cyber_noisy_test.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torchvision.transforms as transforms\n",
    "from torchvision.utils import save_image, make_grid\n",
    "\n",
    "from torch.utils.data import DataLoader\n",
    "from torchvision import datasets\n",
    "from torch.autograd import Variable\n",
    "\n",
    "from models import *\n",
    "from datasets import *\n",
    "from utils import *\n",
    "\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import torch\n",
    "\n",
    "from dataset.time_dataset import TsFinalDataset\n",
    "from pathlib import Path\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from dtw import dtw"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "noise_type='gaussian'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The size of the dataset is:  torch.Size([1, 2045, 33]) torch.Size([1, 2045, 33])\n"
     ]
    }
   ],
   "source": [
    "# Dataset loader\n",
    "datapath = Path('data')\n",
    "dataset = TsFinalDataset(datapath/'CaseI-Attacks without any change.csv', noise_type=noise_type)\n",
    "dataloader = torch.utils.data.DataLoader(dataset, batch_size=1,\n",
    "                                        shuffle=True, num_workers=4)\n",
    "print(\"The size of the dataset is: \", dataset.data_normal.size(), dataset.data_attacked.size())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "normal_seq_len = dataset.data_normal.size(1)\n",
    "normal_feature_len = dataset.data_normal.size(2)\n",
    "abnormal_seq_len = dataset.data_attacked.size(1)\n",
    "abnormal_feature_len = dataset.data_attacked.size(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "G_AB = LSTMFullGenerator(normal_seq_len, normal_feature_len)\n",
    "G_BA = LSTMFullGenerator(abnormal_seq_len, abnormal_feature_len)\n",
    "D_A = LSTMFullDiscriminator(normal_feature_len)\n",
    "D_B = LSTMFullDiscriminator(normal_feature_len)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<All keys matched successfully>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "G_AB.load_state_dict(torch.load(f\"saved_models/true/{noise_type}/G_AB_190.pth\"))\n",
    "G_BA.load_state_dict(torch.load(f\"saved_models/true/{noise_type}/G_BA_190.pth\"))\n",
    "D_A.load_state_dict(torch.load(f\"saved_models/true/{noise_type}/D_A_190.pth\"))\n",
    "D_B.load_state_dict(torch.load(f\"saved_models/true/{noise_type}/D_B_190.pth\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LSTMFullDiscriminator(\n",
       "  (lstm): LSTM(33, 256, batch_first=True)\n",
       "  (linear): Sequential(\n",
       "    (0): Linear(in_features=256, out_features=33, bias=True)\n",
       "    (1): Sigmoid()\n",
       "  )\n",
       ")"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "G_AB.eval()\n",
    "G_BA.eval()\n",
    "D_A.eval()\n",
    "D_B.eval()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "threshold = 1.007"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([15, 1973, 33])"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset_test = TsFinalDataset(datapath/'CaseII.csv', noise_type='none')\n",
    "# dataset_test = TsFullDataset(datapath/'CaseI-Attacks without any change.csv')\n",
    "dataset_test.data_normal.size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Output size: torch.Size([15, 1973, 33]) \t Output Back size: torch.Size([15, 1973, 33])\n"
     ]
    }
   ],
   "source": [
    "with torch.no_grad():\n",
    "    output = G_AB(dataset_test.data_normal)\n",
    "    output_back = G_BA(output)\n",
    "print(f\"Output size: {output.size()} \\t Output Back size: {output_back.size()}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+ 1.000450552578472\n",
      "+ 1.0004505525314251\n",
      "+ 1.0004505525314251\n",
      "+ 1.0004505525314251\n",
      "+ 1.0004505525314251\n",
      "+ 1.0004505524961187\n",
      "+ 1.0004505525314251\n",
      "+ 1.0004505525314251\n",
      "+ 1.0004505525314251\n",
      "+ 1.0004505525314251\n",
      "+ 1.0004505525314251\n",
      "+ 1.0004505525314251\n",
      "+ 1.0004505525314251\n",
