04-패션_아이템을_구분하는_DNN/fashion_mnist.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Fashion MNIST 데이터셋 알아보기"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from torchvision import datasets, transforms, utils\n",
    "from torch.utils import data\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fashion MNIST 데이터셋"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "transform = transforms.Compose([\n",
    "    transforms.ToTensor()\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "trainset = datasets.FashionMNIST(\n",
    "    root      = './.data/', \n",
    "    train     = True,\n",
    "    download  = True,\n",
    "    transform = transform\n",
    ")\n",
    "testset = datasets.FashionMNIST(\n",
    "    root      = './.data/', \n",
    "    train     = False,\n",
    "    download  = True,\n",
    "    transform = transform\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 16\n",
    "\n",
    "train_loader = data.DataLoader(\n",
    "    dataset     = trainset,\n",
    "    batch_size  = batch_size\n",
    ")\n",
    "test_loader = data.DataLoader(\n",
    "    dataset     = testset,\n",
    "    batch_size  = batch_size\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataiter       = iter(train_loader)\n",
    "images, labels = next(dataiter)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 멀리서 살펴보기\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "img   = utils.make_grid(images, padding=0)\n",
    "npimg = img.numpy()\n",
    "plt.figure(figsize=(10, 7))\n",
    "plt.imshow(np.transpose(npimg, (1,2,0)))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([9, 0, 0, 3, 0, 2, 7, 2, 5, 5, 0, 9, 5, 5, 7, 9])\n"
     ]
    }
   ],
   "source": [
    "print(labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ankle boot\n",
      "T-shirt/top\n",
      "T-shirt/top\n",
      "Dress\n",
      "T-shirt/top\n",
      "Pullover\n",
      "Sneaker\n",
      "Pullover\n",
      "Sandal\n",
      "Sandal\n",
      "T-shirt/top\n",
      "Ankle boot\n",
      "Sandal\n",
      "Sandal\n",
      "Sneaker\n",
      "Ankle boot\n"
     ]
    }
   ],
   "source": [
    "CLASSES = {\n",
    "    0: 'T-shirt/top',\n",
    "    1: 'Trouser',\n",
    "    2: 'Pullover',\n",
    "    3: 'Dress',\n",
    "    4: 'Coat',\n",
    "    5: 'Sandal',\n",
    "    6: 'Shirt',\n",
    "    7: 'Sneaker',\n",
    "    8: 'Bag',\n",
    "    9: 'Ankle boot'\n",
    "}\n",
    "\n",
    "\n",
    "for label in labels:\n",
    "    index = label.item()\n",
