demos/link-prediction/node2vec-link-prediction.ipynb from stellargraph/stellargraph

demos/link-prediction/node2vec-link-prediction.ipynb
Summary

Maintainability

Test Coverage

Issues
{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "0",
   "metadata": {},
   "source": [
    "# Link prediction with Node2Vec"
   ]
  },
  {
   "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/link-prediction/node2vec-link-prediction.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/link-prediction/node2vec-link-prediction.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 demo notebook demonstrates how to predict citation links/edges between papers using Node2Vec on the Cora dataset.\n",
    "\n",
    "We're going to tackle link prediction as a supervised learning problem on top of node representations/embeddings. The embeddings are computed with the unsupervised node2vec algorithm. After obtaining embeddings, a binary classifier can be used to predict a link, or not, between any two nodes in the graph. Various hyperparameters could be relevant in obtaining the best link classifier - this demo demonstrates incorporating model selection into the pipeline for choosing the best binary operator to apply on a pair of node embeddings.\n",
    "\n",
    "There are four steps:\n",
    "\n",
    "1. Obtain embeddings for each node\n",
    "2. For each set of hyperparameters, train a classifier\n",
    "3. Select the classifier that performs the best\n",
    "4. Evaluate the selected classifier on unseen data to validate its ability to generalise\n",
    "\n",
    "<a name=\"refs\"></a>\n",
    "**References:** \n",
    "\n",
    "[1] Node2Vec: Scalable Feature Learning for Networks. A. Grover, J. Leskovec. ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), 2016. "
   ]
  },
  {
   "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 matplotlib.pyplot as plt\n",
    "from math import isclose\n",
    "from sklearn.decomposition import PCA\n",
    "import os\n",
    "import networkx as nx\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from stellargraph import StellarGraph, datasets\n",
    "from stellargraph.data import EdgeSplitter\n",
    "from collections import Counter\n",
    "import multiprocessing\n",
    "from IPython.display import display, HTML\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6",
   "metadata": {},
   "source": [
    "## Load the dataset\n",
    "\n",
    "The Cora dataset is a homogeneous network where all nodes are papers and edges between nodes are citation links, e.g. paper A cites paper B."
   ]
  },
  {
   "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",
    "graph, _ = dataset.load(largest_connected_component_only=True, str_node_ids=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "StellarGraph: Undirected multigraph\n",
      " Nodes: 2485, Edges: 5209\n",
      "\n",
      " Node types:\n",
      "  paper: [2485]\n",
      "    Features: float32 vector, length 1433\n",
      "    Edge types: paper-cites->paper\n",
      "\n",
      " Edge types:\n",
      "    paper-cites->paper: [5209]\n"
     ]
    }
   ],
   "source": [
    "print(graph.info())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "10",
   "metadata": {},
   "source": [
    "## Construct splits of the input data\n",
    "\n",
    "We have to carefully split the data to avoid data leakage and evaluate the algorithms correctly:\n",
    "\n",
    "* For computing node embeddings, a **Train Graph** (`graph_train`)\n",
    "* For training classifiers, a classifier **Training Set** (`examples_train`) of positive and negative edges that weren't used for computing node embeddings\n",
    "* For choosing the best classifier, an **Model Selection Test Set** (`examples_model_selection`) of positive and negative edges that weren't used for computing node embeddings or training the classifier \n",
    "* For the final evaluation, a **Test Graph** (`graph_test`) to compute test node embeddings with more edges than the Train Graph, and a **Test Set** (`examples_test`) of positive and negative edges not used for neither computing the test node embeddings or for classifier training or model selection"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "11",
   "metadata": {},
   "source": [
    "###  Test Graph\n",
    "\n",