      "+ 1.0004505525314251\n",
      "+ 1.0004505525196634\n"
     ]
    }
   ],
   "source": [
    "os.makedirs(f'cyber_result/true/{noise_type}/forward', exist_ok=True)\n",
    "forward_res_list = []\n",
    "for i in range(output.size(0)):\n",
    "    plt.plot(output[i, :, 0].numpy())\n",
    "    plt.plot(output_back[i, :, 0].numpy())\n",
    "    plt.plot(dataset_test.data_normal[i, :, 0].numpy())\n",
    "    plt.legend(['Generated', 'Generated Back', 'Original'])\n",
    "    plt.savefig(f'cyber_result/true/{noise_type}/forward/output_{i}.png')\n",
    "    plt.close()\n",
    "\n",
    "    dist1, cost, acc, path = dtw(output[i,:,0].reshape(-1, 1), dataset_test.data_normal[i,:,0].reshape(-1, 1), dist=lambda x, y: np.linalg.norm(x - y, ord=1))\n",
    "    dist2, cost, acc, path = dtw(output_back[i,:,0].reshape(-1, 1), output[i,:,0].reshape(-1, 1), dist=lambda x, y: np.linalg.norm(x - y, ord=1))\n",
    "    dist3, cost, acc, path = dtw(output_back[i,:,0].reshape(-1, 1), dataset_test.data_normal[i,:,0].reshape(-1, 1), dist=lambda x, y: np.linalg.norm(x - y, ord=1))\n",
    "    res = (dist2+dist3)/dist1\n",
    "    print(\"+\" if res < threshold else \"-\", res)\n",
    "    forward_res_list.append(res)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkgAAAHHCAYAAABEEKc/AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/H5lhTAAAACXBIWXMAAA9hAAAPYQGoP6dpAABD1ElEQVR4nO3dfVhUdf7/8dcAMpAKRipIgpCWqRXes2TeJUllmle3lhbh1za/tZZiqWRpu22Rbje6wdqNm2Q3+9Xtu7gWhUuoqYWastTStyyLFG8A/amDgqIy5/fHXIzNkTuJmQF9Pq7rXDVn3pzP+5w5nHl55szBYhiGIQAAADj5eLsBAACAloaABAAAYEJAAgAAMCEgAQAAmBCQAAAATAhIAAAAJgQkAAAAEwISAACACQEJAADAhIAEoFUZMWKERowY4ZZlr1+/XhaLRevXr2907QcffOCWXgB4FwEJOM9kZGTIYrEoICBAe/fuPev5ESNG6KqrrvJCZ63T+++/r0WLFjX7cmtep5rJz89Pl156qR544IE6X7df1v9yuvLKK+scp76f++X0zDPPNMt6/eUvf1FGRkazLAvwJj9vNwDAPaqqqvTCCy/o1Vdf9XYrrcawYcN0/Phx+fv7O+e9//77Kiws1PTp090y5h/+8AdFR0frxIkT2rx5szIyMrRp0yYVFhYqICDApbZr165KTU09axnBwcF1Ln/u3LmaMmWK8/GXX36pP//5z3ryySfVq1cv5/xrrrmmGdbGEZA6duyoBx54oFmWB3gLAQk4T/Xt21dvvvmmUlJSFB4e7pYxDMPQiRMnFBgY6Jble5qPj89ZocTdbrrpJg0cOFCSNGXKFHXs2FELFizQ6tWrddddd7nUBgcHa9KkSee0/BtuuMHlcUBAgP785z/rhhtucNtHlcD5gI/YgPPUk08+qerqar3wwgsN1p4+fVrPPvusunfvLqvVqqioKD355JOqqqpyqYuKitItt9yiNWvWaODAgQoMDNTrr7/uvB5n5cqV+v3vf69LL71U7du31x133CGbzaaqqipNnz5dnTt3Vrt27ZSUlHTWspctW6brr79enTt3ltVqVe/evbVkyZImrfttt92m/v37u8wbO3asLBaLVq9e7Zy3ZcsWWSwWffLJJ5LOvgZpxIgRysrK0q5du5wfRUVFRbks126367nnnlPXrl0VEBCgUaNGaefOnU3qW5KGDh0qSfrxxx+bvIym+OSTTzR06FC1bdtW7du315gxY/TNN9+41JSUlCgpKUldu3aV1WpVly5ddOutt+rnn3+W5Ng/vvnmG3322WfO7UUIQ2vFGSTgPBUdHa37779fb775pubMmVPvWaQpU6bo7bff1h133KGZM2dqy5YtSk1N1bfffqvMzEyX2h07duiee+7RQw89pAcffFA9e/Z0PpeamqrAwEDNmTNHO3fu1Kuvvqo2bdrIx8dHhw8f1jPPPOP8GCk6Olrz5s1z/uySJUvUp08fjRs3Tn5+fvrwww/18MMPy26365FHHjmndR86dKj++c9/qry8XEFBQTIMQ59//rl8fHy0ceNGjRs3TpK0ceNG+fj4aMiQIbUuZ+7cubLZbNqzZ49eeeUVSVK7du1cal544QX5+Pjo8ccfl81m08KFCzVx4kRt2bLlnHquURM2Lr744rOeq66u1sGDB8+aHxgYqLZt2zZpPEl65513lJiYqISEBC1YsECVlZVasmSJrrvuOv373/92hsLbb79d33zzjaZNm6aoqCiVlZUpJydHu3fvVlRUlBYtWqRp06apXbt2mjt3riQpNDS0yX0BXmUAOK8sW7bMkGR8+eWXxo8//mj4+fkZjz76qPP54cOHG3369HE+LigoMCQZU6ZMcVnO448/bkgy1q5d65zXrVs3Q5KRnZ3tUrtu3TpDknHVVVcZJ0+edM6/5557DIvFYtx0000u9XFxcUa3bt1c5lVWVp61LgkJCcZll13mMm/48OHG8OHD690GX375pSHJ+Pjjjw3DMIyvv/7akGTceeedRmxsrLNu3LhxRr9+/c5aj3Xr1jnnjRkz5qxef1nbq1cvo6qqyjl/8eLFhiTjP//5T7091rxOn376qXHgwAGjuLjY+OCDD4xOnToZVqvVKC4uPmu9JdU6PfTQQ/WO9Ut///vfXdbx6NGjRocOHYwHH3zQpa6kpMQIDg52zj98+LAhyfjTn/5U7/L79OnT4OsDtAZ8xAacxy677DLdd999euONN7R///5aaz7++GNJUnJyssv8mTNnSpKysrJc5kdHRyshIaHWZd1///1q06aN83FsbKwMw9DkyZNd6mJjY1VcXKzTp0875/3yOiabzaaDBw9q+PDh+umnn2Sz2RpaVRf9+vVTu3bttGHDBkmOM0Vdu3bV/fffr/z8fFVWVsowDG3atMn5kVZTJSUluVzUXbO8n376qVE/Hx8fr06dOikiIkJ33HGH2rZtq9WrV6tr165n1UZFRSknJ+es6ddcQJ6Tk6MjR47onnvu0cGDB52Tr6+vYmNjtW7dOkmO18ff31/r16/X4cOHmzwe0FrwERtwnnvqqaf0zjvv6IUXXtDixYvPen7Xrl3y8fFRjx49XOaHhYWpQ4cO2rVrl8v86OjoOseKjIx0eVzz7aqIiIiz5tvtdtlsNl1yySWSpM8//1zz589XXl6eKisrXeptNlu939Qy8/X1VVxcnDZu3CjJEZCGDh2q6667TtXV1dq8ebNCQ0N16NChXx2QzOtc89FYY0NEenq6rrjiCtlsNr311lvasGGDrFZrrbVt27ZVfHz8r+rX7IcffpAkXX/99bU+HxQUJEmyWq1asGCBZs6cqdDQUP3mN7/RLbfcovvvv19hYWHN2hPQEhCQgPPcZZddpkmTJumNN97QnDlz6qyzWCyNWl5931jz9fU9p/mGYUhyXJA8atQoXXnllXr55ZcVEREhf39/ffzxx3rllVdkt9sb1dsvXXfddXruued04sQJbdy4UXPnzlWHDh101VVXaePGjc5rY35tQGpo3RoyePBg57fYxo8fr+uuu0733nuvduzYcdb1Tu5Qs23feeedWoOOn9+Zt4np06dr7NixWrVqldasWaOnn35aqampWrt2rfr16+f2XgFPIiABF4CnnnpK7777rhYsWHDWc926dZPdbtcPP/zgcl+c0tJSHTlyRN26dXN7fx9++KGqqqq0evVqlzMyNR/vNMXQoUN18uRJ/e1vf9PevXudQWjYsGHOgHTFFVc0eBFxY4Njc/D19VVqaqpGjhyptLS0egNtc+nevbskqXPnzo06O9W9e3fNnDlTM2fO1A8//KC+ffvqpZde0rvvvivJs9sLcCeuQQIuAN27d9ekSZP0+uuvq6SkxOW5m2++WZLOulv0yy+/LEkaM2aM2/urOQvzy7MuNptNy5Yta/IyY2Nj1aZNGy1YsEAhISHq06ePJEdw2rx5sz777LNGnT1q27btOV8D9WuMGDFCgwcP1qJFi3TixAm3j5eQkKCgoCA9//zzOnXq1FnPHzhwQJJUWVl5Vj/du3dX+/btXW7Z0LZtWx05csStPQOewBkk4AIxd+5cvfPOO9qxY4czLEhSTEyMEhMT9cYbb+jIkSMaPny4tm7dqrffflvjx4/XyJEj3d7b6NGj5e/vr7Fjx+qhhx7SsWPH9Oabb6pz5851XlzekIsuukgDBgzQ5s2bnfdAkhxnkCoqKlRRUdGogDRgwACtWLFCycnJGjRokNq1a6exY8c2qafGeuKJJ3TnnXcqIyNDU6dOdc632WzOMzVm53oDyRpBQUFasmSJ7rvvPvXv318TJkxQp06dtHv3bmVlZWnIkCFKS0vT999/r1GjRumuu+5S79695efnp8zMTJWWlmrChAnO5Q0YMEBLlizRH//4R/Xo0UOdO3eu8/omoCUjIAEXiB49emjSpEl6++23z3pu6dKluuyyy5SRkaHMzEyFhYUpJSVF8+fP90hvPXv21AcffKCnnnpKjz/+uMLCwvTf//3f6tSp01nfgDsXNWeLrrvuOue8sLAw9ejRQzt37mxUQHr44YdVUFCgZcuW6ZVXXlG3bt3cHpBuu+02de/eXS+++KIefPBB5xm2PXv26L777qv1Z5oakCTp3nvvVXh4uF544QX96U9/UlVVlS699FINHTpUSUlJkhwX2t9zzz3Kzc3VO++8Iz8/P1155ZVauXKlbr/9duey5s2bp127dmnhwoU6evSohg8fTkBCq2QxGnslIQAAwAWCa5AAAABMCEgAAAAmBCQAAAATAhIAAIAJAQkAAMCEgAQAAGDCfZCayG63a9++fWrfvj231gcAoJUwDENHjx5VeHi4fHzqPk9EQGqiffv2nfUXygEAQOtQXFysrl271vk8AamJ2rdvL8mxgYOCgrzcDQAAaIzy8nJFREQ438frQkBqopqP1YKCgghIAAC0Mg1dHsNF2gAAACYEJAAAABMCEgAAgAkBCQAAwISABAAAYEJAAgAAMCEgAQAAmBCQAAAATAhIAAAAJgQkAAAAE68GpA0bNmjs2LEKDw+XxWLRqlWrGvyZ9evXq3///rJarerRo4cyMjLOqklPT1dUVJQCAgIUGxurrVu3Op/7+eefZbFYap3+/ve/N+PaAQCA1sqrAamiokIxMTFKT09vVH1RUZHGjBmjkSNHqqCgQNOnT9eUKVO0Zs0aZ82KFSuUnJys+fPnKz8/XzExMUpISFBZWZkkKSIiQvv373eZfv/736tdu3a66aab3LKeAACgdbEYhmF4uwnJ8UfjMjMzNX78+DprZs+eraysLBUWFjrnTZgwQUeOHFF2drYkKTY2VoMGDVJaWpokyW63KyIiQtOmTdOcOXNqXW6/fv3Uv39//fWvf210v+Xl5QoODpZt377a/1itr68UEHDmcUVF3Qvz8ZECA5tWW1kp1fUSWizSRRc1rfb4cclur7uPtm2bVnvihFRd3Ty1F13k6FuSqqqk06ebpzYw0LGdJenkSenUqeapDQhw7BfnWnvqlKO+Llar5Od37rWnTzu2RV38/aU2bc69trra8drVpU0bR/251trtjn2tOWr9/BzbQnL8TlRWNk/tufzec4yovZZjxLnXcoxw/H8jjxHO92+brf4/Nm+0EJKMzMzMemuGDh1qPPbYYy7z3nrrLSMoKMgwDMOoqqoyfH19z1rO/fffb4wbN67WZW7bts2QZHz++ef1jn3ixAnDZrM5p+LiYkOSYXMcTs6ebr7ZdQEXXVR7nWQYw4e71nbsWHftwIGutd261V3bu7drbe/eddd26+ZaO3Bg3bUdO7rWDh9ed+1FF7nW3nxz3bXm3fGOO+qvPXbsTG1iYv21ZWVnah9+uP7aoqIztY8/Xn9tYeGZ2vnz66/duvVM7cKF9deuW3emNi2t/tqPPjpTu2xZ/bUrV56pXbmy/tply87UfvRR/bVpaWdq162rv3bhwjO1W7fWXzt//pnawsL6ax9//ExtUVH9tQ8/fKa2rKz+2sTEM7XHjtVfe8cdhov6ajlGOCaOEWcmjhGOyc3HCJvNZkgybDabUZ9WdZF2SUmJQkNDXeaFhoaqvLxcx48f18GDB1VdXV1rTUlJSa3L/Otf/6pevXrp2muvrXfs1NRUBQcHO6eIiIhftzIAAKDFalUfsV1xxRVKSkpSSkqKc97HH3+sMWPGqLKyUocPH9all16qL774QnFxcc6aWbNm6bPPPtOWLVtclnf8+HF16dJFTz/9tGbOnFlvf1VVVar6xenD8vJyRURE8BHbudZy+vzcazl97vh/PmJrWi3HCMf/c4w499rz9BjR2I/Y/OpeUssTFham0tJSl3mlpaUKCgpSYGCgfH195evrW2tNWFjYWcv74IMPVFlZqfvvv7/Bsa1Wq6w1B8lfatvW9Re2Lo2paUrtLw9YzVn7ywNsc9b+8g2hOWut1jNvYs1Z6+9/5pfPW7Vt2pw5sDRnrZ/fmQNhc9b6+jZ+Hz6XWh8f99RaLO6plVpGLccIB44R5157Ph8jGrO4ZluSB8TFxSk3N9dlXk5OjvNskb+/vwYMGOBSY7fblZub63JGqcZf//pXjRs3Tp06dXJv4wAAoFXx6hmkY8eOaefOnc7HRUVFKigoUEhIiCIjI5WSkqK9e/dq+fLlkqSpU6cqLS1Ns2bN0uTJk7V27VqtXLlSWVlZzmUkJycrMTFRAwcO1ODBg7Vo0SJVVFQoKSnJZeydO3dqw4YN+vjjjz2zsgAAoNXwakDatm2bRo4c6XycnJwsSUpMTFRGRob279+v3bt3O5+Pjo5WVlaWZsyYocWLF6tr165aunSpEhISnDV33323Dhw4oHnz5qmkpER9+/ZVdnb2WRduv/XWW+ratatGjx7t5rUEAACtTYu5SLu1afR9FAAAQIvR2PfvVnUNEgAAgCcQkAAAAEwISAAAACYEJAAAABMCEgAAgAkBCQAAwISABAAAYEJAAgAAMCEgAQAAmBCQAAAATAhIAAAAJgQkAAAAEwISAACACQEJAADAhIAEAABgQkACAAAwISABAACYEJAAAABMCEgAAAAmBCQAAAATAhIAAIAJAQkAAMCEgAQAAGBCQAIAADAhIAEAAJgQkAAAAEwISAAAACYEJAAAABMCEgAAgAkBCQAAwISABAAAYEJAAgAAMCEgAQAAmBCQAAAATAhIAAAAJgQkAAAAEwISAACACQEJAADAhIAEAABgQkACAAAw8WpA2rBhg8aOHavw8HBZLBatWrWqwZ9Zv369+vfvL6vVqh49eigjI+OsmvT0dEVFRSkgIECxsbHaunXrWTV5eXm6/vrr1bZtWwUFBWnYsGE6fvx4M6wVAABo7bwakCoqKhQTE6P09PRG1RcVFWnMmDEaOXKkCgoKNH36dE2ZMkVr1qxx1qxYsULJycmaP3++8vPzFRMTo4SEBJWVlTlr8vLydOONN2r06NHaunWrvvzyS/3ud7+Tjw8n1AAAgGQxDMPwdhOSZLFYlJmZqfHjx9dZM3v2bGVlZamwsNA5b8KECTpy5Iiys7MlSbGxsRo0aJDS0tIkSXa7XREREZo2bZrmzJkjSfrNb36jG264Qc8++2yT+y0vL1dwcLBsNpuCgoKavBwAAOA5jX3/blWnTPLy8hQfH+8yLyEhQXl5eZKkkydPavv27S41Pj4+io+Pd9aUlZVpy5Yt6ty5s6699lqFhoZq+PDh2rRpU71jV1VVqby83GUCAADnp1YVkEpKShQaGuoyLzQ0VOXl5Tp+/LgOHjyo6urqWmtKSkokST/99JMk6ZlnntGDDz6o7Oxs9e/fX6NGjdIPP/xQ59ipqakKDg52ThEREc28dgAAoKVoVQGpOdjtdknSQw89pKSkJPXr10+vvPKKevbsqbfeeqvOn0tJSZHNZnNOxcXFnmoZAAB4mJ+3GzgXYWFhKi0tdZlXWlqqoKAgBQYGytfXV76+vrXWhIWFSZK6dOkiSerdu7dLTa9evbR79+46x7ZarbJarc2xGgAAoIVrVWeQ4uLilJub6zIvJydHcXFxkiR/f38NGDDApcZutys3N9dZExUVpfDwcO3YscNlOd9//726devm5jUAAACtgVfPIB07dkw7d+50Pi4qKlJBQYFCQkIUGRmplJQU7d27V8uXL5ckTZ06VWlpaZo1a5YmT56stWvXauXKlcrKynIuIzk5WYmJiRo4cKAGDx6sRYsWqaKiQklJSZIc35Z74oknNH/+fMXExKhv3756++239d133+mDDz7w7AYAAAAtklcD0rZt2zRy5Ejn4+TkZElSYmKiMjIytH//fpePvaKjo5WVlaUZM2Zo8eLF6tq1q5YuXaqEhARnzd13360DBw5o3rx5KikpUd++fZWdne1y4fb06dN14sQJzZgxQ4cOHVJMTIxycnLUvXt3D6w1AABo6VrMfZBaG+6DBABA63Ne3gcJAADAEwhIAAAAJgQkAAAAEwISAACACQEJAADAhIAEAABgQkACAAAwISABAACYEJAAAABMCEgAAAAmBCQAAAATAhIAAIAJAQkAAMCEgAQAAGBCQAIAADAhIAEAAJgQkAAAAEwISAAAACYEJAAAABMCEgAAgAkBCQAAwISABAAAYEJAAgAAMCEgAQAAmBCQAAAATAhIAAAAJgQkAAAAEwISAACACQEJAADAhIAEAABgQkACAAAwISABAACYEJAAAABMCEgAAAAmBCQAAAATAhIAAIAJAQkAAMCEgAQAAGBCQAIAADAhIAEAAJh4NSBt2LBBY8eOVXh4uCwWi1atWtXgz6xfv179+/eX1WpVjx49lJGRcVZNenq6oqKiFBAQoNjYWG3dutXl+REjRshisbhMU6dObaa1AgAArZ1XA1JFRYViYmKUnp7eqPqioiKNGTNGI0eOVEFBgaZPn64pU6ZozZo1zpoVK1YoOTlZ8+fPV35+vmJiYpSQkKCysjKXZT344IPav3+/c1q4cGGzrhsAAGi9LIZhGN5uQpIsFosyMzM1fvz4Omtmz56trKwsFRYWOudNmDBBR44cUXZ2tiQpNjZWgwYNUlpamiTJbrcrIiJC06ZN05w5cyQ5ziD17dtXixYtanK/5eXlCg4Ols1mU1BQUJOXAwAAPKex79+t6hqkvLw8xcfHu8xLSEhQXl6eJOnkyZPavn27S42Pj4/i4+OdNTXee+89dezYUVdddZVSUlJUWVnp/hUAAACtgp+3GzgXJSUlCg0NdZkXGhqq8vJyHT9+XIcPH1Z1dXWtNd99953z8b333qtu3bopPDxcX3/9tWbPnq0dO3boH//4R51jV1VVqaqqyvm4vLy8mdYKAAC0NK0qIDWX3/72t87/v/rqq9WlSxeNGjVKP/74o7p3717rz6Smpur3v/+9p1oEAABe1Ko+YgsLC1NpaanLvNLSUgUFBSkwMFAdO3aUr69vrTVhYWF1Ljc2NlaStHPnzjprUlJSZLPZnFNxcfGvWBMAANCStaqAFBcXp9zcXJd5OTk5iouLkyT5+/trwIABLjV2u125ubnOmtoUFBRIkrp06VJnjdVqVVBQkMsEAADOT179iO3YsWMuZ22KiopUUFCgkJAQRUZGKiUlRXv37tXy5cslSVOnTlVaWppmzZqlyZMna+3atVq5cqWysrKcy0hOTlZiYqIGDhyowYMHa9GiRaqoqFBSUpIk6ccff9T777+vm2++WZdccom+/vprzZgxQ8OGDdM111zj2Q0AAABaJK8GpG3btmnkyJHOx8nJyZKkxMREZWRkaP/+/dq9e7fz+ejoaGVlZWnGjBlavHixunbtqqVLlyohIcFZc/fdd+vAgQOaN2+eSkpK1LdvX2VnZzsv3Pb399enn37qDE4RERG6/fbb9dRTT3lorQEAQEvXYu6D1NpwHyQAAFqf8/I+SAAAAJ5AQAIAADAhIAEAAJgQkAAAAEwISAAAACYEJAAAABMCEgAAgAkBCQAAwISABAAAYEJAAgAAMCEgAQAAmBCQAAAATAhIAAAAJgQkAAAAEwISAACACQEJAADAhIAEAABgQkACAAAwISABAACYEJAAAABMCEgAAAAmBCQAAAATAhIAAIAJAQkAAMCEgAQAAGBCQAIAADAhIAEAAJgQkAAAAEwISAAAACYEJAAAABMCEgAAgAkBCQAAwISABAAAYEJAAgAAMCEgAQAAmBCQAAAATAhIAAAAJgQkAAAAEwISAACACQEJAADAhIAEAABg4tWAtGHDBo0dO1bh4eGyWCxatWpVgz+zfv169e/fX1arVT169FBGRsZZNenp6YqKilJAQIBiY2O1devWWpdlGIZuuummRo8NAAAuDF4NSBUVFYqJiVF6enqj6ouKijRmzBiNHDlSBQUFmj59uqZMmaI1a9Y4a1asWKHk5GTNnz9f+fn5iomJUUJCgsrKys5a3qJFi2SxWJptfQAAwPnBYhiG4e0mJMlisSgzM1Pjx4+vs2b27NnKyspSYWGhc96ECRN05MgRZWdnS5JiY2M1aNAgpaWlSZLsdrsiIiI0bdo0zZkzx/lzBQUFuuWWW7Rt2zZ16dKlwbHNysvLFRwcLJvNpqCgoHNbWQAA4BWNff9uVdcg5eXlKT4+3mVeQkKC8vLyJEknT57U9u3bXWp8fHwUHx/vrJGkyspK3XvvvUpPT1dYWFijxq6qqlJ5ebnLBAAAzk+tKiCVlJQoNDTUZV5oaKjKy8t1/PhxHTx4UNXV1bXWlJSUOB/PmDFD1157rW699dZGj52amqrg4GDnFBER8etWBgAAtFitKiA1h9WrV2vt2rVatGjROf1cSkqKbDabcyouLnZPgwAAwOv8vN3AuQgLC1NpaanLvNLSUgUFBSkwMFC+vr7y9fWttabmo7S1a9fqxx9/VIcOHVxqbr/9dg0dOlTr16+vdWyr1Sqr1dps6wIAAFquVnUGKS4uTrm5uS7zcnJyFBcXJ0ny9/fXgAEDXGrsdrtyc3OdNXPmzNHXX3+tgoIC5yRJr7zyipYtW+aZFQEAAC2aV88gHTt2TDt37nQ+LioqUkFBgUJCQhQZGamUlBTt3btXy5cvlyRNnTpVaWlpmjVrliZPnqy1a9dq5cqVysrKci4jOTlZiYmJGjhwoAYPHqxFixapoqJCSUlJkhxnoWq7MDsyMlLR0dFuXmMAANAaNOkM0ttvv+0SSmbNmqUOHTro2muv1a5duxq9nG3btqlfv37q16+fJEe46devn+bNmydJ2r9/v3bv3u2sj46OVlZWlnJychQTE6OXXnpJS5cuVUJCgrPm7rvv1osvvqh58+apb9++KigoUHZ29lkXbgMAANSlSfdB6tmzp5YsWaLrr7/e+dX7V155RR999JH8/Pz0j3/8wx29tijcBwkAgNanse/fTfqIrbi4WD169JAkrVq1Srfffrt++9vfasiQIRoxYkSTGgYAAGgpmvQRW7t27fT//t//kyT961//0g033CBJCggI0PHjx5uvOwAAAC9o0hmkG264QVOmTFG/fv30/fff6+abb5YkffPNN4qKimrO/gAAADyuSWeQ0tPTFRcXpwMHDuh///d/dckll0iStm/frnvuuadZGwQAAPC0FvPHalsbLtIGAKD1cfsfq924caMmTZqka6+9Vnv37pUkvfPOO9q0aVNTFwkAANAiNCkg/e///q8SEhIUGBio/Px8VVVVSZJsNpuef/75Zm0QAADA05oUkP74xz/qtdde05tvvqk2bdo45w8ZMkT5+fnN1hwAAIA3NCkg7dixQ8OGDTtrfnBwsI4cOfJrewIAAPCqJgWksLAwl7+hVmPTpk267LLLfnVTAAAA3tSkgPTggw/qscce05YtW2SxWLRv3z699957mjlzpv77v/+7uXsEAADwqCbdKHLOnDmy2+0aNWqUKisrNWzYMFmtVj3xxBOaMmVKc/cIAADgUU06g2SxWDR37lwdOnRIhYWF2rx5sw4cOKDg4GBFR0c3d48AAAAedU4BqaqqSikpKRo4cKCGDBmijz/+WL1799Y333yjnj17avHixZoxY4a7egUAAPCIc/qIbd68eXr99dcVHx+vL774QnfeeaeSkpK0efNmvfTSS7rzzjvl6+vrrl4BAAA84pwC0t///nctX75c48aNU2Fhoa655hqdPn1aX331lSwWi7t6BAAA8Khz+ohtz549GjBggCTpqquuktVq1YwZMwhHAADgvHJOAam6ulr+/v7Ox35+fmrXrl2zNwUAAOBN5/QRm2EYeuCBB2S1WiVJJ06c0NSpU9W2bVuXun/84x/N1yEAAICHnVNASkxMdHk8adKkZm0GAACgJTingLRs2TJ39QEAANBiNOlGkQAAAOczAhIAAIAJAQkAAMCEgAQAAGBCQAIAADAhIAEAAJgQkAAAAEwISAAAACYEJAAAABMCEgAAgAkBCQAAwISABAAAYEJAAgAAMCEgAQAAmBCQAAAATAhIAAAAJgQkAAAAEwISAACACQEJAADAxKsBacOGDRo7dqzCw8NlsVi0atWqBn9m/fr16t+/v6xWq3r06KGMjIyzatLT0xUVFaWAgADFxsZq69atLs8/9NBD6t69uwIDA9WpUyfdeuut+u6775pprQAAQGvn1YBUUVGhmJgYpaenN6q+qKhIY8aM0ciRI1VQUKDp06drypQpWrNmjbNmxYoVSk5O1vz585Wfn6+YmBglJCSorKzMWTNgwAAtW7ZM3377rdasWSPDMDR69GhVV1c3+zoCAIDWx2IYhuHtJiTJYrEoMzNT48ePr7Nm9uzZysrKUmFhoXPehAkTdOTIEWVnZ0uSYmNjNWjQIKWlpUmS7Ha7IiIiNG3aNM2ZM6fW5X799deKiYnRzp071b1790b1W15eruDgYNlsNgUFBTVyLQEAgDc19v27VV2DlJeXp/j4eJd5CQkJysvLkySdPHlS27dvd6nx8fFRfHy8s8asoqJCy5YtU3R0tCIiIuocu6qqSuXl5S4TAAA4P7WqgFRSUqLQ0FCXeaGhoSovL9fx48d18OBBVVdX11pTUlLiMu8vf/mL2rVrp3bt2umTTz5RTk6O/P396xw7NTVVwcHBzqm+MAUAAFq3VhWQmtPEiRP173//W5999pmuuOIK3XXXXTpx4kSd9SkpKbLZbM6puLjYg90CAABP8vN2A+ciLCxMpaWlLvNKS0sVFBSkwMBA+fr6ytfXt9aasLAwl3k1Z4Iuv/xy/eY3v9HFF1+szMxM3XPPPbWObbVaZbVam3eFAABAi9SqziDFxcUpNzfXZV5OTo7i4uIkSf7+/howYIBLjd1uV25urrOmNoZhyDAMVVVVuadxAADQqnj1DNKxY8e0c+dO5+OioiIVFBQoJCREkZGRSklJ0d69e7V8+XJJ0tSpU5WWlqZZs2Zp8uTJWrt2rVauXKmsrCznMpKTk5WYmKiBAwdq8ODBWrRokSoqKpSUlCRJ+umnn7RixQqNHj1anTp10p49e/TCCy8oMDBQN998s2c3AAAAaJG8GpC2bdumkSNHOh8nJydLkhITE5WRkaH9+/dr9+7dzuejo6OVlZWlGTNmaPHixeratauWLl2qhIQEZ83dd9+tAwcOaN68eSopKVHfvn2VnZ3tvHA7ICBAGzdu1KJFi3T48GGFhoZq2LBh+uKLL9S5c2cPrTkAAGjJWsx9kFob7oMEAEDrc17eBwkAAMATCEgAAAAmBCQAAAATAhIAAIAJAQkAAMCEgAQAAGBCQAIAADAhIAEAAJgQkAAAAEwISAAAACYEJAAAABMCEgAAgAkBCQAAwISABAAAYEJAAgAAMCEgAQAAmBCQAAAATAhIAAAAJgQkAAAAEwISAACACQEJAADAhIAEAABgQkACAAAwISABAACYEJAAAABMCEgAAAAmBCQAAAATAhIAAIAJAQkAAMCEgAQAAGBCQAIAADAhIAEAAJgQkAAAAEwISAAAACYEJAAAABMCEgAAgAkBCQAAwISABAAAYEJAAgAAMCEgAQAAmHg1IG3YsEFjx45VeHi4LBaLVq1a1eDPrF+/Xv3795fValWPHj2UkZFxVk16erqioqIUEBCg2NhYbd261fncoUOHNG3aNPXs2VOBgYGKjIzUo48+KpvN1oxrBgAAWjOvBqSKigrFxMQoPT29UfVFRUUaM2aMRo4cqYKCAk2fPl1TpkzRmjVrnDUrVqxQcnKy5s+fr/z8fMXExCghIUFlZWWSpH379mnfvn168cUXVVhYqIyMDGVnZ+u//uu/3LKOAACg9bEYhmF4uwlJslgsyszM1Pjx4+usmT17trKyslRYWOicN2HCBB05ckTZ2dmSpNjYWA0aNEhpaWmSJLvdroiICE2bNk1z5sypdbl///vfNWnSJFVUVMjPz69R/ZaXlys4OFg2m01BQUGNXEsAAOBNjX3/blXXIOXl5Sk+Pt5lXkJCgvLy8iRJJ0+e1Pbt211qfHx8FB8f76ypTc1Gqi8cVVVVqby83GUCAADnp1YVkEpKShQaGuoyLzQ0VOXl5Tp+/LgOHjyo6urqWmtKSkpqXebBgwf17LPP6re//W29Y6empio4ONg5RURE/LqVAQAALVarCkjNrby8XGPGjFHv3r31zDPP1FubkpIim83mnIqLiz3TJAAA8LjGXXDTQoSFham0tNRlXmlpqYKCghQYGChfX1/5+vrWWhMWFuYy7+jRo7rxxhvVvn17ZWZmqk2bNvWObbVaZbVam2dFAABAi9aqziDFxcUpNzfXZV5OTo7i4uIkSf7+/howYIBLjd1uV25urrNGcpw5Gj16tPz9/bV69WoFBAR4ZgUAAECr4NUzSMeOHdPOnTudj4uKilRQUKCQkBBFRkYqJSVFe/fu1fLlyyVJU6dOVVpammbNmqXJkydr7dq1WrlypbKyspzLSE5OVmJiogYOHKjBgwdr0aJFqqioUFJSkqQz4aiyslLvvvuuywXXnTp1kq+vrwe3AAAAaIm8GpC2bdumkSNHOh8nJydLkhITE5WRkaH9+/dr9+7dzuejo6OVlZWlGTNmaPHixeratauWLl2qhIQEZ83dd9+tAwcOaN68eSopKVHfvn2VnZ3tvHA7Pz9fW7ZskST16NHDpZ+ioiJFRUW5a3UBAEAr0WLug9TacB8kAABan/PyPkgAAACeQEACAAAwISABAACYEJAAAABMCEgAAAAmBCQAAAATAhIAAIAJAQkAAMCEgAQAAGBCQAIAADAhIAEAAJgQkAAAAEwISAAAACYEJAAAABMCEgAAgAkBCQAAwISABAAAYEJAAgAAMCEgAQAAmBCQAAAATAhIAAAAJgQkAAAAEwISAACACQEJAADAhIAEAABgQkACAAAwISABAACYEJAAAABMCEgAAAAmBCQAAAATAhIAAIAJAQkAAMCEgAQAAGBCQAIAADAhIAEAAJgQkAAAAEwISAAAACYEJAAAABMCEgAAgAkBCQAAwMSrAWnDhg0aO3aswsPDZbFYtGrVqgZ/Zv369erfv7+sVqt69OihjIyMs2rS09MVFRWlgIAAxcbGauvWrS7Pv/HGGxoxYoSCgoJksVh05MiR5lkhAABwXvBqQKqoqFBMTIzS09MbVV9UVKQxY8Zo5MiRKigo0PTp0zVlyhStWbPGWbNixQolJydr/vz5ys/PV0xMjBISElRWVuasqays1I033qgnn3yy2dcJAAC0fhbDMAxvNyFJFotFmZmZGj9+fJ01s2fPVlZWlgoLC53zJkyYoCNHjig7O1uSFBsbq0GDBiktLU2SZLfbFRERoWnTpmnOnDkuy1u/fr1Gjhypw4cPq0OHDufUb3l5uYKDg2Wz2RQUFHROPwsAALyjse/freoapLy8PMXHx7vMS0hIUF5eniTp5MmT2r59u0uNj4+P4uPjnTUAAAAN8fN2A+eipKREoaGhLvNCQ0NVXl6u48eP6/Dhw6qurq615rvvvvtVY1dVVamqqsr5uLy8/FctDwAAtFyt6gySN6Wmpio4ONg5RUREeLslAADgJq0qIIWFham0tNRlXmlpqYKCghQYGKiOHTvK19e31pqwsLBfNXZKSopsNptzKi4u/lXLAwAALVerCkhxcXHKzc11mZeTk6O4uDhJkr+/vwYMGOBSY7fblZub66xpKqvVqqCgIJcJAACcn7x6DdKxY8e0c+dO5+OioiIVFBQoJCREkZGRSklJ0d69e7V8+XJJ0tSpU5WWlqZZs2Zp8uTJWrt2rVauXKmsrCznMpKTk5WYmKiBAwdq8ODBWrRokSoqKpSUlOSsKSkpUUlJiXPs//znP2rfvr0iIyMVEhLiobUHAAAtluFF69atMySdNSUmJhqGYRiJiYnG8OHDz/qZvn37Gv7+/sZll11mLFu27Kzlvvrqq0ZkZKTh7+9vDB482Ni8ebPL8/Pnz6913NqWVRebzWZIMmw22zmuNQAA8JbGvn+3mPsgtTbcBwkAgNbnvLwPEgAAgCcQkAAAAEwISAAAACYEJAAAABMCEgAAgAkBCQAAwISABAAAYEJAAgAAMCEgAQAAmBCQAAAATAhIAAAAJgQkAAAAEwISAACACQEJAADAhIAEAABgQkACAAAwISABAACYEJBaiEOHpOefl3r0kIKDpT59pMWLpYoKz4x/8qS0dKk0YIBj/MhIae5caf9+z4xvGNJHH0mjR0sXXyx17ixNmSIVFnpmfEnavl267z6pY0cpJEQaO1b69FPPjb9rl/T449Kll0odOki/+Y30zjtSdbVnxrfZpIULpZ49HfvAlVdKL70kHT3qmfFPn5YyMqTBgx3r37WrNGuWtGePZ8aXpDVrpJtvdrz+nTpJDzwgFRR4bvyvvpKSkhxjh4RIN90kffKJ58bfs0eaM0eKiHDsA4MGScuWSadOeWb8o0ell1+Wevd2jH/55dKCBdKRI54Zv7paeu896dprHfvgpZdKycnSzz97ZnxJWrtWGjfO8fp37ChNnCht2+a58b/5Rvrtbx3H4Isvlm64QfrwQ8cx2hNKSqSnnpK6dXPsA/37S2++KVVVeWZ8FwaaxGazGZIMm832q5e1e7dhREYaho+PYTh2Q8OwWBzT1VcbxqFDzdBwPSorDWPECMe4v+zB19cwQkIM4z//ce/4drthPPbYmTFrxvfzc0yZme4d3zAM4+23Hdvbz891fMkwnnnG/eNv2WIY7dq5rn/NazF2rGGcPOne8ffvN4zu3V1f/5oeevY0jLIy945fVWUYN95Y+z4YHGwY27e7d3zDMIwnn6x9H/T1NYy//c39469Y4Rjrl/tgTS+zZrl//H//2zA6dKh9H4yPN4wTJ9w7/oEDhnHllWeOfb/sITraMPbude/4p04Zxvjxte+D7doZRl6ee8c3DMN49tna90GLxTCWLXP/+P/855njrnkf/N3vHMdqd/rmG8O45BLX9a/ZF4YOdbxXNYfGvn8TkJqoOQPSsGGuO+QvJ19fw5g4sRkarsesWWe/Mf5y/B49DKO62n3jf/BB7WPX/HL4+xtGSYn7xt+5s+71r5k+/dR941dVGUZoaN09WCyGkZrqvvENwxFOfnlQ+uXk5+d443CnZ56pfx/s2tXxBuYuH39c/+vv52cYu3a5b/ziYsNo06b+Hj780H3jnz7t+EdaXfuAj49hPP20+8Y3DMO4/fb698EbbnDv+AsXugYz8z7YqZN7Q+K6dfW//j4+hrFjh/vGLyszjICAureB5Ajx7mK3O/4xVt8+OHNm84xFQHKz5gpIhYX1/1LUHBzcFRAqKw0jKKjhHv71L/eMbxiGMWRI3b8UNb8Yzz3nvvFnzqx/fD8/wxg3zn3jr1jR8Pbv0sXxJuYOO3c2PL7F4r6AcPKkYXTs2HAPq1a5Z3zDcLz51rcP+Poaxty57hv/6acbHn/UKPeN/+GHDW//kBBHmHeHPXsa/keK5L6AUF1tGJde2vD477/vnvENw/GPkLr+oVyzD8yY4b7xU1Prfw18fAwjLs594+fmNrz927c3jIqKXz9WY9+/uQbJyzZvbrjm9GkpP9894//wg1ReXn+Nn5+Ul+ee8SVp69b6r7Ox26UvvnDf+Js21T/+6dOOGnfZvFlq06b+mv37pb173TP+1q0N1xiG9OWX7hl/1y7p4MH6a9q0ce8+mJdX/z5QXe3efeCLLxoe393r39A+eOiQ9NNP7hl/2zbH73lDtmxxz/glJQ3/frl7H/z8c8expi7V1dLGje4bf/Nmx+95Xex2x/avr+bXyMtzvNfU5+hRaccO94xfGwKSl/n6Nq6uoR3HneMbhvvGlySfBvZCi6Xhg/ev0Zh1c+f6+/o27qDjzX3A2+O7ex9sTA/e3gcb+zo1Bfugd8dvbA/u3AcbM35L2Afd2YMZAcnLrr/eEQDqExgoxca6Z/yePaWwsPprqqul+Hj3jC85viXR0E7vzvFvvLH+kObn5/g2kbvccEP9/3K0WByvU5cu7hl/+PCGD/z+/tJ117ln/G7dpKio+n8PTp92bCd3ufHG+reBj4/jG5buMnp0/evv5yclJLh3/Pr2QcnxzdbLLnPP+EOGSFZr/TW+vtKIEe4Zv3NnqVev+l+DU6fcuw/edFPD+6A794GG1s3X13Ecbuj9qqlGj274G7udOzu+4egxv/7TvAtTc16kfeed9V+YlpzcDA3X48UX67/+xp2fOxuGYaxfX//n7iEhhtEMm7lOJSWGcdFFdX/+7uNjGPn57hu/utowevWq//qDv/7VfeMbhmE88ED96//QQ+4d/y9/qX8f7NvXvd+g2bKl7otTfXwc1z4cPOi+8Q8dcnxbr74L9b/4wn3j2+2G0b9//fvgq6+6b3zDMIyHH67/Qv1Jk9w7fkZG/ftgz57u/bLKV1/V/z5w0UWGsW+f+8Y/evTsb5CZp9xc941vGI5vqtW3Dy5Y0DzjcJG2mzVnQLLZHCGk5kBQ8wspOb7i7a4LI2tUVxvGgw+6jltzoLrySsdXwN3tL39xvAmYv97ZoYNhbN3q/vHXrTOMtm3P/nqvr69hvPuu+8f/6SfDiIo6s96/fC0ef9z9X689duzMrR5qXoOa/44e3Xxfr62L3W4Yjz5a+z7YvbvjVhjutmyZY0zz19zbtzeMTZvcP/4XXzi+MGHeB318DGPpUvePX1xsGJdf7rrta16Lhx92/z54/PiZWz2Y98Fhwxxv4O5ktxvGnDmu613zuxgZaRg//uje8Q3DcTuJmltLmMORu8OJYRjGtm2Of5D+8h8LNbcZcHdANgzHP1Z79659H0xKar6A2tj3b4thNOZTP5iVl5crODhYNptNQUFBv3p5p09Lq1dLy5c7LsiNipImT3ac9mzoGp3mYBiOiwTfeEP6/nvHTcruuUe6804pIMD940uOi+9ee81xsaDV6rhRY1KSoxdPKCtz3CxzzRrH6fShQ6WHHnLfxwpmFRXS3/4mrVzpuHC+Tx9p6lTHzfo8obpaysqS3n7bccPAiAjH9r/xRs997r95s/T669K33zpu1Hf33dKECY6PmT1h507H+J9/7rje4+abHb+HnTp5ZvyDBx03ZszKcty8NS7OsQ9cfrlnxj9+3LH//e1vjpsz9uolPfig48aJnmC3S9nZjm2we7fjRo2JidItt3huH9y2zXEcKiyUgoIcx8B775XatvXM+D//7NgHN2xwrPPo0Y7XIDTUM+MfPuy4Yevq1dKJE47LO6ZOddw41