exercise2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# This is the second exercise from EPFL EE-559 – DEEP LEARNING course"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## hello all!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*** still under work ***<br>\n",
    "The exercise itself is located  <a href=\"https://fleuret.org/ee559/materials/ee559-practical-2.pdf\"> here</a> .<br>\n",
    "For any questions please refer to my email : tomer@nahshon.net"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "from torchvision import datasets\n",
    "import os\n",
    "import sys"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "os.getcwd()\n",
    "os.chdir('../')\n",
    "import dlc_practical_prologue as prologue\n",
    "os.chdir('exercises practical/')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "c = True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "* Using CIFAR\n",
      "Downloading https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz to ./data/cifar10/cifar-10-python.tar.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100.0%"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Files already downloaded and verified\n",
      "** Reduce the data-set (use --full for the full thing)\n",
      "** Use 1000 train and 1000 test samples\n"
     ]
    }
   ],
   "source": [
    "train_input, train_target, test_input, test_target = prologue.load_data(cifar = c)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### exercise 1 - nearest neighbour"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([1000, 3072])"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_input.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "candidate = test_input[2,:] #choosing random candidate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "distance = (train_input - candidate).pow(2).sum(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "value, ind = torch.min(distance, 0, keepdim=False, out=None) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor(8)"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_target[ind]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### exercise 2 - ERROR"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "# TAKEN FROM SOLUTION\n",
    "def compute_nb_errors(train_input, train_target, test_input, test_target, mean = None, proj = None):\n",
    "\n",
    "    if mean is not None:\n",
    "        train_input = train_input - mean\n",
    "        test_input = test_input - mean\n",
    "\n",
    "    if proj is not None:\n",
    "        train_input = train_input.mm(proj.t())\n",
    "        test_input = test_input.mm(proj.t())\n",
    "\n",
    "    nb_errors = 0\n",
    "    for n in range(test_input.size(0)):\n",
    "        if test_target[n] != nearest_classification(train_input, train_target, test_input[n]):\n",
    "            nb_errors = nb_errors + 1\n",
    "\n",
    "    return nb_errors"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "pytorch-env",
   "language": "python",
   "name": "pytorch-env"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}