jacobians v2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "import numpy as np\n",
    "import os\n",
    "import torch\n",
    "\n",
    "def load_jacobian(density_idx, slope_idx, seed, data_dir):\n",
    "    filepath = f\"jacobian_density-{density_idx}_slope-{slope_idx}_seed-{seed}.pt\"\n",
    "    jac_path = os.path.join(data_dir, \"jacobians_7x7\", filepath)\n",
    "    return torch.load(jac_path).double()[1:, :]\n",
    "\n",
    "def load_jtfs(density_idx, slope_idx, seed, data_dir):\n",
    "    filepath = f\"jtfs_density-{density_idx}_slope-{slope_idx}_seed-{seed}.pt\"\n",
    "    jtfs_path = os.path.join(data_dir, \"jtfs_7x7\", filepath)\n",
    "    return torch.load(jtfs_path).double()[1:, 0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([2])"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Microstructure. Orange: JTFS. Green: SCRAPL')"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "H_jtfs = []\n",
    "H_scrapl = []\n",
    "common_slope_idx = 3\n",
    "y_density_idx = 3\n",
    "data_dir = \"icassp2024_scrapl_data\"\n",
    "\n",
    "for x_density_idx in range(7):\n",
    "\n",
    "    SX, JX, SY = [], [], []\n",
    "    for seed in range(10):\n",
    "        y_coords = (y_density_idx, common_slope_idx)\n",
    "        x_coords = (x_density_idx, common_slope_idx)\n",
    "        Sx = load_jtfs(*x_coords, seed=seed, data_dir=data_dir)\n",
    "        Sy = load_jtfs(*y_coords, seed=seed, data_dir=data_dir)\n",
    "        Jx = load_jacobian(*x_coords, seed=seed, data_dir=data_dir)\n",
    "        SX.append(Sx)\n",
    "        SY.append(Sy)\n",
    "        JX.append(Jx)\n",
    "    SX = torch.stack(SX, dim=0)\n",
    "    SY = torch.stack(SY, dim=0)\n",
    "    JX = torch.stack(JX, dim=0)\n",
    "\n",
    "    G_scrapl = []\n",
    "    for _ in range(100):\n",
    "        Sx_scrapl, Sy_scrapl, Jx_scrapl = [], [], []\n",
    "        for path in range(SX.shape[-1]):\n",
    "            random_seed = torch.randint(SX.shape[0], (1,))\n",
    "            Jx_scrapl.append(JX[random_seed, path, :].squeeze())\n",
    "            Sx_scrapl.append(SX[random_seed, path].squeeze())\n",
    "            Sy_scrapl.append(SY[random_seed, path].squeeze())\n",
    "        Jx_scrapl = torch.stack(Jx_scrapl, dim=0)\n",
    "        Sx_scrapl = torch.stack(Sx_scrapl, dim=0)\n",
    "        Sy_scrapl = torch.stack(Sy_scrapl, dim=0)\n",
    "        g_scrapl = Jx_scrapl.T @ (Sy_scrapl - Sx_scrapl)\n",
    "        G_scrapl.append(g_scrapl)\n",
    "    G_scrapl = torch.stack(G_scrapl, dim=0).squeeze()\n",
    "\n",
    "    G_jtfs = []\n",
    "    for seed in range(10):\n",
    "        y_coords = (y_density_idx, common_slope_idx)\n",
    "        Sy = load_jtfs(*y_coords, seed=seed, data_dir=data_dir)\n",
    "        g_jtfs = JX[seed, :, :].T @ (SY[seed, :] - SX[seed, :])\n",
    "        G_jtfs.append(g_jtfs)\n",
    "    G_jtfs = torch.stack(G_jtfs, dim=0).squeeze()\n",
    "\n",
    "    # Take gradient with respect to coordinate 0 (density)\n",
    "    H_jtfs.append(G_jtfs[:, 0].detach())\n",
    "    H_scrapl.append(G_scrapl[:, 0].detach())\n",
    "\n",
    "H_jtfs = torch.stack(H_jtfs)\n",
    "H_scrapl = torch.stack(H_scrapl)\n",
    "\n",
    "positions = np.arange(-3, 4)\n",
    "plt.boxplot(H_jtfs, positions=positions-1/6, widths=1/3,\n",
    "    patch_artist=True, boxprops=dict(facecolor='orange'));\n",
    "plt.boxplot(H_scrapl, positions=positions+1/6,widths=1/3,\n",
    "    patch_artist=True, boxprops=dict(facecolor='green'));\n",
    "\n",
    "plt.gca().set_xticks(positions)\n",
    "plt.gca().set_xticklabels(positions)\n",
    "plt.grid(linestyle='--')\n",
    "plt.xlabel(\"Relative Density Index\")\n",
    "plt.ylabel(\"Density gradient\")\n",
    "# Microstructure means matched random seeds between\n",
    "# target and prediction.\n",
    "plt.title(\"Microstructure. Orange: JTFS. Green: SCRAPL\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Mesostructure. Orange: JTFS. Green: SCRAPL')"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "H_jtfs = []\n",
    "H_scrapl = []\n",
    "common_slope_idx = 3\n",
    "y_density_idx = 3\n",
    "data_dir = \"icassp2024_scrapl_data\"\n",
    "Sy = load_jtfs(y_density_idx, common_slope_idx, seed=0, data_dir=data_dir)\n",
    "\n",
    "for x_density_idx in range(7):\n",
    "\n",
    "    SX, JX, SY = [], [], []\n",
    "    for seed in range(1, 10):\n",
    "        y_coords = (y_density_idx, common_slope_idx)\n",
    "        x_coords = (x_density_idx, common_slope_idx)\n",
    "        Sx = load_jtfs(*x_coords, seed=seed, data_dir=data_dir)\n",
    "        Jx = load_jacobian(*x_coords, seed=seed, data_dir=data_dir)\n",
    "        SX.append(Sx)\n",
    "        JX.append(Jx)\n",
    "    SX = torch.stack(SX, dim=0)\n",
    "    JX = torch.stack(JX, dim=0)\n",
    "\n",
    "    G_scrapl = []\n",
    "    for _ in range(100):\n",
    "        Sx_scrapl, Sy_scrapl, Jx_scrapl = [], [], []\n",
    "        for path in range(SX.shape[-1]):\n",
    "            random_seed = torch.randint(SX.shape[0], (1,))\n",
    "            Jx_scrapl.append(JX[random_seed, path, :].squeeze())\n",
    "            Sx_scrapl.append(SX[random_seed, path].squeeze())\n",
    "        Jx_scrapl = torch.stack(Jx_scrapl, dim=0)\n",
    "        Sx_scrapl = torch.stack(Sx_scrapl, dim=0)\n",
    "        g_scrapl = Jx_scrapl.T @ (Sy - Sx_scrapl)\n",
    "        G_scrapl.append(g_scrapl)\n",
    "    G_scrapl = torch.stack(G_scrapl, dim=0).squeeze()\n",
    "\n",
    "    G_jtfs = []\n",
    "    for seed in range(SX.shape[0]):\n",
    "        g_jtfs = JX[seed, :, :].T @ (Sy - SX[seed, :])\n",
    "        G_jtfs.append(g_jtfs)\n",
    "    G_jtfs = torch.stack(G_jtfs, dim=0).squeeze()\n",
    "\n",
    "    # Take gradient with respect to coordinate 0 (density)\n",
    "    H_jtfs.append(G_jtfs[:, 0].detach())\n",
    "    H_scrapl.append(G_scrapl[:, 0].detach())\n",
    "\n",
    "H_jtfs = torch.stack(H_jtfs)\n",
    "H_scrapl = torch.stack(H_scrapl)\n",
    "\n",
    "positions = np.arange(-3, 4)\n",
    "plt.boxplot(H_jtfs, positions=positions-1/6, widths=1/3,\n",
    "    patch_artist=True, boxprops=dict(facecolor='orange'));\n",
    "plt.boxplot(H_scrapl, positions=positions+1/6,widths=1/3,\n",
    "    patch_artist=True, boxprops=dict(facecolor='green'));\n",
    "\n",
    "plt.gca().set_xticks(positions)\n",
    "plt.gca().set_xticklabels(positions)\n",
    "plt.grid(linestyle='--')\n",
    "plt.xlabel(\"Relative Density Index\")\n",
    "plt.ylabel(\"Density gradient\")\n",
    "# Mesostructure means that the target has seed=0\n",
    "# whereas predictions have seed>0 between 1 and 9\n",
    "plt.title(\"Mesostructure. Orange: JTFS. Green: SCRAPL\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([10, 233, 2])"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "JX.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Microstructure. Orange: JTFS. Green: SCRAPL')"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlQAAAHHCAYAAAB5gsZZAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAACHYklEQVR4nO3deVxUVeMG8Gdm2GVxAUQUBDfANFHc0FJxyYUy0zRNc8m017Qyte2Xa6aWZWll+fqWpilZufWmlppibrgEuAOpifiqKGhsgsDM3N8fxlwPM8AM2zDX5/v5+CnOnLlzznMvcLj3nHtVkiRJICIiIqJyU1u7AURERES2jgMqIiIiogrigIqIiIiogjigIiIiIqogDqiIiIiIKogDKiIiIqIK4oCKiIiIqII4oCIiIiKqIA6oiIiIiCqIAyqq0VQqFebOnWvtZhAREZWKAyqqct988w1UKhVUKhUOHjxo9LokSfDz84NKpcLjjz9uhRaWz8KFC7F169YHvg3lsW3bNvTr1w/16tWDk5MTWrRogRkzZuDWrVvWbppVFX2v/PHHH0hOTjZ835T1Lzk5Gfv27Svx9eHDhxs+IycnB3PmzEGrVq1Qq1Yt1KtXD6GhoXj11Vdx7dq1crc9Pz8fn332GR555BHUqVMHDg4O8PX1xcCBA/Hdd99Bp9NVRkQ1ys8//4zu3bvD29sbLi4uaNKkCYYNG4Zff/3VqG5WVhbmzZuHNm3awNXVFc7OzmjVqhXefPNNIfexY8cK+87R0REtWrTA7Nmzcffu3RLbMmzYMKhUKrz55psmXy9+fNjb26NJkyYYPXo0/vrrL0O9ouPuo48+qkAyDyY7azeAHhxOTk6IiorCI488IpT//vvv+N///gdHR0ej9+Tl5cHOrmYepgsXLsTTTz+NQYMGPdBtsNSMGTOwZMkStGnTBm+++Sbq1q2LuLg4fP7559iwYQP27NmDoKAgazfT6ry8vPDtt98KZUuWLMH//vc/fPLJJ0Z1k5OTAQCvvPIKOnToILweEBAAACgsLES3bt2QmJiIMWPG4OWXX0ZOTg7Onj2LqKgoPPXUU/D19bW4rWlpaejfvz9iY2PRt29fzJw5E3Xr1kVqaip+++03PPvss7hw4QJmzZpl8bZrqo8++givv/46unfvjrfffhsuLi64cOECfvvtN2zYsAH9+vUz1P3rr7/Qu3dvpKSkYOjQoZg4cSIcHBxw6tQpfP3119iyZQv+/PNPQ31HR0d89dVXAIDMzEz89NNPmD9/Pi5evIj169cbtSUrKws///wzAgIC8N133+H999+HSqUy2e6i46OwsBBxcXFYuXIltm/fjtOnT5dr39N9JKIqtnr1agmANHjwYMnT01MqLCwUXp8wYYIUFhYmNW7cWIqMjKySNuTk5FT6NmvVqiWNGTPGap9vaRvMpdPppLy8vErdZpGoqCgJgPTMM89IWq1WeO3o0aOSi4uL1Lp1a6NjpLg7d+5USfusreh75fjx4yZfj4yMlBo3bmzytejoaAmA9OOPP5a4/R9++EECIK1fv97otby8PCkzM7Nc7e7bt6+kVqulTZs2mXz9+PHj0rp160rdRl5enqTT6cr1+dWtsLBQcnd3l/r06WPy9Rs3bgh127RpI7m4uEgHDhwwqpuZmSn93//9n+HrMWPGSLVq1RLq6PV6qXPnzpJKpZJSU1ONtrFq1SrJ3t5e2rt3rwRA2rdvn1Gdko6PTz/9VAIgLVy4UJIkSbp06ZIEQPrwww9LSYBM4SU/qjYjRozArVu3sHv3bkNZQUEBNm7ciGeffdbke0zNobp69SrGjx8PX19fODo6IjAwEJMmTUJBQQEA+bLJ77//jpdeegne3t5o1KiR4f1ffPEFHnroITg6OsLX1xeTJ09GRkaG8Bnnz5/HkCFD4OPjAycnJzRq1AjDhw9HZmamoV137tzBmjVrDKfQx44dCwCYO3cuVCoVzp07h2effRZ16tQxnJXr0aMHevToYdTPsWPHGs4iFNHr9Vi2bBlat24NJycneHl5oV+/fvjjjz/KbIOp7d3ftuIZT5kyBevXrzfkUnTJ4urVq3j++edRv359ODo64qGHHsKqVatM7itzzJs3D3Xq1MHKlSuh0WiE1zp27Ig333wTp0+fxsaNGw3lPXr0QKtWrRAbG4tu3brBxcUF//d//wcA+OmnnxAZGWk4Fpo2bYr58+cbXV4q2sa5c+cQEREBFxcXNGzYEIsXLzZq4+XLlzFw4EDUqlUL3t7eeO2117Bz506oVCrs27dPqHv06FH069cPHh4ecHFxQffu3XHo0CGjbSYmJiIlJaW8sVWaixcvAgC6du1q9JqTkxPc3d0t3mZMTAx27tyJiRMnYvDgwSbrtG/fHiNHjjR8XXT5acOGDZg5cyYaNmwIFxcXZGVlATA/V3OOz6LP+uGHH7BgwQI0atQITk5O6NWrFy5cuCDUzc3NRWJiItLT00vtc3p6OrKyskzmCADe3t6G/9+0aRNOnjyJd955x+jsPAC4u7tjwYIFpX6eSqXCI488AkmShMtzRdavX48+ffogIiICISEhJs9ilaRnz54AgEuXLpn9HjKtZl5LIUUKCAhAeHg4vvvuO/Tv3x8A8MsvvyAzMxPDhw/Hp59+WuY2rl27ho4dOyIjIwMTJ05EcHAwrl69io0bNyI3NxcODg6Gui+99BK8vLwwe/Zs3LlzB8C9AcW8efPQu3dvTJo0CUlJSfjyyy9x/PhxHDp0CPb29igoKEDfvn2Rn5+Pl19+GT4+Prh69Sq2bduGjIwMeHh44Ntvv8ULL7yAjh07YuLEiQCApk2bCm0dOnQomjdvjoULF0KSJIvzGj9+PL755hv0798fL7zwArRaLQ4cOIAjR46gffv2ZrXBXHv37sUPP/yAKVOmwNPTEwEBAbhx4wY6d+5sGHB5eXnhl19+wfjx45GVlYWpU6da9Bnnz59HUlISxo4dW+Iv7tGjR2POnDnYtm2bMO/n1q1b6N+/P4YPH45Ro0ahfv36AO4Nnl1dXTFt2jS4urpi7969mD17NrKysvDhhx8K2/7777/Rr18/DB48GMOGDcPGjRvx5ptvonXr1obj8c6dO+jZsyeuX7+OV199FT4+PoiKikJ0dLTJzPr374+wsDDMmTMHarUaq1evRs+ePXHgwAF07NjRUDckJATdu3c3GpBVhezsbKMBQd26daFWq9G4cWMAwNq1azFz5swSLwtZ4ueffwYAjBo1yuL3zp8/Hw4ODpgxYwby8/Ph4OBgdq6WHp/vv/8+1Go1ZsyYgczMTCxevBgjR47E0aNHDXWOHTuGiIgIzJkzp9TFMN7e3nB2dsbPP/+Ml19+GXXr1i2x7n//+18AwHPPPWdxPvcruqRbp04dofzatWuIjo7GmjVrANz7w/WTTz7B559/Lvw8LEnRILtevXoVah+Bl/yo6t1/GePzzz+X3NzcpNzcXEmSJGno0KFSRESEJEmSyUt+AKQ5c+YYvh49erSkVqtNXhLR6/XC5z3yyCPCZaWbN29KDg4O0mOPPSZcWvj8888lANKqVaskSZKk+Pj4Mi+dSFLJl9vmzJkjAZBGjBhh9Fr37t2l7t27G5WPGTNGuJRTdOr+lVdeKbGfpbWh+PaKt+1+ACS1Wi2dPXtWKB8/frzUoEEDKT09XSgfPny45OHhYdiH5tq6dasEQPrkk09Krefu7i61a9fO8HX37t0lANKKFSuM6ppqw4svvii5uLhId+/eNdrG2rVrDWX5+fmSj4+PNGTIEEPZkiVLJADS1q1bDWV5eXlScHCwBECKjo6WJOnePmjevLnUt29fYX/k5uZKgYGBRpeCAJjc78VVxiU/U/8uXbpkaF9QUJAEQGrcuLE0duxY6euvvxYuUVnqqaeekgBIGRkZQnleXp6UlpZm+Pf3338btbVJkybCPrQkV3OPz6LPCgkJkfLz8w31li1bJgGQTp8+bdSu+3/mlGT27NkSAKlWrVpS//79pQULFkixsbFG9dq2bSt5eHiUub0iRZf8inK7cOGC9NFHH0kqlUpq1aqVkIskSdJHH30kOTs7S1lZWZIkSdKff/4pAZC2bNki1Cvq26pVq6S0tDTp2rVr0vbt26WAgABJpVIZjjle8is/XvKjajVs2DDk5eVh27ZtyM7OxrZt20q83FecXq/H1q1b8cQTT6B9+/ZGrxf/a3vChAnCZaXffvsNBQUFmDp1KtRqtVDP3d0d27dvBwB4eHgAAHbu3Inc3FyL+1jkX//6V7nfu2nTJqhUKsyZM8fotco4q1Bc9+7d0bJlS8PXkiRh06ZNeOKJJyBJEtLT0w3/+vbti8zMTMTFxVn0GdnZ2QAANze3Uuu5ubkZLv0UcXR0xLhx44zqOjs7C9tPT0/Ho48+arh0cz9XV1fhLIqDgwM6duwoXEL59ddf0bBhQwwcONBQ5uTkhAkTJgjbOnHiBM6fP49nn30Wt27dMmRz584d9OrVC/v374derzfUlySpWs5OAcDs2bOxe/du4Z+Pjw+Ae3kdPXoUr7/+OoB7Z/jGjx+PBg0a4OWXX0Z+fr7Fn1e0r1xdXYXyFStWwMvLy/DP1OWuMWPGCPvQ3FzLc3yOGzdOOGPz6KOPAoCw/3v06AFJksy6Vcu8efMQFRWFtm3bYufOnXjnnXcQFhaGdu3aISEhQcinrGO+uDt37hhya9asGWbMmIGuXbvip59+Mvr+X79+PSIjIw2f0bx5c4SFhZV42e/555+Hl5cXfH19ERkZaZg2YOpnKlmGl/yoWnl5eaF3796IiopCbm4udDodnn76abPem5aWhqysLLRq1cqs+oGBgcLXly9fBgCjFWQODg5o0qSJ4fXAwEBMmzYNH3/8MdavX49HH30UAwcOxKhRowyDrfJ8viUuXrwIX1/fUi8lVKbibU1LS0NGRgZWrlyJlStXmnzPzZs3LfqMoh/4RQOrkmRnZwtzUACgYcOGJi9fnD17FjNnzsTevXuNBmFF892KNGrUyOiXUZ06dXDq1CnD15cvX0bTpk2N6jVr1kz4+vz58wDuDQhKkpmZaXR5pjq0bt0avXv3LvF1Dw8PLF68GIsXL8bly5exZ88efPTRR/j888/h4eGB9957z6LPK9qvOTk5wvfHkCFDDN+r06dPN3nbhOLHnbm5FhYWWnx8+vv7C18X7Zu///67xM8qy4gRIzBixAhkZWXh6NGj+OabbxAVFYUnnngCZ86cMcxLMzXvqTROTk6GS6n/+9//sHjxYty8eVMYfAJAQkIC4uPjMXr0aGE+WI8ePbB8+XJkZWUZXV6fPXs2Hn30UWg0Gnh6eiIkJKTGrqS2NUyRqt2zzz6LCRMmIDU1Ff3790ft2rWr5HOK//CxxJIlSzB27Fj89NNP2LVrF1555RUsWrQIR44cESa4W/r5KpXK5Hyqyr5HT0lnsUr6nOJtLTq7MmrUqBJ/uT388MMWtSkkJAQAhAFMcZcvX0ZWVpZwtsxU+wAgIyMD3bt3h7u7O9599100bdoUTk5OiIuLw5tvvimcIQJgNAm+iKn9UZaibX/44YcIDQ01Waf4GZuaqHHjxnj++efx1FNPoUmTJli/fr3FA6rg4GAAwJkzZ4RJ2n5+fvDz8wNwb/BiaqJ3ScddWbkW3a/MkuOzMvd/ce7u7ujTpw/69OkDe3t7rFmzBkePHkX37t0RHByM+Ph4XLlyxZBHWTQajTAo7tu3L4KDg/Hiiy8a5mQBwLp16wAAr732Gl577TWj7WzatMnozG5ZA24qPw6oqNo99dRTePHFF3HkyBF8//33Zr/Py8sL7u7uOHPmTLk+t2hCblJSEpo0aWIoLygowKVLl4x+yLRu3RqtW7fGzJkzcfjwYXTt2hUrVqww/MIpz6W3OnXqmPxrtejsWJGmTZti586duH37dqlnqUpqQ506dYxWLpr6nJJ4eXnBzc0NOp2u0n74tmjRAi1atMDWrVuxbNkyk5dB1q5dCwBm3eB13759uHXrFjZv3oxu3boZyiuyWqlx48Y4d+4cJEkSsi2+Gqxo8r+7u7sifjnVqVMHTZs2Ldf31uOPP473338f69evL3HVm7nMzbUqjs/K0r59e6xZswbXr18HADzxxBP47rvvsG7dOrz99tvl2maDBg3w2muvYd68eThy5Ag6d+4MSZIQFRWFiIgIvPTSS0bvmT9/PtavX2/yUjlVDc6homrn6uqKL7/8EnPnzsUTTzxh9vvUajUGDRqEn3/+2XDrgPuV9Zdm79694eDggE8//VSo+/XXXyMzMxORkZEA7s150Gq1wntbt24NtVotzDGpVauWyUFLaZo2bYrExESkpaUZyk6ePGm0JHzIkCGQJAnz5s0z2sb9bS+pDU2bNkVmZqZwNuj69evYsmWLWe3UaDQYMmQINm3aZPKX7P3tB8y/LcDs2bPx999/41//+pfR2bLY2Fh88MEHaNWqFYYMGWJWGwExj4KCAnzxxRdlvrckffv2xdWrV4WzAHfv3sV//vMfoV5YWBiaNm2Kjz76CDk5OUbbKW8+Ve3kyZMmzxRdvnwZ586dEy6HZ2ZmIjEx0ejSaXFdu3ZFnz59sHLlSvz0008m65h7FsjcXC09Ps1l7m0TcnNzERMTY/K1X375BYA8teDpp59G69atsWDBApPvyc7OxjvvvFNm215++WW4uLjg/fffBwAcOnQIycnJGDduHJ5++mmjf8888wyio6MrdPd7sgzPUJFVlDZHojQLFy7Erl270L17d0ycOBEhISG4fv06fvzxRxw8eLDUy4deXl54++23MW/ePPTr1w8DBw5EUlISvvjiC3To0MEwYXnv3r2YMmUKhg4dihYtWkCr1eLbb781/BAvEhYWht9++w0ff/wxfH19ERgYiE6dOpXa/ueffx4ff/wx+vbti/Hjx+PmzZtYsWIFHnroIWEOUEREBJ577jl8+umnOH/+PPr16we9Xo8DBw4gIiICU6ZMKbUNw4cPx5tvvomnnnoKr7zyCnJzc/Hll1+iRYsWZk8mf//99xEdHY1OnTphwoQJaNmyJW7fvo24uDj89ttvuH37tqGuubcFGDlyJI4fP45ly5bh3LlzGDlyJOrUqYO4uDisWrUK9erVw8aNG2Fvb19m+7p06YI6depgzJgxeOWVV6BSqfDtt99W6BLOiy++iM8//xwjRozAq6++igYNGmD9+vVwcnICIJ8RVKvV+Oqrr9C/f3889NBDGDduHBo2bIirV68iOjoa7u7uhjkwluRT1Xbv3o05c+Zg4MCB6Ny5M1xdXfHXX39h1apVyM/PFyZjb9myBePGjcPq1asN9zcrybp169CvXz8MGjQI/fv3R+/evVGnTh3DndL3799vuDVFaSzJ1ZLj01zm3jYhNzcXXbp0QefOndGvXz/4+fkhIyMDW7duxYEDBzBo0CC0bdsWAGBvb4/Nmzejd+/e6NatG4YNG4auXbvC3t7ecIf6OnXqlHkvqnr16mHcuHH44osvkJCQgPXr10Oj0Rj+ECxu4MCBeOedd7BhwwZMmzbN4iz27Nlj8lE3gwYNMnse6wOnupcV0oOnrKXgRcy5bYIkSdLly5el0aNHS15eXpKjo6PUpEkTafLkyYYl0WV93ueffy4FBwdL9vb2Uv369aVJkyYJS7r/+usv6fnnn5eaNm0qOTk5SXXr1pUiIiKk3377TdhOYmKi1K1bN8nZ2VkCYLh9QdGtCdLS0kx+/rp166QmTZpIDg4OUmhoqLRz506TtznQarXShx9+KAUHB0sODg6Sl5eX1L9/f2FpdkltkCRJ2rVrl9SqVSvJwcFBCgoKktatW1fibRMmT55ssq03btyQJk+eLPn5+Un29vaSj4+P1KtXL2nlypVG2zDntgBFtm7dKvXp00eqU6eO5OjoKDVr1kyaPn26ycy6d+8uPfTQQya3c+jQIalz586Ss7Oz5OvrK73xxhvSzp07hVsclLYNU7n/9ddfUmRkpOTs7Cx5eXlJ06dPlzZt2iQBkI4cOSLUjY+PlwYPHizVq1dPcnR0lBo3biwNGzZM2rNnj1DP3HxWrVolAZDi4uJMvl7RO6X/9ddf0uzZs6XOnTtL3t7ekp2dneTl5SVFRkZKe/fuFeoWfR+tXr26zHZL0r3bJCxdulQKDw+X3N3dJTs7O8nHx0d6/PHHpfXr1wu3MCmrrebmas7xWdJnFd0e4P7+mXvbhMLCQuk///mPNGjQIKlx48aSo6Oj5OLiIrVt21b68MMPhdszFPn777+l2bNnS61bt5ZcXFwkJycnqVWrVtLbb78tXb9+3VDP1J3Si1y8eFHSaDTSs88+K9WrV0969NFHS21nYGCg1LZt21JzKK4ol5L+ffvtt6W+/0GmkqRKmJFHRKRgS5cuxWuvvYb//e9/aNiwYZV9zqeffopXX30VFy5cKPdNWonIOjiHiojoPnl5ecLXd+/exb///W80b968SgdTAHD8+HHUqlXLsICCiGwH51AREd1n8ODB8Pf3R2hoKDIzM7Fu3TokJiZa9Hw0S23atAn79u3D+vXr8cILL/C+QEQ2iJf8iIjus3TpUnz11VdITk6GTqdDy5Yt8cYbb+CZZ56pss8MDAxEdnY2nnrqKSxduhS1atWqss8ioqrBARURERFRBXEOFREREVEFcUBFREREVEGc+VgN9Ho9rl27Bjc3t3I9roSIiIiqnyRJyM7Ohq+vL9Tq0s9BcUBVDa5du2b2QzGJiIioZrly5QoaNWpUah0OqKpB0UNgr1y5And3dyu3xjxarRaxsbEICwvjEm4wj+KYh4xZiJiHiHnIbDGLrKws+Pn5mXyYe3G20SMbV3SZz93d3WYGVIWFhcjNzYWbm5tZz1VTOuYhYh4yZiFiHiLmIbPlLMyZrsNJ6UREREQVxAEVERERUQVxQEUmaTQahIaGQqPRWLspNQLzEDEPGbMQMQ8R85ApPQveKb0aZGVlwcPDA5mZmTYzh4qIiOhBZ8nvb56hIpO0Wi327t0LrVZr7abUCMxDxDxkzELEPETMQ6b0LDigIpOKbmbGE5j3MA8R85AxCxHzEDEPmdKz4ICKiIiIqII4oCIiIiKqIE5Krwa2OCldr9cjPT0dnp6eZT6/6EHAPETMQ8YsRMxDxDxktpiFJb+/OaCqBrY4oCIiInrQcZUfVVhhYSG2b9+OwsJCazelRmAeIuYhYxYi5iFiHjKlZ8EBFZVIqUtby4t5iJiHjFmImIeIeciUnAUHVEREREQVZGftBlDVyM3NRWJiolCWl5eH5ORkBAQEwNnZ2eg9wcHBcHFxqa4mEhERKQYnpVcDa0xKj4uLQ1hYmEXviY2NRbt27QDIN2Bzc3ODSqWqiibaFOYhYh4yZiFiHiLmIbPFLCz5/c0zVAoVHByM2NhYoSwhIQGjRo3CunXrEBISYvI99zN1FutBxjxEzEPGLETMQ8Q8ZErOgnOoFMrFxQXt2rUT/hUNokJCQoxea9eunXC5T6vVYseOHYqeQGgJ5iFiHjJmIWIeIuYhU3oWPEOlACkpKUhPTy+zXkJCgvDf0mi1WqSlpVW4bURERA8CDqhsXEpKCkKCg5Cbd9fs94waNcqsek6O9oiIiEDTpk3L2zwiIqIHAgdUNi49PR25eXex7iUgxLf0unkFQHI6EOAJODuUXjfhGjDqi0LcunWLAyoiIqIycJVfNajKVX5Fq/li3wPaBVbidi8BYTOBP/74w+LVgkokSRK0Wi3s7OxsZnVKVWIeMmYhYh4i5iGzxSz46BmiKpCXl2ftJtQozEPGLETMQ8Q8ZErOggMqKpVOp7N2E2oErVaL6Ohoxa5OsRTzkDELEfMQMQ+Z0rPggIqIiIiogjigIiIiIqogDqiIzGRnx0Wx92MeMmYhYh4i5iFTchbK7RlVCiUf/Jawt7dHZGSktZtRYzAPGbMQMQ8R85ApPQueoaJS6fV6azehRtDr9bh58ybz+AfzkDELEfMQMQ+Z0rPggIpKpdQD31I6nQ4xMTFc9fgP5iFjFiLmIWIeMqVnwQEVERERUQVxQEVERERUQRxQEZlBpVLBzc3NZh6XUNWYh4xZiJiHiHnIlJ4Fl3BRqYqv8svNzUViYqJRvby8PCQnJyMgIADOzs7Ca8HBwXBxcanSdlY1Ozs79OzZ09rNqDGYh4xZiJiHiHnIlJ4FB1RUquKT0hMTEy1+WHJsbCzatWtXmc2qdnq9HleuXIGfnx/Uap7YZR4yZiFiHiLmIVN6FhxQUamKD6iCg4MRGxtrVC8hIQGjRo3CunXrEBISYvQeW6fT6XDixAn4+voq8geBpZiHjFmImIeIeciUngUHVGQRFxeXUs82hYSE2PzZKCIiIktxQEWlSkxMNOtu6QkJCcJ/y+Lp6Ql/f/8KtY2IiKim4ICKTLqeAUAFjBkzxqL3jRo1yqx6Ts5OSEpMsplBlUqlgpeXl2JXp1iKeciYhYh5iJiHTOlZcEBFJmXnAm3rq4GeAGqb8QYdgGwAbgA0ZdTNALC3ALdS/2czAyo7Ozt06dLF2s2oMZiHjFmImIeIeciUngUHVGSSu50acS+6WvYmHzPreQFoDiTkpACwjW8unU6H8+fPo3nz5tBoyhoxKh/zkDELEfMQMQ+Z0rPggMrGqbR30dZHDWcJwN3K264awLObcpHcTY+7XpW3XQBAGoDNwNcDbOPsFHBvtWNSUhKaNm2qyB8ElmIeMmYhYh4i5iFTehYcUNk4p5yUe2eStABSKm+7/eoA/YYA7c7nIF5rxgOStbh3Ka82yj6q0gGkApKdUwVbSUREVDNwQGXj7rr6o92/c7B+MhDiW3nbPZAEvPotkJimvzdYqmROzk7w9PSs/A0TERFZgc0NqJYvX44PP/wQqampaNOmDT777DN07NixxPo//vgjZs2aheTkZDRv3hwffPABBgwYYHhdkiTMmTMH//nPf5CRkYGuXbviyy+/RPPmzQEAycnJmD9/Pvbu3YvU1FT4+vpi1KhReOedd+Dg4FDl/S2LZOeE+FQ98lQAKvGETy13IP46sHbtWjz00ENl1i/txp6m2NptE9RqNfz9/RV5M7ryYB4yZiFiHiLmIVN6FjY1oPr+++8xbdo0rFixAp06dcLSpUvRt29fJCUlwdvb26j+4cOHMWLECCxatAiPP/44oqKiMGjQIMTFxaFVq1YAgMWLF+PTTz/FmjVrEBgYiFmzZqFv3744d+4cnJyckJiYCL1ej3//+99o1qwZzpw5gwkTJuDOnTv46KOPqjuCavfQQw9ZdKNOpd7YU6PRoG3bttZuRo3BPGTMQsQ8RMxDpvQsbGqY+PHHH2PChAkYN24cWrZsiRUrVsDFxQWrVq0yWX/ZsmXo168fXn/9dYSEhGD+/Plo164dPv/8cwD3zk4tXboUM2fOxJNPPomHH34Ya9euxbVr17B161YAQL9+/bB69Wo89thjaNKkCQYOHIgZM2Zg8+bN1dVtq9LpdNZuQo2g0+kQHx/PPP7BPGTMQsQ8RMxDpvQsbGZAVVBQgNjYWPTu3dtQplar0bt3b8TExJh8T0xMjFAfAPr27Wuof+nSJaSmpgp1PDw80KlTpxK3CQCZmZmoW7duRbpjMyRJEr7Ozc1FXFyc0b/775Re/LXc3FxrNL1S6fV6pKSkGD3b8EHFPGTMQsQ8RMxDpvQsbOaSX3p6OnQ6HerXry+U169fH4mJiSbfk5qaarJ+amqq4fWispLqFHfhwgV89tlnpV7uy8/PR35+vuHrrKwsAEBhYSEKCwsB3BsMajQa6HQ64eAqKtdqtcJgRqPRQK1WG5VXx4Gp1+sNf1GcOXMGnTp1KrGuqTulHzt2DB06dDC7T0XlRVkVKXoEjlarNavc3t5eaDtw7069dnZ2JZaXtD+K6hbff7bep/Iee/f3Qyl9ur/ckj6ZysTW+2Sq3Nw+3f85SunT/eXl6dP9uSilT+XZT0X/lSQJkiTZTJ/MZTMDqprg6tWr6NevH4YOHYoJEyaUWG/RokWYN2+eUfmuXbvg4uICAPD390fbtm1x6tQppKTI9zsICgpCcHAwjh07hrS0NEN5aGgoGjdujP379yM7O9tQXqtWrcroWqmuXLmCEydOALg3WFy1ahXatGmDS5cu4fLlywDunUHU6XTo0aMHLl++LAxIi75JzO1TeHg4vL29sWvXLuFgjoiIgLOzM3bs2CG0b8CAAcjLy0N0dLTwmZGRkUhPTxfONrq5uaFnz55CnwDAy8sLXbp0wfnz55GUlGQoL9pPZ8+eBQDs3r0bgOX7qSb2qaLHXhEl9ak8+yknJweAfGwooU8V3U9FlNSn8u6nuLg4APLxoYQ+VXQ/5eTkwM3NzSb6dOjQIZhLJRW/plNDFRQUwMXFBRs3bsSgQYMM5WPGjEFGRgZ++ukno/f4+/tj2rRpmDp1qqFszpw52Lp1K06ePIm//voLTZs2RXx8PEJDQw11unfvjtDQUCxbtsxQdu3aNfTo0QOdO3fGN998U+oqBVNnqPz8/JCeng53d3cAlfcXwMmTJ9GhQwfEvge0Cyw9Q0vEXQLCZt4bBIWFhSn6L2pz+lRQUIDz588bbkinhD5VZD/pdDpcunQJLVq0MDpLaqt9ur/ckv2k0+mMblZo630yVW5un3Q6Hf766y8EBQUBgCL6dH+5pfspPz8fFy5cMBwfSuhTefeTTqfDxYsXERQUJJy9q8l9un37NurVq4fMzEzD7++S2MwZKgcHB4SFhWHPnj2GAZVer8eePXswZcoUk+8JDw/Hnj17hAHV7t27ER4eDgAIDAyEj48P9uzZYxhQZWVl4ejRo5g0aZLhPVevXkVERATCwsKwevXqMpd8Ojo6wtHR0ajc3t4e9vb2QlnRN1hxRQdWWeVVvfy06KAy9Tkltb2ifSpSPKvylJfUdkv75ODgYPL2Ebbcp4rsJ3t7e8PtMUq647Gt9amsNpZUbmdnZ/LYsOU+lVRuTp/s7e3RsmVLw2tK6NP9LN1Pjo6OFv3ssIU+lXc/2dvbC1nYcp9MsZlJ6QAwbdo0/Oc//8GaNWuQkJCASZMm4c6dOxg3bhwAYPTo0Xj77bcN9V999VX8+uuvWLJkCRITEzF37lz88ccfhgGYSqXC1KlT8d577+G///0vTp8+jdGjR8PX19cwaLt69Sp69OgBf39/fPTRR0hLS0NqamqJc6yUxpLrx0qm1Wpx+PBh5vEP5iFjFiLmIWIeMqVnYTNnqADgmWeeQVpaGmbPno3U1FSEhobi119/NUwqT0lJEUaqXbp0QVRUFGbOnIn/+7//Q/PmzbF161bDPagA4I033sCdO3cwceJEZGRk4JFHHsGvv/4KJ6d7d8ncvXs3Lly4gAsXLqBRo0ZCe2zkailVAkmSkJaWxn3+D+YhYxYi5iFiHjKlZ2FTAyoAmDJlSomX+Pbt22dUNnToUAwdOrTE7alUKrz77rt49913Tb4+duxYjB07tjxNJSIiogeETV3yIyIiIqqJOKCiUin1mUuW0mg0CA0NLXEC9oOGeciYhYh5iJiHTOlZ2NwlP6peHFDdo1ar0bhxY2s3o8ZgHjJmIWIeIuYhU3oW/G1JpVLqagxLabVa7N27l3n8g3nImIWIeYiYh0zpWXBARWQGSZKQnZ2t2NUplmIeMmYhYh4i5iFTeha85KcQCdfKrpNXACSnAwGegLNDxbdHRERE93BAZeM8PT3h4uyEUV/crfRtOznao169epW+XSIiIqWxmWf52bKsrCx4eHiY9Syg8khJSUF6enqZ9RISEjBq1CisW7fO8NiQkuj1eqjVaoSGhnJiOu7lkZ6eDk9PT+YB5nE/ZiFiHiLmIbPFLCz5/c0zVArg7+8Pf39/s+uHhISgXbt2Vdgi5VGr1fD29rZ2M2oM5iFjFiLmIWIeMqVnYRtDRKp2hYWF2L59u9GTzx9UzEPEPGTMQsQ8RMxDpvQsOKCiEil1aWt5MQ8R85AxCxHzEDEPmZKz4ICKiIiIqII4oCIiIiKqIA6oyCQ7OztERETAzo7rFgDmURzzkDELEfMQMQ+Z0rPggIpK5OzsbO0m1CjMQ8Q8ZMxCxDxEzEOm5Cw4oCKTtFotduzYoegJhJZgHiLmIWMWIuYhYh4ypWfBARURERFRBXFARURERFRBHFARERERVRCf5VcNqvpZfuaKi4tDWFgYYmNjy3z0jCRJ0Gq1sLOzg0qlqqYW1lzMQ8Q8ZMxCxDxEzENmi1lY8vubZ6ioRHl5edZuQo3CPETMQ8YsRMxDxDxkSs6CAyoySavVIjo6WrGrMSzFPETMQ8YsRMxDxDxkSs+CAyoiIiKiCuKAioiIiKiCOKCiEin18QDlxTxEzEPGLETMQ8Q8ZErOgqv8qoEtrvIjIiJ60HGVH1WYXq/HzZs3odfrrd2UGoF5iJiHjFmImIeIeciUngUHVGSSTqdDTEwMdDqdtZtSIzAPEfOQMQsR8xAxD5nSs+CAioiIiKiCOKAiIiIiqiAOqMgklUoFNzc3m3k8QFVjHiLmIWMWIuYhYh4ypWfBVX7VgKv8iIiIbI8lv7+Ve0OIB1xubi4SExOFsoSEBOG/xQUHB8PFxQXAvdUYV65cgZ+fH9RqnshkHiLmIWMWIuYhYh4ypWfBAZVCJSYmIiwszORro0aNMll+/5krnU6HEydOwNfXV5EHvqWYh4h5yJiFiHmImIdM6VlwQKVQwcHBiI2NFcry8vKQnJyMgIAAODs7m3wPERERWY4DKoVycXExOU+qa9euVmgNERGRsinvnBtVCpVKBS8vL8WuxrAU8xAxDxmzEDEPEfOQKT0LrvKrBjVllR8RERGZj8/yowrT6XRITExU7CMCLMU8RMxDxixEzEPEPGRKz4IDKjJJr9cjKSlJsQ+xtBTzEDEPGbMQMQ8R85ApPQsOqIiIiIgqiAMqIiIiogrigIpMUqvV8Pf3V+TN18qDeYiYh4xZiJiHiHnIlJ4FV/lVA67yIyIisj18lh9VmE6nw6lTp/Dwww9Do9FYuzkVYuq5hkDpd46//7mGgLLyqAzMQ8YsRMxDxDxkSs+CAyoySa/XIyUlBa1atbL5A7+05xqW5P7nGgLKyqMyMA8ZsxAxDxHzkCk9Cw6oSHFSUlKQnp5u+DovLw/r1q0zqnfp0iXMmjUL8+fPR2BgoPBaXl4e4uLiDF9rtVqkpaVVXaOJiMimcUBFipKSkoKQ4CDk5t01+z2zZs0yq56Toz0iIiLQtGnT8jaPiIgUigMqMkmtViMoKMjmVmOkp6cjN+8u1r0EhPiWXjevAEhOBwI8AWeH0usmXANGfVGI27dvc0AF2z0+qgKzEDEPEfOQKT0LDqjIJI1Gg+DgYGs3o9xCfIF2gWXX6xpk2XaVeN2/PGz9+KhMzELEPETMQ6b0LJQ5TKQK02q1OHz4MLRarbWbUqMwj3t4fMiYhYh5iJiHTOlZcEBFJkmShLS0NPA2ZWQKjw8ZsxAxDxHzkCk9Cw6oiIiIiCqIAyoiIiKiCuKkdDJJo9EgNDSUk7CLKb46xdRd2Eu7AztgfBd2W8TjQ8YsRMxDxDxkSs+CAyoySa1Wo3HjxtZuRo2TlJQkDKoSEhIwatQoi7axbt06hISECGWenp7w9/evlDZWBx4fMmYhYh4i5iFTehY2d8lv+fLlCAgIgJOTEzp16oRjx46VWv/HH39EcHAwnJyc0Lp1a+zYsUN4XZIkzJ49Gw0aNICzszN69+6N8+fPC3UWLFiALl26wMXFBbVr167sLtVIWq0We/fuVexqDEvdyADaNlBjyYzReCGyg+Hfkhmj0dZHbdG/4tt4IbIDnukejCt/JVm7m2bj8SFjFiLmIWIeMqVnYfEZqvz8fBw9ehSXL19Gbm4uvLy80LZtW6NHd1SF77//HtOmTcOKFSvQqVMnLF26FH379kVSUhK8vb2N6h8+fBgjRozAokWL8PjjjyMqKgqDBg1CXFwcWrVqBQBYvHgxPv30U6xZswaBgYGYNWsW+vbti3PnzsHJyQkAUFBQgKFDhyI8PBxff/11lfezJpAkCdnZ2Ta3GkOlvYu2Pmo4SwDMv1l6mdx0asRNrNrLdAnJ8UATC2+MZSW2enxUBWYhYh4i5iFTehZmD6gOHTqEZcuW4eeff0ZhYSE8PDzg7OyM27dvIz8/H02aNMHEiRPxr3/9C25ublXS2I8//hgTJkzAuHHjAAArVqzA9u3bsWrVKrz11ltG9ZctW4Z+/frh9ddfBwDMnz8fu3fvxueff44VK1ZAkiQsXboUM2fOxJNPPgkAWLt2LerXr4+tW7di+PDhAIB58+YBAL755psq6RdVHqecFMS96ApoAaRU3nYfcb/332c35SIxXV95G/6Ho5MjfhjdttK3S0RE1cOsAdXAgQMRFxeHZ599Frt27UL79u2FybZ//fUXDhw4gO+++w4ff/wx1q5diz59+lRqQwsKChAbG4u3337bUKZWq9G7d2/ExMSYfE9MTAymTZsmlPXt2xdbt24FcO/huKmpqejdu7fhdQ8PD3Tq1AkxMTGGARXZjruu/mj37xysn1z2o2cskXANGLkcmL7o35jxcDtDeV5eHv766y9Mnz4djRo1wmuvvYbLly8bHrrcuHFjfPLJJ/jf//6HJUuWQK1Wm5ys7unpCT8bmkNFREQiswZUkZGR2LRpE+zt7U2+3qRJEzRp0gRjxozBuXPncP369UptJHDvGW06nQ7169cXyuvXr2+0yqpIamqqyfqpqamG14vKSqpTHvn5+cjPzzd8nZWVBQAoLCxEYWEhgHuDQY1GA51OB71ePuNRVK7VaoXTohqNBmq1usTyou0WsbO7t2uLX6suqdze3h56vR46nQ4AoNfr0bFjR2g0GqEcAFQqFezs7EpsuzX7VAg7xKfqkacC4IRKk6cC4lOBpsGt0Lp1a0Ofjh07htGjRwMA0tLSDP8PGD90uei1o0ePonXr1sZtv6+/5u4nQN4f1bmf9Ho9OnfuDI1GU+nHnrX6dH+5JX1Sq9Xo0KED9Hq94X223idT5eb26f6fHUrp0/3llvZJkiTh+FBCn8q7n/R6PTp06AC1Wg1JkmymT+Yya0D14osvmr3Bli1bomXLlmbXV6JFixYZLhPeb9euXYbl8v7+/mjbti1OnTqFlBT52lRQUBCCg4Nx7NgxpKWlGcpDQ0PRuHFj7N+/H9nZ2Yby8PBweHt7Y9euXcKOj4iIgLOzs9Ek/AEDBiAvLw/R0dGGMjs7O0RGRiI9PV042+fm5oYGDRrg8uXLOHHihKHcy8sLXbp0wfnz55GUJE+krgl9unjxolHulenw4cO4ceOGoU85OTkYOXIk1q9fj4ULF+Lhhx9G7dq1sX37dri7u8PBwQF3797FO++8gwULFqBWrVpITk42/NFRWfupZ8+euHLlSrXvJ7VajV9++aVKjj1r9ak8x97x48cV16eK7qcGDRogMTFRUX0qz376448/FNcnJe6nkvp06NAhmEslWTg7rEmTJjh+/Djq1asnlGdkZKBdu3b466+/LNmc2QoKCuDi4oKNGzdi0KBBhvIxY8YgIyMDP/30k9F7/P39MW3aNEydOtVQNmfOHGzduhUnT57EX3/9haZNmyI+Ph6hoaGGOt27d0doaCiWLVsmbO+bb77B1KlTkZGRUWpbTZ2h8vPzQ3p6Otzd703Gqel/qRUWFmLv3r3o27evoZ1FavJf1PHx8ejUqRNi3zPv4cjmirsEhM0EDh48iI4dOwp92rdvH/r06YMDBw6gS5cuRn06cuQIunXrhujoaHTt2tXiPt3P2n+pFZXff3wUZ6t9ur/ckmOvsLAQu3btQs+ePQ1n8W29T6bKze3T/ceGWq1WRJ/uL7d0P+Xl5WHPnj2G40MJfSrvfio6Nh577DHY29vbRJ9u376NevXqITMz0/D7uyQWr/JLTk4WGl8kPz8fV69etXRzZnNwcEBYWBj27NljGFDp9Xrs2bMHU6ZMMfme8PBw7NmzRxhQ7d69G+Hh4QCAwMBA+Pj4YM+ePYYBVVZWFo4ePYpJkyaVu62Ojo5wdHQ0Kre3tze6bKrRaEze5KzowDK3vKTLsZaUq9Vq4R5LRfu5eHmRktpuzT6VtK3KUnwf2tnZISIiAgEBAVi8eLFhfl5RO/R6PT788EMEBgbi0UcfLfGGdhXZT2WVV9V+Kjo+quLYK6u8Jh17KpUKOp3O6Niw5T6VVG5un4qODSX1qUh5+mTq+LD1PpliTp90Oh1UKhVUKpVN98lkXXMr/ve//zX8/86dO+Hh4WH4WqfTYc+ePQgICDD7g8tj2rRpGDNmDNq3b4+OHTti6dKluHPnjmHV3+jRo9GwYUMsWrQIAPDqq6+ie/fuWLJkCSIjI7Fhwwb88ccfWLlyJYB7PwinTp2K9957D82bNzfcNsHX11c4C5aSkoLbt28jJSUFOp3OcHqxWbNmcHV1rdI+U82n0WiwZMkSPP300xg0aBDefvtttGrVCmfOnMGiRYuwbds2bNy4UbF3ByYiIgsGVMUvs93P3t4eAQEBWLJkSaU1zJRnnnkGaWlpmD17NlJTUxEaGopff/3VMKk8JSVFGKl26dIFUVFRmDlzJv7v//4PzZs3x9atWw33oAKAN954A3fu3MHEiRORkZGBRx55BL/++qvhHlQAMHv2bKxZs8bwddu295a3R0dHo0ePHlXaZ7INgwcPxsaNGzF9+nR