Datawhale anomalyl detection study/TASK4/TASK4 teaching example.ipynb · ZhouMoqi/GitPythonShare

加入 Gitee

与超过 1200万开发者一起发现、参与优秀开源项目，私有仓库也完全免费：）

克隆/下载

TASK4 teaching example.ipynb 91.48 KB

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Import Library"
   ]
  },
  {
   "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",
    "from sklearn.neighbors import LocalOutlierFactor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.rcParams['font.sans-serif'] = ['SimHei']\n",
    "plt.rcParams['axes.unicode_minus']=False\n",
    "pd.set_option('display.max_columns', None)\n",
    "pd.set_option('display.max_rows', None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96,\n",
       "       97, 98, 99])"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_test=np.arange(100)\n",
    "x_test[80:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(61)#随机数种子，用于复现\n",
    "# 构造两个数据点集群\n",
    "X_inliers1 = 0.2 * np.random.randn(100, 2)#np.random.randn，产生正态分布的数据，是100行2列的\n",
    "X_inliers2 = 0.5 * np.random.randn(100, 2)\n",
    "X_inliers = np.r_[X_inliers1 + 2, X_inliers2 - 2]#np.r_，竖着拼接，要求列数相等\n",
    "# 构造一些离群的点\n",
    "X_outliers = np.random.uniform(low=-4, high=4, size=(20, 2))#均匀分布，-4到4，采集20行2列个点\n",
    "# 拼成训练集\n",
    "X = np.r_[X_inliers, X_outliers]\n",
    "n_outliers = len(X_outliers)\n",
    "ground_truth = np.ones(len(X), dtype=int)\n",
    "# 打标签，群内点构造离群值为1，离群点构造离群值为-1\n",
    "ground_truth[-n_outliers:] = -1#这种[num:]是可以到达最后一个数的"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAW4AAAEGCAYAAABFBX+4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3de5RU5Znv8e/T3UArCKJ0g9qgxrgywbu0hotgt8EwMZpDNBN0zORkBcecCZOTZJDMjDFjTkyy4qyczJyM4okmc0zOMsHMeJkYx9EoICAYbYIXRgjxGC/IrRtBQEHoruf88VbR1UXdu6qrdtXvs1av7l21q/Zr2/z2W89+9/uauyMiItHRUOkGiIhIYRTcIiIRo+AWEYkYBbeISMQouKUszGxUpdtQaWZm1fheEn0KbikJM/u6mS1KeuhBM7siw75/ZGbHxH9uNbO58Z83Z9j/L82s2cyWmtlZZrbIzEab2Q/NbFaG18wzs1MLaH+zmf2LmY0xs6PM7CEza8n39emOD3yx2Nen8V0zm1nC95MIU3BLqRwADsHh3vZZwGMZ9v0ccEv8548B8+M/v2tBo5kNS9q/CbgJ6AVGAle7+x7gEiBt2Mff94R4e3ab2aqkr9VmtiJl/y/G9z8P+BCwDbjazDrMrNPMpufxOyB+vEnA54H/bWafNbMlGfb7czPbYmbbzGxB/LHPmtn++GOJrxuA7wJ/b2aj822H1K6mSjdAoi3+EX444PHN4cCngP3At1M+4d/l7r8DvgmsM7P/BVwJtJnZcqANeJLQofgF8AMzawa2xN9/LNABrIj3tEe5+yvxdoxw9/fiJ42xwAjguHiv+Q/uflFSm0cBv0raPo9wIrkVaAeagauA38W3AXYAq/P8tXwTuMHdD2aqcJjZB+LHm0o4IXWZ2dL40//m7lenec33gYXAzXm2Q2qUglsGqwV4GDgV2At8AJgF/BDYmrTffwceBH7n7vvM7HxCuHcBV8VDbqO7zwIwsyYzayScFGYBFxJCdAPwFvAXQIOZdQGTgL1mdi6hx/yX8X2PjbfnJDNbBRihx/4usCd+nOOBnwM9wKXAQcK/i6OBy+NtHw1ckM8vw8xGACe7+29z7Ho58Ki7b4q/7j8InxJ6srzmfuDLKLjrnoJbBsXddwAXmNnvgDuANcAphEB+w903ApjZJwilkA8RQrwXeI4QVnPMzIEeM3uKELBNwNeBpcArwEmEcsxuYBcwHbjN3W+JlyL+p7vvBVYQeuR3Az8mBPQ17n6RmY0FHnb35LLH0cDXCKWSH8XffyTwA+B78X2GEcJ7Vx6/kvcRTi757Pda0vbrhN9bxuB29z4z22Nmze5+II9jSI1SjVsGzczmAO8H5sa/X0EoPYxNqlUboQb+JnAfoZf8nLtf6O7T3X2Gu18U/z49/vijhPJJD/DV+GtvJoTcMGBK/L0nAf8vqT1NwDhCcH8AOD3eM38COMvMusxsrZl91t3fcPf74i+dSAjPSYQTxynxr1MJpZd8jCW/gG8G3kvaPggcFf/5vyTVt7+Z8rpd8WNIHVOPWwYlXhr4e+D/EkolXwX+APweeAP4nZmdRgjaXnffDGw2sysBj4+U+CVJwUsIy7909yXAccBXgP9B6Gg8RDgBTAeWx3vRx7j7W/H2TAf+DyEYPwP8Fljn7h2J2nb859S//ZGEE85BwklmM/DJ+HPDCD3+hXn8SnYCx+ex37vxNiaMiD8GGWrccccTSkVSxxTcMlj/HXiE0CvuBf6OUN74V0LwHUeoTx9DfzAlew940t3nJh6IlzkOArj7WjO7CHiKMNpjDnCZu+83s/sJdfOlSe+3Efg48Lfx954IvGlmDxPKHWfFL4S+4u6fix9vGKE0cp+7v2tmVwHj3X1xUpuGm5l57lnZXgE+mGOfxH5Tk7ZPBp7J9oJ4O0e6+3vZ9pPap1KJDNbthLBO2E3o5R4khNM9wImE0NyT5vWNwMzk4XrARwm93oQTCTXz/yBc9Fwbf/x+woXLhxI7uvtb8ZErCVPj+49395nufqy7dwBnw+ERJssIw/ceM7N34se42cz6zOw38TYtjbcjK3c/BGyIn2yy+RXwETP7gJm9D/gI4SJvNn9K0mgYqV/qccuguPu7AGbWADTEe6Tzk3b5pZm1EUJvT3ykyBcJdeNbCIH4Y+Bv3D0Wf6+/ALaaWUP8sZeBJYRhdl8Dzo6XWm4CbgDuNLP/6u4rk47bQPgE8Mfx/f40Q/v3WbhR6GLg04Se96vx9v4n8CVCoC4j+4iPZDcD/2pml8W3rzKzfUnP3+juP7Bww9LyeFu/5u6/M7Np6d7QzE4inFw+nGcbpIYpuKVURpDmE1y8LPF+4BvxURGXE4bqzSSM6PhvwAzg5XjduTn+XkcB883sPkJwPgdc6e7bzOzTwBcIN+JsMrO1wN+Z2Vx3f8fMvkDoyb8DeHyfxJDA5PYmetz3A08DC939tfiJY7S7/8LMHiAM3ftzwonm1Vy/CHffbmbfIIxmuRO4O8N+PyKMZEl+7O4M+38O+Ly77891fKl9poUUpJySes0FvYYwfvtgoa/N8p5HKfSkVii4RUQiRhcnRUQiRsEtIhIxCm4RkYgp+6iScePG+SmnnFLuw4iI1JS1a9f2uHvaOeGLDm4zGw/8h7ufl22/U045ha6urmIPIyJSl8zstUzPDaZU8j36J8UREZEhUlRwm9klhJsbtpW2OSIikkvBwR1f4eTrwN+UvjkiIpJLMT3uvwEWu/vuTDuY2fXxOY+7uru7i2+diIgcoZjgng0siM9Bca6Z/Sh1B3e/093b3b29paXohbJFRCSNgkeVJNYEhDCBkLtfV9omiYhINoO6ASc+r7HUA4/B/u2guW1EKk53TkpuHoPHO+HBNni8I2yLSMUouCW3A93Qsxq8N3w/oAvOUsMi8OlSwS25NbfCuOlgTeF7c2ulWyRSHhH5dFndK+B4LPTumlvBLPf+Uh5mMHuZ/l9I7Uv36fKo8ZVu1RGqt8cdkTNf3bCG8Aes0JZyqYYSRUQ+XVZvjzsiZz4RKYFER61ndQjM2ctCZ2GoReTTZfX2uCNy5hOREqimC+AR+HRZvT3uiJz5RKQEEh21RI9bHbWsqje4of/MJyK1TR21glR3cItI/VBHLW/VW+MWEZG0FNwiIhGj4BYRiRgFt4hIxCi4RUQiRsEtIhIxCm4RkYhRcIuIREzRwW1mx5nZpWY2rpQNEhGR7IoKbjMbC/wKuBBYZmZayl1EZIgUe8v72cBfufvT8RA/H3i0dM0SEZFMiupxu/uT8dCeReh1rylts0REJJPB1LgNmAfsAg6lPHe9mXWZWVd3txaWFREppaKD24MFwAvAx1Oeu9Pd2929vaVF5W8RkVIq9uLkX5vZZ+KbxwK7S9ckERHJptge953An5nZCqAReKx0TRIRkWyKGlXi7ruAS0vcFhERyYPunBQRiRgFt4hIxCi4RaQ6eQz2bwf3Srek6ii4RaQ45QxWj8HjnfBgGzzeEbblMAW3iBSu3MF6oBt6VoP3hu8HdCNfMgW3iBSu3MHa3ArjpoM1he/NraV9/4grdpIpEalniWDtWV26YPVYOAE0t4IZzF42cFsOU3CLSOFKHayJ0kviRDB7GVgDHDW+NO2tMSqViEhxEsFait6watoFUXCLSOWppl0QlUpEpGxiMejuhtYs1ZSwj9H64WXYe93EhrfSvcOyvqbeqcctImURi0FnJ7S1wbRp8Pzz4bFM+3R0NtA7bDydl1jY7jhyfwkU3CJSsFgMtue496a7G1avht5e+M1v4NxzYezYsJ1un9WrYePGgdtahyU9BbeIFGRAL7kjhGy6EG9thSlTBj62Z08I5+R9pk+HpqbwffLkgdutGUrdsRhs3QrbttXnHfEKbhEpSGoveeZM0pY20gXqqFHwR3/UH/RmsGwZbN4My5eHfZYsgTfeCNvpatyxWDjWiSeGr3osqSi4RaQgyb3kCy6AZ58dWNpIlFG2bYNnnhn42smTw2tOOCG8R18fNDTA+PEhyDs7YdIkmDcvc086ceKAsE89llQU3CJSkORe8qpVMGNGCPFp00IQd3SEHviVVx4Zvs88A889Fx5/+mn40IfC+2zZAv/5n/DUU9nr27FYeO306f1tyVZSqVUKbhEpWKKX3NAQQvzVV+HQITj5ZFi5MoRvVxdMnRr2GTUq/fusXQsTJ8JJJ8HZZ/c/PnIkjBs3cN/e3nCSaGsL24nAz1RSqWUKbhEZtE9+MvSgEyNGGhpCyK5aFQJ25cr83qevL3zftw96evofj8Vg1qxwjL6+0CNvaKi/wE4odpX3MWb2iJk9ZmYPmNnwUjdMRKpfLAYvvXRkLfucc+CnP4X16+FTn4L2dhg9Ov/3PfdcaGnp3+7uDrX0hPZ2uPrq9BdF60GxPe5rge+7+0eAbcAfl65JIhIFiWGB550HRx898Ll16+CUU0IAr1oVesl79uT3vg0N4fUdHfDCC+E4yRdEp02DBx6o7/HeRQW3uy9291/HN1uAHaVrkoiU3CBXq0m94SbR006E5759+ZUtGlISxwwaG/tr4Yn3jsVCeeWcc+C440LwJy6IPvUUTJgQgryxMfS+k3vn9WBQNW4zmwaMdfenUx6/3sy6zKyru95OhSLVpojVapKDOt0NN52doTfd3Jx0mAznhKOPDgGbeN8BTXM444wQ0hddFPYbPXpgwL/9drhpJ3FB1Cx8PfEEXHhhuAja2Vlf5ZKig9vMjgP+Cfhc6nPufqe7t7t7e0u9nQpFqk2BU6amBvX69elvS+/rg3feObJMctZZA7f/7d/C2O1MXnoplDqWLIHXXw+97ERPHGDMmDD+O9XOnUeOIa8XxV6cHA78C/C37v5aaZskIiVV4JSpyXdGrlwZatgjR6a/LX3mTNi1K9zaniiVJPfCAebMCSF//vn9j40aFV6TeM9588KNN5/8ZP9JAUJ55K23jiyxQBguOHJk+Dnd8MFaZl5EzcvM/gL4DvB8/KE73P3edPu2t7d7V1dX8S0UkcFLXRYsg1gMduwIQZoIbwi93+eeC2UNsyOna926NYzH7usLYTx5criwmElTU+hdNzSEcsnEieFYTU2hZv3ss6GX/tRT6UMbQinnpJPCMRsb4c03QymlVpjZWndvT/dcsRcn73D3se7eEf9KG9oiUiXyWK0mUSKZODGE6WuvhR51U1MYk50ptCFcLJwxI2z39mYP7cQFRQjv0dISthsbw4iRFSvyq123tvbftTljRn3dPamFFEQEGFgiWbMmBOny5QNDOhHuq1eHEseyZf03wvzsZ6Hcke1D/Ic+1H+7+4knhhNDYjuhp+fI2nW6nnTi1vtcCzXUIt05KSLAkVOstrYOHMkBR84MmLggGIvBNddkH9lx4YVw550Db6RZubJ/wigIJ4zE/CO5pnaFI9tXLxTcIgIcOcVqujBMDfdx40KteceOELrQfwPN1Kmh1z5rVqg/NzeHC5KJC4oJiREnicAePz53O+pdURcnC6GLkyK1JVHjHjcOLrmkv2ziHsJ7+vQQuH19Yejg5Mlh/7a2/guQZ58dljJLlFt27AgBXY+950yyXZxUjVtECpIoT2zfPrBskhgl0toaQvzDH+4P9aVLw/fk7Z6e/tr0CSf0j2ipt3p1MVQqEZGipJZNJkzo7zGn1sJ7egaWPxobB/auU2/6SVcrz2edy3qh4BaRomSriedzoTNZatC/9NLAgM4n2OuJgltEipYpjPO50JksOehHjgzzoCQHdKbRLPVKwS0iJZfuJp1sEkG/bl2YaTCxWEIioNP14OuZgltESqrYskZDQ7g7M3E3ZHJAF9qDr3UaVSIiJZWurJHvHCLZ7oZMlGVEPW4RKbHBljXq9W7IQqjHLSIlVc9ziAyQ54yMxVCPW0RKru57zUWsOlQIBbeISKkVuOpQoRTcQ2GQC7WKSMQUuOpQoVTjLrfER6ae1eF/4OxlYVJ7EaldZuHfumrcEVXmj0wiUqXyWHWoWINZ5X28ma0sZWNqUpk/MolI/SmqVGJmY4GfACNz7Vv3yvyRSUTqT7E97j5gHrCnhG2pXWX8yCQi9aeoHre77wEwBZGIyJAry8VJM7vezLrMrKu73udfFBEpsbIEt7vf6e7t7t7e0tJSjkOIiNQtDQcUEYmYQQW3u3eUqB0iIpIn9bhFRCJGwS0iEjEKbhGRiFFwi4gUqsIzfiq4RUQKUeZFEvKh4BYRKUQVzPip4BYRKUQVzPhZPwsplHHhzrKJYptFal0VzPhZHz3uKqhJFSyKbRapFxWe8bM+grsKalIFi2KbRWRI1EdwV0FNqmBRbLOIDIn6qHFXQU2qYFFss4gMifoIbuivSUVJFNssImVXH6USEZEaouAWEYkYBbeISMQouEVEIkbBLSISMQpuEZGIKTq4zezHZrbGzG4qZYNERCS7ooLbzK4EGt19GvA+Mzu9tM0SEZFMiu1xdwC/iP/8GHBRSVojIiI5FRvcI4E34z+/BQy4vc/MrjezLjPr6u7W5EgiIqVUbHDvA46K/zwq9X3c/U53b3f39paWlsG0T0REUhQb3GvpL4+cA7xaktaIiEhOxU4y9SCw0sxOBD4KTC1dk0REJJuietzuvodwgfJpoNPd3y5lo0REJLOip3V19130jywREZEhojsnRUQiRsEtIhIxCm4RkYhRcIuIRIyCW0QkYhTcIiIRo+AWEYkYBbeISMQouEWk/DwG+7eDe6VbUhMU3BEUi8F2/RuQqPAYPN4JD7bB4x1hWwZFwR0xsRh0dkJbG3R0hG2RqnagG3pWg/eG7wc0R/9gKbgjprsbVq+G3t7wXetUSNVrboVx08Gawvfm1kq3KPKKnmRKKqO1FaZPD6E9fXrYFqlqZjB7WehpN7eGbRkUBXfEmMGyZaGn3ap/AxIV1gBHjc+9n+RFwR1BDQ0wXv8GROqWatwiIhGj4BYRiZiCg9vMxpvZynI0phpojLSIVLuCgtvMxgI/AUaWpzmVpTHSIhIFhfa4+4B5wJ4ytKXiNEZapAbV4O32WYPbzH5oZssTX8CXa3lF98QY6aYmjZEWqQk1ert91uGA7v75Yt7UzK4HrgeYNGlSMW9RERojLVJj0t1uXwPjycsyqsTd73T3dndvb2lpKcchyiYxRroaQlsXSkUGqUZvt9cNOFUqcaE0cWv7smXhpCIiBajR2+2LigJ37yhxOySFLpSKlEjidvsaCW3QDThlM9gyR+JCaWMjtLdDJSpOKtWIVCcFdxmUYjy4GTzxBFx4ITz7LMyYAX19JW9qRhrTLlK9FNxlUKoyx86d8MwzIbCffhpmzkwfoOXoGatUI1K9FNxlUOx48NQAbm2FCy7of/7ZZ48M0HL1jDWmXaR6KbhLIDVwE+PBN2+G5cvzuyaSLoDNYNUqmDYtc4CWq2dczH+DiAwNBfcgZerxZhsPnq60kSmAGxtDeGcK0HL2jAsZ064LmSJDR8Gdp0zBlBy4Tz0FL72UPbwyBX22AM4WoNXQM9aFTJGhpeDOQ7ZgSh62N2oUnHde+vBKBP+OHSHgE0Gf6FknB/DSpWG/fHuvlb7bUxcyRYaWgjsP2YIpEbjPPQfvvJN+n+Tg/5M/CQEP4fu4cf37NTSE8dqXXBKt3qsuZIoMLQV3Fr29sH59CNdswdTQAGeckXmf5OBfsyYEPITvPT0D36vQ3mulasvJx00t17ir3i1STgruDHp7Q2CfdRYcfzz8+tfw+utw773p989Wa07tkWY7CRTSex2S2nKauYzTHTdRrnHP3iZdxBQpAXcv69eUKVO8mvX1uW/b5h6LDXz8xRfdQ7yEr+efd581y72pKXzv6yv+OJmOmbpvb2/2/bZtC+2B8H3btsLalFOsz/2xWe4/awrfY305j5vtub6+wf0OReoJ0OUZcrWue9zZeqyTJ8OYMeHnY44Jve/BXIBLvoCY62JivrXusteW081lnOO42Z5LWwaqwdVJRMqtroM7Wz25oSGM7JgyBfbvh2uu6b8RZtq0/r54PoopD+RT6y77UMAMcxlnO24hJaPWltpcnUSk3Oo6uHP1WHftguef7w/Pn/8c1q4NYTRxYn9PODWYk7cTvfqTTgrHSJ4oKlug59ubLnYoYF4nk8RcxnM3w+zlAw6S7biZnjsi1N9L36MXkezqMrgToQXZe6zJ4TltWuh1n38+rFzZH+bbtw8st/T2Dtzevj2M106dKCrXhcVy9aZ7e+GFF8Ix87qoWeK5jAeEeo2uTiJSbuZlri22t7d7V1dXWY+RSSx25PqRha4sk3iPvr4QdMm/rjFjwp2SJ58cArGpCdatCzfhJLbfeAM+8YkQ2hAe27w5/NzW1r/f5s0h0Ir9b0v335oqMVLm7aTlnos5dkl5rOZWJxEpBTNb6+7t6Z6r2R53by9cdNGRvcpMteNMpYNED7Gh4cjn9u0LjyeXNCZPHrg9fvzAiaIuuCBceBzMQgmpvfXUXn6mHvTGjQNDu7GxCm6YqcHVSUTKLtNwk1J9VWI4YF+f+9Sp/UP5koelxWIDh6QlhuflGqYWi7nPnOne0OA+enT/vr297lu2uG/d2j9sL91wv0OHQpsaG/uPceiQ+7Rp6Y+bbchg6pC7F1/Mb1hgX5/7mDFhv9GjQ7szDTUUkcoiy3DAgkIYGAM8AjwGPAAMz/WaSgT3tm0hIBPBPW3awIBKDcV8x0Onjq/u7T0y8DMFbrpjZDpurhNJ6skndf/UYye36dChEPQaQy1S3UoZ3F8ALo3/fAfw8VyvqURwJ4KtsTH0cnOFVLpeeLJ8w3jLlsyBm+4YmY6brkedq02Z2qibXkSiqWTB7QND/F+Bqbn2q9Sdk/nefZi6fyHBlxq8W7dm77mnO0a6x3p7+0soY8YMLK8Uqux3V4pIWWQL7qwXJ83sh2a2POnr7+KPTwPGuvvTpa+6l0ahM+1lGnucz8yAiSF748fnnowq9RjJc3xs3x5Gr1xySVhr8pxzYO/e8Fi+d2umW/5MM/eJ1JhMiZ7pCzgO6AJOzrLP9fF9uiZNmjRUJ6gjbNnSX+vOVHLIJVcZJVWueUgyvSZxjKlTB/aQE9v5Hjvdp4Ni2lSMoTqOSD0gS4+7oHHcZjaccHHyu+7+63xeU6lx3IkhcytXht7n6NFhGtUZM3KP3U73Xokx0u65x0tnem2m/bdv7x/T3dgIF14YFgaePr1/UQWz3HdIJr/PUI/PLnR8vIhkV8px3POB84GvxUsn8wbdujJJlDjcQxju25e+5JDPrd/5TlmaKt9pV5PLGTNmhJNN8hSyV1898Bb7TCpZFtEqOCJDKFNXvFRflbo4mVrimDnzyJJDoSMuCr3Ql2uK00yjQrKVTnKVeypVrii0pCQi2VGqUkkxquWW9+QSR+Lnvr6Bt6vnKi24h15vohyQaw6RTPvnKitkKp2MHBk+ORRT7hkK+ZSFRCQ/2UolVRvc5QqBRGg+9VQIwj17wvvPnJnfZE6Ftivd/rlq0amBv3QpbNgA554bTjYVn19ERMoucnOVlHNJrkQttq8vhDaEnuu992YP4mLm1M4U8rlq0anDDBsbw5qWM2ZoWJ+IVGlwl/NCV/LkTmPG9F8QzNZ7TZ5T+/jj8zuhZJrkCvKbsjV1zHe612j9RpH6VJXBXc7REYkAfPNN2Lkz93zXsViYujUxp/bbb+c+ocRiofSyZk3mfYtZACH5NcmfSi6+GLZuVYCL1IuqDO5SLSKQa6rWxsbs4ZkIx/POg1GjwusSvfRsJ5Tu7nAxMeGCC0pf2kj+VLJyJUyaVMaV3kWkqlRlcEPxS3IllKJOnhyO77wTljHLp5fe2hrKL42NMHVqmI+71KMskks+ZsWXlVRuEYmeqg3uwSp0wYR0Uks2Z5yRu5cOA8sxq1eHk1CpAzL5U8nMmcWVlcp5EVhEyqdmgztdnbzQoBpMySZTPbqUAdnQABMmFN9G3e0oEk01G9xm8MQTYQ3IZcvCdjFBNdiSDRx53JdeKm1potg2auZAkWiq2eCOxeDDHw4XFjs7w3Y5gypbKST5uCNHhhtpqqE0Ua6V5EWkvGo2uJN7uU891T/DXnJQJebAHmzvN1cpJHHcdesyT3ZVKYl5y3fs0AVKkahoqnQDyqW1NaysvnJlCNJ58/rn9xg/vrTTkKYrwaTe0NPQ0H/3Y+KY1VCa0HSsUkmHDh1i8+bNHDhwoNJNqZjm5mba2toYNmxY3q+p2rlKSmHr1jC+Od2cIIOZuzr1VvZCJp+q9ERMqcev5BzeVcVjcKAbmjVD1lD6wx/+wDHHHMPxxx+P1eHv3d3ZuXMne/fu5dRTTx3wXOTmKimVCRMy17SLrXenK4sUUisuxcXO5LakK/Vkezy17bpASQjtxzvhwTZ4vCNsy5A4cOBA3YY2gJlx/PHHF/yJo6aDO1ugFnthLtPIlFyBXOpx3Jnq6tnq7enanvx7SKy2U3e17v3boXsleG/4vn97pVtUV6ohtHfv3p1YdpEdO3Yc8XxfX9+A7Z6eHl5//XUAHnroIQZTuSjmv7+mgxuyB2oxvd9ieqipYdrbO/gQz3QCyTbkMVPbC11YufYYkPif4fFtqSfz589nxYoV7N+/n6uvvpq9e/cOeL6zs5PTTz+dmTNncuONN7Jv3z6+9KUv8fvf/57bbrstbfhee+21PPPMM2Vpb81enCyXRA+1uxvGjQs91Fz16tQRLrNm9a8pWezFwEQIp17ozPR4attT25zPBdaaddR4GDcTelbDuOlhW+rGPffcw9KlS9m5cyd9fX309PRwxRVXHH5+2bJlLF68mCVLljB37lweeugh+vr6WLhwIT/60Y+47LLL2L17N8cee+yA9x02bBhf/OIXOeaYYwA4ePAgK1asKEmbCw5uMzsOmAKsc/eekrSiSmW6kJjooeY7GiM5TNvbQ2gPNiDdYcmSIxcRzhbOibanO162wK95ZnDpcl2crENdXV1897vf5YwzzuDyyy9n79693HfffVx++eX09vYyd+5cHnjgAb71rW/x1ltv8cgjj7Bjxw6OPvpoNm3axOrVq+ns7OT1118/IrgBbr/9dtrb015fHJSCgtvMxgK/Ah4Gvm9ml7h7FYxGLr1cw+QK6aEmh2lq4IkU/IIAAAkFSURBVIfVJAvLinRtSz2xFHoyyBX4Nc8a1NOuQ2eeeSZLlizhiiuuYNWqVbz33nvs3LmTVatWcejQISZOnMjcuXPZtGkTr7zyCmeddRbd3d1cd911bN++nX379nHTTTfR0tICwBtvvMHZZ5/NOeecA8ANN9wAwMaNG3n00UcPPz5Yhfa4zwb+yt2fjof4+cCjJWlJlckVzIX2UJPDdNmyUOOeNy+s3l5oyaRcZY1iAl9kqJVySG1zczMjRoxg4sSJzJ49m71797JlyxZmz54NhPIGwIoVK7jqqqs46aSTmDNnDkuXLuWFF17g5Zdf5u6772b+/PmMHz+eESNGMGXKFB5//PEBx/nsZz/L8OHDB9fYJAUFt7s/CWBms4ALgW+WrCVVJlcwD6aH2tAQvlIXWsg3NOu6rCF1rRw3jB111FF84hOf4Mwzz2TlypVceeWVnHnmmUCoU2/ZsoXf/OY3vPvuuzQ0NPDRj36URYsW0dnZyRtvvMGNN954+L0OHjzI2rVr6ejoGHCMjRs38pWvfGVwDU2S9QYcM/sh8IGkh5YCtwC3AW3A1e6+P83rrgeuB5g0adKU1157rWQNHkrlvFmm0BXjh7JtIkNlw4YNfPCDH8x7/1LfMLZ+/Xq+8IUv0NDQwJtvvsnu3btpbGxkxIgRnHrqqRw6dIjLLruM7du3c+WVVzJhwgTuuusuzIy+vj6WLl3Kxz72MS6++GLmzJnD/fffz/nnn8+7777L5MmT2bBhAyeccAJNTU1s2LCBCy64IO/fQ1lWeTezW4D17n5vtv0qeedktVP4Sr0rNLgH2+FJ57XXXmPJkiU899xzdHR0YGasX7+e/fv3c8011zB16lRuv/127rnnHoYPH86sWbP43ve+R09PDzfccAN33303EG4mmjJlCmvWrOG6665j0aJFvPzyy/z2t7/lpptuYtasWSxfvpyxY8fm9Xso2Z2TZvbXZvaZ+OaxwO5CXi8DlfIuSpF6UOoZLV955RXmz5/Paaedxs9+9jOGDx/OwYMH+cEPfsDChQt5+OGH2bRpE729vfzjP/4jP/3pTw/XvQ8cOMChQ4cOv9eGDRtYsGABo0ePZsGCBbz99tvMmzePUaNGMWbMGBYtWsT69esH1+C4gnrc8QuSvwBGAOuBBZ7jDdTjFpFMCu1x16pCe9yFXpzcBVxafPNERGSwav6WdxGRWqPgFhGJGAW3iEjEKLhFRPJQzqldC6XgFpG6FovFiCXNY3zrrbdy3333Hd4+dOgQfX19VTW1q6Z1FZG69uKLL7Jw4UKamprYsWMHLS0tLF26lIULF/L+978fd+db3/pWVU3tquAWkbp2zjnnMH/+fMyM4cOHc/PNN/Pee+9x2WWXsWjRIm677Ta2bt3KggULqmZqVwV3CejWdZEhVIaFnfv6+nj11Vd58cUXufHGGznttNO46667eOSRR3jnnXeYM2dOVU3tquAepHLMViYiGSQWdk6sVjR7WZhLfZCmTp1Kb28vbW1t/OQnP+Hiiy9mxowZAHz6058GqmtqVwX3IA31kl/q3UtdO9AdQtt7w/cD3SVZAOP6668nFovh7qxbt479+/cfvtjY29vLP/zDP1TV1K7qG3osrOpd5FCeYhYPLla2FdxF6kJza+hpW1P43lyaf3BHH300TzzxBKeffjoPP/wwTz75JNOnT+ff//3fOfPMM9m4cSPXXnst3/jGN1i8eDHbtm3jhhtu4Nvf/jbr1q3jxhtv5NFHw5oyzzzzDOvWrWPx4sUsX76cO+64gwcffJCXX3758ARVg1XfPe4SfOwayiW/6npBXxEI/8BmLyvL+qBPPvkkjz76KM3Nzfzyl79k7969/PznPwfgkksuYdu2bXz5y18+YmrXXbt28Z3vfAcIMwZ+/etfZ82aNXz1q189PLXrP//zP3PTTTdx3XXXZZzatRD1Hdwl+tg1VEt+aeUbEcqyPuiBAwfo6Ojg9ddfHzAe+9Zbb2X16tUcd9xxh6d2nTBhAosXLz78unymdt20adOAqV1nzpw5qPYWvZBCvqp6Wld3eLwjqce9vOoLx6pxSy2p9mld3T3tjTWlVtZpXWtOGT92lYsW9BUZOkMR2sXQxcnEx64q/R8kUuuGco6PalTMf7+CW0Qqprm5mZ07d9ZteLs7O3fupLm5uaDX1XepREQqqq2tjc2bN9Pd3V3pplRMc3MzbW1tBb1GwS0iFTNs2DBOPfXUSjcjclQqERGJGAW3iEjEKLhFRCKm7DfgmFk38FqZ3n4c0FOm9x4Kan9lRbn9UW47qP35ONndW9I9UfbgLicz68p0Z1EUqP2VFeX2R7ntoPYPlkolIiIRo+AWEYmYqAf3nZVuwCCp/ZUV5fZHue2g9g9KpGvcIiL1KOo9bszsODO71MzGVbotIlL7qiFzIh3cZjYW+BVwIbDMzNIOnalGZjbGzB4xs8fM7AEzK80qokPIzMab2cpKt6MYZvZjM1tjZjdVui3FiOrvPup/99WSOZEObuBs4K/c/dvAo8D5FW5PIa4Fvu/uHwG2AX9c4fYUJP4H/BNgZKXbUigzuxJodPdpwPvM7PRKt6kQUf7dE/G/e6okcyId3O7+pLs/bWazCGfANZVuU77cfbG7/zq+2QLsqGR7itAHzAP2VLohRegAfhH/+THgoso1pSiR/d1H/e++WjInUrMDmtkPgQ8kPbQUuIXwR7wLOJTuddUgXdvd/ZtmNg0Y6+5PV6hpecnS/ko1aTBGAm/Gf36LaH1Sw933QPWuzpKPqPzdp2PhF1/RzIlUcLv75zM8tcDMbgE+Dtw7hE3KW7q2m9lxwD8BVw19iwqT5XcfRfuAo+I/jyLinzyjJkp/9+l4GIpX0cyJ9B+smf21mX0mvnkssLuS7SlE/KLMvwB/6+7lmstF0ltLf3nkHODVyjWlvkT9775aMifSwU0YBP9nZrYCaCTUK6NiPuEj+tfMbLmZzat0g+rIg4S/m+8DnwIernB76knU/+6rInN0A47UpfjIjEuBFe6+rdLtESmEgltEJGKiXioREak7Cm4RkYhRcIuIRIyCW0QkYhTcIiIRo+AWEYmY/w+nGYYMdZ844AAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.title('构造数据集 (LOF)')\n",
    "plt.scatter(X[:-n_outliers, 0], X[:-n_outliers, 1], color='b', s=5, label='集群点')\n",
    "plt.scatter(X[-n_outliers:, 0], X[-n_outliers:, 1], color='orange', s=5,label='离群点')\n",
    "#plt.axis('auto')\n",
    "#plt.xlim((-5, 5))\n",
    "#plt.ylim((-5, 5))\n",
    "legend = plt.legend(loc='best')\n",
    "#legend.legendHandles[0]._sizes = [10]\n",
    "#legend.legendHandles[1]._sizes = [20]\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 训练模型（找出每个数据的实际离群值）\n",
    "clf = LocalOutlierFactor(n_neighbors=20, contamination=0.1)\n",
    "# 对单个数据集进行无监督检测时，以1和-1分别表示非离群点与离群点\n",
    "y_pred = clf.fit_predict(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1, -1,  1,\n",
       "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1, -1,  1,  1,  1,\n",
       "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
       "       -1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
       "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
       "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
       "        1,  1,  1, -1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
       "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
       "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
       "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
       "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
       "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1, -1, -1, -1, -1,\n",
       "       -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,  1,  1])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 找出构造离群值与实际离群值不同的点\n",
