00_tryTtips.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[-4.5704e+02,  9.4868e-43,  1.5461e+17],\n",
       "        [ 5.8855e-44, -4.6584e+00,  9.4868e-43],\n",
       "        [ 4.3440e-44,  9.4868e-43, -3.1893e+03]])"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import torch as t\n",
    "a = t.Tensor(3,3)\n",
    "a"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[ 1.3733e-14,  1.2102e+25,  1.6992e-07,  2.9599e+21,  2.8175e+20,\n",
       "          1.7566e+25,  1.7748e+28,  0.0000e+00,  0.0000e+00]])"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# view()函数，将多行Tensor，拼接成一行\n",
    "a.view(1,-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[[ 0.,  0.,  0.,  0.]],\n",
      "\n",
      "        [[ 0.,  0.,  0.,  0.]],\n",
      "\n",
      "        [[ 0.,  0.,  0.,  0.]]]) \n",
      "\n",
      "\t torch.Size([3, 1, 4])\n"
     ]
    }
   ],
   "source": [
    "# squeeze() 和 unsqueeze()\n",
    "b = t.Tensor(3,1,4)\n",
    "print(b,'\\n\\n\\t',b.size())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[[ 0.,  0.,  0.,  0.]],\n",
      "\n",
      "        [[ 0.,  0.,  0.,  0.]],\n",
      "\n",
      "        [[ 0.,  0.,  0.,  0.]]]) \n",
      "\n",
      "\t torch.Size([3, 1, 4])\n"
     ]
    }
   ],
   "source": [
    "b = t.Tensor(3,1,4).squeeze(0)# 参数0代表第一维度,squeeze(0)如果第一维度值为1，则去掉，否则不变\n",
    "print(b,'\\n\\n\\t',b.size())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([3, 1, 4])"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t.Tensor(3,1,4).squeeze(0).size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[[ 0.,  0.,  0.,  0.]],\n",
      "\n",
      "        [[ 0.,  0.,  0.,  0.]],\n",
      "\n",
      "        [[ 0.,  0.,  0.,  0.]]]) \n",
      "\n",
      "\t torch.Size([3, 1, 4])\n"
     ]
    }
   ],
   "source": [
    "b.squeeze(1)# 参数1代表第二维度\n",
    "print(b,'\\n\\n\\t',b.size())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[[-0.1081,  0.9290, -0.0519],\n",
       "         [ 0.0837, -1.1818,  1.6718]]])"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "b.unsqueeze(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[[-0.1081,  0.9290, -0.0519]],\n",
       "\n",
       "        [[ 0.0837, -1.1818,  1.6718]]])"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "b.unsqueeze(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "b.numel()# 返回tensor中元素的个数"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[ 0.91498251,  0.62433221,  0.969374  ],\n",
       "        [ 0.69164461,  0.86542409,  0.93935447],\n",
       "        [ 0.9907821 ,  0.02126888,  0.83379586]],\n",
       "\n",
       "       [[ 0.61679697,  0.27839374,  0.66914182],\n",
       "        [ 0.32646236,  0.22425935,  0.31222748],\n",
       "        [ 0.13508763,  0.58617435,  0.74950824]]])"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "x = np.random.rand(2,3,3)\n",
    "x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([ 1.8631,  1.6089,  1.8233])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t.Tensor(3).uniform_(1,2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "ename": "ValueError",
     "evalue": "invalid literal for int() with base 10: 'hdafdasfafaadj2i'",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-18-21c46473776c>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[0;32m      1\u001b[0m \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;34m'root/hdafdasfafaadj2i.jpg'\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      2\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 3\u001b[1;33m \u001b[0mb\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msplit\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'.'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;33m-\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msplit\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'/'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;33m-\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      4\u001b[0m \u001b[0mb\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mValueError\u001b[0m: invalid literal for int() with base 10: 'hdafdasfafaadj2i'"
     ]
    }
   ],
   "source": [
    "x = 'root/hdafdasfafaadj2i.jpg'\n",
    "\n",
    "b = int(x.split('.')[-2].split('/')[-1])\n",
    "b"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'faadj2i'"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x = 'root/hdafdasfa.faadj2i.jpg'\n",
    "b = x.split('.')[-2].split('/')[-1]\n",
    "b"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "原始tensor_1:\n",
      " tensor([ 2.,  3.,  4.,  5.,  6.,  7.])\n",
      "\n",
      "tensor_1大小 torch.Size([6])\n",
      "\n",
      "改变tensor_1布局，得tensor_2:\n",
      " tensor([[ 2.,  3.,  4.],\n",
      "        [ 5.,  6.,  7.]])\n",
      "此时打印tensor_1:\n",
      " tensor([ 2.,  3.,  4.,  5.,  6.,  7.])\n",
      "\n",
      "改变tensor_2一个值:\n",
      " tensor([[ 2.,  3.,  4.],\n",
      "        [ 5.,  6.,  8.]])\n",
      "此时打印tensor_1:\n",
      " tensor([ 2.,  3.,  4.,  5.,  6.,  8.])\n"
     ]
    }
   ],
   "source": [
    "a = t.arange(2,8)\n",
    "print(\"原始tensor_1:\\n\",a)\n",
    "print(\"\\ntensor_1大小\",a.size())\n",
    "\n",
    "b = a.view(2,3)\n",
    "print(\"\\n改变tensor_1布局，得tensor_2:\\n\",b)\n",
    "print(\"此时打印tensor_1:\\n\",a)\n",
    "\n",
    "b[1,2]=8\n",
    "print(\"\\n改变tensor_2一个值:\\n\",b)\n",
    "print(\"此时打印tensor_1:\\n\",a)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[  0.,   1.,   2.,   3.],\n",
       "        [  4.,   5.,   6.,   7.],\n",
       "        [  8.,   9.,  10.,  11.],\n",
       "        [ 12.,  13.,  14.,  15.]])"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a = t.arange(0, 16).view(4, 4)\n",
    "a"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[ 0],\n",
       "        [ 1],\n",
       "        [ 2],\n",
       "        [ 3]])"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "di = t.LongTensor([[0,1,2,3]]).t()\n",
