c7_衡量回归算法的标准.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 线性回归算法的评测"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### $\\frac{1}{m}\\sum_{i=1}^m(y_{test}^{(i)} - \\hat{y}_{test}^{(i)})^2$\n",
    "\n",
    "**均方误差 MSE (Mean Squared Error)**, 单位量纲是平方;\n",
    "\n",
    "解决方法是将上式开方, 也就是 **均方差误差 RMSE(Root Mean Squared Error)**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### $\\frac{1}{m}\\sum_{i=1}^m \\left | y_{test}^{(i)} - \\hat{y}_{test}^{(i)} \\right |$\n",
    "\n",
    "**平均绝对误差 MAE(Mean Absolute Error)**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 波士顿房产数据"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn import datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Boston House Prices dataset\n",
      "===========================\n",
      "\n",
      "Notes\n",
      "------\n",
      "Data Set Characteristics:  \n",
      "\n",
      "    :Number of Instances: 506 \n",
      "\n",
      "    :Number of Attributes: 13 numeric/categorical predictive\n",
      "    \n",
      "    :Median Value (attribute 14) is usually the target\n",
      "\n",
      "    :Attribute Information (in order):\n",
      "        - CRIM     per capita crime rate by town\n",
      "        - ZN       proportion of residential land zoned for lots over 25,000 sq.ft.\n",
      "        - INDUS    proportion of non-retail business acres per town\n",
      "        - CHAS     Charles River dummy variable (= 1 if tract bounds river; 0 otherwise)\n",
      "        - NOX      nitric oxides concentration (parts per 10 million)\n",
      "        - RM       average number of rooms per dwelling\n",
      "        - AGE      proportion of owner-occupied units built prior to 1940\n",
      "        - DIS      weighted distances to five Boston employment centres\n",
      "        - RAD      index of accessibility to radial highways\n",
      "        - TAX      full-value property-tax rate per $10,000\n",
      "        - PTRATIO  pupil-teacher ratio by town\n",
      "        - B        1000(Bk - 0.63)^2 where Bk is the proportion of blacks by town\n",
      "        - LSTAT    % lower status of the population\n",
      "        - MEDV     Median value of owner-occupied homes in $1000's\n",
      "\n",
      "    :Missing Attribute Values: None\n",
      "\n",
      "    :Creator: Harrison, D. and Rubinfeld, D.L.\n",
      "\n",
      "This is a copy of UCI ML housing dataset.\n",
      "http://archive.ics.uci.edu/ml/datasets/Housing\n",
      "\n",
      "\n",
      "This dataset was taken from the StatLib library which is maintained at Carnegie Mellon University.\n",
      "\n",
      "The Boston house-price data of Harrison, D. and Rubinfeld, D.L. 'Hedonic\n",
      "prices and the demand for clean air', J. Environ. Economics & Management,\n",
      "vol.5, 81-102, 1978.   Used in Belsley, Kuh & Welsch, 'Regression diagnostics\n",
      "...', Wiley, 1980.   N.B. Various transformations are used in the table on\n",