    "n_errors = y_pred != ground_truth"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False, False, False, False,  True, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False,  True, False, False, False, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False, False, False, False,  True, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False, False, False, False,  True, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False, False, False, False, False, False, False, False,\n",
       "       False, False,  True,  True])"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_errors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(220,)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_errors.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(220, 2)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_pred = np.c_[X,n_errors]\n",
    "X_scores = clf.negative_outlier_factor_\n",
    "# 实际离群值有正有负，转化为正数并保留其差异性（不是直接取绝对值）"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "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>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>220.0</td>\n",
       "      <td>-1.441653</td>\n",
       "      <td>1.421518</td>\n",
       "      <td>-12.407542</td>\n",
       "      <td>-1.248249</td>\n",
       "      <td>-1.067362</td>\n",
       "      <td>-0.999912</td>\n",
       "      <td>-0.944325</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   count      mean       std        min       25%       50%       75%  \\\n",
       "0  220.0 -1.441653  1.421518 -12.407542 -1.248249 -1.067362 -0.999912   \n",
       "\n",
       "        max  \n",
       "0 -0.944325  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(X_scores).describe().T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "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>x</th>\n",
       "      <th>y</th>\n",
       "      <th>pred</th>\n",
       "      <th>scores</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.913701</td>\n",
       "      <td>2.087875</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000494</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.999748</td>\n",
       "      <td>2.212225</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.005255</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.040673</td>\n",
       "      <td>2.133115</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.001521</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.791277</td>\n",
       "      <td>1.743218</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.015652</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.991693</td>\n",
       "      <td>1.770405</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.010113</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          x         y  pred    scores\n",
       "0  1.913701  2.087875   0.0  0.000494\n",
