1460 lines (1459 with data), 304.7 kB

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "0c9a8858-c71a-4c48-9d4e-73a5bca4376c",
   "metadata": {},
   "source": [
    "# Get embeddings to predict low EF and one-year mortality"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2add60a7-c016-4280-a3fb-77eb81760953",
   "metadata": {},
   "source": [
    "### Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "b4a2ee4e-b0ae-4a7d-afee-fa23c1f76d41",
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys\n",
    "sys.path.append('../..') \n",
    "\n",
    "import os\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from tqdm.notebook import tqdm\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "import xgboost as xgb\n",
    "from sklearn.metrics import roc_auc_score, average_precision_score, precision_recall_curve, auc\n",
    "\n",
    "import torch\n",
    "from torchvision import transforms\n",
    "from torch.utils.data import DataLoader\n",
    "from torchvision.transforms import Compose\n",
    "\n",
    "import pytorch_lightning as pl\n",
    "from pytorch_lightning.loggers.neptune import NeptuneLogger\n",
    "from pytorch_lightning.callbacks import ModelCheckpoint\n",
    "\n",
    "from ecgxai.network.causalcnn.vae import VAE\n",
    "from ecgxai.utils.dataset import UniversalECGDataset\n",
    "from ecgxai.utils.transforms import ApplyGain, ToTensor, To12Lead, Resample\n",
    "from ecgxai.systems.VAE_system import GaussianVAE\n",
    "from ecgxai.systems.classification_system import ClassificationSystem\n",
    "from ecgxai.utils.loss import CombinedLoss, GaussianVAEReconLoss, KLDivergence, BinaryFocalLoss"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c88e681b-ab78-432b-b862-704efb3bc60a",
   "metadata": {},
   "source": [
    "### Some utils"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "e64660e8-9ec3-4fff-9a5f-6b0267a2d48f",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_latents(model, df):\n",
    "    transform = transforms.Compose([Resample(500), ApplyGain(), ToTensor(), To12Lead()])\n",
    "    \n",
    "    dataset = UniversalECGDataset(\n",
    "        dataset_function=\"umcu\",\n",
    "        waveform_dir=\"/raw_data/umcu_median\",\n",
    "        dataset=df,\n",
    "        transform=transform\n",
    "    )\n",
    "    \n",
    "    dataloader = DataLoader(dataset, batch_size=128, shuffle=False, num_workers=8)\n",
    "    \n",
    "    output = trainer.predict(model, dataloaders=dataloader)\n",
    "    \n",
    "    embedding = torch.cat([x[\"mu\"] for x in output], dim=0).cpu().numpy()\n",
    "    waveforms = torch.cat([x[\"x\"] for x in output], dim=0).cpu().numpy()\n",
    "    reconstructions = torch.cat([x[\"reconstruction\"] for x in output], dim=0).cpu().numpy()\n",
    "    \n",
    "    #scaler = StandardScaler()\n",
    "    #embedding = scaler.fit_transform(embedding)\n",
    "\n",
    "    latent_rep_df = pd.DataFrame(embedding)\n",
    "    latent_rep_df.columns = ['latent_' + str(col+1) for col in latent_rep_df.columns]\n",
    "    latent_dimensions_include = ['latent_' + str(i) for i in [1,5,6,8,9,10,11,12,13,15,16,17,19,22,23,25,26,27,30,31,32]]\n",
    "    latent_rep_df = latent_rep_df[latent_dimensions_include]\n",
    "    \n",
    "    latent_rep_df = pd.concat([df.reset_index(), latent_rep_df], axis=1)\n",
    "    \n",
    "    return latent_rep_df, waveforms, reconstructions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "1af74d9f-33f9-4e36-a8b5-a0b6cdea482d",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_prediction(model, df):\n",
    "    transform = transforms.Compose([Resample(500), ApplyGain(), ToTensor(), To12Lead()])\n",
    "    \n",
    "    dataset = UniversalECGDataset(\n",
    "        dataset_function=\"umcu\",\n",
    "        waveform_dir=\"/raw_data/umcu_median\",\n",
    "        dataset=df,\n",
    "        transform=transform\n",
    "    )\n",
    "    \n",
    "    dataloader = DataLoader(dataset, batch_size=128, shuffle=False, num_workers=8)\n",
    "    \n",
    "    output = trainer.predict(model, dataloaders=dataloader)\n",
    "    \n",
    "    y = torch.cat([x[\"label\"] for x in output], dim=0).cpu().numpy()\n",
    "    y_prob = torch.cat([x[\"y_prob\"] for x in output], dim=0).cpu().numpy()\n",
    "\n",
    "    return y, y_prob"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "08dcee9f-c5c4-4b1d-a0bd-cd65705b655c",
   "metadata": {},
   "outputs": [],
   "source": [
    "def bootstrap_ci(metric, label, y_prob):\n",
    "    boot_metric = []\n",
    "    boot_df = pd.DataFrame({'label': label, 'y_prob': y_prob})\n",
    "\n",
    "    for i in range(2000):\n",
    "        boot = boot_df.sample(frac = 1, replace = True).reset_index(drop=True)\n",
    "        metric_value = metric(boot['label'], boot['y_prob'])\n",
    "        boot_metric.append(metric_value)\n",
    "\n",
