fake_news_detection.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "fake_news_detection.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "# Fraud Detection with Pytorch Geometric"
      ],
      "metadata": {
        "id": "jzJPAMwSY65V"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Installation"
      ],
      "metadata": {
        "id": "1Kcc3IMdZBwA"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import torch\n",
        "vers = torch.__version__\n",
        "print(\"Torch vers: \", vers)\n",
        "\n",
        "# PyG installation\n",
        "!pip install -q torch-scatter -f https://pytorch-geometric.com/whl/torch-${TORCH}+${CUDA}.html\n",
        "!pip install -q torch-sparse -f https://pytorch-geometric.com/whl/torch-${TORCH}+${CUDA}.html\n",
        "!pip install -q git+https://github.com/rusty1s/pytorch_geometric.git\n",
        "\n",
        "import torch_geometric"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gkg-eLyRT3Of",
        "outputId": "7efce579-f20d-4c77-ed05-dad38b0c2aa3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Torch vers:  1.10.0+cu111\n",
            "  Building wheel for torch-scatter (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "\u001b[K     |████████████████████████████████| 43 kB 905 kB/s \n",
            "\u001b[?25h  Building wheel for torch-sparse (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Building wheel for torch-geometric (setup.py) ... \u001b[?25l\u001b[?25hdone\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "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"
      ],
      "metadata": {
        "id": "ZyB3RRbmZD3o"
      }
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tvB33IQ5k7Tr",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "fc9b21fe-d287-457e-f49c-1afa2bcefc11"
      },
      "source": [
        "from torch_geometric.datasets import UPFD\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))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Samples:  1092\n",
            "Test Samples:  3826\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Investigating the News Propagation Graph"
      ],
      "metadata": {
        "id": "RdM8rXX2hqtw"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "sample_id=1\n",
        "train_data[sample_id].edge_index"
      ],
      "metadata": {
        "id": "5qz9MHtnTsvx",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "9cd5696d-46c1-4db6-d0b7-21c057353d7e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "tensor([[  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
              "           0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
              "           0,   0,   0,   0,   0,   0,   0,   0,   0,   0,  15,  15,  17,  17,\n",
              "          22,  22,  22,  22,  23,  26,  26,  26,  26,  26,  26,  26,  26,  27,\n",
              "          27,  27,  28,  28,  28,  28,  28,  28,  28,  28,  28,  28,  28,  28,\n",
              "          28,  28,  28,  28,  28,  28,  28,  28,  28,  28,  28,  28,  28,  28,\n",
              "          28,  28,  28,  28,  28,  28,  28,  28,  30,  34,  34,  38,  38,  38,\n",
              "          38,  39,  40,  57,  59,  65,  65,  65,  66,  68,  69,  77,  77,  79,\n",
              "          79,  80,  83,  83,  83,  83,  84, 101, 110, 115, 116, 117],\n",
              "        [  1,   2,   3,   4,   5,   6,   7,   8,   9,  10,  11,  12,  13,  14,\n",
              "          15,  16,  17,  18,  19,  20,  21,  22,  23,  24,  25,  26,  27,  28,\n",
              "          29,  30,  31,  32,  33,  34,  35,  36,  37,  38,  39,  40,  43,  44,\n",
              "          45,  46,  47,  48,  49,  50,  51,  52,  53,  54,  55,  56,  57,  59,\n",
