CH06 SVM.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# CH6 SVM"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* PRML\n",
    "* Learning with Kernels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:09.941319Z",
     "start_time": "2018-09-28T08:26:08.761929Z"
    }
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## MLiA"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load Data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### readlines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:09.952252Z",
     "start_time": "2018-09-28T08:26:09.943890Z"
    }
   },
   "outputs": [],
   "source": [
    "def loadDataSet(fileName):\n",
    "    dataMat = []; labelMat = []\n",
    "    fr = open(fileName)\n",
    "    for line in fr.readlines():\n",
    "        lineArr = line.strip().split(\"\\t\")\n",
    "        dataMat.append([float(lineArr[0]),float(lineArr[1])])\n",
    "        labelMat.append(float(lineArr[2]))\n",
    "    return dataMat, labelMat\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### pandas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:09.967112Z",
     "start_time": "2018-09-28T08:26:09.955023Z"
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "def loadDataSet(fileName):\n",
    "    df = pd.read_table(fileName, header=None)\n",
    "    dataMat = df[[0,1]].values\n",
    "    labelMat = df[2].values\n",
    "    return dataMat, labelMat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:10.007273Z",
     "start_time": "2018-09-28T08:26:09.969349Z"
    }
   },
   "outputs": [],
   "source": [
    "dataArr, labelArr = loadDataSet(\"./Data/CH6/testSet.txt\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### helper function"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "关于clipAlpha，可以参考《统计学习方法》中的描述\n",
    "$H,L$是$\\alpha$取值的界\n",
    "\n",
    "另外具体的Clipping是源于Platt的文章[2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:10.015525Z",
     "start_time": "2018-09-28T08:26:10.009972Z"
    }
   },
   "outputs": [],
   "source": [
    "def selectJrand(i,m):\n",
    "    j=i\n",
    "    while (j==i):\n",
    "        j = int(np.random.uniform(0,m))\n",
    "    return j\n",
    "\n",
    "def clipAlpha(aj, H, L):\n",
    "    if aj > H:\n",
    "        aj = H\n",
    "    if L > aj:\n",
    "        aj = L\n",
    "    return aj"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### SMO"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### smoSimple\n",
    "C用来控制**函数间隔**以及**保证大部分点的函数间隔小于1.0**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:10.040223Z",
     "start_time": "2018-09-28T08:26:10.018885Z"
    }
   },
   "outputs": [],
   "source": [
    "def smoSimple(dataMatIn, classLabels, C, toler, maxIter):\n",
    "    dataMatrix = np.mat(dataMatIn)             # mxn = 100x2\n",
    "    labelMat = np.mat(classLabels).transpose() # mx1 = 100x1\n",
    "    b = 0; m,n = np.shape(dataMatrix)          # mxn\n",
    "    alphas = np.mat(np.zeros((m,1)))           # mx1 \n",
    "    iter_ = 0\n",
    "    while (iter_ < maxIter):\n",
    "        alphaPairsChanged = 0\n",
    "        # 遍历整个数据集中每个样本，要知道alpha的维度和样本的维度一样\n",
    "        for i in range(m):\n",
    "            # f(x) at i \n",
    "            # 对应了李航书上的g(x),这里还没有用核函数 7.104\n",
    "            #1\n",
    "            fXi = float(np.multiply(alphas,labelMat).T*\\\n",
    "                        (dataMatrix*dataMatrix[i,:].T)) + b #2\n",
    "                        # [1]\n",
    "                        # a * y * x * xi + b \n",
    "                        # multiply(mx1,mx1).T = multipy(100x1,100x1).T = (100x1).T = (1x100)\n",
    "                        #[2]\n",
    "                        # (mxn)x(1xn).T = (100x2)x(1x2).T = (100x2)x(2x1) = (100x1)\n",
    "            # 7.105\n",
    "            Ei = fXi - float(labelMat[i])\n",
    "            # 对应了两边的边界情况            \n",
    "            if ((labelMat[i]*Ei < - toler) and (alphas[i] < C)) or \\\n",
    "               ((labelMat[i]*Ei > toler) and (alphas[i] > 0)):\n",
    "                    # f(x) at j\n",
    "                    j = selectJrand(i,m)\n",
    "                    # 到这里i,j都定了\n",
    "                    fXj = float(np.multiply(alphas,labelMat).T*\\\n",
    "                                (dataMatrix*dataMatrix[j,:].T)) + b\n",
    "                    Ej = fXj - float(labelMat[j])\n",
    "                    \n",
    "                    alphaIold = alphas[i].copy()\n",
    "                    alphaJold = alphas[j].copy()\n",
    "                    \n",
    "                    if (labelMat[i] != labelMat[j]):\n",
    "                        L = max(0, alphas[j] - alphas[i])\n",
    "                        H = min(C, C + alphas[j] - alphas[i])\n",
    "                    else:\n",
    "                        L = max(0, alphas[j] + alphas[i] - C)\n",
    "                        H = min(C, alphas[j] + alphas[i])\n",
    "                    if L==H: \n",
    "                        print(\"L==H\")\n",
    "                        continue\n",
