{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 第三章 线性模型"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "## 3.1 试析在什么情形下式(3.2) 中不必考虑偏置项 b."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "中心化的样本数据，没有截距项，样本重心过原点。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "## 3.2 试证明，对于参数$w$，对率回归的目标函数(3.18) 是非凸的，但其对数似然函数(3.27) 是凸的."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "参考：\n",
    "\n",
    "[\\[转载\\]机器学习中常用的矩阵求导公式](http://blog.sina.com.cn/s/blog_14b5db8290102x1zq.html)\n",
    "\n",
    "[矩阵求导公式有哪些? - 柴士童的回答 - 知乎](https://www.zhihu.com/question/350887271/answer/859865647)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "先推一下附录的公式，熟悉一下结构，根据@柴士童的回答和常见的矩阵求导公式\n",
    "$$\n",
    "\\frac{\\partial w^Tx}{\\partial x} = w\n",
    "\\\\\n",
    "\\frac{\\partial -X^TB^TA}{\\partial t} \n",
    "= -(\\frac{\\partial X^T}{\\partial t})^TB^TA\n",
    "$$\n",
    "有\n",
    "$$\n",
    "\\frac{\\partial w^Tx}{\\partial w} \n",
    "= \\frac{\\partial (x^Tw)^T}{\\partial w} \n",
    "= (\\frac{\\partial x^Tw}{\\partial w})^T \n",
    "= (x^T)^T\n",
    "= x\n",
    "$$\n",
    "利用已有结论\n",
    "$$\n",
    "\\frac{\\partial y}{\\partial w}\n",
    "= \\frac{x \\cdot e^{(w^Tx + b)}}{(1 + e^{(w^Tx + b)})^2}\n",
    "= x \\cdot (y - y^2)\n",
    "\\\\\n",
    "\\frac{\\partial^2 y}{\\partial w^2}\n",
    "= \\frac{\\frac{\\partial y}{\\partial w}}{\\partial w^T}\n",
    "= \\frac{\\partial x \\cdot (y - y^2)}{\\partial y} \\cdot \\frac{\\partial y}{\\partial w^T}\n",
    "= x \\cdot (1 - 2y) \\cdot (\\frac{\\partial y}{\\partial w})^T\n",
    "= x x^T (y - y^2)(1 - 2y)\n",
    "$$\n",
    "显然$(y - y^2)(1 - 2y)$不恒非负，故(3.18)式非凸。\n",
    "\n",
    "PS：这里没搞懂为什么 @四去六进一 [机器学习(周志华) 参考答案 第三章 线性模型](https://blog.csdn.net/icefire_tyh/article/details/52069025)中算的$(x \\cdot (y - y^2))^T$是$x^T(y^2 - y)$而不是$x^T(y - y^2)$\n",
    "\n",
    "同理求(3.27)式关于$w$的二阶偏导\n",
    "$$\n",
    "\\frac{\\partial^2 L}{\\partial w^2}\n",
    "= \\sum_{i = 1}^{m}{x_i x_i^T p_1 (1 - p_1)}\n",
    "$$\n",
    "显然对于$p_1 \\in (0,1)$，上式恒大于0，故(3.27)式为凸函数。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "## 3.3 编程实现对率回归，并给出西瓜数据集 3.0α 上的结果."
