GNN_LRP.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ssib7MOjNvXb",
        "outputId": "95802c1a-ca8a-4a05-9d8a-9c45e45191e7"
      },
      "source": [
        "# Install the igraph python library - see also: https://igraph.org/python/tutorial/latest/tutorial.html \n",
        "# This is a library that allows internal handling of the graph with functionalities supporting nodes, edges additions/removals,\n",
        "# adjacency matrix processing, plotting layouts and so on. Another even more known library is networkx: https://networkx.org/,\n",
        "# with similar functionality and also possibilites of direct conversion from its graph format to PyTorch format for further\n",
        "# GNN processing (although this is not always error-free).\n",
        "\n",
        "!pip install python-igraph"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting python-igraph\n",
            "  Downloading python-igraph-0.10.1.tar.gz (9.5 kB)\n",
            "Collecting igraph==0.10.1\n",
            "  Downloading igraph-0.10.1-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (3.2 MB)\n",
            "\u001b[K     |████████████████████████████████| 3.2 MB 6.9 MB/s \n",
            "\u001b[?25hCollecting texttable>=1.6.2\n",
            "  Downloading texttable-1.6.4-py2.py3-none-any.whl (10 kB)\n",
            "Building wheels for collected packages: python-igraph\n",
            "  Building wheel for python-igraph (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for python-igraph: filename=python_igraph-0.10.1-py3-none-any.whl size=9076 sha256=32c38f1533f4d30ecf6863d7a26bc4ebc1902042755dffc5f74ed3ba56605c91\n",
            "  Stored in directory: /root/.cache/pip/wheels/e2/79/c8/573c0f4e408a66ae133c1aa73c01824cfac9564b6e2b5288e3\n",
            "Successfully built python-igraph\n",
            "Installing collected packages: texttable, igraph, python-igraph\n",
            "Successfully installed igraph-0.10.1 python-igraph-0.10.1 texttable-1.6.4\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "The original paper of the GNN-LRP method is https://arxiv.org/abs/2006.03589"
      ],
      "metadata": {
        "id": "JstWMIDJHEz6"
      }
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "v4ldqw1sONq2"
      },
      "source": [
        "import igraph\n",
        "import numpy as np\n",
        "\n",
        "\n",
        "def relevance_curves_xy_computation(nodes_x_coors_list: list, nodes_y_coors_list: list):\n",
        "    \"\"\"\n",
        "    This method performs the computation of x,y coordinate points for the relevance curves (or equivalently\n",
        "    curved lines) inside the plot. The relevance curves are curves that depict the relevance of a walk,\n",
        "    comprised of 3 nodes (and 2 edges connecting them) by their colour. Since several nodes can be part of\n",
        "    many different walks, it was decided that curved lines, surrounding (\"around\") the graphs will show the \n",
        "    relevances of each walk in the graph (which is depicted in black) - so that there is no overlap.\n",
        "    This is a plot helper method and not a relevance computation method.\n",
        "\n",
        "    :param nodes_x_coors_array: List of the 3 x-coordinates of the nodes in the walk\n",
        "    :param nodes_y_coors_array: List of the 3 y-coordinates of the nodes in the walk\n",
        "    :return:\n",
        "    \"\"\"\n",
        "\n",
        "    relevance_line_curve_x_coors_list = np.array(nodes_x_coors_list)\n",
        "    relevance_line_curve_y_coors_list = np.array(nodes_y_coors_list)\n",
        "\n",
        "    if relevance_line_curve_x_coors_list[0] == relevance_line_curve_x_coors_list[1] == relevance_line_curve_x_coors_list[2] and \\\n",
        "            relevance_line_curve_y_coors_list[0] == relevance_line_curve_y_coors_list[1] == relevance_line_curve_y_coors_list[2]:\n",
        "        relevance_line_curve_x_coors_list = relevance_line_curve_x_coors_list[0] + 0.1 * np.cos(np.linspace(0, 2 * np.pi, 50))\n",
        "        relevance_line_curve_y_coors_list = relevance_line_curve_y_coors_list[0] + 0.1 * np.sin(np.linspace(0, 2 * np.pi, 50))\n",
        "    else:\n",
        "        relevance_line_curve_x_coors_list = 0.75 * relevance_line_curve_x_coors_list + 0.25 * relevance_line_curve_x_coors_list.mean()\n",