    "                    \n",
    "                    # 1x2 x (1x2).T = 1x2x2x1\n",
    "                    # alpha 最优修改量 = xj^2 - xi^2 - 2xixj = (xj - xi)^2 - 2xi^2\n",
    "                    # 7.107\n",
    "                    eta = 2*dataMatrix[i,:]*dataMatrix[j,:].T-\\\n",
    "                          dataMatrix[i,:]*dataMatrix[i,:].T-\\\n",
    "                          dataMatrix[j,:]*dataMatrix[j,:].T\n",
    "                    # 简化处理\n",
    "                    if eta >= 0:\n",
    "                        print(\"eta>=0\")\n",
    "                        continue\n",
    "                    # alphas[j]\n",
    "                    alphas[j] -= labelMat[j]*(Ei - Ej)/eta\n",
    "                    alphas[j] = clipAlpha(alphas[j],H,L)\n",
    "                    \n",
    "                    if (abs(alphas[j] - alphaJold) < 0.00001):\n",
    "                        print(\"j not moving enough\")\n",
    "                        continue\n",
    "                    # alphas[i]\n",
    "                    alphas[i] += labelMat[j]*labelMat[i]*(alphaJold - alphas[j])\n",
    "                    # \n",
    "                    b1 = b - Ei - labelMat[i]*(alphas[i] - alphaIold)*\\\n",
    "                         dataMatrix[i,:]*dataMatrix[i,:].T - \\\n",
    "                         labelMat[j]*(alphas[j]-alphaJold)*\\\n",
    "                         dataMatrix[i,:]*dataMatrix[j,:].T\n",
    "                    \n",
    "                    b2 = b - Ej - labelMat[i]*(alphas[i]-alphaIold)*\\\n",
    "                         dataMatrix[i,:]*dataMatrix[j,:].T -\\\n",
    "                         labelMat[j]*(alphas[j]-alphaJold)*\\\n",
    "                         dataMatrix[j,:]*dataMatrix[j,:].T\n",
    "                    \n",
    "                    if (0 < alphas[i]) and (C > alphas[j]):\n",
    "                        b = b1\n",
    "                    elif (0 < alphas[j]) and (C > alphas[j]):\n",
    "                        b = b2\n",
    "                    else:\n",
    "                        b = (b1 + b2)/2\n",
    "                    alphaPairsChanged += 1\n",
    "                    print(\"iter: %d i: %d, paris changed %d\" % (iter_, i, alphaPairsChanged))\n",
    "        if (alphaPairsChanged == 0):\n",
    "            iter_ += 1\n",
    "        else:\n",
    "            iter_ = 0\n",
    "        print(\"iteration number: %d\" % iter_)\n",
    "    return b, alphas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:13.847348Z",
     "start_time": "2018-09-28T08:26:10.042734Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iter: 0 i: 0, paris changed 1\n",
      "j not moving enough\n",
      "iter: 0 i: 5, paris changed 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "iter: 0 i: 25, paris changed 3\n",
      "L==H\n",
      "L==H\n",
      "j not moving enough\n",
      "iter: 0 i: 34, paris changed 4\n",
      "iter: 0 i: 39, paris changed 5\n",
      "L==H\n",
      "j not moving enough\n",
      "iter: 0 i: 54, paris changed 6\n",
      "iter: 0 i: 55, paris changed 7\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "L==H\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "L==H\n",
      "L==H\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "L==H\n",
      "L==H\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "L==H\n",
      "iter: 0 i: 21, paris changed 1\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 53, paris changed 2\n",
      "j not moving enough\n",
      "iter: 0 i: 55, paris changed 3\n",
      "iter: 0 i: 62, paris changed 4\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 10, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 29, paris changed 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 55, paris changed 3\n",
      "j not moving enough\n",
      "iter: 0 i: 97, paris changed 4\n",
      "iteration number: 0\n",
      "iter: 0 i: 0, paris changed 1\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 29, paris changed 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 43, paris changed 3\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "iter: 0 i: 55, paris changed 4\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 23, paris changed 1\n",
      "iter: 0 i: 25, paris changed 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 52, paris changed 3\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 17, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 97, paris changed 1\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 69, paris changed 1\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 54, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 23, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 54, paris changed 2\n",
      "iter: 0 i: 55, paris changed 3\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 97, paris changed 4\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 23, paris changed 1\n",
      "j not moving enough\n",
