FireAI_025_DBSCANAlgorithm.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "np.random.seed(37) # 使得每次运行得到的随机数都一样"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 250 entries, 0 to 249\n",
      "Data columns (total 2 columns):\n",
      "0    250 non-null float64\n",
      "1    250 non-null float64\n",
      "dtypes: float64(2)\n",
      "memory usage: 4.0 KB\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "# 准备数据集\n",
    "data_path='E:\\PyProjects\\DataSet\\FireAI/data_perf.txt'\n",
    "df=pd.read_csv(data_path,header=None)\n",
    "print(df.info()) # 查看数据信息，确保没有错误\n",
    "# print('-'*100)\n",
    "dataset=df.values # 数据加载没有问题\n",
    "# print(dataset.shape) # (250, 2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DBSCAN(algorithm='auto', eps=0.5, leaf_size=30, metric='euclidean',\n",
       "    metric_params=None, min_samples=5, n_jobs=1, p=None)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 定义一个DBCSCAN模型，并用数据集训练它\n",
    "from sklearn.cluster import DBSCAN\n",
    "model=DBSCAN(eps=0.5,min_samples=5) # 此处的参数是随便指定\n",
    "model.fit(dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "si_score: 0.5134\n"
     ]
    }
   ],
   "source": [
    "# 使用轮廓系数评估模型的优虐\n",
    "from sklearn.metrics import silhouette_score\n",
    "si_score=silhouette_score(dataset,model.labels_)\n",
    "print('si_score: {:.4f}'.format(si_score))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 在定义DBSCAN时，往往我们很难知道最优的eps参数，\n",
    "# 故而可以通过遍历得到最优值\n",
    "def get_optimal_eps(dataset,eps_list):\n",
    "    '''get optimal eps param for DBSCAN \n",
    "       params: \n",
    "            dataset: the whole dataset.\n",
    "            eps_list: must be in np.linspace() format or list format.\n",
    "        return:\n",
    "            three values：optimal eps value, \n",
    "                optimal model with optimal eps\n",
    "                silhouette_scores of all candidate eps.\n",
    "        '''\n",
    "    scores=[]\n",
    "    models=[]\n",
    "    for eps in eps_list:\n",
    "        model=DBSCAN(eps=eps,min_samples=5).fit(dataset)\n",
    "        labels=model.labels_\n",
    "        label_num=len(np.unique(labels))\n",
    "        if label_num>1: # 需要判断label种类，因为如果只有一个label，silhouette_score报错\n",
    "            scores.append(silhouette_score(dataset,model.labels_))\n",
    "            models.append(model)\n",
    "        else:\n",
    "            scores.append(0)\n",
    "            models.append(None)\n",
    "    optimal_id=scores.index(max(scores))\n",
    "    return eps_list[optimal_id],models[optimal_id],scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "optimal eps: 0.8000, \n",
      "candidate eps: [0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.  1.1 1.2 1.3 1.4 1.5 1.6 1.7], \n",
      "scores: [0.12865164017329436, 0.3593618148137507, 0.5134143263329637, 0.616532168834258, 0.6321600450689241, 0.6366395861050828, 0.5141678956134529, 0.5629298661962946, 0.5629298661962946, 0.5629298661962946, 0, 0, 0, 0, 0]\n"
     ]
    }
   ],
   "source": [
    "optimal_eps, optimal_model,scores=get_optimal_eps(dataset,np.linspace(0.3, 1.7, num=15))\n",
    "print('optimal eps: {:.4f}, \\ncandidate eps: {}, \\nscores: {}'.format(optimal_eps,np.linspace(0.3,1.7,15),scores))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "clusters num:  5\n"
     ]
    }
   ],
   "source": [
    "# 上述函数得到了最佳eps参数和该最佳参数下的最佳模型，我们可以从该最佳模型中得到一些属性\n",
    "labels=optimal_model.labels_\n",
    "label_num=len(np.unique(labels))\n",
    "# 但是有标记为-1的样本\n",
    "# 这些样本是没有分配集群的样本，被认为是异常点。\n",
    "if -1 in labels:\n",
    "    label_num-=1 # 需要减一个类别，减去异常点\n",
    "print('clusters num: ',label_num) # 最佳模型划分的簇群数量，\n",
    "\n",
    "\n",
    "# print(labels)\n",
    "# DBSCAN模型中可以得到核心样本的数据点坐标\n",
    "# 首先获取核心样本的坐标索引\n",
    "core_index=optimal_model.core_sample_indices_\n",
    "# print(core_index)\n",
    "mask_core=np.zeros(labels.shape,dtype=np.bool)\n",
    "mask_core[model.core_sample_indices_]=True\n",
    "# print(mask_core)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def visual_2D_dataset(dataset_X,dataset_y):\n",
    "        '''将二维数据集dataset_X和对应的类别dataset_y显示在散点图中'''\n",
    "        assert dataset_X.shape[1]==2,'only support dataset with 2 features'\n",
    "        classes=list(set(dataset_y)) \n",
    "        markers=['.',',','o','v','^','<','>','1','2','3','4','8'\n",
    "                 ,'s','p','*','h','H','+','x','D','d','|']\n",
    "        # colors=['b','c','g','k','m','w','r','y']\n",
    "        colors=['tab:blue', 'tab:orange', 'tab:green', 'tab:red', 'tab:purple', \n",
    "                'tab:brown', 'tab:pink', 'tab:gray', 'tab:olive', 'tab:cyan']\n",
    "        plt.figure()\n",
    "        for class_id in classes:\n",
    "            one_class=np.array([feature for (feature,label) in \n",
    "                       zip(dataset_X,dataset_y) if label==class_id])\n",
    "            plt.scatter(one_class[:,0],one_class[:,1],marker=markers[class_id%len(markers)],\n",
    "                        c=colors[class_id%len(colors)],label='class_'+str(class_id))\n",
    "        plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "visual_2D_dataset(dataset,labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}