11-图像分类数据集.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "fe3e6df5",
   "metadata": {},
   "outputs": [],
   "source": [
    "%config Completer.use_jedi = False\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "9d75727f",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torchvision\n",
    "from torch.utils import data\n",
    "from torchvision import transforms\n",
    "from d2l import torch as d2l\n",
    "\n",
    "d2l.use_svg_display()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "17c24781",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 读取数据集\n",
    "# 通过ToTensor实例将图像数据从PIL类型变换成32位浮点数格式\n",
    "# 并除以255使得所有像素的数值均在0到1之间\n",
    "trans = transforms.ToTensor()\n",
    "mnist_train = torchvision.datasets.FashionMNIST(root=\"../data\", train=True,transform=trans,download=True)\n",
    "mnist_test = torchvision.datasets.FashionMNIST(root=\"../data\", train=False,transform=trans, download=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "47f473f3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(60000, 10000)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(mnist_train), len(mnist_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "42a92d81",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "\"\"\"\n",
    "Fashion-MNIST中包含的10个类别分别为t-shirt（T恤）、trouser（裤子）、pullover（套衫）、dress（连衣裙）\n",
    "coat（外套）、sandal（凉鞋）、shirt（衬衫）、sneaker（运动鞋）、bag（包）和ankle boot（短靴）。\n",
    "以下函数用于在数字标签索引及其文本名称之间进行转换。\n",
    "\"\"\"\n",
    "def get_fashion_mnist_labels(labels): \n",
    "    \"\"\"返回Fashion-MNIST数据集的文本标签。\"\"\"\n",
    "    text_labels = [\n",
    "        't-shirt', 'trouser', 'pullover', 'dress', 'coat', 'sandal', 'shirt',\n",
    "        'sneaker', 'bag', 'ankle boot']\n",
    "    return [text_labels[int(i)] for i in labels]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "a1d550de",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 创建一个函数来可视化这些样本。\n",
    "def show_images(imgs, num_rows, num_cols, titles=None, scale=1.1):\n",
    "    \"\"\"Plot a list of images.\"\"\"\n",
    "    figsize = (num_cols * scale, num_rows * scale)\n",
    "    _, axes = d2l.plt.subplots(num_rows, num_cols, figsize=figsize)\n",
    "    axes = axes.flatten()\n",
    "    for i, (ax, img) in enumerate(zip(axes, imgs)):\n",
    "        if torch.is_tensor(img):\n",
    "            # 图片张量\n",
    "            ax.imshow(img.numpy())\n",
    "        else:\n",
    "            # PIL图片\n",
    "            ax.imshow(img)\n",
    "        ax.axes.get_xaxis().set_visible(False)\n",
    "        ax.axes.get_yaxis().set_visible(False)\n",
    "        if titles:\n",
    "            ax.set_title(titles[i])\n",
    "    return axes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "04711e83",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([5, 8, 4, 2, 0, 9])"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 以下是训练数据集中前几个样本的图像及其相应的标签（文本形式）\n",
    "X, y = next(iter(data.DataLoader(mnist_train,shuffle=True, batch_size=6)))\n",
    "#show_images(X.reshape(6, 28, 28), 2, 3, titles=get_fashion_mnist_labels(y));\n",
    "y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "9159f1fe",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 读取小批量\n",
    "batch_size = 256\n",
    "\n",
    "def get_dataloader_workers():  #@save\n",
    "    \"\"\"使用4个进程来读取数据。\"\"\"\n",
    "    return 4\n",
    "\n",
    "train_iter = data.DataLoader(mnist_train, batch_size, shuffle=True,num_workers=4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "43eae658",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'6.74 sec'"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 看一下读取数据的时间\n",
    "timer = d2l.Timer()\n",
    "for X, y in train_iter:\n",
    "    continue\n",
    "f'{timer.stop():.2f} sec'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "c2e20065",
   "metadata": {},
   "outputs": [],
   "source": [
    " # 完整的读取例子\n",
    "def load_data_fashion_mnist(batch_size, resize=None): \n",
    "    \"\"\"下载Fashion-MNIST数据集，然后将其加载到内存中。\"\"\"\n",
    "    trans = [transforms.ToTensor()]\n",
    "    if resize:\n",
    "        trans.insert(0, transforms.Resize(resize))\n",
    "    trans = transforms.Compose(trans)\n",
    "    # 训练集\n",
    "    train_dataset = torchvision.datasets.FashionMNIST(root=\"../data\",train=True,transform=trans,download=True)\n",
    "    train_dataloader = data.DataLoader(train_dataset, batch_size, shuffle=True,num_workers=get_dataloader_workers())\n",
    "    # 测试集\n",
    "    test_dataset = torchvision.datasets.FashionMNIST(root=\"../data\",train=False,transform=trans,download=True)\n",
    "    test_dataloader = data.DataLoader(test_dataset, batch_size, shuffle=False,num_workers=get_dataloader_workers())\n",
    "    return (train_dataloader,test_dataloader)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "2abd49f8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([6, 1, 28, 28]) torch.float32 torch.Size([6]) torch.int64\n"
     ]
    }
   ],
   "source": [
    "train_iter, test_iter = load_data_fashion_mnist(6, resize=None)\n",
    "for X, y in train_iter:\n",
    "    print(X.shape, X.dtype, y.shape, y.dtype)\n",
    "    break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "7c16b409",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([1, 1, 4, 5, 3, 5])"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "525a4aaa",
   "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}