        "        relevance_line_curve_y_coors_list = 0.75 * relevance_line_curve_y_coors_list + 0.25 * relevance_line_curve_y_coors_list.mean()\n",
        "\n",
        "        relevance_line_curve_x_coors_list = np.concatenate([\n",
        "            np.linspace(relevance_line_curve_x_coors_list[0], relevance_line_curve_x_coors_list[0], 41),\n",
        "            np.linspace(relevance_line_curve_x_coors_list[0], relevance_line_curve_x_coors_list[1], 20),\n",
        "            np.linspace(relevance_line_curve_x_coors_list[1], relevance_line_curve_x_coors_list[2], 20),\n",
        "            np.linspace(relevance_line_curve_x_coors_list[2], relevance_line_curve_x_coors_list[2], 41), ])\n",
        "        relevance_line_curve_y_coors_list = np.concatenate([\n",
        "            np.linspace(relevance_line_curve_y_coors_list[0], relevance_line_curve_y_coors_list[0], 41),\n",
        "            np.linspace(relevance_line_curve_y_coors_list[0], relevance_line_curve_y_coors_list[1], 20),\n",
        "            np.linspace(relevance_line_curve_y_coors_list[1], relevance_line_curve_y_coors_list[2], 20),\n",
        "            np.linspace(relevance_line_curve_y_coors_list[2], relevance_line_curve_y_coors_list[2], 41)])\n",
        "\n",
        "        filt = np.exp(-np.linspace(-2, 2, 41) ** 2)\n",
        "        filt = filt / filt.sum()\n",
        "\n",
        "        relevance_line_curve_x_coors_list = np.convolve(relevance_line_curve_x_coors_list, filt, mode='valid')\n",
        "        relevance_line_curve_y_coors_list = np.convolve(relevance_line_curve_y_coors_list, filt, mode='valid')\n",
        "\n",
        "\n",
        "    return relevance_line_curve_x_coors_list, relevance_line_curve_y_coors_list\n",
        "\n",
        "def compute_walks(adj_matrix: np.ndarray):\n",
        "    \"\"\"\n",
        "    Compute all walks of length 3\n",
        "\n",
        "    :param adj_matrix: Adjacency matrix of graph\n",
        "    :return:\n",
        "    \"\"\"\n",
        "\n",
        "    w = []\n",
        "\n",
        "    for v1 in np.arange(len(adj_matrix)):\n",
        "        for v2 in np.where(adj_matrix[v1])[0]:\n",
        "            for v3 in np.where(adj_matrix[v2])[0]:\n",
        "\n",
        "                w += [(v1, v2, v3)]\n",
        "\n",
        "    return w\n",
        "\n",
        "\n",
        "def set_graph_layout(adj_matrix: np.ndarray, seed):\n",
        "    \"\"\"\n",
        "    Creates a graph and sets the graph layout according to a set of edges and nodes in adj_matrix\n",
        "\n",
        "    :param adj_matrix: Adjacency matrix\n",
        "    :param seed: Random seed of graph\n",
        "    :return:\n",
        "    \"\"\"\n",
        "\n",
        "    graph = igraph.Graph()\n",
        "    graph.add_vertices(len(adj_matrix))\n",
        "    graph.add_edges(zip(*np.where(adj_matrix == 1)))\n",
        "\n",
        "    return np.array(list(graph.layout_kamada_kawai()))"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "KLdYG5p8O06-",
        "outputId": "00d8d8f8-1ac6-4acc-d1fc-5b6b1a31e59b"
      },
      "source": [
        "import random\n",
        "import sys\n",
        "\n",
        "import igraph\n",
        "import matplotlib\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "import torch\n",
        "\n",
        "\n",
        "########################################################################################################################\n",
        "# [A.] Create example graphs and plots thereof =========================================================================\n",
        "########################################################################################################################\n",
        "def create_scale_free_graph(nodes_nr: int, input_seed: int = 0, embed: bool = False, growth=None):\n",
        "    \"\"\"\n",
        "    # Imlementation of the Barabasi-Albert model for the creation of scale-free graphs.\n",
        "    A scale-free graph has a degree distribution that follows a power law - asymptotically. \n",
        "    The fraction P(k) of nodes in the network having k connections to other nodes goes for large values of k:\n",
        "    P(k) ∼ pow(k, −γ)\n",
        "    In this method the γ (gamma) is sampled uniformly being either 1, 2, or 3.\n",
        "\n",
        "    :param nodes_nr: Number of graph nodes\n",
