diff --git a/notebook_必备数学基础/.ipynb_checkpoints/案例：汽车价格预测任务-checkpoint.ipynb b/notebook_必备数学基础/.ipynb_checkpoints/案例：汽车价格预测任务-checkpoint.ipynb
index f22441e..c9cc8a0 100644
--- a/notebook_必备数学基础/.ipynb_checkpoints/案例：汽车价格预测任务-checkpoint.ipynb
+++ b/notebook_必备数学基础/.ipynb_checkpoints/案例：汽车价格预测任务-checkpoint.ipynb
@@ -78,6 +78,8 @@
     "import numpy as np\n",
     "import pandas as pd\n",
     "from pandas import datetime\n",
+    "import warnings  # 忽略普通警告，不打印太多东西\n",
+    "warnings.filterwarnings('ignore')\n",
     "\n",
     "#data visualization and missing values\n",
     "import matplotlib.pyplot as plt\n",
@@ -639,7 +641,7 @@
     {
      "data": {
       "text/plain": [
-       "<matplotlib.axes._subplots.AxesSubplot at 0x25cc901b3c8>"
+       "<matplotlib.axes._subplots.AxesSubplot at 0x1731f11d320>"
       ]
      },
      "execution_count": 5,
@@ -1185,7 +1187,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 28,
+   "execution_count": 10,
    "metadata": {},
    "outputs": [
     {
@@ -1212,15 +1214,19 @@
        "      <th>symboling</th>\n",
        "      <th>normalized-losses</th>\n",
        "      <th>wheel-base</th>\n",
+       "      <th>length</th>\n",
+       "      <th>width</th>\n",
+       "      <th>height</th>\n",
+       "      <th>curb-weight</th>\n",
        "      <th>engine-size</th>\n",
        "      <th>bore</th>\n",
        "      <th>stroke</th>\n",
        "      <th>compress-ratio</th>\n",
        "      <th>horsepower</th>\n",
        "      <th>peak-rpm</th>\n",
+       "      <th>city-mpg</th>\n",
        "      <th>highway-mpg</th>\n",
        "      <th>price</th>\n",
-       "      <th>volume</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
@@ -1229,230 +1235,384 @@
        "      <td>1.000000</td>\n",
        "      <td>0.389010</td>\n",
        "      <td>-0.531954</td>\n",
+       "      <td>-0.357612</td>\n",
+       "      <td>-0.232919</td>\n",
+       "      <td>-0.541038</td>\n",
+       "      <td>-0.227691</td>\n",
        "      <td>-0.105790</td>\n",
        "      <td>-0.160225</td>\n",
        "      <td>-0.020132</td>\n",
        "      <td>-0.178515</td>\n",
        "      <td>0.070421</td>\n",
        "      <td>0.273125</td>\n",
+       "      <td>-0.035823</td>\n",
        "      <td>0.034606</td>\n",
        "      <td>-0.079290</td>\n",
-       "      <td>-0.456263</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>normalized-losses</th>\n",
        "      <td>0.389010</td>\n",
        "      <td>1.000000</td>\n",
        "      <td>0.008200</td>\n",
+       "      <td>0.152832</td>\n",
+       "      <td>0.200870</td>\n",
+       "      <td>-0.287893</td>\n",
+       "      <td>0.218460</td>\n",
        "      <td>0.237114</td>\n",
        "      <td>0.047391</td>\n",
        "      <td>0.070319</td>\n",
        "      <td>-0.073362</td>\n",
        "      <td>0.370384</td>\n",
        "      <td>0.234384</td>\n",
+       "      <td>-0.321647</td>\n",
        "      <td>-0.270414</td>\n",
        "      <td>0.313716</td>\n",
-       "      <td>0.036586</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>wheel-base</th>\n",
        "      <td>-0.531954</td>\n",
        "      <td>0.008200</td>\n",
        "      <td>1.000000</td>\n",
+       "      <td>0.874587</td>\n",
+       "      <td>0.795144</td>\n",
+       "      <td>0.589435</td>\n",
+       "      <td>0.776386</td>\n",
        "      <td>0.569329</td>\n",
        "      <td>0.495108</td>\n",
        "      <td>0.164549</td>\n",
        "      <td>0.249786</td>\n",
        "      <td>0.301696</td>\n",
        "      <td>-0.363355</td>\n",
+       "      <td>-0.470414</td>\n",
        "      <td>-0.544082</td>\n",
        "      <td>0.572348</td>\n",
-       "      <td>0.913669</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>length</th>\n",
+       "      <td>-0.357612</td>\n",
+       "      <td>0.152832</td>\n",
+       "      <td>0.874587</td>\n",
+       "      <td>1.000000</td>\n",
+       "      <td>0.841118</td>\n",
+       "      <td>0.491029</td>\n",
+       "      <td>0.877728</td>\n",
+       "      <td>0.683360</td>\n",
+       "      <td>0.608905</td>\n",
+       "      <td>0.132076</td>\n",
+       "      <td>0.158414</td>\n",
+       "      <td>0.521192</td>\n",
+       "      <td>-0.279406</td>\n",
+       "      <td>-0.670909</td>\n",
+       "      <td>-0.704662</td>\n",
+       "      <td>0.680804</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>width</th>\n",
+       "      <td>-0.232919</td>\n",
+       "      <td>0.200870</td>\n",
+       "      <td>0.795144</td>\n",
+       "      <td>0.841118</td>\n",
+       "      <td>1.000000</td>\n",
+       "      <td>0.279210</td>\n",
+       "      <td>0.867032</td>\n",
+       "      <td>0.735433</td>\n",
+       "      <td>0.556374</td>\n",
+       "      <td>0.183379</td>\n",
+       "      <td>0.181129</td>\n",
+       "      <td>0.596251</td>\n",
+       "      <td>-0.214240</td>\n",
+       "      <td>-0.642704</td>\n",
+       "      <td>-0.677218</td>\n",
+       "      <td>0.765788</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>height</th>\n",
+       "      <td>-0.541038</td>\n",
+       "      <td>-0.287893</td>\n",
+       "      <td>0.589435</td>\n",
+       "      <td>0.491029</td>\n",
+       "      <td>0.279210</td>\n",
+       "      <td>1.000000</td>\n",
+       "      <td>0.295572</td>\n",
+       "      <td>0.067149</td>\n",
+       "      <td>0.199995</td>\n",
+       "      <td>-0.044176</td>\n",
+       "      <td>0.261214</td>\n",
+       "      <td>-0.114968</td>\n",
+       "      <td>-0.322525</td>\n",
+       "      <td>-0.048640</td>\n",
+       "      <td>-0.107358</td>\n",
+       "      <td>0.113942</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>curb-weight</th>\n",
+       "      <td>-0.227691</td>\n",
+       "      <td>0.218460</td>\n",
+       "      <td>0.776386</td>\n",
+       "      <td>0.877728</td>\n",
+       "      <td>0.867032</td>\n",
+       "      <td>0.295572</td>\n",
+       "      <td>1.000000</td>\n",
+       "      <td>0.850594</td>\n",
+       "      <td>0.648219</td>\n",
+       "      <td>0.170425</td>\n",
+       "      <td>0.151362</td>\n",
+       "      <td>0.679865</td>\n",
+       "      <td>-0.264976</td>\n",
+       "      <td>-0.757414</td>\n",
+       "      <td>-0.797465</td>\n",
+       "      <td>0.836802</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>engine-size</th>\n",
        "      <td>-0.105790</td>\n",
        "      <td>0.237114</td>\n",
        "      <td>0.569329</td>\n",
+       "      <td>0.683360</td>\n",
+       "      <td>0.735433</td>\n",
+       "      <td>0.067149</td>\n",
+       "      <td>0.850594</td>\n",
        "      <td>1.000000</td>\n",
        "      <td>0.602516</td>\n",
        "      <td>0.211477</td>\n",
        "      <td>0.028971</td>\n",
        "      <td>0.742119</td>\n",
        "      <td>-0.241031</td>\n",
+       "      <td>-0.653658</td>\n",
        "      <td>-0.677470</td>\n",
        "      <td>0.871189</td>\n",
-       "      <td>0.594351</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>bore</th>\n",
        "      <td>-0.160225</td>\n",
        "      <td>0.047391</td>\n",
        "      <td>0.495108</td>\n",
+       "      <td>0.608905</td>\n",
+       "      <td>0.556374</td>\n",
+       "      <td>0.199995</td>\n",
+       "      <td>0.648219</td>\n",
        "      <td>0.602516</td>\n",
        "      <td>1.000000</td>\n",
        "      <td>-0.049492</td>\n",
        "      <td>0.008511</td>\n",
        "      <td>0.537543</td>\n",
        "      <td>-0.276942</td>\n",
+       "      <td>-0.556570</td>\n",
        "      <td>-0.562065</td>\n",
        "      <td>0.550994</td>\n",
-       "      <td>0.549601</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>stroke</th>\n",
        "      <td>-0.020132</td>\n",
        "      <td>0.070319</td>\n",
        "      <td>0.164549</td>\n",
+       "      <td>0.132076</td>\n",
+       "      <td>0.183379</td>\n",
+       "      <td>-0.044176</td>\n",
+       "      <td>0.170425</td>\n",
        "      <td>0.211477</td>\n",
        "      <td>-0.049492</td>\n",
        "      <td>1.000000</td>\n",
        "      <td>0.187134</td>\n",
        "      <td>0.164722</td>\n",
        "      <td>-0.051970</td>\n",
+       "      <td>-0.033609</td>\n",
        "      <td>-0.036502</td>\n",
        "      <td>0.080531</td>\n",
-       "      <td>0.104859</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>compress-ratio</th>\n",
        "      <td>-0.178515</td>\n",
        "      <td>-0.073362</td>\n",
        "      <td>0.249786</td>\n",
+       "      <td>0.158414</td>\n",
+       "      <td>0.181129</td>\n",
+       "      <td>0.261214</td>\n",
+       "      <td>0.151362</td>\n",
        "      <td>0.028971</td>\n",
        "      <td>0.008511</td>\n",
        "      <td>0.187134</td>\n",
        "      <td>1.000000</td>\n",
        "      <td>-0.202096</td>\n",
        "      <td>-0.436976</td>\n",
+       "      <td>0.324701</td>\n",
        "      <td>0.265201</td>\n",
        "      <td>0.064750</td>\n",
-       "      <td>0.233301</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>horsepower</th>\n",
        "      <td>0.070421</td>\n",
        "      <td>0.370384</td>\n",
        "      <td>0.301696</td>\n",
+       "      <td>0.521192</td>\n",
+       "      <td>0.596251</td>\n",
+       "      <td>-0.114968</td>\n",
+       "      <td>0.679865</td>\n",
        "      <td>0.742119</td>\n",
        "      <td>0.537543</td>\n",
        "      <td>0.164722</td>\n",
        "      <td>-0.202096</td>\n",
        "      <td>1.000000</td>\n",
        "      <td>0.171390</td>\n",
+       "      <td>-0.744246</td>\n",
        "      <td>-0.699361</td>\n",
        "      <td>0.725734</td>\n",
-       "      <td>0.396466</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>peak-rpm</th>\n",
        "      <td>0.273125</td>\n",
        "      <td>0.234384</td>\n",
        "      <td>-0.363355</td>\n",
+       "      <td>-0.279406</td>\n",
+       "      <td>-0.214240</td>\n",
+       "      <td>-0.322525</td>\n",
+       "      <td>-0.264976</td>\n",
        "      <td>-0.241031</td>\n",
        "      <td>-0.276942</td>\n",
        "      <td>-0.051970</td>\n",
        "      <td>-0.436976</td>\n",
        "      <td>0.171390</td>\n",
        "      <td>1.000000</td>\n",
+       "      <td>-0.117108</td>\n",
        "      <td>-0.053351</td>\n",
        "      <td>-0.084937</td>\n",
-       "      <td>-0.325444</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>city-mpg</th>\n",
+       "      <td>-0.035823</td>\n",
+       "      <td>-0.321647</td>\n",
+       "      <td>-0.470414</td>\n",
+       "      <td>-0.670909</td>\n",
+       "      <td>-0.642704</td>\n",
+       "      <td>-0.048640</td>\n",
+       "      <td>-0.757414</td>\n",
+       "      <td>-0.653658</td>\n",
+       "      <td>-0.556570</td>\n",
+       "      <td>-0.033609</td>\n",
+       "      <td>0.324701</td>\n",
+       "      <td>-0.744246</td>\n",
+       "      <td>-0.117108</td>\n",
+       "      <td>1.000000</td>\n",
+       "      <td>0.971337</td>\n",
+       "      <td>-0.693326</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>highway-mpg</th>\n",
        "      <td>0.034606</td>\n",
        "      <td>-0.270414</td>\n",
        "      <td>-0.544082</td>\n",
+       "      <td>-0.704662</td>\n",
+       "      <td>-0.677218</td>\n",
+       "      <td>-0.107358</td>\n",
+       "      <td>-0.797465</td>\n",
        "      <td>-0.677470</td>\n",
        "      <td>-0.562065</td>\n",
        "      <td>-0.036502</td>\n",
        "      <td>0.265201</td>\n",
        "      <td>-0.699361</td>\n",
        "      <td>-0.053351</td>\n",
+       "      <td>0.971337</td>\n",
        "      <td>1.000000</td>\n",
        "      <td>-0.700791</td>\n",
-       "      <td>-0.602410</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>price</th>\n",
        "      <td>-0.079290</td>\n",
        "      <td>0.313716</td>\n",
        "      <td>0.572348</td>\n",
+       "      <td>0.680804</td>\n",
+       "      <td>0.765788</td>\n",
+       "      <td>0.113942</td>\n",
+       "      <td>0.836802</td>\n",
        "      <td>0.871189</td>\n",
        "      <td>0.550994</td>\n",
        "      <td>0.080531</td>\n",
        "      <td>0.064750</td>\n",
        "      <td>0.725734</td>\n",
        "      <td>-0.084937</td>\n",
+       "      <td>-0.693326</td>\n",
        "      <td>-0.700791</td>\n",
        "      <td>1.000000</td>\n",
-       "      <td>0.622139</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>volume</th>\n",
-       "      <td>-0.456263</td>\n",
-       "      <td>0.036586</td>\n",
-       "      <td>0.913669</td>\n",
-       "      <td>0.594351</td>\n",
-       "      <td>0.549601</td>\n",
-       "      <td>0.104859</td>\n",
-       "      <td>0.233301</td>\n",
-       "      <td>0.396466</td>\n",
-       "      <td>-0.325444</td>\n",
-       "      <td>-0.602410</td>\n",
-       "      <td>0.622139</td>\n",
-       "      <td>1.000000</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "text/plain": [
-       "                   symboling  normalized-losses  wheel-base  engine-size  \\\n",
-       "symboling           1.000000           0.389010   -0.531954    -0.105790   \n",
-       "normalized-losses   0.389010           1.000000    0.008200     0.237114   \n",
-       "wheel-base         -0.531954           0.008200    1.000000     0.569329   \n",
-       "engine-size        -0.105790           0.237114    0.569329     1.000000   \n",
-       "bore               -0.160225           0.047391    0.495108     0.602516   \n",
-       "stroke             -0.020132           0.070319    0.164549     0.211477   \n",
-       "compress-ratio     -0.178515          -0.073362    0.249786     0.028971   \n",
-       "horsepower          0.070421           0.370384    0.301696     0.742119   \n",
-       "peak-rpm            0.273125           0.234384   -0.363355    -0.241031   \n",
-       "highway-mpg         0.034606          -0.270414   -0.544082    -0.677470   \n",
-       "price              -0.079290           0.313716    0.572348     0.871189   \n",
-       "volume             -0.456263           0.036586    0.913669     0.594351   \n",
+       "                   symboling  normalized-losses  wheel-base    length  \\\n",
+       "symboling           1.000000           0.389010   -0.531954 -0.357612   \n",
+       "normalized-losses   0.389010           1.000000    0.008200  0.152832   \n",
+       "wheel-base         -0.531954           0.008200    1.000000  0.874587   \n",
+       "length             -0.357612           0.152832    0.874587  1.000000   \n",
+       "width              -0.232919           0.200870    0.795144  0.841118   \n",
+       "height             -0.541038          -0.287893    0.589435  0.491029   \n",
+       "curb-weight        -0.227691           0.218460    0.776386  0.877728   \n",
+       "engine-size        -0.105790           0.237114    0.569329  0.683360   \n",
+       "bore               -0.160225           0.047391    0.495108  0.608905   \n",
+       "stroke             -0.020132           0.070319    0.164549  0.132076   \n",
+       "compress-ratio     -0.178515          -0.073362    0.249786  0.158414   \n",
+       "horsepower          0.070421           0.370384    0.301696  0.521192   \n",
+       "peak-rpm            0.273125           0.234384   -0.363355 -0.279406   \n",
+       "city-mpg           -0.035823          -0.321647   -0.470414 -0.670909   \n",
+       "highway-mpg         0.034606          -0.270414   -0.544082 -0.704662   \n",
+       "price              -0.079290           0.313716    0.572348  0.680804   \n",
        "\n",
-       "                       bore    stroke  compress-ratio  horsepower  peak-rpm  \\\n",
-       "symboling         -0.160225 -0.020132       -0.178515    0.070421  0.273125   \n",
-       "normalized-losses  0.047391  0.070319       -0.073362    0.370384  0.234384   \n",
-       "wheel-base         0.495108  0.164549        0.249786    0.301696 -0.363355   \n",
-       "engine-size        0.602516  0.211477        0.028971    0.742119 -0.241031   \n",
-       "bore               1.000000 -0.049492        0.008511    0.537543 -0.276942   \n",
-       "stroke            -0.049492  1.000000        0.187134    0.164722 -0.051970   \n",
-       "compress-ratio     0.008511  0.187134        1.000000   -0.202096 -0.436976   \n",
-       "horsepower         0.537543  0.164722       -0.202096    1.000000  0.171390   \n",
-       "peak-rpm          -0.276942 -0.051970       -0.436976    0.171390  1.000000   \n",
-       "highway-mpg       -0.562065 -0.036502        0.265201   -0.699361 -0.053351   \n",