06dLFUB4YGIiNGzdi8ODBVmwdERFVNYvnUAUGBuKPP/4wmkNFJcvKyoKHh4dZ12BriqIbsLm5uUGlUlm7OWaLi4tDWFhYlc2h+uOPPxAWFlZiPZ1OhwMHDuD69eto0KBBqZf5bJmtHh9VgVmImIeIechsMQtLfn9bNOGksLAQTZo0MUzSImUz9WBfW5Fwrew6eQVAcjoQ4Ak4O1R8e8C9y38PyllLWz4+KhuzEDEPEfOQKTkLiwZU9vb2OHXqVFW1hWoQrVaLHTt2YMCAASVOHKyJPD094eLshFFfVOJzZ/7h5Gj/wDzLsSy2enxUBWYhYh4i5iFTehYWL4kaNWoUvv76a7z//vtV0R6iCvH390dCYhLS09PLrJuQkIBRo0Zh3bp1CAkJKbWuVqvFqVOn4M+7mRMRkQkWD6i0Wi1WrVqF3377DWFhYahVq5bw+scff1xpjSMqD39/f4sGPiEhIWjXrl2pdQoLC6vkCQBERKQMFg+ozpw5Y/jl8+effwqv2cokMyIiIqLKZPEqP7Kcra7y02q1sLOzU+xA2bAiMDa2zDNUD0IelmAeMmYhYh4i5iGzxSws+f1tfHtRC1y5cgVXrlypyCaoBsvLy7N2E2oU5iFiHjJmIWIeIuYhU3IWFg+otFotZs2aBQ8PDwQEBCAgIAAeHh6YOXOm0TOIyHZptVpER0db9KRtJWMeIuYhYxYi5iFiHjKlZ2HxHKqXX34ZmzdvxuLFiw3PxIuJicHcuXNx69YtfPnll5XeSCKqPLm5uUhMTBTK8vLykJycjICAAJP3iQkODoaLi0t1NZGIyOZYPKCKiorChg0b0L9/f0PZww8/DD8/P4wYMYIDKqIaJCUlxegWEkW3i7BE8VtLaLVapKWlVUobiYiUwOIBlaOjo8mHIAcGBsLBoYzbTZNNseQp2w8CW8sjJSUFIcFByM2r+E1OTQ3AnBztERERgaZNm1Z4+7bO1o6NqsY8RMxDpuQsLF7l9+677yIxMRGrV6+Go6MjACA/Px/jx49H8+bNMWfOnCppqC2zxVV+DwJLVvnZoqL+rXsJCPEtva4lj+EB7j2KZ9QXUGx2RERAFTzLb/DgwcLXv/32Gxo1aoQ2bdoAAE6ePImCggL06tWrnE2mmkav1yM9PR2enp5Qqyu0GFQRbDmPEF/zHhTdNcjybev1esvfpDC2fGxUBeYhYh4ypWdh1oDKw8ND+HrIkCHC135+fpXXIqoRdDodYmJiMGDAAJs/8E1NwgbuzSW6/7/3Kz4JW0l5VCYOqHhsFMc8RMxDpvQszBpQrV69uqrbQVRlEhMTERYWVuLrpuYH8VKW5UoauHIFIRE9CCyeHXbp0iVotVo0b95cKD9//jzs7e1NTlgnsqbg4GDExsYalZf2iz44OLi6mmfTEhMTDZNMy7N6EDBeQejp6cmHUBORzbF4QDV27Fg8//zzRgOqo0eP4quvvsK+ffsqq21kRSqVCm5ubjbzeIDSuLi4lHi2qWvXrmZtQ0l5VIYbGUDbBmosfXOcUN7Wx/LT+EtmjBa+dnRyxA974uHXpByTuqyAx4aIeYiYh0zpWVi8ys/d3R1xcXFo1qyZUH7hwgW0b98eGRkZldk+ReAqP7KG+GOHMf7JR7F+ctmr/Cx18Iwaj7hX3WW6hG7/RkjP4VW2fSIic1T6Kr/7qVQqZGdnG5VnZmZCp9NZujmqofR6Pa5cuQI/Pz9FTh60lC3m4ZSTgrgXXQEtgJTK3fYj//xceXZTLhLTK3diuqOTI34Y3bZSt1mVbPHYqErMQ8Q8ZErPwuIBVbdu3bBo0SJ899130Gg0AO7N3F+0aBEeeeSRSm8gWYdOp8OJEyfg6+uryAPfUraYx11Xf7T7d06VnKFKuAaMXA5MX/RvzHj43uXUypxD5WdDc6hs8dioSsxDxDxkSs/C4gHVBx98gG7duiEoKAiPPvooAODAgQPIysrC3r17K72BRFQ+kp0T4lP1yFMBcKrcbeepgPhUoHnLhw3z08oz+b/ofVzlR0S2zuIBVcuWLXHq1Cl8/vnnOHnyJJydnTF69GhMmTIFdevWrYo2EpENqIzJ/0REtqpcD9Xx9fXFwoULK7stVIOoVCp4eXkpdjWGpZgHlYTHhoh5iJiHTOlZmDWgSklJsei+MFevXkXDhg3L3SiyPjs7O3Tp0sXazagxbDmPhGtl1ynPs/wAZT/o1Fy2fGxUBeYhYh4ypWdh1k/DDh06YNCgQXjhhRfQoUMHk3UyMzPxww8/YNmyZZg4cSJeeeWVSm0oVS+dTofz58+jefPmhsUHDzJbzMPT0xMuzk4Y9cXdKtm+s5MD6tSpUyXbtiW2eGxUJeYhYh4ypWdh1oDq3LlzWLBgAfr06QMnJyeEhYXB19cXTk5O+Pvvv3Hu3DmcPXsW7dq1w+LFizFgwICqbjdVMb1ej6SkJDRt2lSRB76lbDEPf39/JCQmIT09vcy6RSv0iq+4K4lWq8WpU6fQqFGjymiqTbPFY6MqMQ8R85ApPQuzBlT16tXDxx9/jAULFmD79u04ePAgLl++jLy8PHh6emLkyJHo27cvWrVqVdXtJSIL+Pv7W3S5PiQkxKxnGBYWFuL69esVaRoRkaJYNAHC2dkZTz/9NJ5++umqag8RERGRzVHenbWoUqjVavj7+yvy5mvlwTxEzEPGLETMQ8Q8ZErPwuJn+ZHl+Cw/quni4uIQFhaG2NhYsy75ERE9CCz5/a3MYSJVmE6nQ3x8PJ/P+A/mIWIeMmYhYh4i5iFTeha8iQyZpNfrkZKSglatWilyNYallJRHbm4uEhMThbKEhAThv8UVfzyMkvKoKGYhYh4i5iFTehYWD6j279+PLl26GN3QT6vV4vDhw+jWrVulNY6IKl9iYiLCwsJMvlbSw415KZCIqHQWD6giIiJw/fp1eHt7C+WZmZmIiIhQ7Kk8IqUw9RBjcx5gTEREJbN4QCVJksnn8Ny6dQu1atWqlEaR9anVagQFBSl2NYallJRHSQ8xtuQBxkrKo6KYhYh5iJiHTOlZmL3Kb/DgwQCAn376Cf369YOjo6PhNZ1Oh1OnTiEoKAi//vpr1bTUhnGVHxERke2pklV+Hh4e8PDwgCRJcHNzM3zt4eEBHx8fTJw4EevWratw46lmKJoTp9Vqrd2UGoF5iJiHjFmImIeIeciUnoXZl/xWr14NAAgICMCMGTN4eU/hJElCWloaeJuye5iHiHnImIWIeYiYh0zpWVg8h2rOnDlV0Q4iIiIim2XxzLAbN27gueeeg6+vL+zs7KDRaIR/RERERA8ai89QjR07FikpKZg1axYaNGhgcsUf2T6NRoPQ0FAOkv/BPETMQ8YsRMxDxDxkSs/C4mf5ubm54cCBAwgNDa2iJikPV/kRERHZnip9lp+fn59iJ5SRTKvVYu/evYpdjWEp5iFiHjJmIWIeIuYhU3oWFg+oli5dirfeegvJyclV0ByqKSRJQnZ2NgfP/2AeIuYhYxYi5iFiHjKlZ2HxHKpnnnkGubm5aNq0KVxcXGBvby+8fvv27UprHBEREZEtsHhAtXTp0ipoBhEREZHtsnhSOlnOFiel6/V6pKenw9PTU7HPXbIE8xAxDxmzEDEPEfOQ2WIWlvz+LteA6uLFi1i9ejUuXryIZcuWwdvbG7/88gv8/f3x0EMPlbvhSmWLAyoiIqIHXZWu8vv999/RunVrHD16FJs3b0ZOTg4A4OTJk7yLuoIUFhZi+/btKCwstHZTagTmIWIeMmYhYh4i5iFTehYWD6jeeustvPfee9i9ezccHBwM5T179sSRI0cqtXFkXUpd2lpezEPEPGTMQsQ8RMxDpuQsLB5QnT59Gk899ZRRube3N9LT0yulUURERES2xOIBVe3atXH9+nWj8vj4eDRs2LBSGkVERERkSyyelD5jxgwcPXoUP/74I1q0aIG4uDjcuHEDo0ePxujRozmPygRbnJRedAM2Nzc3Pq8RzKM45iFjFiLmIWIeMlvMokonpS9cuBDBwcHw8/NDTk4OWrZsiW7duqFLly6YOXNmuRtNNY+zs7O1m1CjMA8R85AxCxHzEDEPmZKzsHhA5eDggP/85z+4ePEitm3bhnXr1iExMRHffvutYp8g/SDSarXYsWOHoicQWoJ5iJiHjFmImIeIeciUnoXFd0ov4u/vD39//8psCxEREZFNMmtANW3aNMyfPx+1atXCtGnTSq378ccfV0rDiIiIiGyFWZf84uPjDTfiio+PL/HfiRMnqrKtAIDly5cjICAATk5O6NSpE44dO1Zq/R9//BHBwcFwcnJC69atsWPHDuF1SZIwe/ZsNGjQAM7OzujduzfOnz8v1Ll9+zZGjhwJd3d31K5dG+PHjzfc0JSIiIgIkg3ZsGGD5ODgIK1atUo6e/asNGHCBKl27drSjRs3TNY/dOiQpNFopMWLF0vnzp2TZs6cKdnb20unT5821Hn//fclDw8PaevWrdLJkyelgQMHSoGBgVJeXp6hTr9+/aQ2bdpIR44ckQ4cOCA1a9ZMGjFihNntzszMlABImZmZ5e98NdPr9VJBQYGk1+ut3ZQagXmImIeMWYiYh4h5yGwxC0t+f9vUgKpjx47S5MmTDV/rdDrJ19dXWrRokcn6w4YNkyIjI4WyTp06SS+++KIkSfd2ro+Pj/Thhx8aXs/IyJAcHR2l7777TpIkSTp37pwEQDp+/Lihzi+//CKpVCrp6tWrZrXbVgdUmZmZNnXgVyXmIWIeMmYhYh4i5iGzxSws+f1t1hyqwYMHm33Ga/PmzeU5UVamgoICxMbG4u233zaUqdVq9O7dGzExMSbfExMTYzTnq2/fvti6dSsA4NKlS0hNTUXv3r0Nr3t4eKBTp06IiYnB8OHDERMTg9q1a6N9+/aGOr1794ZarcbRo0dN3jU+Pz8f+fn5hq+zsrIA3HuOUdGlU7VaDY1GA51OB71eL/RJo9FAq9VCuu8WYRqNBmq1usTy4s9GsrO7t2uLr6Yoqdze3h56vR46nc7Q1ujoaAwYMMDQziIqlQp2dnYltr2m9un+tpdUXlLb8/PzER0djT59+sDe3l4RfarIfrr/+Ciuxvap4A7SE2Nw69YtqNUqqFRqZGRk4PTp00J9AMI2TJW3bt0abm5uUKvV0BfkIeXUATRqFQ61Qy1DZllZWTh58mSp27m/XJIktGrVCm5uboa+qtVq1K1bF3VbdALsXYz7VAOPvfuPDbVa/cD8jCip/O7du8LPDiX0qbz7qejY6N+/P+zt7W2mT+Yya0Dl4eFh+H9JkrBlyxZ4eHgYBhmxsbHIyMiwaOBlqfT0dOh0OtSvX18or1+/PhITE02+JzU11WT91NRUw+tFZaXV8fb2Fl63s7ND3bp1DXWKW7RoEebNm2dUvmvXLri43Puh6O/vj7Zt2+LUqVNISUkx1AkKCkJwcDCOHTuGtLQ0Q3loaCgaN26M/fv3Izs721AeHh4Ob29v7Nq1S9jxERERcHZ2NpozNmDAAOTl5SE6OlroT2RkJNLT000OTq9cuSLMj/Py8kKXLl1w/vx5JCUlGcptoU9ubm7o2bOnxX06e/YsAGD37t2K6VNF91MRW+lTweXj6JE0Gw2KtT/8/i90MK14+Snxy1YAcNz4j8lOZW2nePkp0y/vC3oXmS4BAGzj2CvyIP2MKKlPcXFxAOSfHUroU0X3U05ODtzc3GyiT4cOHYK5LL5T+ptvvonbt29jxYoVhvtO6XQ6vPTSS3B3d8eHH35oyebMdu3aNTRs2BCHDx9GeLj8I/CNN97A77//jqNHjxq9x8HBAWvWrMGIESMMZV988QXmzZuHGzdu4PDhw+jatSuuXbuGBg3kH7PDhg2DSqXC999/j4ULF2LNmjXCjgHuPbtw3rx5mDRpktHnmjpD5efnh/T0dMOdVmv6X2qFhYXYvXs3z1Dd91fmzp07eYbqvjNURcdHcTW2TzxDZdb+qIwzVEXHBs9QaZCXl4ddu3bxDBXkY8OWzlDdvn0b9erVM+tO6Rbfh2rVqlU4ePCgcBNPjUaDadOmoUuXLlU2oPL09IRGo8GNGzeE8hs3bsDHx8fke3x8fEqtX/TfGzduCAOqGzduIDQ01FDn5s2bwja0Wi1u375d4uc6OjrC0dHRqNze3h729vZCmUajMXlD1KIDy9zy4tstT7larTb8wL//s4qXFymp7TW5T2WVl9YnOzs7o31o632qyH4q+tpm+mTngQbt+hmfoargifXCwkJc1tdHyGOPGfWt06CKbbs0NfnYK2rDg/YzoqQ+WfKzw1b6ZIo5fbKzs4NKpYJKpbLpPpli8Z3StVqtyUtsiYmJRn91VSYHBweEhYVhz549hjK9Xo89e/YIZ6zuFx4eLtQH7p12LaofGBgIHx8foU5WVhaOHj1qqBMeHo6MjAzExsYa6uzduxd6vR6dOhmd0FcMe3t7REZGlnjAP2iYh4h5yJiFiHmImIdM6VlYfIZq3LhxGD9+PC5evIiOHTsCAI4ePYr3338f48aNq/QG3m/atGkYM2YM2rdvj44dO2Lp0qW4c+eO4XNHjx6Nhg0bYtGiRQCAV199Fd27d8eSJUsQGRmJDRs24I8//sDKlSsB3Ds1OHXqVLz33nto3rw5AgMDMWvWLPj6+mLQoEEAgJCQEPTr1w8TJkzAihUrUFhYiClTpmD48OHw9fWt0v5ak16vR3p6Ojw9PU2O/h80zEPEPGTMQsQ8RMxDpvgsLF1CqNPppA8++EDy9fWVVCqVpFKpJF9fX+mDDz6QtFqtpZuz2GeffSb5+/tLDg4OUseOHaUjR44YXuvevbs0ZswYof4PP/wgtWjRQnJwcJAeeughafv27cLrer1emjVrllS/fn3J0dFR6tWrl5SUlCTUuXXrljRixAjJ1dVVcnd3l8aNGydlZ2eb3WZbvG1CQUGBtHXrVqmgoMDaTakRmIeIeciYhYh5iJiHzBazqPTbJtxPrVbjjTfewBtvvGG4HUBZE7Uq05QpUzBlyhSTr+3bt8+obOjQoRg6dGiJ21OpVHj33Xfx7rvvllinbt26iIqKsritRERE9GAo98ORgeodSBERERHVVOUaUG3cuBE//PADUlJSUFBQILxWdM8Nsm0qlQpubm5QqVTWbkqNwDxEzEPGLETMQ8Q8ZErPwuL7UH366ad45513MHbsWKxcuRLjxo3DxYsXcfz4cUyePBkLFiyoqrbarKysLHh4eJh1HwsiIiKqGSz5/W3xNPsvvvgCK1euxGeffQYHBwe88cYb2L17N1555RVkZmaWu9FUs+j1ely+fLlKb4VhS5iHiHnImIWIeYiYh0zpWVg8oEpJSUGXLl0AAM7OzobbzD/33HP47rvvKrd1ZDU6nQ4nTpwQ7kD7IGMeIuYhYxYi5iFiHjKlZ2HxgMrHxwe3b98GcO/ZOEeOHAFw70HDFl49JCIiIlIEiwdUPXv2xH//+18A927y+dprr6FPnz545pln8NRTT1V6A4mIiIhqOotX+a1cudJw/XPy5MmoV68eDh8+jIEDB+LFF1+s9AaSdahUKnh5eSl2NYalmIeIeciYhYh5iJiHTOlZWLTKT6vVYuHChXj++efRqFGjqmyXonCVHxERke2pslV+dnZ2WLx4MbRabYUaSDWfTqdDYmKiYicPWop5iJiHjFmImIeIeciUnoXFc6h69eqF33//vSraQjWIXq9HUlKSYpe3Wop5iJiHjFmImIeIeciUnoXFc6j69++Pt956C6dPn0ZYWBhq1aolvD5w4MBKaxwRERGRLbB4QPXSSy8BAD7++GOj11QqlWJP5RERERGVxOIBlVJP1ZFIrVbD398farXFV4UViXmImIeMWYiYh4h5yJSehcXP8iPLcZUfERGR7bHk97fFZ6g+/fRTk+UqlQpOTk5o1qwZunXrBo1GY+mmqQbR6XQ4deoUHn74Ye5LMI/imIeMWYiYh4h5yJSehcUDqk8++QRpaWnIzc1FnTp1AAB///03XFxc4Orqips3b6JJkyaIjo6Gn59fpTeYqoder0dKSgpatWqlyAPfUsxDxDxkzELEPETMQ6b0LCy+kLlw4UJ06NAB58+fx61bt3Dr1i38+eef6NSpE5YtW4aUlBT4+Pjgtddeq4r2EhEREdU4Fp+hmjlzJjZt2oSmTZsaypo1a4aPPvoIQ4YMwV9//YXFixdjyJAhldpQIiIioprK4jNU169fN3mndK1Wi9TUVACAr68vsrOzK946shq1Wo2goCDFrsawFPMQMQ8ZsxAxDxHzkCk9C4t7FRERgRdffBHx8fGGsvj4eEyaNAk9e/YEAJw+fRqBgYGV10qqdhqNBsHBwYq8zl0ezEPEPGTMQsQ8RMxDpvQsLB5Qff3116hbty7CwsLg6OgIR0dHtG/fHnXr1sXXX38NAHB1dcWSJUsqvbFUfbRaLQ4fPsznNv6DeYiYh4xZiJiHiHnIlJ6FxXOofHx8sHv3biQlJSEpKQkAEBQUhKCgIEOdiIiIymshWYUkSUhLSwNvU3YP8xAxD5mSssjNzUViYqJReV5eHpKTkxEQEABnZ2fhteDgYLi4uBi+VlIelYF5yJSehcUDqiLFB1FERGTbEhMTERYWZtF7YmNj0a5duypqEZHtKPeAioiIlCU4OBixsbFG5QkJCRg1ahTWrVuHkJAQo/cQEQdUVAKNRoPQ0FDFTh60FPMQMQ+ZkrJwcXEp9WxTSEhImWejlJRHZWAeMqVnwQEVmaRWq9G4cWNrN6PGYB4i5iFjFiLmIWIeMqVnocybQVCFabVa7N27V7GrMSzFPETMQ8YsRMxDxDxkSs+iXAOqAwcOYNSoUQgPD8fVq1cBAN9++y0OHjxYqY0j65EkCdnZ2YpdjWEp5iFiHjJmIWIeIuYhU3oWFg+oNm3ahL59+8LZ2Rnx8fHIz88HAGRmZmLhwoWV3kAiIiKims7iOVTvvfceVqxYgdGjR2PDhg2G8q5du+K9996r1MYREVHVSElJQXp6ull1ExIShP+WRqvVIi0trUJtI7JFFg+okpKS0K1bN6NyDw8PZGRkVEabqAbQaDQIDw9X7GoMSzEPEfOQ2WIWKSkpCAkOQm7eXYveN2rUKLPquTg7onfv3ggICChH65TFFo+PqqL0LMp1p/QLFy4YfaMcPHgQTZo0qax2kZWp1Wp4e3tbuxk1BvMQMQ+ZLWaRnp6O3Ly7WPcSEOJbdv28AiA5HQjwBJwdSq+bcA0Y9UU+bt++zQEVbPP4qCpKz8LiAdWECRPw6quvYtWqVVCpVLh27RpiYmIwY8YMzJo1qyraSFZQWFiIXbt24bHHHoO9vb21m2N1zEPEPGS2nEWIL9DOzOfYd7XwwRhKXcllKVs+Piqb0rOweED11ltvQa/Xo1evXsjNzUW3bt3g6OiIGTNm4OWXX66KNpKV8AeiiHmImIeMWZTN1HMCS3tGIGD8nEBbxeNDpuQsLB5QqVQqvPPOO3j99ddx4cIF5OTkoGXLlnB1da2K9hERkQLwOYGkdOW+U7qDgwPc3Nzg5ubGwRQRkQ1Rae+irY8azhIAy+all8lFD4xopUZ6/A4k5PxlKFfn52P7V+Ktda5du4YvvvgCL730Enx9jSdzqW+cRsLeP4Wy2rVro0GrboCD7Z+5ImVRSRbeYUur1WLevHn49NNPkZOTAwBwdXXFyy+/jDlz5ijyumhFZWVlwcPDA5mZmXB3d7d2c8xSdAM2Nzc3qFQqazfH6piHiHnIbDGLhL0bELL/RWs3o9yuD/weDdr1s3YzzGKLx0dVscUsLPn9bfEZqpdffhmbN2/G4sWLER4eDgCIiYnB3LlzcevWLXz55ZflazXVOKbmNDzImIeIechsLYu7rv5o9+8crJ9s3io/SxxKApbvUCM5Q4+7VTBdxtHJET+MNnMmfQ1ha8dHVVJyFhYPqKKiorBhwwb079/fUPbwww/Dz88PI0aM4IBKIbRaLXbs2IEBAwbwrCOYR3HMQ2aLWUh2TohP1SNPBcCpcrft7A58d0aPNWvWoFWrVobyhIQEs+9jVWTdunUICQkRyjw9PeHn718pba0Otnh8VBWlZ2HxgMrR0dHkvUUCAwPh4FDGDUqIiOiBEBwcLEwoDw4ORmxsrFCnpq/ye5BXJpLlLB5QTZkyBfPnz8fq1avh6OgIAMjPz8eCBQswZcqUSm8gERHZPhcXF5Mr9rp27WqF1piHKxPJEhYPqOLj47Fnzx40atQIbdq0AQCcPHkSBQUF6NWrFwYPHmyou3nz5sprKRERVaqEa+bVs/RO6Uph6qxa0aVLU5cji95DDyaLV/mNGzfO7LqrV6+2uEFKZKur/LRaLezs7GxmNUZVYh4i5iGzxSzK+yw/c7k4O+FcQiIaN25cJdu3pri4OISFhZl9JsoWj4+qYotZVOkqPw6SHhx5eXlwc3OzdjNqDOYhYh4yW8vC398fCYlJSE9PF8qL5gcVd+nSJcyaNQvz589HYKC4wq74XCJJkuDk5AR/G5o4XtVs7fioSkrOotw39kxLS0NSUhIAICgoCF5eXpXWKLI+rVaL6Ohoxa7GsBTzEDEPma1m4e/vbzToiYuLK3UlnqnntRY/U1NYWIgdO3agRYsWNpVHVbHV46MqKD0LiwdUd+7cwcsvv4y1a9dCr9cDADQaDUaPHo3PPvuMqxuIiGyUqTlDQOkr2zhniOgeiwdU06ZNw++//46ff/7ZsDrj4MGDeOWVVzB9+nTeh4qIyEaVtBIPqNmr8YhqAosHVJs2bcLGjRvRo0cPQ9mAAQPg7OyMYcOGcUClIHZ25b4irEjMQ8Q8ZMxCxDxEzEOm5CwsXuXn4uKC2NhYo+WiZ8+eRceOHXHnzp1KbaAS2OIqPyIiMmbpKj+ybZb8/lZbuvHw8HDMmTMHd+/Ky23z8vIwb948w7P9yPbp9XrcvHnTME/uQcc8RMxDxixEzEPEPGRKz8LiAdWyZctw6NAhNGrUCL169UKvXr3g5+eHw4cPY9myZVXRRrICnU6HmJgY6HQ6azelRmAeIuYhYxYi5iFiHjKlZ2HxxcxWrVrh/PnzWL9+veEZRyNGjMDIkSMV/RRpIiJSrpSUFKP7cpmSkJAg/LcsWq0WaWlpFWob2YZyzQ5zcXHBhAkTKrstRERE1a48d44v7X5dxTk52iMiIgJNmzYtT/PIRpg1oPrvf/9r9gYHDhxY7sZQzaFSqeDm5mYzjweoasxDxDxkzEJki3mkp6cjN+8u1r0EhPiWXteS5xoC955tOOqLQty6deuBH1DZ4rFhCbNW+anV5k21UqlUir02WhFc5UdEVHMZVu69B7QLLLu+Rdu+BITNNL6jfHE6nQ4HDhzA9evX0aBBAzz66KPQaDSV2xiyWKWv8tPr9Wb9q8rB1O3btzFy5Ei4u7ujdu3aGD9+PHJyckp9z927dzF58mTUq1cPrq6uGDJkCG7cuCHUSUlJQWRkJFxcXODt7Y3XX38dWq3W8Pr169fx7LPPokWLFlCr1Zg6dWpVdK/G0ev1uHz5smJXY1iKeYiYh4xZiJiHaaXlsXnzZjRr1gwRERF49tlnERERgWbNmmHz5s3V2MKqp/Rjw+JVftYycuRInD17Frt378a2bduwf/9+TJw4sdT3vPbaa/j555/x448/4vfff8e1a9cwePBgw+s6nQ6RkZEoKCjA4cOHsWbNGnzzzTeYPXu2oU5+fj68vLwwc+ZMtGnTpsr6V9PodDqcOHGCZxz/wTxEzEPGLETMw7SSBhGbN2/G008/jdatWyMmJgbZ2dmIiYlB69at8fTTTytqUKX0Y8PsAVVMTAy2bdsmlK1duxaBgYHw9vbGxIkTkZ+fX+kNBO6tpvj111/x1VdfoVOnTnjkkUfw2WefYcOGDbh27ZrJ92RmZuLrr7/Gxx9/jJ49eyIsLAyrV6/G4cOHceTIEQDArl27cO7cOaxbtw6hoaHo378/5s+fj+XLl6OgoADAvSepL1u2DKNHj4aHh0eV9I+IiB48Op0O06dPx+OPP46tW7eic+fOcHV1RefOnbF161Y8/vjjmDFjhmIHIEpj9iq/d999Fz169MDjjz8OADh9+jTGjx+PsWPHIiQkBB9++CF8fX0xd+7cSm9kTEwMateujfbt2xvKevfuDbVajaNHj+Kpp54yek9sbCwKCwvRu3dvQ1lwcDD8/f0RExODzp07G/4KqF+/vqFO3759MWnSJJw9exZt27YtV3vz8/OFwWVWVhaAe09hLywsBHBvXppGo4FOpxP+cikq12q1uH96m0ajgVqtLrG8aLtFim7vf//ly9LK7e3thcu292+v+OVclUoFOzu7EtteU/t0f9vL06f7c1FKn8q7n+7vh1L6dH+5JX0ylYmt98lUubl9uv9zbKVPxcurQtHPf41Gg7t37+LMmTOIjY1FcnIyZs6cibi4OOTn5+PixYto3LgxnJ2d8eSTT+Lnn3/GV199hbZt2yIoKAguLi5m9akmHntF/5UkCZIk2czPCHOZPaA6ceIE5s+fb/h6w4YN6NSpE/7zn/8AAPz8/DBnzpwqGVClpqbC29tbKLOzs0PdunWRmppa4nscHBxQu3Ztobx+/fqG96SmpgqDqaLXi14rr0WLFmHevHlG5bt27TJ8M/j7+6Nt27Y4deoUUlJSDHWCgoIQHByMY8eOCfcuCQ0NRePGjbF//35kZ2cbysPDw+Ht7Y1du3YJOz4iIgLOzs7YsWOH0IYBAwYgLy8P0dHRhjI7OztERkYiPT0dMTExhnKNRgOVSoUrV67gxIkThnIvLy906dIF58+fR1JSkqHcFvrk5uaGnj17Wtync+fOAQB2796tmD5VdD95eHhApVJh586diulTefZT0eO2io4NJfSpovvJ3t4eKpXKZvp08OBBVLWjR48iPT0doaGhuHXrFjp16mR47YUXXij1vf/6178AAEuWLDGsFLTl76c7d+7A1dXVJn5GHDp0COYy+1l+Tk5OOH/+PPz8/AAAjzzyCPr374933nkHAJCcnIzWrVsLYZblrbfewgcffFBqnYSEBGzevBlr1qwRwgEAb29vzJs3D5MmTTJ6X1RUFMaNG2d0GbJjx46IiIjABx98gIkTJ+Ly5cvYuXOn4fXc3FzUqlULO3bsQP/+/YX39ujRA6GhoVi6dGmpbTZ1hsrPzw/p6emGVQI19S+1mvhXDfvEPrFP7FNV9umPP/5Ap06dqnSV39GjR9G2bVujM1QvvfQSvvrqK7Rp0wZJSUkYNWoU1qxZg+DgYJw+fRovvPACVqxYoYgzVEVs6di7ffs26tWrZ9YqP7PPUNWvXx+XLl2Cn58fCgoKEBcXJ5yFyc7Ohr29vbmbAwBMnz4dY8eOLbVOkyZN4OPjg5s3bwrlWq0Wt2/fho+Pj8n3+fj4oKCgABkZGcJZqhs3bhje4+Pjg2PHjgnvK1oFWNJ2zeHo6AhHR0ejcnt7e6OMNBqNyaWxJT2Ru6TykrK3pFytVhtukaHT6XD+/Hk0b97ccGAVV1Lba2qfzCkvqe0AcPHiRUMeZbXdFvpUkf2k0+nw559/onnz5orpU1ltLKlcp9OZPDZsuU8llZvTp+I/O2yhT/bQoq2PGs4SAPPv7WkWZwlo6wNcSDgt9MPOzg7t27eHr68v1q5diyVLlhheK/qZu3btWjRs2BDt2rWDRqPBxYsXhW17enrC39/fZJ9q4rFX/Niw5Z8RJuuaW3HAgAGGM0pbt26Fi4sLHn30UcPrp06dsvimZV5eXvDy8iqzXnh4ODIyMhAbG4uwsDAAwN69e6HX64XTpvcLCwuDvb099uzZgyFDhgAAkpKSkJKSYniIc3h4OBYsWICbN28aLinu3r0b7u7uaNmypUV9URq9Xo+kpCQ0bdqU90IB8yiOeciYhcgW83DKSUHci66AFkBKmdUtEgIg7kXg2Y/+hY/SjVf61QeQ/WcqJj5x73dZWx81lswYbXjdG8CLAzub3LajkyN+2BMPvyZBldvoKmKLx4YlzB5QzZ8/H4MHD0b37t3h6uqKNWvWwMFBvk3sqlWr8Nhjj1VJI0NCQtCvXz9MmDABK1asQGFhIaZMmYLhw4fD1/febW2vXr2KXr16Ye3atejYsSM8PDwwfvx4TJs2DXXr1oW7uztefvllhIeHo3PnewfnY489hpYtW+K5557D4sWLkZqaipkzZ2Ly5MnCGaai67M5OTlIS0vDiRMn4ODg8MAPuoiIlOCuqz/a/TsH6yeXfad0Sx08o8Yj7i6IGuJSuRv+R0JyPGAjAyqlM3tA5enpif379yMzMxOurq5Go8sff/wRrq6uld7AIuvXr8eUKVPQq1cvqNVqDBkyBJ9++qnh9cLCQiQlJSE3N9dQ9sknnxjq5ufno2/fvvjiiy8Mr2s0Gmzbtg2TJk1CeHg4atWqhTFjxuDdd98VPvv+1X6xsbGIiopC48aNkZycXGX9JSKi6iHZOSE+VY88FQCnyt12tkaPditzALNmK1vG0ckRP4wu32p0qnwWPxy5pHsx1a1bt8KNKU3dunURFRVV4usBAQEoPr/eyckJy5cvx/Lly0t8X+PGjY1WGhRn5rx9RVGr1fD39zf7sUNKxzxEzEPGLETMQ1S/NhB/XY+1a9fioYceAgDk5eWZ/IP80qVLmDVrFubPn4/AQOPZ8QEBAXB2djZ87enpCT8Tc6hqKqUfG2av8qPy47P8iIhqrup+ll/R51mqrOcBUuWz5Pe3xWeo6MGg0+lw6tQpPPzww4qcPGgp5iFiHjJmIWIept2/tD84OBixsbFGdYrOXBU/E3X/+2yZ0o8NDqjIJL1ej5SUFLRq1UqRB76lmIeIeciYhYh5mHb/xSAXF5cSzzR17dq1uppU7ZR+bCjzQiYRERFRNeKAioiIiKiCOKAik9RqNYKCghS7GsNSzEPEPGTMQsQ8TFOpVNZugtUp/djgHCoySaPR2PwEyMrEPETMQ8YsRLacR8K1suvkFQDJ6UCAJ+DsUHb9om0qcc6QpWz52DAHB1RkklarxbFjx9CxY0eLnmWkVMxDxDxkzEJki3l4enrCxdkJo76o5Af5/cPZyUF4puyDyhaPDUsor0dUKSRJQlpa2gN5U1NTmIeIeciYhcgW8/D390dCYhLS09OFclM34LT05ptarRanTp2Cn59flbTdltjisWEJDqiIiOiB5+/vD/9idx2Pi4vDqFGjTNafNWuWyfLiN98sLCzE9evXK6+hVGNxQEVERGSCqRtwKv3mm1R+HFCRSRqNBqGhoZxI+Q/mIWIeMmYhUlIeJd2A05Kbbyopj4pSehZ8ll814LP8iIiIbI8lv7+VeTMIqjCtVou9e/dCq9Vauyk1AvMQMQ8ZsxAxDxHzkCk9Cw6oyCRJkpCdna3Y1RiWYh4i5iFjFiLmIWIeMqVnwQEVERERUQVxQEVERERUQZyUXg1scVK6Xq9Heno6PD09FfvcJUswDxHzkDELEfMQMQ+ZLWZhye9vDqiqgS0OqIiIiB50XOVHFVZYWIjt27ejsLDQ2k2pEZiHiHnImIWIeYiYh0zpWfDGnlQipS5tLS/mIWIeMmYhYh4iJeSRm5uLxMREo3Jz7hzv4uJi+FoJWZSEAyoiIiIqVWJiIsLCwix+X/FnGyoZB1RERERUKlPPNQSAhIQEjBo1CuvWrUNISIjJ9z0oOCm9GtjipPSiG7C5ublBpVJZuzlWxzxEzEPGLETMQ6T0POLi4hAWFmbWmShbzIKT0qlSmLoe/iBjHiLmIWMWIuYhYh4yJWfBARWZpNVqsWPHDkVPILQE8xAxDxmzEDEPEfOQKT0LDqiIiIiIKogDKiIiIqIK4oCKiIiIqIK4yq8a2OoqP61WCzs7O5tZjVGVmIeIeciYhYh5iJSeh6Wr/GwtC67yo0qRl5dn7SbUKMxDxDxkzELEPETMQ6bkLDigIpO0Wi2io6MVuxrDUsxDxDxkzELEPETMQ6b0LDigIiIiIqogDqiIiIiIKogDKiqRnR0f9Xg/5iFiHjJmIWIeIuYhU3IWXOVXDWxxlR8REVFZLFnlZ4u4yo8qTK/X4+bNm9Dr9dZuSo3APETMQ8YsRMxDxDxkSs+CAyoySafTISYmBjqdztpNqRGYh4h5yJiFiHmImIdM6VlwQEVERERUQRxQEREREVUQB1Rkkkqlgpubm808HqCqMQ8R85AxCxHzEDEPmdKz4Cq/asBVfkREpERc5SfjGSoySa/X4/Lly4pdjWEp5iFiHjJmIWIeIuYhU3oWHFCRSTqdDidOnFDsagxLMQ8R85AxCxHzEDEPmdKz4ICKiIiIqII4oCIiIiKqIOU+VIcqRKVSwcvLS7GrMSzFPETMQ8YsRMxDZKt5pKSkID09vcx6CQkJwn9Lo9VqodPpbC4Lc3GVXzXgKj8iIrIVKSkpCAkOQm7e3UrftouzExISk+Dv71/p264Klvz+5hkqMkmn0+H8+fNo3rw5NBqNtZtjdcxDxDxkzELEPES2mEd6ejpy8+5i3UtAiG/pdfMKgOR0IMATcHYovW7CNWDUF3dx48YNmxlQWYIDKjJJr9cjKSkJTZs2tZkfAlWJeYiYh4xZiJiHyJbzCPEF2gWWXa9rkGXbVeqFMU5KJyIiIqogDqiIiIiIKogDKjJJrVbD398fajUPEYB5FMc8ZMxCxDxEzMOYUlf5cQ4VmaTRaNC2bVtrN6PGYB4i5iFjFiLmIWIexmxtLpm5OGQmk3Q6HeLj4xX7iABLMQ8R85AxCxHzEDEPY0rNggMqMkmv1yMlJUWxD7G0FPMQMQ8ZsxAxDxHzMMZVflZ2+/ZtjBw5Eu7u7qhduzbGjx+PnJycUt9z9+5dTJ48GfXq1YOrqyuGDBmCGzduCHVSUlIQGRkJFxcXeHt74/XXX4dWqzW8vnnzZvTp0wdeXl5wd3dHeHg4du7cWSV9JCIiIttkMwOqkSNH4uzZs9i9eze2bduG/fv3Y+LEiaW+57XXXsPPP/+MH3/8Eb///juuXbuGwYMHG17X6XSIjIxEQUEBDh8+jDVr1uCbb77B7NmzDXX279+PPn36YMeOHYiNjUVERASeeOIJxMfHV1lfiYiIHmQ6nQ779u3Dd999h3379tnEZUKbmJSekJCAX3/9FcePH0f79u0BAJ999hkGDBiAjz76CL6+xrdyzczMxNdff42oqCj07NkTALB69WqEhITgyJEj6Ny5M3bt2oVz587ht99+Q/369REaGor58+fjzTffxNy5c+Hg4IClS5cK2124cCF++ukn/Pzzz4qeaKhWqxEUFMSVKf9gHiLmIWMWIuYhYh7Gylrlt3nzZkyfPh3JycmGsoCAACxZskQ4KVLT2MSAKiYmBrVr1zYMpgCgd+/eUKvVOHr0KJ566imj98TGxqKwsBC9e/c2lAUHB8Pf3x8xMTHo3LkzYmJi0Lp1a9SvX99Qp2/fvpg0aRLOnj1rcsCk1+uRnZ2NunXrltje/Px85OfnG77OysoCABQWFqKwsBDAvW8yjUYDnU4nXFsvKtdqtcJ1Zo1GA7VaXWJ50XaL2Nnd27X3X74srdze3h56vV74K6BZs2bQaDRG5SqVCnZ2diW2vSb3qajtlvYJAJo2bQq9Xg+9Xq+IPlV0P7Vo0UJxfSrPflKr1cKxoYQ+mSq3pE9FPzuU1Kfy7idJkoTjwxb6VNVng9RqNSRJMtmnzZs3Y+jQoRgwYAC+/fZbPPTQQ0hKSsKCBQvw9NNPY8OGDXjqqaeq9fvJXDYxoEpNTYW3t7dQZmdnh7p16yI1NbXE9zg4OKB27dpCef369Q3vSU1NFQZTRa8XvWbKRx99hJycHAwbNqzE9i5atAjz5s0zKt+1axdcXFwAAP7+/mjbti1OnTqFlJQUQ52goCAEBwfj2LFjSEtLM5SHhoaicePG2L9/P7Kzsw3l4eHh8Pb2xq5du4QdHxERAWdnZ+zYsUNow4ABA5CXl4fo6GhDmZ2dHSIjI5Geno6YmBhDuUajQb9+/XD16lWcOHHCUO7l5YUuXbrg/PnzSEpKMpTbQp/c3NzQs2dPXLlyxaI+nTx5EleuXFFUnyq6nzw8PPDII48oqk/l2U85OTnYv3+/ovpU0f1kb2+Pxx57DBcuXFBMn8q7n44ePYr09HSb6tOdO3dQlbKzs6HVao361LdvX0ybNg3t27fH+PHjcevWLRw+fBiRkZH4z3/+g4EDB+KVV16BnZ0dateuXS3fT4cOHTK7XyrJitPt33rrLXzwwQel1klISMDmzZuxZs0aIRwA8Pb2xrx58zBp0iSj90VFRWHcuHHCmSIA6NixIyIiIvDBBx9g4sSJuHz5sjDJPDc3F7Vq1cKOHTvQv39/o21OmDABP/30k3DmqzhTZ6j8/PyQnp5ueFp1Tf9LrbCwELt378aAAQMM7SyixL+oy+rT3bt3sXPnTvTp0wf29vaK6FNF9tP9x0dxttqn+8st2U+FhYX45ZdfDMeGEvpkqtzcPt1/bKjVakX06f5yS/dTXl4edu3aZTg+bKFPJ06cQMeOHRH7nnnP8jNX3CUgbCZw5MgRdOzY0ajtBw8eRM+ePXHgwAF06tTJqE+HDh1Ct27dsHv3bvTo0aNavp9u376NevXqITMz0/D7uyRWPUM1ffp0jB07ttQ6TZo0gY+PD27evCmUa7Va3L59Gz4+Pibf5+Pjg4KCAmRkZAhnqW7cuGF4j4+PD44dOya8r2gVYPHtbtiwAS+88AJ+/PHHUgdTAODo6AhHR0ejcnt7e8MP3CIajcbkTc6KvlnMLS++3fKUq9Vqk9f5Syovqe1K7VPRZ9z/ObbeJyXup+ruU9F8kOLHhi33qaRy9qn8fTL3Z0dN6FNV33hTpVJBpVIZ2p6bm4vExEThrNrp06eRl5eH5ORkBAQEwNnZ2fDasWPH0KVLF8Mld2sceybrml2zCnh5ecHLy6vMeuHh4cjIyEBsbCzCwsIAAHv37oVerxdGsfcLCwuDvb099uzZgyFDhgAAkpKSkJKSgvDwcMN2FyxYgJs3bxouKe7evRvu7u5o2bKlYVvfffcdnn/+eWzYsAGRkZEV6jMRERHJEhMTDb/bAeDRRx8ttf4777yDfv36oV27dlXdNIvYxLKDkJAQ9OvXDxMmTMCxY8dw6NAhTJkyBcOHDzes8Lt69arhOihwb37H+PHjMW3aNERHRyM2Nhbjxo1DeHg4OnfuDAB47LHH0LJlSzz33HM4efIkdu7ciZkzZ2Ly5MmGM0xRUVEYPXo0lixZgk6dOiE1NRWpqanIzMy0ThjVRKPRIDQ0VLGPCLAU8xAxDxmzEDEPEfMwVvyMUnBwMGJjY3Hs2DH4+vqiW7duOH78ONatWwcAWLduHY4fP45u3bqhYcOGOHbsGIKDg63R9FLZxIAKANavX4/g4GD06tULAwYMwCOPPIKVK1caXi8sLERSUhJyc3MNZZ988gkef/xxDBkyBN26dYOPjw82b95seF2j0WDbtm3QaDQIDw/HqFGjMHr0aLz77ruGOitXroRWq8XkyZPRoEEDw79XX321ejpuJWq1Go0bN+ZS338wDxHzkDELEfMQ2WIeKu1dtPVRw1lSA3cr75+zpEZbHzU0+gLh81xcXNCuXTt06NABn332GQ4cOIB3333XMJ+ssLAQ7777Lg4cOIBPP/0UHTp0MCzwqkmsOin9QZGVlQUPDw+zJrXVFFqtFvv370e3bt0suoasVMxDxDxkzELEPES2mEfC3g0I2f9ilW1/W8M34BvWv8TX9+7di08++QTXrl0zlDVs2BBTp0413FfSFE9PT/j7+1dqWy35/W0be5eqnSRJyM7OVuwzlyzFPETMQ8YsRMxDZIt53HX1R7t/52D9ZCDE+L7Z5XYoCVi+Q43kjNm4q51dat36AOr73HdWT3cdUUveRNSSkt/j6OSIH/bEw69JUOU02EIcUBEREZGBZOeE+FQ98lQAnCpvu5l31YgaUrWX6hKS4wEOqIiIiEipsrR6tPt3DtATQG0z3qADkA3ADUBZc/ozAOwFvh5QuZf8LMEBFZlUNFGfK1PuYR4i5iFjFiLmIWIeMjcXIP6GHoiqmu07OTuhnk+jqtm4GTigIpPUarXR434eZMxDxDxkzELEPETMQ9agNgDp3m0QQkJCSqy3bNkyrF+/XrgDukajwciRI0tdYV8Vk9ItwQEVmVRYWIhdu3bhscceK/EOvQ8S5iFiHjJmIWIeIuZhrHnz5iXelPONN97A2rVrjW4zIUkS1q5di/r162Px4sXV0UyL2c6NMajaWfKU7QcB8xAxDxmzEDEPEfMoXW5uLuLi4nDkyBEsWXJvGV+XLl0wd+5cAMDcuXPRpUsXAMCSJUtw5MgR4Z6TNQXPUBEREZHVFH/0DHDvQckHDx4EAMPACgD0ej3Cw8MRGxvLR88QERERFSl69EyPHj0AAEuXLkVsbCwOHjyIdevW4eDBg4iNjcXHH38MAOjRo0eNfPQMz1CRSXZ2doiIiLCZO/tWNeYhYh4yZiFiHiLmYaz4iseiR880atTI8HrR2aeuXbsa6hWdsWrUqFGNfPQMz1BRiZydna3dhBqFeYiYh4xZiJiHiHmY57nnngMAzJkzx2jemVarNVz6K6pX03BARSZptVrs2LGDkyn/wTxEzEPGLETMQ8Q8jN1/O4T79erVC+7u7rh9+zYaNWqElStX4tq1a1i5ciUaNWqEv//+G+7u7ujVq1c1t9g8PAdJREREVqfRaLB69WoMGTIEN2/exIsvyg9oVqlUAIDVq1fX2Juk8gwVERER1QiDBw/Gpk2bjG7Q2bhxY2zatAmDBw+2UsvKxjNUREREVGMMHjwYTz75JA4cOIDr16+jQYMGePTRR2vsmakiKkmSJGs3QumysrLg4eGBzMxMuLu7W7s5ZpEkCVqtFnZ2doZTrQ8y5iFiHjJmIWIeIlvMIy4uDmFhYYh9D2gXWInbvQSEzQT++OMPo/tO1VSW/P7mJT8qUV5enrWbUKMwDxHzkDELEfMQMY8HAwdUZJJWq0V0dDRXpvyDeYiYh4xZiJiHiHkYK2mVn63jgIqIiIiogjigIiIiIqogDqioRHxUgoh5iJiHjFmImIeIeTwYuJfJJHt7e0RGRlq7GTUG8xAxDxmzEDEPEfMwptQBpjJ7RRWm1+uRnp4OT09PqNU8kck8RMxDxixEzENky3kkXCu7Tl4BkJwOBHgCzg7mbU+v11e8cTUQB1Rkkk6nQ0xMDAYMGGBzPwSqAvMQMQ8ZsxAxD5Et5uHp6QkXZyeM+uJupW/bydEederUqfTt1gQcUBEREZGBv78/EhKTkJ6eXmbdhIQEjBo1CuvWrUNISEipdbVaLU6dOmX0WBml4ICKiIiIBP7+/hYNfEJCQtCuXbtS6xQWFuL69esVbVqNxQEVmaRSqeDm5mYzj0qoasxDxDxkzELEPERKySM3NxeJiYlG5QkJCcJ/iwsODoaLiwsA5WRREj7LrxrY4rP8iIiIihQ9389SsbGxZZ65qsks+f3NM1Rkkl6vx5UrV+Dn52czEymrEvMQMQ8ZsxAxD5FS8ggODkZsbKxReV5eHpKTkxEQEABnZ2eT7yuilCxKwgEVmaTT6XDixAn4+voq8sC3FPMQMQ8ZsxAxD5FS8nBxcSnxTFPXrl3N2oZSsiiJ8npEREREVM04oCIiIiKqIA6oyCSVSgUvLy/FrsawFPMQMQ8ZsxAxDxHzkCk9C67yqwZc5UdERGR7LPn9zTNUZJJOp0NiYiJ0Op21m1IjMA8R85AxCxHzEDEPmdKz4ICKTNLr9UhKSlLsQywtxTxEzEPGLETMQ8Q8ZErPggMqIiIiogrigIqIiIiogjigIpPUajX8/f0VefO18mAeIuYhYxYi5iFiHjKlZ8FVftWAq/yIiIhsD1f5UYXpdDrEx8crdjWGpZiHiHnImIWIeYiYh0zpWXBARSbp9XqkpKQodjWGpZiHiHnImIWIeYiYh0zpWXBARURERFRBdtZuwIOgaJpaVlaWlVtivsLCQuTm5iIrKwv29vbWbo7VMQ8R85AxCxHzEDEPmS1mUfR725zp5hxQVYPs7GwAgJ+fn5VbQkRERJbKzs6Gh4dHqXW4yq8a6PV6XLt2DW5ubjbzUMisrCz4+fnhypUrXJkI5lEc85AxCxHzEDEPmS1mIUkSsrOz4evrW+btHniGqhqo1Wo0atTI2s0oF3d3d5s58KsD8xAxDxmzEDEPEfOQ2VoWZZ2ZKsJJ6UREREQVxAEVERERUQVxQEUmOTo6Ys6cOXB0dLR2U2oE5iFiHjJmIWIeIuYhU3oWnJROREREVEE8Q0VERERUQRxQEREREVUQB1REREREFcQBFREREVEFcUBFJs2dOxfBwcGoVasW6tSpg969e+Po0aPWbpZVFBYW4s0330Tr1q1Rq1Yt+Pr6YvTo0bh27Zq1m2Y1mzdvxmOPPYZ69epBpVLhxIkT1m6S1SxfvhwBAQFwcnJCp06dcOzYMWs3ySr279+PJ554Ar6+vlCpVNi6dau1m2Q1ixYtQocOHeDm5gZvb28MGjQISUlJ1m6W1Xz55Zd4+OGHDTf0DA8Pxy+//GLtZlU6DqjIpBYtWuDzzz/H6dOncfDgQQQEBOCxxx5DWlqatZtW7XJzcxEXF4dZs2YhLi4OmzdvRlJSEgYOHGjtplnNnTt38Mgjj+CDDz6wdlOs6vvvv8e0adMwZ84cxMXFoU2bNujbty9u3rxp7aZVuzt37qBNmzZYvny5tZtidb///jsmT56MI0eOYPfu3SgsLMRjjz2GO3fuWLtpVtGoUSO8//77iI2NxR9//IGePXviySefxNmzZ63dtErF2yaQWbKysuDh4YHffvsNvXr1snZzrO748ePo2LEjLl++DH9/f2s3x2qSk5MRGBiI+Ph4hIaGWrs51a5Tp07o0KEDPv/8cwD3ntvp5+eHl19+GW+99ZaVW2c9KpUKW7ZswaBBg6zdlBohLS0N3t7e+P3339GtWzdrN6dGqFu3Lj788EOMHz/e2k2pNDxDRWUqKCjAypUr4eHhgTZt2li7OTVCZmYmVCoVateube2mkJUUFBQgNjYWvXv3NpSp1Wr07t0bMTExVmwZ1TSZmZkA7g0iHnQ6nQ4bNmzAnTt3EB4ebu3mVCo+HJlKtG3bNgwfPhy5ublo0KABdu/eDU9PT2s3y+ru3r2LN998EyNGjLCpB3xS5UpPT4dOp0P9+vWF8vr16yMxMdFKraKaRq/XY+rUqejatStatWpl7eZYzenTpxEeHo67d+/C1dUVW7ZsQcuWLa3drErFM1SE9evXw9XV1fDvwIEDAICIiAicOHEChw8fRr9+/TBs2LAHYm5ISXkA9yaoDxs2DJIk4csvv7RiK6tPaXkQUekmT56MM2fOYMOGDdZuilUFBQXhxIkTOHr0KCZNmoQxY8bg3Llz1m5WpeIZKsLAgQPRqVMnw9cNGzYEANSqVQvNmjVDs2bN0LlzZzRv3hxff/013n77bWs1tVqUlEfRYOry5cvYu3fvA3N2qqQ8HnSenp7QaDS4ceOGUH7jxg34+PhYqVVUk0yZMgXbtm3D/v370ahRI2s3x6ocHBzQrFkzAEBYWBiOHz+OZcuW4d///reVW1Z5OKAiuLm5wc3Nrcx6er0e+fn51dAi6zKVR9Fg6vz584iOjka9evWs1LrqZ+7x8aBxcHBAWFgY9uzZY5h8rdfrsWfPHkyZMsW6jSOrkiQJL7/8MrZs2YJ9+/YhMDDQ2k2qcZT4+4QDKjJy584dLFiwAAMHDkSDBg2Qnp6O5cuX4+rVqxg6dKi1m1ftCgsL8fTTTyMuLg7btm2DTqdDamoqgHuTTB0cHKzcwup3+/ZtpKSkGO7FVXSPHR8fnwfq7My0adMwZswYtG/fHh07dsTSpUtx584djBs3ztpNq3Y5OTm4cOGC4etLly7hxIkTqFu37gO3Enby5MmIiorCTz/9BDc3N8PPCw8PDzg7O1u5ddXv7bffRv/+/eHv74/s7GxERUVh37592Llzp7WbVrkkomLy8vKkp556SvL19ZUcHBykBg0aSAMHDpSOHTtm7aZZxaVLlyQAJv9FR0dbu3lWsXr1apN5zJkzx9pNq3afffaZ5O/vLzk4OEgdO3aUjhw5Yu0mWUV0dLTJY2LMmDHWblq1K+nnxerVq63dNKt4/vnnpcaNG0sODg6Sl5eX1KtXL2nXrl3Wblal432oiIiIiCqIq/yIiIiIKogDKiIiIqIK4oCKiIiIqII4oCIiIiKqIA6oiIiIiCqIAyoiIiKiCuKAioiIiKiCOKAiomq3b98+qFQqZGRk1IjtVNTcuXMRGhpq1TZUprFjxxoep0NE5uGAiogsMnbsWKhUKqhUKtjb2yMwMBBvvPEG7t69W6Wf26NHD0ydOlUo69KlC65fvw4PD48q/ewtW7agc+fO8PDwgJubGx566CGjttQENWWASfQg4rP8iMhi/fr1w+rVq1FYWIjY2FiMGTMGKpUKH3zwQbW2w8HBocqfHbhnzx4888wzhudbqlQqnDt3Drt3767SzyUi28IzVERkMUdHR/j4+MDPzw+DBg1C7969hQGGXq/HokWLEBgYCGdnZ7Rp0wYbN24scXu3bt3CiBEj0LBhQ7i4uKB169b47rvvDK+PHTsWv//+O5YtW2Y4O5acnCyckcnKyoKzszN++eUXYdtbtmyBm5sbcnNzAQBXrlzBsGHDULt2bdStWxdPPvkkkpOTS2zbzz//jK5du+L1119HUFAQWrRogUGDBmH58uUlvkev1+Pdd99Fo0aN4OjoiNDQUPz666+G15OTk6FSqbBhwwZ06dIFTk5OaNWqFX7//XdhO2fOnEH//v3h6uqK+vXr47nnnkN6enqJn1vcN998g9q1a2Pnzp0ICQmBq6sr+vXrh+vXrxvq6HQ6TJs2DbVr10a9evXwxhtvoPgTyUrbn5IkoXfv3ujbt6/hfbdv30ajRo0we/Zss9tKZOs4oCKiCjlz5gwOHz4MBwcHQ9miRYuwdu1arFixAmfPnsVrr72GUaNGGQ0Yity9exdhYWHYvn07zpw5g4kTJ+K5557DsWPHAADLli1DeHg4JkyYgOvXr+P69evw8/MTtuHu7o7HH38cUVFRQvn69esxaNAguLi4oLCwEH379oWbmxsOHDiAQ4cOGQYZBQUFJtvm4+ODs2fP4syZM2ZnsmzZMixZsgQfffQRTp06hb59+2LgwIE4f/68UO/111/H9OnTER8fj/DwcDzxxBO4desWACAjIwM9e/ZE27Zt8ccff+DXX3/FjRs3MGzYMLPbAQC5ubn46KOP8O2332L//v1ISUnBjBkzDK8vWbIE33zzDVatWoWDBw/i9u3b2LJli7CN0vanSqXCmjVrcPz4cXz66acAgH/9619o2LAhB1T0YLHqo5mJyOaMGTNG0mg0Uq1atSRHR0cJgKRWq6WNGzdKkiRJd+/elVxcXKTDhw8L7xs/frw0YsQISZIkKTo6WgIg/f333yV+TmRkpDR9+nTD1927d5deffVVoU7x7WzZskVydXWV7ty5I0mSJGVmZkpOTk7SL7/8IkmSJH377bdSUFCQpNfrDdvIz8+XnJ2dpZ07d5psR05OjjRgwAAJgNS4cWPpmWeekb7++mvp7t27hjpz5syR2rRpY/ja19dXWrBggbCdDh06SC+99JIkSZJ06dIlCYD0/vvvG14vLCyUGjVqJH3wwQeSJEnS/Pnzpccee0zYxpUrVyQAUlJSksm2Fs9j9erVEgDpwoULhjrLly+X6tevb/i6QYMG0uLFi43a8eSTT0qSZN7+lCRJ+uGHHyQnJyfprbfekmrVqiX9+eefJttIpFScQ0VEFouIiMCXX36JO3fu4JNPPoGdnR2GDBkCALhw4QJyc3PRp08f4T0FBQVo27atye3pdDosXLgQP/zwA65evYqCggLk5+fDxcXFonYNGDAA9vb2+O9//4vhw4dj06ZNcHd3R+/evQEAJ0+exIULF+Dm5ia87+7du7h48aLJbdaqVQvbt2/HxYsXER0djSNHjmD69OlYtmwZYmJijNqYlZWFa9euoWvXrkJ5165dcfLkSaEsPDzc8P92dnZo3749EhISDG2Njo6Gq6urUZsuXryIFi1amJWJi4sLmjZtavi6QYMGuHnzJgAgMzMT169fR6dOnYzaIf1z+c7c/Tl06FBs2bIF77//Pr788ks0b97crPYRKQUHVERksVq1aqFZs2YAgFWrVqFNmzb4+uuvMX78eOTk5AAAtm/fjoYNGwrvc3R0NLm9Dz/8EMuWLcPSpUvRunVr1KpVC1OnTi3xMlxJHBwc8PTTTyMqKgrDhw9HVFQUnnnmGdjZ3ftRl5OTg7CwMKxfv97ovV5eXqVuu2nTpmjatCleeOEFvPPOO2jRogW+//57jBs3zqI2misnJwdPPPGEyYn+DRo0MHs79vb2wtcqlcpojlRZ7QDK3p+5ubmIjY2FRqMxurRJ9CDggIqIKkStVuP//u//MG3aNDz77LNo2bIlHB0dkZKSgu7du5u1jUOHDuHJJ5/EqFGjANybBP3nn3+iZcuWhjoODg7Q6XRlbmvkyJHo06cPzp49i7179+K9994zvNauXTt8//338Pb2hru7u4U9lQUEBMDFxQV37twxes3d3R2+vr44dOiQ0P9Dhw6hY8eOQt0jR46gW7duAACtVovY2FhMmTLF0NZNmzYhICDAMCCsbB4eHmjQoAGOHj1q1I527doBgNn7c/r06VCr1fjll18wYMAAREZGomfPnlXSbqKaiJPSiajChg4dCo1Gg+XLl8PNzQ0zZszAa6+9hjVr1uDixYuIi4vDZ599hjVr1ph8f/PmzbF7924cPnwYCQkJePHFF3Hjxg2hTkBAAI4ePYrk5GSkp6dDr9eb3Fa3bt3g4+ODkSNHIjAwULicNXLkSHh6euLJJ5/EgQMHcOnSJezbtw+vvPIK/ve//5nc3ty5c/HGG29g3759uHTpEuLj4/H888+jsLDQ6DJYkddffx0ffPABvv/+eyQlJeGtt97CiRMn8Oqrrwr1li9fji1btiAxMRGTJ0/G33//jeeffx4AMHnyZNy+fRsjRozA8ePHcfHiRezcuRPjxo0za2BprldffRXvv/8+tm7disTERLz00kvCfazM2Z/bt2/HqlWrsH79evTp0wevv/46xowZg7///rvS2klU41l7EhcR2ZYxY8YYJizfb9GiRZKXl5eUk5Mj6fV6aenSpVJQUJBkb28veXl5SX379pV+//13SZKMJ0/funVLevLJJyVXV1fJ29tbmjlzpjR69Gjhc5KSkqTOnTtLzs7OEgDp0qVLJU5uf+ONNyQA0uzZs43aef36dWn06NGSp6en5OjoKDVp0kSaMGGClJmZabK/e/fulYYMGSL5+flJDg4OUv369aV+/fpJBw4cMNQpPildp9NJc+fOlRo2bCjZ29tLbdq0MUyMlyR5UnpUVJTUsWNHycHBQWrZsqW0d+9e4bP//PNP6amnnpJq164tOTs7S8HBwdLUqVOFSfX3MzUp3cPDQ6izZcsW6f4f/YWFhdKrr74qubu7S7Vr15amTZtmlH1p+/PmzZtS/fr1pYULFxrqFxQUSGFhYdKwYcNMtpNIiVSSZMHFdCIiqrDk5GQEBgYiPj5eUY+sIXqQ8ZIfERERUQVxQEVERERUQbzkR0RERFRBPENFREREVEEcUBERERFVEAdURERERBXEARURERFRBXFARURERFRBHFARERERVRAHVEREREQVxAEVERERUQVxQEVERERUQf8PHB1WamfOSmAAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "H_jtfs = []\n",
    "H_scrapl = []\n",
    "common_density_idx = 3\n",
    "y_slope_idx = 3\n",
    "\n",
    "for x_slope_idx in range(7):\n",
    "\n",
    "    data_dir = \"icassp2024_scrapl_data\"\n",
    "    SX, JX, SY = [], [], []\n",
    "    for seed in range(10):\n",
    "        y_coords = (common_density_idx, y_slope_idx)\n",
    "        x_coords = (common_density_idx, x_slope_idx)\n",
    "        Sx = load_jtfs(*x_coords, seed=seed, data_dir=data_dir)\n",
    "        Sy = load_jtfs(*y_coords, seed=seed, data_dir=data_dir)\n",
    "        Jx = load_jacobian(*x_coords, seed=seed, data_dir=data_dir)\n",
    "        SX.append(Sx)\n",
    "        SY.append(Sy)\n",
    "        JX.append(Jx)\n",
    "    SX = torch.stack(SX, dim=0)\n",
    "    SY = torch.stack(SY, dim=0)\n",
    "    JX = torch.stack(JX, dim=0)\n",
    "\n",
    "    G_scrapl = []\n",
    "    for _ in range(100):\n",
    "        Sx_scrapl, Sy_scrapl, Jx_scrapl = [], [], []\n",
    "        for path in range(SX.shape[-1]):\n",
    "            random_seed = torch.randint(SX.shape[0], (1,))\n",
    "            Jx_scrapl.append(JX[random_seed, path, :].squeeze())\n",
    "            Sx_scrapl.append(SX[random_seed, path].squeeze())\n",
    "            Sy_scrapl.append(SY[random_seed, path].squeeze())\n",
    "        Jx_scrapl = torch.stack(Jx_scrapl, dim=0)\n",
    "        Sx_scrapl = torch.stack(Sx_scrapl, dim=0)\n",
    "        Sy_scrapl = torch.stack(Sy_scrapl, dim=0)\n",
    "        g_scrapl = Jx_scrapl.T @ (Sy_scrapl - Sx_scrapl)\n",
    "        G_scrapl.append(g_scrapl)\n",
    "    G_scrapl = torch.stack(G_scrapl, dim=0).squeeze()\n",
    "\n",
    "    G_jtfs = []\n",
    "    for seed in range(10):\n",
    "        y_coords = (common_density_idx, y_slope_idx)\n",
    "        Sy = load_jtfs(*y_coords, seed=seed, data_dir=data_dir)\n",
    "        g_jtfs = JX[seed, :, :].T @ (SY[seed, :] - SX[seed, :])\n",
    "        G_jtfs.append(g_jtfs)\n",
    "    G_jtfs = torch.stack(G_jtfs, dim=0).squeeze()\n",
    "\n",
    "    # Take gradient with respect to coordinate 1 (slope)\n",
    "    H_jtfs.append(G_jtfs[:, 1].detach())\n",
    "    H_scrapl.append(G_scrapl[:, 1].detach())\n",
    "\n",
    "H_jtfs = torch.stack(H_jtfs)\n",
    "H_scrapl = torch.stack(H_scrapl)\n",
    "\n",
    "positions = np.arange(-3, 4)\n",
    "plt.boxplot(np.cbrt(H_jtfs), positions=positions-1/6, widths=1/3,\n",
    "    patch_artist=True, boxprops=dict(facecolor='orange'));\n",
    "plt.boxplot(np.cbrt(H_scrapl), positions=positions+1/6,widths=1/3,\n",
    "    patch_artist=True, boxprops=dict(facecolor='green'));\n",
    "\n",
    "plt.gca().set_xticks(positions)\n",
    "plt.gca().set_xticklabels(positions)\n",
    "plt.grid(linestyle='--')\n",
    "plt.xlabel(\"Relative Slope Index\")\n",
    "plt.ylabel(\"Slope gradient (cbrt)\")\n",
    "# Microstructure means matched random seeds between\n",
    "# target and prediction.\n",
    "plt.title(\"Microstructure. Orange: JTFS. Green: SCRAPL\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
       "       dtype=torch.float64)"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "load_jacobian(density_idx=3, slope_idx=3, seed=1, data_dir=data_dir)[:, 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Microstructure. Orange: JTFS. Green: SCRAPL')"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "H_jtfs = []\n",
    "H_scrapl = []\n",
    "common_density_idx = 3\n",
    "y_slope_idx = 1\n",
    "\n",
    "for x_slope_idx in range(7):\n",
    "\n",
    "    data_dir = \"icassp2024_scrapl_data\"\n",
    "    SX, JX, SY = [], [], []\n",
    "    for seed in range(10):\n",
    "        y_coords = (common_density_idx, y_slope_idx)\n",
    "        x_coords = (common_density_idx, x_slope_idx)\n",
    "        Sx = load_jtfs(*x_coords, seed=seed, data_dir=data_dir)\n",
    "        Sy = load_jtfs(*y_coords, seed=seed, data_dir=data_dir)\n",
    "        Jx = load_jacobian(*x_coords, seed=seed, data_dir=data_dir)\n",
    "        SX.append(Sx)\n",
    "        SY.append(Sy)\n",
    "        JX.append(Jx)\n",
    "    SX = torch.stack(SX, dim=0)\n",
    "    SY = torch.stack(SY, dim=0)\n",
    "    JX = torch.stack(JX, dim=0)\n",
    "\n",
    "    G_scrapl = []\n",
    "    for _ in range(100):\n",
    "        Sx_scrapl, Sy_scrapl, Jx_scrapl = [], [], []\n",
    "        for path in range(SX.shape[-1]):\n",
    "            random_seed = torch.randint(SX.shape[0], (1,))\n",
    "            Jx_scrapl.append(JX[random_seed, path, :].squeeze())\n",
    "            Sx_scrapl.append(SX[random_seed, path].squeeze())\n",
    "            Sy_scrapl.append(SY[random_seed, path].squeeze())\n",
    "        Jx_scrapl = torch.stack(Jx_scrapl, dim=0)\n",
    "        Sx_scrapl = torch.stack(Sx_scrapl, dim=0)\n",
    "        Sy_scrapl = torch.stack(Sy_scrapl, dim=0)\n",
    "        g_scrapl = Jx_scrapl.T @ (Sy_scrapl - Sx_scrapl)\n",
    "        G_scrapl.append(g_scrapl)\n",
    "    G_scrapl = torch.stack(G_scrapl, dim=0).squeeze()\n",
    "\n",
    "    G_jtfs = []\n",
    "    for seed in range(10):\n",
    "        y_coords = (common_density_idx, y_slope_idx)\n",
    "        Sy = load_jtfs(*y_coords, seed=seed, data_dir=data_dir)\n",
    "        g_jtfs = JX[seed, :, :].T @ (SY[seed, :] - SX[seed, :])\n",
    "        G_jtfs.append(g_jtfs)\n",
    "    G_jtfs = torch.stack(G_jtfs, dim=0).squeeze()\n",
    "\n",
    "    # Take gradient with respect to coordinate 1 (slope)\n",
    "    H_jtfs.append(G_jtfs[:, 1].detach())\n",
    "    H_scrapl.append(G_scrapl[:, 1].detach())\n",
    "\n",
    "H_jtfs = torch.stack(H_jtfs)\n",
    "H_scrapl = torch.stack(H_scrapl)\n",
    "\n",
    "positions = np.arange(-1, 6)\n",
    "plt.boxplot(np.cbrt(H_jtfs), positions=positions-1/6, widths=1/3,\n",
    "    patch_artist=True, boxprops=dict(facecolor='orange'));\n",
    "plt.boxplot(np.cbrt(H_scrapl), positions=positions+1/6,widths=1/3,\n",
    "    patch_artist=True, boxprops=dict(facecolor='green'));\n",
    "\n",
    "plt.gca().set_xticks(positions)\n",
    "plt.gca().set_xticklabels(positions)\n",
    "plt.grid(linestyle='--')\n",
    "plt.xlabel(\"Relative Slope Index\")\n",
    "plt.ylabel(\"Slope gradient (cbrt)\")\n",
    "# Microstructure means matched random seeds between\n",
    "# target and prediction.\n",