    "We begin with the full graph and use the `EdgeSplitter` class to produce:\n",
    "\n",
    "* Test Graph\n",
    "* Test set of positive/negative link examples\n",
    "\n",
    "The Test Graph is the reduced graph we obtain from removing the test set of links from the full graph."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "12",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "** Sampled 520 positive and 520 negative edges. **\n",
      "StellarGraph: Undirected multigraph\n",
      " Nodes: 2485, Edges: 4689\n",
      "\n",
      " Node types:\n",
      "  paper: [2485]\n",
      "    Features: float32 vector, length 1433\n",
      "    Edge types: paper-cites->paper\n",
      "\n",
      " Edge types:\n",
      "    paper-cites->paper: [4689]\n"
     ]
    }
   ],
   "source": [
    "# Define an edge splitter on the original graph:\n",
    "edge_splitter_test = EdgeSplitter(graph)\n",
    "\n",
    "# Randomly sample a fraction p=0.1 of all positive links, and same number of negative links, from graph, and obtain the\n",
    "# reduced graph graph_test with the sampled links removed:\n",
    "graph_test, examples_test, labels_test = edge_splitter_test.train_test_split(\n",
    "    p=0.1, method=\"global\"\n",
    ")\n",
    "\n",
    "print(graph_test.info())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "13",
   "metadata": {},
   "source": [
    "### Train Graph\n",
    "\n",
    "This time, we use the `EdgeSplitter` on the Test Graph, and perform a train/test split on the examples to produce:\n",
    "\n",
    "* Train Graph\n",
    "* Training set of link examples\n",
    "* Set of link examples for model selection\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "14",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "** Sampled 468 positive and 468 negative edges. **\n",
      "StellarGraph: Undirected multigraph\n",
      " Nodes: 2485, Edges: 4221\n",
      "\n",
      " Node types:\n",
      "  paper: [2485]\n",
      "    Features: float32 vector, length 1433\n",
      "    Edge types: paper-cites->paper\n",
      "\n",
      " Edge types:\n",
      "    paper-cites->paper: [4221]\n"
     ]
    }
   ],
   "source": [
    "# Do the same process to compute a training subset from within the test graph\n",
    "edge_splitter_train = EdgeSplitter(graph_test, graph)\n",
    "graph_train, examples, labels = edge_splitter_train.train_test_split(\n",
    "    p=0.1, method=\"global\"\n",
    ")\n",
    "(\n",
    "    examples_train,\n",
    "    examples_model_selection,\n",
    "    labels_train,\n",
    "    labels_model_selection,\n",
    ") = train_test_split(examples, labels, train_size=0.75, test_size=0.25)\n",
    "\n",
    "print(graph_train.info())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "15",
   "metadata": {},
   "source": [
    "Below is a summary of the different splits that have been created in this section"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "16",
   "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>Number of Examples</th>\n",
       "      <th>Hidden from</th>\n",
       "      <th>Picked from</th>\n",
       "      <th>Use</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Split</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Training Set</th>\n",
       "      <td>702</td>\n",
       "      <td>Train Graph</td>\n",
       "      <td>Test Graph</td>\n",
       "      <td>Train the Link Classifier</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Model Selection</th>\n",
       "      <td>234</td>\n",
       "      <td>Train Graph</td>\n",
       "      <td>Test Graph</td>\n",
       "      <td>Select the best Link Classifier model</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Test set</th>\n",
       "      <td>1040</td>\n",
       "      <td>Test Graph</td>\n",
       "      <td>Full Graph</td>\n",
       "      <td>Evaluate the best Link Classifier</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Number of Examples  Hidden from Picked from  \\\n",
       "Split                                                          \n",
       "Training Set                    702  Train Graph  Test Graph   \n",
       "Model Selection                 234  Train Graph  Test Graph   \n",
       "Test set                       1040   Test Graph  Full Graph   \n",
       "\n",