      "iter: 0 i: 46, paris changed 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 69, paris changed 3\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "iter: 0 i: 0, paris changed 1\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 39, paris changed 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 17, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "iter: 0 i: 0, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 14, paris changed 2\n",
      "j not moving enough\n",
      "iter: 0 i: 23, paris changed 3\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 54, paris changed 4\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "iter: 0 i: 8, paris changed 1\n",
      "j not moving enough\n",
      "iter: 0 i: 11, paris changed 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "L==H\n",
      "L==H\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "iter: 0 i: 8, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "iter: 1 i: 43, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 1 i: 54, paris changed 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "iter: 0 i: 0, paris changed 1\n",
      "iter: 0 i: 8, paris changed 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 2 i: 54, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 2 i: 94, paris changed 2\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "L==H\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "iter: 1 i: 39, paris changed 1\n",
      "L==H\n",
      "iter: 1 i: 46, paris changed 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 55, paris changed 1\n",
      "L==H\n",
      "j not moving enough\n",
      "L==H\n",
      "L==H\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 23, paris changed 1\n",
      "j not moving enough\n",
      "iter: 0 i: 39, paris changed 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "L==H\n",
      "L==H\n",
      "j not moving enough\n",
      "iter: 0 i: 8, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 47, paris changed 2\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "L==H\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 10, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "L==H\n",
      "iter: 0 i: 46, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 77, paris changed 1\n",
      "j not moving enough\n",
      "L==H\n",
      "iteration number: 0\n",
      "L==H\n",
      "iter: 0 i: 8, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 54, paris changed 1\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 3\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 4\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 4 i: 52, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 31, paris changed 1\n",
      "L==H\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "L==H\n",
      "iteration number: 0\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "iteration number: 1\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 1 i: 17, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 3\n",
      "j not moving enough\n",
      "iter: 3 i: 10, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 24, paris changed 1\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 79, paris changed 2\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 52, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "iter: 1 i: 10, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "iter: 0 i: 52, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "iter: 0 i: 8, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "iter: 0 i: 54, paris changed 1\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 3\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "L==H\n",
      "iter: 3 i: 52, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "iter: 0 i: 23, paris changed 1\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "iter: 1 i: 29, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 2\n",
      "iter: 2 i: 17, paris changed 1\n",
      "iter: 2 i: 23, paris changed 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 3\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 4\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 5\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 6\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 7\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 7 i: 55, paris changed 1\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 3\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 4\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 5\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 6\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 6 i: 52, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 2 i: 54, paris changed 1\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 3\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 4\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 5\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 6\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 7\n",
      "j not moving enough\n",