    "plt.title(\"Microstructure. Orange: JTFS. Green: SCRAPL\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Mesostructure. Orange: JTFS. Green: SCRAPL')"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "H_jtfs = []\n",
    "H_scrapl = []\n",
    "common_density_idx = 3\n",
    "y_slope_idx = 2\n",
    "data_dir = \"icassp2024_scrapl_data\"\n",
    "Sy = load_jtfs(common_density_idx, y_slope_idx, seed=0, data_dir=data_dir)\n",
    "\n",
    "for x_slope_idx in range(7):\n",
    "\n",
    "    SX, JX = [], []\n",
    "    for seed in range(1, 10):\n",
    "        x_coords = (common_density_idx, x_slope_idx)\n",
    "        Sx = load_jtfs(*x_coords, seed=seed, data_dir=data_dir)\n",
    "        Jx = load_jacobian(*x_coords, seed=seed, data_dir=data_dir)\n",
    "        SX.append(Sx)\n",
    "        JX.append(Jx)\n",
    "    SX = torch.stack(SX, dim=0)\n",
    "    JX = torch.stack(JX, dim=0)\n",
    "\n",
    "    G_scrapl = []\n",
    "    for _ in range(100):\n",
    "        Sx_scrapl, Sy_scrapl, Jx_scrapl = [], [], []\n",
    "        for path in range(SX.shape[-1]):\n",
    "            random_seed = torch.randint(SX.shape[0], (1,))\n",
    "            Jx_scrapl.append(JX[random_seed, path, :].squeeze())\n",
    "            Sx_scrapl.append(SX[random_seed, path].squeeze())\n",
    "        Jx_scrapl = torch.stack(Jx_scrapl, dim=0)\n",
    "        Sx_scrapl = torch.stack(Sx_scrapl, dim=0)\n",
    "        g_scrapl = Jx_scrapl.T @ (Sy - Sx_scrapl)\n",
    "        G_scrapl.append(g_scrapl)\n",
    "    G_scrapl = torch.stack(G_scrapl, dim=0).squeeze()\n",
    "\n",
    "    G_jtfs = []\n",
    "    for seed in range(SX.shape[0]):\n",
    "        g_jtfs = JX[seed, :, :].T @ (Sy - SX[seed, :])\n",
    "        G_jtfs.append(g_jtfs)\n",
    "    G_jtfs = torch.stack(G_jtfs, dim=0).squeeze()\n",
    "\n",
    "    # Take gradient with respect to coordinate 1 (slope)\n",
    "    H_jtfs.append(G_jtfs[:, 1].detach())\n",
    "    H_scrapl.append(G_scrapl[:, 1].detach())\n",
    "\n",
    "H_jtfs = torch.stack(H_jtfs)\n",
    "H_scrapl = torch.stack(H_scrapl)\n",
    "\n",
    "positions = np.arange(-3, 4)\n",
    "plt.boxplot(np.cbrt(H_jtfs), positions=positions-1/6, widths=1/3,\n",
    "    patch_artist=True, boxprops=dict(facecolor='orange'));\n",
    "plt.boxplot(np.cbrt(H_scrapl), positions=positions+1/6,widths=1/3,\n",
    "    patch_artist=True, boxprops=dict(facecolor='green'));\n",
    "\n",
    "plt.gca().set_xticks(positions)\n",
    "plt.gca().set_xticklabels(positions)\n",
    "plt.grid(linestyle='--')\n",
    "plt.xlabel(\"Relative Slope Index\")\n",
    "plt.ylabel(\"Slope gradient (cbrt)\")\n",
    "# Mesostructure means that the target has seed=0\n",
    "# whereas predictions have seed>0 between 1 and 9\n",
    "plt.title(\"Mesostructure. Orange: JTFS. Green: SCRAPL\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "ename": "IndexError",
     "evalue": "too many indices for tensor of dimension 1",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mIndexError\u001b[0m                                Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[19], line 6\u001b[0m\n\u001b[1;32m      3\u001b[0m data_dir \u001b[39m=\u001b[39m \u001b[39m\"\u001b[39m\u001b[39micassp2024_scrapl_data\u001b[39m\u001b[39m\"\u001b[39m\n\u001b[1;32m      5\u001b[0m y_coords \u001b[39m=\u001b[39m (\u001b[39m3\u001b[39m, \u001b[39m3\u001b[39m)\n\u001b[0;32m----> 6\u001b[0m Sy \u001b[39m=\u001b[39m load_jtfs(\u001b[39m*\u001b[39;49my_coords, seed\u001b[39m=\u001b[39;49m\u001b[39m0\u001b[39;49m, data_dir\u001b[39m=\u001b[39;49mdata_dir)[:, \u001b[39m0\u001b[39;49m]\n\u001b[1;32m      8\u001b[0m \u001b[39mfor\u001b[39;00m density_idx \u001b[39min\u001b[39;00m \u001b[39mrange\u001b[39m(\u001b[39m7\u001b[39m):\n\u001b[1;32m     10\u001b[0m     SX, JX \u001b[39m=\u001b[39m [], []\n",
      "\u001b[0;31mIndexError\u001b[0m: too many indices for tensor of dimension 1"
     ]
    }
   ],
   "source": [
    "H_jtfs = []\n",
    "H_scrapl = []\n",
    "data_dir = \"icassp2024_scrapl_data\"\n",
    "\n",
    "y_coords = (3, 3)\n",
    "Sy = load_jtfs(*y_coords, seed=0, data_dir=data_dir)[:, 0]\n",
    "\n",
    "for density_idx in range(7):\n",
    "\n",
    "    SX, JX = [], []\n",
    "    for seed in range(1, 10):\n",
    "        x_coords = (3, density_idx)\n",
    "        Sx = load_jtfs(*x_coords, seed=seed, data_dir=data_dir)[:, 0]\n",
    "        Jx = load_jacobian(*x_coords, seed=seed, data_dir=data_dir)\n",
    "        SX.append(Sx)\n",
    "        JX.append(Jx)\n",
    "    SX = torch.stack(SX, dim=0)\n",
    "    JX = torch.stack(JX, dim=0)\n",
    "\n",
    "    G_scrapl = []\n",
    "    for _ in range(100):\n",
    "        Sx_scrapl, Jx_scrapl = [], []\n",
    "        for path in range(SX.shape[-1]):\n",
    "            random_seed = torch.randint(SX.shape[0], (1,))\n",
    "            Jx_scrapl.append(JX[random_seed, path, :])\n",
    "            Sx_scrapl.append(SX[random_seed, path])\n",
    "        Jx_scrapl = torch.stack(Jx_scrapl, dim=0).squeeze()\n",
    "        Sx_scrapl = torch.stack(Sx_scrapl, dim=0).squeeze()\n",
    "        g_scrapl = Jx_scrapl.T @ (Sy - Sx_scrapl)\n",
    "        G_scrapl.append(g_scrapl)\n",
    "    G_scrapl = torch.stack(G_scrapl, dim=0).squeeze()\n",
    "\n",
    "    G_jtfs = []\n",
    "    for seed in range(9):\n",
    "        g_jtfs = JX[seed, :, :].T @ (Sy - SX[seed, :])\n",
    "        G_jtfs.append(g_jtfs)\n",
    "    G_jtfs = torch.stack(G_jtfs, dim=0).squeeze()\n",
    "\n",
    "    H_jtfs.append(G_jtfs[:, 1].detach())\n",
    "    H_scrapl.append(G_scrapl[:, 1].detach())\n",
    "\n",
    "H_jtfs = torch.stack(H_jtfs)\n",
    "H_scrapl = torch.stack(H_scrapl)\n",
    "\n",
    "positions = np.arange(-3, 4)\n",
    "# Show this one in orange\n",
    "plt.boxplot(H_jtfs, positions=positions-1/6, widths=1/3,\n",
    "    patch_artist=True, boxprops=dict(facecolor='orange'));\n",
    "plt.boxplot(H_scrapl, positions=positions+1/6,widths=1/3,\n",
    "    patch_artist=True, boxprops=dict(facecolor='green'));\n",
    "plt.gca().set_xticks(positions)\n",
    "plt.gca().set_xticklabels(positions)\n",
    "plt.grid(linestyle='--')\n",
    "plt.xlabel(\"Relative Density Index\")\n",
    "plt.ylabel(\"Density gradient\")\n",
    "# Mesostructure means that the target has seed=0\n",
    "# whereas predictions have seed>0 between 1 and 9\n",
    "plt.title(\"Mesostructure. Orange: JTFS. Green: SCRAPL\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "SX.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "density_idx = 6\n",
    "\n",
    "data_dir = \"icassp2024_scrapl_data\"\n",
    "SX, JX, SY = [], [], []\n",
    "for seed in range(1, 10):\n",
    "    y_coords = (4, 3)\n",
    "    x_coords = (4, density_idx)\n",
    "    Sx = load_jtfs(*x_coords, seed=seed, data_dir=data_dir)[:, 0]\n",
    "    Sy = load_jtfs(*y_coords, seed=seed, data_dir=data_dir)[:, 0]\n",
    "    Jx = load_jacobian(*x_coords, seed=seed, data_dir=data_dir)\n",
    "    SX.append(Sx)\n",
    "    SY.append(Sy)\n",
    "    JX.append(Jx)\n",
    "SX = torch.stack(SX, dim=0)\n",
    "SY = torch.stack(SY, dim=0)\n",
    "JX = torch.stack(JX, dim=0)\n",
    "\n",
    "G_scrapl = []\n",
    "for _ in range(100):\n",
    "    Sx_scrapl, Sy_scrapl, Jx_scrapl = [], [], []\n",
    "    for path in range(SX.shape[-1]):\n",
    "        random_seed = torch.randint(SX.shape[0], (1,))\n",
    "        Jx_scrapl.append(JX[random_seed, path, :])\n",
    "        Sx_scrapl.append(SX[random_seed, path])\n",
    "        Sy_scrapl.append(SY[random_seed, path])\n",
    "    Jx_scrapl = torch.stack(Jx_scrapl, dim=0).squeeze()\n",
    "    Sx_scrapl = torch.stack(Sx_scrapl, dim=0).squeeze()\n",
    "    Sy_scrapl = torch.stack(Sy_scrapl, dim=0).squeeze()\n",
    "    g_scrapl = Jx_scrapl.T @ (Sy_scrapl - Sx_scrapl)\n",
    "    G_scrapl.append(g_scrapl)\n",
    "G_scrapl = torch.stack(G_scrapl, dim=0).squeeze()\n",
    "\n",
    "G_jtfs = []\n",
    "for seed in range(9):\n",
    "    y_coords = (4, 3)\n",
    "    Sy = load_jtfs(*y_coords, seed=seed, data_dir=data_dir)[:, 0]\n",
    "    g_jtfs = JX[seed, :, :].T @ (Sy - SX[seed, :])\n",
    "    G_jtfs.append(g_jtfs)\n",
    "G_jtfs = torch.stack(G_jtfs, dim=0).squeeze()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Sx.shape"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "scrapl",
   "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": "3.10.11"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}