              "          60,  61,  63,  64,  65,  66,  67,  68,  69,  70,  71,  72,  73,  74,\n",
              "          75,  76,  77,  78,  79,  80,  81,  82,  83,  84,  85,  86,  87,  88,\n",
              "          89,  90,  91,  92,  93,  94,  95,  96, 115, 119, 120, 121, 122, 123,\n",
              "         124,  41,  42,  58,  62,  97,  98,  99, 100, 101, 103, 104, 105, 106,\n",
              "         107, 108, 109, 110, 111, 112, 114, 102, 113, 116, 117, 118]])"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "\"\"\"\n",
        "Had to import this \"manually\" due to some errors.\n",
        "\"\"\"\n",
        "\n",
        "!pip install networkx\n",
        "import networkx as nx\n",
        "\n",
        "# From PyG utils\n",
        "def to_networkx(data, node_attrs=None, edge_attrs=None, to_undirected=False,\n",
        "                remove_self_loops=False):\n",
        "    if to_undirected:\n",
        "        G = nx.Graph()\n",
        "    else:\n",
        "        G = nx.DiGraph()\n",
        "    G.add_nodes_from(range(data.num_nodes))\n",
        "    node_attrs, edge_attrs = node_attrs or [], edge_attrs or []\n",
        "    values = {}\n",
        "    for key, item in data(*(node_attrs + edge_attrs)):\n",
        "        if torch.is_tensor(item):\n",
        "            values[key] = item.squeeze().tolist()\n",
        "        else:\n",
        "            values[key] = item\n",
        "        if isinstance(values[key], (list, tuple)) and len(values[key]) == 1:\n",
        "            values[key] = item[0]\n",
        "    for i, (u, v) in enumerate(data.edge_index.t().tolist()):\n",
        "        if to_undirected and v > u:\n",
        "            continue\n",
        "        if remove_self_loops and u == v:\n",
        "            continue\n",
        "        G.add_edge(u, v)\n",
        "        for key in edge_attrs:\n",
        "            G[u][v][key] = values[key][i]\n",
        "    for key in node_attrs:\n",
        "        for i, feat_dict in G.nodes(data=True):\n",
        "            feat_dict.update({key: values[key][i]})\n",
        "    return G"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7yqF8X6Hin7i",
        "outputId": "3e319015-0d1c-446f-896c-0a3bb6e89379"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Requirement already satisfied: networkx in /usr/local/lib/python3.7/dist-packages (2.6.3)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "nx.draw(to_networkx(train_data[sample_id]))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 319
        },
        "id": "smHrIIV1iQfJ",
        "outputId": "1d585033-2f97-4d40-ed5b-fa7dd0ffd9bf"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Node features"
      ],
      "metadata": {
        "id": "xU_P6S-Zht5A"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "print(train_data[sample_id].x.shape)\n",
        "train_data[sample_id].x"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jmZMuKXv6S0g",
        "outputId": "ca04ec32-64f7-4b7d-aeaf-23d18ba5bda8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "torch.Size([125, 310])\n"
          ]
        },
        {
          "output_type": "execute_result",
          "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]])"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Class distribution"
      ],
      "metadata": {
        "id": "cRzLDYg8Ec_4"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "labels = [data.y.item() for i, data in enumerate(train_data)]\n",
        "df = pd.DataFrame(labels, columns=[\"Labels\"])\n",
        "df[\"Labels\"].hist()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 282
        },
        "id": "Ul5y-1IKEcdU",
        "outputId": "0040120e-418c-4629-8501-a87df2ff8231"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fd1e948c9d0>"