       "                                                   Use  \n",
       "Split                                                   \n",
       "Training Set                 Train the Link Classifier  \n",
       "Model Selection  Select the best Link Classifier model  \n",
       "Test set             Evaluate the best Link Classifier  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(\n",
    "    [\n",
    "        (\n",
    "            \"Training Set\",\n",
    "            len(examples_train),\n",
    "            \"Train Graph\",\n",
    "            \"Test Graph\",\n",
    "            \"Train the Link Classifier\",\n",
    "        ),\n",
    "        (\n",
    "            \"Model Selection\",\n",
    "            len(examples_model_selection),\n",
    "            \"Train Graph\",\n",
    "            \"Test Graph\",\n",
    "            \"Select the best Link Classifier model\",\n",
    "        ),\n",
    "        (\n",
    "            \"Test set\",\n",
    "            len(examples_test),\n",
    "            \"Test Graph\",\n",
    "            \"Full Graph\",\n",
    "            \"Evaluate the best Link Classifier\",\n",
    "        ),\n",
    "    ],\n",
    "    columns=(\"Split\", \"Number of Examples\", \"Hidden from\", \"Picked from\", \"Use\"),\n",
    ").set_index(\"Split\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "17",
   "metadata": {},
   "source": [
    "## Node2Vec\n",
    "\n",
    "We use Node2Vec [[1]](#refs), to calculate node embeddings. These embeddings are learned in such a way to ensure that nodes that are close in the graph remain close in the embedding space. Node2Vec first involves running random walks on the graph to obtain our context pairs, and using these to train a Word2Vec model.\n",
    "\n",
    "These are the set of parameters we can use:\n",
    "\n",
    "* `p` - Random walk parameter \"p\"\n",
    "* `q` - Random walk parameter \"q\"\n",
    "* `dimensions` - Dimensionality of node2vec embeddings\n",
    "* `num_walks` - Number of walks from each node\n",
    "* `walk_length` - Length of each random walk\n",
    "* `window_size` - Context window size for Word2Vec\n",
    "* `num_iter` - number of SGD iterations (epochs)\n",
    "* `workers` - Number of workers for Word2Vec"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "18",
   "metadata": {},
   "outputs": [],
   "source": [
    "p = 1.0\n",
    "q = 1.0\n",
    "dimensions = 128\n",
    "num_walks = 10\n",
    "walk_length = 80\n",
    "window_size = 10\n",
    "num_iter = 1\n",
    "workers = multiprocessing.cpu_count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "19",
   "metadata": {},
   "outputs": [],
   "source": [
    "from stellargraph.data import BiasedRandomWalk\n",
    "from gensim.models import Word2Vec\n",
    "\n",
    "\n",
    "def node2vec_embedding(graph, name):\n",
    "    rw = BiasedRandomWalk(graph)\n",
    "    walks = rw.run(graph.nodes(), n=num_walks, length=walk_length, p=p, q=q)\n",
    "    print(f\"Number of random walks for '{name}': {len(walks)}\")\n",
    "\n",
    "    model = Word2Vec(\n",
    "        walks,\n",
    "        vector_size=dimensions,\n",
    "        window=window_size,\n",
    "        min_count=0,\n",
    "        sg=1,\n",
    "        workers=workers,\n",
    "        epochs=num_iter,\n",
    "    )\n",
    "\n",
    "    def get_embedding(u):\n",
    "        return model.wv[u]\n",
    "\n",
    "    return get_embedding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "20",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of random walks for 'Train Graph': 24850\n"
     ]
    }
   ],
   "source": [
    "embedding_train = node2vec_embedding(graph_train, \"Train Graph\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "21",
   "metadata": {},
   "source": [
    "## Train and evaluate the link prediction model\n",
    "\n",
    "There are a few steps involved in using the Word2Vec model to perform link prediction:\n",
    "1. We calculate link/edge embeddings for the positive and negative edge samples by applying a binary operator on the embeddings of the source and target nodes of each sampled edge.\n",
    "2. Given the embeddings of the positive and negative examples, we train a logistic regression classifier to predict a binary value indicating whether an edge between two nodes should exist or not.\n",