-       "price              0.550994  0.080531        0.064750    0.725734 -0.084937   \n",
-       "volume             0.549601  0.104859        0.233301    0.396466 -0.325444   \n",
+       "                      width    height  curb-weight  engine-size      bore  \\\n",
+       "symboling         -0.232919 -0.541038    -0.227691    -0.105790 -0.160225   \n",
+       "normalized-losses  0.200870 -0.287893     0.218460     0.237114  0.047391   \n",
+       "wheel-base         0.795144  0.589435     0.776386     0.569329  0.495108   \n",
+       "length             0.841118  0.491029     0.877728     0.683360  0.608905   \n",
+       "width              1.000000  0.279210     0.867032     0.735433  0.556374   \n",
+       "height             0.279210  1.000000     0.295572     0.067149  0.199995   \n",
+       "curb-weight        0.867032  0.295572     1.000000     0.850594  0.648219   \n",
+       "engine-size        0.735433  0.067149     0.850594     1.000000  0.602516   \n",
+       "bore               0.556374  0.199995     0.648219     0.602516  1.000000   \n",
+       "stroke             0.183379 -0.044176     0.170425     0.211477 -0.049492   \n",
+       "compress-ratio     0.181129  0.261214     0.151362     0.028971  0.008511   \n",
+       "horsepower         0.596251 -0.114968     0.679865     0.742119  0.537543   \n",
+       "peak-rpm          -0.214240 -0.322525    -0.264976    -0.241031 -0.276942   \n",
+       "city-mpg          -0.642704 -0.048640    -0.757414    -0.653658 -0.556570   \n",
+       "highway-mpg       -0.677218 -0.107358    -0.797465    -0.677470 -0.562065   \n",
+       "price              0.765788  0.113942     0.836802     0.871189  0.550994   \n",
        "\n",
-       "                   highway-mpg     price    volume  \n",
-       "symboling             0.034606 -0.079290 -0.456263  \n",
-       "normalized-losses    -0.270414  0.313716  0.036586  \n",
-       "wheel-base           -0.544082  0.572348  0.913669  \n",
-       "engine-size          -0.677470  0.871189  0.594351  \n",
-       "bore                 -0.562065  0.550994  0.549601  \n",
-       "stroke               -0.036502  0.080531  0.104859  \n",
-       "compress-ratio        0.265201  0.064750  0.233301  \n",
-       "horsepower           -0.699361  0.725734  0.396466  \n",
-       "peak-rpm             -0.053351 -0.084937 -0.325444  \n",
-       "highway-mpg           1.000000 -0.700791 -0.602410  \n",
-       "price                -0.700791  1.000000  0.622139  \n",
-       "volume               -0.602410  0.622139  1.000000  "
+       "                     stroke  compress-ratio  horsepower  peak-rpm  city-mpg  \\\n",
+       "symboling         -0.020132       -0.178515    0.070421  0.273125 -0.035823   \n",
+       "normalized-losses  0.070319       -0.073362    0.370384  0.234384 -0.321647   \n",
+       "wheel-base         0.164549        0.249786    0.301696 -0.363355 -0.470414   \n",
+       "length             0.132076        0.158414    0.521192 -0.279406 -0.670909   \n",
+       "width              0.183379        0.181129    0.596251 -0.214240 -0.642704   \n",
+       "height            -0.044176        0.261214   -0.114968 -0.322525 -0.048640   \n",
+       "curb-weight        0.170425        0.151362    0.679865 -0.264976 -0.757414   \n",
+       "engine-size        0.211477        0.028971    0.742119 -0.241031 -0.653658   \n",
+       "bore              -0.049492        0.008511    0.537543 -0.276942 -0.556570   \n",
+       "stroke             1.000000        0.187134    0.164722 -0.051970 -0.033609   \n",
+       "compress-ratio     0.187134        1.000000   -0.202096 -0.436976  0.324701   \n",
+       "horsepower         0.164722       -0.202096    1.000000  0.171390 -0.744246   \n",
+       "peak-rpm          -0.051970       -0.436976    0.171390  1.000000 -0.117108   \n",
+       "city-mpg          -0.033609        0.324701   -0.744246 -0.117108  1.000000   \n",
+       "highway-mpg       -0.036502        0.265201   -0.699361 -0.053351  0.971337   \n",
+       "price              0.080531        0.064750    0.725734 -0.084937 -0.693326   \n",
+       "\n",
+       "                   highway-mpg     price  \n",
+       "symboling             0.034606 -0.079290  \n",
+       "normalized-losses    -0.270414  0.313716  \n",
+       "wheel-base           -0.544082  0.572348  \n",
+       "length               -0.704662  0.680804  \n",
+       "width                -0.677218  0.765788  \n",
+       "height               -0.107358  0.113942  \n",
+       "curb-weight          -0.797465  0.836802  \n",
+       "engine-size          -0.677470  0.871189  \n",
+       "bore                 -0.562065  0.550994  \n",
+       "stroke               -0.036502  0.080531  \n",
+       "compress-ratio        0.265201  0.064750  \n",
+       "horsepower           -0.699361  0.725734  \n",
+       "peak-rpm             -0.053351 -0.084937  \n",
+       "city-mpg              0.971337 -0.693326  \n",
+       "highway-mpg           1.000000 -0.700791  \n",
+       "price                -0.700791  1.000000  "
       ]
      },
-     "execution_count": 28,
+     "execution_count": 10,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -1471,7 +1631,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 30,
+   "execution_count": 11,
    "metadata": {},
    "outputs": [
     {
@@ -1498,15 +1658,19 @@
        "      <th>symboling</th>\n",
        "      <th>normalized-losses</th>\n",
        "      <th>wheel-base</th>\n",
+       "      <th>length</th>\n",
+       "      <th>width</th>\n",
+       "      <th>height</th>\n",
+       "      <th>curb-weight</th>\n",
        "      <th>engine-size</th>\n",
        "      <th>bore</th>\n",
        "      <th>stroke</th>\n",
        "      <th>compress-ratio</th>\n",
        "      <th>horsepower</th>\n",
        "      <th>peak-rpm</th>\n",
+       "      <th>city-mpg</th>\n",
        "      <th>highway-mpg</th>\n",
        "      <th>price</th>\n",
-       "      <th>volume</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
@@ -1515,45 +1679,133 @@
        "      <td>0.0</td>\n",
        "      <td>0.38901</td>\n",
        "      <td>-0.531954</td>\n",
+       "      <td>-0.357612</td>\n",
+       "      <td>-0.232919</td>\n",
+       "      <td>-0.541038</td>\n",
+       "      <td>-0.227691</td>\n",
        "      <td>-0.105790</td>\n",
        "      <td>-0.160225</td>\n",
        "      <td>-0.020132</td>\n",
        "      <td>-0.178515</td>\n",
        "      <td>0.070421</td>\n",
        "      <td>0.273125</td>\n",
+       "      <td>-0.035823</td>\n",
        "      <td>0.034606</td>\n",
        "      <td>-0.079290</td>\n",
-       "      <td>-0.456263</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>normalized-losses</th>\n",
        "      <td>0.0</td>\n",
        "      <td>0.00000</td>\n",
        "      <td>0.008200</td>\n",
+       "      <td>0.152832</td>\n",
+       "      <td>0.200870</td>\n",
+       "      <td>-0.287893</td>\n",
+       "      <td>0.218460</td>\n",
        "      <td>0.237114</td>\n",
        "      <td>0.047391</td>\n",
        "      <td>0.070319</td>\n",
        "      <td>-0.073362</td>\n",
        "      <td>0.370384</td>\n",
        "      <td>0.234384</td>\n",
+       "      <td>-0.321647</td>\n",
        "      <td>-0.270414</td>\n",
        "      <td>0.313716</td>\n",
-       "      <td>0.036586</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>wheel-base</th>\n",
        "      <td>-0.0</td>\n",
        "      <td>0.00000</td>\n",
        "      <td>0.000000</td>\n",
+       "      <td>0.874587</td>\n",
+       "      <td>0.795144</td>\n",
+       "      <td>0.589435</td>\n",
+       "      <td>0.776386</td>\n",
        "      <td>0.569329</td>\n",
        "      <td>0.495108</td>\n",
        "      <td>0.164549</td>\n",
        "      <td>0.249786</td>\n",
        "      <td>0.301696</td>\n",
        "      <td>-0.363355</td>\n",
+       "      <td>-0.470414</td>\n",
        "      <td>-0.544082</td>\n",
        "      <td>0.572348</td>\n",
-       "      <td>0.913669</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>length</th>\n",
+       "      <td>-0.0</td>\n",
+       "      <td>0.00000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.841118</td>\n",
+       "      <td>0.491029</td>\n",
+       "      <td>0.877728</td>\n",
+       "      <td>0.683360</td>\n",
+       "      <td>0.608905</td>\n",
+       "      <td>0.132076</td>\n",
+       "      <td>0.158414</td>\n",
+       "      <td>0.521192</td>\n",
+       "      <td>-0.279406</td>\n",
+       "      <td>-0.670909</td>\n",
+       "      <td>-0.704662</td>\n",
+       "      <td>0.680804</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>width</th>\n",
+       "      <td>-0.0</td>\n",
+       "      <td>0.00000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.279210</td>\n",
+       "      <td>0.867032</td>\n",
+       "      <td>0.735433</td>\n",
+       "      <td>0.556374</td>\n",
+       "      <td>0.183379</td>\n",
+       "      <td>0.181129</td>\n",
+       "      <td>0.596251</td>\n",
+       "      <td>-0.214240</td>\n",
+       "      <td>-0.642704</td>\n",
+       "      <td>-0.677218</td>\n",
+       "      <td>0.765788</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>height</th>\n",
+       "      <td>-0.0</td>\n",
+       "      <td>-0.00000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.295572</td>\n",
+       "      <td>0.067149</td>\n",
+       "      <td>0.199995</td>\n",
+       "      <td>-0.044176</td>\n",
+       "      <td>0.261214</td>\n",
+       "      <td>-0.114968</td>\n",
+       "      <td>-0.322525</td>\n",
+       "      <td>-0.048640</td>\n",
+       "      <td>-0.107358</td>\n",
+       "      <td>0.113942</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>curb-weight</th>\n",
+       "      <td>-0.0</td>\n",
+       "      <td>0.00000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.850594</td>\n",
+       "      <td>0.648219</td>\n",
+       "      <td>0.170425</td>\n",
+       "      <td>0.151362</td>\n",
+       "      <td>0.679865</td>\n",
+       "      <td>-0.264976</td>\n",
+       "      <td>-0.757414</td>\n",
+       "      <td>-0.797465</td>\n",
+       "      <td>0.836802</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>engine-size</th>\n",
@@ -1561,14 +1813,18 @@
        "      <td>0.00000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
        "      <td>0.602516</td>\n",
        "      <td>0.211477</td>\n",
        "      <td>0.028971</td>\n",
        "      <td>0.742119</td>\n",
        "      <td>-0.241031</td>\n",
+       "      <td>-0.653658</td>\n",
        "      <td>-0.677470</td>\n",
        "      <td>0.871189</td>\n",
-       "      <td>0.594351</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>bore</th>\n",
@@ -1577,13 +1833,17 @@
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
        "      <td>-0.049492</td>\n",
        "      <td>0.008511</td>\n",
        "      <td>0.537543</td>\n",
        "      <td>-0.276942</td>\n",
+       "      <td>-0.556570</td>\n",
        "      <td>-0.562065</td>\n",
        "      <td>0.550994</td>\n",
-       "      <td>0.549601</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>stroke</th>\n",
@@ -1591,14 +1851,18 @@
        "      <td>0.00000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
        "      <td>-0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.187134</td>\n",
        "      <td>0.164722</td>\n",
        "      <td>-0.051970</td>\n",
+       "      <td>-0.033609</td>\n",
        "      <td>-0.036502</td>\n",
        "      <td>0.080531</td>\n",
-       "      <td>0.104859</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>compress-ratio</th>\n",
@@ -1609,11 +1873,15 @@
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
        "      <td>-0.202096</td>\n",
        "      <td>-0.436976</td>\n",
+       "      <td>0.324701</td>\n",
        "      <td>0.265201</td>\n",
        "      <td>0.064750</td>\n",
-       "      <td>0.233301</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>horsepower</th>\n",
@@ -1622,13 +1890,17 @@
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>-0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.171390</td>\n",
+       "      <td>-0.744246</td>\n",
        "      <td>-0.699361</td>\n",
        "      <td>0.725734</td>\n",
-       "      <td>0.396466</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>peak-rpm</th>\n",
@@ -1639,11 +1911,34 @@
        "      <td>-0.000000</td>\n",
        "      <td>-0.000000</td>\n",
        "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
+       "      <td>-0.117108</td>\n",
        "      <td>-0.053351</td>\n",
        "      <td>-0.084937</td>\n",
-       "      <td>-0.325444</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>city-mpg</th>\n",
+       "      <td>-0.0</td>\n",
+       "      <td>-0.00000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.971337</td>\n",
+       "      <td>-0.693326</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>highway-mpg</th>\n",
@@ -1653,12 +1948,16 @@
        "      <td>-0.000000</td>\n",
        "      <td>-0.000000</td>\n",
        "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>-0.000000</td>\n",
        "      <td>-0.000000</td>\n",
        "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
        "      <td>-0.700791</td>\n",
-       "      <td>-0.602410</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>price</th>\n",
@@ -1670,24 +1969,13 @@
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
-       "      <td>-0.000000</td>\n",
-       "      <td>-0.000000</td>\n",
-       "      <td>0.000000</td>\n",
-       "      <td>0.622139</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>volume</th>\n",
-       "      <td>-0.0</td>\n",
-       "      <td>0.00000</td>\n",
-       "      <td>0.000000</td>\n",
-       "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>-0.000000</td>\n",
        "      <td>-0.000000</td>\n",
-       "      <td>0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
@@ -1695,50 +1983,80 @@
        "</div>"
       ],
       "text/plain": [
-       "                   symboling  normalized-losses  wheel-base  engine-size  \\\n",
-       "symboling                0.0            0.38901   -0.531954    -0.105790   \n",
-       "normalized-losses        0.0            0.00000    0.008200     0.237114   \n",
-       "wheel-base              -0.0            0.00000    0.000000     0.569329   \n",
-       "engine-size             -0.0            0.00000    0.000000     0.000000   \n",
-       "bore                    -0.0            0.00000    0.000000     0.000000   \n",
-       "stroke                  -0.0            0.00000    0.000000     0.000000   \n",
-       "compress-ratio          -0.0           -0.00000    0.000000     0.000000   \n",
-       "horsepower               0.0            0.00000    0.000000     0.000000   \n",
-       "peak-rpm                 0.0            0.00000   -0.000000    -0.000000   \n",
-       "highway-mpg              0.0           -0.00000   -0.000000    -0.000000   \n",
-       "price                   -0.0            0.00000    0.000000     0.000000   \n",
-       "volume                  -0.0            0.00000    0.000000     0.000000   \n",
+       "                   symboling  normalized-losses  wheel-base    length  \\\n",
+       "symboling                0.0            0.38901   -0.531954 -0.357612   \n",
+       "normalized-losses        0.0            0.00000    0.008200  0.152832   \n",
+       "wheel-base              -0.0            0.00000    0.000000  0.874587   \n",
+       "length                  -0.0            0.00000    0.000000  0.000000   \n",
+       "width                   -0.0            0.00000    0.000000  0.000000   \n",
+       "height                  -0.0           -0.00000    0.000000  0.000000   \n",
+       "curb-weight             -0.0            0.00000    0.000000  0.000000   \n",
+       "engine-size             -0.0            0.00000    0.000000  0.000000   \n",
+       "bore                    -0.0            0.00000    0.000000  0.000000   \n",
+       "stroke                  -0.0            0.00000    0.000000  0.000000   \n",
+       "compress-ratio          -0.0           -0.00000    0.000000  0.000000   \n",
+       "horsepower               0.0            0.00000    0.000000  0.000000   \n",
+       "peak-rpm                 0.0            0.00000   -0.000000 -0.000000   \n",
+       "city-mpg                -0.0           -0.00000   -0.000000 -0.000000   \n",
+       "highway-mpg              0.0           -0.00000   -0.000000 -0.000000   \n",
+       "price                   -0.0            0.00000    0.000000  0.000000   \n",