            ]
          },
          "metadata": {},
          "execution_count": 7
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Data Loaders"
      ],
      "metadata": {
        "id": "vzAmnPlhl5iO"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from torch_geometric.loader import DataLoader\n",
        "train_loader = DataLoader(train_data, batch_size=128, shuffle=True)\n",
        "test_loader = DataLoader(test_data, batch_size=128, shuffle=False)"
      ],
      "metadata": {
        "id": "c0DhfbZgl44G"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Model and Training\n",
        "\n",
        "--> Because it is a directed graph, it will only share information from the root"
      ],
      "metadata": {
        "id": "C8m-dJ-WZFM1"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from torch_geometric.nn import global_max_pool as gmp\n",
        "from torch_geometric.nn import GATConv\n",
        "from torch.nn import Linear\n",
        "\n",
        "\n",
        "class GNN(torch.nn.Module):\n",
        "    def __init__(self, in_channels, hidden_channels, out_channels):\n",
        "        super().__init__()\n",
        "        \n",
        "        # Graph Convolutions\n",
        "        self.conv1 = GATConv(in_channels, hidden_channels)\n",
        "        self.conv2 = GATConv(hidden_channels, hidden_channels)\n",
        "        self.conv3 = GATConv(hidden_channels, hidden_channels)\n",
        "\n",
        "        # Readout\n",
        "        self.lin_news = Linear(in_channels, hidden_channels)\n",
        "        self.lin0 = Linear(hidden_channels, hidden_channels)\n",
        "        self.lin1 = Linear(2*hidden_channels, out_channels)\n",
        "\n",
        "    def forward(self, x, edge_index, batch):\n",
        "        # Graph Convolutions\n",
        "        h = self.conv1(x, edge_index).relu()\n",
        "        h = self.conv2(h, edge_index).relu()\n",
        "        h = self.conv3(h, edge_index).relu()\n",
        "\n",
        "        # Pooling\n",
        "        h = gmp(h, batch)\n",
        "\n",
        "        # Readout\n",
        "        h = self.lin0(h).relu()\n",
        "\n",
        "        # According to UPFD paper: Include raw word2vec embeddings of news \n",
        "        # This is done per graph in the batch\n",
        "        root = (batch[1:] - batch[:-1]).nonzero(as_tuple=False).view(-1)\n",
        "        root = torch.cat([root.new_zeros(1), root + 1], dim=0)\n",
        "        # root is e.g. [   0,   14,   94,  171,  230,  302, ... ]\n",
        "        news = x[root]\n",
        "        news = self.lin_news(news).relu()\n",
        "        \n",
        "        out = self.lin1(torch.cat([h, news], dim=-1))\n",
        "        return torch.sigmoid(out)\n",
        "\n",
        "GNN(train_data.num_features, 128, 1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "v_W27XGUlrpM",
        "outputId": "05f11a67-46c3-4e8d-ecd2-949e7f85218b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "GNN(\n",
              "  (conv1): GATConv(310, 128, heads=1)\n",
              "  (conv2): GATConv(128, 128, heads=1)\n",
              "  (conv3): GATConv(128, 128, heads=1)\n",
              "  (lin_news): Linear(in_features=310, out_features=128, bias=True)\n",
              "  (lin0): Linear(in_features=128, out_features=128, bias=True)\n",
              "  (lin1): Linear(in_features=256, out_features=1, bias=True)\n",
              ")"
            ]
          },
          "metadata": {},
          "execution_count": 36
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import accuracy_score, f1_score\n",
        "\n",
        "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
