main.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "jzJPAMwSY65V"
   },
   "source": [
    "# Fraud Detection with Pytorch Geometric"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "1Kcc3IMdZBwA"
   },
   "source": [
    "## Installation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "ER0pkl5Q0ZCI",
    "outputId": "6be88e5a-3d59-4743-d58c-78ef79380bfa"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.11.0\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\leok8\\miniconda3\\envs\\net-sci\\lib\\site-packages\\tqdm\\auto.py:22: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n"
     ]
    }
   ],
   "source": [
    "# Install required packages.\n",
    "import os\n",
    "import torch\n",
    "\n",
    "os.environ[\"TORCH\"] = torch.__version__\n",
    "print(torch.__version__)\n",
    "\n",
    "!pip install -q torch-scatter -f https://data.pyg.org/whl/torch-${TORCH}.html\n",
    "!pip install -q torch-sparse -f https://data.pyg.org/whl/torch-${TORCH}.html\n",
    "!pip install -q git+https://github.com/pyg-team/pytorch_geometric.git"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "from src.utils import to_networkx\n",
    "from src.model import GNN, GNNWithAttentionFusion"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ZyB3RRbmZD3o"
   },
   "source": [
    "## Dataset\n",
    "\n",
    "- Contains news propagation graphs extracted from Twitter\n",
    "- Source and raw data: https://github.com/KaiDMML/FakeNewsNet\n",
    "- Preprocessing: https://arxiv.org/pdf/2104.12259.pdf\n",
    "- feature=\"content\" --> Spacy Word2Vec + Profile features\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "tvB33IQ5k7Tr",
    "outputId": "6ad7bebb-8b7a-4782-e269-9cb7a0fe6cf0"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Samples:  1092\n",
      "Test Samples:  3826\n"
     ]
    }
   ],
   "source": [
    "from torch_geometric.datasets import UPFD\n",
    "\n",
    "train_data = UPFD(root=\".\", name=\"gossipcop\", feature=\"content\", split=\"train\")\n",
    "test_data = UPFD(root=\".\", name=\"gossipcop\", feature=\"content\", split=\"test\")\n",
    "print(\"Train Samples: \", len(train_data))\n",
    "print(\"Test Samples: \", len(test_data))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "RdM8rXX2hqtw"
   },
   "source": [
    "### Investigating the News Propagation Graph"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "5qz9MHtnTsvx",
    "outputId": "1810e773-6c1f-4d38-861d-accd06df0beb"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([125, 310])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample_id = 1\n",
    "train_data[sample_id].x.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "-mGLMrSFjAOB",
    "outputId": "d321b3ed-4553-4825-e01c-2652c5a3c706"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([1])"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data[sample_id].y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "8NOGcUvXjEdu",
    "outputId": "14b9a100-f79f-4cfa-a412-85812539b76b"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[0.5220, 0.5120, 0.4817,  ..., 0.6874, 0.1023, 0.1529],\n",
       "        [0.4610, 0.4544, 0.6279,  ..., 0.2055, 0.1667, 0.2500],\n",
       "        [0.4461, 0.4728, 0.4978,  ..., 0.8356, 0.0556, 0.0000],\n",
       "        ...,\n",
       "        [0.5574, 0.4934, 0.5588,  ..., 0.3562, 0.0556, 0.2647],\n",
       "        [0.5274, 0.5889, 0.4425,  ..., 0.9589, 0.1111, 0.1176],\n",
       "        [0.5220, 0.5120, 0.4817,  ..., 0.6874, 0.1023, 0.1529]])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data[sample_id].x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "zYeQSJlIjI50",
    "outputId": "601c0489-ddde-47bd-f3ae-0080698440c8"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[  0,   1],\n",
       "        [  0,   2],\n",
       "        [  0,   3],\n",
       "        [  0,   4],\n",
       "        [  0,   5],\n",
       "        [  0,   6],\n",
       "        [  0,   7],\n",
       "        [  0,   8],\n",
       "        [  0,   9],\n",
       "        [  0,  10],\n",
       "        [  0,  11],\n",
       "        [  0,  12],\n",
       "        [  0,  13],\n",
       "        [  0,  14],\n",
       "        [  0,  15],\n",
       "        [  0,  16],\n",
       "        [  0,  17],\n",
       "        [  0,  18],\n",
       "        [  0,  19],\n",
       "        [  0,  20],\n",
       "        [  0,  21],\n",
       "        [  0,  22],\n",
       "        [  0,  23],\n",
       "        [  0,  24],\n",
       "        [  0,  25],\n",
       "        [  0,  26],\n",
