stellargraph/stellargraph

View on GitHub
demos/embeddings/gcn-unsupervised-graph-embeddings.ipynb

Summary

Maintainability
Test Coverage
{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "0",
   "metadata": {},
   "source": [
    "# Unsupervised graph classification/representation learning via distances"
   ]
  },
  {
   "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/embeddings/gcn-unsupervised-graph-embeddings.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/embeddings/gcn-unsupervised-graph-embeddings.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 demonstrated training a graph classification model without supervision. This model could be used to compute embedding vectors or representations for graphs.\n",
    "\n",
    "The algorithm uses a ground-truth distance between graphs as a metric to train against, by embedding pairs of graphs simultaneously and combining the resulting embedding vectors to match the distance.\n",
    "\n",
    "It is inspired by UGraphEmb[1].\n",
    "\n",
    "[1]: Y. Bai et al., “Unsupervised Inductive Graph-Level Representation Learning via Graph-Graph Proximity,” [arXiv:1904.01098](http://arxiv.org/abs/1904.01098) [cs, stat], Jun. 2019."
   ]
  },
  {
   "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 stellargraph as sg\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import networkx as nx\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "\n",
    "from IPython.display import display, HTML"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6",
   "metadata": {},
   "source": [
    "## Dataset\n",
    "\n",
    "The PROTEINS dataset consists of about one thousand graphs, with binary labels `1` or `2`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "Each graph represents a protein and graph labels represent whether they are are enzymes or non-enzymes. The dataset includes 1113 graphs with 39 nodes and 73 edges on average for each graph. Graph nodes have 4 attributes (including a one-hot encoding of their label), and each graph is labelled as belonging to 1 of 2 classes."
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dataset = sg.datasets.PROTEINS()\n",
    "display(HTML(dataset.description))\n",
    "graphs, graph_labels = dataset.load()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "8",
   "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>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>663</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>450</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   label\n",
       "1    663\n",
       "2    450"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "graph_labels.value_counts().to_frame()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9",
   "metadata": {},
   "source": [
    "The `graphs` value consists of many `StellarGraph` instances:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "10",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "StellarGraph: Undirected multigraph\n",
      " Nodes: 42, Edges: 162\n",
      "\n",
      " Node types:\n",
      "  default: [42]\n",
      "    Features: float32 vector, length 4\n",
      "    Edge types: default-default->default\n",
      "\n",
      " Edge types:\n",
      "    default-default->default: [162]\n",
      "        Weights: all 1 (default)\n",
      "        Features: none\n"
     ]
    }
   ],
   "source": [
    "print(graphs[0].info())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "11",
   "metadata": {},
   "source": [
    "Summary statistics of the sizes of the graphs:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "12",
   "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>nodes</th>\n",
       "      <th>edges</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1113.0</td>\n",
       "      <td>1113.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>39.1</td>\n",
       "      <td>145.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>45.8</td>\n",
       "      <td>169.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>4.0</td>\n",
       "      <td>10.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>15.0</td>\n",
       "      <td>56.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>26.0</td>\n",
       "      <td>98.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>45.0</td>\n",
       "      <td>174.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>620.0</td>\n",
       "      <td>2098.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        nodes   edges\n",
       "count  1113.0  1113.0\n",
       "mean     39.1   145.6\n",
       "std      45.8   169.3\n",
       "min       4.0    10.0\n",
       "25%      15.0    56.0\n",
       "50%      26.0    98.0\n",
       "75%      45.0   174.0\n",
       "max     620.0  2098.0"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "summary = pd.DataFrame(\n",
    "    [(g.number_of_nodes(), g.number_of_edges()) for g in graphs],\n",
    "    columns=[\"nodes\", \"edges\"],\n",
    ")\n",
    "summary.describe().round(1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "13",
   "metadata": {},
   "source": [
    "## Create the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "14",
   "metadata": {},
   "outputs": [],
   "source": [
    "generator = sg.mapper.PaddedGraphGenerator(graphs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "15",
   "metadata": {},
   "outputs": [],
   "source": [
    "gc_model = sg.layer.GCNSupervisedGraphClassification(\n",
    "    [64, 32], [\"relu\", \"relu\"], generator, pool_all_layers=True\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "16",
   "metadata": {},
   "outputs": [],
   "source": [
    "inp1, out1 = gc_model.in_out_tensors()\n",
    "inp2, out2 = gc_model.in_out_tensors()\n",
    "\n",
    "vec_distance = tf.norm(out1 - out2, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "17",
   "metadata": {},
   "outputs": [],
   "source": [
    "pair_model = keras.Model(inp1 + inp2, vec_distance)\n",
    "embedding_model = keras.Model(inp1, out1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "18",
   "metadata": {},
   "source": [
    "## Train the model\n",
    "\n",
    "The model is trained on 100 random pairs of graphs, along with the ground-truth distance between them."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "19",
   "metadata": {},
   "source": [
    "### Similarity measure\n",
    "\n",