   ]
  },
  {
   "attachments": {
    "image.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "![西瓜数据集3.0α](attachment:image.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "参考：\n",
    "\n",
    "[《机器学习（周志华）》 西瓜数据集3.0](https://blog.csdn.net/wiking__acm/article/details/50971461)\n",
    "\n",
    "[pandas的dataframe赋值操作](https://www.cnblogs.com/math98/p/10888893.html)\n",
    "\n",
    "[warning：A value is trying to be set on a copy of a slice from a DataFrame](https://blog.csdn.net/qq_42711381/article/details/90451301)\n",
    "\n",
    "[pandas：Indexing and selecting data](https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html)\n",
    "\n",
    "[Python实现简单的梯度下降法](https://www.cnblogs.com/noluye/p/11108513.html)\n",
    "\n",
    "[性能度量指标](https://blog.csdn.net/hfutdog/article/details/88085878)\n",
    "\n",
    "[使用sklearn进行交叉验证](https://www.cnblogs.com/jiaxin359/p/8552800.html#_label3_0)\n",
    "\n",
    "[sklearn逻辑回归(Logistic Regression,LR)类库使用小结](https://blog.csdn.net/sun_shengyun/article/details/53811483)\n",
    "\n",
    "[通俗地说逻辑回归【Logistic regression】算法（二）sklearn逻辑回归实战](https://www.cnblogs.com/listenfwind/p/11780766.html)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "这里使用梯度下降法和梯度下降法，先求似然函数$L$关于$\\beta$的一阶偏导和二阶偏导：\n",
    "$$\n",
    "grad_{L}(\\beta) \n",
    "= \\frac{\\partial L}{\\partial \\beta}\n",
    "= -\\sum_{i = 1}^{m}(x_i \\cdot (y_i - p_1))\n",
    "\\\\\n",
    "grad^2_{L}(\\beta)\n",
    "= \\frac{\\partial^2 L}{\\partial \\beta^2}\n",
    "= \\sum_{i = 1}^{m}x_i x_i^T p_1 (1 - p_1)\n",
    "$$\n",
    "其中$p_1 = \\frac{1}{1 + e^{-\\beta^T x}}$\n",
    "\n",
    "于是得到更新规则：$\\beta_{new} = \\beta_{old} - \\alpha \\cdot grad_{L}(\\beta_{old})$，拟牛顿法相比于梯度下降法只需将$\\alpha$改为$(\\frac{\\partial^2 L}{\\partial \\beta^2})^{-1}$."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "为了减少多重循环的使用，选择向量化代码，在这里说明一下具体代码中使用的公式。\n",
    "\n",
    "数据集（带截距项）是一个17x3的矩阵$(x_i)_{17 \\times 3}$，记作X，和一个17x1的列向量$(y_i)_{17 \\times 1}$，记作Y，X和Y的每一行构成一组样本，即$(x_{i1},x_{i2},x_{i3})$与$(y_i)$。\n",
    "\n",
    "$p_{i1} = \\frac{1}{1 + e^{-(\\beta^T x_i)}}$其中$\\beta^T x_i$在代码中为$(x_i)_{17 \\times 3} \\cdot \\beta_{3 \\times 1}$，初始化代入$\\beta = beta_{cur}$.\n",
    "\n",
    "一阶偏导公式$grad_{L}(\\beta)$中，$y_i - p_1$是一个17x1的列向量，每一行是$y_i$与$p_{i1}(x_i;\\beta)$的差，$(x_i)_{17 \\times 3}$与$(y_i)_{17 \\times 1}$对应位置相乘，因为单个来看就是向量乘数。对所有样本求和，即对$(x_i)_{17 \\times 3}$每一列求和，得到一个${(grad_L)}_{3 \\times 1}$.\n",
    "\n",
    "二阶偏导公式$grad^2_{L}(\\beta)$类似，向量化的代码与最终写成$(x_i)^T_{3 \\times 17} \\cdot (p_{i1})_{17 \\times 17} \\cdot (x_i)_{17 \\times 3}$，其中$(p_{i1})_{17 \\times 17}$是对角线元为$p_{i1}(x_i,\\beta)$其他位置元素为0的矩阵。最后的$grad^2_{L}$是一个3x3矩阵。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-25T09:51:11.946452Z",
     "start_time": "2020-01-25T09:51:10.109921Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-25T09:51:12.101363Z",
     "start_time": "2020-01-25T09:51:12.003104Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>编号</th>\n",
       "      <th>密度</th>\n",
       "      <th>含糖率</th>\n",
       "      <th>好瓜</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0.697</td>\n",
       "      <td>0.460</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0.774</td>\n",
       "      <td>0.376</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>0.634</td>\n",
       "      <td>0.264</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>0.608</td>\n",
       "      <td>0.318</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0.556</td>\n",
       "      <td>0.215</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>0.403</td>\n",
       "      <td>0.237</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>0.481</td>\n",
       "      <td>0.149</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8</td>\n",
       "      <td>0.437</td>\n",
       "      <td>0.211</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9</td>\n",
       "      <td>0.666</td>\n",
       "      <td>0.091</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10</td>\n",
       "      <td>0.243</td>\n",
       "      <td>0.267</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>11</td>\n",
       "      <td>0.245</td>\n",