+       "\n",
+       "                      width    height  curb-weight  engine-size      bore  \\\n",
+       "symboling         -0.232919 -0.541038    -0.227691    -0.105790 -0.160225   \n",
+       "normalized-losses  0.200870 -0.287893     0.218460     0.237114  0.047391   \n",
+       "wheel-base         0.795144  0.589435     0.776386     0.569329  0.495108   \n",
+       "length             0.841118  0.491029     0.877728     0.683360  0.608905   \n",
+       "width              0.000000  0.279210     0.867032     0.735433  0.556374   \n",
+       "height             0.000000  0.000000     0.295572     0.067149  0.199995   \n",
+       "curb-weight        0.000000  0.000000     0.000000     0.850594  0.648219   \n",
+       "engine-size        0.000000  0.000000     0.000000     0.000000  0.602516   \n",
+       "bore               0.000000  0.000000     0.000000     0.000000  0.000000   \n",
+       "stroke             0.000000 -0.000000     0.000000     0.000000 -0.000000   \n",
+       "compress-ratio     0.000000  0.000000     0.000000     0.000000  0.000000   \n",
+       "horsepower         0.000000 -0.000000     0.000000     0.000000  0.000000   \n",
+       "peak-rpm          -0.000000 -0.000000    -0.000000    -0.000000 -0.000000   \n",
+       "city-mpg          -0.000000 -0.000000    -0.000000    -0.000000 -0.000000   \n",
+       "highway-mpg       -0.000000 -0.000000    -0.000000    -0.000000 -0.000000   \n",
+       "price              0.000000  0.000000     0.000000     0.000000  0.000000   \n",
        "\n",
-       "                       bore    stroke  compress-ratio  horsepower  peak-rpm  \\\n",
-       "symboling         -0.160225 -0.020132       -0.178515    0.070421  0.273125   \n",
-       "normalized-losses  0.047391  0.070319       -0.073362    0.370384  0.234384   \n",
-       "wheel-base         0.495108  0.164549        0.249786    0.301696 -0.363355   \n",
-       "engine-size        0.602516  0.211477        0.028971    0.742119 -0.241031   \n",
-       "bore               0.000000 -0.049492        0.008511    0.537543 -0.276942   \n",
-       "stroke            -0.000000  0.000000        0.187134    0.164722 -0.051970   \n",
-       "compress-ratio     0.000000  0.000000        0.000000   -0.202096 -0.436976   \n",
-       "horsepower         0.000000  0.000000       -0.000000    0.000000  0.171390   \n",
-       "peak-rpm          -0.000000 -0.000000       -0.000000    0.000000  0.000000   \n",
-       "highway-mpg       -0.000000 -0.000000        0.000000   -0.000000 -0.000000   \n",
-       "price              0.000000  0.000000        0.000000    0.000000 -0.000000   \n",
-       "volume             0.000000  0.000000        0.000000    0.000000 -0.000000   \n",
+       "                     stroke  compress-ratio  horsepower  peak-rpm  city-mpg  \\\n",
+       "symboling         -0.020132       -0.178515    0.070421  0.273125 -0.035823   \n",
+       "normalized-losses  0.070319       -0.073362    0.370384  0.234384 -0.321647   \n",
+       "wheel-base         0.164549        0.249786    0.301696 -0.363355 -0.470414   \n",
+       "length             0.132076        0.158414    0.521192 -0.279406 -0.670909   \n",
+       "width              0.183379        0.181129    0.596251 -0.214240 -0.642704   \n",
+       "height            -0.044176        0.261214   -0.114968 -0.322525 -0.048640   \n",
+       "curb-weight        0.170425        0.151362    0.679865 -0.264976 -0.757414   \n",
+       "engine-size        0.211477        0.028971    0.742119 -0.241031 -0.653658   \n",
+       "bore              -0.049492        0.008511    0.537543 -0.276942 -0.556570   \n",
+       "stroke             0.000000        0.187134    0.164722 -0.051970 -0.033609   \n",
+       "compress-ratio     0.000000        0.000000   -0.202096 -0.436976  0.324701   \n",
+       "horsepower         0.000000       -0.000000    0.000000  0.171390 -0.744246   \n",
+       "peak-rpm          -0.000000       -0.000000    0.000000  0.000000 -0.117108   \n",
+       "city-mpg          -0.000000        0.000000   -0.000000 -0.000000  0.000000   \n",
+       "highway-mpg       -0.000000        0.000000   -0.000000 -0.000000  0.000000   \n",
+       "price              0.000000        0.000000    0.000000 -0.000000 -0.000000   \n",
        "\n",
-       "                   highway-mpg     price    volume  \n",
-       "symboling             0.034606 -0.079290 -0.456263  \n",
-       "normalized-losses    -0.270414  0.313716  0.036586  \n",
-       "wheel-base           -0.544082  0.572348  0.913669  \n",
-       "engine-size          -0.677470  0.871189  0.594351  \n",
-       "bore                 -0.562065  0.550994  0.549601  \n",
-       "stroke               -0.036502  0.080531  0.104859  \n",
-       "compress-ratio        0.265201  0.064750  0.233301  \n",
-       "horsepower           -0.699361  0.725734  0.396466  \n",
-       "peak-rpm             -0.053351 -0.084937 -0.325444  \n",
-       "highway-mpg           0.000000 -0.700791 -0.602410  \n",
-       "price                -0.000000  0.000000  0.622139  \n",
-       "volume               -0.000000  0.000000  0.000000  "
+       "                   highway-mpg     price  \n",
+       "symboling             0.034606 -0.079290  \n",
+       "normalized-losses    -0.270414  0.313716  \n",
+       "wheel-base           -0.544082  0.572348  \n",
+       "length               -0.704662  0.680804  \n",
+       "width                -0.677218  0.765788  \n",
+       "height               -0.107358  0.113942  \n",
+       "curb-weight          -0.797465  0.836802  \n",
+       "engine-size          -0.677470  0.871189  \n",
+       "bore                 -0.562065  0.550994  \n",
+       "stroke               -0.036502  0.080531  \n",
+       "compress-ratio        0.265201  0.064750  \n",
+       "horsepower           -0.699361  0.725734  \n",
+       "peak-rpm             -0.053351 -0.084937  \n",
+       "city-mpg              0.971337 -0.693326  \n",
+       "highway-mpg           0.000000 -0.700791  \n",
+       "price                -0.000000  0.000000  "
       ]
      },
-     "execution_count": 30,
+     "execution_count": 11,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -1840,7 +2158,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 16,
+   "execution_count": 13,
    "metadata": {},
    "outputs": [
     {
@@ -2308,7 +2626,7 @@
        "[256 rows x 3 columns]"
       ]
      },
-     "execution_count": 16,
+     "execution_count": 13,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -2322,7 +2640,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 19,
+   "execution_count": 14,
    "metadata": {},
    "outputs": [
     {
@@ -2430,7 +2748,7 @@
        "9   wheel-base        width     0.795144"
       ]
      },
-     "execution_count": 19,
+     "execution_count": 14,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -2459,7 +2777,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 20,
+   "execution_count": 15,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -2472,7 +2790,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 22,
+   "execution_count": 16,
    "metadata": {},
    "outputs": [
     {
@@ -2486,7 +2804,7 @@
        "      dtype='object')"
       ]
      },
-     "execution_count": 22,
+     "execution_count": 16,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -2504,7 +2822,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 33,
+   "execution_count": 17,
    "metadata": {},
    "outputs": [
     {
@@ -2541,16 +2859,16 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 34,
+   "execution_count": 18,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
-       "<seaborn.axisgrid.PairGrid at 0x25cce5c3da0>"
+       "<seaborn.axisgrid.PairGrid at 0x1731f2d8ac8>"
       ]
      },
-     "execution_count": 34,
+     "execution_count": 18,
      "metadata": {},
      "output_type": "execute_result"
     },
@@ -2581,22 +2899,22 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 35,
+   "execution_count": 19,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
-       "<seaborn.axisgrid.FacetGrid at 0x25cd481ab00>"
+       "<seaborn.axisgrid.FacetGrid at 0x1731f283828>"
       ]
      },
-     "execution_count": 35,
+     "execution_count": 19,
      "metadata": {},
      "output_type": "execute_result"
     },
     {
      "data": {
-      "image/png": "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\n",
+      "image/png": "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\n",
       "text/plain": [
        "<Figure size 720x720 with 4 Axes>"
       ]
@@ -2617,6 +2935,865 @@
     "#可以看到价格和马力之间的关系还是很强的"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### 预处理\n",
+    "如果一个特征的方差比其它的大很多，那么它可能支配目标函数，使估计者不能像预期的那样正确的从其它特征中学习。这就是为什么要坐数据缩放"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "对连续值进行标准化，让数据落在大致的区域范围内，而不是差异非常大\n",
+    "<br>\n",
+    "你也可以尝试某个值不做归一化，后续的特征重要性可以看到不做归一化的重要性极高"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 71,
+   "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>symboling</th>\n",
+       "      <th>normalized-losses</th>\n",
+       "      <th>make</th>\n",
+       "      <th>fuel-type</th>\n",
+       "      <th>aspiration</th>\n",
+       "      <th>num-of-doors</th>\n",
+       "      <th>body-style</th>\n",
+       "      <th>drive-wheels</th>\n",
+       "      <th>engine-location</th>\n",
+       "      <th>wheel-base</th>\n",
+       "      <th>...</th>\n",
+       "      <th>engine-size</th>\n",
+       "      <th>fuel-system</th>\n",
+       "      <th>bore</th>\n",
+       "      <th>stroke</th>\n",
+       "      <th>compress-ratio</th>\n",
+       "      <th>horsepower</th>\n",
+       "      <th>peak-rpm</th>\n",
+       "      <th>highway-mpg</th>\n",
+       "      <th>output</th>\n",
+       "      <th>volume</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>1.743470</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>alfa-romero</td>\n",
+       "      <td>gas</td>\n",
+       "      <td>std</td>\n",
+       "      <td>two</td>\n",
+       "      <td>convertible</td>\n",
+       "      <td>rwd</td>\n",
+       "      <td>front</td>\n",
+       "      <td>-1.690772</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0.074449</td>\n",
+       "      <td>mpfi</td>\n",
+       "      <td>0.532789</td>\n",
+       "      <td>-1.830840</td>\n",
+       "      <td>-0.288349</td>\n",
+       "      <td>0.114182</td>\n",
+       "      <td>-0.27402</td>\n",
+       "      <td>-0.546059</td>\n",
+       "      <td>no</td>\n",
+       "      <td>-1.144195</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1.743470</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>alfa-romero</td>\n",
+       "      <td>gas</td>\n",
+       "      <td>std</td>\n",
+       "      <td>two</td>\n",
+       "      <td>convertible</td>\n",
+       "      <td>rwd</td>\n",
+       "      <td>front</td>\n",
+       "      <td>-1.690772</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0.074449</td>\n",
+       "      <td>mpfi</td>\n",
+       "      <td>0.532789</td>\n",
+       "      <td>-1.830840</td>\n",
+       "      <td>-0.288349</td>\n",
+       "      <td>0.114182</td>\n",
+       "      <td>-0.27402</td>\n",
+       "      <td>-0.546059</td>\n",
+       "      <td>no</td>\n",
+       "      <td>-1.144195</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>0.133509</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>alfa-romero</td>\n",
+       "      <td>gas</td>\n",
+       "      <td>std</td>\n",
+       "      <td>two</td>\n",
+       "      <td>hatchback</td>\n",
+       "      <td>rwd</td>\n",
+       "      <td>front</td>\n",
+       "      <td>-0.708596</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0.604046</td>\n",
+       "      <td>mpfi</td>\n",
+       "      <td>-2.367552</td>\n",
+       "      <td>0.691744</td>\n",
+       "      <td>-0.288349</td>\n",
+       "      <td>1.105818</td>\n",
+       "      <td>-0.27402</td>\n",
+       "      <td>-0.691627</td>\n",
+       "      <td>no</td>\n",
+       "      <td>-0.392670</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>0.938490</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>audi</td>\n",
+       "      <td>gas</td>\n",
+       "      <td>std</td>\n",
+       "      <td>four</td>\n",
+       "      <td>sedan</td>\n",
+       "      <td>fwd</td>\n",
+       "      <td>front</td>\n",
+       "      <td>0.173698</td>\n",
+       "      <td>...</td>\n",
+       "      <td>-0.431076</td>\n",
+       "      <td>mpfi</td>\n",
+       "      <td>-0.495180</td>\n",
+       "      <td>0.468224</td>\n",
+       "      <td>-0.035973</td>\n",
+       "      <td>-0.093370</td>\n",
+       "      <td>0.76817</td>\n",
+       "      <td>-0.109354</td>\n",
+       "      <td>no</td>\n",
+       "      <td>0.203076</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>0.938490</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>audi</td>\n",
+       "      <td>gas</td>\n",
+       "      <td>std</td>\n",
+       "      <td>four</td>\n",
+       "      <td>sedan</td>\n",
+       "      <td>4wd</td>\n",
+       "      <td>front</td>\n",
+       "      <td>0.107110</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0.218885</td>\n",
+       "      <td>mpfi</td>\n",
+       "      <td>-0.495180</td>\n",
+       "      <td>0.468224</td>\n",
+       "      <td>-0.540725</td>\n",
+       "      <td>0.206427</td>\n",
+       "      <td>0.76817</td>\n",
+       "      <td>-1.273900</td>\n",
+       "      <td>no</td>\n",
+       "      <td>0.227271</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>5 rows × 22 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   symboling  normalized-losses         make fuel-type aspiration  \\\n",
+       "0   1.743470            0.89088  alfa-romero       gas        std   \n",
+       "1   1.743470            0.89088  alfa-romero       gas        std   \n",
+       "2   0.133509            0.89088  alfa-romero       gas        std   \n",
+       "3   0.938490            0.89088         audi       gas        std   \n",
+       "4   0.938490            0.89088         audi       gas        std   \n",
+       "\n",
+       "  num-of-doors   body-style drive-wheels engine-location  wheel-base  ...  \\\n",
+       "0          two  convertible          rwd           front   -1.690772  ...   \n",
+       "1          two  convertible          rwd           front   -1.690772  ...   \n",
+       "2          two    hatchback          rwd           front   -0.708596  ...   \n",
+       "3         four        sedan          fwd           front    0.173698  ...   \n",
+       "4         four        sedan          4wd           front    0.107110  ...   \n",
+       "\n",
+       "  engine-size fuel-system      bore    stroke  compress-ratio  horsepower  \\\n",
+       "0    0.074449        mpfi  0.532789 -1.830840       -0.288349    0.114182   \n",
+       "1    0.074449        mpfi  0.532789 -1.830840       -0.288349    0.114182   \n",
+       "2    0.604046        mpfi -2.367552  0.691744       -0.288349    1.105818   \n",
+       "3   -0.431076        mpfi -0.495180  0.468224       -0.035973   -0.093370   \n",
+       "4    0.218885        mpfi -0.495180  0.468224       -0.540725    0.206427   \n",
+       "\n",
+       "   peak-rpm  highway-mpg  output    volume  \n",
+       "0  -0.27402    -0.546059      no -1.144195  \n",
+       "1  -0.27402    -0.546059      no -1.144195  \n",
+       "2  -0.27402    -0.691627      no -0.392670  \n",
+       "3   0.76817    -0.109354      no  0.203076  \n",
+       "4   0.76817    -1.273900      no  0.227271  \n",
+       "\n",
+       "[5 rows x 22 columns]"
+      ]
+     },
+     "execution_count": 71,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "target = data.price  #预测标签\n",
+    "\n",
+    "regressors = [x for x in data.columns if x not in ['price']]  #训练数据\n",
+    "features = data.loc[:, regressors]\n",
+    "\n",
+    "num = ['symboling','normalized-losses','volume',\n",
+    "       'horsepower','wheel-base','bore','engine-size',\n",
+    "       'stroke','compress-ratio','peak-rpm','highway-mpg']  #获取连续值特征列\n",
+    "\n",
+    "#标准化\n",
+    "standard_scaler = StandardScaler()\n",
+    "features[num] = standard_scaler.fit_transform(features[num])\n",
+    "\n",
+    "features.head()  #数值已经标准化"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 72,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# features['highway-mpg']"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "**对分类属性进行one-hot编码**\n",
+    "<br>\n",
+    "比如fuel-type里面只有gas和diesel，机器是不认识的需要转化成0/1类型的"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 73,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "In total: (205, 73)\n"
+     ]