        "model = GNN(train_data.num_features, 128, 1).to(device)\n",
        "optimizer = torch.optim.Adam(model.parameters(), lr=0.01, weight_decay=0.01)\n",
        "loss_fnc = torch.nn.BCELoss()\n",
        "\n",
        "def train(epoch):\n",
        "    model.train()\n",
        "    total_loss = 0\n",
        "    for data in train_loader:\n",
        "        data = data.to(device)\n",
        "        optimizer.zero_grad()\n",
        "        out = model(data.x, data.edge_index, data.batch)\n",
        "        loss = loss_fnc(torch.reshape(out, (-1,)), data.y.float())\n",
        "        loss.backward()\n",
        "        optimizer.step()\n",
        "        total_loss += float(loss) * data.num_graphs\n",
        "    return total_loss / len(train_loader.dataset)\n",
        "\n",
        "@torch.no_grad()\n",
        "def test(epoch):\n",
        "    model.eval()\n",
        "    total_loss = 0\n",
        "    all_preds = []\n",
        "    all_labels = []\n",
        "    for data in test_loader:\n",
        "        data = data.to(device)\n",
        "        out = model(data.x, data.edge_index, data.batch)\n",
        "        loss = loss_fnc(torch.reshape(out, (-1,)), data.y.float())\n",
        "        total_loss += float(loss) * data.num_graphs\n",
        "        all_preds.append(torch.reshape(out, (-1,)))\n",
        "        all_labels.append(data.y.float())\n",
        "\n",
        "    # Calculate Metrics\n",
        "    accuracy, f1 = metrics(all_preds, all_labels)\n",
        "\n",
        "    return total_loss / len(test_loader.dataset), accuracy, f1\n",
        "\n",
        "\n",
        "def metrics(preds, gts):\n",
        "    preds = torch.round(torch.cat(preds))\n",
        "    gts = torch.cat(gts)\n",
        "    acc = accuracy_score(preds, gts)\n",
        "    f1 = f1_score(preds, gts)\n",
        "    return acc, f1"
      ],
      "metadata": {
        "id": "gKHbOXHJMUwd"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "for epoch in range(40):\n",
        "    train_loss = train(epoch)\n",
        "    test_loss, test_acc, test_f1 = test(epoch)\n",
        "    print(f'Epoch: {epoch:02d} |  TrainLoss: {train_loss:.2f} | '\n",
        "          f'TestLoss: {test_loss:.2f} | TestAcc: {test_acc:.2f} | TestF1: {test_f1:.2f}')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vxYpZCPpMaQh",
        "outputId": "9b94f5a4-540c-4013-83ad-f5d8c0b8b53e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch: 00 |  TrainLoss: 0.77 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
            "Epoch: 01 |  TrainLoss: 0.70 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.67\n",
            "Epoch: 02 |  TrainLoss: 0.70 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.01\n",
            "Epoch: 03 |  TrainLoss: 0.70 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
            "Epoch: 04 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.67\n",
            "Epoch: 05 |  TrainLoss: 0.70 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.67\n",
            "Epoch: 06 |  TrainLoss: 0.69 | TestLoss: 0.72 | TestAcc: 0.50 | TestF1: 0.00\n",
            "Epoch: 07 |  TrainLoss: 0.71 | TestLoss: 0.70 | TestAcc: 0.50 | TestF1: 0.00\n",
            "Epoch: 08 |  TrainLoss: 0.69 | TestLoss: 0.70 | TestAcc: 0.50 | TestF1: 0.67\n",
            "Epoch: 09 |  TrainLoss: 0.70 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
            "Epoch: 10 |  TrainLoss: 0.70 | TestLoss: 0.80 | TestAcc: 0.50 | TestF1: 0.67\n",
            "Epoch: 11 |  TrainLoss: 0.72 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.67\n",
            "Epoch: 12 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
            "Epoch: 13 |  TrainLoss: 0.69 | TestLoss: 0.73 | TestAcc: 0.50 | TestF1: 0.00\n",
