notebooks/label_freq.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "a3010bb5-6676-4823-820a-b6b6df675703",
   "metadata": {},
   "outputs": [],
   "source": [
    "# import glob\n",
    "# import os\n",
    "# import sigmf\n",
    "# import torch\n",
    "# import torchsig\n",
    "\n",
    "# import numpy as np\n",
    "# import torch.nn.functional as F\n",
    "# import torchsig.transforms as ST\n",
    "\n",
    "# from datetime import datetime\n",
    "# from models import CustomNetwork\n",
    "# from pathlib import Path\n",
    "# from pytorch_lightning import LightningModule, Trainer\n",
    "# from pytorch_lightning.callbacks import ModelCheckpoint\n",
    "# from sigmf_db_dataset import SigMFDB\n",
    "# from sigmf_pytorch_dataset import SigMFDataset\n",
    "# from sklearn.metrics import classification_report\n",
    "# from torchsig.models.iq_models.efficientnet.efficientnet import efficientnet_b4\n",
    "# from torchsig.transforms import (\n",
    "#     Compose,\n",
    "#     IQImbalance,\n",
    "#     Normalize,\n",
    "#     RandomApply,\n",
    "#     RandomFrequencyShift,\n",
    "#     RandomPhaseShift,\n",
    "#     RandomResample,\n",
    "#     RandomTimeShift,\n",
    "#     RayleighFadingChannel,\n",
    "#     TargetSNR,\n",
    "#     ComplexTo2D,\n",
    "# )\n",
    "# from torchsig.utils.cm_plotter import plot_confusion_matrix\n",
    "# from torchsig.utils.dataset import SignalDataset\n",
    "# from torchsig.utils.visualize import IQVisualizer, SpectrogramVisualizer, two_channel_to_complex\n",
    "# from tqdm import tqdm\n",
    "# from typing import List\n",
    "\n",
    "\n",
    "# import data as data_class\n",
    "# import annotation_utils"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "072ab439-5251-46fb-a399-5fad2369ee91",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ltindall/.pyenv/versions/3.10.9/envs/rfml-dev/lib/python3.10/site-packages/cupyx/jit/_interface.py:173: FutureWarning: cupyx.jit.rawkernel is experimental. The interface can change in the future.\n",
      "  cupy._util.experimental('cupyx.jit.rawkernel')\n"
     ]
    }
   ],
   "source": [
    "\n",
    "\n",
    "import glob\n",
    "\n",
    "from pathlib import Path\n",
    "from tqdm import tqdm\n",
    "\n",
    "import data as data_class\n",
    "import annotation_utils"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "17bd1f76-0fae-499d-a87c-5bc5eab8670f",
   "metadata": {},
   "source": [
    "# Parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "84e058ab-4ac9-40a5-8b1f-2713a5c1f1e9",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_directory = \"data/gamutrf/gamutrf-sd-gr-ieee-wifi/test_offline/\"\n",
    "# model_file = \"weights/checkpoint-v17_torchscript.pt\"\n",
    "# num_iq_samples = 1024\n",
    "# only_use_start_of_burst = True\n",
    "# idx_to_class = {\n",
    "#   0: 'anom_wifi',\n",
    "#   1: 'mini2_telem',\n",
    "#   2: 'mini2_video',\n",
    "#   3: 'wifi6',\n",
    "#   4: 'wifi',\n",
    "#   5: 'fhss_css'\n",
    "# }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "b3521e0a-848e-4c54-b644-1b6a62127f49",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|                                                                                       | 0/1 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Loaded \n",
      " Data file: data/gamutrf/gamutrf-sd-gr-ieee-wifi/test_offline/anomwifi-encoding-0-sendsps-10000000.s16.zst.sigmf-data \n",
      " SigMF-Meta file: data/gamutrf/gamutrf-sd-gr-ieee-wifi/test_offline/anomwifi-encoding-0-sendsps-10000000.s16.zst.sigmf-meta\n",
      "\n",
      "np.max(avg_pwr_db)=-52.91393975212567\n",
      "np.mean(avg_pwr_db)=-63.217670300631006\n",
      "median absolute deviation threshold = -61.75284272288487\n",
      "using threshold = -61.75284272288487\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "  0%|                                                                                    | 0/1119 [00:00<?, ?it/s]\u001b[A\n",
      "  1%|▍                                                                           | 6/1119 [00:00<00:19, 56.23it/s]\u001b[A\n",
      "  1%|▊                                                                          | 13/1119 [00:00<00:17, 62.26it/s]\u001b[A\n",
      "  2%|█▎                                                                         | 20/1119 [00:00<00:17, 61.33it/s]\u001b[A\n",
      "  3%|██                                                                         | 30/1119 [00:00<00:14, 73.35it/s]\u001b[A\n",
      "  3%|██▌                                                                        | 38/1119 [00:00<00:15, 69.05it/s]\u001b[A\n",
      "  4%|███                                                                        | 45/1119 [00:00<00:16, 65.43it/s]\u001b[A\n",
      "  5%|███▊                                                                       | 57/1119 [00:00<00:13, 78.90it/s]\u001b[A\n",
      "  6%|████▎                                                                      | 65/1119 [00:00<00:14, 72.76it/s]\u001b[A\n",
      "  7%|████▉                                                                      | 73/1119 [00:01<00:14, 69.92it/s]\u001b[A\n",
      "  7%|█████▍                                                                     | 81/1119 [00:01<00:16, 64.85it/s]\u001b[A\n",
      "  8%|█████▉                                                                     | 88/1119 [00:01<00:16, 63.82it/s]\u001b[A\n",
      "  8%|██████▎                                                                    | 95/1119 [00:01<00:17, 60.08it/s]\u001b[A\n",
      "  9%|██████▉                                                                   | 105/1119 [00:01<00:14, 68.83it/s]\u001b[A\n",
      " 10%|███████▍                                                                  | 113/1119 [00:01<00:14, 69.33it/s]\u001b[A\n",
      " 11%|████████▏                                                                 | 123/1119 [00:01<00:13, 75.84it/s]\u001b[A\n",
