data/superpixels/prepare_superpixels_CIFAR.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Notebook for preparing and saving CIFAR graphs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import torch\n",
    "import pickle\n",
    "import time\n",
    "import os\n",
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Download CIFAR dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "File already downloaded\n"
     ]
    }
   ],
   "source": [
    "if not os.path.isfile('superpixels.zip'):\n",
    "    print('downloading..')\n",
    "    !curl https://www.dropbox.com/s/y2qwa77a0fxem47/superpixels.zip?dl=1 -o superpixels.zip -J -L -k\n",
    "    !tar -xvf superpixels.zip -C ../\n",
    "else:\n",
    "    print('File already downloaded')\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Convert to DGL format and save with pickle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/Users/xbresson/Documents/Dropbox/06_NTU_2017_now/03_my_codes/34_benchmark20/GITHUB_benchmark_project/benchmarking-gnn\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "os.chdir('../../') # go to root folder of the project\n",
    "print(os.getcwd())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "\n",
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "\n",
    "from data.superpixels import SuperPixDatasetDGL \n",
    "\n",
    "from data.data import LoadData\n",
    "from torch.utils.data import DataLoader\n",
    "from data.superpixels import SuperPixDataset\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Adj matrix defined from super-pixel locations (only)\n",
      "preparing 10000 graphs for the TEST set...\n",
      "preparing 50000 graphs for the TRAIN set...\n",
      "[I] Data load time: 639.9484s\n",
      "Time (sec): 639.9550220966339\n"
     ]
    }
   ],
   "source": [
    "start = time.time()\n",
    "\n",
    "DATASET_NAME = 'CIFAR10'\n",
    "dataset = SuperPixDatasetDGL(DATASET_NAME) \n",
    "\n",
    "print('Time (sec):',time.time() - start) # 636s=10min\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "nb/min/max : 45000 85 150\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "nb/min/max : 5000 95 145\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "nb/min/max : 10000 94 139\n"
     ]
    }
   ],
   "source": [
    "def plot_histo_graphs(dataset, title):\n",
    "    # histogram of graph sizes\n",
    "    graph_sizes = []\n",
    "    for graph in dataset:\n",
    "        graph_sizes.append(graph[0].number_of_nodes())\n",
    "        #graph_sizes.append(graph[0].number_of_edges())\n",
    "    plt.figure(1)\n",
    "    plt.hist(graph_sizes, bins=20)\n",
    "    plt.title(title)\n",
    "    plt.show()\n",
    "    graph_sizes = torch.Tensor(graph_sizes)\n",
    "    print('nb/min/max :',len(graph_sizes),graph_sizes.min().long().item(),graph_sizes.max().long().item())\n",
    "    \n",
    "plot_histo_graphs(dataset.train,'trainset')\n",
    "plot_histo_graphs(dataset.val,'valset')\n",
    "plot_histo_graphs(dataset.test,'testset')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "45000\n",
      "5000\n",
      "10000\n",
      "(DGLGraph(num_nodes=110, num_edges=880,\n",
      "         ndata_schemes={'feat': Scheme(shape=(5,), dtype=torch.float16)}\n",
      "         edata_schemes={'feat': Scheme(shape=(1,), dtype=torch.float16)}), tensor(6))\n",
      "(DGLGraph(num_nodes=123, num_edges=984,\n",
      "         ndata_schemes={'feat': Scheme(shape=(5,), dtype=torch.float16)}\n",
      "         edata_schemes={'feat': Scheme(shape=(1,), dtype=torch.float16)}), tensor(6))\n",
      "(DGLGraph(num_nodes=118, num_edges=944,\n",
      "         ndata_schemes={'feat': Scheme(shape=(5,), dtype=torch.float16)}\n",
      "         edata_schemes={'feat': Scheme(shape=(1,), dtype=torch.float16)}), tensor(3))\n"
     ]
    }
   ],
   "source": [
    "print(len(dataset.train))\n",
    "print(len(dataset.val))\n",
    "print(len(dataset.test))\n",
    "\n",
    "print(dataset.train[0])\n",
    "print(dataset.val[0])\n",
    "print(dataset.test[0])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Time (sec): 56.585689067840576\n"
     ]
    }
   ],
   "source": [
    "start = time.time()\n",
    "\n",
    "with open('data/superpixels/CIFAR10.pkl','wb') as f:\n",
    "        pickle.dump([dataset.train,dataset.val,dataset.test],f)\n",
    "        \n",
    "print('Time (sec):',time.time() - start) # 58s\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Test load function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[I] Loading dataset CIFAR10...\n",
      "train, test, val sizes : 45000 5000 10000\n",
      "[I] Finished loading.\n",
      "[I] Data load time: 76.3279s\n"
     ]
    }
   ],
   "source": [
    "DATASET_NAME = 'CIFAR10'\n",
    "dataset = LoadData(DATASET_NAME) # 54s\n",
    "trainset, valset, testset = dataset.train, dataset.val, dataset.test\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Time (sec): 0.002046823501586914\n"
     ]
    }
   ],
   "source": [
    "start = time.time()\n",
    "\n",
    "batch_size = 10\n",
    "collate = SuperPixDataset.collate\n",
    "train_loader = DataLoader(trainset, batch_size=batch_size, shuffle=True, collate_fn=collate)\n",
    "\n",
    "print('Time (sec):',time.time() - start) # 0.0001s\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}