+    },
+    {
+     "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>symboling</th>\n",
+       "      <th>normalized-losses</th>\n",
+       "      <th>wheel-base</th>\n",
+       "      <th>engine-size</th>\n",
+       "      <th>bore</th>\n",
+       "      <th>stroke</th>\n",
+       "      <th>compress-ratio</th>\n",
+       "      <th>horsepower</th>\n",
+       "      <th>peak-rpm</th>\n",
+       "      <th>highway-mpg</th>\n",
+       "      <th>...</th>\n",
+       "      <th>num-of-cylinders_twelve</th>\n",
+       "      <th>num-of-cylinders_two</th>\n",
+       "      <th>fuel-system_1bbl</th>\n",
+       "      <th>fuel-system_2bbl</th>\n",
+       "      <th>fuel-system_4bbl</th>\n",
+       "      <th>fuel-system_idi</th>\n",
+       "      <th>fuel-system_mfi</th>\n",
+       "      <th>fuel-system_mpfi</th>\n",
+       "      <th>fuel-system_spdi</th>\n",
+       "      <th>fuel-system_spfi</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>1.743470</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>-1.690772</td>\n",
+       "      <td>0.074449</td>\n",
+       "      <td>0.532789</td>\n",
+       "      <td>-1.830840</td>\n",
+       "      <td>-0.288349</td>\n",
+       "      <td>0.114182</td>\n",
+       "      <td>-0.27402</td>\n",
+       "      <td>-0.546059</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>1</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1.743470</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>-1.690772</td>\n",
+       "      <td>0.074449</td>\n",
+       "      <td>0.532789</td>\n",
+       "      <td>-1.830840</td>\n",
+       "      <td>-0.288349</td>\n",
+       "      <td>0.114182</td>\n",
+       "      <td>-0.27402</td>\n",
+       "      <td>-0.546059</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>1</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>0.133509</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>-0.708596</td>\n",
+       "      <td>0.604046</td>\n",
+       "      <td>-2.367552</td>\n",
+       "      <td>0.691744</td>\n",
+       "      <td>-0.288349</td>\n",
+       "      <td>1.105818</td>\n",
+       "      <td>-0.27402</td>\n",
+       "      <td>-0.691627</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>1</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>0.938490</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>0.173698</td>\n",
+       "      <td>-0.431076</td>\n",
+       "      <td>-0.495180</td>\n",
+       "      <td>0.468224</td>\n",
+       "      <td>-0.035973</td>\n",
+       "      <td>-0.093370</td>\n",
+       "      <td>0.76817</td>\n",
+       "      <td>-0.109354</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>1</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>0.938490</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>0.107110</td>\n",
+       "      <td>0.218885</td>\n",
+       "      <td>-0.495180</td>\n",
+       "      <td>0.468224</td>\n",
+       "      <td>-0.540725</td>\n",
+       "      <td>0.206427</td>\n",
+       "      <td>0.76817</td>\n",
+       "      <td>-1.273900</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>1</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>5 rows × 73 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   symboling  normalized-losses  wheel-base  engine-size      bore    stroke  \\\n",
+       "0   1.743470            0.89088   -1.690772     0.074449  0.532789 -1.830840   \n",
+       "1   1.743470            0.89088   -1.690772     0.074449  0.532789 -1.830840   \n",
+       "2   0.133509            0.89088   -0.708596     0.604046 -2.367552  0.691744   \n",
+       "3   0.938490            0.89088    0.173698    -0.431076 -0.495180  0.468224   \n",
+       "4   0.938490            0.89088    0.107110     0.218885 -0.495180  0.468224   \n",
+       "\n",
+       "   compress-ratio  horsepower  peak-rpm  highway-mpg  ...  \\\n",
+       "0       -0.288349    0.114182  -0.27402    -0.546059  ...   \n",
+       "1       -0.288349    0.114182  -0.27402    -0.546059  ...   \n",
+       "2       -0.288349    1.105818  -0.27402    -0.691627  ...   \n",
+       "3       -0.035973   -0.093370   0.76817    -0.109354  ...   \n",
+       "4       -0.540725    0.206427   0.76817    -1.273900  ...   \n",
+       "\n",
+       "   num-of-cylinders_twelve  num-of-cylinders_two  fuel-system_1bbl  \\\n",
+       "0                        0                     0                 0   \n",
+       "1                        0                     0                 0   \n",
+       "2                        0                     0                 0   \n",
+       "3                        0                     0                 0   \n",
+       "4                        0                     0                 0   \n",
+       "\n",
+       "   fuel-system_2bbl  fuel-system_4bbl  fuel-system_idi  fuel-system_mfi  \\\n",
+       "0                 0                 0                0                0   \n",
+       "1                 0                 0                0                0   \n",
+       "2                 0                 0                0                0   \n",
+       "3                 0                 0                0                0   \n",
+       "4                 0                 0                0                0   \n",
+       "\n",
+       "   fuel-system_mpfi  fuel-system_spdi  fuel-system_spfi  \n",
+       "0                 1                 0                 0  \n",
+       "1                 1                 0                 0  \n",
+       "2                 1                 0                 0  \n",
+       "3                 1                 0                 0  \n",
+       "4                 1                 0                 0  \n",
+       "\n",
+       "[5 rows x 73 columns]"
+      ]
+     },
+     "execution_count": 73,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "classes = ['make','fuel-type','aspiration','engine-type',\n",
+    "           'num-of-doors','body-style','drive-wheels','output',\n",
+    "           'engine-location','num-of-cylinders','fuel-system']  #获取所需特征列\n",
+    "\n",
+    "dummies = pd.get_dummies(features[classes])  #进行one-hot\n",
+    "#加入one-hot后的数值，并去掉原有的字符串类型的数据\n",
+    "features = features.join(dummies).drop(classes, axis=1)\n",
+    "\n",
+    "print('In total:', features.shape)\n",
+    "features.head(5)  #可以看到结果已经没有字符串类型的数据了，且新增的只有0或者1"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "**划分数据集**\n",
+    "划分成训练集和验证集，训练出来的结果在验证集上测试效果"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 74,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Train (143, 73) Test: (62, 73)\n"
+     ]
+    }
+   ],
+   "source": [
+    "# tain为训练集，test为测试集\n",
+    "#test_size切分比例，训练占总数据0.7，测试0.3，random_state随机切分的种子\n",
+    "X_train, X_test, y_train, y_test = train_test_split(features,target,\n",
+    "                                                   test_size=0.3,\n",
+    "                                                   random_state=seed)\n",
+    "print(\"Train\", X_train.shape, \"Test:\", X_test.shape)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### 回归求解\n",
+    "Lass回归，多加了一个绝对值项来惩罚过大的系数，alphas=0就是最小二乘"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "为什么需要惩罚项：\n",
+    "<br>\n",
+    "假设有数据`x=[1,1,1]`,` w1[1/3,1/3,1/3]`,`w2[1/3,0,0]`\n",
+    "<br>\n",
+    "计算x和w1，数据分别相乘并相加，结果为1，\n",
+    "<br>\n",
+    "而x和w2只有第一个数据点和w1相同，后面两个为0，结果为1/3；\n",
+    "<br>\n",
+    "<br>\n",
+    "x和w1的3个数据点结果平稳，而x和w2的结果非常不平稳，\n",
+    "<br>\n",
+    "我们希望的结果是x和w1这种每个结果数据点都比较稳定的。"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "**惩罚项**\n",
+    "因为0乘以任何结果都为0，我们引入惩罚项：\n",
+    "<br>\n",
+    "原本我们的计算是 真实的y - 预测的y得到值越小越好\n",
+    "<br>\n",
+    "现在是loss = (真实的y - 预测的y) + λ(w) 越小越好"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 90,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "CV results: 0.9297963022389589 53.35375452969131\n"
+     ]
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 720x288 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "alphas = 3. ** np.arange(2,12)  #自选惩罚项参数，现在是手动调，也有自动包\n",
+    "scores = np.empty_like(alphas)  #计算不同参数的不同得分\n",
+    "\n",
+    "#第一种方法\n",
+    "for i, a in enumerate(alphas):\n",
+    "    lasso = Lasso(random_state=seed)  #指定Lasso模型\n",
+    "    lasso.set_params(alpha=a)  #指定惩罚力度\n",
+    "    lasso.fit(X_train, y_train)  #训练\n",
+    "    scores[i] = lasso.score(X_test, y_test)  #预测\n",
+    "    \n",
+    "#第二种方法，交叉验证，训练集再切分几分交叉训练并自验证，最后再拿预测集预测\n",
+    " #cv=10，平均切分10次，即10份，每次训练9份拿另外的1份验证\n",
+    "lassocv = LassoCV(cv=10, random_state=seed)\n",
+    "lassocv.fit(features, target)  #训练\n",
+    "lassocv_score = lassocv.score(features, target)  #预测\n",
+    "lassocv_alpha = lassocv.alpha_  #获取alpha值\n",
+    "\n",
+    "plt.figure(figsize=(10,4))\n",
+    "plt.plot(alphas, scores, '-ko')\n",
+    "plt.axhline(lassocv_score, color=c)\n",
+    "plt.xlabel(r'$\\alpha$')\n",
+    "plt.ylabel('CV score')\n",
+    "plt.xscale('log', basex=2)\n",
+    "sns.despine(offset=15)\n",
+    "\n",
+    "#查看不同alpha值下的得分\n",
+    "print('CV results:', lassocv_score, lassocv_alpha)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "**查看特征重要性**"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 76,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Lasso picked28features and eliminated the other45features.\n"
+     ]
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 720x288 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "coefs = pd.Series(lassocv.coef_, index=features.columns)\n",
+    "\n",
+    "print(\"Lasso picked\" + str(sum(coefs != 0)) + \\\n",
+    "      \"features and eliminated the other\"+ \\\n",
+    "     str(sum(coefs == 0)) + \"features.\")\n",
+    "\n",
+    "coefs = pd.concat([coefs.sort_values().head(5), coefs.sort_values().tail(5)])\n",
+    "\n",
+    "plt.figure(figsize=(10,4))\n",
+    "coefs.plot(kind=\"barh\", color=c)\n",
+    "plt.title(\"Coefficients\")\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 85,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "0.870697563184287"
+      ]
+     },
+     "execution_count": 85,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "#进行训练和预测，并检验得分，得分越高表明模型越好\n",
+    "#alpha选取我们手动找到的\n",
+    "model_l1 = LassoCV(alphas=alphas,cv=10,random_state=seed).fit(X_train,y_train)\n",
+    "y_pred_l1 = model_l1.predict(X_test)\n",
+    "\n",
+    "model_l1.score(X_test,y_test)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "**画出残差点**\n",
+    "<br>\n",
+    "预测值和真实值之间的差异，y轴真实值和预测值之间的差异，x轴为预测值\n",
+    "<br>\n",
+    "差异在0左右是最好的"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 93,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<matplotlib.axes._subplots.AxesSubplot at 0x1733cb98828>"
+      ]
+     },
+     "execution_count": 93,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 432x432 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plt.rcParams['figure.figsize'] = (6.0,6.0)\n",
+    "\n",
+    "preds = pd.DataFrame({'preds':model_l1.predict(X_train), 'true':y_train})\n",
+    "preds['residuals'] = preds['true'] - preds['preds']\n",
+    "preds.plot(x='preds', y='residuals', kind='scatter', color=c)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "常用的评测方法：\n",
+    "<br>\n",
+    "MSE均方误差（越小越好）\n",
+    "<br>R2决定系数（0-1之间，越大表示模型拟合效果越好）"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 94,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "MSE: 4232851.141\n",
+      "R2: 0.871\n"
+     ]
+    }
+   ],
+   "source": [
+    "def MSE(y_true, y_pred):\n",
+    "    mse = mean_squared_error(y_true, y_pred)\n",
+    "    print(\"MSE: %2.3f\" % mse)\n",
+    "    return mse\n",
+    "\n",
+    "def R2(y_true, y_pred):\n",
+    "    r2 = r2_score(y_true, y_pred)\n",
+    "    print(\"R2: %2.3f\" % r2)\n",
+    "    return r2\n",
+    "\n",
+    "MSE(y_test, y_pred_l1);R2(y_test, y_pred_l1);"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 95,
+   "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>true</th>\n",
+       "      <th>predicted</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>9279</td>\n",
+       "      <td>8053.973904</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>35056</td>\n",
+       "      <td>32509.124398</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>17075</td>\n",
+       "      <td>19477.432690</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>7898</td>\n",
+       "      <td>6878.526602</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>8058</td>\n",
+       "      <td>7740.960545</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    true     predicted\n",
+       "0   9279   8053.973904\n",
+       "1  35056  32509.124398\n",
+       "2  17075  19477.432690\n",
+       "3   7898   6878.526602\n",
+       "4   8058   7740.960545"
+      ]
+     },
+     "execution_count": 95,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# 打印真实值和预测值\n",
+    "p = {'true':list(y_test),\n",
+    "    'predicted':pd.Series(y_pred_l1)\n",
+    "    }\n",
+    "pd.DataFrame(p).head()"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
diff --git a/notebook_必备数学基础/案例：汽车价格预测任务.ipynb b/notebook_必备数学基础/案例：汽车价格预测任务.ipynb
index f22441e..380ab11 100644
--- a/notebook_必备数学基础/案例：汽车价格预测任务.ipynb
+++ b/notebook_必备数学基础/案例：汽车价格预测任务.ipynb
@@ -78,6 +78,8 @@
     "import numpy as np\n",
     "import pandas as pd\n",
     "from pandas import datetime\n",
+    "import warnings  # 忽略普通警告，不打印太多东西\n",
+    "warnings.filterwarnings('ignore')\n",
     "\n",
     "#data visualization and missing values\n",
     "import matplotlib.pyplot as plt\n",
@@ -639,7 +641,7 @@
     {
      "data": {
       "text/plain": [
-       "<matplotlib.axes._subplots.AxesSubplot at 0x25cc901b3c8>"
+       "<matplotlib.axes._subplots.AxesSubplot at 0x1731f11d320>"
       ]
      },
      "execution_count": 5,
@@ -1185,7 +1187,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 28,
+   "execution_count": 10,
    "metadata": {},
    "outputs": [
     {
@@ -1212,15 +1214,19 @@
        "      <th>symboling</th>\n",
        "      <th>normalized-losses</th>\n",
        "      <th>wheel-base</th>\n",
+       "      <th>length</th>\n",
+       "      <th>width</th>\n",
+       "      <th>height</th>\n",
+       "      <th>curb-weight</th>\n",
        "      <th>engine-size</th>\n",
        "      <th>bore</th>\n",
        "      <th>stroke</th>\n",
        "      <th>compress-ratio</th>\n",
        "      <th>horsepower</th>\n",
        "      <th>peak-rpm</th>\n",
+       "      <th>city-mpg</th>\n",
        "      <th>highway-mpg</th>\n",
        "      <th>price</th>\n",
-       "      <th>volume</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
@@ -1229,230 +1235,384 @@
        "      <td>1.000000</td>\n",
        "      <td>0.389010</td>\n",
        "      <td>-0.531954</td>\n",
+       "      <td>-0.357612</td>\n",
+       "      <td>-0.232919</td>\n",
+       "      <td>-0.541038</td>\n",
+       "      <td>-0.227691</td>\n",
        "      <td>-0.105790</td>\n",
        "      <td>-0.160225</td>\n",
        "      <td>-0.020132</td>\n",
        "      <td>-0.178515</td>\n",
        "      <td>0.070421</td>\n",
        "      <td>0.273125</td>\n",
+       "      <td>-0.035823</td>\n",
        "      <td>0.034606</td>\n",
        "      <td>-0.079290</td>\n",
-       "      <td>-0.456263</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>normalized-losses</th>\n",
        "      <td>0.389010</td>\n",
        "      <td>1.000000</td>\n",
        "      <td>0.008200</td>\n",
+       "      <td>0.152832</td>\n",
+       "      <td>0.200870</td>\n",
+       "      <td>-0.287893</td>\n",
+       "      <td>0.218460</td>\n",
        "      <td>0.237114</td>\n",
        "      <td>0.047391</td>\n",
        "      <td>0.070319</td>\n",
        "      <td>-0.073362</td>\n",
        "      <td>0.370384</td>\n",
        "      <td>0.234384</td>\n",
+       "      <td>-0.321647</td>\n",
        "      <td>-0.270414</td>\n",
        "      <td>0.313716</td>\n",
-       "      <td>0.036586</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>wheel-base</th>\n",
        "      <td>-0.531954</td>\n",
        "      <td>0.008200</td>\n",
        "      <td>1.000000</td>\n",
+       "      <td>0.874587</td>\n",
+       "      <td>0.795144</td>\n",
+       "      <td>0.589435</td>\n",
+       "      <td>0.776386</td>\n",
        "      <td>0.569329</td>\n",
        "      <td>0.495108</td>\n",
        "      <td>0.164549</td>\n",
        "      <td>0.249786</td>\n",
        "      <td>0.301696</td>\n",
        "      <td>-0.363355</td>\n",
+       "      <td>-0.470414</td>\n",
        "      <td>-0.544082</td>\n",
        "      <td>0.572348</td>\n",
-       "      <td>0.913669</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>length</th>\n",