        "    :param input_seed: Seed for the random generation\n",
        "    :param embed: Parameter that will be used for visualization\n",
        "    :param growth: Growth parameter of the Barabasi-Albert (BA) method. The scale-free graph increases over time.\n",
        "    At each timestep a new node with \"growth\" number of links that connect the new node to \"growth\" number of nodes\n",
        "    already in the network is added.\n",
        "    :return:\n",
        "    \"\"\"\n",
        "\n",
        "    # [1.] Define number of nodes and the adjacency matrix -------------------------------------------------------------\n",
        "    adj_matrix = np.zeros([nodes_nr, nodes_nr])     # Adjacency Matrix\n",
        "    adj_matrix[1, 0] = 1\n",
        "    adj_matrix[0, 1] = 1\n",
        "\n",
        "    # The generated graph that increases over time. At each timestep a new node with \"growth\" number of links that -----\n",
        "    # connect the new node to \"growth\" number of nodes already in the network is added ---------------------------------\n",
        "    random = np.random.mtrand.RandomState(input_seed)\n",
        "    growth = growth if growth is not None else random.randint(1, 3)\n",
        "    N0 = 2\n",
        "\n",
        "    # [2.] Create the edges according to the growth parameter ----------------------------------------------------------\n",
        "    #      The number of nodes in the graph is specified here. What is not specified is the edges. \n",
        "    #      The code implementation defines the process of creating the scale-free graph from scratch.\n",
        "    #      Assuming that we are in a step of this process, and we already have a graph of some sort,\n",
        "    #      we can think of incrementing this graph by inserting a new node, that will be connected to \n",
        "    #      some other in the already existing graph. The decision of which node will the new one \"dock to\"\n",
        "    #      (have an edge with) is specified by the growth factor. It defines with what probability the \n",
        "    #      new node will be connected to the nodes the graph already has.\n",
        "    for i in range(N0, nodes_nr):\n",
        "        if growth == 1 : tt = 1     # Barabasi-Albert with growth 1 \n",
        "        elif growth == 2 : tt = 2   # Barabasi-Albert with growth 2\n",
        "        else:\n",
        "            tt = 1 + 1*((growth-1) > random.uniform(0, 1))\n",
        "\n",
        "        p = adj_matrix.sum(axis=0)/adj_matrix.sum()\n",
        "\n",
        "        for j in random.choice(nodes_nr, tt, p=p, replace=False):\n",
        "            adj_matrix[i, j] = 1\n",
        "            adj_matrix[j, i] = 1\n",
        "\n",
        "    r = random.permutation(len(adj_matrix))\n",
        "    adj_matrix = adj_matrix[r][:, r]*1.0\n",
        "\n",
        "    # [3.] Add Self-Connections in the Adjacency matrix ----------------------------------------------------------------\n",
        "    adj_matrix = adj_matrix + np.identity(len(adj_matrix))\n",
        "\n",
        "    # [4.] Compute Laplacian of the graph ------------------------------------------------------------------------------\n",
        "    D = adj_matrix.sum(axis=1)\n",
        "    laplacian_matrix = torch.FloatTensor(adj_matrix/(np.outer(D, D)**.5+1e-9))\n",
        "\n",
        "    # [5.] Compute the graph information data structure (dictionary) ---------------------------------------------------\n",
        "    #      The dataset is comprised by pairs of (graph_of_growth_factor, growth_factor) \n",
        "    #      the second one being the target, as one sees in the returned data structure\n",
        "    return {\n",
        "        'adjacency': torch.FloatTensor(adj_matrix),\n",
        "        'laplacian': laplacian_matrix,\n",
        "        'target': growth,\n",
        "        'layout': set_graph_layout(adj_matrix, input_seed) if embed else None,\n",
        "        'walks': compute_walks(adj_matrix)\n",
        "    }\n",
        "\n",
        "\n",
        "def visualize_graph(graph_data_structure: dict, ax):\n",
        "    \"\"\"\n",
        "    Function to visualise a graph\n",
        "\n",
        "    :param graph_data_structure: Graph information data structure (dictionary)\n",
        "    :param ax: Axis of plot\n",
