demos/node-classification/sgc-node-classification.ipynb from stellargraph/stellargraph

demos/node-classification/sgc-node-classification.ipynb
Summary

Maintainability

Test Coverage

Issues
{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "0",
   "metadata": {},
   "source": [
    "# Node classification with Simplified Graph Convolutions (SGC)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1",
   "metadata": {
    "nbsphinx": "hidden",
    "tags": [
     "CloudRunner"
    ]
   },
   "source": [
    "<table><tr><td>Run the latest release of this notebook:</td><td><a href=\"https://mybinder.org/v2/gh/stellargraph/stellargraph/master?urlpath=lab/tree/demos/node-classification/sgc-node-classification.ipynb\" alt=\"Open In Binder\" target=\"_parent\"><img src=\"https://mybinder.org/badge_logo.svg\"/></a></td><td><a href=\"https://colab.research.google.com/github/stellargraph/stellargraph/blob/master/demos/node-classification/sgc-node-classification.ipynb\" alt=\"Open In Colab\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\"/></a></td></tr></table>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2",
   "metadata": {},
   "source": [
    "This notebook demonstrates the use of `StellarGraph`'s GCN [[2]](#refs), class for training the simplified graph convolution (SGC) model in introduced in [[1]](#refs).\n",
    "\n",
    "We show how to use `StellarGraph` to perform node attribute inference on the Cora citation network using SGC by creating a single layer GCN model with softmax activation.\n",
    "\n",
    "SGC simplifies GCN in the following ways,\n",
    "\n",
    " - It removes the non-linearities in the graph convolutional layers.\n",
    " - It smooths the node input features using powers of the normalized adjacency matrix with self loops (see [[2]](#refs)).\n",
    " - It uses a single softmax layer such that GCN is simplified to logistic regression on smoothed node features.\n",
    "\n",
    "For a graph with $N$ nodes, $F$-dimensional node features and $C$ number of classes, SGC simplifies GCN using the following logistic regression classifier with smoothed features,\n",
    "\n",
    "$\\hat{\\boldsymbol{Y}}_{SGC} = \\mathtt{softmax}(\\boldsymbol{S}^K \\boldsymbol{X}\\; \\boldsymbol{\\Theta})$\n",
    "\n",
    "where $\\hat{\\boldsymbol{Y}}_{SGC} \\in \\mathbb{R}^{N\\times C}$ are the class predictions; $\\boldsymbol{S}^K \\in \\mathbb{R}^{N\\times N}$ is the normalised graph adjacency matrix with self loops raised to the K-th power; $\\boldsymbol{X}\\in \\mathbb{R}^{N\\times F}$ are the node input features; and $\\boldsymbol{\\Theta} \\in \\mathbb{R}^{F\\times C}$ are the classifier's parameters to be learned.\n",
    "\n",
    "<a name=\"refs\"></a>\n",
    "**References**\n",
    "\n",
    "[1] Simplifying Graph Convolutional Networks. F. Wu, T. Zhang, A. H. de Souza Jr., C. Fifty, T. Yu, and K. Q. Weinberger, arXiv: 1902.07153. [link](https://arxiv.org/abs/1902.07153)\n",
    "\n",
    "[2] Semi-Supervised Classification with Graph Convolutional Networks. T. N. Kipf and M. Welling, ICLR 2016. [link](https://arxiv.org/abs/1609.02907)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "3",
   "metadata": {
    "nbsphinx": "hidden",
    "tags": [
     "CloudRunner"
    ]
   },
   "outputs": [],
   "source": [
    "# install StellarGraph if running on Google Colab\n",
    "import sys\n",
    "if 'google.colab' in sys.modules:\n",
    "  %pip install -q stellargraph[demos]==1.3.0b"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "4",
   "metadata": {
    "nbsphinx": "hidden",
    "tags": [
     "VersionCheck"
    ]
   },
   "outputs": [],
   "source": [
    "# verify that we're using the correct version of StellarGraph for this notebook\n",
    "import stellargraph as sg\n",
    "\n",
    "try:\n",
    "    sg.utils.validate_notebook_version(\"1.3.0b\")\n",
    "except AttributeError:\n",
    "    raise ValueError(\n",
    "        f\"This notebook requires StellarGraph version 1.3.0b, but a different version {sg.__version__} is installed.  Please see <https://github.com/stellargraph/stellargraph/issues/1172>.\"\n",
    "    ) from None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "5",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import os\n",
    "\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.manifold import TSNE\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "import stellargraph as sg\n",
    "from stellargraph.mapper import FullBatchNodeGenerator\n",
    "from stellargraph.layer import GCN\n",
    "\n",
    "from tensorflow.keras import layers, optimizers, losses, metrics, Model, regularizers\n",
    "from sklearn import preprocessing, feature_extraction, model_selection\n",
    "from stellargraph import datasets\n",
    "from IPython.display import display, HTML\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6",
   "metadata": {},
   "source": [
    "## Loading the CORA network\n",
    "\n",
    "Create graph using the `Cora` loader from `datasets`."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7",
   "metadata": {
    "tags": [
     "DataLoadingLinks"
    ]
   },
   "source": [
    "(See [the \"Loading from Pandas\" demo](../basics/loading-pandas.ipynb) for details on how data can be loaded.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "8",
   "metadata": {
    "tags": [
     "DataLoading"
    ]
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "The Cora dataset consists of 2708 scientific publications classified into one of seven classes. The citation network consists of 5429 links. Each publication in the dataset is described by a 0/1-valued word vector indicating the absence/presence of the corresponding word from the dictionary. The dictionary consists of 1433 unique words."