    "di"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[  0.,   5.,  10.,  15.]])"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 选取对角线的元素\n",
    "index = t.LongTensor([[0,1,2,3]])\n",
    "a.gather(0, index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[  3.],\n",
       "        [  6.],\n",
       "        [  9.],\n",
       "        [ 12.]])"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 选取反对角线上的元素\n",
    "index = t.LongTensor([[3,2,1,0]]).t()\n",
    "a.gather(1, index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[[ 1.,  1.],\n",
       "         [ 1.,  1.],\n",
       "         [ 1.,  1.]],\n",
       "\n",
       "        [[ 1.,  1.],\n",
       "         [ 1.,  1.],\n",
       "         [ 1.,  1.]]])"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a = t.ones(3, 2)\n",
    "b = t.zeros(2, 3,1)\n",
    "\n",
    "# 自动广播法则\n",
    "# 第一步：a是2维,b是3维，所以先在较小的a前面补1 ，\n",
    "# 即：a.unsqueeze(0)，a的形状变成（1，3，2），b的形状是（2，3，1）,\n",
    "# 第二步: a和b在第一维和第三维形状不一样，其中一个为1 ，\n",
    "# 可以利用广播法则扩展，两个形状都变成了（2，3，2）\n",
    "a+b"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[[ 1.,  1.],\n",
       "         [ 1.,  1.],\n",
       "         [ 1.,  1.]],\n",
       "\n",
       "        [[ 1.,  1.],\n",
       "         [ 1.,  1.],\n",
       "         [ 1.,  1.]]])"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 手动广播法则\n",
    "# 或者 a.view(1,3,2).expand(2,3,2)+b.expand(2,3,2)\n",
    "a.unsqueeze(0).expand(2, 3, 2) + b.expand(2,3,2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[ 1.6229],\n",
      "        [ 1.9575],\n",
      "        [ 1.4342],\n",
      "        [ 1.9799],\n",
      "        [ 1.2583],\n",
      "        [ 1.1112],\n",
      "        [ 1.9413],\n",
      "        [ 1.1765]])\n",
      "tensor([[ 4.8688],\n",
      "        [ 5.8725],\n",
      "        [ 4.3025],\n",
      "        [ 5.9396],\n",
      "        [ 3.7749],\n",
      "        [ 3.3335],\n",
      "        [ 5.8240],\n",
      "        [ 3.5294]])\n"
     ]
    }
   ],
   "source": [
    "a = t.rand(8,1).uniform_(1,2)\n",
    "print(a)\n",
    "print(a*3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[ 1.5830,  1.0162],\n",
       "        [ 1.5806,  1.1798]])"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "def f(x):\n",
    "    '''计算y'''\n",
    "    y = x**2 * t.exp(x)\n",
    "    return y\n",
    "def gradf(x):\n",
    "    '''手动求导函数'''\n",
    "    dx = 2*x*t.exp(x) + x**2*t.exp(x)\n",
    "    return dx\n",
    "\n",
    "from torch.autograd import Variable as V\n",
    "x = V(t.Tensor(2,2).uniform_(1,2), requires_grad = True)\n",
    "x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[ 12.2015,   2.8532],\n",
       "        [ 12.1359,   4.5288]])"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y = f(x)\n",
    "y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[ 27.6176,   8.4684],\n",
       "        [ 27.4921,  12.2061]])"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y.backward(t.ones(y.size())) # grad_variables形状与y一致\n",
    "x.grad"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([ 26.7246])"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x = t.Tensor([1.5658])\n",
    "c = 2*x*t.exp(x) + x**2*t.exp(x)\n",
    "c"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "26.724593728555057"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x = 1.5658\n",
    "c = 2*x*(2.71828**x) + x**2*(2.71828**x)\n",
    "c"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.2.4 小试牛刀: 用Variable实现线性回归\n",
    "在上一节中讲解了利用tensor实现线性回归，在这一小节中，将讲解如何利用autograd/Variable实现线性回归，以此感受autograd的便捷之处。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import torch as t\n",
    "import time\n",
    "from torch.autograd import Variable as V\n",
    "%matplotlib inline\n",
    "from matplotlib import pyplot as plt\n",
    "from IPython import display"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 设置随机数种子，为了在不同人电脑上运行时下面的输出一致\n",
    "t.manual_seed(1000)\n",
    "def get_fake_data(batch_size=8):\n",
    "    ''' 产生随机数据：y = x*2 + 3，加上了一些噪声'''\n",
    "    x = t.rand(batch_size,1) * 20\n",
    "    y = x * 2 + (1 + t.randn(batch_size, 1))*3\n",
    "    return x, y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x272320a2e48>"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAD8CAYAAABn919SAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x27230e586d8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 来看看产生x-y分布是什么样的\n",
    "x, y = get_fake_data()\n",
    "plt.scatter(x.squeeze().numpy(), y.squeeze().numpy())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAD8CAYAAAB0IB+mAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIhFCgS8iEiH+D6kxd7GLKah9AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x272320d1fd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "ename": "NameError",
     "evalue": "name 'time' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-95-e7ed419f7b8d>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[0;32m     29\u001b[0m         \u001b[0mplt\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mylim\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m41\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     30\u001b[0m         \u001b[0mplt\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mshow\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 31\u001b[1;33m         \u001b[0mplt\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpause\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m0.5\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     32\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     33\u001b[0m \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mw\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdata\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msqueeze\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mb\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdata\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msqueeze\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32mC:\\ProgramData\\Anaconda3\\lib\\site-packages\\matplotlib\\pyplot.py\u001b[0m