4VES8EzgG+OzMhSl1xVw+XMz5w8Xjg+HhsYQkdWm7qgPQVuaV/fUdxjeGz28MyMzPR8QeFC+oXQD8FHh6Zv5ypoKUNPSm/FsEHA5c2vT/WWX3L8DSHsYmDTpzeT34Ozf4uimjV5T9b7ZMexZwZk+iUkM35bMP8Jmyvx74MfC0zPz6TAQoTYO5fLiY84eLxwfDw2MJSeqSFXtDIjPPo3jhrCRVIjMvA6LqOKS6MpcPPn/nBl9mWj4DLDOXVh2DNJPM5cPDnD9cPD4YHh5LSFLBpjgHT+NulHkdxjeGr5v5UCTNYv4WSd3z+1MPltPgs4wGm+WjYebne7hYnsPF8hwelqUkdcmKvcFzXdlf1GH8gWW/U/vTktQL/hZJ3fP7Uw+W0+CzjAab5aNh5ud7uFiew8XyHB6WpSR1yYq9wdNow/2EiNiifCJiLnAUcDdwZb8DkzSr+Fskdc/vTz1YToPPMhpslo+GmZ/v4WJ5DhfLc3hYlpLUJSv2Bkxm3ghcDCwETmsZfRawM/DpzNzY59AkzSL+Fknd8/tTD5bT4LOMBpvlo2Hm53u4WJ7DxfIcHpalJHUvMrPqGIZeRJwEnFT+uw/wVGAVsLwc9tvMfF3T9AcAVwB7ARcB1wJHAEsoHj///cxc24/YJQ0Pf4uk7vn9qQfLafBZRoPN8tEw8/M9XCzP4WJ5Dg/LUpL6w4q9PoiIM4G3jjPJLzNzYcs8DwPOBk4E9gBWA/8JnJWZd8xMpJKGmb9FUvf8/tSD5TT4LKPBZvlomPn5Hi6W53CxPIeHZSlJ/WHFniRJkiRJkiRJklQDvmNPkiRJkiRJkiRJqgEr9iRJkiRJkiRJkqQasGJPkiRJkiRJkiRJqgEr9iRJkiRJkiRJkqQasGJPkiRJkiRJkiRJqgEr9iRJkiRJkiRJkqQasGJPkiRJkiRJkiRJqgEr9iRJkiRJkiRJkqQasGJPkiRJkiRJkiRJqgEr9iRJkiRJkiRJkqQasGJPkiRJkiRJkiRJqgEr9iRJkiRJkiRJkqQasGJPkiRJkiRJkiRJqgEr9iRJkiRJkiRJkqQasGJPkiRJkiRJkiRJqgEr9iRJkiRJkiRJkqQa2K7qACRpPBGxJ3AbEBNMei8wLzPvmfmoJEnSZJnLJUmqN3O5JEmDxYo9SYNuF+Dspv8XAqcCVwP/1TT8N548SJI0kMzlkiTVm7lckqQBEplZdQySNGkR8f+AjwOvz8z3VB2PJEmaGnO5JEn1Zi6XJKlavmNPUt0cWvZ/WGEMkiSpe4eW/R9WGIMkSereoWX/hxXGIEnSrGXFnqS6eUzZ/1GlUUiSpG6ZyyVJqjdzuSRJFbIpTkm1EREBrAPuysyHVhyOJEmaInO5JEn1Zi6XJKl6PrEnqU4eAeyKdwVKklRX5nJJkurNXC5JUsWs2JNUJ4vL/k8rjUKSJHXLXC5JUr2ZyyVJqpgVe5LqZNeyf2elUUiSpG6ZyyVJqjdzuSRJFduu6gAkaQquL/uvjojdge9m5ueqDEiSJE2JuVySpHozl0uSVDGf2JNUG5n5feAtwH3Aq4HDKw1IkiRNiblckqR6M5dLklS9yMyqY5AkSZIkSZIkSZI0AZ/YkyRJkiRJkiRJkmrAij1JkiRJkiRJkiSpBqzYkyRJkiRJkiRJkmrAij1JkiRJkiRJkiSpBqzYkyRJkiRJkiRJkmrAij1JkiRJkiRJkiSpBqzYkyRJkiRJkiRJkmrAij1JkiRJkiRJkiSpBqzYkyRJkiRJkiRJkmrAij1JkiRJkiRJkiSpBqzYkyRJkiRJkiRJkmrAij1JkiRJkiRJkiSpBqzYkyRJkiRJkiRJkmrAij1JkiRJkiRJkiSpBqzYkyRJkiRJkiRJkmrAij1JkiRJkiRJkiSpBv4/OJ05UvpEfqsAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 900x288 with 6 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 282,
       "width": 891
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(2, 3, figsize=((3.5*3) + 2, 4.0))\n",
    "\n",
    "\n",
    "\n",
    "ntotal_task = 10\n",
    "nrows = 4\n",
    "rowlist = [0, 1, 2, 3]\n",
    "tasklist = [0, 1, 2, 3]\n",
    "assert len(tasklist) <= 7\n",
    "ntask = len(tasklist)\n",
    "inds = np.array([[t + (ntotal_task * row) for t in tasklist] for row in rowlist]).reshape(-1).tolist()\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "snuba_perfs = [\n",
    "        0.2694, 0.1624, 0.0805, 0.2308, 0.2648, 0.5235, 0.7082, 0.7520, 0.6034, 0.5936,\n",
    "        0.2313, 0.0924, 0.1526, 0.2640, 0.2598, 0.8811, 0.4905, 0.6640, 0.7187, 0.5560,\n",
    "        0.4219, 0.1520, 0.1554, 0.1804, 0.2506, 0.9774, 0.6994, 0.7400, 0.5674, 0.5804,\n",
    "        0.6617, 0.7550, 0.3103, 0.2419, 0.0000, 0.8414, 0.0000, 0.6280, 0.5608, 0.5699,\n",
    "]\n",
    "snub_multi_perfs = [\n",
    "        0.2374, 0.1386, 0.0991, 0.2343, 0.2596, 0.5421, 0.7264, 0.7080, 0.6484, 0.6072,\n",
    "        0.5625, 0.1390, 0.2354, 0.5625, 0.2644, 0.8968, 0.5143, 0.6280, 0.7218, 0.5020,\n",
    "        0.3866, 0.2036, 0.1588, 0.2283, 0.3702, 0.9742, 0.7559, 0.7320, 0.5526, 0.5864,\n",
    "        0.3541, 0.6788, 0.2479, 0.2785, 0.0000, 0.7916, 0.0000, 0.7759, 0.5435, 0.5681,\n",
    "]\n",
    "\n",
    "snuba_perfs = np.array(snuba_perfs)[inds].tolist()\n",
    "snub_multi_perfs = np.array(snub_multi_perfs)[inds].tolist()\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(len(snuba_perfs)).tolist() + \\\n",