       "1  1.999748  2.212225   0.0  0.005255\n",
       "2  2.040673  2.133115   0.0  0.001521\n",
       "3  1.791277  1.743218   0.0  0.015652\n",
       "4  1.991693  1.770405   0.0  0.010113"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_scores_nor = (X_scores.max() - X_scores) / (X_scores.max() - X_scores.min())\n",
    "X_pred = np.c_[X_pred,X_scores_nor]\n",
    "X_pred = pd.DataFrame(X_pred,columns=['x','y','pred','scores'])\n",
    "X_pred_same = X_pred[X_pred['pred'] == False]\n",
    "X_pred_different = X_pred[X_pred['pred'] == True]\n",
    "# 直观地看一看数据\n",
    "X_pred.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig=plt.figure(figsize=(15,8))\n",
    "plt.title('局部离群因子检测 (LOF)')\n",
    "plt.scatter(X[:-n_outliers, 0], X[:-n_outliers, 1], color='b', s=5, label='集群点')\n",
    "plt.scatter(X[-n_outliers:, 0], X[-n_outliers:, 1], color='orange', s=5,label='离群点')\n",
    "# 以标准化之后的局部离群值为半径画圆，以圆的大小直观表示出每个数据点的离群程度\n",
    "plt.scatter(X_pred_same.values[:,0]\n",
    "            , X_pred_same.values[:, 1]\n",
    "            ,s=1000 * X_pred_same.values[:, 3]\n",
    "            , edgecolors='g'\n",
    "            ,facecolor=\"none\"\n",
    "            , label='标签一致'\n",
    "           )\n",
    "#,facecolors='none'\n",
    "            \n",
    "plt.scatter(X_pred_different.values[:,0]\n",
    "            , X_pred_different.values[:, 1]\n",
    "            ,s=1000 * X_pred_different.values[:, 3]\n",
    "            , edgecolors='r'\n",
    "            ,facecolor=\"none\"\n",
    "            , label='标签不同'\n",
    "           )       \n",
    "#,label=\"标签不同\"\n",
    "           \n",
    "#,facecolors='none'\n",
    "            \n",
    "#plt.axis('tight')\n",
    "#plt.xlim((-5, 5))\n",
    "#plt.ylim((-5, 5))\n",
    "legend = plt.legend(loc='lower right')\n",
    "#legend.legendHandles[0]._sizes = [10]\n",
    "#legend.legendHandles[1]._sizes = [20]\n",
    "plt.show()"
   ]
  },
  {
   "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": "3.8.3"
  },
  "latex_envs": {
   "LaTeX_envs_menu_present": true,
   "autoclose": false,
   "autocomplete": true,
   "bibliofile": "biblio.bib",
   "cite_by": "apalike",
   "current_citInitial": 1,
   "eqLabelWithNumbers": true,
   "eqNumInitial": 1,
   "hotkeys": {
    "equation": "Ctrl-E",
    "itemize": "Ctrl-I"
   },
   "labels_anchors": false,
   "latex_user_defs": false,
   "report_style_numbering": false,
   "user_envs_cfg": false
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}

一键复制编辑原始数据按行查看历史

提交于 2021-01-21 20:39 . pyod lof toy example 的jupyter文件

Python

https://gitee.com/zhou-moqi/GitPythonShare.git

git@gitee.com:zhou-moqi/GitPythonShare.git

zhou-moqi

GitPythonShare

master

ZhouMoqi / GitPythonShare

简介

发行版

贡献者

近期动态

ZhouMoqi / GitPythonShare .gitee-modal { width: 500px !important; }

简介

发行版

贡献者

近期动态

搜索帮助

ZhouMoqi / GitPythonShare