    "    print('C-stat CI: ' + str(np.quantile(boot_metric,q=0.025,axis=0)) + ' ' + str(np.quantile(boot_metric,q=0.975,axis=0)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "6c9e16da-4a3b-4888-8923-46d82ca8cc10",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "GPU available: True, used: True\n",
      "TPU available: False, using: 0 TPU cores\n",
      "IPU available: False, using: 0 IPUs\n"
     ]
    }
   ],
   "source": [
    "trainer = pl.Trainer(logger=False, gpus=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4b9482d9-36ef-4c81-a482-86cd5734237e",
   "metadata": {},
   "source": [
    "### Initialize models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "2b044b99-8ea0-480d-b2b9-73325b641066",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'reconstruction_mean', 'mu', 'reconstruction_std', 'std', 'x'}\n"
     ]
    }
   ],
   "source": [
    "vae = GaussianVAE.load_from_checkpoint(checkpoint_path=\"/training/factorecg/final_vae_epoch_37_2.ckpt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "2fddfdb4-4bd8-4c2a-b1fb-e1ce88ab8300",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['label', 'y_hat']\n"
     ]
    }
   ],
   "source": [
    "ef_vencoder = ClassificationSystem.load_from_checkpoint(checkpoint_path=\"/training/factorecg/ef_vencoder_40.ckpt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "f39da3f5-4359-4364-a21e-4d85cde6464b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['label', 'y_hat']\n"
     ]
    }
   ],
   "source": [
    "mort_vencoder = ClassificationSystem.load_from_checkpoint(checkpoint_path=\"/training/factorecg/mortality_vencoder.ckpt\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "22dd3a2c-d7d6-4c04-8664-56abe04187eb",
   "metadata": {},
   "source": [
    "## Reduced EF model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b022e21d-29f7-4a44-927f-61261cf2ce19",
   "metadata": {},
   "source": [
    "### Get data and dataloaders"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "3a2fea4a-9ea4-4ffb-ae6d-5cb2ef020072",
   "metadata": {},
   "outputs": [],
   "source": [
    "trainset_path = '/training/secondary_task/train_ef.csv'\n",
    "df_train = pd.read_csv(trainset_path)\n",
    "df_train['Gain'] = 0.00488\n",
    "testset_path = '/training/secondary_task/test_ef.csv'\n",
    "df_test = pd.read_csv(testset_path)\n",
    "df_test['Gain'] = 0.00488\n",
    "df_test = df_test.sort_values('AcquisitionDateTime').groupby('PseudoID').head(1).reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "76457532-c33d-4dad-984a-5cf4e5ffb53d",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train['Label'] = df_train['EF_best_method'] <= 40\n",
    "df_test['Label'] = df_test['EF_best_method'] <= 40"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "5419fcea-af47-4583-a2a5-7fd677469261",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train.to_csv('/training/secondary_task/train_ef_40.csv')\n",
    "df_test.to_csv('/training/secondary_task/test_ef_40.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c1fc736d-a067-49cb-8b4a-7a10e28d5cb8",
   "metadata": {},
   "source": [
    "### Perform FactorECG-based prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "99d29748-135f-4fba-814a-3cf854ad7595",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0,1]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ddf466cde5d24902bce60916bb839aee",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Predicting: 0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n"
     ]
    }
   ],
   "source": [
    "latent_rep_df_train, waveforms_train, reconstructions_train = get_latents(vae, df_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "8976e1ff-4744-4a70-a642-7b34c9d23863",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0,1]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5f3ea06389d94f689f52678719babae3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Predicting: 0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "latent_rep_df_test, waveforms_test, reconstructions_test = get_latents(vae, df_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "05f3ef68-f79a-48e8-a03b-fc270abc8b16",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAD4CAYAAADvsV2wAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8rg+JYAAAACXBIWXMAAAsTAAALEwEAmpwYAAAscklEQVR4nO3deXxc5Xno8d8z58yiGa2W5QWvGNvYZnfEvoNZQloISwKUBJNCndykN+QmLYXyaZsmbS9tkjbJzepCElKSZiEBzBYWAyFpCCB2bGNssA3eZcnapdEsz/3jHMmyNJLGGlny0Tzfz2c+c5ZXc95jn/PMO895z3tEVTHGGDPxhca7AsYYY8aGBXxjjCkSFvCNMaZIWMA3xpgiYQHfGGOKhDveFRjK5MmTde7cueNdDWOMCYyXXnppj6rW5Fp3SAf8uXPnUldXN97VMMaYwBCRLYOts5SOMcYUCQv4xhhTJCzgG2NMkbCAb4wxRcICvjHGFAkL+MYYUyQs4BtjTJEYlYAvIheLyHoR2Sgit+ZY/3kRWSsir4vIahGZMxrbNRPfxt2tPP3W7vGuhjETQsEBX0Qc4NvAB4ElwLUisqRfsVeAWlU9FrgX+LdCt2uKw9Xf/yOf+NGL1Lcmx7sqxgTeaLTwTwI2quq7qtoN/Ay4rG8BVX1aVTv82T8CM0dhu2aCU1Ua2rsB+OeH145zbYwJvtEI+DOA9/vMb/WXDeZG4NHBVorIChGpE5G6+vr6UaieCaqeVr0IPPj6DrpSmXGukTHBNqYXbUXkY0At8JXByqjqSlWtVdXampqc4/+YIrGrxQv4H/nATDJZZc325nGukTHBNhoBfxswq8/8TH/ZfkRkGXA7cKmqWkLWDKuh3TtMzjlyCgBrtreMZ3WMCbzRCPgvAgtE5HARiQDXAKv6FhCRE4Dv4wV763JxCEqmM3znmY38bsOhk0ZraPPy94unlxN2hB3NXeNcI2OCreCAr6pp4C+Bx4B1wC9UdY2IfElELvWLfQUoBX4pIq+KyKpBPs6Mk1+/vI1/+816Pn7XCzz79qER9Hta+DVlUaaUxdhlAd+YgozKePiq+gjwSL9lf99netlobMccHKrK3X/YzOxJccKO8C+PrOOsheN//aShrZuIGyIRcZhaHmVniwV8Ywphd9oant/UyFs7W/n0OUdw0VHT2Li7je50dryrRUN7N5MTEUSE6RUl7LQWvjEFsYBv+PFzm6koCXPZ8TM4cloZ6ayyaU/7eFeLhrYk1aVRwEvr7Gmza/3GFMICfpHbvKedx9bs4uoTZ1EScTiiphSAd+vbxrlmXgt/UiICQCLq0N6dQVXHuVbGBJcF/CJ3y72vk4g43HDaXACmVcQA2H0IDGXQ0NZNdakX8OMRl0xWSR4CqSZjgsoCfhF7c1szL2xu5OZlCzmssgSAqngEJyTjPnaNN6xCksmJCGSzlEa9/gXtyfS41suYILOAX8R+/Nxm4hGHqz6wb2gjJyRUJyLjHvA7ujN0pbJc8f6/wNcWMj35LgDtSRtewZiRsoBfpDJZ5bE1u7j46GlUlIT3W1dTFqV+nC+Q1rcmWSTvsWjng9Bez4Jt9wPQZi18Y0bMAn6Reuqt3TR3pjjXH7agryllUXa3jm8XyPcaO7gwVIciMO0YptT/AYCObgv4xoyUBfwilM5k+eKqNSyYUsqyxVMHrJ9ZFee9ho5x7RGzpaGdU0NrSU85BuZfQLxlEyGy1sI3pgAW8IvQ6rd2s62pky9cuJCSiDNg/eGTE7R0pXvHoh8PW/a0cGzoHdw5J0PVXETTTKfBcvjGFMACfpHpTmf5ztMbOawilrN1DzCvJgEwrjdfpXesJSFJZNZJUOU9EXN2aLf10jGmABbwi4iqctuv3+C1rc3cdsli3HQHvPQjqPsB7N3SW27eZO/mq0314xfwKxpe9SZmngiVXsCfKfW0Ww7fmBEblcHTTDA8/MYOfvXyVj63bAF/OmkbfOt8aN3urQwn4LpfwtzTmVFVQtgR3h2nFn4mq8zpXEN7pIpE1VxIdQJQQ/MhMcaPMUFlLfwisX5nK197/G0On5zgs8dk4J4rwY3CJx6Fz7wAFTPg3k9AVzNOSJhTnRi34RW2N3VyHBvYO+k47/mGkTgajlMlrRbwjSmAtfCLwGvvN/HR7z9HdybLj66cSeinH4VwDJavgsrZXqErVsJ/ngdP/RNc8hXmTU6MWw7/7c1bOD+0g+2zPrZvYbyaSV2tbM4cxIDf0Ui6fS+NXRmqK8rZ+f5G3NatbH5vK29sb4Y+nZZCAkvnVDGzKjHgY1QEdWOoW0I2Uko2Uk42UkY2Wo5GyognElSWRNjW5P1yiUccqhMR9rR1U1MW5b3GjgGf6YaEWZPiNHekaOzwLqZPSkSoKAmjqmzd20k6O7JeVSVhp3dIDTOxjUrAF5GLgW8ADnCnqt7Rb30U+DHwAaABuFpVN4/Gts3QdjZ3cdOP66gpi3LPVTOY+/A10NEANzy0L9gDHHYCLF3u5fTP+DyH1yR4Zn09mazihGRM69y07hkAqpec3btM4tVU721j/Wi18FOdsOl30LmX1ObnSK55mNLu3bhAz50JM/z3qcDJOSs6sk0n1WUvpXRoBc2aoIUErcRp1jitJGjWOC0ap4VE73sXYY6eWsKOPXtxM13EpJsup5SlxxzNq3sjPLepsOf9XrhkKoumlbH8tLm9I5SaiafggC8iDvBt4AJgK/CiiKxS1bV9it0I7FXV+SJyDfCvwNWFbtsM73u/fYemjm4eu2EOcx+8Cjqb4Pr7YcbSgYVP/yy8fDc8/z3mTb6R7kyWbXs7mV0dH7P6ZrOKu+kZuiRGbE6fMBuvpjq0eXRSOq//En30FqSzEYCURvlt9jjelAtpdyqZXAIdnZ0cuXAxFdOPQOOTOG1+DSXhfV1YW5Mp/rCxAc3RqhYyhDJJnEwX4XQbbqqNcKqVcKoVN93Kpq07yLQ3cGxlN1Npp6t1D26qhXI6iNM5eL2b8M7YvmftWkhriLayyaTL51A/+SR2TT2bpsqjvHTYMBTl0Td28vuNe3hy3S5+u2EP/+vsIzh3UQ1Rd2CXXRNso9HCPwnYqKrvAojIz4DLgL4B/zLgi/70vcC3RET0IN3Zs2lPO1XxMJXxyH7LtzV1kkzt68c9a1Kchrbu/e7ejEdcqksjvO//rI6FHQ6rLCGTVbY0jP8Y8Qdi0552fvr8e3z+iO3Me+BzkE7C9Q/kDvYAk+bBksug7gfMv/LPAXhnT9uYBvxVr2zizNT/0DDjbGa4ff7/EpOp4o3CR8v8w7fg8dvZVnoMt3Wv4D2dgpZO4+8vr+VvluTupppLWRlcNHnWiKpw1FArMylItkJXE3Q173ulusBxIRz3XyXQuRdatuE2b6OyZTvUr2Py+u+weP23oXwmLPoQLP4TmH2a97eDuPwEbyylx9bs5NM/eZlP3fMSc6vj/MOlR+W8E9sE12gE/BnA+33mtzLwF3BvGVVNi0gzUA3sGYXtD/DUNz/JLq1kyvylZOI1dLhVbNzTwZp33yNCmgwhYnh50K1aQzMJoG9rSImQJk4XYTKcPH8qe5pb2L6niRKSCNBOlA6N0UEUAcKkiZAmTJqopKigHZc0Wf+6eJYQWYQsgu43LThkccgQJ0mpdBLxP6tHhhApHLpxcchSQTul0kmDVlAiSaawly6iNGkCR7JMopUjZDuzQ7tZHWlm1nvboXq+F+ynLB76H++0z8Ka+1i87dfAQt7Z3XZQT/rWrhQ/+P1mWrtSdOzdxcL136HabSV1zqf2L1hSRSUFPolr/aPw+O3smHERZ7/zMS44agY/uPhIZlSWEAsfIq1ZJwzxSd5rJNob4O3fwFsPeb/WXvg+lFTBwg96wf+I87wvixwuOmoaf7ztfB58bTtfe3w9n/jhi5y3aApf/vDRzKjM/TejRpVkexO/evZVmts7cTVJNNNBLN1KLNNKItvMidUpyjNN0NkIiSnQ3QodjV6KUkIQciDk9nk54Ma8zgm710EkAZEy74syVg5l07wvT4BsGtJdXnknDBWzAPF+JcWrIdPtbaNiJkTLINnm/R85EUhMBnG8zwzHvXLZjF+nQ6tfzCF30VZEVgArAGbPnj1M6RzS3VybeJF4507Y9F/7rxskNZl2SkhFKlBAk+1Esx249Lmjc+vQf38o6opOpq10DuVVc2H+zbD0494BP5wZS2HumcRfXsm0xNd4e1frgW9cFZq3wq43oXWH12JNJ8l0d7JldxNkksQ6d1PSuYNQy1aWZ1LE6CYmKXBh71HXU7XwvP0/M1pOnE6SqRH2w296H+77FB3VR3Puu9fiumH++uIjex/4MmEkquGE67xXdztsfBLeehjWPwyv/dQLSEdfCefeDuXTB/x5TVmUPz/jcC49/jDuePQt7n1pK9t/9CKPfPZMQiO8lvPu9nreX/8ypS3vsH3jqyRa3mEO23sbNWV0UEoHUcnyZ0N8TteWMDspo4lSqniRdkpoopQmylEgRBaXNA7dOGRwyRAlRYJOdoamEnJcIqndtBOjLNRIjWyGTIqwGyIejZB1IoQy3WST7ZRl9o5oXzMqOOIlLlLq0EAFitBEmV+3LFFSZAiRxmtkdOPSTYQkYboJk8ah063g4r97YER1GMpoBPxtQN/ftjP9ZbnKbBURF6jAu3g7gKquBFYC1NbWHnjKx40Q/5v10LYb6tdDe73XAkgnoXSK962fSXu9VCQETe/hNm/F7Wr2AlW0FCKlEIl77yHX+7Z2I+CWeH+HQKrDO6G627zPcSL+K+y1EmKV3s9oVe+Fep+Dgmb7vHRfyyRc4rUenKj3WT052Gza+6mf6faWxSohWg5tu7wgXjbNuwjZ1eS1NEqqiMXKGXG/i9M/Bz+5kuWVL/GbXRX5/1392/CHb8Lbj0H77gGrBYfp6pDCZadW8bZOZgcncOLcaubPmg6JGph9KlUzTxz42dEyQiiSGkFaTRXu+xTpTJor96wgHI3z3G3n946xP2FFEl6Kbsll3vGz+Xew9gF49aew5n646J9h6fU5c/2TS6N89SPHcfr8av7Pz1/jZy++z7UnzUKGuS7Qlkyzq76elrWr2fnGUxzR9SaHd29knngNqGPUoSE6k+bEQjpCMVBll5sg6ZTS6ZQzedosjp8z2TsXwnEoqYSSKt5pDfOLN5oZye+7rMKDr22nvi3JJcdMZ3p5jN+s2cnWvZ2cMLuS195vItvvsDp7fhVHTisjpBni6SZSoSghzRDr2MHz67bQmI6wtEY5akqUknQzQpZouo3OlgY2NXZSEY8xvzpKaWoP4UwX4WwXWXHIAh2hGCHd16B0NEVCuynPpnCzSUKaIhXZfwTb0TIaR/yLwAIRORwvsF8DA76oVwHLgeeAq4CnDlb+vlfpFO81kZXW7JsOl4w8DdDf/PNhylFc1fxLvtu4lGxWh27dqcIfvomu/hIpwrxXczbzzlpG6LDjofwwMtFyLvrWC2zc08Vxsyr5j48ehwvMBhZHXaaW5/HVFC0DwEmN4N6AtQ/Alt/zd6kbeUem8I+XLp74wb4/J+ylc444D06/GVZ9Fh78LLyzGv70G17aJ4cPHXMY33vmXf72vjd46PXt/MvlxzB38v6/FN9v7OCeP25m6t6XmbLhv7lInyMsGZIaZlN0IasnXc38488