        "    :return:\n",
        "    \"\"\"\n",
        "\n",
        "    # [1.] Arrange graph layout for plot -------------------------------------------------------------------------------\n",
        "    r = graph_data_structure['layout']\n",
        "    r = r - r.min(axis=0)\n",
        "    r = r / r.max(axis=0) * 2 - 1\n",
        "\n",
        "    # [2.] Plot the edges ----------------------------------------------------------------------------------------------\n",
        "    nodes_nr = len(graph_data_structure['adjacency'])\n",
        "    for i in np.arange(nodes_nr):\n",
        "        for j in np.arange(nodes_nr):\n",
        "            if graph_data_structure['adjacency'][i, j] > 0 and i != j: plt.plot([r[i, 0], r[j, 0]], [r[i, 1], r[j, 1]],\n",
        "                                                                                color='gray',\n",
        "                                                                                lw=0.5,\n",
        "                                                                                ls='dotted')\n",
        "\n",
        "    # [3.] Plot the nodes ----------------------------------------------------------------------------------------------\n",
        "    ax.plot(r[:, 0], r[:, 1], 'o', color='black', ms=3)\n",
        "\n",
        "\n",
        "# Plot some example graphs ---------------------------------------------------------------------------------------------\n",
        "sample_ids = [1, 3, 4, 5]\n",
        "nodes_nr = 10\n",
        "\n",
        "plt.figure(figsize=(3 * len(sample_ids), 3))\n",
        "for ids, seed in enumerate(sample_ids):\n",
        "\n",
        "    ax = plt.subplot(1, len(sample_ids), ids + 1)\n",
        "    example_graph = create_scale_free_graph(nodes_nr, input_seed=seed, embed=True)\n",
        "\n",
        "    visualize_graph(example_graph, ax=ax)\n",
        "    plt.subplots_adjust(left=0, right=1, bottom=0, top=1)\n",
        "\n",
        "    plt.axis('off')\n",
        "    plt.xlim(-1.2, 1.2)\n",
        "    plt.ylim(-1.2, 1.2)\n",
        "    ax.set_title(f\"growth={example_graph['target']}\")\n",
        "\n",
        "plt.show()\n",
        "plt.close()\n",
        "\n",
        "\n",
        "########################################################################################################################\n",
        "# [B.] Create GraphNet GNN ::: Define architecture, forward and lrp methods ============================================\n",
        "########################################################################################################################\n",
        "class GraphNet:\n",
        "    \"\"\"\n",
        "    Graph Net - The Graph Neural Network (GNN)\n",
        "\n",
        "    The adjacency matrix is not enough in general. Graph nodes do have features \n",
        "    (think of values representing size, color and so on) and edges also having \n",
        "    corresponding ones (distance, weight...). So only in this case, this very \n",
        "    simple graph where in principle only connectivity is used for the \n",
        "    classification, the adjacency matrix is enough.\n",
        "\n",
        "    \"\"\"\n",
        "\n",
        "    def __init__(self, input_layer_size: int, hidden_layer_size: int, output_layer_size: int):\n",
        "        \"\"\"\n",
        "        GNN init architecture\n",
        "\n",
        "        :param input_layer_size: Input layer size\n",
        "        :param hidden_layer_size: Hidden layer size\n",
        "        :param output_layer_size: Output layer size\n",
        "        \"\"\"\n",
        "\n",
        "        # [1.] Architecture of the GNN ---------------------------------------------------------------------------------\n",
        "        #      Weights/parameters of the GNN ---------------------------------------------------------------------------\n",
        "        #      U the weights/parameters of the input layer, V the weights parameters of the last layer -----------------\n",
        "        self.U = torch.nn.Parameter(torch.FloatTensor(np.random.normal(0, input_layer_size ** -.5,\n",
        "                                                                       [input_layer_size, hidden_layer_size])))\n",