      " 12%|████████▋                                                                 | 132/1119 [00:01<00:12, 78.43it/s]\u001b[A\n",
      " 13%|█████████▎                                                                | 140/1119 [00:01<00:12, 77.43it/s]\u001b[A\n",
      " 13%|█████████▊                                                                | 148/1119 [00:02<00:13, 72.86it/s]\u001b[A\n",
      " 14%|██████████▌                                                               | 160/1119 [00:02<00:11, 82.26it/s]\u001b[A\n",
      " 15%|███████████▏                                                              | 169/1119 [00:02<00:11, 79.59it/s]\u001b[A\n",
      " 16%|███████████▊                                                              | 178/1119 [00:02<00:12, 74.16it/s]\u001b[A\n",
      " 17%|████████████▎                                                             | 186/1119 [00:02<00:13, 69.96it/s]\u001b[A\n",
      " 17%|████████████▉                                                             | 195/1119 [00:02<00:12, 73.12it/s]\u001b[A\n",
      " 18%|█████████████▍                                                            | 203/1119 [00:02<00:13, 67.93it/s]\u001b[A\n",
      " 19%|█████████████▉                                                            | 211/1119 [00:02<00:13, 68.84it/s]\u001b[A\n",
      " 19%|██████████████▍                                                           | 218/1119 [00:03<00:13, 65.87it/s]\u001b[A\n",
      " 20%|██████████████▉                                                           | 226/1119 [00:03<00:13, 67.56it/s]\u001b[A\n",
      " 21%|███████████████▍                                                          | 233/1119 [00:03<00:13, 66.45it/s]\u001b[A\n",
      " 21%|███████████████▊                                                          | 240/1119 [00:03<00:14, 62.70it/s]\u001b[A\n",
      " 22%|████████████████▎                                                         | 247/1119 [00:03<00:13, 63.30it/s]\u001b[A\n",
      " 23%|████████████████▊                                                         | 254/1119 [00:03<00:14, 59.71it/s]\u001b[A\n",
      " 24%|█████████████████▍                                                        | 264/1119 [00:03<00:12, 68.37it/s]\u001b[A\n",
      " 24%|█████████████████▉                                                        | 271/1119 [00:03<00:12, 66.44it/s]\u001b[A\n",
      " 25%|██████████████████▌                                                       | 281/1119 [00:04<00:11, 73.87it/s]\u001b[A\n",
      " 26%|███████████████████▏                                                      | 290/1119 [00:04<00:10, 77.05it/s]\u001b[A\n",
      " 27%|███████████████████▊                                                      | 299/1119 [00:04<00:10, 79.50it/s]\u001b[A\n",
      " 28%|████████████████████▎                                                     | 308/1119 [00:04<00:10, 80.29it/s]\u001b[A\n",
      " 28%|████████████████████▉                                                     | 317/1119 [00:04<00:10, 75.32it/s]\u001b[A\n",
      " 29%|█████████████████████▍                                                    | 325/1119 [00:04<00:10, 74.90it/s]\u001b[A\n",
      " 30%|██████████████████████                                                    | 333/1119 [00:04<00:10, 72.22it/s]\u001b[A\n",
      " 30%|██████████████████████▌                                                   | 341/1119 [00:04<00:10, 72.80it/s]\u001b[A\n",
      " 31%|███████████████████████▎                                                  | 352/1119 [00:04<00:09, 80.07it/s]\u001b[A\n",
      " 32%|███████████████████████▊                                                  | 361/1119 [00:05<00:09, 76.77it/s]\u001b[A\n",
      " 33%|████████████████████████▍                                                 | 370/1119 [00:05<00:09, 77.32it/s]\u001b[A\n",
      " 34%|█████████████████████████▎                                                | 383/1119 [00:05<00:08, 88.59it/s]\u001b[A\n",
      " 35%|█████████████████████████▉                                                | 392/1119 [00:05<00:08, 88.41it/s]\u001b[A\n",
      " 36%|██████████████████████████▌                                               | 401/1119 [00:05<00:08, 82.14it/s]\u001b[A\n",
      " 37%|███████████████████████████                                               | 410/1119 [00:05<00:09, 75.97it/s]\u001b[A\n",
      " 37%|███████████████████████████▋                                              | 418/1119 [00:05<00:09, 73.92it/s]\u001b[A\n",
      " 38%|████████████████████████████▏                                             | 426/1119 [00:05<00:09, 73.87it/s]\u001b[A\n",
      " 39%|████████████████████████████▋                                             | 434/1119 [00:06<00:09, 70.25it/s]\u001b[A\n",
      " 39%|█████████████████████████████▏                                            | 442/1119 [00:06<00:09, 68.03it/s]\u001b[A\n",
      " 40%|█████████████████████████████▋                                            | 449/1119 [00:06<00:10, 65.76it/s]\u001b[A\n",
      " 41%|██████████████████████████████▍                                           | 460/1119 [00:06<00:08, 74.96it/s]\u001b[A\n",
      " 42%|██████████████████████████████▉                                           | 468/1119 [00:06<00:09, 70.45it/s]\u001b[A\n",
      " 43%|███████████████████████████████▌                                          | 477/1119 [00:06<00:08, 73.08it/s]\u001b[A\n",
      " 44%|████████████████████████████████▏                                         | 487/1119 [00:06<00:08, 77.67it/s]\u001b[A\n",
      " 44%|████████████████████████████████▋                                         | 495/1119 [00:06<00:08, 76.92it/s]\u001b[A\n",
      " 45%|█████████████████████████████████▎                                        | 503/1119 [00:06<00:08, 68.71it/s]\u001b[A\n",