hOqqqQPPnDcsPPQIce+/+CDjuNxc32819j3bwJSEzV3QAIAAO7X2PdvLtIGAAAwISABAACYEJAAAABMCEgAAAAmBCQAAAATAhIAAIAJAQkAAMCEgAQAAGBCQAIAADAhIAEAAJj4ebuB1qrmL7SUl5d7uRMAANBYNe/bDf2lNQJSEx09elSSFBER4eVOAADAuTp69KiCg4PrfJ4/VttEdrtd+/btU/v27WVprj8xLEeyjYiIUHFx8QX7R3Av9G1woa+/xDa40NdfYhuw/u5bf8MwdPToUYWHh8vHp+4rjTiD1EQ+Pj7q2rWr25YfFBR0Qf5S/NKFvg0u9PWX2AYX+vpLbAPW3z3rX9+ZoxpcpA0AAGBCQAIAADAhILUwVqtV8+fPl9Vq9XYrXnOhb4MLff0ltsGFvv4S24D19/76c5E2AACACWeQAAAATAhIAAAAJgQkAAAAEwISAACACQGphUlPT1dUVJQCAgIUGxurrVu3erslj0hNTdWgQYPUvn17de7cWePHj9eOHTu83ZZXvfDCC7JYLJo+fbq3W/GYvXv3atKkSbrkkksUGBioq6++Wtu2bfN2Wx5TXV2tp59+WtHR0QoMDFT37t317LPPNvg3o1qrDRs2aOzYsQoPD5fFYtGqVatcnjcMQ/PmzVOXLl0UGBio+Ph4/fDDD95p1k3q2wanTp3S7NmzdfXVV6tt27YKDw/X/fffr3379nmv4WbW0D7wS1OnTpXFYtGiRYs80hsBqQVZsWKFkpOTNX/+fOXn5ysmJkYJCQkqKyvzdmtu99lnn+mRRx7R5s2blZOTo1OnTmn06NGqqKjwdmte8eWXX+r111/XNddc4+1WPObw4cMaMmSI2rRpo08++UT/93//p5deekkXX3yxt1vzmAULFmjJkiVKS0vTt99+qwULFmjhwoV69dVXvd2aW1RUVCgmJkbp6em1Pr9w4UL9+c9/1muvvaYtW7aobdu2SkhI0IkTJzzcqfvUtw0qKyuVn5+vp59+Wvn5+frHP/6hHTt2aNy4cV7o1D0a2gdqZGZmavPmzQoPD/dQZ5IMtBiDBw82HnnkEefj6upqIzw83EhNTfViV95RVlZmSDI+++wzb7ficUePHjUuv/xyIycnxxg+fLjx2GOPebslj5g9e7Zx3XXXebsNrxozZowxefJkl3m33XabMXHiRC915DmSjMzMTOdju91uhIWFGX/605+c844cOWJYrVbjb3/7mxc6dD/zNqjN1q1bDUnGrl27PNOUB9W1/nv27DEuvfRSo7Cw0OjWrZvxyiuveKQfziC1ECdPntT27dsVHx/vnOfj46P4+Hjl5eV5sTPvsNlskqSQkBAvd+J5jzzyiMaMGeOyL1wIVq9erYEDB+rOO+9U586d1a9fP7355pvebsujrr32WuXm5ur777+XJH311VfatGmTbrrpJi935nlFRUUqKSlx+T0IDg5WbGzsBXlMrGGz2WSxWNShQwdvt+IRdrtd9913n5544gn16dPHo2Pzx2pbiIMHD6q6ulqhoaEu80NDQ/Xdd995qSvvsNvtmj59uoYMGaKrrrrK2+141P/8z/8oPz9fX375pbdb8biffvpJS5YsUXJysp588kl9+eWXevTRR+Xv76/ExERvt+cRc+bMUXl5ua688kr5+vqqurpazz33nCZOnOjt1jyupKREkmo9JtY8d6E5ceKEZs+erXvuueeC+QO2CxYskJ+fnx599FGPj01AQovzyCOPqLCwUJs2bfJ2Kx5VXFysxx57TDk5OQoICPB2Ox5nt9s1cOBAPf/885Kkfv36qbCwUK+99toFE5BWrlyp9957T++//7769OmjgoICTZ8+XeHh4RfMNkDtTp06pbvuukuGYWjJkiXebscjtm/frsWLFys/P18Wi8Xj4/MRWwvRsWNH+fr6qrS01GV+aWmpwsLCvNSV5/3ud7/TRx99pHXr1qlr167ebsejtm/frrKyMvXv319+fn7y8/PTZ599pj//+c/y8/NTdXW1t1t0qy5duqh3794u83r16qXdu3d7qSPPe+KJJzRnzhxNmDBBV199te677z7NmDFDqamp3m7N42qOexf6MVE6E4527dqlnJycC+bs0caNG1VWVqbIyEjnMXHXrl2aOXOmoqKi3D4+AamF8Pf314ABA5Sbm+ucZ7fblZubq7i4OC925hmGYeh3v/udMjMztXbtWkVHR3u7JY8bNWqU/vOf/6igoMA5DRw4UBMnTlRBQYF8fX293aJbDRky5KxbO3z//ffq1q2blzryvMrKSvn4uB6WfX19ZbfbvdSR90RHRyssLMzlmFheXq4tW7ZcEMfEGjXh6IcfftCnn36qSy65xNstecx9992nr7/+2uWYGB4erieeeEJr1qxx+/h8xNaCJCcnKzExUQMHDtTgwYO1aNEiVVRUKCkpydutud0jjzyi999/X//85z/Vvn175zUGwcHBCgwM9HJ3ntG+ffuzrrlq27atLrnkkgviWqwZM2bo2muv1fPPP6+77rpLW7du1RtvvKE33njD2615zNixY/Xcc88pMjJSffr00b///W+9/PLLmjx5srdbc4tjx45p586dzsdFRUUqKChQSEiIIiMjNX36dP3xj3/U5ZdfrujoaD399NMKDw/X+PHjvdd0M6tvG3Tp0kV33HGH8vPz9dFHH6m6utp5bAwJCZG/v7+32m42De0D5kDYpk0bhYWFqWfPnu5vziPflUOjvfrqq0ZkZKTh7+9vDB482Ni8ebO3W/IISbVOy5Yt83ZrXnUhfc3fMAzjww8/NK666irDarUaV155pfHGG294uyWPKi8vNx577DEjMjLSCAgIMC677DJj7ty5RlVVlbdbc4t169bV+nufmJhoGIbjq/5PP/20ERoaalitVmPUqFHGjh07vNt0M6tvGxQVFdV5bFy3bp23W28WDe0DZp78mr/FMM7TW7QCAAA0EdcgAQAAmBCQAAAATAhIAAAAJgQkAAAAEwISAACACQEJAADAhIAEAABgQkACABOLxaJVq1Z5uw0AXkRAAnBeeeCBB86rP0UBwDsISAAAACYEJADnrREjRujRRx/VrFmzFBISorCwMD3zzDMuNT/88IOGDRumgIAA9e7dWzk5OWctp7i4WHfddZc6dOigkJAQ3Xrrrfr5558lSd99950uuugivf/++876lStXKjAwUP/3f//nztUD4EYEJADntbfffltt27bVli1btHDhQv3hD39whiC73a7bbrtN/v7+2rJli1577TXNnj3b5edPnTqlhIQEtW/fXhs3btTnn3+udu3a6cYbb9TJkyd15ZVX6sUXX9TDDz+s3bt3a8+ePZo6daoWLFig3r17e2OVATQD/lgtgPPKAw88oCNHjmjVqlUaMWKEqqurtXHjRufzgwcP1vXXX68XXnhB//rXvzRmzBjt2rVL4eHhkqTs7GzddNNNyszM1Pjx4/Xuu+/qj3/8o7799ltZLBZJ0smTJ9WhQwetWrVKo0ePliTdcsstKi8vl7+/v3x9fZWdne2sB9D6+Hm7AQBwp2uuucblcZcuXVRWViZJ+vbbbxUREeEMR5IUFxfnUv/VV19p586dat++vcv8EydO6Mcff3Q+fuutt3TFFVfIx8dH33zzDeEIaOUISADOa23atHF5bLFYZLfbG/3zx44d04ABA/Tee++d9VynTp2c///VV1+poqJCPj4+2r9/v7p06dL0pgF4HQEJwAWrV69eKi4udgk0mzdvdqnp37+/VqxYoc6dOysoKKjW5Rw6dEgPPPCA5s6dq/3792vixInKz89XYGCg29cBgHtwkTaAC1Z8fLyuuOIKJSYm6quvvtLGjRs1d+5cl5qJEyeqY8eOuvXWW7Vx40YVFRVp/fr1evTRR7Vnzx5J0tSpUxUREaGnnnpKL7/8sqqrq/X44497Y5UANBMCEoALlo+PjzIzM3X8+HENHjxYU6ZM0XPPPedSc9FFF2nDhg2KjIzUbbfdpl69eum//uu/dOLECQUFBWn58uX6+OOP9c4778jPz09t27bVu+++qzfffFOffPKJl9YMwK/Ft9gAAABMOIMEAABgQkACAAAwISABAACYEJAAAABMCEgAAAAmBCQAAAATAhIAAIAJAQkAAMCEgAQAAGBCQAIAADAhIAEAAJgQkAAAAEz+P7gI+6EPSOnYAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure()\n",