      "iter: 7 i: 29, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 3\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 4\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 5\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 6\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 7\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 8\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 9\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 10\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 11\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 12\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 13\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 13 i: 54, paris changed 1\n",
      "iteration number: 0\n",
      "j not moving enough\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 2\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 3\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 4\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 5\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 6\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 7\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 8\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 9\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 10\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 11\n",
      "iter: 11 i: 17, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 0 i: 55, paris changed 1\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 3\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 4\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 5\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 6\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 7\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 8\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 9\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 10\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 11\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 12\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 13\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 14\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 15\n",
      "j not moving enough\n",
      "L==H\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 16\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 16 i: 52, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 3\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 4\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 5\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 6\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iter: 6 i: 55, paris changed 1\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 3\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 4\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 5\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 6\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 7\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 8\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 9\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 10\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 11\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 12\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 13\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 14\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 15\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 16\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 17\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 18\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 19\n",
      "iter: 19 i: 17, paris changed 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 3\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 4\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 5\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 6\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 7\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 8\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 9\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 10\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 11\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 12\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 13\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 14\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 15\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 16\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 17\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 18\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 19\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 20\n",
      "j not moving enough\n",
      "iter: 20 i: 55, paris changed 1\n",
      "iteration number: 0\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 1\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 2\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 3\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 4\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 5\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 6\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 7\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 8\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 9\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 10\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 11\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 12\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 13\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 14\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 15\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 16\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 17\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 18\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 19\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 20\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 21\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 22\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 23\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 24\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 25\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 26\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 27\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 28\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 29\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 30\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 31\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 32\n",
      "j not moving enough\n",
      "j not moving enough\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration number: 33\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 34\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 35\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 36\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 37\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 38\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 39\n",
      "j not moving enough\n",
      "j not moving enough\n",
      "iteration number: 40\n"
     ]
    }
   ],
   "source": [
    "b, alphas = smoSimple(dataArr, labelArr, 0.6, 0.001, 40)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### 拉格朗日乘子"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:13.869394Z",
     "start_time": "2018-09-28T08:26:13.850756Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "matrix([[0.11448958, 0.24978708, 0.00919759, 0.35507907]])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "alpha = alphas[alphas > 0]\n",
    "alpha"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### 支持向量\n",
    "$\\alpha > 0$的对应支持向量， 注意支持向量可能在边界，可能在边界中，也可能错分一侧"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:13.892234Z",
     "start_time": "2018-09-28T08:26:13.872575Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[4.658191 3.507396] -1\n",
      "[ 3.457096 -0.082216] -1\n",
      "[ 5.286862 -2.358286] 1\n",
      "[6.080573 0.418886] 1\n"
     ]
    }
   ],
   "source": [
    "for i in range(100):\n",
    "    if alphas[i] > 0:\n",
    "        print(dataArr[i],labelArr[i])      "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:13.908683Z",
     "start_time": "2018-09-28T08:26:13.896717Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[4.658191 3.507396] -1\n",
      "[ 3.457096 -0.082216] -1\n",
      "[ 5.286862 -2.358286] 1\n",
      "[6.080573 0.418886] 1\n",
      "[[5.55111512e-17]]\n"
     ]
    }
   ],
   "source": [
    "result = 0\n",
    "for i in range(100):\n",
    "    if alphas[i] > 0:\n",
    "        result += labelArr[i]*alphas[i]\n",
    "        print(dataArr[i],labelArr[i])\n",
    "print(result) # sum(alpha*y) = 0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### 超平面"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:13.951341Z",
     "start_time": "2018-09-28T08:26:13.911674Z"
    },
    "scrolled": 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>x0</th>\n",
       "      <th>x1</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3.542485</td>\n",
       "      <td>1.977398</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3.018896</td>\n",
       "      <td>2.556416</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>7.551510</td>\n",
       "      <td>-1.580030</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2.114999</td>\n",
       "      <td>-0.004466</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>8.127113</td>\n",
       "      <td>1.274372</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         x0        x1  y\n",
       "0  3.542485  1.977398 -1\n",
       "1  3.018896  2.556416 -1\n",
       "2  7.551510 -1.580030  1\n",
       "3  2.114999 -0.004466 -1\n",
       "4  8.127113  1.274372  1"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_table(\"./Data/CH6/testSet.txt\", header=None)\n",
    "df.columns = [\"x0\",\"x1\",\"y\"]\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:13.990511Z",
     "start_time": "2018-09-28T08:26:13.956643Z"
    }
   },
   "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>x0</th>\n",
       "      <th>x1</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>100.000000</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>100.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>4.790926</td>\n",