    "This method can use any notion of distance or similarity between two graphs. In this case, we use something efficient, but not particularly accurate: the distance between the spectrum (or eigenvalues) of the [Laplacian matrix](https://en.wikipedia.org/wiki/Laplacian_matrix) of the graphs.\n",
    "\n",
    "Other options include graph edit distance and minimum common subgraph, but these are NP-hard to compute and are too slow for this demonstration."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "20",
   "metadata": {},
   "outputs": [],
   "source": [
    "def graph_distance(graph1, graph2):\n",
    "    spec1 = nx.laplacian_spectrum(graph1.to_networkx(feature_attr=None))\n",
    "    spec2 = nx.laplacian_spectrum(graph2.to_networkx(feature_attr=None))\n",
    "    k = min(len(spec1), len(spec2))\n",
    "    return np.linalg.norm(spec1[:k] - spec2[:k])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "21",
   "metadata": {},
   "source": [
    "### Training examples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "22",
   "metadata": {},
   "outputs": [],
   "source": [
    "graph_idx = np.random.RandomState(0).randint(len(graphs), size=(100, 2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "23",
   "metadata": {},
   "outputs": [],
   "source": [
    "targets = [graph_distance(graphs[left], graphs[right]) for left, right in graph_idx]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "24",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_gen = generator.flow(graph_idx, batch_size=10, targets=targets)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "25",
   "metadata": {},
   "source": [
    "### Training procedure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "26",
   "metadata": {},
   "outputs": [],
   "source": [
    "pair_model.compile(keras.optimizers.Adam(1e-2), loss=\"mse\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "27",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  ['...']\n",
      "CPU times: user 3min 30s, sys: 1min 40s, total: 5min 11s\n",
      "Wall time: 1min 17s\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%time\n",
    "history = pair_model.fit(train_gen, epochs=500, verbose=0)\n",
    "sg.utils.plot_history(history)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "28",
   "metadata": {},
   "source": [
    "## Compute embeddings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "29",
   "metadata": {},
   "outputs": [],
   "source": [
    "embeddings = embedding_model.predict(generator.flow(graphs))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "30",
   "metadata": {},
   "source": [
    "## Downstream tasks\n",
    "\n",
    "Now that we've computed some embedding vectors in an unsupervised fashion, we can use them for other supervised, semi-supervised and unsupervised tasks."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "31",
   "metadata": {},
   "source": [
    "### Supervised graph classification\n",
    "\n",
    "We can use the embedding vectors to perform logistic regression classification, using the labels."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "32",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn import model_selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "33",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test classification accuracy: 0.6846307385229541\n"
     ]
    }
   ],
   "source": [
    "train_labels, test_labels = model_selection.train_test_split(\n",
    "    graph_labels, train_size=0.1, test_size=None, stratify=graph_labels\n",
    ")\n",
    "\n",
    "test_embeddings = embeddings[test_labels.index - 1]\n",
    "train_embeddings = embeddings[train_labels.index - 1]\n",
    "\n",
    "lr = LogisticRegression(multi_class=\"auto\", solver=\"lbfgs\")\n",
    "lr.fit(train_embeddings, train_labels)\n",
    "\n",
    "y_pred = lr.predict(test_embeddings)\n",
    "gcn_acc = (y_pred == test_labels).mean()\n",
    "print(f\"Test classification accuracy: {gcn_acc}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "34",
   "metadata": {},
   "source": [
    "#### Confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "35",
   "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>predicted</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>true</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>506</td>\n",
       "      <td>91</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>225</td>\n",
       "      <td>180</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "predicted    1    2\n",
       "true               \n",
       "1          506   91\n",
       "2          225  180"
      ]
     },
     "execution_count": 152,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(test_labels, y_pred, rownames=[\"true\"], colnames=[\"predicted\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "36",
   "metadata": {},
   "source": [
    "### Visualising embeddings\n",
    "\n",
    "We can also get a qualitative measure of the embeddings, using dimensionality reduction."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "37",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.manifold import TSNE\n",
    "\n",
    "tsne = TSNE(2)\n",
    "two_d = tsne.fit_transform(embeddings)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "38",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x163db1a20>"
      ]
     },
     "execution_count": 154,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "\n",
    "plt.scatter(two_d[:, 0], two_d[:, 1], c=graph_labels.cat.codes, cmap=\"jet\", alpha=0.4)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "39",
   "metadata": {},
   "source": [
    "## Conclusion\n",
    "\n",
    "This demo demonstrated training a graph classification model without supervision. This model could be used to compute embedding vectors or representations for graphs. \n",
    "\n",
    "The algorithm works with three components:\n",
    "\n",
    "- a ground truth distance or similarity between two graphs such as graph edit distance, or, in this case, Laplacian spectrum distance (for efficiency)\n",
    "- a model that encodes graphs into embedding vectors\n",
    "- a data generator that yields pairs of graphs and the corresponding ground truth distance\n",
    "\n",
    "This model is inspired by UGraphEmb[1]."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "40",
   "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/embeddings/gcn-unsupervised-graph-embeddings.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/embeddings/gcn-unsupervised-graph-embeddings.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
}