ch9_4_k_means.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ch9.4 k_means\n",
    "试编程实现 K 均值算法设置三组不同的K值、三组不同初始中心点，在西瓜数据集4.0上进行实验比较，并讨论什么样的初始中心有利于取得好结果。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "dataset = pd.read_csv('data/table_9_1_watermelon_4_0.csv')\n",
    "dataset = dataset.drop(['Idx'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def get_nearest_dist_num(x_j, u):\n",
    "    dist = np.zeros(len(u))\n",
    "    for i in range(len(u)):\n",
    "        dist[i] = np.sum((x_j-u[i])**2)\n",
    "    return np.argmin(dist)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def k_means(dataset, k=3):\n",
    "    random_int = np.random.randint(0,len(dataset),size=k)\n",
    "    while len(np.unique(random_int)) < k:\n",
    "        random_int = np.random.randint(0,len(dataset),size=k)\n",
    "    count = 0\n",
    "    u = np.zeros([k, 2])\n",
    "    print('k-means: k=',end=\"\")\n",
    "    print(k)\n",
    "    for i in range(k):\n",
    "        u[i] = list(dataset.ix[random_int[i],:-1])\n",
    "    print(u)\n",
    "\n",
    "    for t in [221,222,223,224]:\n",
    "        C = dict()\n",
    "        for i in range(k):\n",
    "            C[i] = []\n",
    "        for j in range(len(dataset)):\n",
    "            lambda_j = get_nearest_dist_num(np.array(dataset.ix[j,:-1]), u)\n",
    "            C[lambda_j].append(list(dataset.ix[j,:-1]))\n",
    "        for i in range(k):\n",
    "            u[i] = np.mean(C[i], axis=0)\n",
    "\n",
    "        dataset0 = dataset[dataset['label']==0]\n",
    "        dataset1 = dataset[dataset['label']==1]\n",
    "        color = ['g', 'b', 'r']\n",
    "        fig = plt.subplot(t)\n",
    "        for i in range(k):\n",
    "            fig.plot(np.array(C[i])[:,0], np.array(C[i])[:,1], color[i]+'.')\n",
    "            fig.plot(u[i,0],u[i,1], color[i]+'+')\n",
    "\n",
    "        plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "k-means: k=2\n",
      "[[ 0.725  0.445]\n",
      " [ 0.774  0.376]]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAMUAAACFCAYAAADrYDycAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nbK/oF1vvI4QAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x75edb00>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAMUAAACFCAYAAADrYDycAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x75d7da0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAMUAAACFCAYAAADrYDycAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7649710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAMUAAACFCAYAAADrYDycAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x766f470>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "k_means(dataset, k=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "k-means: k=2\n",
      "[[ 0.478  0.437]\n",
      " [ 0.525  0.369]]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAMUAAACFCAYAAADrYDycAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nUXPogasAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xadf06a0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAMUAAACFCAYAAADrYDycAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIikKCpEUBYVIioJCJEVBIZLyf94WF99rCJnuAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xadae278>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAMUAAACFCAYAAADrYDycAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nkRQFhUiKgkIkRUEhkqKgEElRUIik/B+ocqc6CnzjAAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xae5de80>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAMUAAACFCAYAAADrYDycAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x9ad34e0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "k_means(dataset, k=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda root]",
   "language": "python",
   "name": "conda-root-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  },
  "toc": {
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": "block",
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}