    "    print(CLASSES[index])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 가까이서 살펴보기\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(28, 28)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPsAAAEICAYAAACZA4KlAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAUMElEQVR4nO3de7CdVXnH8e/P3OAkQm4lhnAJN7VpBiNkGBhBLE4R0zpI6SCUobGVBq1CneoUxv4h0w5TpuJ17FhjQUJHUBkF6QxFCdMZYFBjgBTCpQlYMDcSCInkai48/WO/cTbh7LWO+923nPX7zJw5e+/nvaz9Js95372fd62liMDMRr+39LsBZtYbTnazQjjZzQrhZDcrhJPdrBBOdrNCONkLIulhSR9tETtR0vYeN8l6yMk+4CRtb/p5XdKupueXd2o/EfHLiJiUacuwfywknSPpQUljJYWk2Z1ql3XO2H43wNKaE1DSC8CVEbG0l22QlDsp/DFwby/aYu3zmX2UkTQk6XZJmyVtlbRM0vSmRU6Q9IikbZLukzS1Wu9kSdG0nYcl/ZOknwI7gDuAs4B/q64qvtK0zQU0kv3B6vlT1TIXV9v6uKTnqjbdLWlm9fqBK4GrJf2fpFck3TiCPy7WBh/U0ecvgSHgGGAa8DfA7qb4nwMLgRnARODvEtu6Avgr4AjgcuCnwMcjYlJEfBpA0rHA5Ih4Anhvtd4fVMv8QNL5wD8CfwbMAtYD3zloPxcCpwGnV8v9RRvv2zKc7KPPXmA6cHJE7I+I5RHR/MXbzRGxOiJ2AncC8xLbuiUinomIvRGxr8UyC4D/SmzjcuDfI2JFROwGrgPOlXRM0zI3RsSWiHgR+BpwWeY9Whuc7IcwSWMO+gLvaOBWYCnwfUnrqsvi5u9mXmp6vBNIfSm3ZgTNOHAJ38rRwIsHnkTEa8AWGmf54fbzYrWOdZiT/RBWnbknNf2sj4g9EXF9RPw+cDZwEY2za1u7SD2XNL7ax9IWy0Pjsv34pnXeCkwB1jUtc2zT4+OqdazDnOyjjKTzJM2tvuR6jcZl/esd2vxG4MSm5+cCj0bEDmj88QE2H7TMHcDHJJ0qaQLwz8BDEbG2aZm/lzRZ0nHANcD3OtRea+JkH32OBn5II9GfonHWvb1D2/4KcFn1Lf+XGL7k9nng9mqZP42I+2h8QXcXsIHGmfvgK43/BFYAj1fL3dqh9loTefAKa5ekVcCfRMSqNtcfS+PK44SIeKGTbbM385nd2iLpMBrf7LeV6NZ7PrNb3/jM3ltOdrNC+DLerBA97QjTfO+1jdxhhx2WjB933HEtY6+++mpy3Z07dybjuSu/XPzwww9vGZsyZUpy3d27dyfjGzduTMb379+fjI9WEaHhXq+V7JIuAL4KjKFxS+SNdbbXT9Kwx+e3+vlxZ/bs2cn417/+9ZaxO++8M7nu448/nozv2bMnGd+7d28yPnfu3Jaxiy66KLnu888/n4x/4QtfSMa3bt2ajJem7ct4SWOAfwU+CMyhUX+d06mGmVln1fnMfgbwXDXowR7guzR6L5nZAKqT7LN4YweGtbyxcwMAkhZJWi5peY19mVlNXf+CLiIWA4vBX9CZ9VOdM/s63thb6Rje2JPJzAZInWT/BXCKpBOqro6XAvd0pllm1mm17qCTtIBGT6gxNEY1uSGzfNcu4/tZOps3LzXYC1x66aXJ+MUXX5yM5+rFEydObBlL1bkBpk2blox306pV6dvqX3893TP3He94RzKeqsP/+Mc/Tq570003JeMrV65MxvupK3X2iLgXjypqdkjw7bJmhXCymxXCyW5WCCe7WSGc7GaFcLKbFaKnI9UM8u2yRxxxRDJ+2223tYydeuqpyXXf8pb039Rt27Yl47l+3aluprka/bhx45LxI488MhnfsWNHMp6qlXf7/15qHIDc/Qfjx49Pxh966KFk/IorrkjGu6lVnd1ndrNCONnNCuFkNyuEk92sEE52s0I42c0K4dJbZenSpcn48ccf3zK2efPm5Lq5rppjx6Y7H+7bty8Zz3XvTcmVBXOjy44ZM6Zr++6mul2iZ86cmYx/4AMfSMafffbZZLwOl97MCudkNyuEk92sEE52s0I42c0K4WQ3K4ST3awQPZ2yuZ9OP/30ZDxVRwd45ZVXWsZydfJcLTo3JfOsWW+aVesNhoaGWsZytezcLKy595brQpuqZ+e61+buL8h1DV67dm3b287Jve8rr7wyGf/sZz9ba//t8JndrBBOdrNCONnNCuFkNyuEk92sEE52s0I42c0KUUx/9lxd85prrknGU3X2XH/1XJ09V7P95je/mYyvX7++ZSxVawY4+uijk/ENGzYk43X6w0+YMCG57qRJk5Lx0047LRm/+uqrW8ZS/