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dataset = datasets.Cora()\n",
    "display(HTML(dataset.description))\n",
    "G, node_subjects = dataset.load()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "StellarGraph: Undirected multigraph\n",
      " Nodes: 2708, Edges: 5429\n",
      "\n",
      " Node types:\n",
      "  paper: [2708]\n",
      "    Edge types: paper-cites->paper\n",
      "\n",
      " Edge types:\n",
      "    paper-cites->paper: [5429]\n"
     ]
    }
   ],
   "source": [
    "print(G.info())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "10",
   "metadata": {},
   "source": [
    "We aim to train a graph-ML model that will predict the \"subject\" attribute on the nodes. These subjects are one of 7 categories:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "11",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'Case_Based',\n",
       " 'Genetic_Algorithms',\n",
       " 'Neural_Networks',\n",
       " 'Probabilistic_Methods',\n",
       " 'Reinforcement_Learning',\n",
       " 'Rule_Learning',\n",
       " 'Theory'}"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "set(node_subjects)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "12",
   "metadata": {},
   "source": [
    "### Splitting the data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "13",
   "metadata": {},
   "source": [
    "For machine learning we want to take a subset of the nodes for training, and use the rest for validation and testing. We'll use scikit-learn again to do this.\n",
    "\n",
    "Here we're taking 140 node labels for training, 500 for validation, and the rest for testing."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "14",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_subjects, test_subjects = model_selection.train_test_split(\n",
    "    node_subjects, train_size=140, test_size=None, stratify=node_subjects\n",
    ")\n",
    "val_subjects, test_subjects = model_selection.train_test_split(\n",
    "    test_subjects, train_size=500, test_size=None, stratify=test_subjects\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "15",
   "metadata": {},
   "source": [
    "Note using stratified sampling gives the following counts:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "16",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Counter({'Theory': 18,\n",
       "         'Neural_Networks': 42,\n",
       "         'Genetic_Algorithms': 22,\n",
       "         'Probabilistic_Methods': 22,\n",
       "         'Case_Based': 16,\n",
       "         'Rule_Learning': 9,\n",
       "         'Reinforcement_Learning': 11})"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from collections import Counter\n",
    "\n",
    "Counter(train_subjects)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "17",
   "metadata": {},
   "source": [
    "The training set has class imbalance that might need to be compensated, e.g., via using a weighted cross-entropy loss in model training, with class weights inversely proportional to class support. However, we will ignore the class imbalance in this example, for simplicity."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "18",
   "metadata": {},
   "source": [
    "### Converting to numeric arrays"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "19",
   "metadata": {},
   "source": [
    "For our categorical target, we will use one-hot vectors that will be fed into a soft-max Keras layer during training. To do this conversion ..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "20",
   "metadata": {},
   "outputs": [],
   "source": [
    "target_encoding = preprocessing.LabelBinarizer()\n",
    "\n",
    "train_targets = target_encoding.fit_transform(train_subjects)\n",
    "val_targets = target_encoding.transform(val_subjects)\n",
    "test_targets = target_encoding.transform(test_subjects)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "21",
   "metadata": {},
   "source": [
    "We now do the same for the node attributes we want to use to predict the subject. These are the feature vectors that the Keras model will use as input. The CORA dataset contains attributes 'w_x' that correspond to words found in that publication. If a word occurs more than once in a publication the relevant attribute will be set to one, otherwise it will be zero."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "22",
   "metadata": {},
   "source": [
    "## Prepare node generator"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "23",
   "metadata": {},
   "source": [
    "To feed data from the graph to the Keras model we need a generator. Since SGC is a full-batch model, we use the `FullBatchNodeGenerator` class to feed node features and graph adjacency matrix to the model. \n",
    "\n",
    "For SGC, we need to tell the generator to smooth the node features by some power of the normalised adjacency matrix with self loops before multiplying by the model parameters.\n",
    "\n",