       "      <td>0.057</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>12</td>\n",
       "      <td>0.343</td>\n",
       "      <td>0.099</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>13</td>\n",
       "      <td>0.639</td>\n",
       "      <td>0.161</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>14</td>\n",
       "      <td>0.657</td>\n",
       "      <td>0.198</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>15</td>\n",
       "      <td>0.360</td>\n",
       "      <td>0.370</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>16</td>\n",
       "      <td>0.593</td>\n",
       "      <td>0.042</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>17</td>\n",
       "      <td>0.719</td>\n",
       "      <td>0.103</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    编号     密度    含糖率  好瓜\n",
       "0    1  0.697  0.460   1\n",
       "1    2  0.774  0.376   1\n",
       "2    3  0.634  0.264   1\n",
       "3    4  0.608  0.318   1\n",
       "4    5  0.556  0.215   1\n",
       "5    6  0.403  0.237   1\n",
       "6    7  0.481  0.149   1\n",
       "7    8  0.437  0.211   1\n",
       "8    9  0.666  0.091   0\n",
       "9   10  0.243  0.267   0\n",
       "10  11  0.245  0.057   0\n",
       "11  12  0.343  0.099   0\n",
       "12  13  0.639  0.161   0\n",
       "13  14  0.657  0.198   0\n",
       "14  15  0.360  0.370   0\n",
       "15  16  0.593  0.042   0\n",
       "16  17  0.719  0.103   0"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "read_path = 'D:\\JWE\\Files\\Courses\\数据挖掘\\melon_data.csv'\n",
    "melon_data = pd.read_csv(read_path)\n",
    "# Y = melon_data.loc[:,['好瓜']].replace(['是','否'],[1,0]) # replace方法不改变原dataframe\n",
    "data0 = melon_data.loc[:,['编号','密度','含糖率','好瓜']][:]\n",
    "data0.loc[data0['好瓜'] == '是',['好瓜']] = 1\n",
    "data0.loc[data0['好瓜'] == '否',['好瓜']] = 0\n",
    "data0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-25T09:51:22.922060Z",
     "start_time": "2020-01-25T09:51:22.369286Z"
    },
    "hidden": true,
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 2.89968798, 11.45957527, -4.07142452]])"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "array([[ 3.15832966, 12.52119579, -4.42886451]])"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "梯度下降法，准确率:0.7058823529411765\n",
      "[[25.05472993]\n",
      " [11.96199189]\n",
      " [ 2.20848229]\n",
      " [ 3.80277366]\n",
      " [ 1.00462133]\n",
      " [ 0.82950103]\n",
      " [ 0.37938723]\n",
      " [ 0.67957309]\n",
      " [ 0.33373401]\n",
      " [ 0.73558164]\n",
      " [ 0.066683  ]\n",
      " [ 0.14336899]\n",
      " [ 0.68829877]\n",
      " [ 1.10812166]\n",
      " [ 3.36189904]\n",
      " [ 0.15403008]\n",
      " [ 0.44654611]]\n",
      "牛顿法，准确率:0.7058823529411765\n",
      "[[34.20053642]\n",
      " [15.23586619]\n",
      " [ 2.40876022]\n",
      " [ 4.36291402]\n",
      " [ 1.01940985]\n",
      " [ 0.82817214]\n",
      " [ 0.35202617]\n",
      " [ 0.66584249]\n",
      " [ 0.30544421]\n",
      " [ 0.72743464]\n",
      " [ 0.05279359]\n",
      " [ 0.12172887]\n",
      " [ 0.67382871]\n",
      " [ 1.13355062]\n",
      " [ 3.82277505]\n",
      " [ 0.13132438]\n",
      " [ 0.41964661]]\n"
     ]
    }
   ],
   "source": [
    "X = np.c_[data0[['密度','含糖率']].values,np.ones(len(data0))]\n",
    "Y = data0['好瓜'].values\n",
    "\n",
    "p_1 = lambda x,beta:1/(1 + np.exp(-np.dot(x,beta.T)))\n",
    "grad_L = lambda x,beta,y:-sum(x * (y.reshape(len(y),1) - p_1(x=x,beta=beta)))\n",
    "grad_L_2 = lambda x,beta,y:np.dot(np.dot(x.T,np.diag(p_1(x=x,beta=beta) * (1 - p_1(x=x,beta=beta)))),x)\n",
    "\n",
    "def gradient_descent(x,y,beta_cur=np.array([[0,0,1]]),alpha=10**(-2),precision=10**(-4),max_iters=10**(4)):\n",
    "    for i in range(max_iters):\n",
    "        grad_cur = grad_L(x=x,y=y,beta=beta_cur)\n",
    "        if np.abs(grad_cur).all() < precision:\n",
    "            break\n",
    "        beta_cur = beta_cur - alpha * grad_cur\n",
    "    return beta_cur\n",
    "\n",
    "def newton(x,y,beta_cur=np.array([[0,0,1]]),alpha=np.linalg.inv(grad_L_2(x=X,y=Y,beta=np.array([[0,0,1]]).reshape((3,)))),precision=10**(-4),max_iters=10**(4)):\n",