       "      <td>0.059057</td>\n",
       "      <td>-0.080000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>3.103636</td>\n",
       "      <td>1.465117</td>\n",
       "      <td>1.001817</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-0.743036</td>\n",
       "      <td>-5.766721</td>\n",
       "      <td>-1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.976116</td>\n",
       "      <td>-0.573050</td>\n",
       "      <td>-1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>3.539520</td>\n",
       "      <td>-0.009497</td>\n",
       "      <td>-1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>8.034324</td>\n",
       "      <td>0.810767</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>9.854303</td>\n",
       "      <td>3.964960</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               x0          x1           y\n",
       "count  100.000000  100.000000  100.000000\n",
       "mean     4.790926    0.059057   -0.080000\n",
       "std      3.103636    1.465117    1.001817\n",
       "min     -0.743036   -5.766721   -1.000000\n",
       "25%      1.976116   -0.573050   -1.000000\n",
       "50%      3.539520   -0.009497   -1.000000\n",
       "75%      8.034324    0.810767    1.000000\n",
       "max      9.854303    3.964960    1.000000"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:13.999934Z",
     "start_time": "2018-09-28T08:26:13.992989Z"
    }
   },
   "outputs": [],
   "source": [
    "def calcWs(alphas, dataArr, classLabels):\n",
    "    X = np.mat(dataArr)\n",
    "    labelMat = np.mat(classLabels).T\n",
    "    m,n = np.shape(X)\n",
    "    w = np.zeros((n,1))\n",
    "    for i in range(m):\n",
    "        w += np.multiply(alphas[i]*labelMat[i],X[i,:].T)\n",
    "    return w"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:14.020137Z",
     "start_time": "2018-09-28T08:26:14.003357Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.8108583535261407, -0.2539767167836068] [-3.8240961202888877]\n",
      "[-2, 12] [-21.442173504357527, 23.254903822766572]\n"
     ]
    }
   ],
   "source": [
    "HP_ws = calcWs(alphas, dataArr, labelArr).reshape(2,).tolist()\n",
    "HP_b = b.tolist()[0]\n",
    "HP_x = [-2,12]\n",
    "HP_y_L = [-HP_ws[0]*x/HP_ws[1] - (HP_b[0]+1)/HP_ws[1] for x in HP_x]\n",
    "HP_y_H = [-HP_ws[0]*x/HP_ws[1] - (HP_b[0]-1)/HP_ws[1] for x in HP_x]\n",
    "HP_y = [-HP_ws[0]*x/HP_ws[1] - (HP_b[0]+0)/HP_ws[1] for x in HP_x]\n",
    "\n",
    "print(HP_ws,HP_b)\n",
    "print(HP_x,HP_y)\n",
    "# HP_y -> x1, HP_x -> x0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Visualization\n",
    "这是个线性可分的数据集"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:14.353759Z",
     "start_time": "2018-09-28T08:26:14.022322Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# sns.pointplot(x=0,y=1,data=df,hue=2,linestyles=\"-\")\n",
    "# sns.pairplot(data=df,vars = [0,1],hue=2)\n",
    "# g = sns.jointplot(x=0,y=1,data=df,kind=\"kde\")\n",
    "# g.plot_joint(plt.scatter, c=\"w\", s=30, linewidth=1, marker=\"+\")\n",
    "\n",
    "%matplotlib inline\n",
    "plt.figure(figsize = (5,5))\n",
    "\n",
    "supportVactor = df[list(map(lambda x:x[0]>0,alphas.tolist()))]\n",
    "plt.scatter(x=df[df[\"y\"] == 1][\"x0\"],y=df[df[\"y\"] == 1][\"x1\"],\n",
    "            s = 50, marker = \"v\",c=\"red\")\n",
    "plt.scatter(x=df[df[\"y\"] == -1][\"x0\"],y=df[df[\"y\"] == -1][\"x1\"],\n",
    "            s = 50, c=\"blue\")\n",
    "plt.scatter(x=supportVactor[\"x0\"],y=supportVactor[\"x1\"],\n",
    "            s = 300, marker=\"o\",c = \"none\",edgecolors=\"black\")\n",
    "\n",
    "plt.plot(HP_x,HP_y)\n",
    "plt.plot(HP_x,HP_y_H,\"--\")\n",
    "plt.plot(HP_x,HP_y_L,\"--\")\n",
    "plt.xlim(-2,12)\n",
    "plt.ylim(-8,6)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### smoP"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 支持向量求解"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### w求解"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 可视化"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 核函数"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 实例"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Mnist"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Sklearn"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:14.370329Z",
     "start_time": "2018-09-28T08:26:14.363582Z"
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "def loadDataSet(fileName):\n",
    "    df = pd.read_table(fileName, header=None)\n",
    "    dataMat = df[[0,1]].values\n",
    "    labelMat = df[2].values\n",
    "    return dataMat, labelMat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:14.400843Z",
     "start_time": "2018-09-28T08:26:14.390715Z"
    }
   },
   "outputs": [],
   "source": [
    "dataMat, labelMat = loadDataSet(fileName=\"./Data/CH6/testSet.txt\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Modeling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:14.538847Z",
     "start_time": "2018-09-28T08:26:14.405711Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.svm import SVC\n",
    "clf = SVC(kernel=\"linear\")\n",
    "# default rbf kernel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:14.551346Z",
     "start_time": "2018-09-28T08:26:14.542302Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape='ovr', degree=3, gamma='auto', kernel='linear',\n",
       "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "  tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = dataMat\n",
    "y = labelMat\n",
    "clf.fit(X,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:14.562671Z",
     "start_time": "2018-09-28T08:26:14.554723Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([2, 1], dtype=int32)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.n_support_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:14.585672Z",
     "start_time": "2018-09-28T08:26:14.565194Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([17, 29, 55], dtype=int32)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.support_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:14.603633Z",
     "start_time": "2018-09-28T08:26:14.588453Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 4.658191,  3.507396],\n",
       "       [ 3.457096, -0.082216],\n",
       "       [ 6.080573,  0.418886]])"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.support_vectors_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:14.833459Z",
     "start_time": "2018-09-28T08:26:14.605626Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plt.scatter(X[np.where(y ==1)][:,0],X[np.where(y ==1)][:,1])\n",
    "# plt.scatter(X[np.where(y ==-1)][:,0],X[np.where(y ==-1)][:,1])\n",
    "X0,X1 = X[:,0],X[:,1]\n",
    "plt.scatter(X0,X1,c=y,cmap=plt.cm.coolwarm)\n",
    "plt.scatter(clf.support_vectors_[:,0],clf.support_vectors_[:,1],marker=\"o\",s=100,edgecolors=\"black\",c=\"\")\n",
    "\n",
    "# plot the decision function\n",
    "ax = plt.gca()\n",
    "xlim = ax.get_xlim()\n",
    "ylim = ax.get_ylim()\n",
    "\n",
    "# create grid to evaluate model\n",
    "xx = np.linspace(xlim[0], xlim[1], 30)\n",
    "yy = np.linspace(ylim[0], ylim[1], 30)\n",
    "YY, XX = np.meshgrid(yy, xx)\n",
    "xy = np.vstack([XX.ravel(), YY.ravel()]).T\n",
    "Z = clf.decision_function(xy).reshape(XX.shape)\n",
    "\n",
    "# plot decision boundary and margins\n",
    "ax.contour(XX, YY, Z, colors='k', levels=[-1, 0, 1], alpha=0.5,\n",
    "           linestyles=['--', '-', '--'])\n",
    "# plot support vectors\n",
    "ax.scatter(clf.support_vectors_[:, 0], clf.support_vectors_[:, 1], s=100,\n",
    "           linewidth=1, facecolors='none')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Comprasion"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:15.364928Z",
     "start_time": "2018-09-28T08:26:14.836798Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x720 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "C = 1\n",
    "models = [SVC(kernel=\"linear\",C=C),\n",
    "          SVC(kernel=\"rbf\",gamma=0.7,C=C),\n",
    "          SVC(kernel=\"poly\",degree=3,C=C)]\n",
    "models = [clf.fit(X,y) for clf in models]\n",
    "# title for the plots\n",
    "titles = ['SVC with linear kernel',\n",
    "          'SVC with RBF kernel',\n",
    "          'SVC with polynomial (degree 3) kernel']\n",
    "\n",
    "fig, sub = plt.subplots(3, 1)\n",
    "fig.set_figheight(10)\n",
    "plt.subplots_adjust(wspace=0.4, hspace=0.4)\n",
    "X0, X1 = X[:, 0], X[:, 1]\n",
    "h = 0.1\n",
    "x_min, x_max = X0.min() - 1, X0.max() + 1\n",
    "y_min, y_max = X1.min() - 1, X1.max() + 1\n",
    "xx, yy = np.meshgrid(np.arange(x_min, x_max, h),\n",
    "                     np.arange(y_min, y_max, h))\n",
    "xy = np.c_[xx.ravel(), yy.ravel()]\n",
    "\n",
    "Z = (clf.predict(xy).reshape(xx.shape) for clf in models)\n",
    "\n",
    "for clf, title, ax, z in zip(models, titles, sub.flatten(),Z):\n",
    "    ax.contourf(xx, yy, z, levels=[-1, 0, 1], alpha=0.5,cmap=plt.cm.coolwarm)\n",
    "    ax.scatter(X0,X1, c=y, cmap=plt.cm.coolwarm, s=20, edgecolors='k')\n",
    "    ax.set_xlim(xx.min(), xx.max())\n",
    "    ax.set_ylim(yy.min(), yy.max())\n",
    "    ax.set_title(title)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### EX1 : RBF Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:15.384871Z",
     "start_time": "2018-09-28T08:26:15.367228Z"
    }
   },
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.214824</td>\n",
       "      <td>0.662756</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.061569</td>\n",
       "      <td>-0.091875</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.406933</td>\n",
       "      <td>0.648055</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.223650</td>\n",
       "      <td>0.130142</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.231317</td>\n",
       "      <td>0.766906</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1    2\n",
       "0 -0.214824  0.662756 -1.0\n",
       "1 -0.061569 -0.091875  1.0\n",