kOn0J0ZojmF5VxsDfFkM7ogpumz1j+IKD+OuLjqSlM8UU/1j7mw8uoqGtm2kVMepbk7R2pXrLRtwQM6sGv251XVeKxrZu5lTHc34Bbm/qZGp5bMx7t+Wj4KPez8n/JfAYXrfMH6jqGhH5ElCnqquAu4D/EpGNQCPel4I5VInAuX9Lzc+v49LME2xrOpdZkwY5ATIpWPW/4bX/5rWys7hpz7Xs2VLBvPYEDe176U43oChdqSyfOfcIrjt5Tu/TtQ5IrBwA90ADviqZ1V9mo87m3VlX8Oy1tdbnfNI8uH6V92vsqS/Dtpfhyjth9ikDikbcED//5Cl89fH13PPH9zjnq89QGnW55JhpvL61me0NLZyf+T03uY9wVGgLncTYOPcaGmdfyMKl57GoqpxF47CL/cXCzn7XacJOqPc4qCmLUlOWf762PBamPDZ4C3xEx/cYGZVmjqo+AjzSb9nf95nuAj4yGtsyY2TRh2idejK37Pw5azd+lFknnTawTCYFv7oJ1t5P22m38JHfHs91p87FDQl3/n4Ti6eXc+aCyQBUlIT55FnzcJ0RXsSKegE/nDrAawrbXsJp3Mid6RV88bJjLdj3CIXgjM/B3DPhV38OP/wgnPF5b5n/a6pHZTzCP334GG44bS6/fGkrdZv3cl/dZpZHnuKz0YcoT+8hWbUAzvgmJcdcxeJ8rhWZcVFkv2tN3kTQy79H8rtnc/TqG2D6T2BWn9x647vw4M2w6Vm48J+5q+MiUpm3uf7UOcyrKeXGMw9nallsxBf6BugJ+OkDy+E3v3I/cXWYfspHWTy9fHTqMpHM/AB88nfwyF/D777q3Xh39i3wgRu83i19zJ9Sxm2nV6Clj5LpuBO3dRvMOB1O/y7R+csOuR4pZiAL+GZQ5dPm8fHwP/CN7FfgrmUw7Vjv7tzWnbDtJe+i9mXfpvuYP+Mn//oUZy+sYZ7f62V6xSj/rPVbnZHMgaV0Um8/ycu6gGvOPHp06zORxMrhiu/DSX8BT34RHr0Fnv0KLLwIph4DmoGW7bD9FdjyBwTFnXMGXPpN73pPHjd4mUODBXwzJDnsWD7Z/HV+eeJ6ePdpr2UfLYNzbvN6eJRP57HXtrO7NckdV845eBXxA340cwAt/FQnVa3rWRe+gpMP4bzqIWNmLSx/0LuQ+8pPYO2D8Mo93jo35t3DcfYtcOzVUH3E+NbVjIgFfDOk42ZW8O0N9fxu6nU8tfcCFi0u4+oT990foar88H82Mac6zjkLD2KvqBEE/PSON3DJkpxyzMGq1cQjAvOXeS9VaN/j9fCJlnndh02gWcA3Qzp7YQ3/76mNfPyuF3qXvbWzlS9ceCRtXWk++v3neK+xg3/68NGjl6/Pxb8Q6Gbzf7j6xld/zyJg0dKzDlKlJjiR/bv+msCzgG+GtHR2FbdfspjWZJprTpzFlx5cyw//ZzP3v7KNqOvQ1NnNP/zpEq47eQR3RR8IJ0wGh3B2iMHF+mndVEejlnHaCccdxIoZExwW8M2QQiHhL86a1zv/3Y8t5f5Xt/Hrl7cREuGTZ8/jtCMmj0ldUk4J0VQy7/KVTWvYEV/EJBv10RjAAr45QCLC5SfM7B1hcSylQiVEyS+lk0qnmZXdxpuVOe4fMKZIWcdZExhpp4SYJslnVI76bZuJSQqpnjdsWWOKhQV8ExhpJ0acZF6P8mvcuh6A2NT5B7taxgSGBXwTGBmnhBK6SOXxXNuOXRsBqJp5KIzkYsyhwQK+CYy0W0JckqQyw7fwteFdUuowZabdIGRMDwv4JjCybgkldOfVwo+0bGFXqAY3HBm2rDHFwgK+CYyMG6eEJOk8Wvjlne+zJzLy8dONmYgs4JvA0N6UzvAt/MnpnXTGx77rqDGHMgv4JjCyfgt/uICf7Gyjgja07ECfq2TMxFZQwBeRSSLyhIhs8N8HPCdNRI4XkedEZI2IvC4iVxeyTVO8NFxCnC7SwwT8Xdu2ABCdZC18Y/oqtIV/K7BaVRcAq/35/jqA61X1KOBi4OsiUlngdk0R0nACR5RU99B32zbu9AJ+YvKssaiWMYFRaMC/DLjbn74b+HD/Aqr6tqpu8Ke3A7sBG4LPHLiI91zdbHfHkMXa92wFoGraQR7QzZiAKTTgT1XVHf70TmDqUIVF5CQgArwzRJkVIlInInX19fUFVs9MKGHvISbZ5NBPvZJW75Asq7GAb0xfww6eJiJPAtNyrLq974yqqogM2l9ORKYD/wUsV9VBk7CquhJYCVBbWzt8/ztTPMLemPjZ5NAPQQl37KRLw5SUTRqLWhkTGMMGfFVdNtg6EdklItNVdYcf0HcPUq4ceBi4XVX/OOLamqIWinopHR0mpRPt3M1umcRse6i2Mfsp9IxYBSz3p5cDD/QvICIR4D7gx6p6b4HbM0VM/KdeaffQLfx4dz2NoeqxqJIxgVJowL8DuEBENgDL/HlEpFZE7vTLfBQ4C7hBRF71X8cXuF1ThPYF/KFb+GXde2hyxuahLMYESUEPQFHVBuD8HMvrgJv86XuAewrZjjEAjp/SYZiAX55ppD1+0hjUyJhgsSSnCYxQtNSbSA2R0kl1UaKdJCMD7gE0puhZwDeB4US9lI6kh3iQeUcDAOmo9dAxpj8L+CYw3FhPC3+IlI4f8DMlFvCN6c8CvgkM18/hh9KDD62g7Xu8ibj10jGmPwv4JjDccJikukhq8JROssW7OzuUsF46xvRnAd8EhhsSkkRwMoMH/O4W794/t8yGazKmPwv4JjCckNBJhFBm8JROqrWejAoRG1bBmAEs4JvAcEMhOjWKM0QOP9u2hyZKKSuJjmHNjAkGC/gmMJyQ0DVMSkc7GmjUcspi4TGsmTHBYAHfBEoXUZxMctD1oc4GGimjPFbQTeTGTEgW8E2gJCWCmx08peN27aVRy6yFb0wOFvBNoCSJ4g5x0TaSbGSvllIZt4BvTH8W8E2gdEl08Ba+KtF0C+1OObGwM7YVMyYALOCbQOkeKuCnOnE0TSZSMbaVMiYgLOCbQElKlPBgAb+rGQCNlY9hjYwJDgv4JlBSEiWcHaSXjh/wQzFr4RuTS8EBX0QmicgTIrLBfx90IHIRKReRrSLyrUK3a4pTt8S8gK85nm/vB3w3bmPhG5PLaLTwbwVWq+oCYLU/P5gvA8+OwjZNkeoORXHIQCY1YJ127gUgbMMqGJPTaAT8y4C7/em7gQ/nKiQiHwCmAo+PwjZNkeoWf8iEHGPid7Y0AhCzgG9MTqMR8Keq6g5/eideUN+PiISArwF/NdyHicgKEakTkbr6+vpRqJ6ZSFKhmD8xcHiF9lYv4JdW2Fj4xuSS1/3nIvIkMC3Hqtv7zqiqikiO5CqfBh5R1a0iMuS2VHUlsBKgtrY212eZIpZ2Bm/hJ/2AX15pLXxjcskr4KvqssHWicguEZmuqjtEZDqwO0exU4EzReTTQCkQEZE2VR0q32/MAOmeFn6OETNT7Xvp1AiTyq1bpjG5jMYIU6uA5cAd/vsD/Quo6nU90yJyA1Brwd6MRHqIlE66o4kW4lSXRsa4VsYEw2jk8O8ALhCRDcAyfx4RqRWRO0fh843ptS+Hn+NB5p3NNGuCqrgFfGNyKbiFr6oNwPk5ltcBN+VY/iPgR4Vu1xSnjDt4Cz+UbKYlVErEtfsJjcnFzgwTKOkhWvhuqoWkUzrGNTImOCzgm0DJOoO38KPpVlLhsjGukTHBYQHfBEraKfEmcgT8kmybjZRpzBAs4JtA0cFy+KoktN1GyjRmCBbwTaBk3EFa+N3tuGTBRso0ZlAW8E2ghJwwKdwBF22T7d5dthqrHIdaGRMMFvBNoDghIUlkQAu/yx84TayFb8ygLOCbQHFDQheRAS38Ln8cHSmpHIdaGRMMFvBNoDghoYvogLF0Uu3+WPiJynGolTHBYAHfBErYCdGZo4XfE/DdhD3typjBWMA3geKEhE6NDsjhpzu8xxtGSy3gGzMYC/gmULwcfnhAwNfOJgBiFvCNGZQFfBMoTkjo0OiAlI52NdOmMRIlJeNUM2MOfRbwTaC4IaFTB3bLDHV5Y+EnoqPxiAdjJiYL+CZQnFCIDiJo/4Df3UKLJkhEnXGqmTGHvoICvohMEpEnRGSD/54zgSois0XkcRFZJyJrRWRuIds1xct1hK4cF23d7hZaiRN1LeAbM5hCW/i3AqtVdQGw2p/P5cfAV1R1MXASuZ97a8ywBrvxKpxqpSNkY+EbM5RCA/5lwN3+9N3Ah/sXEJElgKuqTwCoapuq5ng+nTHDc0JCB/5FW9Xe5dF0K50W8I0ZUqEBf6qq7vCndwJTc5RZCDSJyK9F5BUR+YqIDPq7W0RWiEidiNTV19cXWD0z0bghoUNjiGb3u9s2lmmjy7WHnxgzlGG7NIjIk8C0HKtu7zujqioimqOcC5wJnAC8B/wcuAG4K9f2VHUlsBKgtrY21+eZIuY4IdqJejPd7RAugWyWkmwb3RbwjRnSsAFfVZcNtk5EdonIdFXdISLTyZ2b3wq8qqrv+n9zP3AKgwR8Y4bihoQO/IegdLdDYjIkmwmhpCL28BNjhlJoSmcVsNyfXg48kKPMi0CliNT48+cBawvcrilSTkho1z4BH8C/y7Y7bEMjGzOUQgP+HcAFIrIBWObPIyK1InIngKpmgL8CVovIG4AA/1ngdk2RGtDCB+j0Bk7L2sNPjBlSQbclqmoDcH6O5XXATX3mnwCOLWRbxgC4Toh27cnht3nvXU2APe3KmOHYnbYmUHK38JsAC/jGDMcCvgkUJyS09wv46qd0NFo5TrUyJhgs4JtAcXtGy4TelE7af/iJlthFW2OGYgHfBIqTI6WT7WikS8O40cQ41syYQ58FfBMobihEB1EUgWQrANmOvTRRStS1w9mYodgZYgLFdQQlRCZcCskWwMvhN2vCAr4xw7AzxASKGxIA0uEy6PICPp1NNJMgGrahkY0ZigV8EyjOfgHfe3C5JJtpVkvpGDMcO0NMoLgh75BNhct6UzpOl9/Ct4BvzJDsDDGB0tPCT/Vp4TvJZj+HbykdY4ZiAd8Eiut4Ab/bLfUCfqoTJ91Og5YRC9vhbMxQ7AwxgdLTwu92/ZROmzcidz2V1sI3ZhgW8E2ghP0cftIt9XrptO4EoF4riFoL35gh2RliAsXxUzpd4UmgGah/C4A9WmEXbY0Zhp0hJlB6+uG3R6q9BbveBKBeLaVjzHAs4JtA6cnht4cneQt2vI4iNFBuKR1jhlHwGSIik0TkCRHZ4L9XDVLu30RkjYisE5FviogUum1TfHpa+G1hv4X//vO0RaeSxrWUjjHDGI0z5FZgtaouAFb78/sRkdOA0/GeenU0cCJw9ihs2xSZnhZ+m+O38FH2xmYCEHEs4BszlNE4Qy4D7van7wY+nKOMAjEgAkSBMLBrFLZtikzPnbYdoQTEJwPQEJlB1A1hPxqNGdpoBPypqrrDn94JTO1fQFWfA54Gdvivx1R1Xa4PE5EVIlInInX19fWjUD0zkfS08LMAca+VvyV+lKVzjMlDXg8xF5EngWk5Vt3ed0ZVVUQ0x9/PBxYDM/1FT4jImar6u/5lVXUlsBKgtrZ2wGeZ4tY7WmZG4eL/Cy//mBedC4mGG8a5ZsYc+vIK+Kq6bLB1IrJLRKar6g4RmQ7szlHscuCPqtrm/82jwKnAgIBvzFBCIUEEMtkszF8G85fR+YtXrYVvTB5G4yxZBSz3p5cDD+Qo8x5wtoi4IhLGu2CbM6VjzHAcEdLZfT/+kumsBXxj8jAaZ8kdwAUisgFY5s8jIrUicqdf5l7gHeAN4DXgNVV9cBS2bYqQExIy2ifgp7J205UxecgrpTMUVW0Azs+xvA64yZ/OAJ8sdFvGgJfHz2T6tvAzdtOVMXmws8QETihkKR1jRsLOEhM4bkjIDAj4ltIxZjgW8E3gOKFQvxx+xlr4xuTBzhITOANz+FmiYWvhGzMcC/gmcJz+OXxr4RuTFztLTOA4IfFuvPLZRVtj8mNniQkcNyT0yejYRVtj8mQB3wTOwBZ+hoi18I0Zlp0lJnCckHiDpwHZrJLKKDG78cqYYdlZYgLH6dMPvzvjtfQtpWPM8Czgm8Bx+4ylk0z1BHw7lI0Zjp0lJnD6tvCT6QyA5fCNyYOdJSZw+ubwk2lr4RuTLztLTOD0HR65N+DbnbbGDMsCvgkcNxQakNKxFr4xwyvoLBGRj4jIGhHJikjtEOUuFpH1IrJRRG4tZJvG9B1aoaeFbzl8Y4ZX6FnyJnAF8OxgBUTEAb4NfBBYAlwrIksK3K4pYn1vvLJeOsbkr6AnXqnqOgARGarYScBGVX3XL/sz4DJgbSHbNsXLC/jetPXDNyZ/Y9EsmgG832d+q78sJxFZISJ1IlJXX19/0Ctngsfdr4VvOXxj8jVsC19EngSm5Vh1u6o+MNoVUtWVwEqA2tpaHaa4KUK5cvgW8I0Z3rABX1WXFbiNbcCsPvMz/WXGjMj+N15ZSseYfI1Fs+hFYIGIHC4iEeAaYNUYbNdMUPuNpdPbD99a+MYMp9BumZeLyFbgVOBhEXnMX36YiDwCoKpp4C+Bx4B1wC9UdU1h1TbFzM0xtIKldIwZXqG9dO4D7suxfDtwSZ/5R4BHCtmWMT2cUMj64RszAnaWmMBxQuxr4fv98COOHcrGDMfOEhM4fYdW6M5kcEOCawHfmGHZWWICZ79eOil7gLkx+bIzxQSOGxJS/h22yXTW8vfG5MnOFBM4/R+AYn3wjcmPBXwTOK5/p62q0p3OWh98Y/JkZ4oJnJ4LtJmskkxbDt+YfNmZYgLHdbzRWdN+wLccvjH5sTPFBE445B22qUzWcvjGHAAL+CZwelv4GT+Hby18Y/JiZ4oJnJ4cfiqbtRy+MQfAzhQTOOHQvhZ+MmU5fGPyZWeKCZyeFn46o5bDN+YAWMA3gRP2c/ipbNZy+MYcADtTTOC4ob4tfLvxyph82ZliAqenl47XLTNLxLGUjjH5KPSJVx8RkTUikhWR2kHKzBKRp0VkrV/25kK2aUx4vxuvMtbCNyZPhZ4pbwJXAM8OUSYNfEFVlwCnAJ8RkSUFbtcUsZ6UTjKVIZVRy+Ebk6dCH3G4DkBEhiqzA9jhT7eKyDpgBrC2kG2b4tWT0mntSgOQiBR0GBtTNMa0aSQic4ETgOeHKLNCROpEpK6+vn7M6maCo6eF39KVAqAkYjl8Y/IxbNNIRJ4EpuVYdbuqPpDvhkSkFPgV8DlVbRmsnKquBFYC1NbWar6fb4pHTwu/udML+HEL+MbkZdiAr6rLCt2IiITxgv1PVPXXhX6eKW49g6e1dHopHQv4xuTnoKd0xEvw3wWsU9V/P9jbMxPfvhx+T0rHcvjG5KPQbpmXi8hW4FTgYRF5zF9+mIg84hc7Hfg4cJ6IvOq/Limo1qao9XTL7MnhWwvfmPwU2kvnPuC+HMu3A5f4078HBu/GY8wBcvuldErCFvCNyYd1YDaB4/Zr4SeiltIxJh8W8E3ghJ39u2VaSseY/FjAN4Hj+uPh96Z0LOAbkxcL+CZw3P4tfMvhG5MXC/gmcMJ9bryKOKHeLwBjzNDsTDGB05PDV7V0jjEHwgK+CZywE+rN49sFW2PyZwHfBFJP33tr4RuTPwv4JpB6Ar0NjWxM/izgm0DqCfjWwjcmfxbwTSD1pHQsh29M/izgm0DqadlbwDcmfxbwTSD1XrQNWw7fmHxZwDeBZCkdYw6cBXwTSJbSMebAFfoAlI+IyBoRyYpI7TBlHRF5RUQeKmSbxvRVakMjG5O3Qlv4bwJXAM/mUfZmYF2B2zMGgHfq2wFYPL18nGtiTHAUFPBVdZ2qrh+unIjMBD4E3FnI9ozp0Z70hkY+blbl+FbEmAAZq9/DXwduAcqGKygiK4AVALNnzz64tTKB9b2PfYBnN9RTUxYd76oYExjDBnwReRKYlmPV7ar6QB5//yfAblV9SUTOGa68qq4EVgLU1tbqcOVNcVpyWDlLDrN0jjEHYtiAr6rLCtzG6cClInIJEAPKReQeVf1YgZ9rjDHmABz0bpmqepuqzlTVucA1wFMW7I0xZuwV2i3zchHZCpwKPCwij/nLDxORR0ajgsYYY0ZHQRdtVfU+4L4cy7cDl+RY/gzwTCHbNMYYMzJ2p60xxhQJC/jGGFMkLOAbY0yRsIBvjDFFQlQP3XubRKQe2DLCP58M7BnF6oynibIvE2U/wPblUGX7AnNUtSbXikM64BdCROpUdcgRPINiouzLRNkPsH05VNm+DM1SOsYYUyQs4BtjTJGYyAF/5XhXYBRNlH2ZKPsBti+HKtuXIUzYHL4xxpj9TeQWvjHGmD4s4BtjTJGYcAFfRC4WkfUislFEbh3v+gxHRH4gIrtF5M0+yyaJyBMissF/r/KXi4h809+310Vk6fjVfCARmSUiT4vIWv/h9jf7ywO3PyISE5EXROQ1f1/+0V9+uIg879f55yIS8ZdH/fmN/vq547oD/YiIIyKviMhD/nxQ92OziLwhIq+KSJ2/LHDHF4CIVIrIvSLyloisE5FTD/a+TKiALyIO8G3gg8AS4FoRWTK+tRrWj4CL+y27FVitqguA1f48ePu1wH+tAL47RnXMVxr4gqouAU4BPuP/+wdxf5LAeap6HHA8cLGInAL8K/Afqjof2Avc6Je/EdjrL/8Pv9yh5GZgXZ/5oO4HwLmqenyfPupBPL4AvgH8RlUXAcfh/f8c3H1R1QnzwhuX/7E+87cBt413vfKo91zgzT7z64Hp/vR0YL0//X3g2lzlDsUX8ABwQdD3B4gDLwMn49356PY/3oDHgFP9adcvJ+Ndd78+M/3gcR7wECBB3A+/TpuByf2WBe74AiqATf3/bQ/2vkyoFj4wA3i/z/xWf1nQTFXVHf70TmCqPx2Y/fNTAScAzxPQ/fHTIK8Cu4EngHeAJlVN+0X61rd3X/z1zUD1mFZ4cF8HbgGy/nw1wdwPAAUeF5GXRGSFvyyIx9fhQD3wQz/VdqeIJDjI+zLRAv6Eo97XeaD6zopIKfAr4HOq2tJ3XZD2R1Uzqno8Xgv5JGDR+NbowInInwC7VfWl8a7LKDlDVZfipTg+IyJn9V0ZoOPLBZYC31XVE4B29qVvgIOzLxMt4G8DZvWZn+kvC5pdIjIdwH/f7S8/5PdPRMJ4wf4nqvprf3Fg9wdAVZuAp/FSH5Ui0vOkuL717d0Xf30F0DC2Nc3pdOBSEdkM/AwvrfMNgrcfAKjqNv99N97T9k4imMfXVmCrqj7vz9+L9wVwUPdlogX8F4EFfg+ECN5D01eNc51GYhWw3J9ejpcL71l+vX/F/hSguc/Pv3EnIgLcBaxT1X/vsypw+yMiNSJS6U+X4F2LWIcX+K/yi/Xfl559vAp4ym+hjStVvU1VZ6rqXLzz4SlVvY6A7QeAiCREpKxnGrgQeJMAHl+quhN4X0SO9BedD6zlYO/LeF+8OAgXQy4B3sbLt94+3vXJo77/DewAUnjf+jfi5UxXAxuAJ4FJflnB64X0DvAGUDve9e+3L2fg/QR9HXjVf10SxP0BjgVe8fflTeDv/eXzgBeAjcAvgai/PObPb/TXzxvvfcixT+cADwV1P/w6v+a/1vSc30E8vvz6HQ/U+cfY/UDVwd4XG1rBGGOKxERL6RhjjBmEBXxjjCkSFvCNMaZIWMA3xpgiYQHfGGOKhAV8Y4wpEhbwjTGmSPx/vzE+vyV8ZmkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for _ in range(15):\n",
    "    idx_test = np.random.randint(df_train.shape[0])\n",