    "We achieve this by specifying `model='sgc'` and `k=2`, in this example, to use the SGC method and take the square of the adjacency matrix. For the setting `k=2` we are considering a 2-hop neighbourhood that is equivalent to a 2-layer GCN. We can set `k` larger to consider larger node neighbourhoods but this carries an associated computational penalty."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "24",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Calculating 2-th power of normalized A...\n"
     ]
    }
   ],
   "source": [
    "generator = FullBatchNodeGenerator(G, method=\"sgc\", k=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "25",
   "metadata": {},
   "source": [
    "For training we map only the training nodes returned from our splitter and the target values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "26",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_gen = generator.flow(train_subjects.index, train_targets)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "27",
   "metadata": {},
   "source": [
    "## Creating the SGC model in Keras"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "28",
   "metadata": {},
   "source": [
    "Now we can specify our machine learning model, we need a few more parameters for this:\n",
    "\n",
    " * the `layer_sizes` is a list of hidden feature sizes of each layer in the model. For SGC, we use a single hidden layer with output dimensionality equal to the number of classes.\n",
    " * `activations` is the activation function for the output layer. For SGC the output layer is the classification layer and for multi-class classification it should be a `softmax` activation. \n",
    " * Arguments such as `bias` and `dropout` are internal parameters of the model, execute `?GCN` for details. \n",
    " \n",
    "**Note:** The SGC model is a single layer GCN model with `softmax` activation and the full batch generator we created above that smooths the node features based on the graph structure. So, our SGC model is declared as a `StellarGraph.layer.GCN` model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "29",
   "metadata": {},
   "outputs": [],
   "source": [
    "sgc = GCN(\n",
    "    layer_sizes=[train_targets.shape[1]],\n",
    "    generator=generator,\n",
    "    bias=True,\n",
    "    dropout=0.5,\n",
    "    activations=[\"softmax\"],\n",
    "    kernel_regularizer=regularizers.l2(5e-4),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "30",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Expose the input and output tensors of the SGC model for node prediction,\n",
    "# via GCN.in_out_tensors() method:\n",
    "x_inp, predictions = sgc.in_out_tensors()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "31",
   "metadata": {},
   "source": [
    "### Training the model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "32",
   "metadata": {},
   "source": [
    "Now let's create the actual Keras model with the input tensors `x_inp` and output tensors being the predictions `predictions` from the final dense layer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "33",
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Model(inputs=x_inp, outputs=predictions)\n",
    "model.compile(\n",
    "    optimizer=optimizers.Adam(lr=0.2),\n",
    "    loss=losses.categorical_crossentropy,\n",
    "    metrics=[\"acc\"],\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "34",
   "metadata": {},
   "source": [
    "Train the model, keeping track of its loss and accuracy on the training set, and its generalisation performance on the validation set (we need to create another generator over the validation data for this)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "35",
   "metadata": {},
   "outputs": [],
   "source": [
    "val_gen = generator.flow(val_subjects.index, val_targets)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "36",
   "metadata": {},
   "source": [
    "Create callbacks for early stopping (if validation accuracy stops improving) and best model checkpoint saving:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "37",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.callbacks import EarlyStopping, ModelCheckpoint\n",
    "\n",
    "if not os.path.isdir(\"logs\"):\n",
    "    os.makedirs(\"logs\")\n",
    "es_callback = EarlyStopping(\n",
    "    monitor=\"val_acc\", patience=50\n",
    ")  # patience is the number of epochs to wait before early stopping in case of no further improvement\n",
    "mc_callback = ModelCheckpoint(\n",
    "    \"logs/best_model.h5\", monitor=\"val_acc\", save_best_only=True, save_weights_only=True\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "38",
   "metadata": {},
   "source": [
    "Train the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "39",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": []
    }
   ],
   "source": [
    "history = model.fit(\n",
    "    train_gen,\n",
    "    epochs=50,\n",
    "    validation_data=val_gen,\n",
    "    verbose=0,\n",
    "    shuffle=False,  # this should be False, since shuffling data means shuffling the whole graph\n",
    "    callbacks=[es_callback, mc_callback],\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "40",
   "metadata": {},
   "source": [
    "Plot the training history:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "41",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sg.utils.plot_history(history)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "42",