        "        self.W1 = torch.nn.Parameter(torch.FloatTensor(np.random.normal(0, hidden_layer_size ** -.5,\n",
        "                                                                        [hidden_layer_size, hidden_layer_size])))\n",
        "        self.W2 = torch.nn.Parameter(torch.FloatTensor(np.random.normal(0, hidden_layer_size ** -.5,\n",
        "                                                                        [hidden_layer_size, hidden_layer_size])))\n",
        "        self.V = torch.nn.Parameter(torch.FloatTensor(np.random.normal(0, hidden_layer_size ** -.5,\n",
        "                                                                       [hidden_layer_size, output_layer_size])))\n",
        "\n",
        "        # [2.] Weights/parameters of the GNN ---------------------------------------------------------------------------\n",
        "        self.params = [self.U, self.W1, self.W2, self.V]\n",
        "\n",
        "    def forward_pass(self, adj_matrix: torch.Tensor):\n",
        "        \"\"\"\n",
        "        Forward pass of the GNN\n",
        "\n",
        "        :param adj_matrix: Adjacency matrix of the GNN\n",
        "        :return:\n",
        "        \"\"\"\n",
        "\n",
        "        H = torch.eye(len(adj_matrix))\n",
        "        H = H.matmul(self.U).clamp(min=0)\n",
        "        H = (adj_matrix.transpose(1, 0).matmul(H.matmul(self.W1))).clamp(min=0)\n",
        "        H = (adj_matrix.transpose(1, 0).matmul(H.matmul(self.W2))).clamp(min=0)\n",
        "        H = H.matmul(self.V).clamp(min=0)\n",
        "\n",
        "        return H.mean(dim=0)\n",
        "\n",
        "    def lrp_computation(self, adj_matrix: torch.Tensor, gamma: float, target: int, indexes: tuple):\n",
        "        \"\"\"\n",
        "        Compute the LRP relevance, as presented in the original paper: https://arxiv.org/abs/2006.03589\n",
        "\n",
        "        :param adj_matrix: Adjacency matrix of the graph\n",
        "        :param gamma: LRP gamma hyperparameter\n",
        "        :param target: Target class\n",
        "        :param indexes: Indexes of nodes in the graph\n",
        "        \"\"\"\n",
        "\n",
        "        # [1.] Create a column array which contains only one non-zero element ------------------------------------------\n",
        "        #      in the position of the node index contained in the walk -------------------------------------------------\n",
        "        if indexes is not None:\n",
        "\n",
        "            j, k = indexes\n",
        "            M_j = torch.FloatTensor(np.eye(len(adj_matrix))[j][:, np.newaxis])\n",
        "            M_k = torch.FloatTensor(np.eye(len(adj_matrix))[k][:, np.newaxis])\n",
        "\n",
        "        # [2.] Weights of the GNN (U, W1, W2, V) plus a weighted version of their positive part ------------------------\n",
        "        #      The ReLu of equation 9 in page 5, Table 1 of the original paper is implemented by the clamp() method ----\n",
        "        W1p = self.W1 + gamma * self.W1.clamp(min=0)\n",
        "        W2p = self.W2 + gamma * self.W2.clamp(min=0)\n",
        "        Vp = self.V + gamma * self.V.clamp(min=0)\n",
        "\n",
        "        # [3.] Define a tensor that will contain the results of the LRP ------------------------------------------------\n",
        "        #      Enable the automatic differentiation mechanism for this tensor ------------------------------------------\n",
        "        X = torch.eye(len(adj_matrix))\n",
        "        X.requires_grad_(True)\n",
        "\n",
        "        # [4.] Multiply the identity matrix (starting values of the LRP results tensor) --------------------------------\n",
        "        #      with the weights of the input layer ---------------------------------------------------------------------\n",
        "        H = X.matmul(self.U).clamp(min=0)                           # H_0\n",
        "\n",
        "        ################################################################################################################\n",
        "        # [5.] Perform the computation of equations for the 1st layer as described the original paper ------------------\n",
        "        # [5.1.] (Lambda * H_0) * W_1 = Z_1 * W_1                   # Page 5 of original paper -------------------------\n",
