{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 5.2 填充和步幅"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.4.1\n"
     ]
    }
   ],
   "source": [
    "import torch\n",
    "from torch import nn\n",
    "\n",
    "print(torch.__version__)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5.2.1 填充"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([8, 8])"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 定义一个函数来计算卷积层。它对输入和输出做相应的升维和降维\n",
    "def comp_conv2d(conv2d, X):\n",
    "    # (1, 1)代表批量大小和通道数（“多输入通道和多输出通道”一节将介绍）均为1\n",
    "    X = X.view((1, 1) + X.shape)\n",
    "    Y = conv2d(X)\n",
    "    return Y.view(Y.shape[2:])  # 排除不关心的前两维：批量和通道\n",
    "\n",
    "# 注意这里是两侧分别填充1行或列，所以在两侧一共填充2行或列\n",
    "conv2d = nn.Conv2d(in_channels=1, out_channels=1, kernel_size=3, padding=1)\n",
    "\n",
    "X = torch.rand(8, 8)\n",
    "comp_conv2d(conv2d, X).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([8, 8])"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 使用高为5、宽为3的卷积核。在高和宽两侧的填充数分别为2和1\n",
    "conv2d = nn.Conv2d(in_channels=1, out_channels=1, kernel_size=(5, 3), padding=(2, 1))\n",
    "comp_conv2d(conv2d, X).shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5.2.2 步幅"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([4, 4])"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "conv2d = nn.Conv2d(1, 1, kernel_size=3, padding=1, stride=2)\n",
    "comp_conv2d(conv2d, X).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([2, 2])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "conv2d = nn.Conv2d(1, 1, kernel_size=(3, 5), padding=(0, 1), stride=(3, 4))\n",
    "comp_conv2d(conv2d, X).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [default]",
   "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
}