       "        [  0,  27],\n",
       "        [  0,  28],\n",
       "        [  0,  29],\n",
       "        [  0,  30],\n",
       "        [  0,  31],\n",
       "        [  0,  32],\n",
       "        [  0,  33],\n",
       "        [  0,  34],\n",
       "        [  0,  35],\n",
       "        [  0,  36],\n",
       "        [  0,  37],\n",
       "        [  0,  38],\n",
       "        [ 15,  39],\n",
       "        [ 15,  40],\n",
       "        [ 17,  43],\n",
       "        [ 17,  44],\n",
       "        [ 22,  45],\n",
       "        [ 22,  46],\n",
       "        [ 22,  47],\n",
       "        [ 22,  48],\n",
       "        [ 23,  49],\n",
       "        [ 26,  50],\n",
       "        [ 26,  51],\n",
       "        [ 26,  52],\n",
       "        [ 26,  53],\n",
       "        [ 26,  54],\n",
       "        [ 26,  55],\n",
       "        [ 26,  56],\n",
       "        [ 26,  57],\n",
       "        [ 27,  59],\n",
       "        [ 27,  60],\n",
       "        [ 27,  61],\n",
       "        [ 28,  63],\n",
       "        [ 28,  64],\n",
       "        [ 28,  65],\n",
       "        [ 28,  66],\n",
       "        [ 28,  67],\n",
       "        [ 28,  68],\n",
       "        [ 28,  69],\n",
       "        [ 28,  70],\n",
       "        [ 28,  71],\n",
       "        [ 28,  72],\n",
       "        [ 28,  73],\n",
       "        [ 28,  74],\n",
       "        [ 28,  75],\n",
       "        [ 28,  76],\n",
       "        [ 28,  77],\n",
       "        [ 28,  78],\n",
       "        [ 28,  79],\n",
       "        [ 28,  80],\n",
       "        [ 28,  81],\n",
       "        [ 28,  82],\n",
       "        [ 28,  83],\n",
       "        [ 28,  84],\n",
       "        [ 28,  85],\n",
       "        [ 28,  86],\n",
       "        [ 28,  87],\n",
       "        [ 28,  88],\n",
       "        [ 28,  89],\n",
       "        [ 28,  90],\n",
       "        [ 28,  91],\n",
       "        [ 28,  92],\n",
       "        [ 28,  93],\n",
       "        [ 28,  94],\n",
       "        [ 28,  95],\n",
       "        [ 28,  96],\n",
       "        [ 30, 115],\n",
       "        [ 34, 119],\n",
       "        [ 34, 120],\n",
       "        [ 38, 121],\n",
       "        [ 38, 122],\n",
       "        [ 38, 123],\n",
       "        [ 38, 124],\n",
       "        [ 39,  41],\n",
       "        [ 40,  42],\n",
       "        [ 57,  58],\n",
       "        [ 59,  62],\n",
       "        [ 65,  97],\n",
       "        [ 65,  98],\n",
       "        [ 65,  99],\n",
       "        [ 66, 100],\n",
       "        [ 68, 101],\n",
       "        [ 69, 103],\n",
       "        [ 77, 104],\n",
       "        [ 77, 105],\n",
       "        [ 79, 106],\n",
       "        [ 79, 107],\n",
       "        [ 80, 108],\n",
       "        [ 83, 109],\n",
       "        [ 83, 110],\n",
       "        [ 83, 111],\n",
       "        [ 83, 112],\n",
       "        [ 84, 114],\n",
       "        [101, 102],\n",
       "        [110, 113],\n",
       "        [115, 116],\n",
       "        [116, 117],\n",
       "        [117, 118]])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data[sample_id].edge_index.t()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "cu2L49DKh8AJ",
    "outputId": "c328dc6b-662b-4cb8-fbed-35f3e23045f6"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([1])\n",
      "tensor([1])\n",
      "tensor([0])\n",
      "tensor([0])\n"
     ]
    }
   ],
   "source": [
    "for data in train_data:\n",
    "    print(data.y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "7yqF8X6Hin7i",
    "outputId": "fd33aa44-653b-4186-8539-b5417a467c18"
   },
   "outputs": [],
   "source": [
    "import networkx as nx"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 319
    },
    "id": "smHrIIV1iQfJ",
    "outputId": "d0bfafcd-55d3-4be3-d02f-5aa2ff51ee06"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "nx.draw(to_networkx(train_data[sample_id]), with_labels=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "xU_P6S-Zht5A"
   },
   "source": [
    "### Node features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "jmZMuKXv6S0g",
    "outputId": "bb4014f2-5d7d-4cee-fd54-c9e36a77a3e8"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([125, 310])\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "tensor([[0.5220, 0.5120, 0.4817,  ..., 0.6874, 0.1023, 0.1529],\n",
       "        [0.4610, 0.4544, 0.6279,  ..., 0.2055, 0.1667, 0.2500],\n",
       "        [0.4461, 0.4728, 0.4978,  ..., 0.8356, 0.0556, 0.0000],\n",
       "        ...,\n",