        "        # This is a component of the right part of the equation computing Q_t  -----------------------------------------\n",
        "        P = adj_matrix.transpose(1, 0).matmul(H.matmul(self.W1))\n",
        "\n",
        "        # [5.2.] (Lambda * H_0) * W_1^ = Z_1 * W_1^ = P_1           # Page 5 of original paper -------------------------\n",
        "        Pt = adj_matrix.transpose(1, 0).matmul(H.matmul(W1p))\n",
        "\n",
        "        # [5.3.] Q_1 = P_1 :::element_wise_mult::: [RelU(Z_1 * W_1) / P_t]cst. -----------------------------------------\n",
        "        Qt = (Pt * (P / (Pt + 1e-6)).data).clamp(min=0)\n",
        "\n",
        "        # [5.4.] H_1 = Q_1 :::element_wise_mult::: M_j + [Q_t]cst. :::element_wise_mult::: (1 - M_j) -------------------\n",
        "        #        Mutliplication with the column array which contains only one non-zero element in the position of the --\n",
        "        #        node index contained in the walk, as defined in step [1.] of this method ------------------------------\n",
        "        if indexes is not None: H = Qt * M_j + (1 - M_j) * (Qt.data)\n",
        "\n",
        "        ################################################################################################################\n",
        "        # [6.] Perform the computation of equations for the 2nd layer as described the original paper ------------------\n",
        "        #      Corresponds to the equations for the 1st layer ----------------------------------------------------------\n",
        "        P = adj_matrix.transpose(1, 0).matmul(H.matmul(self.W2))\n",
        "        Pt = adj_matrix.transpose(1, 0).matmul(H.matmul(W2p))\n",
        "        Qt = (Pt * (P / (Pt + 1e-6)).data).clamp(min=0)\n",
        "\n",
        "        if indexes is not None: H = Qt * M_k + (1 - M_k) * (Qt.data)\n",
        "\n",
        "        ################################################################################################################\n",
        "        # [7.] Perform the computation of equations for the 3d layer as described the original paper -------------------\n",
        "        #      Corresponds to the equations for the 1st and 2nd layer --------------------------------------------------\n",
        "        P = H.matmul(self.V)\n",
        "        Pt = H.matmul(Vp)\n",
        "        Qt = (Pt * (P / (Pt + 1e-6)).data).clamp(min=0)\n",
        "\n",
        "        ################################################################################################################\n",
        "        # [8.] Compute the mean w.r.t. the target class and apply the automatic differentiation ------------------------\n",
        "        Y = Qt.mean(dim=0)[target]\n",
        "        Y.backward()\n",
        "\n",
        "        return X.data * X.grad\n",
        "\n",
        "\n",
        "########################################################################################################################\n",
        "# [C.] Train GNN model =================================================================================================\n",
        "########################################################################################################################\n",
        "def train_gnn_model():\n",
        "    \"\"\"\n",
        "    Train the GNN model\n",
        "\n",
        "    :return:\n",
        "    \"\"\"\n",
        "\n",
        "    # [1.] Train the model with size of input layer: 10, size of hidden layer: 64, size of output layer: 2 -------------\n",
        "    hidden_layer_size = 64\n",
        "    output_layer_size = 2\n",
        "    gnn_model = GraphNet(nodes_nr, hidden_layer_size, output_layer_size)\n",
        "\n",
        "    optimizer = torch.optim.SGD(gnn_model.params, lr=0.001, momentum=0.99)  # Use SGD optimizer ------------------------\n",
        "    loss_avg = 0.5    # Running average of the MSE -----------------------------------------------------------------------\n",
        "\n",
        "    print('Training model:')\n",
        "    print('   iter | loss')\n",
        "    print('   -----------')\n",
        "\n",
        "    train_iterations = 20001\n",
        "\n",
        "    for iteration in range(0, train_iterations):\n",
        "        optimizer.zero_grad()\n",
        "\n",
        "        # [2.] Create a new scale-free graph ---------------------------------------------------------------------------\n",