       "2  0.406933  0.648055 -1.0\n",
       "3  0.223650  0.130142  1.0\n",
       "4  0.231317  0.766906 -1.0"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "RBFTestData = pd.read_table(\"./Data/CH6/testSetRBF.txt\",header=None)\n",
    "RBFTestData.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:15.409510Z",
     "start_time": "2018-09-28T08:26:15.387154Z"
    }
   },
   "outputs": [],
   "source": [
    "X = RBFTestData[RBFTestData.columns.tolist()[:-1]].values\n",
    "y = RBFTestData[RBFTestData.columns.tolist()[-1]].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:16.151481Z",
     "start_time": "2018-09-28T08:26:15.412655Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x1080 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "C = 1\n",
    "models = [SVC(kernel=\"linear\",C=C),\n",
    "          SVC(kernel=\"rbf\",gamma=0.4,C=C),\n",
    "          SVC(kernel=\"poly\",degree=3,C=C)]\n",
    "models = [clf.fit(X,y) for clf in models]\n",
    "# title for the plots\n",
    "titles = ['SVC with linear kernel',\n",
    "          'SVC with RBF kernel',\n",
    "          'SVC with polynomial (degree 3) kernel']\n",
    "fig.clf()\n",
    "fig, sub = plt.subplots(3, 1)\n",
    "fig.set_figheight(15)\n",
    "X0, X1 = X[:, 0], X[:, 1]\n",
    "h = 0.1\n",
    "x_min, x_max = X0.min() - 1, X0.max() + 1\n",
    "y_min, y_max = X1.min() - 1, X1.max() + 1\n",
    "xx, yy = np.meshgrid(np.arange(x_min, x_max, h),\n",
    "                     np.arange(y_min, y_max, h))\n",
    "xy = np.c_[xx.ravel(), yy.ravel()]\n",
    "\n",
    "Z = (clf.predict(xy).reshape(xx.shape) for clf in models)\n",
    "\n",
    "for clf, title, ax, z in zip(models, titles, sub.flatten(),Z):\n",
    "    ax.contourf(xx, yy, z, levels=[-1, 0, 1], alpha=0.5,cmap=plt.cm.coolwarm)\n",
    "    ax.scatter(X0,X1, c=y, cmap=plt.cm.coolwarm, s=20, edgecolors='k')\n",
    "    ax.scatter(clf.support_vectors_[:,0],clf.support_vectors_[:,1],c=\"\",edgecolors=\"black\",s=100)\n",
    "    ax.set_xlim(xx.min(), xx.max())\n",
    "    ax.set_ylim(yy.min(), yy.max())\n",
    "    ax.set_title(title)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### EX2 : MNIST"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Load Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:16.157950Z",
     "start_time": "2018-09-28T08:26:16.154081Z"
    }
   },
   "outputs": [],
   "source": [
    "import glob,os\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:16.171768Z",
     "start_time": "2018-09-28T08:26:16.161784Z"
    }
   },
   "outputs": [],
   "source": [
    "def loadImage(files):\n",
    "    lst = glob.glob(files)\n",
    "    samples = []\n",
    "    labels = []\n",
    "    for filepath in lst:\n",
    "        num = pd.read_csv(filepath,header=None).values\n",
    "        fpath,fname = os.path.split(filepath)\n",
    "        labels.append(int(fname.split(\"_\")[0]))\n",
    "        samples.append([int(x) for x in \"\".join([x[0] for x in num])])\n",
    "    return samples, labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:20.566610Z",
     "start_time": "2018-09-28T08:26:16.174896Z"
    }
   },
   "outputs": [],
   "source": [
    "X,y = loadImage(\"./Data/CH2/digits/trainingDigits/*.txt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:20.576121Z",
     "start_time": "2018-09-28T08:26:20.568930Z"
    }
   },
   "outputs": [],
   "source": [
    "X_train,X_test, y_train, y_test = train_test_split(X,y,test_size = 0.1,stratify=y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Modeling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:21.954647Z",
     "start_time": "2018-09-28T08:26:20.578557Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=200, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape='ovr', degree=3, gamma='auto', kernel='rbf',\n",
       "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "  tol=0.0001, verbose=False)"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf = SVC(kernel=\"rbf\",C=200,tol=0.0001)\n",
    "clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:22.187867Z",
     "start_time": "2018-09-28T08:26:21.957579Z"
    }
   },
   "outputs": [],
   "source": [
    "y_pred = clf.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-28T08:26:22.197512Z",
     "start_time": "2018-09-28T08:26:22.191319Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9845360824742269"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(y_pred == y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ref\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. [Decision Function](http://scikit-learn.org/stable/auto_examples/svm/plot_separating_hyperplane.html#sphx-glr-auto-examples-svm-plot-separating-hyperplane-py)\n",
    "1. [Fast training of support vector machines using sequential minimal optimization](http://www.cs.utsa.edu/~bylander/cs6243/smo-book.pdf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [default]",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": true,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}