56Qv78gN/R4bv3Zs2cn43V0ZcpmSS8A24D9wL6ImF9ne2bWPZ24g+4PIyL9Z9LM+s6f2c0KUTfZA/iJpEclLRpuAUmLJC2XtLzmvsyshrqX8WdHxDpJRwH3S3o2Ih5sXiAiFgOLYbAHnDQb7Wqd2SNiXfV7E3AXcEYnGmVmndd2skuaKOmtBx4D5wMrO9UwM+ustuvskk6kcTaHxseB2yPihsw6fbuM/9nPfpaMH3XUUcl4qu90bmz1XL3417/+dTJ+5plnJuPnn39+y1iuL/y3v/3tZPyqq65KxleuTP99T02NnLv/YOPGjcn4ihUrkvHVq1e3jOX6wufGGMj1h3/nO9+ZjM+dO7dlbNWqVcl1czpeZ4+IXwLvartFZtZTLr2ZFcLJblYIJ7tZIZzsZoVwspsVopihpN/1rnThYM2aNcl4qitnrqtmTq67ZM59993XMrZjx47kunPmzEnGc12D77rrrmT8Qx/6UMtYrhvoY489loznhgdPlccmTpyYXDfX7TjXrflXv/pVMn7WWWe1jNUtvbXiM7tZIZzsZoVwspsVwsluVggnu1khnOxmhXCymxVi1NTZU10GAV5++eVkPNdlMdUdMzUtMaS7eQJs3rw5Gc9Jvfff/OY3yXVnzpyZjN9wQ7LXcva9p6aEzq2bqkWPRGqI7VzX37p19l27diXj55xzTsvYkiVLkuu2y2d2s0I42c0K4WQ3K4ST3awQTnazQjjZzQrhZDcrxKips1977bXJeK7WvX379mQ8VXfNbXv37t3JeK7GP39+enLcadOmtYxNnTo1ue64ceOS8RkzZiTjqTo6pN/7+PHjk+tOnjw5Gf/IRz6SjE+ZMqVlLFcHP/LII5Px3Pq595b7N+0Gn9nNCuFkNyuEk92sEE52s0I42c0K4WQ3K4ST3awQo6bO/sgjjyTjb3vb25Lxk08+ORlPje2eG4M8NXUw5PtO56abTvWtzvW7zu07N61ybuz3VJ/13L5TY/VDftrl1PjrQ0NDyXVz7zvXtlRfeoC77747Ge+G7Jld0i2SNkla2fTaVEn3S1pd/W5994KZDYSRXMbfClxw0GvXAQ9ExCnAA9VzMxtg2WSPiAeBVw96+ULgwNg5S4APd7hdZtZh7X5mnxERG6rHLwEtb6CWtAhY1OZ+zKxDan9BFxEhKRLxxcBigNRyZtZd7ZbeNkqaCVD93tS5JplZN7Sb7PcAC6vHC4EfdaY5ZtYtikhfWUu6A3gfMB3YCHweuBv4PnAc8CJwSUQc/CXecNsa2Mv4VN9ngFNOOaVl7BOf+ERy3XPPPTcZz80Nn+tbvXXr1paxXH/1XD25m3Ljxudq2blxAlLH7cknn0yue/nllyfjgywihj2w2c/sEXFZi9D7a7XIzHrKt8uaFcLJblYIJ7tZIZzsZoVwspsVYtR0ca1ry5YtyfiyZctaxnLTIp933nnJeK78mRuWONXFNlday3WBzcmVz1Lx3L4nTJiQjO/ZsycZP+yww1rGcl2iRyOf2c0K4WQ3K4ST3awQTnazQjjZzQrhZDcrhJPdrBDF1Nlz9eBcV9BUTTdXJ3/ttdeS8VwtPDfkcm7/KbnjUmfb3Vane26qW3An9p27h6Afx9VndrNCONnNCuFkNyuEk92sEE52s0I42c0K4WQ3K0QxdfZcXXPv3r1tb/v5559PxnN19ty0x7l+2ykjGCq81vo5ue2n5N537t6IlNy/SU5umOvcvRH94DO7WSGc7GaFcLKbFcLJblYIJ7tZIZzsZoVwspsVopg6e06duumuXbuS6+bqxbnx0fft25eMp+r0devodcaFh/Rxze07Nx7/0NBQMp5qW+6YjkbZM7ukWyRtkrSy6bXrJa2TtKL6WdDdZppZXSO5jL8VuGCY178cEfOqn3s72ywz67RsskfEg8CrPWiLmXVRnS/oPiXpieoyf0qrhSQtkrRc0vIa+zKzmtpN9m8AJwHzgA3AF1stGBGLI2J+RMxvc19m1gFtJXtEbIyI/RHxOvAt4IzONsvMOq2tZJc0s+npRcDKVsua2WDI1tkl3QG8D5guaS3weeB9kuYBAbwAXNXFNvZEnX7buTHC6477novn7hFIybW9ztjskK5159qde9+5ttep8ecM8nj6rWSTPSIuG+blm7vQFjPrIt8ua1YIJ7tZIZzsZoVwspsVwsluVgh3ce2BWbNmJeNbtmxJxnPlr1QZKFfeqjPUc7fl2p4b/jv13uqWFA9FPrObFcLJblYIJ7tZIZzsZoVwspsVwsluVggnu1khXGevdLPLYt1hi8ePH5+Mp7rQ1h0KuptDUee6qOamZM4NNZ1qW53pnnPbHlQ+s5sVwsluVggnu1khnOxmhXCymxXCyW5WCCe7WSFcZ++BXD0417c6V6dPrZ+rZefqxbm25aajTm0/NdV0bl2AnTt3JuMpkydPbnvdQ5XP7GaFcLKbFcLJblYIJ7tZIZzsZoVwspsVwsluVoiRTNl8LHAbMIPGFM2LI+KrkqYC3wNm