       "        [0.5574, 0.4934, 0.5588,  ..., 0.3562, 0.0556, 0.2647],\n",
       "        [0.5274, 0.5889, 0.4425,  ..., 0.9589, 0.1111, 0.1176],\n",
       "        [0.5220, 0.5120, 0.4817,  ..., 0.6874, 0.1023, 0.1529]])"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(train_data[sample_id].x.shape)\n",
    "train_data[sample_id].x"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "cRzLDYg8Ec_4"
   },
   "source": [
    "### Class distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 282
    },
    "id": "Ul5y-1IKEcdU",
    "outputId": "325ad0f8-b21a-4891-9d91-3ed59cd8d434"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "labels = [data.y.item() for i, data in enumerate(train_data)]\n",
    "df = pd.DataFrame(labels, columns=[\"Labels\"])\n",
    "df[\"Labels\"].hist()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "vzAmnPlhl5iO"
   },
   "source": [
    "### Data Loaders"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "id": "c0DhfbZgl44G"
   },
   "outputs": [],
   "source": [
    "from torch_geometric.loader import DataLoader\n",
    "\n",
    "train_loader = DataLoader(train_data, batch_size=128, shuffle=True)\n",
    "test_loader = DataLoader(test_data, batch_size=128, shuffle=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "C8m-dJ-WZFM1"
   },
   "source": [
    "## Model and Training\n",
    "\n",
    "--> Because it is a directed graph, it will only share information from the root"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cuda\n"
     ]
    }
   ],
   "source": [
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "# device = torch.device(\"mps\" if torch.backends.mps.is_available() else \"cpu\")\n",
    "print(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 00 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 01 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 02 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.54 | TestF1: 0.16\n",
      "Epoch: 03 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 04 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.73 | TestF1: 0.64\n",
      "Epoch: 05 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 06 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.84 | TestF1: 0.82\n",
      "Epoch: 07 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.68 | TestF1: 0.53\n",
      "Epoch: 08 |  TrainLoss: 0.68 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 09 |  TrainLoss: 0.68 | TestLoss: 0.68 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 10 |  TrainLoss: 0.68 | TestLoss: 0.68 | TestAcc: 0.76 | TestF1: 0.80\n",
      "Epoch: 11 |  TrainLoss: 0.68 | TestLoss: 0.68 | TestAcc: 0.61 | TestF1: 0.35\n",
      "Epoch: 12 |  TrainLoss: 0.68 | TestLoss: 0.68 | TestAcc: 0.73 | TestF1: 0.78\n",
      "Epoch: 13 |  TrainLoss: 0.68 | TestLoss: 0.68 | TestAcc: 0.50 | TestF1: 0.01\n",
      "Epoch: 14 |  TrainLoss: 0.68 | TestLoss: 0.68 | TestAcc: 0.64 | TestF1: 0.43\n",
      "Epoch: 15 |  TrainLoss: 0.68 | TestLoss: 0.68 | TestAcc: 0.86 | TestF1: 0.84\n",
      "Epoch: 16 |  TrainLoss: 0.68 | TestLoss: 0.68 | TestAcc: 0.54 | TestF1: 0.15\n",
      "Epoch: 17 |  TrainLoss: 0.68 | TestLoss: 0.67 | TestAcc: 0.88 | TestF1: 0.89\n",
      "Epoch: 18 |  TrainLoss: 0.67 | TestLoss: 0.67 | TestAcc: 0.61 | TestF1: 0.37\n",
      "Epoch: 19 |  TrainLoss: 0.67 | TestLoss: 0.67 | TestAcc: 0.66 | TestF1: 0.49\n",
      "Epoch: 20 |  TrainLoss: 0.67 | TestLoss: 0.67 | TestAcc: 0.86 | TestF1: 0.88\n",
      "Epoch: 21 |  TrainLoss: 0.67 | TestLoss: 0.67 | TestAcc: 0.55 | TestF1: 0.18\n",
      "Epoch: 22 |  TrainLoss: 0.67 | TestLoss: 0.67 | TestAcc: 0.91 | TestF1: 0.91\n",
      "Epoch: 23 |  TrainLoss: 0.67 | TestLoss: 0.67 | TestAcc: 0.58 | TestF1: 0.29\n",
      "Epoch: 24 |  TrainLoss: 0.67 | TestLoss: 0.66 | TestAcc: 0.92 | TestF1: 0.92\n",
      "Epoch: 25 |  TrainLoss: 0.66 | TestLoss: 0.66 | TestAcc: 0.78 | TestF1: 0.73\n",
      "Epoch: 26 |  TrainLoss: 0.66 | TestLoss: 0.66 | TestAcc: 0.85 | TestF1: 0.83\n",
      "Epoch: 27 |  TrainLoss: 0.66 | TestLoss: 0.66 | TestAcc: 0.83 | TestF1: 0.80\n",
      "Epoch: 28 |  TrainLoss: 0.66 | TestLoss: 0.65 | TestAcc: 0.77 | TestF1: 0.71\n",
      "Epoch: 29 |  TrainLoss: 0.65 | TestLoss: 0.65 | TestAcc: 0.80 | TestF1: 0.75\n",
      "Epoch: 30 |  TrainLoss: 0.65 | TestLoss: 0.65 | TestAcc: 0.86 | TestF1: 0.84\n",
      "Epoch: 31 |  TrainLoss: 0.65 | TestLoss: 0.64 | TestAcc: 0.91 | TestF1: 0.91\n",
      "Epoch: 32 |  TrainLoss: 0.64 | TestLoss: 0.64 | TestAcc: 0.72 | TestF1: 0.62\n",
      "Epoch: 33 |  TrainLoss: 0.64 | TestLoss: 0.64 | TestAcc: 0.86 | TestF1: 0.84\n",
      "Epoch: 34 |  TrainLoss: 0.63 | TestLoss: 0.63 | TestAcc: 0.92 | TestF1: 0.92\n",
      "Epoch: 35 |  TrainLoss: 0.63 | TestLoss: 0.63 | TestAcc: 0.88 | TestF1: 0.87\n",
      "Epoch: 36 |  TrainLoss: 0.62 | TestLoss: 0.62 | TestAcc: 0.83 | TestF1: 0.79\n",
      "Epoch: 37 |  TrainLoss: 0.62 | TestLoss: 0.61 | TestAcc: 0.89 | TestF1: 0.88\n",
      "Epoch: 38 |  TrainLoss: 0.61 | TestLoss: 0.61 | TestAcc: 0.76 | TestF1: 0.68\n",
      "Epoch: 39 |  TrainLoss: 0.60 | TestLoss: 0.60 | TestAcc: 0.73 | TestF1: 0.63\n",
      "Epoch: 40 |  TrainLoss: 0.59 | TestLoss: 0.59 | TestAcc: 0.79 | TestF1: 0.74\n",
      "Epoch: 41 |  TrainLoss: 0.58 | TestLoss: 0.58 | TestAcc: 0.72 | TestF1: 0.62\n",
      "Epoch: 42 |  TrainLoss: 0.56 | TestLoss: 0.57 | TestAcc: 0.71 | TestF1: 0.60\n",
      "Epoch: 43 |  TrainLoss: 0.55 | TestLoss: 0.53 | TestAcc: 0.92 | TestF1: 0.92\n",
      "Epoch: 44 |  TrainLoss: 0.52 | TestLoss: 0.51 | TestAcc: 0.89 | TestF1: 0.90\n",
      "Epoch: 45 |  TrainLoss: 0.49 | TestLoss: 0.47 | TestAcc: 0.90 | TestF1: 0.90\n",
      "Epoch: 46 |  TrainLoss: 0.46 | TestLoss: 0.43 | TestAcc: 0.91 | TestF1: 0.91\n",
      "Epoch: 47 |  TrainLoss: 0.42 | TestLoss: 0.43 | TestAcc: 0.82 | TestF1: 0.78\n",
      "Epoch: 48 |  TrainLoss: 0.40 | TestLoss: 0.39 | TestAcc: 0.88 | TestF1: 0.89\n",
      "Epoch: 49 |  TrainLoss: 0.37 | TestLoss: 0.38 | TestAcc: 0.84 | TestF1: 0.81\n",
      "Epoch: 50 |  TrainLoss: 0.33 | TestLoss: 0.31 | TestAcc: 0.91 | TestF1: 0.91\n",
      "Epoch: 51 |  TrainLoss: 0.30 | TestLoss: 0.33 | TestAcc: 0.86 | TestF1: 0.84\n",
      "Epoch: 52 |  TrainLoss: 0.30 | TestLoss: 0.35 | TestAcc: 0.86 | TestF1: 0.88\n",
      "Epoch: 53 |  TrainLoss: 0.28 | TestLoss: 0.32 | TestAcc: 0.85 | TestF1: 0.82\n",
      "Epoch: 54 |  TrainLoss: 0.28 | TestLoss: 0.29 | TestAcc: 0.90 | TestF1: 0.91\n",
      "Epoch: 55 |  TrainLoss: 0.25 | TestLoss: 0.25 | TestAcc: 0.91 | TestF1: 0.91\n",
      "Epoch: 56 |  TrainLoss: 0.23 | TestLoss: 0.24 | TestAcc: 0.92 | TestF1: 0.92\n",
      "Epoch: 57 |  TrainLoss: 0.22 | TestLoss: 0.22 | TestAcc: 0.92 | TestF1: 0.92\n",
      "Epoch: 58 |  TrainLoss: 0.20 | TestLoss: 0.21 | TestAcc: 0.94 | TestF1: 0.93\n",
      "Epoch: 59 |  TrainLoss: 0.19 | TestLoss: 0.20 | TestAcc: 0.94 | TestF1: 0.94\n"
     ]
    }
   ],
   "source": [
    "from src.trainer import Trainer\n",
    "\n",
    "model = GNN(train_data.num_features, 128, 1).to(device)\n",
    "optimizer = torch.optim.Adam(model.parameters(), lr=0.0002, weight_decay=0.01)\n",
    "loss_fnc = torch.nn.BCELoss()\n",
    "\n",
    "trainer = Trainer(model, loss_func=loss_fnc, optimizer=optimizer, device=device)\n",
    "\n",
    "train_losses_concat_lr_2e_4 = []\n",
    "test_losses_concat_lr_2e_4 = []\n",
    "test_acces_concat_lr_2e_4 = []\n",
    "test_f1score_concat_lr_2e_4 = []\n",
    "\n",
    "for epoch in range(60):\n",
    "    train_loss = trainer.train(train_loader)\n",
    "    test_loss, test_acc, test_f1 = trainer.test(test_loader)\n",
    "\n",
    "    train_losses_concat_lr_2e_4.append(train_loss)\n",
    "    test_losses_concat_lr_2e_4.append(test_loss)\n",
    "    test_acces_concat_lr_2e_4.append(test_acc)\n",
    "    test_f1score_concat_lr_2e_4.append(test_f1)\n",
    "\n",
    "    print(\n",
    "        f\"Epoch: {epoch:02d} |  TrainLoss: {train_loss:.2f} | \"\n",
    "        f\"TestLoss: {test_loss:.2f} | TestAcc: {test_acc:.2f} | TestF1: {test_f1:.2f}\"\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 235
    },
    "id": "W_5IZcgCFyav",
    "outputId": "314cebb5-5136-4d55-eb74-e1556e3c5ebc"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   pred_logit  pred  true\n",
      "0    0.576764   1.0     1\n",
      "1    0.503087   1.0     1\n",
      "2    0.430423   0.0     0\n",
      "3    0.512809   1.0     1\n",
      "4    0.339395   0.0     0\n",
      "5    0.436462   0.0     0\n",
      "6    0.420225   0.0     0\n",
      "7    0.617786   1.0     1\n",
      "8    0.338772   0.0     0\n",
      "9    0.413669   0.0     0\n"
     ]
    }
   ],
   "source": [
    "for data in test_loader:\n",
    "    data = data.to(device)\n",
    "    pred = trainer.model(data.x, data.edge_index, data.batch)\n",
    "    df = pd.DataFrame()\n",
    "    df[\"pred_logit\"] = pred.cpu().detach().numpy()[:, 0]\n",
    "    df[\"pred\"] = torch.round(pred).cpu().detach().numpy()[:, 0]\n",
    "    df[\"true\"] = data.y.cpu().numpy()\n",
    "    print(df.head(10))\n",
    "    break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 00 |  TrainLoss: 0.75 | TestLoss: 0.70 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 01 |  TrainLoss: 0.69 | TestLoss: 0.71 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 02 |  TrainLoss: 0.69 | TestLoss: 0.76 | TestAcc: 0.50 | TestF1: 0.67\n",
      "Epoch: 03 |  TrainLoss: 0.72 | TestLoss: 0.70 | TestAcc: 0.50 | TestF1: 0.67\n",
      "Epoch: 04 |  TrainLoss: 0.70 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 05 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 06 |  TrainLoss: 0.70 | TestLoss: 0.68 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 07 |  TrainLoss: 0.69 | TestLoss: 0.68 | TestAcc: 0.50 | TestF1: 0.67\n",
      "Epoch: 08 |  TrainLoss: 0.68 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.67\n",
      "Epoch: 09 |  TrainLoss: 0.68 | TestLoss: 0.68 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 10 |  TrainLoss: 0.68 | TestLoss: 0.68 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 11 |  TrainLoss: 0.66 | TestLoss: 0.66 | TestAcc: 0.57 | TestF1: 0.25\n",
      "Epoch: 12 |  TrainLoss: 0.66 | TestLoss: 0.66 | TestAcc: 0.50 | TestF1: 0.67\n",
      "Epoch: 13 |  TrainLoss: 0.73 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 14 |  TrainLoss: 0.67 | TestLoss: 0.66 | TestAcc: 0.54 | TestF1: 0.13\n",
      "Epoch: 15 |  TrainLoss: 0.66 | TestLoss: 0.66 | TestAcc: 0.51 | TestF1: 0.06\n",
      "Epoch: 16 |  TrainLoss: 0.66 | TestLoss: 0.64 | TestAcc: 0.52 | TestF1: 0.67\n",
      "Epoch: 17 |  TrainLoss: 0.65 | TestLoss: 0.65 | TestAcc: 0.50 | TestF1: 0.67\n",
      "Epoch: 18 |  TrainLoss: 0.66 | TestLoss: 0.67 | TestAcc: 0.50 | TestF1: 0.01\n",
      "Epoch: 19 |  TrainLoss: 0.66 | TestLoss: 0.63 | TestAcc: 0.53 | TestF1: 0.12\n",
      "Epoch: 20 |  TrainLoss: 0.61 | TestLoss: 0.60 | TestAcc: 0.66 | TestF1: 0.49\n",
      "Epoch: 21 |  TrainLoss: 0.59 | TestLoss: 0.57 | TestAcc: 0.76 | TestF1: 0.69\n",
      "Epoch: 22 |  TrainLoss: 0.60 | TestLoss: 0.69 | TestAcc: 0.51 | TestF1: 0.03\n",
      "Epoch: 23 |  TrainLoss: 0.66 | TestLoss: 0.63 | TestAcc: 0.53 | TestF1: 0.13\n",
      "Epoch: 24 |  TrainLoss: 0.60 | TestLoss: 0.58 | TestAcc: 0.54 | TestF1: 0.69\n",
      "Epoch: 25 |  TrainLoss: 0.58 | TestLoss: 0.53 | TestAcc: 0.89 | TestF1: 0.90\n",
      "Epoch: 26 |  TrainLoss: 0.58 | TestLoss: 0.63 | TestAcc: 0.55 | TestF1: 0.17\n",
      "Epoch: 27 |  TrainLoss: 0.56 | TestLoss: 0.54 | TestAcc: 0.69 | TestF1: 0.55\n",
      "Epoch: 28 |  TrainLoss: 0.51 | TestLoss: 0.51 | TestAcc: 0.82 | TestF1: 0.84\n",
      "Epoch: 29 |  TrainLoss: 0.51 | TestLoss: 0.57 | TestAcc: 0.54 | TestF1: 0.68\n",
      "Epoch: 30 |  TrainLoss: 0.49 | TestLoss: 0.46 | TestAcc: 0.90 | TestF1: 0.91\n",
      "Epoch: 31 |  TrainLoss: 0.46 | TestLoss: 0.57 | TestAcc: 0.61 | TestF1: 0.36\n",
      "Epoch: 32 |  TrainLoss: 0.52 | TestLoss: 0.43 | TestAcc: 0.92 | TestF1: 0.91\n",
      "Epoch: 33 |  TrainLoss: 0.46 | TestLoss: 0.42 | TestAcc: 0.87 | TestF1: 0.86\n",
      "Epoch: 34 |  TrainLoss: 0.45 | TestLoss: 0.61 | TestAcc: 0.51 | TestF1: 0.67\n",
      "Epoch: 35 |  TrainLoss: 0.46 | TestLoss: 0.39 | TestAcc: 0.92 | TestF1: 0.91\n",
      "Epoch: 36 |  TrainLoss: 0.41 | TestLoss: 0.38 | TestAcc: 0.91 | TestF1: 0.91\n",
      "Epoch: 37 |  TrainLoss: 0.39 | TestLoss: 0.37 | TestAcc: 0.91 | TestF1: 0.91\n",
      "Epoch: 38 |  TrainLoss: 0.38 | TestLoss: 0.45 | TestAcc: 0.77 | TestF1: 0.81\n",
      "Epoch: 39 |  TrainLoss: 0.39 | TestLoss: 0.36 | TestAcc: 0.93 | TestF1: 0.93\n",
      "Epoch: 40 |  TrainLoss: 0.36 | TestLoss: 0.35 | TestAcc: 0.92 | TestF1: 0.93\n",
      "Epoch: 41 |  TrainLoss: 0.34 | TestLoss: 0.35 | TestAcc: 0.89 | TestF1: 0.88\n",
      "Epoch: 42 |  TrainLoss: 0.35 | TestLoss: 0.54 | TestAcc: 0.65 | TestF1: 0.47\n",
      "Epoch: 43 |  TrainLoss: 0.41 | TestLoss: 0.33 | TestAcc: 0.93 | TestF1: 0.93\n",
      "Epoch: 44 |  TrainLoss: 0.35 | TestLoss: 0.35 | TestAcc: 0.90 | TestF1: 0.91\n",
      "Epoch: 45 |  TrainLoss: 0.44 | TestLoss: 0.52 | TestAcc: 0.67 | TestF1: 0.75\n",
      "Epoch: 46 |  TrainLoss: 0.37 | TestLoss: 0.34 | TestAcc: 0.90 | TestF1: 0.91\n",
      "Epoch: 47 |  TrainLoss: 0.33 | TestLoss: 0.34 | TestAcc: 0.89 | TestF1: 0.90\n",
      "Epoch: 48 |  TrainLoss: 0.33 | TestLoss: 0.32 | TestAcc: 0.92 | TestF1: 0.93\n",
      "Epoch: 49 |  TrainLoss: 0.30 | TestLoss: 0.35 | TestAcc: 0.88 | TestF1: 0.89\n",
      "Epoch: 50 |  TrainLoss: 0.33 | TestLoss: 0.35 | TestAcc: 0.87 | TestF1: 0.88\n",
      "Epoch: 51 |  TrainLoss: 0.33 | TestLoss: 0.32 | TestAcc: 0.92 | TestF1: 0.92\n",
      "Epoch: 52 |  TrainLoss: 0.30 | TestLoss: 0.30 | TestAcc: 0.93 | TestF1: 0.93\n",
