__PLA.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "plt.style.use('seaborn-white')\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "weights 是一个列表，里面存储的是我们用来产生随机数据的那条直线的法向量。\n"
    "weights=[4,3]\n",
    "numLines=100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "w=np.array(weights)\n",
    "num_features=len(weights)\n",
    "dataset=np.zeros((numLines,num_features+1))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(numLines):\n",
    "    x=np.random.rand(1,num_features)*20-10\n",
    "    # 计算內积\n",
    "    inner_product=np.sum(w*x)\n",
    "    # 小于0则是代表负实例，大于0则是代表正实例\n",
    "    if inner_product<=0:\n",
    "        dataset[i]=np.append(x,-1)\n",
    "    else:\n",
    "        dataset[i]=np.append(x,1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x393f690>"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "# 可视化我们的数据\n",
    "fig,ax=plt.subplots(figsize=(8,8))\n",
    "plt.title(\"Liner sepaeable data set\")\n",
    "plt.xlabel(\"X\")\n",
    "plt.ylabel(\"Y\")\n",
    "x1=dataset[dataset[:,2]==1][:,:2]\n",
    "# x1[:,0]代表第一个特征，x1[:,1]代表第二个特征\n",
    "plt.scatter(x1[:,0],x1[:,1],marker='o',color='g',label=1,s=20,cmap='viridis')\n",
    "\n",
    "x2=dataset[dataset[:,2]==-1][:,:2]\n",
    "plt.scatter(x2[:,0],x2[:,1],marker='x',color='r',label=-1,s=20,cmap='viridis')\n",
    "plt.legend(loc=\"upper right\",frameon=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "_w=np.zeros((1,num_features-1))  # initialize weights\n",
    "separated=False\n",
    "i=0\n",
    "while not separated and i<numLines:\n",
    "    if dataset[i][-1]*np.sum(_w*dataset[i,0:-1])<=0:  #假如数据分类错误\n",
    "        _w=_w+dataset[i][-1]*dataset[i,0:-1]   #更新权值\n",
    "        separated=False\n",
    "        i=0\n",
    "    else:\n",
    "        i+=1   #分类正确则更新下一个数据点\n",
    "_w"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}