      "pages 244-261 of the latter.\n",
      "\n",
      "The Boston house-price data has been used in many machine learning papers that address regression\n",
      "problems.   \n",
      "     \n",
      "**References**\n",
      "\n",
      "   - Belsley, Kuh & Welsch, 'Regression diagnostics: Identifying Influential Data and Sources of Collinearity', Wiley, 1980. 244-261.\n",
      "   - Quinlan,R. (1993). Combining Instance-Based and Model-Based Learning. In Proceedings on the Tenth International Conference of Machine Learning, 236-243, University of Massachusetts, Amherst. Morgan Kaufmann.\n",
      "   - many more! (see http://archive.ics.uci.edu/ml/datasets/Housing)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "boston = datasets.load_boston()\n",
    "print(boston.DESCR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['CRIM', 'ZN', 'INDUS', 'CHAS', 'NOX', 'RM', 'AGE', 'DIS', 'RAD',\n",
       "       'TAX', 'PTRATIO', 'B', 'LSTAT'], dtype='<U7')"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "boston.feature_names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(506,)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x = boston.data[:, 5] # 只使用房间数量 RM 作为数据\n",
    "x.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(506,)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y = boston.target\n",
    "y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAD8CAYAAABn919SAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJztnX+QHOWZ37/Pjho0i8+MwGsHBmThS0q643RizZYhpypXJF9QbGy8Eb9M4SuScoX84UqMTe1ZTjlGXJGgi3KB++PKV5SdHCl8WCDsNZg6y1eGS+qoAkdi2eN0oLrYgOQRMfKhwUYaxOzukz9mejQz22/32z39e76fKtWuZqa7n+7e+b5vP8/zPo+oKgghhBSfiawNIIQQEg8UdEIIKQkUdEIIKQkUdEIIKQkUdEIIKQkUdEIIKQkUdEIIKQkUdEIIKQkUdEIIKQlr0jzY+973Pt2wYUOahySEkMJz6NChX6jqVNDnUhX0DRs24ODBg2kekhBCCo+IvGbzObpcCCGkJFDQCSGkJFDQCSGkJFDQCSGkJFDQCSGkJFhluYjIqwB+BWAZwJKqzojIBQD2AdgA4FUAN6nqyWTMJKMyv9DA3gNHcLzZwsW1KuZ2bMTsdD1rs0Yi7nNK8xrFeay0721cx3P302i2UBHBsmrv56QzgdbSClSBighuuepS3DO7eeTjD2+7bdMUnn75RKz3of986il/18SmY1FX0GdU9Rd9r/0XAG+q6h4R2QVgnap+2W8/MzMzyrTF9JlfaOAr33kRrfZy77WqU8G9OzcXVtTjPqc0r1Gcx0r73sZ1PK/9BPHZq9dj5oMXRD6+zTHjvA+j7rMfETmkqjNBnxvF5fJpAA92f38QwOwI+yIJsvfAkVV/aK32MvYeOJKRRaMT9zmleY3iPFba9zau43ntJ4iHnzs20vFtjhnnfRh1n1GwFXQF8EMROSQit3df+4Cqvg4A3Z/v99pQRG4XkYMicvDEiROjW0xCc7zZCvV6EYj7nNK8RnEeK+17G9fxoti3rDrS8W2PGed9GGWfUbAV9K2q+mEAHwfweRH5qO0BVPUBVZ1R1ZmpqcCVqyQBLq5VQ71eBOI+pzSvUZzHSvvexnW8KPZVREY6vu0x47wPo+wzClaCrqrHuz/fAPBdAB8B8HMRuQgAuj/fSMpIMhpzOzai6lQGXqs6Fczt2JiRRaMT9zmleY3iPFba9zau43ntJ4hbrrp0pOPbHDPO+zDqPqMQmOUiIucBmFDVX3V/vwbAHwB4HMBtAPZ0f34vSUNJdNxgTJmyXOI+pzSvUZzHSvvexnW8/v2EzXKJenwv290sF9eGfn93mHMynU/uslxE5EPozMqBzgDw56r6n0TkQgCPAFgP4CiAG1X1Tb99McuFEOJFlmm1RcgCs81yCZyhq+pPAWzxeP0fAHwsmnmEENJhWFAbzRa+8p0XAYSbJfvt32+w8MucyYug28KVooSQTEky9dIdLBrNFhRnB4v5hUbvM2XKAqOgE0IyJUlBtRksypQFRkEnhGRKkoJqM1iUKQuMgk4IyZQkBdVmsJidruPenZtRr1UhAOq1aq4ComFItQUdIYQMk2Tq5dyOjZ4ZLMODxex0vZACPgwFnRCSOUkJahnXYPhBQSeEJEYeyjaXZfZtAwWdEJIISeeXk9UwKEoISYQylm3OOxR0QkjszC800CjRgp2iQEEnhMSK62oxUcQFO0WBPnRCSKz4de8ZThnMQ9C0TFDQCSGx4udS6V+ww6Bp/NDlQgiJFZNLpV6rWlc5JNGgoBNCYsV2KX+aVQ7nFxrYuucpXLbrSWzd89RAtcUyQZcLISRWbFdnXlyrembCxB00HSfXDgWdEBI7NqszbeusjEqZGlgEQUEnhGRCWnVWytTAIggKOiHEiiRSDNOos5KWaycPMChKCAnEppVb1P0mHawsUwOLICjohJBATH7o3Y8fjrzPpAaJYcrUwCIIulwIIYGY/M3NVhvzC41I4phmsHJcSuhyhk4ICcTP33zHvhciuUvGKViZFhR0QkggQf7mKO6SJJtDDzMuC4so6ISQQGan61g36fh+Juyy/bSClWn56vMABZ0QYsVdn7p8lQAP02i2rGfCaQUrx6lmDIOihBAr+hcCmZpXCNB7z2aJfRrBynHy1XOGTgixZna6jmd2bcf9N1+xarYuAHTo83mYCafpq88aCjohJDRe7pJhMXfJeibs5at3KoJTZ5ZKFySly4WQgpGXLj/D7pKte57K5RL74ZoxtUkHb7+zhGarDaBc1Rc5QyekQOQ5YyPPS+xdV9Ere67F5Dlr0F4ZfJ7Ig2soDijohBSIPGdsFGWJfZmDpHS5EFIg8i5GRVhiX+bqi5yhE1IgxiljIyny7BoaFQo6IQWizGKUFkVxDUXB2uUiIhUABwE0VPWTInIZgG8DuADA8wB+T1XfTcZMQgiQXpefslME11AUwvjQvwDgJQDv7f7/DwHcp6rfFpE/BfA5AF+P2T5CyBBFF6O8pF2WESuXi4hcAuBaAN/o/l8AbAewv/uRBwHMJmEgIaQ85DntsgzY+tDvB/D7AFa6/78QQFNVl7r//xkAzyFWRG4XkYMicvDEiRMjGUsIKTZ5TrssA4GCLiKfBPCGqh7qf9njo54rf1X1AVWdUdWZqampiGYSQspA3tMui46ND30rgOtE5BMA1qLjQ78fQE1E1nRn6ZcAOJ6cmYSQMlDmHPA8EDhDV9WvqOolqroBwGcAPKWqtwJ4GsAN3Y/dBuB7iVlJCCkFcaZdjksXojCMkof+ZQBfEpH/i45P/ZvxmEQIKStx5YAzuOqNqJqKXsbPzMyMHjx4MLXjEULKiamyY71WxTO7tmdgUbKIyCFVnQn6HGu5EFJyypj3zeCqN1z6T0iJKatrgjVtvKGgE1Jiypj3Pb/QwOl3l1a9zpo2dLkQUlhsXCllc024TxzDg1St6mD3dZcX3pU0KhR0QgrIV+dfxLeePdpbzWdqo3Z+1em1Wuvn/KqThpmx4/XEAQDnnbtm7MUcoMuFkMIxv9AYEHMXL1eKeK3p9nk975TtiSNuKOiEFIy9B45419nAamFrnl49O/d7Pe8wGOoPBZ2QguE3Gx0WNpPQKYDpP/hh4bJd2ODDHwo6GVuKunTcJNICrBK2uR0b4VS8/SsnT7dxx74XCiXsZe42FAcMipKxZDhbwhRUzCNzOzauyvQQALdevd7b9oDF4CdPt3vnDuS/G1LRG3wkCQWdjCV++dl5F4swbej2HjiC9kpweY9Wexl3P3EY77RXCjnIkQ50uZCxpOjZErPTdczt2IiLa1Ucb7aw98ART7dJmPM5ebpdukVI4wZn6GQsKXpdbluXkek8wzDqIFfGWjJ5hTN0MpbkPVsiKGBru6Tf6zzDMsogV9ZaMnmFgk7GkjxnS9iIoGnW3Gi2BgaA/vOMwqiDXBlryeQZulzI2JLXbAmbgG1t0sFJw+KgYffL7HQdB197Ew89ezTw2FVnAhecd25s7pGixyqKBgWdkJxhI4JBfWmGB4CHnztmdeylFY3Vx130WEXRoMuFkJxhs7z9LY+CW8P0DwDLlp3J2ssaqzsk77GKskFBJyRn2IigzQy3/zOVENW44nSH5DlWUUbociEkZ9gsHPJaLdrP8ABwy1WXWvnQgfjdIXmNVZQRCjohOSRIBN337n7icC84Kuis8q97DAD3zG7Gd59v4NS73gOAi5c7hHnkxYGCTkgOsRHRg6+9OVAGV3FWkL0E93SAmJ/dy6Adc/sX0V7uvN5otjC3fxEAywHkEQo6ITnDZhWoTZOLvQeOoNFsoSKCZdXeTz9a7RXMPXpWsO9+4nBPzF3ay4q7nzhMQc8hDIoSkjNsFuP4NblwBwA3XdAVcetMl5WzmS6mXHfT6yRbOEMnJCGi+p5t8tD9MlEqIsZgKXDW1x7FBpJvKOiEJEDYeuv94j9hcI30Z5+YFuwIgmfiNvN091g1Q5PpWkGbTJcdulwISYAwNUyGa7eYBPnUmaVejRavXHW3ycWoYutMSC/TZfd1l8OZkFXv777u8pGOQZKBM3RCEiBMDRMv8fei2WqvmuV7uXSe/JvXjfuoOhWsdSaMPvBa1cHu6y7v7T9MMw2SPRR0QhIgTA2TMP7q/hotplz1pk/AstVexrlrJuBUZCB7pepUPFdwMge9WFDQCUmAbZumVqUVmmqYhG1CETQABO2v2WrDmRCsm3TQPN1eJdSuiDearYEAKlvS5R/60AmJmfmFBh471BgQcwFw/ZXeM+qwTSjOD/CR2+yvvaKYPGcNXtlzLZ7ZtX1AzPtTHv3y3En+oKATEjNePnEF8PTLJzw/7xawWjdpF8w89e6Sb8ef4YJYJoZn+vMLDdz5yGKgP58pjfmFgk5IzNh2E+pndrqOha9dYyXqwyVuvdrVzU7X8cyu7Xhlz7XGbkX9/nx3Zm6z+Ii1zPMLBZ2QmPETPK92cv2CbLsC0x00bNrV2ZTjtc20YS3zfBMo6CKyVkR+LCKLInJYRO7uvn6ZiDwnIn8vIvtE5JzkzSUk/wT5sPv90MOCbIs7aPjlu7sDxRf3vYBz10xg3aRjrEnu50Zx3TasZZ5/bLJczgDYrqpvi4gD4K9F5C8AfAnAfar6bRH5UwCfA/D1BG0lpBD0526bsk3c121nxv30z5L93Dv9K1WbrTaqTgX33XyFpyCbMmMqIvijm7ZQxAtC4AxdO7zd/a/T/acAtgPY3339QQCziVhISAGZna5jbsdGY6cgQWd2HiZdEQDOO+dsvvj8QgMThv171XPxy1AxuWUo5sXCKg9dRCoADgH4xwD+BMBPADRVdan7kZ8B4F0nhSLJRTNBQUZFZ3ZuU9K2H7em+VfnX/Qsnwt0hNg06zfN6LkitBxYCbqqLgO4QkRqAL4L4De8Pua1rYjcDuB2AFi/fn1EMwmJl7DFs8Jy9xOHrdL/wvjNgc6XbPfjh/FWq+25bUUE9+7cbHT3+AVs2Squ+IRaKaqqTRH5KwBXA6iJyJruLP0SAMcN2zwA4AEAmJmZCfv3S0gi+AUTo4ha/2y/NulYZau44mryXZtm7l7VD12WVfHFfS+gNunAmRC0VwaX9zNDpdzYZLlMdWfmEJEqgN8F8BKApwHc0P3YbQC+l5SRhMRNmOJZQQxnqtiIuVPpVDTctmnK8/2rP7TOd1GQHz0bpFNsy5TZQsqHzQz9IgAPdv3oEwAeUdXvi8jfAfi2iNwDYAHANxO0k5BYCVM8y0R/zZOwtJcVdz6yaJyFv/oPLdx69XrPejB+1RKHj/Grd5aMmS2kfAQKuqr+DYBpj9d/CuAjSRhFSNLM7dg44EMHwrkkhn3wUfALhh5vtnDP7GYAwMPPHev1BL3+yjpmPniB9bGXVVlQa4zgSlEylgzXO7F1SbiLde7Y98JIYh7ExbUq5hca2PfjYwM9Qff9+BgArLLdr2QAC2qND6IhUqZGZWZmRg8ePJja8Uj+KVK97TCzcqciOO+cNb4BTL9t996wBbsfP2xs//bCXdeEsk0AvLLn2tC2kHwgIodUdSboc6yHTjIj6dTBuLFd1VkfGpg27Hoy3IG6cyzTYOD1unssk1+eBbXGg0IIepFmccSeuFMHkyYoA8bU9acesoFFe0UjuUjc444SGyDFJvc+dJtqcqSYxJk6mAZ+s1w/H3wUMW00W5h0zF9P099/1NgAKQe5n6EXbRZH7IkjdXBUvJ7+AO8l8KbMGJtenFVnAq32irVdFRGc61Rw2rCN398/V3yOL7kX9KLN4og9o6YOjoqXD3/u0UVA0Gug7OXXD3L/ee3XqYjnyk2TT35Z1bfZM//+iRe5F/Q8zOJIMmRdEMrr6a9fcF36nwhtZr+e+11WnHdOBSvtlYGc8qdfPmH0r4sApiQ0kbPB1lrVwe7rLuesnOTfh27TbYUUk6yD3WFmuTafdXPUTQJ96t3lgZzyxw41sG3TlLEZhsfY4vles9XG3KOLjCuR/As6gzzlJA/B7jBPeUGf7T8fW1rtZXx/8XWs9Ql+Ap3ZeO93w2eiZsaQcpF7lwvAIE8ZyUOw28uH70zIgA8dsHsijNJ5CPCvnNhDgVe7i4Iu88lpp1+dFELQSfnIQ7Db5MP3ei1okPGzu16r4tSZpUirRoHBpwNTTGn4c2Q8oaCTTMhLsNv09Bf2KcF0PvVaFc/s2h65mJdbZtdlbsdGzO1fHHiCADpPFowrEQo6yYSsUxZdhgOz2zZN4emXT4QO1Aadj9fTwOl3l3zL4K6bdHDXpwazV9zf737icG9bZrkQFwo6yYSsUhaHOwu9/c5SL1Wx0WzhoWeP9j5rqi0zv9DwFNR7d24eeP3cNf7Bzmt/+yI8dqgxMAgIOqVchuvB9OMXU8o6c4hkC6stklLhJ2hR3R4VEayo9mbwD//4GJY9cgonnQm0l3Ugl90V6HVDgwfQmcH356K7n+1//96dnZroNiLtdX6mlaykWNhWW6Sgk1IwPGt26Rc0vxzxrHB97CbbalUHZ5ZWrETatA/3GKS4sHwuGRv8Zt6t9jJ2P34YB197M3diDpxtEG3KkvHKjDGld+Yhc4hkS+4XFhESRFAOeLPVHvCN54lKd9VQ2OweL5E27YPpjOMDZ+iksIzSpDkvLKvisl1PojbpeBbvMjWEdlvUDWfoDAdZWSZjvKCgk1hJK8sijibNeUGBVaLtZs4A3g0rtm2aWlXR8aFnj6LqTGDdpIPm6TazXMYQCjqJjTRbykVdal8Uzix16qCb0jtN59+puS647+YrKORjyNgIOvNzk8e2PovpXoS5R2UP9AWV7P3ivhestiXjxVgIetGaERcVmywL0704+NqbA/7foHvkV9OkLPgNWkHnX/YBj3gzFlkufjNHEh82WRame/Hwc8dC3aNtm6ZGtDb/+GWnePUJsN2WlJexmKEzPzcdbOqzmGaVy4YFbu49ml9oYPfjhyNXLCwaQdkpXjVdbLcl5WUsZujMz00Hm2YkFTG1aPDGTc+be3RxbMS8ImK1XH92uo6Fr12D+2++gg1gCIAxWfrPGhf5YYNPg4bhpsnuPSp6rnkUBIgcMCblg0v/+8i6GXEeyIsg1H3qhrvpeI1mCxURtNrLni6FcaC/LV/YgDHJD2l/78Zihj7u5OkJJciWPCwYqohgWbX3Myr333wF7nxk0bgPpyKAYmB1qJ89w7DoVr6J83tnO0MfCx/6uJOnLJ8gP3seFgy54jmKmAOdc/Xbx94btmDvjVt61yLInmEY1M83WXzvxsLlMu7kLcvHr0FDWURq3aQDwN/F5F4D96ep/K1phs6gfr7J4nvHGfoYkHSWz/xCA1v3PIXLdj2JrXuewvxCI/I+bObE6yYdTIRLlkkVpyK461OdOixe+eJORXDqzNKq6+X12apTwS1XXer5OlMT800W2XUU9DHAJBRxCILrJ2w0WwOBvDCi3r+PIKpOBWfaywhwO2dGRQR7b9gyMPvudzGtm3QA7ZT0Hb5eJnfUPbObA9NBSf5I8ntngkHRMSGpaHscXXL8Ogmtm3SgCrzVOls98A6fOiZZYhPwYleh8SKu711saYsicimA/wngHwFYAfCAqv6xiFwAYB+ADQBeBXCTqp4MbSlJBT+/9Sj4+Qlt/5hN+xAAC1+7BsDZL4ZfUaq0EAEuPr/aS69cVu2lXQId0Tadc97iGSRZkvrembAJii4BuFNVnxeRXwNwSET+EsC/AvAjVd0jIrsA7ALw5eRMJXmkNul45onXJh3rgmimQlMTItiw60lMCHLlYlGF52zapgic6VwZ4CRxEOhDV9XXVfX57u+/AvASgDqATwN4sPuxBwHMJmUkySfzCw28/c6S53vNVts6ZctUaMrN7MiTmAMd98gw8wsN3PnIYuA5Z+FXJeNDqLRFEdkAYBrAcwA+oKqvAx3RF5H3x24dyTV7DxwxLooxhWa8XAvDK3knRlzQkzQbLhwUdHdmbpMvzlXLJEmsBV1E3gPgMQB3qOovxbLIkojcDuB2AFi/fn0UG0lOieL37XctePnYAeQ26OnyzE/exFfnX8Q9s5sBBC+GGnanpO1XJeODVdqiiDjoiPm3VPU73Zd/LiIXdd+/CMAbXtuq6gOqOqOqM1NT5a9hPU6E9fv2uxa80h3n9i/iSzkXc5eHnzvW+90v3ZLuFJImgYIunan4NwG8pKr/re+txwHc1v39NgDfi988kmeCmiz0Y7PEv72sWIndymRw3SvzCw3jsn3bMriExIWNy2UrgN8D8KKIuNOn/wBgD4BHRORzAI4CuDEZE0le6RfnoEVBw1khRU/TmxD//HkB8Ec3baGYk1QJFHRV/WvAOAn5WLzmkKLR7w/+zf/4FzjdXj3Hduua9FP4nqDq72rJb0iXlBku/Sex8Z93/nanJGwf/XVN+sljT9CKT4GY4XdsXENhSyAQMiqstphj8tKUwpYwKXlPv3wibfN8mXQmPJ8uRsHNQc/zPSPlgoKeU2xWHaZhQ9gBxTYlL28+9CAxj+pCydt5hqFoEwpCl0tuybopRRxVFP32PRGyWXRRKeqS/iTvP0kOCnpOybqIU1IDStCqyjwSdejxykGPo3Z8GmQ9oSDRoKDnlCyK4/eT1IAStcWcMyG9bJlRmltE2fTWq9db59u7eNUsL9KsN+sJBYkGBT2nZF3EKakBJYogCICbP3Ip7vrU5ajXqiMV61IAterqNEoT9Vq112CiYlvuAp0snmF/c5FmvVlPKEg0GBTNKVkXcZrbsdGzY/moA0qU/HMF8P3F17Hv/xxDezk9V03/+brXffiaeKEAHnr2KL6/+PpAY44izXqTuv8kWdixiKzCzW7wauAw6oAynL2TN+q1qu8AOr/QiFQ8rOpUsNaZ8Kwd7x43b1kkzHLJD7F1LCLlweYLOiy4y6q9mVkcX+Yw5QKiIOjMkNcZGm/44dUGzlQRMiyt9jLOXTOBqlPxHMyySEsNglUhiwd96GOCbUAuDT/v7HQdz+zajlf3XBvL/lzfdr1Wxa1Xr0e9VkUzQMxt4hOma1Z1on1t3mq1e82evcirP50UBwr6mGAr1HH6eW1S9MIEKL2oVR385N5P4P6br8CpM0t46NmjPQH22+benZsHjr3WQ6RN12xtyIwXl4tr1d5gZgqv5tGfTooDBX1MsBXquLIbbJ4Ivjr/IpqtcG6RfpwJwe7rLu8dy2Zf7jYAcGbp7OrQk6fbq+wzXbOg2b8Xw08AzCIhSUBBHxNMQjEhMiBicaVLBj0RzC808K1nj4baZz/1WhV7b+yUp7XJbReLbYafWPxE1+Q2GT6ma6ubk+4+tTSarVWzdGaRkFFhUHRM8EpDAzpBz/5gXFzpkqaAp/v63gNHItdHEQzWVw9yU3gFO4PsAzrXbO7RxVV9U493nzrcAKwJHTr2cMC5fx8VkYEBJYtgJLNaig8FfUxwv5h3PrK4atn9cFXAOLIbKoZGz24AcxRf8flDfne/3PaqU8G2TVPYuuepAaEKsq+Hh7Nb+34GiXr/eXo9Fbj7cG3JKtslD8XgyOjQ5VJy+gOTew8csepMHwem47ivj+IrbrbaA0FWUyu8dZMOrr+yjscONQZ8+Xfse8HXPne/ew8cCVzI5M7CTS6Y/vM0XePhI2SR7VKkVazEDGfoJcZr1mWaUcYRjOt/ZDfNgF3hM7mAbPGaQXrli3s9kQTh7td2kHOfDoavrVcg1Db3Pu1slyKtYiVmKOglxvSI78WoHYS8FiQN47WU/u4nDq9aABTkxnDpdxUNu4lGqero7jdsmYIgn7jXIOY3wKbp0zadK7NuigVdLpYUpexpP2HEaNQOQkGZJl7VBwHgl62lVZ91hdEG0wwyalVHl0azZXTl+GHyic8vNDA7Xcf1V9Z7fvqKCH7n1y/wzCratmkq1cqMWReDI/FAQbegSGVPXeYXGqFKxY76aO23/XDpgPmFBq64+4e+vmzXNw14BCr78JpBzi80Yisr4Ley04TJJz6/0MBjhxq9c15WxfNH38L1V9ZRr1V7qZX37tyMp18+kapPe3a63jvXfjsYEC0WLM5lgZs3PIxXOlxeMNlsYtRzCTqeu3/b4lzD9nht57or6n0+892PHx5psZLJBr/j2yDoZOd42ebaP1wQzbSfV2IqmUCKA4tzxUgRA0ZBM+aoZVFNfl1TzvawPTauEKciOHVmCZftenKV79gVvX4xbTRbmHt0ESsAln2KpYcR4H6bAW+f/1pnAksralXSt+ZTLMx94vOLP7jQp038oMvFgiIu0zbZ5j5KR3m09nM9zU7X8Z615vmBa0/QIChd1W222p7HeGbXdtRr1VXC3F5RXzHv7DzwFD1t7uedvmbSrfaKlZhXnQr8HoTdAKrNfujTJn5Q0C0oYsDIz2ZXGF/Zcy2e2bXd2k8alKvsV+PEvVZBg6AAq2b5rfYy7n7icO//UZ6M6rWqr6h62dFotgYC4FECre6A+ZaPG8gmE4c+bWIDBd2CIgaMkrA5yPVkEuta1ekdNyhzxDTJPnm63RPWKE9G2zZNWbeQAwbdOe4TQtiBxC1RMDtdN9q8btIJDLq6vvw8/72RfMCgKLEmKDjsFTisOhXcu3MzgLMLf2qTDlQ79cEnfAKAYY5js+22TVN4KKAgmMnP7opu1EBz0LUxnY/7GYr5eGMbFOUMnVhjml2ffnep5+P2eioAMOB7P3m6jTNLK7jv5iuwEmJC4c6Q3eOE4XizhXtmN+OzV68fyAPf+usXDNhrsuZ4yLz0YZec3xNT/3uuXUAxngRJvuAM3ZIiVqJLwub5hYZnaqDfTNIvpdEvRW+Y4VTGKKmZpmvQ30fV79g2PUVrVQe7r7s8938fpDhwhh4jRV1YlITNs9N1nHfu6mwWN3DptZrWz/fsJeZOReBMDPq7vYLQXjNmZ0LgVLx95aZr0H+tvBguWRDki+9vnEFImlDQLShiJbokbTYJ9MnTbc8BxCaIWRE524Tihi3Ye+OWwICulxtj741bsPeGLaH6dvplr3gdO+iJIu9/G6S8cGGRBWVaWBSHzbZFq1xhm9uxEXP7F31ztldUV62AtHFZmGq3z07XcdmuJz194sPXwHRNhhtpuNQtzj/PfxukvHCGbkGZFhbFYXOY4ODxZqvjpjlcMgupAAAJD0lEQVTHf+4QpWdpULE022sQ9lrZnH9t0ilcMTdSfCjoFpRtYdGoeLk6akNdhFxcUfRbWBPWLtv4gO012HCht3CbXh/OShn2qDsVwdvvLBUq5kLKAV0uFsTVZzMKUTNV0rb5k1suwmOHGsYaMSY3TUUkVGre/ELDqo0eYH8Nnv3pSc9jmV53991fPbL/GKfOLK3KAvKyb1SKmHlFkiUwbVFE/juATwJ4Q1V/q/vaBQD2AdgA4FUAN6mq+a+/S5HTFrPAbzGK3xc36S+6ya7rr6zj6ZdPeB436rkEHbefqJUIN+x60vjeqxH2Z/Ldx1kpMY7rSYpDnGmLfwbgXwy9tgvAj1T1nwD4Uff/JGaiZKqkkWJpsuvpl08Ya8TEUYogqJZK1PiAXxpilOuWRsyliJlXJHkCXS6q+r9FZMPQy58G8M+6vz8I4K8AfDlGuwiiZar4fdGHW7RFncVHzaAxZaTY4rd/ASLHB2656lJjSYAobhKvVnNxx1yKmHlFkieqD/0Dqvo6AKjq6yLy/hhtIl2i9Hm0+aJ7NY8ebrgct11x4JcuqbCz3Yt7ZjcbBX34enoNhMBqP/29Ozcn6vZiD1DiReJZLiJyu4gcFJGDJ06M1rdy3IiSqWLzuD/q43pWWT9zOzYaS5qHbRNnu33/dfNyZ809uoi5/YurXFwAIpUotqWImVckeaIK+s9F5CIA6P58w/RBVX1AVWdUdWZqarTO8uNGFL+zzRd91Mf1rMoJz07XcevV61eJehxCZnPdvAbCtkfHojR82UUs6UySJ6rL5XEAtwHY0/35vdgsKhFxZJuE9TvbpOrF8bg+qj88KvfMbsbMBy8wnl+SaZ5h/NNp+LKzugckvwQKuog8jE4A9H0i8jMAd6Ej5I+IyOcAHAVwY5JGFpFR/dSjEPRFTyNo50UUsTVtY6qYGOaae+3br1G2bckD97OEpI1Nlssthrc+FrMtpcI22yQL4lx0ZCvSUQY4m236jw+sbk5huuZR7PEaCJ0JAQQDbpc4BkcuGiJR4ErRhMh7WtnwLNetjRJ29mwrilEGuKBtbDsXeV3zKPaYBkKv10YR3yyf7kixoaAnRJHSyqIKSBhRjDLABW1j27TZ65rHnUsfp9Dm+emO5BsW50qIIqWVRU1jDCOKUVZPBm1j+7Tjdc3zXEEz7093JL9Q0BMir2llXmVnowpIGFGMMsAFbWMjvusmHc9rntcBd36hgQlDKYI8DDYk39DlkiB5SyszuVZqkw5Onl5d3tZLQPqDdedXHTgVsQoIRgnEBm3jFaTsp+pUcNenLo+0b7/z7v9snMFL9/54dUTKw2BD8g+bRGdMmtkMpqbKtaqDM0srgZX7vIKQzoTgPWvXoHm6jYtrVWzbNGWsuJgEwwOMCHq2xHVsv+qSXiWDoz6Jme5PRQR/dNOWXE0OSLrYVlvkDD1D0s5mMLlQ3mq1cd/NVwQOLKaVkpPnrMHC167JJDsjjacgU4zh4eeOWdVlt8V0f1ZUKebECgp6hqSdzeCXeWMjjFGyTsqQnWE6b1Oz6KjByyJlRpF8wqBohqSdzTBqIDBq1kkS52PTUzQuTOdtqqMeVYDzGqglxYGCniFpp86NmnkTNesk7vNJo4lHP6bzvuWqS2MV4LxmRpHiQJdLhmRRU2UUn3OUrJMkzidt147fefsVCot6LAo4iQqzXDKmbDU70jgfm56dZbuuZLxhlktBKNuMLI3zCQoeshYKGVfoQyeFI8iXzwbKZFzhDJ0kRpz1z/sJ8uWzFgoZVyjoJBGSqn/u4ufaYT43GVfociGJEMXtEZerhPncZFzhDJ0kQhL1z22JsyMTIUWCgk4SIYrbI05XSdmyhwixgS4XkghJ1D8nhPjDGTpJhCTqnxNC/OFKUUIIyTm2K0XpciGEkJJAQSeEkJJAQSeEkJJAQSeEkJJAQSeEkJKQapaLiJwA8FpqB4zG+wD8ImsjUoDnWS7G5TyB8TnX/vP8oKpOBW2QqqAXARE5aJMeVHR4nuViXM4TGJ9zjXKedLkQQkhJoKATQkhJoKCv5oGsDUgJnme5GJfzBMbnXEOfJ33ohBBSEjhDJ4SQkkBB70NEKiKyICLfz9qWJBGRV0XkRRF5QURKWy1NRGoisl9EXhaRl0Tkn2ZtU9yIyMbufXT//VJE7sjariQQkS+KyGER+VsReVhE1mZtUxKIyBe653g47L1k+dxBvgDgJQDvzdqQFNimqmXP5f1jAD9Q1RtE5BwAk1kbFDeqegTAFUBnQgKgAeC7mRqVACJSB/DvAfymqrZE5BEAnwHwZ5kaFjMi8lsA/g2AjwB4F8APRORJVf17m+05Q+8iIpcAuBbAN7K2hYyOiLwXwEcBfBMAVPVdVW1ma1XifAzAT1Q174v3orIGQFVE1qAzOB/P2J4k+A0Az6rqaVVdAvC/APxL240p6Ge5H8DvA1jJ2pAUUAA/FJFDInJ71sYkxIcAnADwP7putG+IyHlZG5UwnwHwcNZGJIGqNgD8VwBHAbwO4C1V/WG2ViXC3wL4qIhcKCKTAD4B4FLbjSnoAETkkwDeUNVDWduSEltV9cMAPg7g8yLy0awNSoA1AD4M4OuqOg3gFIBd2ZqUHF2X0nUAHs3aliQQkXUAPg3gMgAXAzhPRD6brVXxo6ovAfhDAH8J4AcAFgEs2W5PQe+wFcB1IvIqgG8D2C4iD2VrUnKo6vHuzzfQ8bd+JFuLEuFnAH6mqs91/78fHYEvKx8H8Lyq/jxrQxLidwG8oqonVLUN4DsAfidjmxJBVb+pqh9W1Y8CeBOAlf8coKADAFT1K6p6iapuQOex9SlVLd3oDwAicp6I/Jr7O4Br0HnMKxWq+v8AHBMRt8P0xwD8XYYmJc0tKKm7pctRAFeLyKSICDr386WMbUoEEXl/9+d6ADsR4r4yy2X8+ACA73a+E1gD4M9V9QfZmpQY/w7At7ruiJ8C+NcZ25MIXV/rPwfwb7O2JSlU9TkR2Q/geXRcEAso74rRx0TkQgBtAJ9X1ZO2G3KlKCGElAS6XAghpCRQ0AkhpCRQ0AkhpCRQ0AkhpCRQ0AkhpCRQ0AkhpCRQ0AkhpCRQ0AkhpCT8f9F55igLg2gbAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1361b3feda0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 看下数据\n",
    "plt.scatter(x, y)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(490,) (490,)\n"
     ]
    }
   ],
   "source": [
    "# 去除掉离群点\n",
    "x = x[y < 50]\n",
    "y = y[y < 50]\n",
    "print(x.shape, y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1361987d6d8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(x, y)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 使用简单线性回归法"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(392,) (98,)\n",
      "(392,) (98,)\n"
     ]
    }
   ],
   "source": [
    "from c2_model_selection import train_test_split\n",
    "\n",
    "x_train, y_train, x_test, y_test = train_test_split(x, y, seed=666)\n",
    "print(x_train.shape, x_test.shape)\n",
    "print(y_train.shape, y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from c6_SimpleLinearRegression import SimpleLinearRegression2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Simple Linear Regression2()"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reg = SimpleLinearRegression2()\n",
    "reg.fit(x_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "y = ax + b, a is 7.860854, b is -27.459343\n"
     ]
    }
   ],
   "source": [
    "print('y = ax + b, a is %f, b is %f' % (reg.a_, reg.b_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1361b561f60>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(x_train, y_train)\n",
    "plt.plot(x_train, reg.predict(x_train), color='red')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_predict = reg.predict(x_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# MSE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "24.156602134387438"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mse_test = np.sum((y_predict - y_test)**2)/len(y_test)\n",
    "mse_test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# RMSE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4.914936635846635"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from math import sqrt\n",
    "\n",
    "rmse_test = sqrt(mse_test)\n",