    "        np.arange(len(snub_multi_perfs)).tolist(),\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        [r'Snuba'] * len(snuba_perfs) + \\\n",
    "        [r'Snuba (multipolar)'] * len(snub_multi_perfs),\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1.0 - np.array(\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs,\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[0, 0], \n",
    "        xlim=None, log=False, frameon=False, legend=False)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "snuba_perfs = [\n",
    "        0.6842, 0.6409, 0.8387, 0.8105, 0.9090, 0.5379, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.9608, 0.7183, 0.9557, 0.9716, 0.9956, 0.9999, 0.8072, 0.9999, 0.9971, 0.9999,\n",
    "        0.8761, 0.9772, 0.9230, 0.9320, 0.8998, 0.9999, 0.5912, 0.9999, 0.9884, 0.9999,\n",
    "        0.7667, 0.7262, 0.1521, 0.1323, 0.0000, 0.4116, 0.0000, 0.9999, 0.4781, 0.2716,\n",
    "        ]\n",
    "snub_multi_perfs = [\n",
    "        0.7422, 0.8129, 0.4470, 0.5407, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.3472, 0.5807, 0.4763, 0.3472, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.6185, 0.6337, 0.8232, 0.7193, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.5027, 0.3372, 0.2033, 0.7235, 0.0000, 0.9999, 0.0000, 0.7320, 0.5926, 0.2816,\n",
    "]\n",
    "\n",
    "snuba_perfs = np.array(snuba_perfs)[inds].tolist()\n",
    "snub_multi_perfs = np.array(snub_multi_perfs)[inds].tolist()\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(len(snuba_perfs)).tolist() + \\\n",
    "        np.arange(len(snub_multi_perfs)).tolist(),\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        [r'Snuba'] * len(snuba_perfs) + \\\n",
    "        [r'Snuba (multipolar)'] * len(snub_multi_perfs),\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1.0 - np.array(\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs,\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[1, 0], \n",
    "        xlim=None, log=True, frameon=False, legend=False)\n",
    "\n",
    "        \n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "ntotal_task = 10\n",
    "nrows = 4\n",
    "rowlist = [0, 1, 2, 3]\n",
    "tasklist = [4, 5, 6]\n",
    "assert len(tasklist) <= 7\n",
    "ntask = len(tasklist)\n",
    "inds = np.array([[t + (ntotal_task * row) for t in tasklist] for row in rowlist]).reshape(-1).tolist()\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "snuba_perfs = [\n",
    "        0.2694, 0.1624, 0.0805, 0.2308, 0.2648, 0.5235, 0.7082, 0.7520, 0.6034, 0.5936,\n",
    "        0.2313, 0.0924, 0.1526, 0.2640, 0.2598, 0.8811, 0.4905, 0.6640, 0.7187, 0.5560,\n",
    "        0.4219, 0.1520, 0.1554, 0.1804, 0.2506, 0.9774, 0.6994, 0.7400, 0.5674, 0.5804,\n",
    "        0.6617, 0.7550, 0.3103, 0.2419, 0.0000, 0.8414, 0.0000, 0.6280, 0.5608, 0.5699,\n",
    "]\n",
    "snub_multi_perfs = [\n",
    "        0.2374, 0.1386, 0.0991, 0.2343, 0.2596, 0.5421, 0.7264, 0.7080, 0.6484, 0.6072,\n",
    "        0.5625, 0.1390, 0.2354, 0.5625, 0.2644, 0.8968, 0.5143, 0.6280, 0.7218, 0.5020,\n",
    "        0.3866, 0.2036, 0.1588, 0.2283, 0.3702, 0.9742, 0.7559, 0.7320, 0.5526, 0.5864,\n",
    "        0.3541, 0.6788, 0.2479, 0.2785, 0.0000, 0.7916, 0.0000, 0.7759, 0.5435, 0.5681,\n",
    "]\n",
    "\n",
    "snuba_perfs = np.array(snuba_perfs)[inds].tolist()\n",
    "snub_multi_perfs = np.array(snub_multi_perfs)[inds].tolist()\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(len(snuba_perfs)).tolist() + \\\n",