      " 46%|█████████████████████████████████▊                                        | 512/1119 [00:07<00:08, 71.40it/s]\u001b[A\n",
      " 46%|██████████████████████████████████▍                                       | 520/1119 [00:07<00:08, 70.69it/s]\u001b[A\n",
      " 47%|██████████████████████████████████▉                                       | 528/1119 [00:07<00:08, 70.93it/s]\u001b[A\n",
      " 48%|███████████████████████████████████▍                                      | 536/1119 [00:07<00:08, 68.55it/s]\u001b[A\n",
      " 49%|███████████████████████████████████▉                                      | 543/1119 [00:07<00:09, 62.48it/s]\u001b[A\n",
      " 49%|████████████████████████████████████▍                                     | 551/1119 [00:07<00:08, 64.66it/s]\u001b[A\n",
      " 50%|████████████████████████████████████▉                                     | 558/1119 [00:07<00:08, 64.01it/s]\u001b[A\n",
      " 51%|█████████████████████████████████████▍                                    | 566/1119 [00:07<00:08, 66.24it/s]\u001b[A\n",
      " 51%|██████████████████████████████████████                                    | 575/1119 [00:08<00:07, 69.70it/s]\u001b[A\n",
      " 52%|██████████████████████████████████████▌                                   | 583/1119 [00:08<00:08, 63.95it/s]\u001b[A\n",
      " 53%|███████████████████████████████████████                                   | 590/1119 [00:08<00:09, 58.55it/s]\u001b[A\n",
      " 54%|███████████████████████████████████████▋                                  | 600/1119 [00:08<00:07, 67.00it/s]\u001b[A\n",
      " 54%|████████████████████████████████████████▏                                 | 607/1119 [00:08<00:07, 67.44it/s]\u001b[A\n",
      " 55%|████████████████████████████████████████▌                                 | 614/1119 [00:08<00:07, 66.66it/s]\u001b[A\n",
      " 56%|█████████████████████████████████████████▏                                | 623/1119 [00:08<00:06, 72.74it/s]\u001b[A\n",
      " 56%|█████████████████████████████████████████▋                                | 631/1119 [00:08<00:07, 66.53it/s]\u001b[A\n",
      " 57%|██████████████████████████████████████████▏                               | 638/1119 [00:09<00:07, 62.67it/s]\u001b[A\n",
      " 58%|██████████████████████████████████████████▋                               | 645/1119 [00:09<00:07, 60.20it/s]\u001b[A\n",
      " 58%|███████████████████████████████████████████▏                              | 654/1119 [00:09<00:07, 66.14it/s]\u001b[A\n",
      " 59%|███████████████████████████████████████████▊                              | 662/1119 [00:09<00:06, 68.18it/s]\u001b[A\n",
      " 60%|████████████████████████████████████████████▍                             | 672/1119 [00:09<00:06, 74.07it/s]\u001b[A\n",
      " 61%|████████████████████████████████████████████▉                             | 680/1119 [00:09<00:06, 71.77it/s]\u001b[A\n",
      " 61%|█████████████████████████████████████████████▍                            | 688/1119 [00:09<00:05, 73.16it/s]\u001b[A\n",
      " 62%|██████████████████████████████████████████████                            | 696/1119 [00:09<00:05, 72.63it/s]\u001b[A\n",
      " 63%|██████████████████████████████████████████████▌                           | 704/1119 [00:09<00:06, 68.42it/s]\u001b[A\n",
      " 64%|███████████████████████████████████████████████                           | 712/1119 [00:10<00:05, 68.69it/s]\u001b[A\n",
      " 64%|███████████████████████████████████████████████▌                          | 720/1119 [00:10<00:05, 69.55it/s]\u001b[A\n",
      " 65%|████████████████████████████████████████████████▏                         | 729/1119 [00:10<00:05, 73.95it/s]\u001b[A\n",
      " 66%|████████████████████████████████████████████████▋                         | 737/1119 [00:10<00:05, 69.47it/s]\u001b[A\n",
      " 67%|█████████████████████████████████████████████████▎                        | 745/1119 [00:10<00:05, 71.39it/s]\u001b[A\n",
      " 67%|█████████████████████████████████████████████████▊                        | 754/1119 [00:10<00:04, 74.56it/s]\u001b[A\n",
      " 68%|██████████████████████████████████████████████████▍                       | 762/1119 [00:10<00:04, 72.98it/s]\u001b[A\n",
      " 69%|██████████████████████████████████████████████████▉                       | 770/1119 [00:10<00:05, 67.12it/s]\u001b[A\n",
      " 70%|███████████████████████████████████████████████████▋                      | 781/1119 [00:11<00:04, 76.32it/s]\u001b[A\n",
      " 71%|████████████████████████████████████████████████████▏                     | 789/1119 [00:11<00:04, 73.81it/s]\u001b[A\n",
      " 71%|████████████████████████████████████████████████████▋                     | 797/1119 [00:11<00:04, 65.66it/s]\u001b[A\n",
      " 72%|█████████████████████████████████████████████████████▏                    | 804/1119 [00:11<00:04, 64.26it/s]\u001b[A\n",
      " 72%|█████████████████████████████████████████████████████▋                    | 811/1119 [00:11<00:04, 63.96it/s]\u001b[A\n",
      " 73%|██████████████████████████████████████████████████████▏                   | 819/1119 [00:11<00:04, 66.18it/s]\u001b[A\n",
      " 74%|██████████████████████████████████████████████████████▌                   | 826/1119 [00:11<00:04, 64.86it/s]\u001b[A\n",
      " 74%|███████████████████████████████████████████████████████                   | 833/1119 [00:11<00:04, 65.90it/s]\u001b[A\n",
      " 75%|███████████████████████████████████████████████████████▌                  | 840/1119 [00:11<00:04, 62.92it/s]\u001b[A\n",
      " 76%|████████████████████████████████████████████████████████                  | 847/1119 [00:12<00:04, 59.76it/s]\u001b[A\n",
      " 76%|████████████████████████████████████████████████████████▌                 | 855/1119 [00:12<00:04, 64.38it/s]\u001b[A\n",
      " 77%|█████████████████████████████████████████████████████████                 | 863/1119 [00:12<00:03, 68.52it/s]\u001b[A\n",