    "3. We evaluate the performance of the link classifier for each of the 4 operators on the training data with node embeddings calculated on the **Train Graph** (`graph_train`), and select the best classifier.\n",
    "4. The best classifier is then used to calculate scores on the test data with node embeddings calculated on the **Test Graph** (`graph_test`).\n",
    "\n",
    "Below are a set of helper functions that let us repeat these steps for each of the binary operators."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "22",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.linear_model import LogisticRegressionCV\n",
    "from sklearn.metrics import roc_auc_score\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "\n",
    "# 1. link embeddings\n",
    "def link_examples_to_features(link_examples, transform_node, binary_operator):\n",
    "    return [\n",
    "        binary_operator(transform_node(src), transform_node(dst))\n",
    "        for src, dst in link_examples\n",
    "    ]\n",
    "\n",
    "\n",
    "# 2. training classifier\n",
    "def train_link_prediction_model(\n",
    "    link_examples, link_labels, get_embedding, binary_operator\n",
    "):\n",
    "    clf = link_prediction_classifier()\n",
    "    link_features = link_examples_to_features(\n",
    "        link_examples, get_embedding, binary_operator\n",
    "    )\n",
    "    clf.fit(link_features, link_labels)\n",
    "    return clf\n",
    "\n",
    "\n",
    "def link_prediction_classifier(max_iter=2000):\n",
    "    lr_clf = LogisticRegressionCV(Cs=10, cv=10, scoring=\"roc_auc\", max_iter=max_iter)\n",
    "    return Pipeline(steps=[(\"sc\", StandardScaler()), (\"clf\", lr_clf)])\n",
    "\n",
    "\n",
    "# 3. and 4. evaluate classifier\n",
    "def evaluate_link_prediction_model(\n",
    "    clf, link_examples_test, link_labels_test, get_embedding, binary_operator\n",
    "):\n",
    "    link_features_test = link_examples_to_features(\n",
    "        link_examples_test, get_embedding, binary_operator\n",
    "    )\n",
    "    score = evaluate_roc_auc(clf, link_features_test, link_labels_test)\n",
    "    return score\n",
    "\n",
    "\n",
    "def evaluate_roc_auc(clf, link_features, link_labels):\n",
    "    predicted = clf.predict_proba(link_features)\n",
    "\n",
    "    # check which class corresponds to positive links\n",
    "    positive_column = list(clf.classes_).index(1)\n",
    "    return roc_auc_score(link_labels, predicted[:, positive_column])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "23",
   "metadata": {},
   "source": [
    "We consider 4 different operators: \n",
    "\n",
    "* *Hadamard*\n",
    "* $L_1$\n",
    "* $L_2$\n",
    "* *average*\n",
    "\n",
    "The paper [[1]](#refs) provides a detailed description of these operators. All operators produce link embeddings that have equal dimensionality to the input node embeddings (128 dimensions for our example). "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "24",
   "metadata": {},
   "outputs": [],
   "source": [
    "def operator_hadamard(u, v):\n",
    "    return u * v\n",
    "\n",
    "\n",
    "def operator_l1(u, v):\n",
    "    return np.abs(u - v)\n",
    "\n",
    "\n",
    "def operator_l2(u, v):\n",
    "    return (u - v) ** 2\n",
    "\n",
    "\n",
    "def operator_avg(u, v):\n",
    "    return (u + v) / 2.0\n",
    "\n",
    "\n",
    "def run_link_prediction(binary_operator):\n",
    "    clf = train_link_prediction_model(\n",
    "        examples_train, labels_train, embedding_train, binary_operator\n",
    "    )\n",
    "    score = evaluate_link_prediction_model(\n",
    "        clf,\n",
    "        examples_model_selection,\n",
    "        labels_model_selection,\n",
    "        embedding_train,\n",
    "        binary_operator,\n",
    "    )\n",
    "\n",
    "    return {\n",
    "        \"classifier\": clf,\n",
    "        \"binary_operator\": binary_operator,\n",
    "        \"score\": score,\n",
    "    }\n",
    "\n",
    "\n",
    "binary_operators = [operator_hadamard, operator_l1, operator_l2, operator_avg]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "25",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best result from 'operator_l1'\n"
     ]
    },
    {
