notebooks/bladerf_eda.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "a3010bb5-6676-4823-820a-b6b6df675703",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\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 matplotlib import pyplot as plt\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 torch import optim\n",
    "from torch.utils.data import DataLoader\n",
    "from torchsig.datasets.sig53 import Sig53\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"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "02127963-ab00-4540-82ab-65d82bef68c0",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "#dataset_path = \"./dev_data/torchsig_test/gamutrf-wifi-and-anom-bladerf\"\n",
    "dataset_path = \"./data/gamutrf/gamutrf-wifi-and-anom-bladerf\"\n",
    "num_iq_samples = 1024\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": 3,
   "id": "2dcd1c9b-e536-4949-88f9-2f901e54d4b9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Class List: ['anom_wifi', 'mini2_telem', 'mini2_video', 'wifi6', 'wifi', 'fhss_css']\n",
      "len(dataset)=92530\n",
      "{'anom_wifi': 46768, 'mini2_telem': 0, 'mini2_video': 0, 'wifi6': 0, 'wifi': 45762, 'fhss_css': 0}\n"
     ]
    }
   ],
   "source": [
    "transform = ST.Compose([\n",
    "    ST.Normalize(norm=np.inf),\n",
    "    ST.ComplexTo2D(),\n",
    "])\n",
    "dataset = SigMFDataset( \n",
    "    root=dataset_path, \n",
    "    sample_count= num_iq_samples, \n",
    "    class_list = list(idx_to_class.values()),\n",
    "    transform = transform,\n",
    "    only_first_samples=True,\n",
    ")\n",
    "dataset_class_counts = {class_name:0 for class_name in dataset.class_list}\n",
    "for data,label in dataset:\n",
    "    dataset_class_counts[dataset.class_list[label]] += 1\n",
    "print(f\"{len(dataset)=}\")\n",
    "print(dataset_class_counts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "8ad64afc-c4fc-4130-bc37-388d1759e468",
   "metadata": {},
   "outputs": [],
   "source": [
    "dataloader = DataLoader(\n",
    "    dataset=dataset,\n",
    "    batch_size=180,\n",
    "    num_workers=16,\n",
    "    shuffle=False,\n",
    "    drop_last=True,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "2f2f7196-c55e-4d45-a71a-9540e6369515",
   "metadata": {},
   "outputs": [],
   "source": [
    "model = torch.jit.load('weights/checkpoint-v16_torchscript.pt')\n",
    "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
    "model = model.to(device)\n",
    "\n",
    "example_model = CustomNetwork(model, dataloader)\n",
    "\n",
    "example_model = example_model.eval()\n",
    "example_model = example_model.cuda() if torch.cuda.is_available() else example_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "8b9a96a2-6b3e-4f38-88b8-2b36acc18712",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████| 514/514 [00:45<00:00, 11.41it/s]\n"
     ]
    }
   ],
   "source": [
    "# Infer results over validation set\n",
    "num_test_examples = len(dataset)\n",
    "y_preds = np.zeros((num_test_examples,))\n",
    "y_true = np.zeros((num_test_examples,))\n",
    "y_true_list = []\n",
    "y_preds_list = []\n",
    "with torch.no_grad():\n",
    "    example_model.eval()\n",
    "    for data, label in tqdm(dataloader):\n",
    "        # Infer\n",
    "        data = data.float()\n",
    "\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",
    "\n",
    "        y_preds_list.extend(np.argmax(pred_tmp, axis=1).tolist())\n",
    "        y_true_list.extend(label.tolist())\n",
    "        \n",
    "y_preds = y_preds_list\n",
    "y_true = y_true_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "4e0e698f-2bca-418c-b35e-f3f3a28b15e2",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ltindall/.pyenv/versions/3.10.9/envs/rfml-dev/lib/python3.10/site-packages/torchsig/utils/cm_plotter.py:31: RuntimeWarning: invalid value encountered in divide\n",
      "  cm = cm.astype(\"float\") / cm.sum(axis=1)[:, np.newaxis]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 2600x2000 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "acc = np.sum(np.asarray(y_preds)==np.asarray(y_true))/len(y_true)\n",
    "plot_confusion_matrix(\n",
    "    y_true, \n",
    "    y_preds, \n",
    "    classes=np.unique(np.concatenate((np.arange(len(dataset.class_list)),np.unique(y_preds)))),\n",
    "    normalize=True,\n",
    "    title=\"Example Modulations Confusion Matrix\\nTotal Accuracy: {:.2f}%\".format(acc*100),\n",
    "    text=True,\n",
    "    rotate_x_text=90,\n",
    "    figsize=(26,20),\n",
    ")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5ef988d4-a4c1-4103-9bf3-070fcb142a0b",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "71e05124-2106-46b1-8eda-e05317a91ee7",
   "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
}