      " 78%|█████████████████████████████████████████████████████████▋                | 873/1119 [00:12<00:03, 74.26it/s]\u001b[A\n",
      " 79%|██████████████████████████████████████████████████████████▎               | 881/1119 [00:12<00:03, 72.56it/s]\u001b[A\n",
      " 80%|██████████████████████████████████████████████████████████▊               | 890/1119 [00:12<00:03, 75.04it/s]\u001b[A\n",
      " 80%|███████████████████████████████████████████████████████████▍              | 898/1119 [00:12<00:03, 70.96it/s]\u001b[A\n",
      " 81%|███████████████████████████████████████████████████████████▉              | 906/1119 [00:12<00:03, 63.73it/s]\u001b[A\n",
      " 82%|████████████████████████████████████████████████████████████▌             | 915/1119 [00:13<00:02, 68.86it/s]\u001b[A\n",
      " 82%|█████████████████████████████████████████████████████████████             | 923/1119 [00:13<00:03, 62.87it/s]\u001b[A\n",
      " 83%|█████████████████████████████████████████████████████████████▌            | 930/1119 [00:13<00:03, 62.94it/s]\u001b[A\n",
      " 84%|██████████████████████████████████████████████████████████████▏           | 940/1119 [00:13<00:02, 71.60it/s]\u001b[A\n",
      " 85%|██████████████████████████████████████████████████████████████▋           | 948/1119 [00:13<00:02, 69.00it/s]\u001b[A\n",
      " 85%|███████████████████████████████████████████████████████████████▏          | 956/1119 [00:13<00:02, 62.92it/s]\u001b[A\n",
      " 86%|███████████████████████████████████████████████████████████████▋          | 963/1119 [00:13<00:02, 64.63it/s]\u001b[A\n",
      " 87%|████████████████████████████████████████████████████████████████▏         | 970/1119 [00:13<00:02, 62.56it/s]\u001b[A\n",
      " 87%|████████████████████████████████████████████████████████████████▌         | 977/1119 [00:14<00:02, 61.23it/s]\u001b[A\n",
      " 88%|█████████████████████████████████████████████████████████████████         | 984/1119 [00:14<00:02, 60.54it/s]\u001b[A\n",
      " 89%|█████████████████████████████████████████████████████████████████▌        | 992/1119 [00:14<00:01, 65.20it/s]\u001b[A\n",
      " 89%|██████████████████████████████████████████████████████████████████        | 999/1119 [00:14<00:01, 62.36it/s]\u001b[A\n",
      " 90%|█████████████████████████████████████████████████████████████████▋       | 1006/1119 [00:14<00:01, 62.72it/s]\u001b[A\n",
      " 91%|██████████████████████████████████████████████████████████████████       | 1013/1119 [00:14<00:01, 62.19it/s]\u001b[A\n",
      " 91%|██████████████████████████████████████████████████████████████████▌      | 1020/1119 [00:14<00:01, 59.12it/s]\u001b[A\n",
      " 92%|██████████████████████████████████████████████████████████████████▉      | 1026/1119 [00:14<00:01, 57.02it/s]\u001b[A\n",
      " 92%|███████████████████████████████████████████████████████████████████▌     | 1035/1119 [00:14<00:01, 64.25it/s]\u001b[A\n",
      " 93%|███████████████████████████████████████████████████████████████████▉     | 1042/1119 [00:15<00:01, 65.70it/s]\u001b[A\n",
      " 94%|████████████████████████████████████████████████████████████████████▍    | 1049/1119 [00:15<00:01, 66.63it/s]\u001b[A\n",
      " 94%|████████████████████████████████████████████████████████████████████▉    | 1057/1119 [00:15<00:00, 68.98it/s]\u001b[A\n",
      " 95%|█████████████████████████████████████████████████████████████████████▍   | 1064/1119 [00:15<00:00, 66.98it/s]\u001b[A\n",
      " 96%|█████████████████████████████████████████████████████████████████████▊   | 1071/1119 [00:15<00:00, 64.93it/s]\u001b[A\n",
      " 96%|██████████████████████████████████████████████████████████████████████▎  | 1078/1119 [00:15<00:00, 64.60it/s]\u001b[A\n",
      " 97%|██████████████████████████████████████████████████████████████████████▉  | 1088/1119 [00:15<00:00, 73.03it/s]\u001b[A\n",
      " 98%|███████████████████████████████████████████████████████████████████████▋ | 1098/1119 [00:15<00:00, 78.68it/s]\u001b[A\n",
      " 99%|████████████████████████████████████████████████████████████████████████▏| 1107/1119 [00:15<00:00, 80.31it/s]\u001b[A\n",
      "100%|█████████████████████████████████████████████████████████████████████████| 1119/1119 [00:16<00:00, 69.61it/s]\u001b[A\n",
      "100%|███████████████████████████████████████████████████████████████████████████████| 1/1 [00:43<00:00, 43.24s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Writing 904 annotations to data/gamutrf/gamutrf-sd-gr-ieee-wifi/test_offline/anomwifi-encoding-0-sendsps-10000000.s16.zst.sigmf-meta\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|                                                                                       | 0/1 [00:00<?, ?it/s]/home/ltindall/.pyenv/versions/3.10.9/envs/rfml-dev/lib/python3.10/site-packages/sigmf/sigmffile.py:914: UserWarning: Compliant Dataset `data/gamutrf/gamutrf-sd-gr-ieee-wifi/test_offline/wifi-encoding-0-sendsps-10000000.s16.zst.sigmf-data` exists but \"core:dataset\" is also defined; using `data/gamutrf/gamutrf-sd-gr-ieee-wifi/test_offline/wifi-encoding-0-sendsps-10000000.s16.zst.sigmf-data`\n",
      "  warnings.warn(f'Compliant Dataset `{compliant_data_fn}` exists but '\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Loaded \n",
      " Data file: data/gamutrf/gamutrf-sd-gr-ieee-wifi/test_offline/wifi-encoding-0-sendsps-10000000.s16.zst.sigmf-data \n",
      " SigMF-Meta file: data/gamutrf/gamutrf-sd-gr-ieee-wifi/test_offline/wifi-encoding-0-sendsps-10000000.s16.zst.sigmf-meta\n",
      "\n",
      "np.max(avg_pwr_db)=-52.644283524819464\n",
      "np.mean(avg_pwr_db)=-62.98255600272792\n",
      "median absolute deviation threshold = -61.525868729847154\n",
      "using threshold = -61.525868729847154\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "  0%|                                                                                    | 0/1170 [00:00<?, ?it/s]\u001b[A\n",