   "metadata": {},
   "source": [
    "Reload the saved weights of the best model found during the training (according to validation accuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "43",
   "metadata": {},
   "outputs": [],
   "source": [
    "model.load_weights(\"logs/best_model.h5\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "44",
   "metadata": {},
   "source": [
    "Evaluate the best model on the test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "45",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_gen = generator.flow(test_subjects.index, test_targets)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "46",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Test Set Metrics:\n",
      "\tloss: 0.8700\n",
      "\tacc: 0.7863\n"
     ]
    }
   ],
   "source": [
    "test_metrics = model.evaluate(test_gen)\n",
    "print(\"\\nTest Set Metrics:\")\n",
    "for name, val in zip(model.metrics_names, test_metrics):\n",
    "    print(\"\\t{}: {:0.4f}\".format(name, val))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "47",
   "metadata": {},
   "source": [
    "### Making predictions with the model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "48",
   "metadata": {},
   "source": [
    "Now let's get the predictions for all nodes:\n",
    "\n",
    "Note that the `predict` function now operates differently to the `GraphSAGE` or `HinSAGE` models\n",
    "in that if you give it a set of nodes, it will still return predictions for **all** nodes in the graph, and in a fixed order defined by the order of nodes in `X` and `A` (which is defined by the order of `G.nodes()`)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "49",
   "metadata": {},
   "outputs": [],
   "source": [
    "all_nodes = node_subjects.index\n",
    "all_gen = generator.flow(all_nodes)\n",
    "all_predictions = model.predict(all_gen)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "50",
   "metadata": {},
   "source": [
    "Note that for full-batch methods the batch size is 1 and the predictions have shape $(1, N_{nodes}, N_{classes})$ so we we remove the batch dimension to obtain predictions of shape $(N_{nodes}, N_{classes})$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "51",
   "metadata": {},
   "outputs": [],
   "source": [
    "all_predictions = all_predictions.squeeze()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "52",
   "metadata": {},
   "source": [
    "These predictions will be the output of the softmax layer, so to get final categories we'll use the `inverse_transform` method of our target attribute specification to turn these values back to the original categories"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "53",
   "metadata": {},
   "outputs": [],
   "source": [
    "node_predictions = target_encoding.inverse_transform(all_predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "54",
   "metadata": {},
   "source": [
    "Let's have a look at a few:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "55",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Predicted</th>\n",
       "      <th>True</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>31336</td>\n",
       "      <td>Probabilistic_Methods</td>\n",
       "      <td>Neural_Networks</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1061127</td>\n",
       "      <td>Rule_Learning</td>\n",
       "      <td>Rule_Learning</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1106406</td>\n",
       "      <td>Reinforcement_Learning</td>\n",
       "      <td>Reinforcement_Learning</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>13195</td>\n",
       "      <td>Reinforcement_Learning</td>\n",
       "      <td>Reinforcement_Learning</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>37879</td>\n",
       "      <td>Probabilistic_Methods</td>\n",
       "      <td>Probabilistic_Methods</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1126012</td>\n",
       "      <td>Probabilistic_Methods</td>\n",
       "      <td>Probabilistic_Methods</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1107140</td>\n",
       "      <td>Theory</td>\n",
       "      <td>Theory</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1102850</td>\n",
       "      <td>Neural_Networks</td>\n",
       "      <td>Neural_Networks</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>31349</td>\n",
       "      <td>Probabilistic_Methods</td>\n",
       "      <td>Neural_Networks</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1106418</td>\n",
       "      <td>Theory</td>\n",
       "      <td>Theory</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1123188</td>\n",
       "      <td>Neural_Networks</td>\n",
       "      <td>Neural_Networks</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1128990</td>\n",
       "      <td>Genetic_Algorithms</td>\n",
       "      <td>Genetic_Algorithms</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>109323</td>\n",
       "      <td>Probabilistic_Methods</td>\n",
       "      <td>Probabilistic_Methods</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>217139</td>\n",
       "      <td>Theory</td>\n",