    "        np.arange(len(snub_multi_perfs)).tolist(),\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        [r'Snuba'] * len(snuba_perfs) + \\\n",
    "        [r'Snuba (multipolar)'] * len(snub_multi_perfs),\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1.0 - np.array(\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs,\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[0, 1], \n",
    "        xlim=None, log=False, frameon=False, legend=False)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "snuba_perfs = [\n",
    "        0.6842, 0.6409, 0.8387, 0.8105, 0.9090, 0.5379, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.9608, 0.7183, 0.9557, 0.9716, 0.9956, 0.9999, 0.8072, 0.9999, 0.9971, 0.9999,\n",
    "        0.8761, 0.9772, 0.9230, 0.9320, 0.8998, 0.9999, 0.5912, 0.9999, 0.9884, 0.9999,\n",
    "        0.7667, 0.7262, 0.1521, 0.1323, 0.0000, 0.4116, 0.0000, 0.9999, 0.4781, 0.2716,\n",
    "        ]\n",
    "snub_multi_perfs = [\n",
    "        0.7422, 0.8129, 0.4470, 0.5407, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.3472, 0.5807, 0.4763, 0.3472, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.6185, 0.6337, 0.8232, 0.7193, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.5027, 0.3372, 0.2033, 0.7235, 0.0000, 0.9999, 0.0000, 0.7320, 0.5926, 0.2816,\n",
    "]\n",
    "\n",
    "snuba_perfs = np.array(snuba_perfs)[inds].tolist()\n",
    "snub_multi_perfs = np.array(snub_multi_perfs)[inds].tolist()\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(len(snuba_perfs)).tolist() + \\\n",
    "        np.arange(len(snub_multi_perfs)).tolist(),\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        [r'Snuba'] * len(snuba_perfs) + \\\n",
    "        [r'Snuba (multipolar)'] * len(snub_multi_perfs),\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1.0 - np.array(\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs,\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[1, 1], \n",
    "        xlim=None, log=True, frameon=False, legend=False)\n",
    "\n",
    "        \n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "ntotal_task = 10\n",
    "nrows = 4\n",
    "rowlist = [0, 1, 2, 3]\n",
    "tasklist = [7, 8, 9]\n",
    "assert len(tasklist) <= 7\n",
    "ntask = len(tasklist)\n",
    "inds = np.array([[t + (ntotal_task * row) for t in tasklist] for row in rowlist]).reshape(-1).tolist()\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "snuba_perfs = [\n",
    "        0.2694, 0.1624, 0.0805, 0.2308, 0.2648, 0.5235, 0.7082, 0.7520, 0.6034, 0.5936,\n",
    "        0.2313, 0.0924, 0.1526, 0.2640, 0.2598, 0.8811, 0.4905, 0.6640, 0.7187, 0.5560,\n",
    "        0.4219, 0.1520, 0.1554, 0.1804, 0.2506, 0.9774, 0.6994, 0.7400, 0.5674, 0.5804,\n",
    "        0.6617, 0.7550, 0.3103, 0.2419, 0.0000, 0.8414, 0.0000, 0.6280, 0.5608, 0.5699,\n",
    "]\n",
    "snub_multi_perfs = [\n",
    "        0.2374, 0.1386, 0.0991, 0.2343, 0.2596, 0.5421, 0.7264, 0.7080, 0.6484, 0.6072,\n",
    "        0.5625, 0.1390, 0.2354, 0.5625, 0.2644, 0.8968, 0.5143, 0.6280, 0.7218, 0.5020,\n",
    "        0.3866, 0.2036, 0.1588, 0.2283, 0.3702, 0.9742, 0.7559, 0.7320, 0.5526, 0.5864,\n",
    "        0.3541, 0.6788, 0.2479, 0.2785, 0.0000, 0.7916, 0.0000, 0.7759, 0.5435, 0.5681,\n",
    "]\n",
    "\n",
    "snuba_perfs = np.array(snuba_perfs)[inds].tolist()\n",
    "snub_multi_perfs = np.array(snub_multi_perfs)[inds].tolist()\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(len(snuba_perfs)).tolist() + \\\n",