        "        input_graph = create_scale_free_graph(nodes_nr, input_seed=iteration, embed=False)\n",
        "\n",
        "        # [3.] Forward the input graph's laplacian to the model --------------------------------------------------------\n",
        "        y = gnn_model.forward_pass(input_graph['laplacian'])\n",
        "\n",
        "        # [4.] Compute the error between the true value and the predicted one ------------------------------------------\n",
        "        loss = (y[0]-(input_graph['target'] == 1)*1.0)**2 + (y[1]-(input_graph['target'] == 2)*1.0)**2\n",
        "\n",
        "        #      This is a running average of the MSE --------------------------------------------------------------------\n",
        "        loss_avg = 0.999*loss_avg + 0.001*loss.data.numpy()\n",
        "\n",
        "        # [5.] Backpropagate a weighted sum of the average error and the currently computed error ----------------------\n",
        "        loss.backward()\n",
        "        optimizer.step()\n",
        "\n",
        "        # [6.] Print the average error at each iteration ---------------------------------------------------------------\n",
        "        if iteration % 1000 == 0:\n",
        "            print('% 8d %.3f' % (iteration, loss_avg))\n",
        "\n",
        "    return gnn_model\n",
        "\n",
        "\n",
        "model = train_gnn_model()\n",
        "\n",
        "########################################################################################################################\n",
        "# [D.] Test GNN model ==================================================================================================\n",
        "########################################################################################################################\n",
        "test_size = 200\n",
        "num_false = 0\n",
        "\n",
        "# Create new graphs for the test set -----------------------------------------------------------------------------------\n",
        "for it in range(20001, 20001 + test_size):\n",
        "\n",
        "    test_graph = create_scale_free_graph(nodes_nr, input_seed=it, embed=False)\n",
        "    y = model.forward_pass(test_graph['laplacian'])\n",
        "    prediction = int(y.data.argmax()) + 1\n",
        "\n",
        "    if prediction != test_graph['target'] : num_false += 1\n",
        "\n",
        "print('For {} test sample graphs, the model predicts the growth parameter with an accuracy of {} %'.\n",
        "      format(test_size, 100 * (test_size - num_false)/test_size))\n",
        "\n",
        "\n",
        "########################################################################################################################\n",
        "# [E.] Explain and plot relevances =====================================================================================\n",
        "########################################################################################################################\n",
        "def explain_graph_LRP(input_graph: dict, nn: GraphNet, target: int, gamma=None, ax=None):\n",
        "    \"\"\"\n",
        "    Explain a graph prediction through GNN-LRP\n",
        "\n",
        "    :param input_graph: Dictionary of input graph information\n",
        "    :param nn: GNN\n",
        "    :param target: target class\n",
        "    :param gamma: LRP gamma hyperparameter\n",
        "    :param ax: Plot axis\n",
        "    :return:\n",
        "    \"\"\"\n",
        "\n",
        "    # [1.] Plot the underlying graph in black --------------------------------------------------------------------------\n",
        "    r = input_graph['layout']\n",
        "    r = r - r.min(axis=0)\n",
        "    r = r / r.max(axis=0) * 2 - 1\n",
        "\n",
        "    N = len(input_graph['adjacency'])\n",
        "    for i in np.arange(N):\n",
        "        for j in np.arange(N):\n",
        "            if input_graph['adjacency'][i, j] > 0 and i != j: plt.plot([r[i, 0], r[j, 0]], [r[i, 1], r[j, 1]],\n",
        "                                                                       color='gray',\n",
        "                                                                       lw=0.5,\n",
        "                                                                       ls='dotted')\n",