       "      <td>Case_Based</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>31353</td>\n",
       "      <td>Probabilistic_Methods</td>\n",
       "      <td>Neural_Networks</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>32083</td>\n",
       "      <td>Neural_Networks</td>\n",
       "      <td>Neural_Networks</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1126029</td>\n",
       "      <td>Reinforcement_Learning</td>\n",
       "      <td>Reinforcement_Learning</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1118017</td>\n",
       "      <td>Neural_Networks</td>\n",
       "      <td>Neural_Networks</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>49482</td>\n",
       "      <td>Neural_Networks</td>\n",
       "      <td>Neural_Networks</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>753265</td>\n",
       "      <td>Neural_Networks</td>\n",
       "      <td>Neural_Networks</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      Predicted                    True\n",
       "31336     Probabilistic_Methods         Neural_Networks\n",
       "1061127           Rule_Learning           Rule_Learning\n",
       "1106406  Reinforcement_Learning  Reinforcement_Learning\n",
       "13195    Reinforcement_Learning  Reinforcement_Learning\n",
       "37879     Probabilistic_Methods   Probabilistic_Methods\n",
       "1126012   Probabilistic_Methods   Probabilistic_Methods\n",
       "1107140                  Theory                  Theory\n",
       "1102850         Neural_Networks         Neural_Networks\n",
       "31349     Probabilistic_Methods         Neural_Networks\n",
       "1106418                  Theory                  Theory\n",
       "1123188         Neural_Networks         Neural_Networks\n",
       "1128990      Genetic_Algorithms      Genetic_Algorithms\n",
       "109323    Probabilistic_Methods   Probabilistic_Methods\n",
       "217139                   Theory              Case_Based\n",
       "31353     Probabilistic_Methods         Neural_Networks\n",
       "32083           Neural_Networks         Neural_Networks\n",
       "1126029  Reinforcement_Learning  Reinforcement_Learning\n",
       "1118017         Neural_Networks         Neural_Networks\n",
       "49482           Neural_Networks         Neural_Networks\n",
       "753265          Neural_Networks         Neural_Networks"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame({\"Predicted\": node_predictions, \"True\": node_subjects})\n",
    "df.head(20)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "56",
   "metadata": {},
   "source": [
    "## Node representations\n",
    "\n",
    "Evaluate node representations as activations of the output layer and visualise them, coloring nodes by their true subject label. We expect to see nice clusters of papers in the node representation space, with papers of the same subject belonging to the same cluster.\n",
    "\n",
    "We are going to project the node representations to 2d using either TSNE or PCA transform, and visualise them, coloring nodes by their true subject label."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "57",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = all_predictions\n",
    "y = np.argmax(target_encoding.transform(node_subjects), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "58",
   "metadata": {},
   "outputs": [],
   "source": [
    "if X.shape[1] > 2:\n",
    "    transform = TSNE  # PCA\n",
    "\n",
    "    trans = transform(n_components=2)\n",
    "    emb_transformed = pd.DataFrame(trans.fit_transform(X), index=list(G.nodes()))\n",
    "    emb_transformed[\"label\"] = y\n",
    "else:\n",
    "    emb_transformed = pd.DataFrame(X, index=list(G.nodes()))\n",
    "    emb_transformed = emb_transformed.rename(columns={\"0\": 0, \"1\": 1})\n",
    "    emb_transformed[\"label\"] = y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "59",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "alpha = 0.7\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(7, 7))\n",
    "ax.scatter(\n",
    "    emb_transformed[0],\n",
    "    emb_transformed[1],\n",
    "    c=emb_transformed[\"label\"].astype(\"category\"),\n",
    "    cmap=\"jet\",\n",
    "    alpha=alpha,\n",
    ")\n",
    "ax.set(aspect=\"equal\", xlabel=\"$X_1$\", ylabel=\"$X_2$\")\n",
    "plt.title(\n",
    "    \"{} visualization of SGC activations for cora dataset\".format(transform.__name__)\n",
    ")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "60",
   "metadata": {
    "nbsphinx": "hidden",
    "tags": [
     "CloudRunner"
    ]
   },
   "source": [
    "<table><tr><td>Run the latest release of this notebook:</td><td><a href=\"https://mybinder.org/v2/gh/stellargraph/stellargraph/master?urlpath=lab/tree/demos/node-classification/sgc-node-classification.ipynb\" alt=\"Open In Binder\" target=\"_parent\"><img src=\"https://mybinder.org/badge_logo.svg\"/></a></td><td><a href=\"https://colab.research.google.com/github/stellargraph/stellargraph/blob/master/demos/node-classification/sgc-node-classification.ipynb\" alt=\"Open In Colab\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\"/></a></td></tr></table>"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}