            "Epoch: 14 |  TrainLoss: 0.69 | TestLoss: 0.68 | TestAcc: 0.54 | TestF1: 0.68\n",
            "Epoch: 15 |  TrainLoss: 0.68 | TestLoss: 0.68 | TestAcc: 0.50 | TestF1: 0.01\n",
            "Epoch: 16 |  TrainLoss: 0.68 | TestLoss: 0.67 | TestAcc: 0.68 | TestF1: 0.54\n",
            "Epoch: 17 |  TrainLoss: 0.70 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.00\n",
            "Epoch: 18 |  TrainLoss: 0.69 | TestLoss: 0.68 | TestAcc: 0.81 | TestF1: 0.84\n",
            "Epoch: 19 |  TrainLoss: 0.68 | TestLoss: 0.68 | TestAcc: 0.50 | TestF1: 0.67\n",
            "Epoch: 20 |  TrainLoss: 0.69 | TestLoss: 0.69 | TestAcc: 0.50 | TestF1: 0.67\n",
            "Epoch: 21 |  TrainLoss: 0.67 | TestLoss: 0.66 | TestAcc: 0.50 | TestF1: 0.67\n",
            "Epoch: 22 |  TrainLoss: 0.67 | TestLoss: 0.65 | TestAcc: 0.90 | TestF1: 0.89\n",
            "Epoch: 23 |  TrainLoss: 0.65 | TestLoss: 0.64 | TestAcc: 0.59 | TestF1: 0.31\n",
            "Epoch: 24 |  TrainLoss: 0.64 | TestLoss: 0.62 | TestAcc: 0.68 | TestF1: 0.53\n",
            "Epoch: 25 |  TrainLoss: 0.62 | TestLoss: 0.67 | TestAcc: 0.50 | TestF1: 0.02\n",
            "Epoch: 26 |  TrainLoss: 0.62 | TestLoss: 0.60 | TestAcc: 0.62 | TestF1: 0.39\n",
            "Epoch: 27 |  TrainLoss: 0.58 | TestLoss: 0.57 | TestAcc: 0.72 | TestF1: 0.78\n",
            "Epoch: 28 |  TrainLoss: 0.56 | TestLoss: 0.58 | TestAcc: 0.60 | TestF1: 0.33\n",
            "Epoch: 29 |  TrainLoss: 0.55 | TestLoss: 0.54 | TestAcc: 0.68 | TestF1: 0.54\n",
            "Epoch: 30 |  TrainLoss: 0.52 | TestLoss: 0.50 | TestAcc: 0.88 | TestF1: 0.86\n",
            "Epoch: 31 |  TrainLoss: 0.51 | TestLoss: 0.49 | TestAcc: 0.78 | TestF1: 0.73\n",
            "Epoch: 32 |  TrainLoss: 0.48 | TestLoss: 0.47 | TestAcc: 0.87 | TestF1: 0.88\n",
            "Epoch: 33 |  TrainLoss: 0.46 | TestLoss: 0.45 | TestAcc: 0.88 | TestF1: 0.89\n",
            "Epoch: 34 |  TrainLoss: 0.44 | TestLoss: 0.43 | TestAcc: 0.92 | TestF1: 0.92\n",
            "Epoch: 35 |  TrainLoss: 0.43 | TestLoss: 0.42 | TestAcc: 0.90 | TestF1: 0.91\n",
            "Epoch: 36 |  TrainLoss: 0.41 | TestLoss: 0.43 | TestAcc: 0.80 | TestF1: 0.76\n",
            "Epoch: 37 |  TrainLoss: 0.41 | TestLoss: 0.38 | TestAcc: 0.93 | TestF1: 0.93\n",
            "Epoch: 38 |  TrainLoss: 0.42 | TestLoss: 0.43 | TestAcc: 0.82 | TestF1: 0.84\n",
            "Epoch: 39 |  TrainLoss: 0.43 | TestLoss: 0.36 | TestAcc: 0.92 | TestF1: 0.92\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "for data in test_loader:\n",
        "    data = data.to(device)\n",
        "    pred = model(data.x, data.edge_index, data.batch)\n",
        "    df = pd.DataFrame()\n",
        "    df[\"pred_logit\"] = pred.detach().numpy()[:,0]\n",
        "    df[\"pred\"] = torch.round(pred).detach().numpy()[:,0]\n",
        "    df[\"true\"] = data.y.numpy()\n",
        "    print(df.head(10))\n",
        "    break"
      ],
      "metadata": {
        "id": "W_5IZcgCFyav",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "8023a73f-3ecf-494d-a7a7-b023b1db182e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "   pred_logit  pred  true\n",
            "0    0.712935   1.0     1\n",
            "1    0.546704   1.0     1\n",
            "2    0.337851   0.0     0\n",
            "3    0.572155   1.0     1\n",
            "4    0.143253   0.0     0\n",
            "5    0.337286   0.0     0\n",
            "6    0.283256   0.0     0\n",
            "7    0.805957   1.0     1\n",
            "8    0.147197   0.0     0\n",
            "9    0.272142   0.0     0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        ""
      ],
      "metadata": {
        "id": "2rKHflkhUpaT"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}