    "    for i in range(max_iters):\n",
    "        grad_cur = grad_L(x=x,y=y,beta=beta_cur)\n",
    "        if np.abs(grad_cur).all() < precision:\n",
    "            break\n",
    "        beta_cur = beta_cur - np.dot(alpha,grad_cur)\n",
    "    return beta_cur    \n",
    "\n",
    "beta_hat_gd = gradient_descent(x=X,y=Y,max_iters=10**4)\n",
    "beta_hat_gd\n",
    "beta_hat_nt = newton(x=X,y=Y,max_iters=10**2) # 相比之下，牛顿法需要的迭代次数更少。\n",
    "beta_hat_nt\n",
    "Y_gd = p_1(x=X,beta=beta_hat_gd)\n",
    "predict_gd = Y_gd/(1 - Y_gd)\n",
    "accuracy_gd = sum([1 if i == j else 0 for i,j in zip([1 if each >= 1 else 0 for each in predict_gd],Y)])/len(Y)\n",
    "Y_nt = p_1(x=X,beta=beta_hat_nt)\n",
    "predict_nt = Y_nt/(1 - Y_nt)\n",
    "accuracy_nt = sum([1 if i == j else 0 for i,j in zip([1 if each >= 1 else 0 for each in predict_nt],Y)])/len(Y)\n",
    "print(f'梯度下降法，准确率:{accuracy_gd}')\n",
    "print(f'{predict_gd}')\n",
    "print(f'牛顿法，准确率:{accuracy_nt}')\n",
    "print(f'{predict_nt}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-25T09:51:32.557095Z",
     "start_time": "2020-01-25T09:51:23.022526Z"
    },
    "hidden": true,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "回归参数（全集）:[ 0.17158985  0.44667882 -0.24623819  0.        ]\n",
      "平均准确率:0.4104\n"
     ]
    }
   ],
   "source": [
    "# 使用sklearn包训练logit模型\n",
    "# 划分数据集，20次2拆交叉验证\n",
    "from sklearn.model_selection import RepeatedKFold\n",
    "# X = data0[['密度','含糖率']].values\n",
    "X = np.c_[data0[['密度','含糖率']].values,np.ones(len(data0))]\n",
    "Y = data0['好瓜'].values\n",
    "rkf = RepeatedKFold(n_splits=2, n_repeats=20)\n",
    "X_index = rkf.split(X)\n",
    "# 训练logit模型，计算accuracy\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import accuracy_score\n",
    "# 全集，求解器参数：lbfgs为牛顿法，liblinear为坐标下降法\n",
    "all_logitmodel = LogisticRegression(solver='lbfgs',fit_intercept=False).fit(X,Y)\n",
    "my_logitmodel = []\n",
    "haogua_predict = []\n",
    "model_accuracy = []\n",
    "for train_index,test_index in X_index:\n",
    "    cur_model = LogisticRegression(solver='liblinear',fit_intercept=False).fit(X[train_index],Y[train_index])\n",
    "    my_logitmodel.append(cur_model)\n",
    "    haogua_predict.append(cur_model.predict(X[test_index]))\n",
    "    model_accuracy.append(accuracy_score(Y[test_index],cur_model.predict(X[test_index])))\n",
    "model_accuracy0 = np.mean(model_accuracy)\n",
    "# print(f'回归参数（平均）:{np.mean([np.append(each.coef_,each.intercept_) for each in my_logitmodel],axis=0)}')\n",
    "print(f'回归参数（全集）:{np.append(all_logitmodel.coef_,all_logitmodel.intercept_)}')\n",
    "print(f'平均准确率:{model_accuracy0:.4f}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-25T09:51:32.785172Z",
     "start_time": "2020-01-25T09:51:32.767126Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.17158985,  0.44667882, -0.24623819]])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "array([[1.08202274],\n",
       "       [1.05603739],\n",
       "       [0.9806624 ],\n",
       "       [1.00013232],\n",
       "       [0.94667609],\n",
       "       [0.93125271],\n",
       "       [0.9074214 ],\n",
       "       [0.92588597],\n",
       "       [0.91273328],\n",
       "       [0.91825642],\n",
       "       [0.836325  ],\n",
       "       [0.86661396],\n",
       "       [0.9373702 ],\n",
       "       [0.95593893],\n",
       "       [0.98098837],\n",
       "       [0.88185742],\n",
       "       [0.92602211]])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "array([1, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], dtype=int64)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "0.7058823529411765"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 查看sklearn训练logit模型的结果\n",
    "beta_hat_sk = all_logitmodel.coef_\n",
    "beta_hat_sk\n",
    "Y_sk = 1/(1 + np.exp(-np.dot(X,beta_hat_sk.T)))\n",
    "Y_sk/(1 - Y_sk)\n",
    "all_logitmodel.predict(X)\n",