      "Epoch: 53 |  TrainLoss: 0.31 | TestLoss: 0.29 | TestAcc: 0.93 | TestF1: 0.94\n",
      "Epoch: 54 |  TrainLoss: 0.30 | TestLoss: 0.28 | TestAcc: 0.94 | TestF1: 0.94\n",
      "Epoch: 55 |  TrainLoss: 0.30 | TestLoss: 0.30 | TestAcc: 0.91 | TestF1: 0.90\n",
      "Epoch: 56 |  TrainLoss: 0.29 | TestLoss: 0.33 | TestAcc: 0.85 | TestF1: 0.83\n",
      "Epoch: 57 |  TrainLoss: 0.30 | TestLoss: 0.38 | TestAcc: 0.80 | TestF1: 0.75\n",
      "Epoch: 58 |  TrainLoss: 0.36 | TestLoss: 0.32 | TestAcc: 0.87 | TestF1: 0.85\n",
      "Epoch: 59 |  TrainLoss: 0.32 | TestLoss: 0.28 | TestAcc: 0.93 | TestF1: 0.93\n"
     ]
    }
   ],
   "source": [
    "model_2 = GNN(train_data.num_features, 128, 1).to(device)\n",
    "optimizer = torch.optim.Adam(model_2.parameters(), lr=0.01, weight_decay=0.01)\n",
    "loss_fnc = torch.nn.BCELoss()\n",
    "\n",
    "trainer_2 = Trainer(\n",
    "    model=model_2, loss_func=loss_fnc, optimizer=optimizer, device=device\n",
    ")\n",
    "\n",
    "train_losses_concat_lr_1e_2 = []\n",
    "test_losses_concat_lr_1e_2 = []\n",
    "test_acces_concat_lr_1e_2 = []\n",
    "test_f1score_concat_lr_1e_2 = []\n",
    "\n",
    "for epoch in range(60):\n",
    "    train_loss = trainer_2.train(train_loader)\n",
    "    test_loss, test_acc, test_f1 = trainer_2.test(test_loader)\n",
    "\n",
    "    train_losses_concat_lr_1e_2.append(train_loss)\n",
    "    test_losses_concat_lr_1e_2.append(test_loss)\n",
    "    test_acces_concat_lr_1e_2.append(test_acc)\n",
    "    test_f1score_concat_lr_1e_2.append(test_acc)\n",
    "\n",
    "    print(\n",
    "        f\"Epoch: {epoch:02d} |  TrainLoss: {train_loss:.2f} | \"\n",
    "        f\"TestLoss: {test_loss:.2f} | TestAcc: {test_acc:.2f} | TestF1: {test_f1:.2f}\"\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "uf2YD4IG2kbO",
    "outputId": "abf6e525-1193-4935-f904-9cbcdf47d4a8"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 00 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 01 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 02 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 03 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 04 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 05 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 06 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 07 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 08 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 09 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 10 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 11 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 12 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.85 | TestF1: 0.86\n",
      "Epoch: 13 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 14 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.52 | TestF1: 0.07\n",
      "Epoch: 15 |  TrainLoss: 0.69 | TestLoss: 0.68 | TestAcc: 0.91 | TestF1: 0.91\n",
      "Epoch: 16 |  TrainLoss: 0.68 | TestLoss: 0.68 | TestAcc: 0.61 | TestF1: 0.35\n",
      "Epoch: 17 |  TrainLoss: 0.68 | TestLoss: 0.67 | TestAcc: 0.85 | TestF1: 0.83\n",
      "Epoch: 18 |  TrainLoss: 0.67 | TestLoss: 0.68 | TestAcc: 0.50 | TestF1: 0.00\n",
      "Epoch: 19 |  TrainLoss: 0.67 | TestLoss: 0.67 | TestAcc: 0.50 | TestF1: 0.67\n",
      "Epoch: 20 |  TrainLoss: 0.67 | TestLoss: 0.66 | TestAcc: 0.51 | TestF1: 0.05\n",
      "Epoch: 21 |  TrainLoss: 0.66 | TestLoss: 0.64 | TestAcc: 0.73 | TestF1: 0.63\n",
      "Epoch: 22 |  TrainLoss: 0.65 | TestLoss: 0.64 | TestAcc: 0.55 | TestF1: 0.69\n",
      "Epoch: 23 |  TrainLoss: 0.63 | TestLoss: 0.62 | TestAcc: 0.66 | TestF1: 0.49\n",
      "Epoch: 24 |  TrainLoss: 0.62 | TestLoss: 0.61 | TestAcc: 0.78 | TestF1: 0.72\n",
      "Epoch: 25 |  TrainLoss: 0.60 | TestLoss: 0.60 | TestAcc: 0.67 | TestF1: 0.75\n",
      "Epoch: 26 |  TrainLoss: 0.59 | TestLoss: 0.58 | TestAcc: 0.75 | TestF1: 0.80\n",
      "Epoch: 27 |  TrainLoss: 0.57 | TestLoss: 0.55 | TestAcc: 0.90 | TestF1: 0.91\n",
      "Epoch: 28 |  TrainLoss: 0.56 | TestLoss: 0.54 | TestAcc: 0.91 | TestF1: 0.91\n",
      "Epoch: 29 |  TrainLoss: 0.53 | TestLoss: 0.51 | TestAcc: 0.92 | TestF1: 0.92\n",
      "Epoch: 30 |  TrainLoss: 0.50 | TestLoss: 0.49 | TestAcc: 0.89 | TestF1: 0.89\n",
      "Epoch: 31 |  TrainLoss: 0.49 | TestLoss: 0.52 | TestAcc: 0.68 | TestF1: 0.53\n",
      "Epoch: 32 |  TrainLoss: 0.47 | TestLoss: 0.46 | TestAcc: 0.86 | TestF1: 0.84\n",
      "Epoch: 33 |  TrainLoss: 0.44 | TestLoss: 0.46 | TestAcc: 0.79 | TestF1: 0.73\n",
      "Epoch: 34 |  TrainLoss: 0.45 | TestLoss: 0.43 | TestAcc: 0.85 | TestF1: 0.83\n",
