make-training-set.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import skimage as ski\n",
    "from skimage import io, transform,draw\n",
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "plt.rcParams[\"figure.figsize\"] =(12,9)\n",
    "import os\n",
    "import copy\n",
    "from IPython.display import clear_output\n",
    "import compression as comp\n",
    "\n",
    "#run these if you want to reload imported modules every time the code is run \n",
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# hide the skimage low contrast warning\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "##### make a bunch of blank 250x250 images with black dots in them for testing and training the neural network \n",
    "N = 100 # length of compressed projective representation \n",
    "d = 150 #dimension of image \n",
    "M = int(round(np.sqrt(2*d**2)))# length of uncompressed projective representation\n",
    "L = 5 # number of projective representations  MUST BE ODD \n",
    "\n",
    "if not os.path.isdir('./data'):\n",
    "    os.mkdir('./data')\n",
    "if not os.path.isdir('./data/training'):\n",
    "    os.mkdir('./data/training')\n",
    "if not os.path.isdir('./data/testing'):\n",
    "    os.mkdir('./data/testing')\n",
    "    \n",
    "S = comp.sensing_matrix(N,M) \n",
    "np.save('./data/sensing_matrix.npy',S)\n",
    "\n",
    "for i in range(15000):\n",
    "\n",
    "    points = np.random.randint(0,d,(np.random.randint(0,20),2))\n",
    "    np.save('./data/training/train-uncomp-%05d.npy'%i,points)\n",
    "\n",
    "    im = np.zeros((d,d),dtype=float)    \n",
    "    for p in points:\n",
    "        im[ski.draw.circle(*p,3,shape=im.shape)] = 1.0\n",
    "    ski.io.imsave('./data/training/train-%05d.jpg'%i,im)\n",
    "    \n",
    "    F = comp.project(points,shape=(d,d),L=L,M=M)\n",
    "    Y = comp.encode(F,N,S) #encode into length N vectors. S is the sensing matrix\n",
    "    np.save('./data/training/train-comp-%05d.npy'%i,Y)  \n",
    "    \n",
    "for i in range(500):\n",
    "\n",
    "    points = np.random.randint(0,d,(np.random.randint(0,20),2))\n",
    "    np.save('./data/testing/test-uncomp-%05d.npy'%i,points)\n",
    "\n",
    "    im = np.zeros((d,d),dtype=float)    \n",
    "    for p in points:\n",
    "        im[ski.draw.circle(*p,3,shape=im.shape)] = 1.0\n",
    "    ski.io.imsave('./data/testing/test-%05d.jpg'%i,im)\n",
    "    \n",
    "    F = comp.project(points,shape=(d,d),L=L,M=M)\n",
    "    Y = comp.encode(F,N,S) #encode into length N vectors. S is the sensing matrix\n",
    "    np.save('./data/testing/test-comp-%05d.npy'%i,Y)      "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# check that the data encode the positions of the points in the compressed representation \n",
    "\n",
    "def check_comp_vs_uncomp(n):\n",
    "    \"\"\" the compressed->uncompressed points are blue plusses\n",
    "    ground truth is pink circles\"\"\"\n",
    "    im = ski.io.imread('./data/training/train-%04d.jpg'%n)\n",
    "    pts_u = np.load('./data/training/train-uncomp-%04d.npy'%n)\n",
    "    pts_c = np.load('./data/training/train-comp-%04d.npy'%n)\n",
    "    pts_u_ = comp.decompress(pts_c,S,im.shape,L,M)\n",
    "    im = np.stack((im,)*3,-1)\n",
    "    for p in pts_u:\n",
    "        r,c = ski.draw.circle_perimeter(p[0],p[1],5,shape = im.shape)\n",
    "        im[r,c]=[255,0,255]\n",
    "    for p in pts_u_:\n",
    "        im = comp.draw_plus(p,im,col=[0,0,255])\n",
    "    plt.imshow(im)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "N = np.random.randint(0,15000,size=5)\n",
    "for n in N:\n",
    "    check_comp_vs_uncomp(n)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}