    "accuracy_score(Y,all_logitmodel.predict(X))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "## 3.4 选择两个 UCI 数据集，比较 10 折交叉验证法和留出法所估计出的对率回归的错误率."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "日后再补。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-22T07:32:46.499175Z",
     "start_time": "2020-01-22T07:32:46.494161Z"
    },
    "heading_collapsed": true
   },
   "source": [
    "## 3.5 编辑实现线性判别分析，并给出西瓜数据集 3.0α 上的结果."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "参考：\n",
    "\n",
    "[机器学习——线性判别分析](https://blog.csdn.net/dhaiuda/article/details/84325203)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "LDA中$w$的求解最后转换成了求$S_w^{-1}(\\mu_0 - \\mu_1) = V \\Sigma^{-1} U^T(\\mu_0 - \\mu_1)$的数值解。\n",
    "\n",
    "其中$S_w = \\sum_{x \\in X_0}(x - \\mu_0)(x - \\mu_0)^T + \\sum_{x \\in X_1}(x - \\mu_0)(x - \\mu_0)^T$，$S_w = U \\Sigma V^T$为$S_w$的奇异值分解。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-25T09:51:32.931059Z",
     "start_time": "2020-01-25T09:51:32.927550Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-25T09:51:33.077949Z",
     "start_time": "2020-01-25T09:51:33.019795Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>编号</th>\n",
       "      <th>密度</th>\n",
       "      <th>含糖率</th>\n",
       "      <th>好瓜</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0.697</td>\n",
       "      <td>0.460</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0.774</td>\n",
       "      <td>0.376</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>0.634</td>\n",
       "      <td>0.264</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>0.608</td>\n",
       "      <td>0.318</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0.556</td>\n",
       "      <td>0.215</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>0.403</td>\n",
       "      <td>0.237</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>0.481</td>\n",
       "      <td>0.149</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8</td>\n",
       "      <td>0.437</td>\n",
       "      <td>0.211</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9</td>\n",
       "      <td>0.666</td>\n",
       "      <td>0.091</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10</td>\n",
       "      <td>0.243</td>\n",
       "      <td>0.267</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>11</td>\n",
       "      <td>0.245</td>\n",
       "      <td>0.057</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>12</td>\n",
       "      <td>0.343</td>\n",
       "      <td>0.099</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>13</td>\n",
       "      <td>0.639</td>\n",
       "      <td>0.161</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>14</td>\n",
       "      <td>0.657</td>\n",
       "      <td>0.198</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>15</td>\n",
       "      <td>0.360</td>\n",
       "      <td>0.370</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>16</td>\n",
       "      <td>0.593</td>\n",
       "      <td>0.042</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>17</td>\n",
       "      <td>0.719</td>\n",
       "      <td>0.103</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    编号     密度    含糖率  好瓜\n",
       "0    1  0.697  0.460   1\n",
       "1    2  0.774  0.376   1\n",
       "2    3  0.634  0.264   1\n",
       "3    4  0.608  0.318   1\n",
       "4    5  0.556  0.215   1\n",
       "5    6  0.403  0.237   1\n",
       "6    7  0.481  0.149   1\n",
       "7    8  0.437  0.211   1\n",
       "8    9  0.666  0.091   0\n",
       "9   10  0.243  0.267   0\n",
       "10  11  0.245  0.057   0\n",
       "11  12  0.343  0.099   0\n",
       "12  13  0.639  0.161   0\n",
       "13  14  0.657  0.198   0\n",
       "14  15  0.360  0.370   0\n",
       "15  16  0.593  0.042   0\n",
       "16  17  0.719  0.103   0"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "read_path = 'D:\\JWE\\Files\\Courses\\数据挖掘\\melon_data.csv'\n",
    "melon_data = pd.read_csv(read_path)\n",
    "data0 = melon_data.loc[:,['编号','密度','含糖率','好瓜']][:]\n",
    "data0.loc[data0['好瓜'] == '是',['好瓜']] = 1\n",
    "data0.loc[data0['好瓜'] == '否',['好瓜']] = 0\n",
    "data0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-25T09:51:33.213314Z",