        "    ax.plot(r[:, 0], r[:, 1], 'o', color='black', ms=3)\n",
        "\n",
        "    # [2.] For all walks of length 3 compute and plot the relevances ---------------------------------------------------\n",
        "    for (i, j, k) in input_graph['walks']:\n",
        "\n",
        "        R = nn.lrp_computation(input_graph['laplacian'], gamma, target, (j, k))[i].sum()\n",
        "        tx, ty = relevance_curves_xy_computation([r[i, 0], r[j, 0], r[k, 0]], [r[i, 1], r[j, 1], r[k, 1]])\n",
        "\n",
        "        # Positive (red) or negative (blue) relevance defines the explanation colorbar ----------------------------------------------------------\n",
        "        # Depending on the intensity of the relevance, define the transparency parameter alpha -------------------------\n",
        "        if R > 0.0:\n",
        "            alpha = np.clip(20 * R.data.numpy(), 0, 1)\n",
        "            ax.plot(tx, ty, alpha=alpha, color='red', lw=1.2)\n",
        "\n",
        "        if R < -0.0:\n",
        "            alpha = np.clip(-20 * R.data.numpy(), 0, 1)\n",
        "            ax.plot(tx, ty, alpha=alpha, color='blue', lw=1.2)\n",
        "\n",
        "\n",
        "########################################################################################################################\n",
        "# Generate new graph - new test sample ---------------------------------------------------------------------------------\n",
        "# Pass this example through the GNN and compute the relevance of the paths of this graph -------------------------------\n",
        "#\n",
        "# Red denotes positive relevance/contribution to the predicted class, even if it's a wrong prediction ------------------\n",
        "# Blue denotes negative relevance/contribution to the predicted class, even it it's a wrong prediction -----------------\n",
        "########################################################################################################################\n",
        "gamma = 0.1\n",
        "\n",
        "for target in [0, 1]:\n",
        "\n",
        "    plt.figure(figsize=(3 * len(sample_ids), 3))\n",
        "\n",
        "    for ids, seed in enumerate(sample_ids):\n",
        "        ax = plt.subplot(1, len(sample_ids), ids + 1)\n",
        "        example_graph = create_scale_free_graph(nodes_nr, input_seed=seed, embed=True)\n",
        "\n",
        "        # Explain with LRP ---------------------------------------------------------------------------------------------\n",
        "        explain_graph_LRP(example_graph, model, target, gamma=gamma, ax=ax)\n",
        "        plt.subplots_adjust(left=0, right=1, bottom=0, top=1)\n",
        "\n",
        "        plt.axis('off')\n",
        "        plt.xlim(-1.2, 1.2)\n",
        "        plt.ylim(-1.2, 1.2)\n",
        "\n",
        "    plt.suptitle(f\"Evidence for growth factor ={target + 1} with $\\gamma={gamma}$ (LRP relevances in red -positive- and blue -negative-)\", size=14)\n",
        "    plt.show()\n",
        "    plt.close()\n"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 864x216 with 4 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Training model:\n",
            "   iter | loss\n",
            "   -----------\n",
            "       0 0.500\n",
            "    1000 0.655\n",
            "    2000 0.417\n",
            "    3000 0.191\n",
            "    4000 0.082\n",
            "    5000 0.039\n",
            "    6000 0.020\n",
            "    7000 0.013\n",
            "    8000 0.009\n",
            "    9000 0.007\n",
            "   10000 0.006\n",
            "   11000 0.006\n",
            "   12000 0.006\n",
            "   13000 0.005\n",
            "   14000 0.005\n",
            "   15000 0.006\n",
            "   16000 0.006\n",
            "   17000 0.005\n",
            "   18000 0.005\n",
            "   19000 0.005\n",
            "   20000 0.005\n",
            "For 200 test sample graphs, the model predicts the growth parameter with an accuracy of 100.0 %\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 864x216 with 4 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 864x216 with 4 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "u33yPS8u1C2w"
      },
      "source": [],
      "execution_count": null,
      "outputs": []
    }
  ]
}