in \u001b[0;36mpause\u001b[1;34m(interval)\u001b[0m\n\u001b[0;32m    292\u001b[0m         \u001b[0mshow\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mblock\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mFalse\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    293\u001b[0m         \u001b[0mcanvas\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstart_event_loop\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0minterval\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 294\u001b[1;33m     \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    295\u001b[0m         \u001b[0mtime\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msleep\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0minterval\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    296\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mNameError\u001b[0m: name 'time' is not defined"
     ]
    }
   ],
   "source": [
    "# 随机初始化参数\n",
    "w = V(t.rand(1,1), requires_grad=True)\n",
    "b = V(t.zeros(1,1), requires_grad=True)\n",
    "lr =0.001 # 学习率\n",
    "for ii in range(8000):\n",
    "    x, y = get_fake_data()\n",
    "    x, y = V(x), V(y)\n",
    "    # forward：计算loss\n",
    "    y_pred = x.mm(w) + b.expand_as(y)\n",
    "    loss = 0.5 * (y_pred - y) ** 2\n",
    "    loss = loss.sum()\n",
    "    # backward：手动计算梯度\n",
    "    loss.backward()\n",
    "    # 更新参数\n",
    "    w.data.sub_(lr * w.grad.data)\n",
    "    b.data.sub_(lr * b.grad.data)\n",
    "    # 梯度清零\n",
    "    w.grad.data.zero_()\n",
    "    b.grad.data.zero_()\n",
    "    if ii%1000 ==0:\n",
    "        # 画图\n",
    "        display.clear_output(wait=True)\n",
    "        x = t.arange(0, 20).view(-1, 1)\n",
    "        y = x.mm(w.data) + b.data.expand_as(x)\n",
    "        plt.plot(x.numpy(), y.numpy()) # predicted\n",
    "        x2, y2 = get_fake_data(batch_size=20)\n",
    "        plt.scatter(x2.numpy(), y2.numpy()) # true data\n",
    "        plt.xlim(0,20)\n",
    "        plt.ylim(0,41)\n",
    "        plt.show()\n",
    "        plt.pause(0.5)\n",
    "        \n",
    "print(w.data.squeeze()[0], b.data.squeeze()[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# 全连接层（神经网络工具箱）"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import torch as t\n",
    "from torch import nn\n",
    "from torch.autograd import Variable as V"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "class Linear(nn.Module): # 继承nn.Module\n",
    "    def __init__(self, in_features, out_features):\n",
    "        super(Linear, self).__init__() # 等价于nn.Module.__init__(self)\n",
    "        self.w = nn.Parameter(t.randn(in_features, out_features))# 将w封装成Parameter\n",
    "        self.b = nn.Parameter(t.randn(out_features))# 将b封装成Parameter\n",
    "    \n",
    "    def forward(self, x):\n",
    "        x = x.mm(self.w) # x.@(self.w) 这段什么意思\n",
    "        return x + self.b.expand_as(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[-1.8143,  0.4692, -0.4410],\n",
       "        [-0.7742, -0.7019, -1.1366]])"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "layer = Linear(4,3)# 4个输入，3个输出\n",
    "input = V(t.randn(2,4))# 4个输入，输入了2次\n",
    "output = layer(input)\n",
    "output"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "w Parameter containing:\n",
      "tensor([[ 2.1541, -1.5201, -0.0853],\n",
      "        [-0.4178, -1.6190, -0.7864],\n",
      "        [-0.1170,  0.6966, -0.2373],\n",
      "        [-0.4831, -0.8131, -0.8169]])\n",
      "b Parameter containing:\n",
      "tensor([-0.4744, -0.3817,  0.2188])\n"
     ]
    }
   ],
   "source": [
    "for name, parameter in layer.named_parameters():\n",
    "    print(name, parameter) # w and b "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 多层感知机"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "class Perceptron(nn.Module):\n",
    "    def __init__(self, in_features, hidden_features, out_features):\n",
    "        nn.Module.__init__(self)\n",
    "        self.layer1 = Linear(in_features, hidden_features) # 此处的Linear是前面自定义的全连接层\n",
    "        self.layer2 = Linear(hidden_features, out_features)\n",
    "    def forward(self,x):\n",
    "        x = self.layer1(x)\n",
    "        x = t.sigmoid(x)\n",
    "        return self.layer2(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "layer1.w torch.Size([3, 4])\n",
      "layer1.b torch.Size([4])\n",
      "layer2.w torch.Size([4, 1])\n",
      "layer2.b torch.Size([1])\n"
     ]
    }
   ],
   "source": [
    "perceptron = Perceptron(3,4,1)# 3个in 4个hi 1个out\n",
    "for name, param in perceptron.named_parameters():\n",
    "    print(name, param.size())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.1 常用神经网络层\n",
    "#### 4.1.1 图像相关层\n",
    "\n",
    "图像相关层主要包括卷积层（Conv）、池化层（Pool）等，这些层在实际使用中可分为一维(1D)、二维(2D)、三维（3D），池化方式又分为平均池化（AvgPool）、最大值池化（MaxPool）、自适应池化（AdaptiveAvgPool）等。而卷积层除了常用的前向卷积之外，还有逆卷积（TransposeConv）。下面举例说明一些基础的使用。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAMAAAAEHCAAAAADdLRZjAAB3qUlEQVR4nJz9d5CkWZIfiLn7e+8T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\njbL/BeiX53mrl5xTAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<PIL.PngImagePlugin.PngImageFile image mode=L size=192x263 at 0x2A5C723FDD8>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from PIL import Image\n",
    "from torchvision.transforms import ToTensor, ToPILImage\n",
    "to_tensor = ToTensor() # img -> tensor\n",
    "to_pil = ToPILImage()\n",
    "anni = Image.open('C:/Users/yefci/Desktop/lab/sgaliGray.png')\n",