+       "      <td>-0.357612</td>\n",
+       "      <td>0.152832</td>\n",
+       "      <td>0.874587</td>\n",
+       "      <td>1.000000</td>\n",
+       "      <td>0.841118</td>\n",
+       "      <td>0.491029</td>\n",
+       "      <td>0.877728</td>\n",
+       "      <td>0.683360</td>\n",
+       "      <td>0.608905</td>\n",
+       "      <td>0.132076</td>\n",
+       "      <td>0.158414</td>\n",
+       "      <td>0.521192</td>\n",
+       "      <td>-0.279406</td>\n",
+       "      <td>-0.670909</td>\n",
+       "      <td>-0.704662</td>\n",
+       "      <td>0.680804</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>width</th>\n",
+       "      <td>-0.232919</td>\n",
+       "      <td>0.200870</td>\n",
+       "      <td>0.795144</td>\n",
+       "      <td>0.841118</td>\n",
+       "      <td>1.000000</td>\n",
+       "      <td>0.279210</td>\n",
+       "      <td>0.867032</td>\n",
+       "      <td>0.735433</td>\n",
+       "      <td>0.556374</td>\n",
+       "      <td>0.183379</td>\n",
+       "      <td>0.181129</td>\n",
+       "      <td>0.596251</td>\n",
+       "      <td>-0.214240</td>\n",
+       "      <td>-0.642704</td>\n",
+       "      <td>-0.677218</td>\n",
+       "      <td>0.765788</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>height</th>\n",
+       "      <td>-0.541038</td>\n",
+       "      <td>-0.287893</td>\n",
+       "      <td>0.589435</td>\n",
+       "      <td>0.491029</td>\n",
+       "      <td>0.279210</td>\n",
+       "      <td>1.000000</td>\n",
+       "      <td>0.295572</td>\n",
+       "      <td>0.067149</td>\n",
+       "      <td>0.199995</td>\n",
+       "      <td>-0.044176</td>\n",
+       "      <td>0.261214</td>\n",
+       "      <td>-0.114968</td>\n",
+       "      <td>-0.322525</td>\n",
+       "      <td>-0.048640</td>\n",
+       "      <td>-0.107358</td>\n",
+       "      <td>0.113942</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>curb-weight</th>\n",
+       "      <td>-0.227691</td>\n",
+       "      <td>0.218460</td>\n",
+       "      <td>0.776386</td>\n",
+       "      <td>0.877728</td>\n",
+       "      <td>0.867032</td>\n",
+       "      <td>0.295572</td>\n",
+       "      <td>1.000000</td>\n",
+       "      <td>0.850594</td>\n",
+       "      <td>0.648219</td>\n",
+       "      <td>0.170425</td>\n",
+       "      <td>0.151362</td>\n",
+       "      <td>0.679865</td>\n",
+       "      <td>-0.264976</td>\n",
+       "      <td>-0.757414</td>\n",
+       "      <td>-0.797465</td>\n",
+       "      <td>0.836802</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>engine-size</th>\n",
        "      <td>-0.105790</td>\n",
        "      <td>0.237114</td>\n",
        "      <td>0.569329</td>\n",
+       "      <td>0.683360</td>\n",
+       "      <td>0.735433</td>\n",
+       "      <td>0.067149</td>\n",
+       "      <td>0.850594</td>\n",
        "      <td>1.000000</td>\n",
        "      <td>0.602516</td>\n",
        "      <td>0.211477</td>\n",
        "      <td>0.028971</td>\n",
        "      <td>0.742119</td>\n",
        "      <td>-0.241031</td>\n",
+       "      <td>-0.653658</td>\n",
        "      <td>-0.677470</td>\n",
        "      <td>0.871189</td>\n",
-       "      <td>0.594351</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>bore</th>\n",
        "      <td>-0.160225</td>\n",
        "      <td>0.047391</td>\n",
        "      <td>0.495108</td>\n",
+       "      <td>0.608905</td>\n",
+       "      <td>0.556374</td>\n",
+       "      <td>0.199995</td>\n",
+       "      <td>0.648219</td>\n",
        "      <td>0.602516</td>\n",
        "      <td>1.000000</td>\n",
        "      <td>-0.049492</td>\n",
        "      <td>0.008511</td>\n",
        "      <td>0.537543</td>\n",
        "      <td>-0.276942</td>\n",
+       "      <td>-0.556570</td>\n",
        "      <td>-0.562065</td>\n",
        "      <td>0.550994</td>\n",
-       "      <td>0.549601</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>stroke</th>\n",
        "      <td>-0.020132</td>\n",
        "      <td>0.070319</td>\n",
        "      <td>0.164549</td>\n",
+       "      <td>0.132076</td>\n",
+       "      <td>0.183379</td>\n",
+       "      <td>-0.044176</td>\n",
+       "      <td>0.170425</td>\n",
        "      <td>0.211477</td>\n",
        "      <td>-0.049492</td>\n",
        "      <td>1.000000</td>\n",
        "      <td>0.187134</td>\n",
        "      <td>0.164722</td>\n",
        "      <td>-0.051970</td>\n",
+       "      <td>-0.033609</td>\n",
        "      <td>-0.036502</td>\n",
        "      <td>0.080531</td>\n",
-       "      <td>0.104859</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>compress-ratio</th>\n",
        "      <td>-0.178515</td>\n",
        "      <td>-0.073362</td>\n",
        "      <td>0.249786</td>\n",
+       "      <td>0.158414</td>\n",
+       "      <td>0.181129</td>\n",
+       "      <td>0.261214</td>\n",
+       "      <td>0.151362</td>\n",
        "      <td>0.028971</td>\n",
        "      <td>0.008511</td>\n",
        "      <td>0.187134</td>\n",
        "      <td>1.000000</td>\n",
        "      <td>-0.202096</td>\n",
        "      <td>-0.436976</td>\n",
+       "      <td>0.324701</td>\n",
        "      <td>0.265201</td>\n",
        "      <td>0.064750</td>\n",
-       "      <td>0.233301</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>horsepower</th>\n",
        "      <td>0.070421</td>\n",
        "      <td>0.370384</td>\n",
        "      <td>0.301696</td>\n",
+       "      <td>0.521192</td>\n",
+       "      <td>0.596251</td>\n",
+       "      <td>-0.114968</td>\n",
+       "      <td>0.679865</td>\n",
        "      <td>0.742119</td>\n",
        "      <td>0.537543</td>\n",
        "      <td>0.164722</td>\n",
        "      <td>-0.202096</td>\n",
        "      <td>1.000000</td>\n",
        "      <td>0.171390</td>\n",
+       "      <td>-0.744246</td>\n",
        "      <td>-0.699361</td>\n",
        "      <td>0.725734</td>\n",
-       "      <td>0.396466</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>peak-rpm</th>\n",
        "      <td>0.273125</td>\n",
        "      <td>0.234384</td>\n",
        "      <td>-0.363355</td>\n",
+       "      <td>-0.279406</td>\n",
+       "      <td>-0.214240</td>\n",
+       "      <td>-0.322525</td>\n",
+       "      <td>-0.264976</td>\n",
        "      <td>-0.241031</td>\n",
        "      <td>-0.276942</td>\n",
        "      <td>-0.051970</td>\n",
        "      <td>-0.436976</td>\n",
        "      <td>0.171390</td>\n",
        "      <td>1.000000</td>\n",
+       "      <td>-0.117108</td>\n",
        "      <td>-0.053351</td>\n",
        "      <td>-0.084937</td>\n",
-       "      <td>-0.325444</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>city-mpg</th>\n",
+       "      <td>-0.035823</td>\n",
+       "      <td>-0.321647</td>\n",
+       "      <td>-0.470414</td>\n",
+       "      <td>-0.670909</td>\n",
+       "      <td>-0.642704</td>\n",
+       "      <td>-0.048640</td>\n",
+       "      <td>-0.757414</td>\n",
+       "      <td>-0.653658</td>\n",
+       "      <td>-0.556570</td>\n",
+       "      <td>-0.033609</td>\n",
+       "      <td>0.324701</td>\n",
+       "      <td>-0.744246</td>\n",
+       "      <td>-0.117108</td>\n",
+       "      <td>1.000000</td>\n",
+       "      <td>0.971337</td>\n",
+       "      <td>-0.693326</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>highway-mpg</th>\n",
        "      <td>0.034606</td>\n",
        "      <td>-0.270414</td>\n",
        "      <td>-0.544082</td>\n",
+       "      <td>-0.704662</td>\n",
+       "      <td>-0.677218</td>\n",
+       "      <td>-0.107358</td>\n",
+       "      <td>-0.797465</td>\n",
        "      <td>-0.677470</td>\n",
        "      <td>-0.562065</td>\n",
        "      <td>-0.036502</td>\n",
        "      <td>0.265201</td>\n",
        "      <td>-0.699361</td>\n",
        "      <td>-0.053351</td>\n",
+       "      <td>0.971337</td>\n",
        "      <td>1.000000</td>\n",
        "      <td>-0.700791</td>\n",
-       "      <td>-0.602410</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>price</th>\n",
        "      <td>-0.079290</td>\n",
        "      <td>0.313716</td>\n",
        "      <td>0.572348</td>\n",
+       "      <td>0.680804</td>\n",
+       "      <td>0.765788</td>\n",
+       "      <td>0.113942</td>\n",
+       "      <td>0.836802</td>\n",
        "      <td>0.871189</td>\n",
        "      <td>0.550994</td>\n",
        "      <td>0.080531</td>\n",
        "      <td>0.064750</td>\n",
        "      <td>0.725734</td>\n",
        "      <td>-0.084937</td>\n",
+       "      <td>-0.693326</td>\n",
        "      <td>-0.700791</td>\n",
        "      <td>1.000000</td>\n",
-       "      <td>0.622139</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>volume</th>\n",
-       "      <td>-0.456263</td>\n",
-       "      <td>0.036586</td>\n",
-       "      <td>0.913669</td>\n",
-       "      <td>0.594351</td>\n",
-       "      <td>0.549601</td>\n",
-       "      <td>0.104859</td>\n",
-       "      <td>0.233301</td>\n",
-       "      <td>0.396466</td>\n",
-       "      <td>-0.325444</td>\n",
-       "      <td>-0.602410</td>\n",
-       "      <td>0.622139</td>\n",
-       "      <td>1.000000</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "text/plain": [
-       "                   symboling  normalized-losses  wheel-base  engine-size  \\\n",
-       "symboling           1.000000           0.389010   -0.531954    -0.105790   \n",
-       "normalized-losses   0.389010           1.000000    0.008200     0.237114   \n",
-       "wheel-base         -0.531954           0.008200    1.000000     0.569329   \n",
-       "engine-size        -0.105790           0.237114    0.569329     1.000000   \n",
-       "bore               -0.160225           0.047391    0.495108     0.602516   \n",
-       "stroke             -0.020132           0.070319    0.164549     0.211477   \n",
-       "compress-ratio     -0.178515          -0.073362    0.249786     0.028971   \n",
-       "horsepower          0.070421           0.370384    0.301696     0.742119   \n",
-       "peak-rpm            0.273125           0.234384   -0.363355    -0.241031   \n",
-       "highway-mpg         0.034606          -0.270414   -0.544082    -0.677470   \n",
-       "price              -0.079290           0.313716    0.572348     0.871189   \n",
-       "volume             -0.456263           0.036586    0.913669     0.594351   \n",
+       "                   symboling  normalized-losses  wheel-base    length  \\\n",
+       "symboling           1.000000           0.389010   -0.531954 -0.357612   \n",
+       "normalized-losses   0.389010           1.000000    0.008200  0.152832   \n",
+       "wheel-base         -0.531954           0.008200    1.000000  0.874587   \n",
+       "length             -0.357612           0.152832    0.874587  1.000000   \n",
+       "width              -0.232919           0.200870    0.795144  0.841118   \n",
+       "height             -0.541038          -0.287893    0.589435  0.491029   \n",
+       "curb-weight        -0.227691           0.218460    0.776386  0.877728   \n",
+       "engine-size        -0.105790           0.237114    0.569329  0.683360   \n",
+       "bore               -0.160225           0.047391    0.495108  0.608905   \n",
+       "stroke             -0.020132           0.070319    0.164549  0.132076   \n",
+       "compress-ratio     -0.178515          -0.073362    0.249786  0.158414   \n",
+       "horsepower          0.070421           0.370384    0.301696  0.521192   \n",
+       "peak-rpm            0.273125           0.234384   -0.363355 -0.279406   \n",
+       "city-mpg           -0.035823          -0.321647   -0.470414 -0.670909   \n",
+       "highway-mpg         0.034606          -0.270414   -0.544082 -0.704662   \n",
+       "price              -0.079290           0.313716    0.572348  0.680804   \n",
        "\n",
-       "                       bore    stroke  compress-ratio  horsepower  peak-rpm  \\\n",
-       "symboling         -0.160225 -0.020132       -0.178515    0.070421  0.273125   \n",
-       "normalized-losses  0.047391  0.070319       -0.073362    0.370384  0.234384   \n",
-       "wheel-base         0.495108  0.164549        0.249786    0.301696 -0.363355   \n",
-       "engine-size        0.602516  0.211477        0.028971    0.742119 -0.241031   \n",
-       "bore               1.000000 -0.049492        0.008511    0.537543 -0.276942   \n",
-       "stroke            -0.049492  1.000000        0.187134    0.164722 -0.051970   \n",
-       "compress-ratio     0.008511  0.187134        1.000000   -0.202096 -0.436976   \n",
-       "horsepower         0.537543  0.164722       -0.202096    1.000000  0.171390   \n",
-       "peak-rpm          -0.276942 -0.051970       -0.436976    0.171390  1.000000   \n",
-       "highway-mpg       -0.562065 -0.036502        0.265201   -0.699361 -0.053351   \n",
-       "price              0.550994  0.080531        0.064750    0.725734 -0.084937   \n",
-       "volume             0.549601  0.104859        0.233301    0.396466 -0.325444   \n",
+       "                      width    height  curb-weight  engine-size      bore  \\\n",
+       "symboling         -0.232919 -0.541038    -0.227691    -0.105790 -0.160225   \n",
+       "normalized-losses  0.200870 -0.287893     0.218460     0.237114  0.047391   \n",
+       "wheel-base         0.795144  0.589435     0.776386     0.569329  0.495108   \n",
+       "length             0.841118  0.491029     0.877728     0.683360  0.608905   \n",
+       "width              1.000000  0.279210     0.867032     0.735433  0.556374   \n",
+       "height             0.279210  1.000000     0.295572     0.067149  0.199995   \n",
+       "curb-weight        0.867032  0.295572     1.000000     0.850594  0.648219   \n",
+       "engine-size        0.735433  0.067149     0.850594     1.000000  0.602516   \n",
+       "bore               0.556374  0.199995     0.648219     0.602516  1.000000   \n",
+       "stroke             0.183379 -0.044176     0.170425     0.211477 -0.049492   \n",
+       "compress-ratio     0.181129  0.261214     0.151362     0.028971  0.008511   \n",
+       "horsepower         0.596251 -0.114968     0.679865     0.742119  0.537543   \n",
+       "peak-rpm          -0.214240 -0.322525    -0.264976    -0.241031 -0.276942   \n",
+       "city-mpg          -0.642704 -0.048640    -0.757414    -0.653658 -0.556570   \n",
+       "highway-mpg       -0.677218 -0.107358    -0.797465    -0.677470 -0.562065   \n",
+       "price              0.765788  0.113942     0.836802     0.871189  0.550994   \n",
        "\n",
-       "                   highway-mpg     price    volume  \n",
-       "symboling             0.034606 -0.079290 -0.456263  \n",
-       "normalized-losses    -0.270414  0.313716  0.036586  \n",
-       "wheel-base           -0.544082  0.572348  0.913669  \n",
-       "engine-size          -0.677470  0.871189  0.594351  \n",
-       "bore                 -0.562065  0.550994  0.549601  \n",
-       "stroke               -0.036502  0.080531  0.104859  \n",
-       "compress-ratio        0.265201  0.064750  0.233301  \n",
-       "horsepower           -0.699361  0.725734  0.396466  \n",
-       "peak-rpm             -0.053351 -0.084937 -0.325444  \n",
-       "highway-mpg           1.000000 -0.700791 -0.602410  \n",
-       "price                -0.700791  1.000000  0.622139  \n",
-       "volume               -0.602410  0.622139  1.000000  "
+       "                     stroke  compress-ratio  horsepower  peak-rpm  city-mpg  \\\n",
+       "symboling         -0.020132       -0.178515    0.070421  0.273125 -0.035823   \n",
+       "normalized-losses  0.070319       -0.073362    0.370384  0.234384 -0.321647   \n",
+       "wheel-base         0.164549        0.249786    0.301696 -0.363355 -0.470414   \n",
+       "length             0.132076        0.158414    0.521192 -0.279406 -0.670909   \n",
+       "width              0.183379        0.181129    0.596251 -0.214240 -0.642704   \n",
+       "height            -0.044176        0.261214   -0.114968 -0.322525 -0.048640   \n",
+       "curb-weight        0.170425        0.151362    0.679865 -0.264976 -0.757414   \n",
+       "engine-size        0.211477        0.028971    0.742119 -0.241031 -0.653658   \n",
+       "bore              -0.049492        0.008511    0.537543 -0.276942 -0.556570   \n",
+       "stroke             1.000000        0.187134    0.164722 -0.051970 -0.033609   \n",
+       "compress-ratio     0.187134        1.000000   -0.202096 -0.436976  0.324701   \n",
+       "horsepower         0.164722       -0.202096    1.000000  0.171390 -0.744246   \n",