     "start_time": "2020-01-25T09:51:33.192254Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "# 正例1，反例0\n",
    "X = data0[['密度','含糖率']].values\n",
    "Y = data0['好瓜'].values\n",
    "X1 = data0.loc[data0['好瓜'] == 1,['密度','含糖率']].values\n",
    "X0 = data0.loc[data0['好瓜'] == 0,['密度','含糖率']].values \n",
    "Y1 = data0.loc[data0['好瓜'] == 1,['好瓜']].values\n",
    "Y0 = data0.loc[data0['好瓜'] == 0,['好瓜']].values\n",
    "\n",
    "X1_mean = np.mean(X1,axis=0)\n",
    "X0_mean = np.mean(X0,axis=0)\n",
    "Sigma1 = np.dot((X1 - X1_mean).T,X1 - X1_mean)\n",
    "Sigma0 = np.dot((X0 - X0_mean).T,X0 - X0_mean)\n",
    "S_w = Sigma1 + Sigma0\n",
    "U,Sigma,VT = np.linalg.svd(S_w)\n",
    "Sigma = np.diag(Sigma)\n",
    "w = np.dot(np.dot(np.dot(VT.T,np.linalg.inv(Sigma)),U.T),X0_mean - X1_mean)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-01-25T09:52:01.482660Z",
     "start_time": "2020-01-25T09:52:00.891586Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e15208>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e32278>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e15278>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e32828>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e32b00>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e32da0>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e32e48>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e38400>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e38748>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e38a20>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e38cc0>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e38f98>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e32e10>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e512e8>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e51668>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e51940>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e51c88>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, '投影前')"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e51e48>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e7c358>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e7c748>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e7ca20>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e7cda0>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e7cd30>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e911d0>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e91518>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e917b8>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e91c18>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e91ef0>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e9a198>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e9a5f8>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e9a908>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e9ac50>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e9af98>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x25350e91be0>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, '投影后')"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "\n",
    "Xw = np.dot(X,w.reshape(2,1))\n",
    "\n",
    "# 显示中字\n",
    "plt.rcParams['font.sans-serif']=['SimHei']\n",
    "\n",
    "# 绘图，红色正例，蓝色反例\n",
    "fig = plt.figure()\n",
    "ax1 = fig.add_subplot(2,2,1)\n",
    "for i in range(len(X)):\n",
    "    if Y[i] == 1:\n",
    "        plt.scatter(X[i,0],X[i,1],c='r')\n",
    "    if Y[i] == 0:\n",
    "        plt.scatter(X[i,0],X[i,1],c='b')\n",
    "# plt.scatter(X[:,0],X[:,1])\n",
    "plt.title('投影前')\n",
    "\n",
    "ax2 = fig.add_subplot(2,2,2)\n",
    "for i in range(len(X)):\n",
    "    if Y[i] == 1:\n",
    "        plt.scatter(Xw[i],0,c='r')\n",
    "    if Y[i] == 0:\n",
    "        plt.scatter(Xw[i],0,c='b')\n",
    "# plt.scatter(Xw,[0 for i in range(len(Xw))])\n",
    "plt.title('投影后')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "## 多分类问题日后再补"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "hide_input": false,
  "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": "3.7.3"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}