    "anni"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAL4AAAEFCAAAAAB4ZQJDAAB0jElEQVR4nJT9abBlSXIeiLl7RJzt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\nOr+k/wEGJ6oBIeI/gQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<PIL.Image.Image image mode=L size=190x261 at 0x2A5C7A35940>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 卷积层\n",
    "\n",
    "# 输入是一个batch，batch_size＝1\n",
    "input = to_tensor(anni).unsqueeze(0)\n",
    "\n",
    "# 定义锐化卷积核\n",
    "kernel = t.ones(3, 3)/-9.\n",
    "kernel[1][1] = 1\n",
    "\n",
    "conv = nn.Conv2d(1, 1, (3, 3), 1, bias=False)\n",
    "conv.weight.data = kernel.view(1, 1, 3, 3)\n",
    "out = conv(V(input))\n",
    "to_pil(out.data.squeeze(0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAGAAAACDCAAAAABXI0HIAAAiOklEQVR4nHV6WZCk2XXWOXf5t9wr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\n3hkM7sxfhDTwRxOXQqUBrvCH9P8B7D721/TjUdgAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<PIL.Image.Image image mode=L size=96x131 at 0x2A5C5959A58>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 池化层\n",
    "\n",
    "pool = nn.AvgPool2d(2,2)# 变小了\n",
    "list(pool.parameters())\n",
    "\n",
    "out = pool(V(input))\n",
    "to_pil(out.data.squeeze(0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[-0.0551,  0.1574,  0.3943,  1.2670],\n",
       "        [ 0.9518,  0.0048, -0.3682, -0.4939]])"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 输入 batch_size=2，维度3\n",
    "input = V(t.randn(2, 3))\n",
    "linear = nn.Linear(3, 4)\n",
    "h = linear(input)\n",
    "h"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(tensor([ 0.,  0.,  0.,  0.]),\n",
       " tensor([ 15.9994,  15.9726,  15.9989,  15.9998]))"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 4 channel，初始化标准差为4，均值为0\n",
    "bn = nn.BatchNorm1d(4)\n",
    "bn.weight.data = t.ones(4) * 4\n",
    "bn.bias.data = t.zeros(4)\n",
    "bn_out = bn(h)\n",
    "# 注意输出的均值和方差\n",
    "# 方差是标准差的平方，计算无偏方差分母会减1\n",
    "# 使用unbiased=False 分母不减1\n",
    "bn_out.mean(0), bn_out.var(0, unbiased=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[-7.9998,  7.9931,  7.9997,  7.9999],\n",
       "        [ 7.9998, -0.0000, -7.9997, -7.9999]])"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 每个元素以0.5的概率舍弃\n",
    "dropout = nn.Dropout(0.5)\n",
    "o = dropout(bn_out)\n",
    "o # 有一半左右的数变为0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 激活函数ReLU"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[-0.2305, -0.3784,  0.4586],\n",
      "        [ 2.1690, -0.6063,  2.2945]])\n",
      "tensor([[ 0.0000,  0.0000,  0.4586],\n",
      "        [ 2.1690,  0.0000,  2.2945]])\n"
     ]
    }
   ],
   "source": [
    "relu = nn.ReLU(inplace=True)\n",
    "input = V(t.randn(2, 3))\n",
    "print(input)\n",
    "output = relu(input)\n",
    "print(output) # 小于0的都被截断为0\n",
    "# 等价于input.clamp(min=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "net1: Sequential(\n",
      "  (conv): Conv2d(3, 3, kernel_size=(3, 3), stride=(1, 1))\n",
      "  (batchnorm): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "  (activation_layer): ReLU()\n",
      ")\n",
      "net2: Sequential(\n",
      "  (0): Conv2d(3, 3, kernel_size=(3, 3), stride=(1, 1))\n",
      "  (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "  (2): ReLU()\n",
      ")\n",
      "net3: Sequential(\n",
      "  (conv1): Conv2d(3, 3, kernel_size=(3, 3), stride=(1, 1))\n",
      "  (bn1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "  (relu1): ReLU()\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "# Sequential的三种写法\n",
    "net1 = nn.Sequential()\n",
    "net1.add_module('conv', nn.Conv2d(3, 3, 3))\n",
    "net1.add_module('batchnorm', nn.BatchNorm2d(3))\n",
    "net1.add_module('activation_layer', nn.ReLU())\n",
    "\n",
    "net2 = nn.Sequential(\n",
    "    nn.Conv2d(3, 3, 3),\n",
    "    nn.BatchNorm2d(3),\n",
    "    nn.ReLU()\n",
    ")\n",
    "\n",
    "from collections import OrderedDict\n",
    "net3= nn.Sequential(OrderedDict([\n",
    "    ('conv1', nn.Conv2d(3, 3, 3)),\n",
    "    ('bn1', nn.BatchNorm2d(3)),\n",
    "    ('relu1', nn.ReLU())\n",
    "]))\n",
    "\n",
    "print('net1:', net1)\n",
    "print('net2:', net2)\n",
    "print('net3:', net3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(Conv2d(3, 3, kernel_size=(3, 3), stride=(1, 1)),\n",
       " Conv2d(3, 3, kernel_size=(3, 3), stride=(1, 1)),\n",
       " Conv2d(3, 3, kernel_size=(3, 3), stride=(1, 1)))"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 可根据名字或序号取出子module\n",
    "net1.conv, net2[0], net3.conv1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "input = V(t.rand(1, 3, 4, 4))\n",
    "output = net1(input)\n",
    "output = net2(input)\n",
    "output = net3(input)\n",
    "output = net3.relu1(net1.batchnorm(net1.conv(input)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "modellist = nn.ModuleList([nn.Linear(3,4), nn.ReLU(), nn.Linear(4,2)])\n",
    "input = V(t.randn(1, 3))\n",
    "for model in modellist:\n",
    "    input = model(input)\n",
    "# 下面会报错,因为modellist没有实现forward方法\n",
    "# output = modelist(input)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 激活函数"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import torch as t\n",
    "from torch.autograd import Variable\n",
    "%matplotlib inline\n",
    "from matplotlib import pyplot as plt\n",
    "from torch import nn\n",
    "import torch.nn.functional as F"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x24e45754d30>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXYAAAD8CAYAAABjAo9vAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nhMz/A/Gfd7W0c2OpAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24e457322b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# fake data\n",
    "x = t.linspace(-5, 5, 200)# x data(tensor), shape=(100,1)\n",
    "x = Variable(x)\n",
    "x_np = x.data.numpy()\n",
    "\n",
    "y_relu = F.relu(x).data.numpy()\n",
    "y_sigmoid = F.sigmoid(x).data.numpy()\n",
    "y_tanh = F.tanh(x).data.numpy()\n",