+       "peak-rpm          -0.051970       -0.436976    0.171390  1.000000 -0.117108   \n",
+       "city-mpg          -0.033609        0.324701   -0.744246 -0.117108  1.000000   \n",
+       "highway-mpg       -0.036502        0.265201   -0.699361 -0.053351  0.971337   \n",
+       "price              0.080531        0.064750    0.725734 -0.084937 -0.693326   \n",
+       "\n",
+       "                   highway-mpg     price  \n",
+       "symboling             0.034606 -0.079290  \n",
+       "normalized-losses    -0.270414  0.313716  \n",
+       "wheel-base           -0.544082  0.572348  \n",
+       "length               -0.704662  0.680804  \n",
+       "width                -0.677218  0.765788  \n",
+       "height               -0.107358  0.113942  \n",
+       "curb-weight          -0.797465  0.836802  \n",
+       "engine-size          -0.677470  0.871189  \n",
+       "bore                 -0.562065  0.550994  \n",
+       "stroke               -0.036502  0.080531  \n",
+       "compress-ratio        0.265201  0.064750  \n",
+       "horsepower           -0.699361  0.725734  \n",
+       "peak-rpm             -0.053351 -0.084937  \n",
+       "city-mpg              0.971337 -0.693326  \n",
+       "highway-mpg           1.000000 -0.700791  \n",
+       "price                -0.700791  1.000000  "
       ]
      },
-     "execution_count": 28,
+     "execution_count": 10,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -1471,7 +1631,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 30,
+   "execution_count": 11,
    "metadata": {},
    "outputs": [
     {
@@ -1498,15 +1658,19 @@
        "      <th>symboling</th>\n",
        "      <th>normalized-losses</th>\n",
        "      <th>wheel-base</th>\n",
+       "      <th>length</th>\n",
+       "      <th>width</th>\n",
+       "      <th>height</th>\n",
+       "      <th>curb-weight</th>\n",
        "      <th>engine-size</th>\n",
        "      <th>bore</th>\n",
        "      <th>stroke</th>\n",
        "      <th>compress-ratio</th>\n",
        "      <th>horsepower</th>\n",
        "      <th>peak-rpm</th>\n",
+       "      <th>city-mpg</th>\n",
        "      <th>highway-mpg</th>\n",
        "      <th>price</th>\n",
-       "      <th>volume</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
@@ -1515,45 +1679,133 @@
        "      <td>0.0</td>\n",
        "      <td>0.38901</td>\n",
        "      <td>-0.531954</td>\n",
+       "      <td>-0.357612</td>\n",
+       "      <td>-0.232919</td>\n",
+       "      <td>-0.541038</td>\n",
+       "      <td>-0.227691</td>\n",
        "      <td>-0.105790</td>\n",
        "      <td>-0.160225</td>\n",
        "      <td>-0.020132</td>\n",
        "      <td>-0.178515</td>\n",
        "      <td>0.070421</td>\n",
        "      <td>0.273125</td>\n",
+       "      <td>-0.035823</td>\n",
        "      <td>0.034606</td>\n",
        "      <td>-0.079290</td>\n",
-       "      <td>-0.456263</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>normalized-losses</th>\n",
        "      <td>0.0</td>\n",
        "      <td>0.00000</td>\n",
        "      <td>0.008200</td>\n",
+       "      <td>0.152832</td>\n",
+       "      <td>0.200870</td>\n",
+       "      <td>-0.287893</td>\n",
+       "      <td>0.218460</td>\n",
        "      <td>0.237114</td>\n",
        "      <td>0.047391</td>\n",
        "      <td>0.070319</td>\n",
        "      <td>-0.073362</td>\n",
        "      <td>0.370384</td>\n",
        "      <td>0.234384</td>\n",
+       "      <td>-0.321647</td>\n",
        "      <td>-0.270414</td>\n",
        "      <td>0.313716</td>\n",
-       "      <td>0.036586</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>wheel-base</th>\n",
        "      <td>-0.0</td>\n",
        "      <td>0.00000</td>\n",
        "      <td>0.000000</td>\n",
+       "      <td>0.874587</td>\n",
+       "      <td>0.795144</td>\n",
+       "      <td>0.589435</td>\n",
+       "      <td>0.776386</td>\n",
        "      <td>0.569329</td>\n",
        "      <td>0.495108</td>\n",
        "      <td>0.164549</td>\n",
        "      <td>0.249786</td>\n",
        "      <td>0.301696</td>\n",
        "      <td>-0.363355</td>\n",
+       "      <td>-0.470414</td>\n",
        "      <td>-0.544082</td>\n",
        "      <td>0.572348</td>\n",
-       "      <td>0.913669</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>length</th>\n",
+       "      <td>-0.0</td>\n",
+       "      <td>0.00000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.841118</td>\n",
+       "      <td>0.491029</td>\n",
+       "      <td>0.877728</td>\n",
+       "      <td>0.683360</td>\n",
+       "      <td>0.608905</td>\n",
+       "      <td>0.132076</td>\n",
+       "      <td>0.158414</td>\n",
+       "      <td>0.521192</td>\n",
+       "      <td>-0.279406</td>\n",
+       "      <td>-0.670909</td>\n",
+       "      <td>-0.704662</td>\n",
+       "      <td>0.680804</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>width</th>\n",
+       "      <td>-0.0</td>\n",
+       "      <td>0.00000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.279210</td>\n",
+       "      <td>0.867032</td>\n",
+       "      <td>0.735433</td>\n",
+       "      <td>0.556374</td>\n",
+       "      <td>0.183379</td>\n",
+       "      <td>0.181129</td>\n",
+       "      <td>0.596251</td>\n",
+       "      <td>-0.214240</td>\n",
+       "      <td>-0.642704</td>\n",
+       "      <td>-0.677218</td>\n",
+       "      <td>0.765788</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>height</th>\n",
+       "      <td>-0.0</td>\n",
+       "      <td>-0.00000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.295572</td>\n",
+       "      <td>0.067149</td>\n",
+       "      <td>0.199995</td>\n",
+       "      <td>-0.044176</td>\n",
+       "      <td>0.261214</td>\n",
+       "      <td>-0.114968</td>\n",
+       "      <td>-0.322525</td>\n",
+       "      <td>-0.048640</td>\n",
+       "      <td>-0.107358</td>\n",
+       "      <td>0.113942</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>curb-weight</th>\n",
+       "      <td>-0.0</td>\n",
+       "      <td>0.00000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.850594</td>\n",
+       "      <td>0.648219</td>\n",
+       "      <td>0.170425</td>\n",
+       "      <td>0.151362</td>\n",
+       "      <td>0.679865</td>\n",
+       "      <td>-0.264976</td>\n",
+       "      <td>-0.757414</td>\n",
+       "      <td>-0.797465</td>\n",
+       "      <td>0.836802</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>engine-size</th>\n",
@@ -1561,14 +1813,18 @@
        "      <td>0.00000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
        "      <td>0.602516</td>\n",
        "      <td>0.211477</td>\n",
        "      <td>0.028971</td>\n",
        "      <td>0.742119</td>\n",
        "      <td>-0.241031</td>\n",
+       "      <td>-0.653658</td>\n",
        "      <td>-0.677470</td>\n",
        "      <td>0.871189</td>\n",
-       "      <td>0.594351</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>bore</th>\n",
@@ -1577,13 +1833,17 @@
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
        "      <td>-0.049492</td>\n",
        "      <td>0.008511</td>\n",
        "      <td>0.537543</td>\n",
        "      <td>-0.276942</td>\n",
+       "      <td>-0.556570</td>\n",
        "      <td>-0.562065</td>\n",
        "      <td>0.550994</td>\n",
-       "      <td>0.549601</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>stroke</th>\n",
@@ -1591,14 +1851,18 @@
        "      <td>0.00000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
        "      <td>-0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.187134</td>\n",
        "      <td>0.164722</td>\n",
        "      <td>-0.051970</td>\n",
+       "      <td>-0.033609</td>\n",
        "      <td>-0.036502</td>\n",
        "      <td>0.080531</td>\n",
-       "      <td>0.104859</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>compress-ratio</th>\n",
@@ -1609,11 +1873,15 @@
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
        "      <td>-0.202096</td>\n",
        "      <td>-0.436976</td>\n",
+       "      <td>0.324701</td>\n",
        "      <td>0.265201</td>\n",
        "      <td>0.064750</td>\n",
-       "      <td>0.233301</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>horsepower</th>\n",
@@ -1622,13 +1890,17 @@
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>-0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.171390</td>\n",
+       "      <td>-0.744246</td>\n",
        "      <td>-0.699361</td>\n",
        "      <td>0.725734</td>\n",
-       "      <td>0.396466</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>peak-rpm</th>\n",
@@ -1639,11 +1911,34 @@
        "      <td>-0.000000</td>\n",
        "      <td>-0.000000</td>\n",
        "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
+       "      <td>-0.117108</td>\n",
        "      <td>-0.053351</td>\n",
        "      <td>-0.084937</td>\n",
-       "      <td>-0.325444</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>city-mpg</th>\n",
+       "      <td>-0.0</td>\n",
+       "      <td>-0.00000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>0.000000</td>\n",
+       "      <td>0.971337</td>\n",
+       "      <td>-0.693326</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>highway-mpg</th>\n",
@@ -1653,12 +1948,16 @@
        "      <td>-0.000000</td>\n",
        "      <td>-0.000000</td>\n",
        "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>-0.000000</td>\n",
        "      <td>-0.000000</td>\n",
        "      <td>0.000000</td>\n",
+       "      <td>0.000000</td>\n",
        "      <td>-0.700791</td>\n",
-       "      <td>-0.602410</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>price</th>\n",
@@ -1670,24 +1969,13 @@
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
-       "      <td>-0.000000</td>\n",
-       "      <td>-0.000000</td>\n",
-       "      <td>0.000000</td>\n",
-       "      <td>0.622139</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>volume</th>\n",
-       "      <td>-0.0</td>\n",
-       "      <td>0.00000</td>\n",
-       "      <td>0.000000</td>\n",
-       "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "      <td>-0.000000</td>\n",
        "      <td>-0.000000</td>\n",
-       "      <td>0.000000</td>\n",
+       "      <td>-0.000000</td>\n",
        "      <td>0.000000</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
@@ -1695,50 +1983,80 @@
        "</div>"
       ],
       "text/plain": [
-       "                   symboling  normalized-losses  wheel-base  engine-size  \\\n",
-       "symboling                0.0            0.38901   -0.531954    -0.105790   \n",
-       "normalized-losses        0.0            0.00000    0.008200     0.237114   \n",
-       "wheel-base              -0.0            0.00000    0.000000     0.569329   \n",
-       "engine-size             -0.0            0.00000    0.000000     0.000000   \n",
-       "bore                    -0.0            0.00000    0.000000     0.000000   \n",
-       "stroke                  -0.0            0.00000    0.000000     0.000000   \n",
-       "compress-ratio          -0.0           -0.00000    0.000000     0.000000   \n",
-       "horsepower               0.0            0.00000    0.000000     0.000000   \n",
-       "peak-rpm                 0.0            0.00000   -0.000000    -0.000000   \n",
-       "highway-mpg              0.0           -0.00000   -0.000000    -0.000000   \n",
-       "price                   -0.0            0.00000    0.000000     0.000000   \n",
-       "volume                  -0.0            0.00000    0.000000     0.000000   \n",
+       "                   symboling  normalized-losses  wheel-base    length  \\\n",
+       "symboling                0.0            0.38901   -0.531954 -0.357612   \n",
+       "normalized-losses        0.0            0.00000    0.008200  0.152832   \n",
+       "wheel-base              -0.0            0.00000    0.000000  0.874587   \n",
+       "length                  -0.0            0.00000    0.000000  0.000000   \n",
+       "width                   -0.0            0.00000    0.000000  0.000000   \n",
+       "height                  -0.0           -0.00000    0.000000  0.000000   \n",
+       "curb-weight             -0.0            0.00000    0.000000  0.000000   \n",
+       "engine-size             -0.0            0.00000    0.000000  0.000000   \n",
+       "bore                    -0.0            0.00000    0.000000  0.000000   \n",
+       "stroke                  -0.0            0.00000    0.000000  0.000000   \n",
+       "compress-ratio          -0.0           -0.00000    0.000000  0.000000   \n",
+       "horsepower               0.0            0.00000    0.000000  0.000000   \n",
+       "peak-rpm                 0.0            0.00000   -0.000000 -0.000000   \n",
+       "city-mpg                -0.0           -0.00000   -0.000000 -0.000000   \n",
+       "highway-mpg              0.0           -0.00000   -0.000000 -0.000000   \n",
+       "price                   -0.0            0.00000    0.000000  0.000000   \n",
+       "\n",
+       "                      width    height  curb-weight  engine-size      bore  \\\n",
+       "symboling         -0.232919 -0.541038    -0.227691    -0.105790 -0.160225   \n",
+       "normalized-losses  0.200870 -0.287893     0.218460     0.237114  0.047391   \n",
+       "wheel-base         0.795144  0.589435     0.776386     0.569329  0.495108   \n",
+       "length             0.841118  0.491029     0.877728     0.683360  0.608905   \n",
+       "width              0.000000  0.279210     0.867032     0.735433  0.556374   \n",
+       "height             0.000000  0.000000     0.295572     0.067149  0.199995   \n",
+       "curb-weight        0.000000  0.000000     0.000000     0.850594  0.648219   \n",
+       "engine-size        0.000000  0.000000     0.000000     0.000000  0.602516   \n",
+       "bore               0.000000  0.000000     0.000000     0.000000  0.000000   \n",
+       "stroke             0.000000 -0.000000     0.000000     0.000000 -0.000000   \n",
+       "compress-ratio     0.000000  0.000000     0.000000     0.000000  0.000000   \n",
+       "horsepower         0.000000 -0.000000     0.000000     0.000000  0.000000   \n",
+       "peak-rpm          -0.000000 -0.000000    -0.000000    -0.000000 -0.000000   \n",
+       "city-mpg          -0.000000 -0.000000    -0.000000    -0.000000 -0.000000   \n",
+       "highway-mpg       -0.000000 -0.000000    -0.000000    -0.000000 -0.000000   \n",
+       "price              0.000000  0.000000     0.000000     0.000000  0.000000   \n",
        "\n",
-       "                       bore    stroke  compress-ratio  horsepower  peak-rpm  \\\n",
-       "symboling         -0.160225 -0.020132       -0.178515    0.070421  0.273125   \n",
-       "normalized-losses  0.047391  0.070319       -0.073362    0.370384  0.234384   \n",
-       "wheel-base         0.495108  0.164549        0.249786    0.301696 -0.363355   \n",
-       "engine-size        0.602516  0.211477        0.028971    0.742119 -0.241031   \n",
-       "bore               0.000000 -0.049492        0.008511    0.537543 -0.276942   \n",
-       "stroke            -0.000000  0.000000        0.187134    0.164722 -0.051970   \n",
-       "compress-ratio     0.000000  0.000000        0.000000   -0.202096 -0.436976   \n",
-       "horsepower         0.000000  0.000000       -0.000000    0.000000  0.171390   \n",
-       "peak-rpm          -0.000000 -0.000000       -0.000000    0.000000  0.000000   \n",
-       "highway-mpg       -0.000000 -0.000000        0.000000   -0.000000 -0.000000   \n",
-       "price              0.000000  0.000000        0.000000    0.000000 -0.000000   \n",
-       "volume             0.000000  0.000000        0.000000    0.000000 -0.000000   \n",
+       "                     stroke  compress-ratio  horsepower  peak-rpm  city-mpg  \\\n",
+       "symboling         -0.020132       -0.178515    0.070421  0.273125 -0.035823   \n",
+       "normalized-losses  0.070319       -0.073362    0.370384  0.234384 -0.321647   \n",
+       "wheel-base         0.164549        0.249786    0.301696 -0.363355 -0.470414   \n",
+       "length             0.132076        0.158414    0.521192 -0.279406 -0.670909   \n",
+       "width              0.183379        0.181129    0.596251 -0.214240 -0.642704   \n",
+       "height            -0.044176        0.261214   -0.114968 -0.322525 -0.048640   \n",
+       "curb-weight        0.170425        0.151362    0.679865 -0.264976 -0.757414   \n",
+       "engine-size        0.211477        0.028971    0.742119 -0.241031 -0.653658   \n",
+       "bore              -0.049492        0.008511    0.537543 -0.276942 -0.556570   \n",