    "    plt.plot(waveforms_train[idx_test,6,:])\n",
    "    plt.plot(reconstructions_train[idx_test,6,:])\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "fe1cc2ca-dc6a-4410-83a1-d68c8aab142f",
   "metadata": {},
   "outputs": [],
   "source": [
    "x_cols = latent_rep_df_train.columns[latent_rep_df_train.columns.str.startswith('latent_')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "22807336-ff6e-4db7-bb19-31c18c896954",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['latent_1', 'latent_5', 'latent_6', 'latent_8', 'latent_9', 'latent_10',\n",
       "       'latent_11', 'latent_12', 'latent_13', 'latent_15', 'latent_16',\n",
       "       'latent_17', 'latent_19', 'latent_22', 'latent_23', 'latent_25',\n",
       "       'latent_26', 'latent_27', 'latent_30', 'latent_31', 'latent_32'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "c3754575-236f-49ec-af5c-e8811e07543a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# filled with best values based on hyperopt\n",
    "params = {'colsample_bytree': 0.88, 'gamma': 0.21, 'learning_rate': 0.06, 'max_depth': 5, 'min_child_weight': 8.0, 'n_estimators': 260, 'subsample': 0.74}\n",
    "\n",
    "ef_xgboost = xgb.XGBRegressor(objective = \"binary:logistic\",\n",
    "                             #scale_pos_weight=1/0.075073, \n",
    "                             **params,\n",
    "                             eval_metric='logloss'\n",
    "                        ).fit(latent_rep_df_train[x_cols], df_train['Label'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "6afd48b2-9ee4-406a-bb1d-231fce4ab389",
   "metadata": {},
   "outputs": [],
   "source": [
    "latent_rep_df_test['Prob'] = ef_xgboost.predict(latent_rep_df_test[x_cols])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "4de1d83a-3c7e-4f96-a053-c01a353c92ba",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>Label</th>\n",
       "      <th>False</th>\n",
       "      <th>True</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Prob</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>False</th>\n",
       "      <td>3172</td>\n",
       "      <td>926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>True</th>\n",
       "      <td>1219</td>\n",
       "      <td>352</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Label  False  True \n",
       "Prob               \n",
       "False   3172    926\n",
       "True    1219    352"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(latent_rep_df_test['Prob'] > 0.2, df_train['Label'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "dfdc1547-f42d-438a-bfa6-71494cb33a16",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.8894241781340604\n",
      "C-stat CI: 0.8772661424954669 0.9008777372247756\n"
     ]
    }
   ],
   "source": [
    "print(roc_auc_score(df_test['Label'], latent_rep_df_test['Prob']))\n",
    "bootstrap_ci(roc_auc_score, df_test['Label'], latent_rep_df_test['Prob'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "8cff9270-6045-4411-a181-32b86a8aae34",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.6629096190972908\n",
      "C-stat CI: 0.6305438283430667 0.6940583596378074\n"
     ]
    }
   ],
   "source": [
    "print(average_precision_score(df_test['Label'], latent_rep_df_test['Prob']))\n",
    "bootstrap_ci(average_precision_score, df_test['Label'], latent_rep_df_test['Prob'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "25289694-933d-4cca-bfe3-6f19c6f70c2f",
   "metadata": {},
   "outputs": [],
   "source": [
    "ef_xgboost.save_model('/training/secondary_task/xgboost_ef.save')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "69556d59-fa3f-4b2a-9014-ff1fabe1c44a",
   "metadata": {},
   "source": [
    "### Perform 'black box'-based prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "caed591c-1004-4171-a6a0-ba7557c09800",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0,1]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a357fee36ed94c39a747c9938de1bb90",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Predicting: 0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y, y_prob = get_prediction(ef_vencoder, df_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "8b0332e8-abb0-40a1-8928-592ce8180711",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.906157260739824\n",
      "C-stat CI: 0.8954855726623766 0.9163456154854548\n"
     ]
    }
   ],
   "source": [
    "print(roc_auc_score(y, y_prob))\n",
    "bootstrap_ci(roc_auc_score, y, y_prob.squeeze())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "6e5edc3c-469b-4ff1-a05b-3901f3f702ec",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7088961608679234\n",
      "C-stat CI: 0.6792007818187509 0.7396276267331521\n"
     ]
    }
   ],
   "source": [
    "print(average_precision_score(y, y_prob))\n",