    "y_softplus = F.softplus(x).data.numpy()\n",
    "\n",
    "plt.subplot(221)\n",
    "plt.plot(x_np, y_relu, c='red', label='relu')\n",
    "plt.ylim((-1, 5))\n",
    "plt.legend(loc='best')\n",
    "\n",
    "plt.subplot(222)\n",
    "plt.plot(x_np, y_sigmoid, c='red', label='sigmoid')\n",
    "plt.ylim((-0.2, 1.2))\n",
    "plt.legend(loc='best')\n",
    "\n",
    "plt.subplot(223)\n",
    "plt.plot(x_np, y_tanh, c='red', label='tanh')\n",
    "plt.ylim((-1.2, 1.2))\n",
    "plt.legend(loc='best')\n",
    "\n",
    "plt.subplot(224)\n",
    "plt.plot(x_np, y_softplus, c='red', label='softplus')\n",
    "plt.ylim((-0.2, 1.2))\n",
    "plt.legend(loc='best')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAD8CAYAAAB0IB+mAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAIABJREFUeJztnXtsXPd15z9nHhw+hpJIamjJomjJ\ntOTYidcvxqsiQdZO7I3tJlaapIW7aes0AYxuEjRBt6jtGttu0waoW6BNdrdtIDRFXCCp6yYtbDfJ\npnaaNEggx5H8iO3IskU/ZImyRFGU+JzhPM7+MXeoS2oeJOdx7517PgDBmXvmzjkc3vnOmfM791xR\nVQzDMIz2J+J1AIZhGEZrMME3DMMICSb4hmEYIcEE3zAMIySY4BuGYYQEE3zDMIyQYIJvGIYREkzw\nDcMwQoIJvmEYRkiIeR2Am82bN+uOHTu8DsMwDCNQHDx48LSqpmo9zleCv2PHDg4cOOB1GIZhGIFC\nRN5YzeOspGMYhhESTPANwzBCggm+YRhGSDDBNwzDCAkm+IZhGCHBBN8wDCMk+Kot02g+YxOzPDk2\niQJDfV0cn1pYup0vKNGILNtWy54vKDs29zCSSnr6dxmNYWxiltdPzxGNyLqOh3J2O0b8gwl+m7BS\nyCu9Gb9x8BhvTM6RLygI4FzhMhoVtvR2cmI6TaHgbKxhj0aFLRs6ScSi/Mo7hxjs7awoAL8wMmBv\neI+pdYxEI/DwT4+RzuY5PZshlUys6XgoZ5+YzZDqTdgx4hMaJvgiEgUOAMdV9QMishN4COgHngZ+\nXVUXG+UvjFTKvk7NZPjhyxNLQl5JnHviUcbPpVGFQpnnX8jkWMhVvsZxOfv0QhYBXjk1zXBfT1kB\niEaEHZt7+Mh1Q2wrkwVa9tc43MdItQ/7csdINl8gmy+gCnmF2XR2zcfDSns6p0zNL9ox4hMameF/\nFjgEbHDuPwD8pao+JCJfBj4J/E0D/bU97oysWvaVzRdI5wrgEvJyb8a5aJZ8lWvWp6u8eSvZc47D\nbKbAa6dnywpABHju2FkmptPMLuaXfQjt6O+hqyPKO3f0s7k3YW/wNbAyATg1k146Rt5axYd9LcFe\nz/Gw0q7Uf4yU+xZpx8j6aIjgi8gQ8IvAF4DfEREB3gv8N+chDwL/CxP8VTM2McsD3znE4bdmyebz\nzKZzZAtKQZVcXklnc8xmCkvfole+hTJl3lTZfHWf1d++te3lfIIjMAU4M5dhPquI6/kiMsdcJs+h\nE9N0xiMowtVDG+lJxPjUTZfZG7oCYxOz/PX3jzCXyfHzEzMM93fx4vg06WyegiqZnC51ZCiQiecu\n+LCv9P8qUe/xUM6+1mOkAMxn8ojAn/2/w7z94g0cPbPAlVt77RhZB43K8L8I/B7Q69wfAM6qas65\nfwzYVm5HEbkbuBtgeHi4QeEEE3fG9vQbU7w5NU8mlyedzbOYL5BXUFUKFLOm0lun3FsoGoFCofY2\nN/EIZOuw13r+AoI60SrFb/MLiwVyeSUbKVDQonX87ALRSITHnj3OVUObLKNzcH/jOz2T5rWJObKF\nAulsjpPTabK5PLmConr+NS69zgV3gd2h2cdDOftajxGAXKGAiJDNFzg5nSadzV1wjIT92FgtdQu+\niHwAOKWqB0XkxtLmMg8t+9GuqvuAfQCjo6O1koa2oyTy7q/jp2czbOiMc/TM3DJhd1PIF981lV6w\ncv22tXpwm21fGbMCuVyebAFymfySOL04Pk1EhFdPz7JtY1fZhb8wvMFX1uPdNfi8KplsYSkBmEvn\nmM8Wlh0P7te59Nq78eJ4WOsxAqWSkJIr5BmfWmAhV2BydvH8MbKpK3THxnppRIb/LuAOEbkd6KRY\nw/8isElEYk6WPwSMN8BXW1BO5I+emWMxt3zBrFr2VMVU0b6efZptL21zi1NeIa/KuYUci9nZCxf+\n+nva9g1eKQHojEWqLrgv5gtVSyx+OR7qeU7V4t9ZWu8tHSNzmZmlY2P34IZla0LW+bMcKX39a8iT\nFTP833W6dP4J+KZr0fZnqvrX1fYfHR3Vdh2PXPo6fmomw4HXzzCfzfPyW9NLIr+ytHnhF/D67c14\nTi9iikWK2xPxSFu8wVeTAHREIVNlDSYox0OzfW5IREnnCsQiQmc8uqzzp53LgiJyUFVHaz2umX34\n9wAPicifAM8AX2miL99SEvrS1/F0tlhnTcQizC1WzmeasWDWjOes176efdxdH0dOzZDOFXjm6NTS\nG/w9u1K+F/+1JgCLTVhwD8r/ey322UyeArCYV9LZwlLnTzpXWCoLfvbmXdx4+WCNZ2pPGprh10u7\nZPgr38zj5xY4Ojm/7Ot4hNpfb43auF/HCCACiVjEl9md+7gonTfhTgBmqqXwxrpJRIsfmNEIREQY\n6uvijqu3+T4pWAt+yPBDR6VsPh6RC1riTOwbQ2HlbYX5bMFX2V2542LleRO5vIl9s1jM4zofQHn9\n9Dz7fji2LCnY0ybCXwsT/Abxg8On+NITrzAxk+bEisW1XLWznYymUOrrPjWTZjEPU/PFro7PP/Yi\nx6cWmp71r+Zb3gXxGk1h5buvwPKk4Oz8It9+/gS3X7W17YXfSjp1Unpj/+2PXuX41AL5vGK5mn9Y\nucgXEdi2sbNpWX+1b3lz1dquDM/oiUfIFZTBDQk2JzsDWeO3kk4TKZe9vXF63rI0H3JBdqeVs/56\nsrvSMfHt50+UzebtW55/WcgWKADjZ9OcnM7w+cde5NmjZ9uqxl/CMvw1YNlbe1A269/Uue7szl3O\nm5jJELPjIdBEgM54JFA1fsvwG4zV6NuHclm/O7tbbbZfqZyXteMh0LRzjd8EfxWUBpmNTcyVrdFb\nLhd88gr5vPL65Dxf+dGrNd/g7gRg/Gz6wm4hI/CUFv5PnktzfGqBE+cW+I+XJ7jntrcFVvRN8FfB\n/rFJxs8ukM1rzRM/jGBTUDgzv8hb0xkWFnM8fXTqgomMYxOzfOmJl3n55CyL+bwJfJuTzhXHOUwv\nZDn81jT7xyYDK/h2TdsajE3McvjENGVmTy0Rr/EqlrOvZ58g+vRjTLX2yeaUfEGZzeR4dWKO/WOT\nS7axiVkeeeY459JZ8qqUWwLzy9/px5iC6DPijILM5ApMp3NMzGSq7+BjLMOvgLvr4sz8Iulsfk2T\nKWvZ/TipMCwx1donl8uzWICxieKkykwuz+mZDAoceP0MZ+YXeXNyvuJwO7/8nbXs5nP1dqU4tiEW\nyfP4z99isDcRyHq+CX4ZVnZddMQiF8w2cROUyZV+npbpN5+lqZ2FvHL4rRlePz0HsDQGYa2TTO21\nD77P0tm6h0/O8OX/OMI3Dh4LXM++Cb6Lyl0X1U+lytU4YsrZ17NPEH36Maa17FMS/nmXwtcag+CX\nv9OPMQXdpwL5wvq6uvyACb5DsRPnJV45NXPBSVS1MgA/Tg30i89a9iD6XM/xYK99e/lc2dUVlO4d\nW7R1eOzZ4xx4/QynzqVrvqENwzCg2NU1MZPh2Ten+OqPX2NsYtbrkKoS+gy/VMb5p4PHOLuQXbqa\njmEYxmqYX8wzt1hczH3h+LSv6/qhFnx3GefEWcvsDcNYO6UTqydmFjkzl+WB7xxie3+3L8s7oS3p\njE3M8tUfv8azb05xeiZjYm8YRl3ktXilrTcm531b3qlb8EWkU0SeEpHnRORFEfkjZ/tOEfmJiLwi\nIv8oIh31h9sYSpn9v790itOzi8zXun6cYRjGKpnPFvj+S6d44Dsv+U70G5HhZ4D3qurVwDXArSKy\nB3gA+EtV3QVMAZ9sgK+GUFqgPTu3SEG54GpUhmEY9TA1v8iB18/w2LPHvQ5lGXULvhYpfYzFnR8F\n3gt8w9n+IPChen01grGJWR59bpyzC1kbYWsYRlOYWyxwdiHLo8+N+yrLb0gNX0SiIvIscAp4HBgD\nzqpqznnIMWBbI3zVQ6lu/9a5BevGMQyjqRQU3jqX9lU9vyGCr6p5Vb0GGAJuAK4o97By+4rI3SJy\nQEQOTExMNCKcsrjr9vNZU3vDMJqP3+r5De3SUdWzwA+APcAmESm1fQ4B4xX22aeqo6o6mkqlGhnO\nMvaPTXL4rWkyWVugNQyjdaSz+aWxyl7TiC6dlIhscm53ATcDh4DvAx91HnYX8Ei9vtZLacTxQjZf\ndm5GEEe2BsWnH2MKi08/xhQWn25yBWUhm+fwiWnPs/xGnHi1FXhQRKIUP0AeVtV/FZGfAw+JyJ8A\nzwBfaYCvNVMq5bw5Nc/UfLatRrYGwWctu/n0zm4+W2NPO2fiHnxjipMzGU9n7tQt+Kr6M+DaMttf\npVjP95RSC2Y2XyCXV2Jl/jtBHtnqd5+17ObTO7v5bI09r0quAMfOzvPWdJrHnj3O5265vMZezaGt\nz7R1t2DOZPJL86xXEvSRrX726ceYwuLTjzGFxedKuwLT6bznrZptLfj7xyY5NZ1e1oLZziNb/eiz\nlt18emc3n623FxROTac9W8BtW8Efm5jlqdcmWbCTqwzD8BEL2QJPvTbpSZbfloJfWqg9+MaUjU0w\nDMNX5BWefmPKk978thT8/WOTvHD8HNMLWa9DMQzDuIBzC1leOH6u5aWdtpyHPzGTYWp+kWzeyjmG\nYfiP4jlBysRMpqV+2zLDn5rLkMsXML03DMOPFArFE7JeOz3b0rJO2wn+2MQsPzpyeqkVyjAMw28o\nkMtry2v5bSf4pVZME3vDMPyKOj+truW3XQ1/YiZD2loxDcMIAK2u5bddhm8YhhEUtMWliLYS/LGJ\nWc7MZugoNzDHwS/T9MLi048xhcWnH2MKi8/VxhSPCvFo62S4bQS/dLLVgaNTLFYZbuGXaXph8VnL\nbj69s5tP7+ylbZmcsrCY44cvT7Rk4bZtBL90stWJswvkqnxN8ss0vbD4rGU3n97Zzad39tK20kDH\nVydmWrJw2zaLtqs92cov0/TC4tOPMYXFpx9jCovPtcSkwNxiviULt22T4QMUVCnUePH9Mk0vLD5r\n2c2nd3bz6Z195bZWLd62leAbhmEYlTHBNwzDCAl1C76IbBeR74vIIRF5UUQ+62zvF5HHReQV53df\n/eEahmEY66URGX4O+B+qegWwB/i0iFwJ3At8T1V3Ad9z7jeFsYlZXjs9u+zKVoZhGEFBlZYMUqtb\n8FX1hKo+7dyeAQ4B24C9wIPOwx4EPlSvr3K4L3aSy6vN0DEMI3AUaM1FURpawxeRHcC1wE+Ai1T1\nBBQ/FIDBRvoq4b7YSWkgkWEYRtBoxSC1hvXhi0gS+CbwOVWdFpHV7nc3cDfA8PDwmv3axU4Mw2gH\nWjFIrSEZvojEKYr911T1n53NJ0Vkq2PfCpwqt6+q7lPVUVUdTaVS6/K/mv57wzAMP1MoFLWsmTSi\nS0eArwCHVPUvXKZHgbuc23cBj9TryzAMw1g/jSjpvAv4deB5EXnW2fb7wJ8CD4vIJ4GjwC83wFdZ\nIiKIqBXwDcMILCJFLWsmdQu+qv4IqBTl++p9/loM9iboTcSYJcd8tkA8AtWuf1LOvp596rWHxacf\nYwqLTz/GFBaf64kpEYuQTMQY7E1UfrI6CfyZttv6upYl9kEZnxoWn7Xs5tM7u/n0zl5pH6Woac0i\n8IJ/bGph2deLoIxPDYvPWnbz6Z3dfHpnr7SPUNS0ZhH48cgTMxlmMrmltsygjE8Ni08/xhQWn36M\nKSw+1xPTYr7ATCbn/7ZMr3G3ZQZlfGpYfNaym0/v7ObTO3u5bYFoyzQMwzCCQVsIfrEt0+soDMMw\n1k8r2jIDL/iDvQn6ujvo7gj8n2IYRojp7ojS191hbZnV2DMywDu2bWBjVwdC5RMCDMMw/Mym7jjv\n2LaBPSMDTfMReMEfSSW557YruHa4j1jUBN8wjOARAa7Z3sc9t13BSCrZVD+BZySV5NJUsun1L8Mw\njGYgApemkk0Ve2gTwTcMwzBqY4JvGIYREtpK8K090zCMINIq2WobwS9NzUzEqv9J8TLmctuabQ+L\nTz/GFBaffowpLD7XGlNnPNrUdswSbSP42/q6iEYjxKPVPyv9Mk0vLD5r2c2nd3bz6Z195bZ4LNLU\nKZnVYgkk+YJy9dBGhgd6qv5RfpmmFxaftezm0zu7+fTOXtomQEcEtmzoJF9o/hWc2kbwd2zuoScR\nI5fXqv8Av0zTC4tPP8YUFp9+jCksPtcSU97R+R2be6o/UQNoG8EfSSX51E2Xcdlgsuof5ZdpemHx\nWctuPr2zm0/v7O5tIvDOHf1N78GHBgm+iPydiJwSkRdc2/pF5HERecX53dcIX9UonYBljTqGYQQB\nAaIRoa+noyX+GpXhfxW4dcW2e4Hvqeou4HvO/ZZgrZmGYQSBVkzIdNMQwVfVHwJnVmzeCzzo3H4Q\n+FAjfNVisDdBZzzaCleGYRh10YoJmW6aWcO/SFVPADi/B5voa4k9IwMMbuhshSvDMIx1UZrs24oJ\nmW48X7QVkbtF5ICIHJiYmKj7+UZSSd592WZiYpMzDcPwJwLEoq2ZkOmmmYJ/UkS2Aji/T5V7kKru\nU9VRVR1NpVINcdzX00E8FiFiim8Yhg8RgVgk0pIJmW6aKfiPAnc5t+8CHmmir2XYVbAMw/Azra7d\nl2hUW+Y/APuBy0XkmIh8EvhT4BYReQW4xbnfEtxXwTIMw/Abra7dl4g14klU9VcrmN7XiOdfK6Wr\nYH3x8ZcZP3ui5qnPhmEYraJ0davP3bK7peWcku+25PxVsLyOxDAM4zyRFl3dqqzvlntsIYO9CboT\nsWXdOn4ZnxoWn36MKSw+/RhTWHxWs3d3xlpeuy/R1oK/Z2SAkVSSqJz/Q/0yPjUsPmvZzad3dvPZ\nWrsAsQiMbE62vHa/Mpa2ZCSV5D27U/QkYnTEinm+X8anhsVnLbv59M5uPltrT8SEnniU9+xOeVLO\ngTYX/BK5glIaNe2X8alh8enHmMLi048xhcVnOXtBIdf8kfdVaXvBH+xNsLErTsLJ8P0yPjUsPmvZ\nzad3dvPZWnsiFmFjV9yz+j2EQPDdPfml+RWGYRitwqu5OeVoe8Ev9eRfO9xHR0ysTdMwjJYSEeiI\nScvn5pSNxTPPLWQklWTPpQMM9CRs3IJhGC2luyPKQE+CPZcOeCr2EBLBBxu3YBiGN/ihlFMiNIJf\nKu3c9LZBumudKWEYhtEAuuMRbrx80PNSTolQKd9IKsnH37WTLRu7vA7FMIwQsGVjFx9/105fiD2E\nTPDh/AVS4pEQ/vGGYbSECMWRCu++bLNvxB5CqnmXb+kl1dtJTyKUf75hGE2mJxEl1dvJ5Vt6vQ5l\nGaFUvJW9+aF8EQzDaBp+Wqh1E0qtcy/gppIdbOisflmAdpng57epgeYzfDGFwWdPIuqrhVo3oRR8\nOL+Ae83wJrb2dRKtckJWu0zw88vUQPPpD7v5bKy9NA1z92CvrxZq3YRW8OF8pn/9Jf10VfkYb5cJ\nfn6ZGmg+/WE3n42zC5BMRNnQGfN0GmYtmi74InKriBwWkSMicm+z/a2VkVSSK7ZsIBqJVJyz0y4T\n/PwyNdB8+sNuPhtrT8Sj9Ha2/sLka6Gpgi8iUeCvgNuAK4FfFZErm+lzPewZGWAgmag49a5dJvj5\nZWqg+fSH3Xw2xl5KFLviES7f4t3FTVZDszP8G4Ajqvqqqi4CDwF7m+xzzYykktxx9cX0dcftLFzD\nMNZETyJKKtnh24VaN81Wt23Am677x5xtS4jI3SJyQEQOTExMNDmcynzwmot5544++nts1o5hGKtn\nU3eca4Y3+Xah1k2zBb9cWXzZNyRV3aeqo6o6mkqlmhxOZWzWjmEYa8Vvs3Jq0WxlOwZsd90fAsab\n7HPd2KwdwzDWgt9m5dSi2YL/U2CXiOwUkQ7gTuDRJvusC6vnG4ZRi554hP7uOHdcfXFgxB6aLPiq\nmgM+A3wXOAQ8rKovNtNnI3DX8230gmEYK+lPdjC6o48PXnOx16GsiaZrmap+W1V3q+qIqn6h2f4a\nwcrRCzZkzTCMEkGr27sxJauAe/SCXSXLMIwSQavbuzHBr4J17hiGUSKodXs3pmI1sM4dwwg3EYr9\n5UGt27sxwV8F7s6d3jL1/CCMbPXKpx9jCotPP8YURJ8bumKBOZO2Fib4q6TUuTPU301sxavm55Gt\nXvusZTef3tnNZ217LAJbN3UG5kzaWlS/8oexRKme/8gzx1nMjnP0zDz5QnE0ql9HtvrBZy27+fTO\nbj4r24Vi5j/U38MtV2xh77XbAi/2YBn+mhhJJdl77TY6YlEKyNKMCD+PbPXapx9jCotPP8YUFJ8i\nkFdIxCJtI/Zggr9mRlJJ3v/2LWzojFW9aIqbsIyJDUpMYfFZy24+K9u7YhE2dsV5/9u3tI3Ygwn+\nuijV8y/a2EkiJnR12MtoGO1CFLhoY2fgO3LKYTX8dVCq5+8fm+Q7z4/zxuQ884tpr8MyDKMBbN3U\nxSfefSm/MDLQVtk9mOCvm5FUkpFUkl8YGeCrP36Nbx48xny21vKQYRh+QyiWdXriERLxKB+9fohf\n23OJ12E1BatF1EnpxKzhgW7iEXtBDSNodHVESMSkbcs4bizDbwAjqST33nYFX3r8ZcbPpTk5k/E6\nJMMwVslATwfD/d3cdtXFbVnGcWOC3yBuvHyQ7f3dVt4xjABRmnzZDidVrQarQDSQSuWdctd5NAzD\nO6JSPLHqkoGe0Ig9mOA3nFJ55x0Xb+TiPmvbNAy/EQW2buzkHds2cs9tbwuN2IOVdJpCqbxjbZuG\n4T+2burit24caft6fTnqSj1F5JdF5EURKYjI6ArbfSJyREQOi8j76wszeIykkvzankv4/Ieu4qa3\nDdLTEV1W2gni1MB2iSksPv0Yk1c+BehNROnvji+1XYZN7KH+ks4LwIeBH7o3isiVFC9Y/nbgVuCv\nRSRap69AUqrr776ol1gEIo7qB21qYLPs5tM7e1h8JuIROmLCUH9X27dd1qKuko6qHgIQuWBZci/w\nkKpmgNdE5AhwA7C/Hn9BZSSV5KPXDwEwt5jljcl5oiJQqNzJ45epgc22m0/v7GHwGRPYuTnJxq44\n1w33tdUgtPXQrNXEbcCbrvvHnG2hZc/IAJemeuhJxIlFInTEq3/h8cvUwHaMKSw+/RhTq32KCPFI\nhIs2dIZe7GEVGb6IPAFsKWO6X1UfqbRbmW1lh9aJyN3A3QDDw8O1wgksI6kkn7rpsmULuWfnsxUf\n75epgc22m0/v7GHwedGGTj58/VAoF2jLUVPwVfXmdTzvMWC76/4QMF7h+fcB+wBGR0dr/T8Djc3f\nMYzmEqa5OOuhWSWdR4E7RSQhIjuBXcBTTfIVOGz+jmE0hzDNxVkPdS3aisgvAf8HSAHfEpFnVfX9\nqvqiiDwM/BzIAZ9W1Xz94bYP7vk7E3MZTkylsRfIMNZPFNiyoZOtGztDMRdnPYiqf6ooo6OjeuDA\nAa/DaCljE7PsH5vkKz96lWOu6+SWvpoahlGdqBTbnbf39/CJd+8MpdCLyEFVHa31ODvT1mNKdf2h\nvq6lbH9iJkM0IswvWn3fMKpRGpOQ6k3w2Zt3c+Plg16H5GtM8H3CynEM4+fSHD09b2UewyiDZfXr\nwwTfR7i7ePaPTfLlH4xx7OyC12EZhm+IUCx1pno72Lqxy7L6NWKC70NKwn96JsOD+18nk81b+6Zh\nAD2JKF3xKLdcuSVUY40bhXUE+pgPXnMx79zRx0UbOwnlICLDcBEFUr0JrhneZGK/TizD9zEjqST3\n3HaFdfEYocbq9Y3DBN/nVOri6YgKM5nKZZ54BFZWgcpt89puPsMV01r3iQlssS6chmGCHxBWdvGc\nmV/klZOzFYdJ+WU0bS27+fTO7mefpaz+ssFePrbnEsvqG4QJfoBwd/E88sxxFrPjHD+7wGJeKayo\n7/hhNO1q7ObTO7sffUajERIR5ZL+bpKJGB+5fjsfs1k4DcMWbQPISCrJ3mu30dsVJxqJEItcOJzU\nD6Np/RpTWHz6MaZa+3TEo8QiEZKJODtTSfaMDFTfwVgTJvgBZSSV5HM372b3YJLBDYkLunj8MJp2\nNXbz6Z3dbz6jQH93nGu2b+TD1w/xqZsuszJOg7GSToBx1/Wti8cIKtaF0zpM8AOOzeIxgozNwmkt\nJvhtgs3iMYJA6ZunZfXeYILfRqycxWNlHsNvdHVEyBeUVDJhWb0HmOC3IVbmMfyIXaDEe0zw25iy\nZZ7JeQpqmb7RfBIRyBSKZ8uKlW98gQl+m7OyzPPNg2/y+uQ86WyeTLZQ80QYw1gPUWBwQyfpXIFU\nsoNEPGrlGx9Q7zVt/xz4ILAIjAG/qapnHdt9wCeBPPDbqvrdOmM16sAt/E+OTTIxk+aR58atxm80\nlJWLsUN9XeQLyo7NPZbV+4B6M/zHgftUNSciDwD3AfeIyJXAncDbgYuBJ0Rkt13I3HtKwg9wzXCf\n1fiNhmEtlv6nLsFX1X9z3X0S+Khzey/wkKpmgNdE5AhwA7C/Hn9GY6lU489XSPP9MD3RfPovJqvR\nB4dG1vA/Afyjc3sbxQ+AEsecbYbPWFnj/9qTb3Dk1Ax55YKBbGGb2Bh0n7Xs9T5nIh4hkle29XWx\nsTNuWX0AqCn4IvIEsKWM6X5VfcR5zP1ADvhaabcyjy+bN4rI3cDdAMPDw6sI2WgGJeEX4BsHjzG3\nmGXs5NyyE7fCMrGxXXzWstfznDGBnZuTbOyKc91wH3uv3WZZfQCoKfiqenM1u4jcBXwAeJ+qlkT9\nGLDd9bAhYLzC8+8D9gGMjo7amqHH7BkZ4OmjU7x6eo5IVBCFfEFRLsz4V9IuExvbxWczYopIMfOP\nRIR4JMJFGzpN7ANEvV06twL3AP9FVeddpkeBr4vIX1BctN0FPFWPL6M1jKSSfOqmy9g/NgmqSy2c\nuUKBeDRSdVG3HSY2tpPPWva17hMBOmNCZzzKJQNJPnz9kNXrA0a9Nfz/CySAx0UE4ElV/S1VfVFE\nHgZ+TrHU82nr0AkO5Vo4f/r6GZvPEzLKzb354NUXk+pNmNAHlHq7dC6rYvsC8IV6nt/wFncL59jE\n7PL5PGUWdY32IpmIsJi3uTfthJ1pa6yKlfN50vk8RycXyGTzFNRO3GoXok6LZSIWZddgkq6OqM29\naSNM8I01Uerdf/30HKdm0jyPVFmLAAAJpklEQVT81Jt24lab4D5x6lfeuZ3B3k47Q7bNMME31oy7\n1HPDzgEbzhZQbLhZ+DDBN+rChrMFExtuFk5M8I2GYMPZ/E2lK03ZcLNwYYJvNJRqw9k6osJMxnJ+\nL7COGwNM8I0msnI428JinsMnZ62zp0VYx42xEhN8o6m4Sz3lOnss628M7g9OoSj2W6zjxliBCb7R\nEip19izL+rnwZC4/jwX2k89kIkI6VyAeOT/64CM2+sBYgQm+0XIqZf3n0lmOnplfNrTLj2OB/eYz\nFoGh/m76uzsY3TFgow+MipjgG56xMut/5JnjfOtn4xw/u8BiXikoSISqc3qDMqq4Gc8ZE4hGhG19\n3dxyxRabWmnUpFZiYRgtYSSVZO+12+jtihONRIhHhahAVKofokEZVdyM54zFIsSiETZ2xU3sjVVh\ngm/4hpFUks/dvJvdg0kuTfXQ39NBf08H0Sr7BGVUcb3PWSIqxRJOT0eUSwe62X1Rks/evMvE3lgV\nVtIxfIV7Vk80IhybWrAJnVzYR2+dN8Z6MME3fIe7tg+EdkKn9dEbjcYE3/A91SZ0tksfv/XRG63A\nBN8IBKvq49fa3S5+pasjQjZvffRGczHBNwJHtbN3T5xNLxvP7PeSz1I2v6GTrRs7rY/eaCr1XsT8\nj4G9FBOrU8DHVXVcihe4/RJwOzDvbH+63mANw025rN89nnk1F173igjQGbds3mgt9Wb4f66q/xNA\nRH4b+APgt4DbgF3Oz38G/sb5bRhNoeqF11dclMWLrL/cxUbsguBGq6n3IubTrrs9nH8f7QX+XlUV\neFJENonIVlU9UY8/w6hFuQuve53128VGDL9Qdw1fRL4A/AZwDrjJ2bwNeNP1sGPONhN8o2VUzfpP\nz5Nvkl+72IjhV2oKvog8AWwpY7pfVR9R1fuB+0XkPuAzwB9SPOZXUvZbtIjcDdwNMDw8vNq4DWPV\nlMv6q53M1YjJlXaxEcOP1BR8Vb15lc/1deBbFAX/GLDdZRsCxis8/z5gH8Do6KifGyqMNqAk/u6T\nuY5PpZnP5JbGM69ncmVUAIGejhiXpXrsJCnDl9TbpbNLVV9x7t4BvOTcfhT4jIg8RHGx9pzV7w0/\n4T6Z6/lj5/iPlyeYW8zyxuQ8UREoVE7hV1piApemekgmYrxn9yBXDW20so3hS+qt4f+piFxO8T3w\nBsUOHYBvU2zJPEKxLfM36/RjGA2nlO3v2NzDm1PzvHp6jlgkQjwWYb5KzWbl5EoRIZmIszPVwwev\nudiE3vAt9XbpfKTCdgU+Xc9zG0arGEkl+dRNly2dvTt+Ls3MfLbmou75RdluPmx99EYAsDNtDYPl\nHT3LFnULF5Zwujoi5Au2KGsEDxN8w3CxclF3/FyakzOZZY8Z6OlguL/bFmWNwGGCbxhlKC3qfvXH\nr/HPTx9nfjGPAF2JKDdePsjH37XThN4IHCb4hlGBkVSSj79rJ69OzPLzEzMIyhVbN5rYG4HFBN8w\nqjCSSvL5D13Fk2OTKFgJxwg0JviGUYOVV+AyjKBiFzE3DMMICSb4hmEYIcEE3zAMIySY4BuGYYQE\nE3zDMIyQYIJvGIYREqQ458wfiMgExamb5dgMnG5hOGvF4qsPv8cH/o/R4quPIMd3iaqmaj2BrwS/\nGiJyQFVHvY6jEhZfffg9PvB/jBZffYQhPivpGIZhhAQTfMMwjJAQJMHf53UANbD46sPv8YH/Y7T4\n6qPt4wtMDd8wDMOojyBl+IZhGEYdBEbwReR3RURFZLNzX0Tkf4vIERH5mYhc51Fcf+z4f1ZE/k1E\nLvZZfH8uIi85MfyLiGxy2e5z4jssIu/3KL5fFpEXRaQgIqMrbJ7H58RxqxPDERG516s43IjI34nI\nKRF5wbWtX0QeF5FXnN99HsW2XUS+LyKHnP/tZ/0UnxNLp4g8JSLPOTH+kbN9p4j8xInxH0Wkw8MY\noyLyjIj8a8NiU1Xf/wDbge9S7NHf7Gy7HfgOIMAe4CcexbbBdfu3gS/7LL7/CsSc2w8ADzi3rwSe\nAxLATmAMiHoQ3xXA5cAPgFHXdr/EF3V8Xwp0ODFd6cX/ckVc7wGuA15wbfsz4F7n9r2l/7UHsW0F\nrnNu9wIvO/9PX8Tn+Bcg6dyOAz9x3qcPA3c6278M/HcPY/wd4OvAvzr3644tKBn+XwK/B7gXHPYC\nf69FngQ2icjWVgemqtOuuz2cj9Ev8f2bquacu08CQ674HlLVjKq+BhwBbvAgvkOqeriMyRfxOT6P\nqOqrqroIPOTE5imq+kPgzIrNe4EHndsPAh9qaVAOqnpCVZ92bs8Ah4BtfonPiUtVdda5G3d+FHgv\n8A1nu2cxisgQ8IvA3zr3pRGx+V7wReQO4LiqPrfCtA1403X/mLOt5YjIF0TkTeBjwB84m30Tn4tP\nUPzWAf6Mz41f4vNLHKvhIlU9AUXRBQY9jgcR2QFcSzGD9lV8TsnkWeAU8DjFb3JnXQmSl//rL1JM\ncgvO/QEaEJsvrnglIk8AW8qY7gd+n2JZ4oLdymxrSstRtfhU9RFVvR+4X0TuAz4D/KGf4nMecz+Q\nA75W2s1P8ZXbrcw2L1rK/BJH4BCRJPBN4HOqOl1MUv2DquaBa5x1rX+hWF684GGtjQpE5APAKVU9\nKCI3ljaXeeiaY/OF4KvqzeW2i8hVFOu3zzkHyxDwtIjcQPETbrvr4UPAeCvjK8PXgW9RFHzfxCci\ndwEfAN6nTgHQT/FVoGXxBSSO1XBSRLaq6gmnfHjKq0BEJE5R7L+mqv/st/jcqOpZEfkBxRr+JhGJ\nOZm0V//rdwF3iMjtQCewgWLGX3dsvi7pqOrzqjqoqjtUdQfFN991qvoW8CjwG043zB7gXOnrYisR\nkV2uu3cALzm3/RLfrcA9wB2qOu8yPQrcKSIJEdkJ7AKeanV8VfBLfD8FdjkdEh3AnU5sfuRR4C7n\n9l1ApW9PTcWpN38FOKSqf+Ey+SI+ABFJlTrWRKQLuJniWsP3gY86D/MkRlW9T1WHHM27E/h3Vf1Y\nQ2LzagV6navWr3O+S0eAv6JYd3seV4dHi2P6JvAC8DPgMWCbz+I7QrEG/azz82WX7X4nvsPAbR7F\n90sUP8gzwEngu36Kz4njdoqdJmMUy1CexLEipn8ATgBZ5/X7JMU67/eAV5zf/R7F9m6K5YafuY67\n2/0SnxPjfwKecWJ8AfgDZ/ulFBOLI8A/AQmP/883cr5Lp+7Y7ExbwzCMkODrko5hGIbROEzwDcMw\nQoIJvmEYRkgwwTcMwwgJJviGYRghwQTfMAwjJJjgG4ZhhAQTfMMwjJDw/wE4HspszllroAAAAABJ\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1af3d63de10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 画心心\n",
    "# (x**2 + y**2 - 1)**3 - (x**2) * (y**3) = 0\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "x_coords = np.linspace(-100, 100, 500)\n",
    "y_coords = np.linspace(-100, 100, 500)\n",
    "\n",
    "points = []\n",
    "\n",
    "for y in y_coords:\n",
    "    for x in x_coords:\n",
    "        if (((x*0.03)**2 + (y*0.03)**2 - 1)**3 - (x*0.03)**2 * (y*0.03)**3)<= 0:\n",
    "            points.append({\"x\":x, \"y\":y})\n",
    "\n",
    "heart_x = list(map(lambda point:point[\"x\"],points))\n",
    "heart_y = list(map(lambda point:point[\"y\"],points))\n",
    "\n",
    "lenth = len(heart_x)\n",
    "\n",
    "# plt.scatter(heart_x, heart_y, s=10, alpha=0.5)\n",
    "# plt.show()\n",
    "\n",
    "print(\"紫色 magma \\n彩虹 rainbow \\n青春 spring \\n五彩 gist_rainbow \\n橘黄 autumn \\n红 Reds\")\n",
    "colormap = str(input(\"Enter the color of your heart:\"))\n",
    "plt.scatter(heart_x, heart_y, s=10, alpha=0.5, c=range(lenth), cmap=colormap)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sequential(\n",
      "  (0): Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "  (1): ReLU(inplace)\n",
      "  (2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "  (3): ReLU(inplace)\n",
      "  (4): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "  (5): ReLU(inplace)\n",
      "  (6): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "import sys\n",
    "sys.path.append('..')\n",
    "\n",
    "import numpy as np\n",
    "import torch\n",
    "from torch import nn\n",
    "from torch.autograd import Variable\n",
    "from torchvision.datasets import CIFAR10\n",
    "\n",
    "def vgg_block(num_convs, in_channels, out_channels):\n",
    "    net = [nn.Conv2d(in_channels, out_channels, kernel_size=3, padding=1), nn.ReLU(True)] # 定义第一层\n",
    "    \n",
    "    for i in range(num_convs-1): # 定义后面的很多层\n",
    "        net.append(nn.Conv2d(out_channels, out_channels, kernel_size=3, padding=1))\n",
    "        net.append(nn.ReLU(True))\n",
    "        \n",
    "    net.append(nn.MaxPool2d(2, 2)) # 定义池化层\n",
    "    return nn.Sequential(*net)\n",
    "\n",
    "block_demo = vgg_block(3, 64, 128)\n",
    "print(block_demo)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([1, 64, 300, 300])\n",