+       "stroke             0.000000        0.187134    0.164722 -0.051970 -0.033609   \n",
+       "compress-ratio     0.000000        0.000000   -0.202096 -0.436976  0.324701   \n",
+       "horsepower         0.000000       -0.000000    0.000000  0.171390 -0.744246   \n",
+       "peak-rpm          -0.000000       -0.000000    0.000000  0.000000 -0.117108   \n",
+       "city-mpg          -0.000000        0.000000   -0.000000 -0.000000  0.000000   \n",
+       "highway-mpg       -0.000000        0.000000   -0.000000 -0.000000  0.000000   \n",
+       "price              0.000000        0.000000    0.000000 -0.000000 -0.000000   \n",
        "\n",
-       "                   highway-mpg     price    volume  \n",
-       "symboling             0.034606 -0.079290 -0.456263  \n",
-       "normalized-losses    -0.270414  0.313716  0.036586  \n",
-       "wheel-base           -0.544082  0.572348  0.913669  \n",
-       "engine-size          -0.677470  0.871189  0.594351  \n",
-       "bore                 -0.562065  0.550994  0.549601  \n",
-       "stroke               -0.036502  0.080531  0.104859  \n",
-       "compress-ratio        0.265201  0.064750  0.233301  \n",
-       "horsepower           -0.699361  0.725734  0.396466  \n",
-       "peak-rpm             -0.053351 -0.084937 -0.325444  \n",
-       "highway-mpg           0.000000 -0.700791 -0.602410  \n",
-       "price                -0.000000  0.000000  0.622139  \n",
-       "volume               -0.000000  0.000000  0.000000  "
+       "                   highway-mpg     price  \n",
+       "symboling             0.034606 -0.079290  \n",
+       "normalized-losses    -0.270414  0.313716  \n",
+       "wheel-base           -0.544082  0.572348  \n",
+       "length               -0.704662  0.680804  \n",
+       "width                -0.677218  0.765788  \n",
+       "height               -0.107358  0.113942  \n",
+       "curb-weight          -0.797465  0.836802  \n",
+       "engine-size          -0.677470  0.871189  \n",
+       "bore                 -0.562065  0.550994  \n",
+       "stroke               -0.036502  0.080531  \n",
+       "compress-ratio        0.265201  0.064750  \n",
+       "horsepower           -0.699361  0.725734  \n",
+       "peak-rpm             -0.053351 -0.084937  \n",
+       "city-mpg              0.971337 -0.693326  \n",
+       "highway-mpg           0.000000 -0.700791  \n",
+       "price                -0.000000  0.000000  "
       ]
      },
-     "execution_count": 30,
+     "execution_count": 11,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -1840,7 +2158,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 16,
+   "execution_count": 13,
    "metadata": {},
    "outputs": [
     {
@@ -2308,7 +2626,7 @@
        "[256 rows x 3 columns]"
       ]
      },
-     "execution_count": 16,
+     "execution_count": 13,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -2322,7 +2640,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 19,
+   "execution_count": 14,
    "metadata": {},
    "outputs": [
     {
@@ -2430,7 +2748,7 @@
        "9   wheel-base        width     0.795144"
       ]
      },
-     "execution_count": 19,
+     "execution_count": 14,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -2459,7 +2777,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 20,
+   "execution_count": 15,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -2472,7 +2790,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 22,
+   "execution_count": 16,
    "metadata": {},
    "outputs": [
     {
@@ -2486,7 +2804,7 @@
        "      dtype='object')"
       ]
      },
-     "execution_count": 22,
+     "execution_count": 16,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -2504,7 +2822,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 33,
+   "execution_count": 17,
    "metadata": {},
    "outputs": [
     {
@@ -2541,16 +2859,16 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 34,
+   "execution_count": 18,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
-       "<seaborn.axisgrid.PairGrid at 0x25cce5c3da0>"
+       "<seaborn.axisgrid.PairGrid at 0x1731f2d8ac8>"
       ]
      },
-     "execution_count": 34,
+     "execution_count": 18,
      "metadata": {},
      "output_type": "execute_result"
     },
@@ -2581,22 +2899,22 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 35,
+   "execution_count": 19,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
-       "<seaborn.axisgrid.FacetGrid at 0x25cd481ab00>"
+       "<seaborn.axisgrid.FacetGrid at 0x1731f283828>"
       ]
      },
-     "execution_count": 35,
+     "execution_count": 19,
      "metadata": {},
      "output_type": "execute_result"
     },
     {
      "data": {
-      "image/png": "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\n",
+      "image/png": "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\n",
       "text/plain": [
        "<Figure size 720x720 with 4 Axes>"
       ]
@@ -2617,11 +2935,868 @@
     "#可以看到价格和马力之间的关系还是很强的"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### 预处理\n",
+    "如果一个特征的方差比其它的大很多，那么它可能支配目标函数，使估计者不能像预期的那样正确的从其它特征中学习。这就是为什么要坐数据缩放"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "对连续值进行标准化，让数据落在大致的区域范围内，而不是差异非常大\n",
+    "<br>\n",
+    "你也可以尝试某个值不做归一化，后续的特征重要性可以看到不做归一化的重要性极高"
+   ]
+  },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 71,
+   "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>symboling</th>\n",
+       "      <th>normalized-losses</th>\n",
+       "      <th>make</th>\n",
+       "      <th>fuel-type</th>\n",
+       "      <th>aspiration</th>\n",
+       "      <th>num-of-doors</th>\n",
+       "      <th>body-style</th>\n",
+       "      <th>drive-wheels</th>\n",
+       "      <th>engine-location</th>\n",
+       "      <th>wheel-base</th>\n",
+       "      <th>...</th>\n",
+       "      <th>engine-size</th>\n",
+       "      <th>fuel-system</th>\n",
+       "      <th>bore</th>\n",
+       "      <th>stroke</th>\n",
+       "      <th>compress-ratio</th>\n",
+       "      <th>horsepower</th>\n",
+       "      <th>peak-rpm</th>\n",
+       "      <th>highway-mpg</th>\n",
+       "      <th>output</th>\n",
+       "      <th>volume</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>1.743470</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>alfa-romero</td>\n",
+       "      <td>gas</td>\n",
+       "      <td>std</td>\n",
+       "      <td>two</td>\n",
+       "      <td>convertible</td>\n",
+       "      <td>rwd</td>\n",
+       "      <td>front</td>\n",
+       "      <td>-1.690772</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0.074449</td>\n",
+       "      <td>mpfi</td>\n",
+       "      <td>0.532789</td>\n",
+       "      <td>-1.830840</td>\n",
+       "      <td>-0.288349</td>\n",
+       "      <td>0.114182</td>\n",
+       "      <td>-0.27402</td>\n",
+       "      <td>-0.546059</td>\n",
+       "      <td>no</td>\n",
+       "      <td>-1.144195</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1.743470</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>alfa-romero</td>\n",
+       "      <td>gas</td>\n",
+       "      <td>std</td>\n",
+       "      <td>two</td>\n",
+       "      <td>convertible</td>\n",
+       "      <td>rwd</td>\n",
+       "      <td>front</td>\n",
+       "      <td>-1.690772</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0.074449</td>\n",
+       "      <td>mpfi</td>\n",
+       "      <td>0.532789</td>\n",
+       "      <td>-1.830840</td>\n",
+       "      <td>-0.288349</td>\n",
+       "      <td>0.114182</td>\n",
+       "      <td>-0.27402</td>\n",
+       "      <td>-0.546059</td>\n",
+       "      <td>no</td>\n",
+       "      <td>-1.144195</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>0.133509</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>alfa-romero</td>\n",
+       "      <td>gas</td>\n",
+       "      <td>std</td>\n",
+       "      <td>two</td>\n",
+       "      <td>hatchback</td>\n",
+       "      <td>rwd</td>\n",
+       "      <td>front</td>\n",
+       "      <td>-0.708596</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0.604046</td>\n",
+       "      <td>mpfi</td>\n",
+       "      <td>-2.367552</td>\n",
+       "      <td>0.691744</td>\n",
+       "      <td>-0.288349</td>\n",
+       "      <td>1.105818</td>\n",
+       "      <td>-0.27402</td>\n",
+       "      <td>-0.691627</td>\n",
+       "      <td>no</td>\n",
+       "      <td>-0.392670</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>0.938490</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>audi</td>\n",
+       "      <td>gas</td>\n",
+       "      <td>std</td>\n",
+       "      <td>four</td>\n",
+       "      <td>sedan</td>\n",
+       "      <td>fwd</td>\n",
+       "      <td>front</td>\n",
+       "      <td>0.173698</td>\n",
+       "      <td>...</td>\n",
+       "      <td>-0.431076</td>\n",
+       "      <td>mpfi</td>\n",
+       "      <td>-0.495180</td>\n",
+       "      <td>0.468224</td>\n",
+       "      <td>-0.035973</td>\n",
+       "      <td>-0.093370</td>\n",
+       "      <td>0.76817</td>\n",
+       "      <td>-0.109354</td>\n",
+       "      <td>no</td>\n",
+       "      <td>0.203076</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>0.938490</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>audi</td>\n",
+       "      <td>gas</td>\n",
+       "      <td>std</td>\n",
+       "      <td>four</td>\n",
+       "      <td>sedan</td>\n",
+       "      <td>4wd</td>\n",
+       "      <td>front</td>\n",
+       "      <td>0.107110</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0.218885</td>\n",
+       "      <td>mpfi</td>\n",
+       "      <td>-0.495180</td>\n",
+       "      <td>0.468224</td>\n",
+       "      <td>-0.540725</td>\n",
+       "      <td>0.206427</td>\n",
+       "      <td>0.76817</td>\n",
+       "      <td>-1.273900</td>\n",
+       "      <td>no</td>\n",
+       "      <td>0.227271</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>5 rows × 22 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   symboling  normalized-losses         make fuel-type aspiration  \\\n",
+       "0   1.743470            0.89088  alfa-romero       gas        std   \n",
+       "1   1.743470            0.89088  alfa-romero       gas        std   \n",
+       "2   0.133509            0.89088  alfa-romero       gas        std   \n",
+       "3   0.938490            0.89088         audi       gas        std   \n",
+       "4   0.938490            0.89088         audi       gas        std   \n",
+       "\n",
+       "  num-of-doors   body-style drive-wheels engine-location  wheel-base  ...  \\\n",
+       "0          two  convertible          rwd           front   -1.690772  ...   \n",
+       "1          two  convertible          rwd           front   -1.690772  ...   \n",
+       "2          two    hatchback          rwd           front   -0.708596  ...   \n",
+       "3         four        sedan          fwd           front    0.173698  ...   \n",
+       "4         four        sedan          4wd           front    0.107110  ...   \n",
+       "\n",
+       "  engine-size fuel-system      bore    stroke  compress-ratio  horsepower  \\\n",
+       "0    0.074449        mpfi  0.532789 -1.830840       -0.288349    0.114182   \n",
+       "1    0.074449        mpfi  0.532789 -1.830840       -0.288349    0.114182   \n",
+       "2    0.604046        mpfi -2.367552  0.691744       -0.288349    1.105818   \n",
+       "3   -0.431076        mpfi -0.495180  0.468224       -0.035973   -0.093370   \n",
+       "4    0.218885        mpfi -0.495180  0.468224       -0.540725    0.206427   \n",
+       "\n",
+       "   peak-rpm  highway-mpg  output    volume  \n",
+       "0  -0.27402    -0.546059      no -1.144195  \n",
+       "1  -0.27402    -0.546059      no -1.144195  \n",
+       "2  -0.27402    -0.691627      no -0.392670  \n",
+       "3   0.76817    -0.109354      no  0.203076  \n",
+       "4   0.76817    -1.273900      no  0.227271  \n",
+       "\n",
+       "[5 rows x 22 columns]"
+      ]
+     },
+     "execution_count": 71,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "target = data.price  #预测标签\n",
+    "\n",
+    "regressors = [x for x in data.columns if x not in ['price']]  #训练数据\n",
+    "features = data.loc[:, regressors]\n",
+    "\n",
+    "num = ['symboling','normalized-losses','volume',\n",
+    "       'horsepower','wheel-base','bore','engine-size',\n",
+    "       'stroke','compress-ratio','peak-rpm','highway-mpg']  #获取连续值特征列\n",
+    "\n",
+    "#标准化\n",
+    "standard_scaler = StandardScaler()\n",
+    "features[num] = standard_scaler.fit_transform(features[num])\n",
+    "\n",
+    "features.head()  #数值已经标准化"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 72,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# features['highway-mpg']"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "**对分类属性进行one-hot编码**\n",
+    "<br>\n",
+    "比如fuel-type里面只有gas和diesel，机器是不认识的需要转化成0/1类型的"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 73,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "In total: (205, 73)\n"
+     ]
+    },
+    {
+     "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>symboling</th>\n",
+       "      <th>normalized-losses</th>\n",
+       "      <th>wheel-base</th>\n",
+       "      <th>engine-size</th>\n",
+       "      <th>bore</th>\n",
+       "      <th>stroke</th>\n",
+       "      <th>compress-ratio</th>\n",
+       "      <th>horsepower</th>\n",
+       "      <th>peak-rpm</th>\n",
+       "      <th>highway-mpg</th>\n",
+       "      <th>...</th>\n",
+       "      <th>num-of-cylinders_twelve</th>\n",
+       "      <th>num-of-cylinders_two</th>\n",
+       "      <th>fuel-system_1bbl</th>\n",
+       "      <th>fuel-system_2bbl</th>\n",
+       "      <th>fuel-system_4bbl</th>\n",
+       "      <th>fuel-system_idi</th>\n",
+       "      <th>fuel-system_mfi</th>\n",
+       "      <th>fuel-system_mpfi</th>\n",
+       "      <th>fuel-system_spdi</th>\n",
+       "      <th>fuel-system_spfi</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>1.743470</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>-1.690772</td>\n",
+       "      <td>0.074449</td>\n",
+       "      <td>0.532789</td>\n",
+       "      <td>-1.830840</td>\n",
+       "      <td>-0.288349</td>\n",
+       "      <td>0.114182</td>\n",
+       "      <td>-0.27402</td>\n",
+       "      <td>-0.546059</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>1</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1.743470</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>-1.690772</td>\n",
+       "      <td>0.074449</td>\n",
+       "      <td>0.532789</td>\n",
+       "      <td>-1.830840</td>\n",
+       "      <td>-0.288349</td>\n",
+       "      <td>0.114182</td>\n",
+       "      <td>-0.27402</td>\n",
+       "      <td>-0.546059</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>1</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>0.133509</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>-0.708596</td>\n",
+       "      <td>0.604046</td>\n",
+       "      <td>-2.367552</td>\n",
+       "      <td>0.691744</td>\n",
+       "      <td>-0.288349</td>\n",
+       "      <td>1.105818</td>\n",
+       "      <td>-0.27402</td>\n",
+       "      <td>-0.691627</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>1</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>0.938490</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>0.173698</td>\n",
+       "      <td>-0.431076</td>\n",
+       "      <td>-0.495180</td>\n",
+       "      <td>0.468224</td>\n",
+       "      <td>-0.035973</td>\n",
+       "      <td>-0.093370</td>\n",
+       "      <td>0.76817</td>\n",
+       "      <td>-0.109354</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>1</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>0.938490</td>\n",
+       "      <td>0.89088</td>\n",
+       "      <td>0.107110</td>\n",
+       "      <td>0.218885</td>\n",
+       "      <td>-0.495180</td>\n",
+       "      <td>0.468224</td>\n",
+       "      <td>-0.540725</td>\n",
+       "      <td>0.206427</td>\n",
+       "      <td>0.76817</td>\n",
+       "      <td>-1.273900</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>1</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>5 rows × 73 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   symboling  normalized-losses  wheel-base  engine-size      bore    stroke  \\\n",
+       "0   1.743470            0.89088   -1.690772     0.074449  0.532789 -1.830840   \n",
+       "1   1.743470            0.89088   -1.690772     0.074449  0.532789 -1.830840   \n",
+       "2   0.133509            0.89088   -0.708596     0.604046 -2.367552  0.691744   \n",