      "Epoch: 35 |  TrainLoss: 0.42 | TestLoss: 0.40 | TestAcc: 0.92 | TestF1: 0.92\n",
      "Epoch: 36 |  TrainLoss: 0.39 | TestLoss: 0.38 | TestAcc: 0.92 | TestF1: 0.92\n",
      "Epoch: 37 |  TrainLoss: 0.38 | TestLoss: 0.37 | TestAcc: 0.91 | TestF1: 0.91\n",
      "Epoch: 38 |  TrainLoss: 0.36 | TestLoss: 0.36 | TestAcc: 0.92 | TestF1: 0.92\n",
      "Epoch: 39 |  TrainLoss: 0.35 | TestLoss: 0.36 | TestAcc: 0.90 | TestF1: 0.91\n",
      "Epoch: 40 |  TrainLoss: 0.36 | TestLoss: 0.33 | TestAcc: 0.92 | TestF1: 0.92\n",
      "Epoch: 41 |  TrainLoss: 0.32 | TestLoss: 0.32 | TestAcc: 0.92 | TestF1: 0.92\n",
      "Epoch: 42 |  TrainLoss: 0.32 | TestLoss: 0.33 | TestAcc: 0.88 | TestF1: 0.86\n",
      "Epoch: 43 |  TrainLoss: 0.30 | TestLoss: 0.30 | TestAcc: 0.92 | TestF1: 0.92\n",
      "Epoch: 44 |  TrainLoss: 0.29 | TestLoss: 0.29 | TestAcc: 0.92 | TestF1: 0.92\n",
      "Epoch: 45 |  TrainLoss: 0.28 | TestLoss: 0.29 | TestAcc: 0.91 | TestF1: 0.91\n",
      "Epoch: 46 |  TrainLoss: 0.27 | TestLoss: 0.30 | TestAcc: 0.91 | TestF1: 0.91\n",
      "Epoch: 47 |  TrainLoss: 0.27 | TestLoss: 0.27 | TestAcc: 0.92 | TestF1: 0.91\n",
      "Epoch: 48 |  TrainLoss: 0.26 | TestLoss: 0.30 | TestAcc: 0.87 | TestF1: 0.86\n",
      "Epoch: 49 |  TrainLoss: 0.27 | TestLoss: 0.26 | TestAcc: 0.93 | TestF1: 0.93\n",
      "Epoch: 50 |  TrainLoss: 0.26 | TestLoss: 0.25 | TestAcc: 0.93 | TestF1: 0.93\n",
      "Epoch: 51 |  TrainLoss: 0.24 | TestLoss: 0.25 | TestAcc: 0.92 | TestF1: 0.92\n",
      "Epoch: 52 |  TrainLoss: 0.24 | TestLoss: 0.24 | TestAcc: 0.93 | TestF1: 0.93\n",
      "Epoch: 53 |  TrainLoss: 0.25 | TestLoss: 0.29 | TestAcc: 0.90 | TestF1: 0.91\n",
      "Epoch: 54 |  TrainLoss: 0.25 | TestLoss: 0.23 | TestAcc: 0.93 | TestF1: 0.93\n",
      "Epoch: 55 |  TrainLoss: 0.24 | TestLoss: 0.25 | TestAcc: 0.91 | TestF1: 0.91\n",
      "Epoch: 56 |  TrainLoss: 0.23 | TestLoss: 0.23 | TestAcc: 0.93 | TestF1: 0.93\n",
      "Epoch: 57 |  TrainLoss: 0.23 | TestLoss: 0.23 | TestAcc: 0.93 | TestF1: 0.93\n",
      "Epoch: 58 |  TrainLoss: 0.22 | TestLoss: 0.22 | TestAcc: 0.93 | TestF1: 0.93\n",
      "Epoch: 59 |  TrainLoss: 0.22 | TestLoss: 0.22 | TestAcc: 0.93 | TestF1: 0.93\n"
     ]
    }
   ],
   "source": [
    "fusion_model = GNNWithAttentionFusion(train_data.num_features, 128, 1).to(device)\n",
    "optimizer = torch.optim.Adam(fusion_model.parameters(), lr=0.0002, weight_decay=0.01)\n",
    "loss_fnc = torch.nn.BCELoss()\n",
    "\n",
    "trainer_attention_fusion = Trainer(\n",
    "    model=fusion_model, loss_func=loss_fnc, optimizer=optimizer, device=device\n",
    ")\n",
    "\n",
    "train_losses_fusion = []\n",
    "test_losses_fusion = []\n",
    "test_acces_fusion = []\n",
    "test_f1score_fusion = []\n",
    "\n",
    "for epoch in range(60):\n",
    "    train_loss = trainer_attention_fusion.train(train_loader)\n",
    "    test_loss, test_acc, test_f1 = trainer_attention_fusion.test(test_loader)\n",
    "\n",
    "    train_losses_fusion.append(train_loss)\n",
    "    test_losses_fusion.append(test_loss)\n",
    "    test_acces_fusion.append(test_acc)\n",
    "    test_f1score_fusion.append(test_f1)\n",
    "\n",
    "    print(\n",
    "        f\"Epoch: {epoch:02d} |  TrainLoss: {train_loss:.2f} | \"\n",
    "        f\"TestLoss: {test_loss:.2f} | TestAcc: {test_acc:.2f} | TestF1: {test_f1:.2f}\"\n",
    "    )"
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Results"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 282
    },
    "id": "O0GQzuPy2meW",
    "outputId": "e89543d1-a20d-4edb-d5da-04b4daced029"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.plot(test_losses_fusion, label=\"Attention fusion lr=2e-4\")\n",
    "plt.plot(test_losses_concat_lr_1e_2, label=\"Concatenation lr=1e-2\")\n",
    "plt.plot(test_losses_concat_lr_2e_4, label=\"Concatenation lr=2e-4\")\n",
    "plt.xlabel(\"Epochs\")\n",
    "plt.ylabel(\"Loss\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 282
    },
    "id": "N_R67kpCeukH",
    "outputId": "971553dc-ecfb-48f8-d9bc-720e8ab0d9cc"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(test_acces_fusion, label=\"Attention fusion lr=2e-4\")\n",
    "plt.plot(test_acces_concat_lr_1e_2, label=\"Concatenation lr=1e-2\")\n",
    "plt.plot(test_acces_concat_lr_2e_4, label=\"Concatenation lr=2e-4\")\n",
    "plt.xlabel(\"Epochs\")\n",
    "plt.ylabel(\"Accuracy\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(test_f1score_fusion, label=\"Attention fusion lr=2e-4\")\n",
    "plt.plot(test_f1score_concat_lr_1e_2, label=\"Concatenation lr=1e-2\")\n",