     "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>ROC AUC score</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>name</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>operator_hadamard</th>\n",
       "      <td>0.883362</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>operator_l1</th>\n",
       "      <td>0.891015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>operator_l2</th>\n",
       "      <td>0.886894</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>operator_avg</th>\n",
       "      <td>0.666127</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   ROC AUC score\n",
       "name                            \n",
       "operator_hadamard       0.883362\n",
       "operator_l1             0.891015\n",
       "operator_l2             0.886894\n",
       "operator_avg            0.666127"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results = [run_link_prediction(op) for op in binary_operators]\n",
    "best_result = max(results, key=lambda result: result[\"score\"])\n",
    "\n",
    "print(f\"Best result from '{best_result['binary_operator'].__name__}'\")\n",
    "\n",
    "pd.DataFrame(\n",
    "    [(result[\"binary_operator\"].__name__, result[\"score\"]) for result in results],\n",
    "    columns=(\"name\", \"ROC AUC score\"),\n",
    ").set_index(\"name\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "26",
   "metadata": {},
   "source": [
    "### Evaluate the best model using the test set\n",
    "\n",
    "Now that we've trained and selected our best model, we use a test set of embeddings and calculate a final evaluation score."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "27",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of random walks for 'Test Graph': 24850\n"
     ]
    }
   ],
   "source": [
    "embedding_test = node2vec_embedding(graph_test, \"Test Graph\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "28",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ROC AUC score on test set using 'operator_l1': 0.9120266272189348\n"
     ]
    }
   ],
   "source": [
    "test_score = evaluate_link_prediction_model(\n",
    "    best_result[\"classifier\"],\n",
    "    examples_test,\n",
    "    labels_test,\n",
    "    embedding_test,\n",
    "    best_result[\"binary_operator\"],\n",
    ")\n",
    "print(\n",
    "    f\"ROC AUC score on test set using '{best_result['binary_operator'].__name__}': {test_score}\"\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "29",
   "metadata": {},
   "source": [
    "### Visualise representations of link embeddings\n",
    "\n",
    "Learned link embeddings have 128 dimensions but for visualisation we project them down to 2 dimensions using the PCA algorithm ([link](https://en.wikipedia.org/wiki/Principal_component_analysis)). \n",
    "\n",
    "Blue points represent positive edges and red points represent negative (no edge should exist between the corresponding vertices) edges."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "30",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x1586ffa50>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Calculate edge features for test data\n",
    "link_features = link_examples_to_features(\n",
    "    examples_test, embedding_test, best_result[\"binary_operator\"]\n",
    ")\n",
    "\n",
    "# Learn a projection from 128 dimensions to 2\n",
    "pca = PCA(n_components=2)\n",
    "X_transformed = pca.fit_transform(link_features)\n",
    "\n",
    "# plot the 2-dimensional points\n",
    "plt.figure(figsize=(16, 12))\n",
    "plt.scatter(\n",
    "    X_transformed[:, 0],\n",
    "    X_transformed[:, 1],\n",
    "    c=np.where(labels_test == 1, \"b\", \"r\"),\n",
    "    alpha=0.5,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "31",
   "metadata": {},
   "source": [
    "This example has demonstrated how to use the `stellargraph` library to build a link prediction algorithm for homogeneous graphs using the Node2Vec, [1], representation learning algorithm."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "32",
   "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/link-prediction/node2vec-link-prediction.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/link-prediction/node2vec-link-prediction.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": "3.7.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}