      "  1%|▌                                                                           | 9/1170 [00:00<00:14, 78.88it/s]\u001b[A\n",
      "  2%|█▎                                                                         | 20/1170 [00:00<00:12, 93.39it/s]\u001b[A\n",
      "  3%|█▉                                                                         | 30/1170 [00:00<00:13, 82.50it/s]\u001b[A\n",
      "  3%|██▌                                                                        | 39/1170 [00:00<00:14, 76.56it/s]\u001b[A\n",
      "  4%|███▎                                                                       | 51/1170 [00:00<00:12, 86.61it/s]\u001b[A\n",
      "  5%|███▊                                                                       | 60/1170 [00:00<00:13, 84.31it/s]\u001b[A\n",
      "  6%|████▍                                                                      | 69/1170 [00:00<00:13, 82.92it/s]\u001b[A\n",
      "  7%|█████                                                                      | 79/1170 [00:00<00:12, 85.95it/s]\u001b[A\n",
      "  8%|█████▋                                                                     | 88/1170 [00:01<00:13, 80.16it/s]\u001b[A\n",
      "  8%|██████▏                                                                    | 97/1170 [00:01<00:15, 68.74it/s]\u001b[A\n",
      "  9%|██████▋                                                                   | 105/1170 [00:01<00:14, 71.48it/s]\u001b[A\n",
      " 10%|███████▏                                                                  | 113/1170 [00:01<00:14, 73.35it/s]\u001b[A\n",
      " 10%|███████▋                                                                  | 121/1170 [00:01<00:14, 73.65it/s]\u001b[A\n",
      " 11%|████████▎                                                                 | 131/1170 [00:01<00:13, 78.99it/s]\u001b[A\n",
      " 12%|████████▊                                                                 | 140/1170 [00:01<00:14, 72.86it/s]\u001b[A\n",
      " 13%|█████████▍                                                                | 149/1170 [00:01<00:13, 76.95it/s]\u001b[A\n",
      " 13%|█████████▉                                                                | 157/1170 [00:02<00:13, 75.24it/s]\u001b[A\n",
      " 14%|██████████▍                                                               | 165/1170 [00:02<00:14, 71.03it/s]\u001b[A\n",
      " 15%|███████████                                                               | 174/1170 [00:02<00:13, 73.53it/s]\u001b[A\n",
      " 16%|███████████▌                                                              | 183/1170 [00:02<00:13, 75.12it/s]\u001b[A\n",
      " 16%|████████████▏                                                             | 193/1170 [00:02<00:12, 81.19it/s]\u001b[A\n",
      " 17%|████████████▊                                                             | 202/1170 [00:02<00:11, 81.75it/s]\u001b[A\n",
      " 18%|█████████████▎                                                            | 211/1170 [00:02<00:13, 73.24it/s]\u001b[A\n",
      " 19%|█████████████▊                                                            | 219/1170 [00:02<00:14, 66.68it/s]\u001b[A\n",
      " 19%|██████████████▍                                                           | 228/1170 [00:02<00:13, 72.17it/s]\u001b[A\n",
      " 20%|██████████████▉                                                           | 236/1170 [00:03<00:13, 70.08it/s]\u001b[A\n",
      " 21%|███████████████▍                                                          | 244/1170 [00:03<00:13, 67.61it/s]\u001b[A\n",
      " 22%|████████████████                                                          | 253/1170 [00:03<00:12, 71.31it/s]\u001b[A\n",
      " 22%|████████████████▌                                                         | 261/1170 [00:03<00:12, 69.99it/s]\u001b[A\n",
      " 23%|█████████████████                                                         | 270/1170 [00:03<00:12, 72.80it/s]\u001b[A\n",
      " 24%|█████████████████▊                                                        | 281/1170 [00:03<00:11, 80.54it/s]\u001b[A\n",
      " 25%|██████████████████▎                                                       | 290/1170 [00:03<00:11, 79.43it/s]\u001b[A\n",
      " 26%|██████████████████▉                                                       | 299/1170 [00:03<00:11, 72.66it/s]\u001b[A\n",
      " 26%|███████████████████▍                                                      | 308/1170 [00:04<00:11, 75.05it/s]\u001b[A\n",
      " 27%|███████████████████▉                                                      | 316/1170 [00:04<00:11, 75.69it/s]\u001b[A\n",
      " 28%|████████████████████▌                                                     | 325/1170 [00:04<00:11, 75.30it/s]\u001b[A\n",
      " 28%|█████████████████████                                                     | 333/1170 [00:04<00:11, 71.15it/s]\u001b[A\n",
      " 29%|█████████████████████▌                                                    | 341/1170 [00:04<00:12, 68.57it/s]\u001b[A\n",
      " 30%|██████████████████████▍                                                   | 354/1170 [00:04<00:09, 82.32it/s]\u001b[A\n",
      " 31%|██████████████████████▉                                                   | 363/1170 [00:04<00:10, 79.35it/s]\u001b[A\n",
      " 32%|███████████████████████▌                                                  | 372/1170 [00:04<00:10, 77.00it/s]\u001b[A\n",
      " 33%|████████████████████████▎                                                 | 384/1170 [00:05<00:09, 85.99it/s]\u001b[A\n",
      " 34%|████████████████████████▊                                                 | 393/1170 [00:05<00:09, 80.46it/s]\u001b[A\n",
      " 34%|█████████████████████████▍                                                | 402/1170 [00:05<00:09, 77.18it/s]\u001b[A\n",
      " 35%|█████████████████████████▉                                                | 410/1170 [00:05<00:10, 75.80it/s]\u001b[A\n",