    "rmse_test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# MAE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.5430974409463873"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mae_test = np.sum(np.absolute(y_predict - y_test)) /len(y_test)\n",
    "mae_test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 用自定义的评价函数"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "from c3_accuracy_score import mean_squared_error, mean_absolute_error, root_mean_squared_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "24.156602134387438"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_squared_error(y_test, y_predict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4.914936635846635"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "root_mean_squared_error(y_test, y_predict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.5430974409463873"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_absolute_error(y_test, y_predict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# scikit-learn中的MSE和MAE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.metrics import mean_absolute_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "24.156602134387438"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_squared_error(y_test, y_predict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.5430974409463873"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_absolute_error(y_test, y_predict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 最好的衡量线性回归法的指标 R Squared"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### $R^2 = 1 - \\frac{SS_{residual}}{SS_{total}}$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### $R^2 = 1 - \\frac{\\sum_i{(\\hat y^{(i)} - y^{(i)})^2}}{\\sum_i{(\\overline y - y^{(i)})^2}}\n",
    "= 1 - \\frac{(\\sum_i{(\\hat y^{(i)} - y^{(i)})^2})/m}{(\\sum_i{(\\overline y - y^{(i)})^2})/m}\n",
    "= 1 - \\frac{MSE(\\hat y,\\  y)}{Var(y)}$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* R^2 <= 1, 且越大越好\n",
    "* R^2 < 0 说明我们学习到的模型还不如基准模型, 此时很有可能我们的数据不存在任何线性关系 \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6129316803937322"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "R_square = 1 - mean_squared_error(y_test, y_predict)/np.var(y_test)\n",
    "R_square"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6129316803937322"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from c3_accuracy_score import r2_score\n",
    "R_square2 = r2_score(y_test, y_predict)\n",
    "R_square2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# scikit-learn中的R Square"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import r2_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6129316803937324"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r2_score(y_test, y_predict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6129316803937322"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reg.score(x_test, y_test)"
   ]
  },
  {
   "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}