      "torch.Size([1, 128, 150, 150])\n"
     ]
    }
   ],
   "source": [
    "# 首先定义输入为 (1, 64, 300, 300)\n",
    "input_demo = Variable(torch.zeros(1, 64, 300, 300))\n",
    "print(input_demo.shape)\n",
    "output_demo = block_demo(input_demo)\n",
    "print(output_demo.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch:  0 | Step:  0 | batch x:  [  2.   7.  10.   1.   4.] | batch y:  [  9.   4.   1.  10.   7.]\n",
      "Epoch:  0 | Step:  1 | batch x:  [ 3.  5.  6.  8.  9.] | batch y:  [ 8.  6.  5.  3.  2.]\n",
      "Epoch:  1 | Step:  0 | batch x:  [ 7.  8.  2.  3.  9.] | batch y:  [ 4.  3.  9.  8.  2.]\n",
      "Epoch:  1 | Step:  1 | batch x:  [  5.   1.   6.   4.  10.] | batch y:  [  6.  10.   5.   7.   1.]\n",
      "Epoch:  2 | Step:  0 | batch x:  [  8.   1.   4.   9.  10.] | batch y:  [  3.  10.   7.   2.   1.]\n",
      "Epoch:  2 | Step:  1 | batch x:  [ 7.  6.  5.  2.  3.] | batch y:  [ 4.  5.  6.  9.  8.]\n"
     ]
    }
   ],
   "source": [
    "\"\"\"\n",
    "View more, visit my tutorial page: https://morvanzhou.github.io/tutorials/\n",
    "My Youtube Channel: https://www.youtube.com/user/MorvanZhou\n",
    "\n",
    "Dependencies:\n",
    "torch: 0.1.11\n",
    "\"\"\"\n",
    "import torch\n",
    "import torch.utils.data as Data\n",
    "\n",
    "torch.manual_seed(1)    # reproducible\n",
    "\n",
    "# 对于一大批数据量，分成小批训练\n",
    "\n",
    "BATCH_SIZE = 5#每小批中的训练数量\n",
    "# BATCH_SIZE = 8\n",
    "\n",
    "x = torch.linspace(1, 10, 10)       # this is x data (torch tensor)用于训练\n",
    "y = torch.linspace(10, 1, 10)       # this is y data (torch tensor)用于计算误差\n",
    "\n",
    "torch_dataset = Data.TensorDataset(x, y)\n",
    "loader = Data.DataLoader(\n",
    "    dataset=torch_dataset,      # torch TensorDataset format\n",
    "    batch_size=BATCH_SIZE,      # mini batch size\n",
    "    shuffle=True,               # random shuffle for training是否打乱每次训练数据的顺序\n",
    "    num_workers=2,              # subprocesses for loading data线程\n",
    ")\n",
    "\n",
    "for epoch in range(3):   # train entire dataset 3 times\n",
    "    for step, (batch_x, batch_y) in enumerate(loader):  # for each training step\n",
    "        # train your data...\n",
    "        print('Epoch: ', epoch, '| Step: ', step, '| batch x: ',\n",
    "              batch_x.numpy(), '| batch y: ', batch_y.numpy())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "ename": "TypeError",
     "evalue": "'Tensor' object is not callable",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-16-0aa026fbe5de>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[0;32m      3\u001b[0m \u001b[0ma\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mt\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mlinspace\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m16\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m16\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      4\u001b[0m \u001b[0mb\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0ma\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mview\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m4\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m4\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 5\u001b[1;33m \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mb\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      6\u001b[0m \u001b[0mc\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mt\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmax\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mb\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      7\u001b[0m \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mc\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mTypeError\u001b[0m: 'Tensor' object is not callable"
     ]
    }
   ],
   "source": [
    "import torch as t\n",
    "\n",
    "a = t.linspace(1,16,16)\n",
    "b = a.view(4,4)\n",
    "print(b)\n",
    "c = t.max(b,1)\n",
    "print(c)\n",
    "d = c[1]\n",
    "print(d)\n",
    "e = d.data.squeeze()\n",
    "print(e)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[  8.,   8.,   8.,   8.],\n",
       "        [  8.,   8.,   8.,   8.],\n",
       "        [  9.,  10.,  11.,  12.],\n",
       "        [ 13.,  14.,  15.,  16.]])"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f = t.clamp(b,8)\n",
    "f"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "ename": "TypeError",
     "evalue": "'Tensor' object is not callable",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-13-5d91aebeae46>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[0;32m      1\u001b[0m \u001b[0mg\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m(\u001b[0m\u001b[0mb\u001b[0m\u001b[1;33m==\u001b[0m\u001b[0mf\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msum\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;31m# b和f中有多少个相同的元素\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      2\u001b[0m \u001b[0mh\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mfloat\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mb\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msize\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 3\u001b[1;33m \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mg\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      4\u001b[0m \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mh\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      5\u001b[0m \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mg\u001b[0m\u001b[1;33m/\u001b[0m\u001b[0mh\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mTypeError\u001b[0m: 'Tensor' object is not callable"
     ]
    }
   ],
   "source": [
    "g = (b==f).sum()# b和f中有多少个相同的元素\n",
    "h = float(b.size(0))\n",
    "print(g)\n",
    "print(h)\n",
    "print(g/h)"
   ]
  },
  {
   "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
}