      " 36%|██████████████████████████▍                                               | 418/1170 [00:05<00:10, 74.89it/s]\u001b[A\n",
      " 36%|██████████████████████████▉                                               | 426/1170 [00:05<00:10, 74.11it/s]\u001b[A\n",
      " 37%|███████████████████████████▌                                              | 436/1170 [00:05<00:09, 80.61it/s]\u001b[A\n",
      " 38%|████████████████████████████▏                                             | 445/1170 [00:05<00:09, 76.91it/s]\u001b[A\n",
      " 39%|████████████████████████████▋                                             | 453/1170 [00:05<00:09, 74.31it/s]\u001b[A\n",
      " 40%|█████████████████████████████▎                                            | 463/1170 [00:06<00:09, 78.48it/s]\u001b[A\n",
      " 40%|█████████████████████████████▊                                            | 471/1170 [00:06<00:09, 77.46it/s]\u001b[A\n",
      " 41%|██████████████████████████████▎                                           | 479/1170 [00:06<00:09, 74.70it/s]\u001b[A\n",
      " 42%|██████████████████████████████▊                                           | 487/1170 [00:06<00:09, 70.24it/s]\u001b[A\n",
      " 42%|███████████████████████████████▎                                          | 495/1170 [00:06<00:09, 68.10it/s]\u001b[A\n",
      " 43%|███████████████████████████████▉                                          | 505/1170 [00:06<00:08, 76.37it/s]\u001b[A\n",
      " 44%|████████████████████████████████▌                                         | 515/1170 [00:06<00:08, 79.54it/s]\u001b[A\n",
      " 45%|█████████████████████████████████▏                                        | 524/1170 [00:06<00:08, 80.08it/s]\u001b[A\n",
      " 46%|█████████████████████████████████▊                                        | 534/1170 [00:06<00:07, 83.31it/s]\u001b[A\n",
      " 46%|██████████████████████████████████▎                                       | 543/1170 [00:07<00:07, 79.56it/s]\u001b[A\n",
      " 47%|██████████████████████████████████▉                                       | 552/1170 [00:07<00:08, 73.51it/s]\u001b[A\n",
      " 48%|███████████████████████████████████▌                                      | 562/1170 [00:07<00:07, 77.46it/s]\u001b[A\n",
      " 49%|████████████████████████████████████                                      | 571/1170 [00:07<00:07, 79.51it/s]\u001b[A\n",
      " 50%|████████████████████████████████████▋                                     | 580/1170 [00:07<00:07, 77.08it/s]\u001b[A\n",
      " 50%|█████████████████████████████████████▎                                    | 590/1170 [00:07<00:07, 79.26it/s]\u001b[A\n",
      " 51%|█████████████████████████████████████▉                                    | 600/1170 [00:07<00:06, 81.87it/s]\u001b[A\n",
      " 52%|██████████████████████████████████████▌                                   | 609/1170 [00:07<00:07, 73.90it/s]\u001b[A\n",
      " 53%|███████████████████████████████████████                                   | 617/1170 [00:08<00:07, 71.97it/s]\u001b[A\n",
      " 54%|███████████████████████████████████████▌                                  | 626/1170 [00:08<00:07, 73.98it/s]\u001b[A\n",
      " 54%|████████████████████████████████████████▏                                 | 635/1170 [00:08<00:06, 76.52it/s]\u001b[A\n",
      " 55%|████████████████████████████████████████▋                                 | 644/1170 [00:08<00:06, 78.13it/s]\u001b[A\n",
      " 56%|█████████████████████████████████████████▏                                | 652/1170 [00:08<00:07, 71.25it/s]\u001b[A\n",
      " 56%|█████████████████████████████████████████▋                                | 660/1170 [00:08<00:07, 68.91it/s]\u001b[A\n",
      " 57%|██████████████████████████████████████████▏                               | 667/1170 [00:08<00:07, 67.16it/s]\u001b[A\n",
      " 58%|██████████████████████████████████████████▋                               | 675/1170 [00:08<00:07, 69.19it/s]\u001b[A\n",
      " 58%|███████████████████████████████████████████▎                              | 684/1170 [00:09<00:06, 72.61it/s]\u001b[A\n",
      " 59%|███████████████████████████████████████████▊                              | 692/1170 [00:09<00:06, 69.67it/s]\u001b[A\n",
      " 60%|████████████████████████████████████████████▎                             | 700/1170 [00:09<00:06, 68.05it/s]\u001b[A\n",
      " 60%|████████████████████████████████████████████▋                             | 707/1170 [00:09<00:07, 64.86it/s]\u001b[A\n",
      " 61%|█████████████████████████████████████████████▏                            | 715/1170 [00:09<00:06, 67.60it/s]\u001b[A\n",
      " 62%|█████████████████████████████████████████████▊                            | 724/1170 [00:09<00:06, 70.85it/s]\u001b[A\n",
      " 63%|██████████████████████████████████████████████▌                           | 736/1170 [00:09<00:05, 82.69it/s]\u001b[A\n",
      " 64%|███████████████████████████████████████████████                           | 745/1170 [00:09<00:05, 77.68it/s]\u001b[A\n",
      " 64%|███████████████████████████████████████████████▋                          | 753/1170 [00:09<00:05, 72.82it/s]\u001b[A\n",
      " 65%|████████████████████████████████████████████████▏                         | 761/1170 [00:10<00:05, 70.42it/s]\u001b[A\n",
      " 66%|████████████████████████████████████████████████▋                         | 769/1170 [00:10<00:05, 69.16it/s]\u001b[A\n",
      " 67%|█████████████████████████████████████████████████▎                        | 780/1170 [00:10<00:04, 78.45it/s]\u001b[A\n",
      " 67%|█████████████████████████████████████████████████▊                        | 788/1170 [00:10<00:05, 74.19it/s]\u001b[A\n",
      " 68%|██████████████████████████████████████████████████▍                       | 798/1170 [00:10<00:04, 79.21it/s]\u001b[A\n",
      " 69%|███████████████████████████████████████████████████                       | 807/1170 [00:10<00:04, 79.64it/s]\u001b[A\n",