05i2+ZKISA+AXqhcrbuuVJ/xuv2uuznufJ2+8CNZP3V/wuGHH55cN2e09mffB3wmIuYAZwKflDQHuA54ICJOAR6onpvZgMome0RsiIjHqsfbgGeAWcCFwJJqsSXAh7vVSDOr73f6zC5pNvBu4OfAjIjYUIVeonGZb2YDasT3xkuaBPwA+HREvNb8eSoiQtKwH2IkLQIW1W2omdUzojO7pHE0Ev07EfHD6uWNkmZW8ZnApuHWjYjFETE/IuZ3osFm1p5ssqtxCr8ZeCYivtQUugdYWD1eCPyo880zs04ZyWX8e4ArgCclrahe+xxwI/B9SR8DXgQu6U4TD3258lVd3SwD9bP0ltt3ndLb0NBQct3RKJvsEfEw0Opf9P2dbY6ZdYvvoDMrhJPdrBBOdrNCONnNCuFkNyuEk92sEB5KutLPLou54ZrrqNuNNKdO27vd/TY1lXU3j/mg8pndrBBOdrNCONnNCuFkNyuEk92sEE52s0I42c0K4Tp7pe6wxSm5aY272bc6N4x13emiu3nc6upmnX20DiVtZqOAk92sEE52s0I42c0K4WQ3K4ST3awQTnazQrjOPgDq9MuGdK07t+268Vwdv5/jyqe4P7uZjVpOdrNCONnNCuFkNyuEk92sEE52s0I42c0Kka2zSzoWuA2YAQSwOCK+Kul64K+Bl6tFPxcR93arod3Wzf7J69evT8bf/va3J+O5PuWpWneuDj5u3Li2tz2SeOq45u4fGDu23m0gqX2X2J99JEdzH/CZiHhM0luBRyXdX8W+HBE3da95ZtYp2WSPiA3AhurxNknPALO63TAz66zf6TO7pNnAu4GfVy99StITkm6RNKXFOoskLZe0vFZLzayWESe7pEnAD4BPR8RrwDeAk4B5NM78XxxuvYhYHBHzI2J+B9prZm0aUbJLGkcj0b8TET8EiIiNEbE/Il4HvgWc0b1mmlld2WRXo9vSzcAzEfGlptdnNi12EbCy880zs04Zybfx7wGuAJ6UtKJ67XPAZZLm0SjHvQBc1ZUWjgKTJ09OxidOnJiM50pQ06dPbxmr24U1V5qrI1d6y5XH1qxZk4ynhug+6aSTkuvm1O362w8j+Tb+YWC4TsmHbE3drES+g86sEE52s0I42c0K4WQ3K4ST3awQTnazQngo6Uo3px5+/PHHk/Gnn346Gd+6dWsyXqcWnqsXb9++PRnPHZfUca3TdRfyU2FPmTJsdw0Ali1bllw3ZxDr6Dk+s5sVwsluVggnu1khnOxmhXCymxXCyW5WCCe7WSHUyyFxJb0MvNj00nTglZ414HczqG0b1HaB29auTrbt+Ij4veECPU32N+1cWj6oY9MNatsGtV3gtrWrV23zZbxZIZzsZoXod7Iv7vP+Uwa1bYPaLnDb2tWTtvX1M7uZ9U6/z+xm1iNOdrNC9CXZJV0g6X8lPSfpun60oRVJL0h6UtKKfs9PV82ht0nSyqbXpkq6X9Lq6nfrTtu9b9v1ktZVx26FpAV9atuxkv5b0tOSnpL0t9XrfT12iXb15Lj1/DO7pDHAKuCPgLXAL4DLIiI9gkOPSHoBmB8Rfb8BQ9J7ge3AbRExt3rtX4BXI+LG6g/llIi4dkDadj2wvd/TeFezFc1snmYc+DDwUfp47BLtuoQeHLd+nNnPAJ6LiF9GxB7gu8CFfWjHwIuIB4FXD3r5QmBJ9XgJjf8sPdeibQMhIjZExGPV423AgWnG+3rsEu3qiX4k+yyged6etQzWfO8B/ETSo5IW9bsxw5gRERuqxy8BM/rZmGFkp/HupYOmGR+YY9fO9Od1+Qu6Nzs7Ik4DPgh8srpcHUjR+Aw2SLXTEU3j3SvDTDP+W/08du1Of15XP5J9HXBs0/NjqtcGQkSsq35vAu5i8Kai3nhgBt3q96Y+t+e3Bmka7+GmGWcAjl0/pz/vR7L/AjhF0gmSxgOXAvf0oR1vImli9cUJkiYC5zN4U1HfAyysHi8EftTHtrzBoEzj3Wqacfp87Po+/XlE9PwHWEDjG/nngX/oRxtatOtE4H+qn6f63TbgDhqXdXtpfLfxMWAa8ACwGlgKTB2gtv0H8CTwBI3Emtmntp1N4xL9CWBF9bOg38cu0a6eHDffLmtWCH9BZ1YIJ7tZIZzsZoVwspsVwsluVggnu1khnOxmhfh/EdMNwZWEwv4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "idx = 1\n",
    "\n",
    "item_img = images[idx]\n",
    "item_npimg = item_img.squeeze().numpy()\n",
    "plt.title(CLASSES[labels[idx].item()])\n",
    "print(item_npimg.shape)\n",
    "plt.imshow(item_npimg, cmap='gray')\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}