ch01-导论/d2-1cluster聚类(手写数字集).ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 无监督学习：聚类\n",
    "- 聚类分析\n",
    "- 手写数据集\n",
    "- 在训练之前，对数字进行聚类，减少我们做标注的时间！"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.cluster import KMeans\n",
    "from sklearn.datasets import load_digits\n",
    "\n",
    "from sklearn.preprocessing import scale\n",
    "from sklearn import metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.  0.  5. ...  0.  0.  0.]\n",
      " [ 0.  0.  0. ... 10.  0.  0.]\n",
      " [ 0.  0.  0. ... 16.  9.  0.]\n",
      " ...\n",
      " [ 0.  0.  1. ...  6.  0.  0.]\n",
      " [ 0.  0.  2. ... 12.  0.  0.]\n",
      " [ 0.  0. 10. ... 12.  1.  0.]]\n",
      "[0 1 2 ... 8 9 8]\n"
     ]
    }
   ],
   "source": [
    "digits = load_digits()\n",
    "print(digits.data)\n",
    "print(digits.target)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.  0.  0. 12. 13.  5.  0.  0.]\n",
      " [ 0.  0.  0. 11. 16.  9.  0.  0.]\n",
      " [ 0.  0.  3. 15. 16.  6.  0.  0.]\n",
      " [ 0.  7. 15. 16. 16.  2.  0.  0.]\n",
      " [ 0.  0.  1. 16. 16.  3.  0.  0.]\n",
      " [ 0.  0.  1. 16. 16.  6.  0.  0.]\n",
      " [ 0.  0.  1. 16. 16.  6.  0.  0.]\n",
      " [ 0.  0.  0. 11. 16. 10.  0.  0.]]\n",
      "1\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 建立一张图\n",
    "figure = plt.figure()\n",
    "# 0~15 16个点，8x8的位图，每个像素是一个灰度\n",
    "img = digits.data[1].reshape((8, 8))\n",
    "label = digits.target[1]\n",
    "print(img)\n",
    "print(label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 显示一张灰度图\n",
    "plt.imshow(img, cmap=plt.cm.gray)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 显示一张图\n",
    "def show_digits(digits_data, digits_label):\n",
    "    figure = plt.figure()\n",
    "    for row in range(4):\n",
    "        for col in range(4):\n",
    "            # 显示4x4图形\n",
    "            # 方格从 1 开始算\n",
    "            ax = figure.add_subplot(4, 4, row*4+ col + 1)\n",
    "            img = digits_data[row*4 + col].reshape((8, 8))\n",
    "            label = digits_label[row * 4 + col]\n",
    "            plt.imshow(img, cmap=plt.cm.gray)\n",
    "            # 一直在subplot上显示\n",
    "            \n",
    "            # 不显示 x、y标签\n",
    "            plt.xticks([])\n",
    "            plt.yticks([])\n",
    "            # 每张图对应的标签\n",
    "            plt.xlabel(label)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "digits = load_digits()\n",
    "datas = digits.data\n",
    "labels = digits.target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 16 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "show_digits(datas, labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- 聚类之前，对数据进行标准化\n",
    "- 灰度图0～15之间，标准化到0～1之间"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.         -0.33501649 -0.04308102 ... -1.14664746 -0.5056698\n",
      "  -0.19600752]\n",
      " [ 0.         -0.33501649 -1.09493684 ...  0.54856067 -0.5056698\n",
      "  -0.19600752]\n",
      " [ 0.         -0.33501649 -1.09493684 ...  1.56568555  1.6951369\n",
      "  -0.19600752]\n",
      " ...\n",
      " [ 0.         -0.33501649 -0.88456568 ... -0.12952258 -0.5056698\n",
      "  -0.19600752]\n",
      " [ 0.         -0.33501649 -0.67419451 ...  0.8876023  -0.5056698\n",
      "  -0.19600752]\n",
      " [ 0.         -0.33501649  1.00877481 ...  0.8876023  -0.26113572\n",
      "  -0.19600752]]\n"
     ]
    }
   ],
   "source": [
    "scale_data = scale(digits.data)\n",
    "print(scale_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "样点量：1797 \t 特征数：64 \t 数字：10\n"
     ]
    }
   ],
   "source": [
    "# 多少样本、多少特征、多少数字\n",
    "sample_count, feature_count = scale_data.shape\n",
    "digit_count = len(np.unique(labels))\n",
    "print('样点量：%d \\t 特征数：%d \\t 数字：%d' %(sample_count, feature_count, digit_count))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1 2 2 ... 2 3 3]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 16 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 开始选中心点 聚成几类 尝试做10次聚类\n",
    "# 随机生产质心，每次结果都不一样\n",
    "estimator = KMeans(init='k-means++', n_clusters=digit_count, n_init=10)\n",
    "estimator.fit(scale_data)\n",
    "\n",
    "# 给聚出来的东西，标上一个标签\n",
    "print(estimator.labels_)\n",
    "# 聚类出来的标签, 结果不一定匹配；聚成同一类\n",
    "show_digits(digits.data, estimator.labels_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "均方差： 69687.7202266693\n",
      "均一性： 0.6778238757139822\n",
      "完整性： 0.7170676325554005\n",
      "综合评估： 0.6968937139072262\n"
     ]
    }
   ],
   "source": [
    "# 评估聚类出来的结果 —— 好不好  —— 分数\n",
    "print('均方差：', estimator.inertia_)\n",
    "print('均一性：', metrics.homogeneity_score(labels, estimator.labels_))\n",
    "print('完整性：', metrics.completeness_score(labels, estimator.labels_))\n",
    "print('综合评估：', metrics.v_measure_score(labels, estimator.labels_))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "<br><br>\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![](images/d2-1.png)\n",
    "![](images/d2-2.png)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4rc1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}