    "        np.arange(len(snub_multi_perfs)).tolist(),\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        [r'Snuba'] * len(snuba_perfs) + \\\n",
    "        [r'Snuba (multipolar)'] * len(snub_multi_perfs),\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1.0 - np.array(\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs,\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[0, 2], \n",
    "        xlim=None, log=False, frameon=False)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "snuba_perfs = [\n",
    "        0.6842, 0.6409, 0.8387, 0.8105, 0.9090, 0.5379, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.9608, 0.7183, 0.9557, 0.9716, 0.9956, 0.9999, 0.8072, 0.9999, 0.9971, 0.9999,\n",
    "        0.8761, 0.9772, 0.9230, 0.9320, 0.8998, 0.9999, 0.5912, 0.9999, 0.9884, 0.9999,\n",
    "        0.7667, 0.7262, 0.1521, 0.1323, 0.0000, 0.4116, 0.0000, 0.9999, 0.4781, 0.2716,\n",
    "        ]\n",
    "snub_multi_perfs = [\n",
    "        0.7422, 0.8129, 0.4470, 0.5407, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.3472, 0.5807, 0.4763, 0.3472, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.6185, 0.6337, 0.8232, 0.7193, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999, 0.9999,\n",
    "        0.5027, 0.3372, 0.2033, 0.7235, 0.0000, 0.9999, 0.0000, 0.7320, 0.5926, 0.2816,\n",
    "]\n",
    "\n",
    "snuba_perfs = np.array(snuba_perfs)[inds].tolist()\n",
    "snub_multi_perfs = np.array(snub_multi_perfs)[inds].tolist()\n",
    "\n",
    "problems = pd.Series(\n",
    "        np.arange(len(snuba_perfs)).tolist() + \\\n",
    "        np.arange(len(snub_multi_perfs)).tolist(),\n",
    "        dtype=int, name='problem')\n",
    "methods = pd.Series(\n",
    "        [r'Snuba'] * len(snuba_perfs) + \\\n",
    "        [r'Snuba (multipolar)'] * len(snub_multi_perfs),\n",
    "        dtype=str, name='method')\n",
    "\n",
    "\n",
    "objective_vals = pd.Series(\n",
    "        1.0 - np.array(\n",
    "                snuba_perfs + \\\n",
    "                snub_multi_perfs,\n",
    "                ), \n",
    "        dtype=float, name='obj')\n",
    "\n",
    "df = pd.DataFrame([problems, methods, objective_vals]).T\n",
    "\n",
    "taus, solver_vals, solvers, transformed_data = optperfprofpy.calc_perprof(\n",
    "    df, ['problem'], ['obj'], ['method'])\n",
    "\n",
    "\n",
    "my_draw_simple_pp(\n",
    "        taus, solver_vals, solvers, axs[1, 2], \n",
    "        xlim=None, log=True, frameon=False, legend=False)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "axs[0, 0].set_title('Accuracy on image tasks')\n",
    "axs[1, 0].set_title('Coverage on image tasks')\n",
    "\n",
    "axs[0, 1].set_title('Accuracy on diverse tasks')\n",
    "axs[1, 1].set_title('Coverage on diverse tasks')\n",
    "\n",
    "axs[0, 2].set_title('Accuracy on text tasks')\n",
    "axs[1, 2].set_title('Coverage on text tasks')\n",
    "\n",
    "plt.suptitle('Accuracy and coverage performance profiles across task types and Snuba polarity')\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"neurips/snuba_polarity.pdf\", bbox_inches='tight')\n",
    "plt.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.7.11 ('FWRENCH')",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.11"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "806ed366974b6be290ad49c0cf4ac974a5c82640c8d0846094220742baddf83d"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}