    "bootstrap_ci(average_precision_score, y, y_prob.squeeze())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "879d2a28-121f-4d06-805f-6a1ee77ae9d0",
   "metadata": {},
   "source": [
    "## Mortality model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eab69e6b-4c50-49a1-bde5-001de574b0fd",
   "metadata": {},
   "source": [
    "### Get data and dataloaders"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "517060d1-4990-4c24-a812-28d6d23823f5",
   "metadata": {},
   "outputs": [],
   "source": [
    "trainset_path = '/training/secondary_task/train_mortality.csv'\n",
    "df_train = pd.read_csv(trainset_path)\n",
    "df_train['Gain'] = 0.00488\n",
    "testset_path = '/training/secondary_task/test_mortality_single.csv'\n",
    "df_test = pd.read_csv(testset_path)\n",
    "df_test['Gain'] = 0.00488\n",
    "df_test = df_test.sort_values('AcquisitionDateTime').groupby('PseudoID').head(1).reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c89baa45-62a7-40fa-8827-6e4d69f055f2",
   "metadata": {},
   "source": [
    "### Perform age/sex based prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "e578add3-d981-420a-bda4-31d051e727c3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# filled with best values based on hyperopt\n",
    "params = {\n",
    "    'colsample_bytree': 0.66,\n",
    "    'gamma': 0.44,\n",
    "    'learning_rate': 0.18,\n",
    "    'max_depth': 8,\n",
    "    'min_child_weight': 7.0,\n",
    "    'n_estimators': 50,\n",
    "    'subsample': 0.93\n",
    "}\n",
    "\n",
    "mort_xgboost = xgb.XGBRegressor(objective = \"binary:logistic\",\n",
    "                         **params,\n",
    "                         eval_metric='logloss'\n",
    "                        ).fit(df_train[['Age', 'Gender']], df_train['Label'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "0e0d7af6-c842-4a34-8a77-36c2e612d3f2",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_test['Prob_AS'] = mort_xgboost.predict(df_test[['Age', 'Gender']])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "e710108e-b4e2-4bce-9d46-3fa56b5d77b0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>Label</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Prob_AS</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>False</th>\n",
       "      <td>39155</td>\n",
       "      <td>2082</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>True</th>\n",
       "      <td>26490</td>\n",
       "      <td>3252</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Label        0     1\n",
       "Prob_AS             \n",
       "False    39155  2082\n",
       "True     26490  3252"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(df_test['Prob_AS'] > 0.08, df_test['Label'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "ea255e12-f88a-453b-935b-32a4130ee8ed",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.647798313427746\n",
      "C-stat CI: 0.6407608599771043 0.6554021053956709\n"
     ]
    }
   ],
   "source": [
    "print(roc_auc_score(df_test['Label'], df_test['Prob_AS']))\n",
    "bootstrap_ci(roc_auc_score, df_test['Label'], df_test['Prob_AS'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "dd58f9d2-7129-4ace-94e5-d2895a5252bf",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.12013478927501976\n",
      "C-stat CI: 0.11508818685871797 0.12568509610599216\n"
     ]
    }
   ],
   "source": [
    "print(average_precision_score(df_test['Label'], df_test['Prob_AS']))\n",
    "bootstrap_ci(average_precision_score, df_test['Label'], df_test['Prob_AS'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b6918e87-2b12-4a4f-a9cd-65827f8b82d7",
   "metadata": {},
   "source": [
    "### Perform FactorECG-based prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "5ac0d35e-44be-4a4f-849f-4339f3bc20df",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0,1]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5876ea61e55c47d0a32bcc6ddb43e4d3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Predicting: 0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n"
     ]
    }
   ],
   "source": [
    "latent_rep_df_train, waveforms_train, reconstructions_train = get_latents(vae, df_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "f613f3dc-2416-4109-bd9d-2b2b3ec5ad71",
   "metadata": {},
   "outputs": [],
   "source": [
    "latent_rep_df_train['Gender'] = df_train['Gender']\n",
    "latent_rep_df_train['Age'] = df_train['Age']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "8d09bb81-cec1-42da-a165-b709d504b93d",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0,1]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d5cb53e959a3442aa88c5d28152bf8b3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Predicting: 0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n"
     ]
    }
   ],
   "source": [
    "latent_rep_df_test, waveforms_test, reconstructions_test = get_latents(vae, df_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "657a213b-a777-463d-b0d1-0a11adaef3a4",