    "# Assuming res_list and threshold are defined as in your code\n",
    "colors = ['red' if res > threshold else 'blue' for res in forward_res_list]\n",
    "# Scatter plot with individual colors for each point\n",
    "plt.scatter(range(len(forward_res_list)), forward_res_list, color=colors)\n",
    "plt.axhline(y=threshold, color='red', linestyle='--')\n",
    "plt.xlabel('Index')\n",
    "plt.ylabel('Res')\n",
    "plt.title('Normal with RE Test')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([15, 1973, 33])"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset_test.data_attacked.size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([1, 2045, 33])"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_noisy_test = dataset.data_attacked\n",
    "data_noisy_test.size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Output size: torch.Size([1, 2045, 33]) \t Output Back size: torch.Size([1, 2045, 33])\n"
     ]
    }
   ],
   "source": [
    "with torch.no_grad():\n",
    "    output = G_AB(data_noisy_test)\n",
    "    output_back = G_BA(output)\n",
    "print(f\"Output size: {output.size()} \\t Output Back size: {output_back.size()}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+ 1.000041228553356\n"
     ]
    }
   ],
   "source": [
    "os.makedirs(f'cyber_result/true/{noise_type}/detection', exist_ok=True)\n",
    "detection_res_list = []\n",
    "for i in range(output.size(0)):\n",
    "    plt.plot(output[i, :, 0].numpy())\n",
    "    plt.plot(output_back[i, :, 0].numpy())\n",
    "    plt.plot(data_noisy_test[i, :, 0].numpy())\n",
    "    plt.legend(['Generated', 'Generated Back', 'Original'])\n",
    "    plt.savefig(f'cyber_result/true/{noise_type}/detection/output_{i}.png')\n",
    "    plt.close()\n",
    "\n",
    "    dist1, cost, acc, path = dtw(output[i,0,:].reshape(-1, 1), data_noisy_test[i, :].reshape(-1, 1), dist=lambda x, y: np.linalg.norm(x - y, ord=1))\n",
    "    dist2, cost, acc, path = dtw(output_back[i,0,:].reshape(-1, 1), output[i,0,:].reshape(-1, 1), dist=lambda x, y: np.linalg.norm(x - y, ord=1))\n",
    "    dist3, cost, acc, path = dtw(output_back[i,0,:].reshape(-1, 1), data_noisy_test[i, :].reshape(-1, 1), dist=lambda x, y: np.linalg.norm(x - y, ord=1)) \n",
    "    res = (dist2+dist3)/dist1   \n",
    "    print(\"+\" if res < threshold else \"-\", res)\n",
    "    detection_res_list.append(res)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure()\n",
    "# Assuming res_list and threshold are defined as in your code\n",
    "colors = ['red' if res > threshold else 'blue' for res in detection_res_list]\n",
    "# Scatter plot with individual colors for each point\n",
    "plt.scatter(range(len(detection_res_list)), detection_res_list, color=colors)\n",
    "plt.axhline(y=threshold, color='red', linestyle='--')\n",
    "plt.xlabel('Index')\n",
    "plt.ylabel('Res')\n",
    "plt.title('Test: Attacked data with RE')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[1.000450552578472,\n",
       " 1.0004505525314251,\n",
       " 1.0004505525314251,\n",
       " 1.0004505525314251,\n",
       " 1.0004505525314251,\n",
       " 1.0004505525314251,\n",
       " 1.0004505525314251,\n",
       " 1.0004505525314251,\n",
       " 1.0004505525314251,\n",
       " 1.0004505525314251,\n",
       " 1.0004505525314251,\n",
       " 1.0004505525314251,\n",
       " 1.0004505525314251,\n",
       " 1.0004505525196634,\n",
       " 1.0004505524961187]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sorted(forward_res_list, reverse=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[1.000041228553356]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sorted(detection_res_list, reverse=True)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "GAN",
   "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": "3.12.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}