      " 70%|███████████████████████████████████████████████████▋                      | 817/1170 [00:10<00:04, 82.40it/s]\u001b[A\n",
      " 71%|████████████████████████████████████████████████████▏                     | 826/1170 [00:10<00:04, 76.70it/s]\u001b[A\n",
      " 71%|████████████████████████████████████████████████████▉                     | 836/1170 [00:11<00:04, 81.15it/s]\u001b[A\n",
      " 72%|█████████████████████████████████████████████████████▍                    | 845/1170 [00:11<00:04, 75.18it/s]\u001b[A\n",
      " 73%|█████████████████████████████████████████████████████▉                    | 853/1170 [00:11<00:04, 75.25it/s]\u001b[A\n",
      " 74%|██████████████████████████████████████████████████████▍                   | 861/1170 [00:11<00:04, 73.17it/s]\u001b[A\n",
      " 74%|██████████████████████████████████████████████████████▉                   | 869/1170 [00:11<00:04, 70.36it/s]\u001b[A\n",
      " 75%|███████████████████████████████████████████████████████▌                  | 878/1170 [00:11<00:03, 73.38it/s]\u001b[A\n",
      " 76%|████████████████████████████████████████████████████████                  | 886/1170 [00:11<00:03, 72.10it/s]\u001b[A\n",
      " 76%|████████████████████████████████████████████████████████▌                 | 894/1170 [00:11<00:04, 66.35it/s]\u001b[A\n",
      " 77%|████████████████████████████████████████████████████████▉                 | 901/1170 [00:12<00:04, 63.75it/s]\u001b[A\n",
      " 78%|█████████████████████████████████████████████████████████▍                | 909/1170 [00:12<00:03, 65.46it/s]\u001b[A\n",
      " 79%|██████████████████████████████████████████████████████████                | 919/1170 [00:12<00:03, 72.21it/s]\u001b[A\n",
      " 79%|██████████████████████████████████████████████████████████▋               | 927/1170 [00:12<00:03, 68.18it/s]\u001b[A\n",
      " 80%|███████████████████████████████████████████████████████████▏              | 936/1170 [00:12<00:03, 71.29it/s]\u001b[A\n",
      " 81%|███████████████████████████████████████████████████████████▋              | 944/1170 [00:12<00:03, 67.54it/s]\u001b[A\n",
      " 81%|████████████████████████████████████████████████████████████▏             | 952/1170 [00:12<00:03, 69.32it/s]\u001b[A\n",
      " 82%|████████████████████████████████████████████████████████████▋             | 960/1170 [00:12<00:03, 66.06it/s]\u001b[A\n",
      " 83%|█████████████████████████████████████████████████████████████▏            | 968/1170 [00:12<00:03, 66.53it/s]\u001b[A\n",
      " 83%|█████████████████████████████████████████████████████████████▋            | 975/1170 [00:13<00:02, 66.23it/s]\u001b[A\n",
      " 84%|██████████████████████████████████████████████████████████████▏           | 983/1170 [00:13<00:02, 67.56it/s]\u001b[A\n",
      " 85%|██████████████████████████████████████████████████████████████▌           | 990/1170 [00:13<00:02, 66.32it/s]\u001b[A\n",
      " 85%|███████████████████████████████████████████████████████████████           | 997/1170 [00:13<00:02, 66.58it/s]\u001b[A\n",
      " 86%|██████████████████████████████████████████████████████████████▉          | 1008/1170 [00:13<00:02, 76.58it/s]\u001b[A\n",
      " 87%|███████████████████████████████████████████████████████████████▌         | 1018/1170 [00:13<00:01, 80.23it/s]\u001b[A\n",
      " 88%|████████████████████████████████████████████████████████████████         | 1027/1170 [00:13<00:01, 73.88it/s]\u001b[A\n",
      " 89%|████████████████████████████████████████████████████████████████▋        | 1036/1170 [00:13<00:01, 75.81it/s]\u001b[A\n",
      " 89%|█████████████████████████████████████████████████████████████████▏       | 1044/1170 [00:14<00:01, 67.79it/s]\u001b[A\n",
      " 90%|█████████████████████████████████████████████████████████████████▋       | 1052/1170 [00:14<00:01, 69.34it/s]\u001b[A\n",
      " 91%|██████████████████████████████████████████████████████████████████▏      | 1060/1170 [00:14<00:01, 64.27it/s]\u001b[A\n",
      " 91%|██████████████████████████████████████████████████████████████████▌      | 1067/1170 [00:14<00:01, 65.43it/s]\u001b[A\n",
      " 92%|███████████████████████████████████████████████████████████████████      | 1074/1170 [00:14<00:01, 63.42it/s]\u001b[A\n",
      " 93%|███████████████████████████████████████████████████████████████████▋     | 1085/1170 [00:14<00:01, 74.10it/s]\u001b[A\n",
      " 94%|████████████████████████████████████████████████████████████████████▎    | 1095/1170 [00:14<00:00, 80.55it/s]\u001b[A\n",
      " 94%|████████████████████████████████████████████████████████████████████▉    | 1104/1170 [00:14<00:00, 76.36it/s]\u001b[A\n",
      " 95%|█████████████████████████████████████████████████████████████████████▍   | 1112/1170 [00:14<00:00, 75.32it/s]\u001b[A\n",
      " 96%|█████████████████████████████████████████████████████████████████████▉   | 1120/1170 [00:15<00:00, 71.71it/s]\u001b[A\n",
      " 96%|██████████████████████████████████████████████████████████████████████▍  | 1128/1170 [00:15<00:00, 69.88it/s]\u001b[A\n",
      " 97%|██████████████████████████████████████████████████████████████████████▉  | 1136/1170 [00:15<00:00, 69.54it/s]\u001b[A\n",
      " 98%|███████████████████████████████████████████████████████████████████████▌ | 1146/1170 [00:15<00:00, 74.88it/s]\u001b[A\n",
      " 99%|████████████████████████████████████████████████████████████████████████ | 1154/1170 [00:15<00:00, 74.80it/s]\u001b[A\n",
      " 99%|████████████████████████████████████████████████████████████████████████▌| 1162/1170 [00:15<00:00, 70.55it/s]\u001b[A\n",
      "100%|█████████████████████████████████████████████████████████████████████████| 1170/1170 [00:15<00:00, 73.88it/s]\u001b[A\n",