   "metadata": {},
   "outputs": [],
   "source": [
    "latent_rep_df_test['Gender'] = df_test['Gender']\n",
    "latent_rep_df_test['Age'] = df_test['Age']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "60b03a9d-a627-455f-9d28-6f58b85574c3",
   "metadata": {},
   "outputs": [],
   "source": [
    "x_cols = latent_rep_df_train.columns[latent_rep_df_train.columns.str.startswith('latent_')].tolist() + ['Age', 'Gender']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "c4ccb31f-d4cf-469c-95fd-27b4d4520f92",
   "metadata": {},
   "outputs": [],
   "source": [
    "# filled with best values based on hyperopt\n",
    "params = {\n",
    "    'colsample_bytree': 0.66,\n",
    "    'gamma': 0.44,\n",
    "    'learning_rate': 0.18,\n",
    "    'max_depth': 8,\n",
    "    'min_child_weight': 7.0,\n",
    "    'n_estimators': 50,\n",
    "    'subsample': 0.93\n",
    "}\n",
    "\n",
    "mort_xgboost = xgb.XGBRegressor(objective = \"binary:logistic\",\n",
    "                         **params,\n",
    "                         eval_metric='logloss'\n",
    "                        ).fit(latent_rep_df_train[x_cols], latent_rep_df_train['Label'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "fbd11ca2-389c-4d20-9670-fb02b77901a3",
   "metadata": {},
   "outputs": [],
   "source": [
    "latent_rep_df_test['Prob'] = mort_xgboost.predict(latent_rep_df_test[x_cols])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "ba822930-1a3d-4a0b-9419-95a469a45a46",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>Label</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Prob</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>False</th>\n",
       "      <td>49894</td>\n",
       "      <td>2017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>True</th>\n",
       "      <td>15751</td>\n",
       "      <td>3317</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Label      0     1\n",
       "Prob              \n",
       "False  49894  2017\n",
       "True   15751  3317"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(latent_rep_df_test['Prob'] > 0.08, latent_rep_df_test['Label'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "815ec5c8-07cf-42cb-96bc-5939d0a18a8f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7630942563743246\n",
      "C-stat CI: 0.756466058714179 0.769252746221119\n"
     ]
    }
   ],
   "source": [
    "print(roc_auc_score(df_test['Label'], latent_rep_df_test['Prob']))\n",
    "bootstrap_ci(roc_auc_score, df_test['Label'], latent_rep_df_test['Prob'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "2d98b1f8-c119-4db1-992c-0c8e2c1167d1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.21325298157725914\n",
      "C-stat CI: 0.2042005518976031 0.22357932281136683\n"
     ]
    }
   ],
   "source": [
    "print(average_precision_score(df_test['Label'], latent_rep_df_test['Prob']))\n",
    "bootstrap_ci(average_precision_score, df_test['Label'], latent_rep_df_test['Prob'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "1b168cf9-63fa-4d59-b4c7-8609e0a5fde2",
   "metadata": {},
   "outputs": [],
   "source": [
    "mort_xgboost.save_model('/training/secondary_task/xgboost_mortality.save')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "17360d1e-c5e3-41fb-932c-8f7ef81f9446",
   "metadata": {},
   "source": [
    "### Perform 'black box'-based prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "facb8d24-9d24-4eee-a221-a2f7a8696bbf",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0,1]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4220ef80712c4a558a16b8a0fe3070d8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Predicting: 0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n",
      "The To12Lead transform only works with 8 channel input ECGs, please check.\n",
      "Above error was caught in dataloader, returning neighbouring sample to continue training\n"
     ]
    }
   ],
   "source": [
    "y, y_prob = get_prediction(mort_vencoder, df_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "81fb3aa3-16b8-441e-b58d-2dd537db60c5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7506415437173628\n",
      "C-stat CI: 0.7443623460004999 0.7572009336697407\n"
     ]
    }
   ],
   "source": [
    "print(roc_auc_score(y, y_prob))\n",
    "bootstrap_ci(roc_auc_score, y, y_prob.squeeze())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "e7c5fe95-a74f-4542-ac2f-c71bbe7aa41b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.20804906608577498\n",
      "C-stat CI: 0.19898104798618582 0.21778075748195366\n"
     ]
    }
   ],
   "source": [
    "print(average_precision_score(y, y_prob))\n",
    "bootstrap_ci(average_precision_score, y, y_prob.squeeze())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f00ad5b2-d06d-4558-922a-76ce8a0d34ab",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}