    "plt.plot(test_f1score_concat_lr_2e_4, label=\"Concatenation lr=2e-4\")\n",
    "plt.xlabel(\"Epochs\")\n",
    "plt.ylabel(\"F1 Score\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "# Visualization with T-SNE"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "id": "r5d-8CIQe6FK"
   },
   "outputs": [],
   "source": [
    "from sklearn.manifold import TSNE\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "# a parameter for dimension reduction\n",
    "n_components = 2"
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Concatenation (lr=0.0002)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 517
    },
    "id": "ybdBSFm0hRoM",
    "outputId": "99f9f4ab-9db9-478d-c7a6-15a19fffc036"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\leok8\\miniconda3\\envs\\net-sci\\lib\\site-packages\\sklearn\\manifold\\_t_sne.py:780: FutureWarning: The default initialization in TSNE will change from 'random' to 'pca' in 1.2.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# load the node embedding vector that has been trained. (34, 4) dimension\n",
    "node_emb = np.copy(model.emb.cpu().detach().numpy())\n",
    "\n",
    "# transform the node embedding vector size from (34, 4) into (34, 2) with a T-SNE model\n",
    "model_tsne = TSNE(n_components=n_components, learning_rate=100, perplexity=3)\n",
    "node_emb_2d = model_tsne.fit_transform(node_emb)\n",
    "\n",
    "# separate the x-axis values and y-axis values\n",
    "x = node_emb_2d[:, 0]\n",
    "y = node_emb_2d[:, 1]\n",
    "\n",
    "# set color options from the labels.\n",
    "# node_colours = labels[:, -1]\n",
    "\n",
    "# scatter plot\n",
    "plt.figure(figsize=(10, 8))\n",
    "plt.scatter(x, y, cmap=\"jet\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Concatenation (lr=0.01)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\leok8\\miniconda3\\envs\\net-sci\\lib\\site-packages\\sklearn\\manifold\\_t_sne.py:780: FutureWarning: The default initialization in TSNE will change from 'random' to 'pca' in 1.2.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# load the node embedding vector that has been trained. (34, 4) dimension\n",
    "node_emb = np.copy(model_2.emb.cpu().detach().numpy())\n",
    "\n",
    "# transform the node embedding vector size from (34, 4) into (34, 2) with a T-SNE model\n",
    "model_tsne = TSNE(n_components=n_components, learning_rate=100, perplexity=3)\n",
    "node_emb_2d = model_tsne.fit_transform(node_emb)\n",
    "\n",
    "# separate the x-axis values and y-axis values\n",
    "x = node_emb_2d[:, 0]\n",
    "y = node_emb_2d[:, 1]\n",
    "\n",
    "# set color options from the labels.\n",
    "# node_colours = labels[:, -1]\n",
    "\n",
    "# scatter plot\n",
    "plt.figure(figsize=(10, 8))\n",
    "plt.scatter(x, y, cmap=\"jet\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Attention fusion (lr=0.0002)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 517
    },
    "id": "F4g13AXUhlMh",
    "outputId": "4673ef2b-28ae-44dd-c70d-7a433d40ca44"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\leok8\\miniconda3\\envs\\net-sci\\lib\\site-packages\\sklearn\\manifold\\_t_sne.py:780: FutureWarning: The default initialization in TSNE will change from 'random' to 'pca' in 1.2.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# load the node embedding vector that has been trained. (34, 4) dimension\n",
    "node_emb = np.copy(fusion_model.emb.cpu().detach().numpy())\n",
    "\n",
    "# transform the node embedding vector size from (34, 4) into (34, 2) with a T-SNE model\n",
    "model_tsne = TSNE(n_components=n_components, learning_rate=100, perplexity=3)\n",
    "node_emb_2d = model_tsne.fit_transform(node_emb)\n",
    "\n",
    "# separate the x-axis values and y-axis values\n",
    "x = node_emb_2d[:, 0]\n",
    "y = node_emb_2d[:, 1]\n",
    "\n",
    "# set color options from the labels.\n",
    "# node_colours = labels[:, -1]\n",
    "\n",
    "# scatter plot\n",
    "plt.figure(figsize=(10, 8))\n",
    "plt.scatter(x, y, cmap=\"jet\")\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [],
   "name": "fake_news_detection.ipynb의 사본",
   "provenance": []
  },
  "gpuClass": "standard",
  "interpreter": {
   "hash": "061f8ce379c074f0236f397706a31c1eff37be2453e3164063365ec65c21205c"
  },
  "kernelspec": {
   "display_name": "Python 3.8.13 ('net-sci')",
   "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": "3.8.13"
  },
  "vscode": {
   "interpreter": {
    "hash": "061f8ce379c074f0236f397706a31c1eff37be2453e3164063365ec65c21205c"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}