      "100%|███████████████████████████████████████████████████████████████████████████████| 1/1 [00:46<00:00, 46.09s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Writing 906 annotations to data/gamutrf/gamutrf-sd-gr-ieee-wifi/test_offline/wifi-encoding-0-sendsps-10000000.s16.zst.sigmf-meta\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "for f in tqdm(glob.glob(str(Path(data_directory,\"anom*.sigmf-data\")))):\n",
    "    annotation_utils.annotate(f, label=\"anom_wifi\", avg_window_len=256, avg_duration=2, debug=True, dry_run=False)\n",
    "\n",
    "for f in tqdm(glob.glob(str(Path(data_directory,\"wifi*.sigmf-data\")))):\n",
    "    annotation_utils.annotate(f, label=\"wifi\", avg_window_len=256, avg_duration=3, debug=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "32a1152e-8c8a-4088-bd3e-bc250b8c79c3",
   "metadata": {},
   "source": [
    "# Option 2: (unknown labels) Create SigMF-Meta annotations without core:label"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "62ada507-d435-45dc-832a-7136c7744f30",
   "metadata": {},
   "source": [
    "### Calling annotation_utils.annotate() with label=None will create unlabelled SigMF annotations for each burst that satisfies the median absolute deviation threshold criteria (see annotation_utils for more details)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4f78a707-4552-4f10-9da9-a4f00b8500c3",
   "metadata": {},
   "outputs": [],
   "source": [
    "for f in tqdm(glob.glob(str(Path(data_directory,\"*.sigmf-data\")))):\n",
    "    annotation_utils.annotate(f, label=None, avg_window_len=256, avg_duration=3, debug=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e78534eb-e774-4455-a780-c6c03336ac25",
   "metadata": {},
   "source": [
    "# Create SigMF Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2dcd1c9b-e536-4949-88f9-2f901e54d4b9",
   "metadata": {},
   "outputs": [],
   "source": [
    "transform = ST.Compose([\n",
    "    ST.Normalize(norm=np.inf),\n",
    "    ST.ComplexTo2D(),\n",
    "])\n",
    "dataset = SigMFDataset( \n",
    "    root=data_directory, \n",
    "    sample_count= num_iq_samples, \n",
    "    # class_list = list(idx_to_class.values()),\n",
    "    transform = transform,\n",
    "    only_first_samples=only_use_start_of_burst,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9f9f55d1-cc37-4bb4-8954-f57576af206e",
   "metadata": {},
   "source": [
    "# Load Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2f2f7196-c55e-4d45-a71a-9540e6369515",
   "metadata": {},
   "outputs": [],
   "source": [
    "model = torch.jit.load(model_file)\n",
    "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
    "model = model.to(device)\n",
    "\n",
    "example_model = CustomNetwork(model)\n",
    "\n",
    "example_model = example_model.eval()\n",
    "example_model = example_model.cuda() if torch.cuda.is_available() else example_model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "643a15cc-6cec-41a5-b548-83a24feb4a7f",
   "metadata": {},
   "source": [
    "# Create/Reset prediction_* SigMF-Meta files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d6e83c0a-a7d1-4681-b503-cdbbff971e96",
   "metadata": {},
   "outputs": [],
   "source": [
    "annotation_utils.reset_predictions_sigmf(dataset)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a5874218-44f6-419f-9315-2b4bb285793e",
   "metadata": {},
   "source": [
    "# Run inference on dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "be6fe885-089c-49df-aa47-cda265609b88",
   "metadata": {},
   "outputs": [],
   "source": [
    "example_model.eval()\n",
    "\n",
    "data_obj_dict = {}\n",
    "sigmf_obj_dict = {}\n",
    "for i in tqdm(range(len(dataset))):\n",
    "\n",
    "    # Run inference\n",
    "    data = torch.tensor(dataset[i][0]).float().reshape(1,2,-1)\n",
    "    data = data.cuda() if torch.cuda.is_available() else data\n",
    "    pred_tmp = example_model.predict(data)\n",
    "    pred_tmp = pred_tmp.cpu().numpy() if torch.cuda.is_available() else pred_tmp\n",
    "    prediction_class = idx_to_class[np.argmax(pred_tmp, axis=1)[0]]\n",
    "\n",
    "    # Prepare SigMF data \n",
    "    data_file = dataset.index[i][1].absolute_path\n",
    "    if data_file not in data_obj_dict:\n",
    "        data_obj = data_class.Data(data_file)\n",
    "        data_obj_dict[data_file] = data_obj\n",
    "    else:\n",
    "        data_obj = data_obj_dict[data_file]\n",
    "    \n",
    "    sample_size = dataset.index[i][1].item_type.itemsize * (\n",
    "        2 if \"c\" in data_obj.metadata[\"global\"][\"core:datatype\"] else 1\n",
    "    )\n",
    "    sample_start = int(dataset.index[i][1].byte_offset/sample_size)\n",
    "    metadata = {\n",
    "        \"core:label\": prediction_class,\n",
    "        \"core:freq_lower_edge\": data_obj.metadata[\"captures\"][0][\"core:frequency\"]\n",
    "        - data_obj.metadata[\"global\"][\"core:sample_rate\"] / 2,\n",
    "        \"core:freq_upper_edge\": data_obj.metadata[\"captures\"][0][\"core:frequency\"]\n",
    "        + data_obj.metadata[\"global\"][\"core:sample_rate\"] / 2,\n",
    "    }\n",
    "\n",
    "    # Add new annotations \n",
    "    prediction_meta_path = Path(Path(data_obj.sigmf_meta_filename).parent, f\"prediction_{Path(data_obj.sigmf_meta_filename).name}\")\n",
    "    if prediction_meta_path not in sigmf_obj_dict:\n",
    "        temp_sigmf_obj = sigmf.sigmffile.fromfile(prediction_meta_path)\n",
    "        sigmf_obj_dict[prediction_meta_path] = temp_sigmf_obj\n",
    "    else:\n",
    "        temp_sigmf_obj = sigmf_obj_dict[prediction_meta_path]\n",
    "    \n",
    "    temp_sigmf_obj.add_annotation(sample_start, length=num_iq_samples, metadata=metadata)\n",
    "\n",
    "# Write prediction_* SigMF-Meta \n",
    "for prediction_meta_path, temp_sigmf_obj in sigmf_obj_dict.items():\n",
    "    temp_sigmf_obj.tofile(prediction_meta_path, skip_validate=True)\n",
    "    print(f\"Writing prediction annotation to {prediction_meta_path}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d8f75d82-9219-4668-aece-da6f816c562c",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}