+       "3   0.938490            0.89088    0.173698    -0.431076 -0.495180  0.468224   \n",
+       "4   0.938490            0.89088    0.107110     0.218885 -0.495180  0.468224   \n",
+       "\n",
+       "   compress-ratio  horsepower  peak-rpm  highway-mpg  ...  \\\n",
+       "0       -0.288349    0.114182  -0.27402    -0.546059  ...   \n",
+       "1       -0.288349    0.114182  -0.27402    -0.546059  ...   \n",
+       "2       -0.288349    1.105818  -0.27402    -0.691627  ...   \n",
+       "3       -0.035973   -0.093370   0.76817    -0.109354  ...   \n",
+       "4       -0.540725    0.206427   0.76817    -1.273900  ...   \n",
+       "\n",
+       "   num-of-cylinders_twelve  num-of-cylinders_two  fuel-system_1bbl  \\\n",
+       "0                        0                     0                 0   \n",
+       "1                        0                     0                 0   \n",
+       "2                        0                     0                 0   \n",
+       "3                        0                     0                 0   \n",
+       "4                        0                     0                 0   \n",
+       "\n",
+       "   fuel-system_2bbl  fuel-system_4bbl  fuel-system_idi  fuel-system_mfi  \\\n",
+       "0                 0                 0                0                0   \n",
+       "1                 0                 0                0                0   \n",
+       "2                 0                 0                0                0   \n",
+       "3                 0                 0                0                0   \n",
+       "4                 0                 0                0                0   \n",
+       "\n",
+       "   fuel-system_mpfi  fuel-system_spdi  fuel-system_spfi  \n",
+       "0                 1                 0                 0  \n",
+       "1                 1                 0                 0  \n",
+       "2                 1                 0                 0  \n",
+       "3                 1                 0                 0  \n",
+       "4                 1                 0                 0  \n",
+       "\n",
+       "[5 rows x 73 columns]"
+      ]
+     },
+     "execution_count": 73,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "classes = ['make','fuel-type','aspiration','engine-type',\n",
+    "           'num-of-doors','body-style','drive-wheels','output',\n",
+    "           'engine-location','num-of-cylinders','fuel-system']  #获取所需特征列\n",
+    "\n",
+    "dummies = pd.get_dummies(features[classes])  #进行one-hot\n",
+    "#加入one-hot后的数值，并去掉原有的字符串类型的数据\n",
+    "features = features.join(dummies).drop(classes, axis=1)\n",
+    "\n",
+    "print('In total:', features.shape)\n",
+    "features.head(5)  #可以看到结果已经没有字符串类型的数据了，且新增的只有0或者1"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "**划分数据集**\n",
+    "划分成训练集和验证集，训练出来的结果在验证集上测试效果"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 74,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Train (143, 73) Test: (62, 73)\n"
+     ]
+    }
+   ],
+   "source": [
+    "# tain为训练集，test为测试集\n",
+    "#test_size切分比例，训练占总数据0.7，测试0.3，random_state随机切分的种子\n",
+    "X_train, X_test, y_train, y_test = train_test_split(features,target,\n",
+    "                                                   test_size=0.3,\n",
+    "                                                   random_state=seed)\n",
+    "print(\"Train\", X_train.shape, \"Test:\", X_test.shape)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### 回归求解\n",
+    "Lass回归，多加了一个绝对值项来惩罚过大的系数，alphas=0就是最小二乘"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "为什么需要惩罚项：\n",
+    "<br>\n",
+    "假设有数据`x=[1,1,1]`,` w1[1/3,1/3,1/3]`,`w2[1/3,0,0]`\n",
+    "<br>\n",
+    "计算x和w1，数据分别相乘并相加，结果为1，\n",
+    "<br>\n",
+    "而x和w2只有第一个数据点和w1相同，后面两个为0，结果为1/3；\n",
+    "<br>\n",
+    "<br>\n",
+    "x和w1的3个数据点结果平稳，而x和w2的结果非常不平稳，\n",
+    "<br>\n",
+    "我们希望的结果是x和w1这种每个结果数据点都比较稳定的。"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "**惩罚项**\n",
+    "因为0乘以任何结果都为0，我们引入惩罚项：\n",
+    "<br>\n",
+    "原本我们的计算是 真实的y - 预测的y得到值越小越好\n",
+    "<br>\n",
+    "现在是loss = (真实的y - 预测的y) + λ(w) 越小越好"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 90,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "CV results: 0.9297963022389589 53.35375452969131\n"
+     ]
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 720x288 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "alphas = 3. ** np.arange(2,12)  #自选惩罚项参数，现在是手动调，也有自动包\n",
+    "scores = np.empty_like(alphas)  #计算不同参数的不同得分\n",
+    "\n",
+    "#第一种方法\n",
+    "for i, a in enumerate(alphas):\n",
+    "    lasso = Lasso(random_state=seed)  #指定Lasso模型\n",
+    "    lasso.set_params(alpha=a)  #指定惩罚力度\n",
+    "    lasso.fit(X_train, y_train)  #训练\n",
+    "    scores[i] = lasso.score(X_test, y_test)  #预测\n",
+    "    \n",
+    "#第二种方法，交叉验证，训练集再切分几分交叉训练并自验证，最后再拿预测集预测\n",
+    " #cv=10，平均切分10次，即10份，每次训练9份拿另外的1份验证\n",
+    "lassocv = LassoCV(cv=10, random_state=seed)\n",
+    "lassocv.fit(features, target)  #训练\n",
+    "lassocv_score = lassocv.score(features, target)  #预测\n",
+    "lassocv_alpha = lassocv.alpha_  #获取alpha值\n",
+    "\n",
+    "plt.figure(figsize=(10,4))\n",
+    "plt.plot(alphas, scores, '-ko')\n",
+    "plt.axhline(lassocv_score, color=c)\n",
+    "plt.xlabel(r'$\\alpha$')\n",
+    "plt.ylabel('CV score')\n",
+    "plt.xscale('log', basex=2)\n",
+    "sns.despine(offset=15)\n",
+    "\n",
+    "#查看不同alpha值下的得分\n",
+    "print('CV results:', lassocv_score, lassocv_alpha)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "**查看特征重要性**"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 76,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Lasso picked28features and eliminated the other45features.\n"
+     ]
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 720x288 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "coefs = pd.Series(lassocv.coef_, index=features.columns)\n",
+    "\n",
+    "print(\"Lasso picked\" + str(sum(coefs != 0)) + \\\n",
+    "      \"features and eliminated the other\"+ \\\n",
+    "     str(sum(coefs == 0)) + \"features.\")\n",
+    "\n",
+    "coefs = pd.concat([coefs.sort_values().head(5), coefs.sort_values().tail(5)])\n",
+    "\n",
+    "plt.figure(figsize=(10,4))\n",
+    "coefs.plot(kind=\"barh\", color=c)\n",
+    "plt.title(\"Coefficients\")\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 85,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "0.870697563184287"
+      ]
+     },
+     "execution_count": 85,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "#进行训练和预测，并检验得分，得分越高表明模型越好\n",
+    "#alpha选取我们手动找到的\n",
+    "model_l1 = LassoCV(alphas=alphas,cv=10,random_state=seed).fit(X_train,y_train)\n",
+    "y_pred_l1 = model_l1.predict(X_test)\n",
+    "\n",
+    "model_l1.score(X_test,y_test)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "**画出残差点**\n",
+    "<br>\n",
+    "预测值和真实值之间的差异，y轴真实值和预测值之间的差异，x轴为预测值\n",
+    "<br>\n",
+    "差异在0左右是最好的"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 93,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<matplotlib.axes._subplots.AxesSubplot at 0x1733cb98828>"
+      ]
+     },
+     "execution_count": 93,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZMAAAF2CAYAAACrlXVQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjAsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+17YcXAAAgAElEQVR4nO3de3RTZbo/8G+apGmh7UAxIZUiR8WFR1BxUUA8TquivVAKtOoRqcCMvxmR4yg4Wqdc5CJUpPYAwxRYg3qEUfAMorTAlAIWwYEiUsYZFwNeBqWHWtqmtpC29JLL/v1RE5s2bZPsZGcn+X7W6oLsJtnv3mneZ7/Pe9kKQRAEEBERiRDm7wIQEVHgYzAhIiLRGEyIiEg0BhMiIhKNwYSIiERT+bsA/tDW1oazZ89Cq9VCqVT6uzhERAHBYrHAYDBgzJgxiIiIcPhdSAaTs2fPIjs729/FICIKSDt27EBCQoLDtpAMJlqtFkDnCdHr9X4uDRFRYKipqUF2dra9Du0qJIOJLbWl1+sRHx/v59IQEQUWZ90D7IAnIiLRGEyIiEg0BhMiIhKNwYSIiERjMCEiItEYTIiISDQGEyIiEo3BhIiIRGMwISIi0RhMiChgNDZZca7ShMYmq7+LQt2E5HIqRBR4Dla04tX3jFApFTBbBCx5PAbJCZH+Lhb9iC0TIpK9xiYrXn3PiHYT0NImoN0E5L1nZAtFRhhMiEj2LjdYoFIqHLaplApcbrD4qUTUHYMJEcleXKwSZovgsM1sERAXy5vbyQWDCRHJ3uDoMCx5PAYaNTAwQgGNGljyeAwGR7MKkwt2wBNRQEhOiMT4URpcbrAgLlbJQCIzDCZEFDAGR4cxiMgUPxUiIhKNwYSIiERjMCEiItEYTIiISDQGEyIiEo3BhIiIRGMwISIi0RhMiIhINAYTIiISjcGEiIhEYzAhIiLRGEyIiEg0BhMiIhKNwYSIiERjMCEiItEYTIiISDS/BJPi4mKkp6cjPT0da9euBQCcP38eWVlZSElJwZIlS2A2mwEA1dXVyM7ORmpqKubPn4+WlhYAgNFoxFNPPYW0tDRkZ2fDYDD441CIiAh+CCatra3Iy8vDO++8g+LiYlRUVKC8vBw5OTlYtmwZDh48CEEQsGvXLgDAypUrMWvWLJSWlmLMmDHYvHkzAGDDhg1ISEjAgQMH8OijjyIvL0/qQyEioh9JHkwsFgusVitaW1thNpthNpuhUqnQ1taGsWPHAgCysrJQWloKk8mE06dPIyUlxWE7ABw9ehQZGRkAgKlTp+KTTz6ByWSS+nCIiAh+uAd8VFQUFixYgLS0NERGRmL8+PFQq9XQarX252i1WtTW1qKxsRFRUVFQqVQO2wGgrq7O/hqVSoWoqCg0NDRg6NChDvszGo0wGo0O22pqanx5iEREIUfyYPLll1/igw8+wMcff4zo6Gi8+OKLOHHiBBQKhf05giBAoVDY/+2q++OurwkL69nQ2r59OwoLC717EERE5EDyYHL8+HFMmjQJQ4YMAdCZunrrrbccOtDr6+uh0+kQGxuLpqYmWCwWKJVKGAwG6HQ6AIBOp0N9fT30ej3MZjNaWlowaNCgHvubO3cuMjMzHbbV1NQgOzvbh0dJRBRaJO8zufXWW1FeXo5r165BEAQcOXIEEyZMgEajwZkzZwB0jvZKTEyEWq1GQkICSkpKAABFRUVITEwEACQlJaGoqAgAUFJSgoSEBKjV6h77i4mJQXx8vMOPXq+X6GiJiEKD5C2Te++9F+fOnUNWVhbUajVuv/12PPXUU3jooYewdOlSNDc3Y/To0ZgzZw4AYPny5cjNzcWWLVsQFxeHdevWAQAWLFiA3NxcpKenIzo6GgUFBVIfChER/UghCILg70JIraqqCpMnT0ZZWRni4+P9XRwiooDQV93JGfBERCQagwkREYnGYEJERKIxmBARkWgMJkREJBqDCRERicZgQkREojGYEBGRaAwmREQkGoMJERGJxmBCRESiMZgQEZFoDCZERCQagwkREYnGYEJERKIxmBARkWgMJkREJBqDCRERicZgQkREojGYEBGRaAwmREQkGoMJERGJxmBCRESiMZgQEZFoDCZERCQagwkREYnGYEJERKIxmBARkWgMJkREJBqDCRERicZgQkREojGYEBGRaAwmREQkGoMJERGJxmBCRESi+SWYHDlyBFlZWUhLS8Pq1asBAOXl5cjIyEBycjLWr19vf+758+eRlZWFlJQULFmyBGazGQBQXV2N7OxspKamYv78+WhpafHHoRAREfwQTC5duoTly5dj8+bN2Lt3L86dO4djx45h8eLF2Lx5M0pKSnD27FkcO3YMAJCTk4Nly5bh4MGDEAQBu3btAgCsXLkSs2bNQmlpKcaMGYPNmzdLfShERPQjyYPJ4cOHMWXKFOj1eqjVaqxfvx6RkZEYMWIEhg8fDpVKhYyMDJSWluL7779HW1sbxo4dCwDIyspCaWkpTCYTTp8+jZSUFIftRETkHyqpd1hZWQm1Wo2nn34aly9fxn333YdbbrkFWq3W/hydTofa2lrU1dU5bNdqtaitrUVjYyOioqKgUqkctjtjNBphNBodttXU1PjgyIiIQpfkwcRisaCiogLvvPMOBgwYgPnz5yMiIgIKhcL+HEEQoFAoYLVanW63/dtV98c227dvR2FhoW8OhoiIAPghmFx33XWYNGkSYmNjAQAPPvggSktLoVQq7c8xGAzQ6XTQ6/UwGAz27fX19dDpdIiNjUVTUxMsFguUSqX9+c7MnTsXmZmZDttqamqQnZ3tg6MjIgpNkveZ3H///Th+/DiMRiMsFgv++te/IjU1Fd999x0qKythsViwf/9+JCYmYtiwYdBoNDhz5gwAoLi4GImJiVCr1UhISEBJSQkAoKioCImJiU73FxMTg/j4eIcfvV4v2fESEYUCyVsmd955J371q19h1qxZMJlM+I//+A88/vjjuOmmm/Dss8+ivb0dSUlJSE1NBQAUFBRg6dKlaG5uxujRozFnzhwAwPLly5Gbm4stW7YgLi4O69atk/pQiIjoRwpBEAR/F0JqVVVVmDx5MsrKyhAfH+/v4hARBYS+6k7OgCciItEYTIiISDQGEyIiEo3BhIiIRGMwISIi0RhMiIhINAYTIiISjcGEiIhEYzAhIiLRGEyIiEg0BhMiogDR2GTFuUoTGpus/i5KD5Iv9EhERO47WNGKV98zQqVUwGwRsOTxGCQnRPq7WHZsmRARyVxjkxWvvmdEuwloaRPQbgLy3jPKqoXCYEJEJHOXGyxQKR3vJqtSKnC5weKnEvXEYEJEJHNxsUqYLY53CzFbBMTFKnt5hfQYTIiIZG5wdBiWPB4DjRoYGKGARg0seTwGg6PlU4WzA56IKAAkJ0Ri/CgNLjdYEBerlFUgARhMiIgCxuDoMNkFERt5loqIiAIKgwkREYnGYEJERKIxmBARkWgMJkREJBqDCRERicZgQkREojGYEBGRaAwmJBk534uBiMThDHiShNzvxUBE4rBlQj4XCPdiICJxGEzI5wLhXgxEJA6DCflcINyLgYjEYTAhnwuEezEQkTjsgCdJyP1eDEQkDoMJSUbO92IgInH8+s1eu3YtcnNzAQDnz59HVlYWUlJSsGTJEpjNZgBAdXU1srOzkZqaivnz56OlpQUAYDQa8dRTTyEtLQ3Z2dkwGAx+Ow4iolDnt2By8uRJ7Nmzx/44JycHy5Ytw8GDByEIAnbt2gUAWLlyJWbNmoXS0lKMGTMGmzdvBgBs2LABCQkJOHDgAB599FHk5eX55TiIyHc40TVw+CWYXLlyBevXr8fTTz8NAPj+++/R1taGsWPHAgCysrJQWloKk8mE06dPIyUlxWE7ABw9ehQZGRkAgKlTp+KTTz6ByWTyw9EQkS8crGjFjBUGPLupETNWGHCootXfRaI++KXPZNmyZXj++edx+fJlAEBdXR20Wq3991qtFrW1tWhsbERUVBRUKpXD9u6vUalUiIqKQkNDA4YOHeqwL6PRCKPR6LCtpqbGZ8dGROJ1nejabuocVp73nhHjR2nY7yZTkgeT999/H3FxcZg0aRI+/PBDAIDVaoVC8dOkNkEQoFAo7P921f1x19eEhfX8I9u+fTsKCwu9eARE5Gu2ia62QAL8NNGVwUSeJA8mJSUlMBgMmD59Oq5evYpr165BoVA4dKDX19dDp9MhNjYWTU1NsFgsUCqVMBgM0Ol0AACdTof6+nro9XqYzWa0tLRg0KBBPfY3d+5cZGZmOmyrqalBdna2bw+UiDzGia6BR/IQ//bbb2P//v0oLi7Gc889hwceeABr1qyBRqPBmTNnAADFxcVITEyEWq1GQkICSkpKAABFRUVITEwEACQlJaGoqAhAZ4BKSEiAWq3usb+YmBjEx8c7/Oj1eomOlog8wYmugUc280wKCgqwdOlSNDc3Y/To0ZgzZw4AYPny5cjNzcWWLVsQFxeHdevWAQAWLFiA3NxcpKenIzo6GgUFBf4sPhF5GSe6BhaFIAhC/08LLlVVVZg8eTLKysoQHx/v7+IQEQWEvupOhnoiIhKNwYSIiERjMCEiItEYTIiISDQGEyIiEo3BhIiIRGMwkRhXQSWiYCSbSYuh4GBFK159zwiVUgGzRcCSx2OQnBDp72IREYnGlolEuq6C2tImoN3UuQoqWyhEFAxcDiaXLl0C0HkfkU2bNqGpqclnhQpGtlVQu7KtgkpEFOhcCibLli3DG2+8gQsXLmDp0qWoqqrC4sWLfV22oMJVUIkomLkUTM6ePYsVK1bg8OHDyMzMxJo1a/D999/7umxBhaugElEwc6kD3nbjqRMnTthvtdvW1ubTggUjroJKJL3GJiu/cxJwKZjccMMN+PWvf42qqipMmDABL7zwAkaNGuXrsgWlwdFh/IMmkghHUErHpWCyZs0aHD58GOPGjbPfsGrGjBm+LhsRkcd4H3lp9RlM/vnPf9r/P3LkSFy9ehVXr17FHXfcgW+//RajR4/2eQGJiDzB+8hLq89g8uyzz/b6O4VCgbKyMq8XiIjIGziCUlp9BpMjR45IVQ4iIq+yjaDM69ZnwlaJb7jUZ9LQ0IC9e/eipaUFgiDAarWisrIS//3f/+3r8pHEOPKFgglHUErHpWCycOFCRERE4F//+hfuuecelJeXY9y4cb4uG0mMI18oGHEEpTRcOsPV1dXYunUrEhMT8cQTT+C9997Dt99+6+uykYTkvHYYV1omkj+Xgsl1110HAPi3f/s3fP311xg6dCjMZrNPC0bS8tbaYf1V/O4GhoMVrZixwoBnNzVixgoDDlW0ulUeIpKGS2muIUOG4M0338TYsWPxhz/8AVFRUZwBH2S8MfKlvzSZu2k0zhMgChwufSNfeeUVhIeHIyEhAWPGjMHGjRvx4osv+rpsJCFna4ctzIzG5QaLS62I/tJknqTR5LjSMlNuRM653DKZM2cOACAnJwc5OTk+LRT5R9eRL19e6sCGPU0utyL6myDmyQQyuc0T4AAFot65FEwyMjKcbt+3b59XCxMoxAyflfvQW1uZ5m9sdiu91F/F70lgkNM8AVdTbnL/fIl8xaVg8vLLL9v/bzKZ8Je//AXDhw/3WaHkTMzVaaBc2XrSihgcHYaMiRHYffynvrSMuyPsz/c0MMhlnoAr5yRQPl8iX3ApmEyYMMHh8T333IOZM2di/vz5PimUXInpEA6kzmRPWhGNTVbsO+U4KGPfp234f6nR9uPzNDDIYZ5Af+ckkD5fIl/w6K+8sbERdXV13i6L7InpEJZjZ3JvPLmRl6vHNzg6DLeNUAdcBdvfOQmkz5fIFzzqM6mursZjjz3mkwLJmZgOYbl1JvfH3VaE1Mfnj76Jvs5JoH2+RN7mdp+JQqFAbGwsbr75Zp8VSq7Edgj/4qGBePtQC9SqwFh0zp30krvnRkww8GffRG/nRE6DBYj8oc9gUl1dDQCIj493+rvrr7/eN6WSMU/y/l0rP4UCyJ48AJn3DAi6isbVcyMmGPTWN3FLvBqt7YJfO+nlMliAyB/6DCbp6elQKBQQBAFtbW0YOHAglEoljEYjhgwZguPHj0tVTllx54rdWeW3/VALMu8Z4Msi+k1/50ZsR7WzUVWCAMzN/8GhxeevUVRyGCxA5A99BpPPP/8cALBs2TJMnDgR6enpAICysjJ89NFHvi9dEODd3hyJPR/O+iY6zLZ/5TuKivNPKNi59Fd99uxZeyABgMmTJ+PLL7/0eKeFhYVIT09Heno68vPzAQDl5eXIyMhAcnIy1q9fb3/u+fPnkZWVhZSUFCxZssS+wGR1dTWys7ORmpqK+fPno6WlxePy+JK3O2YDfTkPseej+6gqtRLQqB2fI7dRVFyskkKBS8HEarXi1KlT9seffPIJFApFH6/oXXl5OY4fP449e/agqKgI//znP7F//34sXrwYmzdvRklJCc6ePYtjx44B6Fy+ZdmyZTh48CAEQcCuXbsAACtXrsSsWbNQWlqKMWPGYPPmzR6Vx9c8GWbbm2ColLxxPpITIlG0Qos/PDMYf/rdkB6/l9MoKjkv7U/kTS6N5lq6dCkWLlwItVoNq7XzS1BYWOjRDrVaLXJzcxEeHg4AuPnmm3Hx4kWMGDHCPqs+IyMDpaWlGDlyJNra2jB27FgAQFZWFjZu3IhHH30Up0+fxqZNm+zbn3jiCdmuGeaNjtlgmhTnjfPRtW9CzqOomOakUOFSMElISMDHH3+Mr7/+GgAwatQoqFQuvbSHW265xf7/ixcv4sCBA3jiiSeg1Wrt23U6HWpra1FXV+ewXavVora2Fo2NjYiKirKXwbbdGaPRCKPR6LCtpqbGo7KLIbZj1t+Vkrdz/t7sqJbzKCrOP6FQ0WdEeOONN/DrX/8aq1atcprWWrp0qcc7/uabbzBv3jy89NJLUCqVuHjxov13giBAoVDAarU67Ne23fZvV72l3bZv3+5xK0oKrlbSziolk1mA8ZoVjU1Wn1aggbDmlFxHUXH+CYWKPoNJdHQ0AGDw4MFe3emZM2fw3HPPYfHixUhPT8dnn30Gg8Fg/73BYIBOp4Ner3fYXl9fD51Oh9jYWDQ1NcFisUCpVNqf78zcuXORmZnpsK2mpgbZ2dlePSZPuFNJd6+U2joEWAVgydtXfVrByzW9Fkijo+TccpJCIH1W5Lk+g8nMmTMBAL/5zW/s22pra3Hp0iUkJCR4tMPLly/jmWeewfr16zFp0iQAwJ133onvvvsOlZWViI+Px/79+/Hwww9j2LBh0Gg0OHPmDMaNG4fi4mIkJiZCrVYjISEBJSUlyMjIQFFRERITE53uLyYmBjExMR6V1Zc8qaRtldJXVSb87s0raDfB3lrxVQXv7/SaM4HQUupOri0nX/PFZ+Xr4MTg5xmXOj7ee+89VFRUYMmSJcjKykJUVBSSk5PxwgsvuL3Dt956C+3t7Xjttdfs22bOnInXXnsNzz77LNrb25GUlITU1FQAQEFBAZYuXYrm5maMHj3afpOu5cuXIzc3F1u2bEFcXBzWrVvndln8ydNKenB0GGIGhElWwcst5y/XlhL15IvPytcXEoF4oSIXLgWT999/H1u3bkVpaSkeeOABLF++HP/5n//pUTBZunRpr30te/fu7bHt1ltvxe7du3tsHzZsGN555x239y8XgbJopNxy/nJsKZFz3v6sfH0hwQsVcVwKJgqFAtdddx1OnjyJtLQ0qFQq+xBh8oyYSlrqCl6qnL8r6QW5tZSod97+rHx9IcELFXFcCibh4eF444038Nlnn2H16tXYuXMnIiPZ9BNLTCUtdaeur3P+rqYXBkeHYWFmFNZ90Ay1CrBYPZ8ESr7l7YseX19I8EJFHJeCSV5eHt566y2sXbsWP/vZz3DmzBmsXr3a12ULCb1V0q5cpQdLp6476YWDFa3YsKfZvqjj81nRzGnLmDcvenzdIpdbSjfQuBRMbrrpJrz88suorKyEIAhYvXo1WyY+FGqdgK6mF7oGHaDzuRv2NOG+OyP4hZexQJqgGurDuMVw6Uz9/e9/x4MPPoh58+ahtrYW9913H/72t7/5umwhKRTXcnI1vcBb4xLg+1s/B+qtpf3NpbOVn5+Pbdu2YdCgQdDr9cjPz0deXp6vyxYU3F3lNxQrTFcXf2ROm0i+XEpztbW1YeTIkfbHSUlJDsvEk3OepKtCtcJ0Jb3AnDaRfLkUTFQqFa5evWpf/+rbb7/1aaGCgadj1kO5wnQlt86cNpE8uRRM5s2bhyeeeAL19fX47W9/ixMnTuCVV17xddkC2ldVJoR1W3vS1THrrDD7Fiyj2IiCiUvBZOPGjSgsLMTx48chCAKeeeYZ3Hzzzb4umyy5MmT3YEUr8nYa7beTtTFbBERqFDhXaeo3SLDCJKJA4lIwiYyMhEajkcVKu/7kSh+ILb3VPZBo1MDUuyPwi9d/CJkhv0QUOlwKJq2trZg8eTL0ej0GDBhg375v3z6fFUxuXO0DcTZnIiIcWDorBqt2cN0fIgpOLgWTJUuW+LocsufqxDpno7EEAYiKlG6lXyIiqbkUTCZMmODrcsieq0N2B0eHIWNiBHYfb7Nvy7g7AqPi1SE55JeIQgMviV3k6sS6xiYr9p1qc9i279POx668nogoELnUMqFOrgzZ7SsdFupDfnkHO6LgxWDipv6G7PaXDrO91rY8SqhUqqG2eCVRqAmNmkxC/aXDDla0YsYKA57d1IgZKww4VNHq5xL7nj8Xr3R3bTQi8gxbJj7QWzqrr+HFAHySApJDaslfd7Bja4hIOgwmPuIsHdZbpbrnxDVsP9zi9UpPLpWpPxav5P28iaTFb5WE4mKVMJkdK1WTWcD2wy1eTwF9V2NG3k553BfF1ZFw3hSKS/kT+RNbJhL67Kt2WLvEEmUY8MvkgXj3yDWvpoAOVrRi9Q4jTN3qTX9OkpR6JFuoLuUfbOSQpiXX8NORiC3tYu5SwauUwH1jI7xa6dn20z2QiH1fb5DyDnb+aA2Rd4XiYJVAxpaJRHrrL2ltFzD17gh88FfHGfOeVnrO9gMA4arQq0xDfV6PN/irZcA+r8DDYCKR3tIukRoF9n/ac8b8/0uN9uhL42w/aiWw/aUhuFHv34/bWxWTO+8j5VL+wZaS8ecADn+NACTPMZhIpLc7KH789za0mxyf6+xL42pF1dt++gokUlSC3qqY5DJCLVDK5Sl/twzY5xV4GEwk1D3tAgCrdxp6PK/7l8bdisqd9I4UlaC3Kiax7+OroOnvitcXnLUMFIrOO4je/e8an+8/lG9fHagYTCTWNe1yrtIEtUqBjm7DhecmD3RpoqPYOzVKVQl6K2Uh5n18GTSDMSXjrGXQ1gH87s0rkrW62OcVWPjp+IFtiY9IjaLHF1YVBtx/Z4T9sS/nS0g1F8NbKQtP36ev5Vy8sdxKMKZkbC2D8G6Xm1LPV5JyBCCJw09IYh8ev4bpyw34TWEj5ub/gP8YHY5wFexfWqUS+MXrP9iHQfqyopKqEvTWMF1P36e3oLnnxDWvDD0N1mHIyQmRyP/1IESGO2739+RPrrcmT0xz+Yiz/PyHx1uQv6sZAOyprSN/74BKCVh//F7YOuO7ppt8lTuWMi/trZSFJ+/TW9DcdqgFHWbvpPiCNSUzKl7tMNEW8G+rK9gGOgQTBhMfcPYHP36UBus/aHb6fLOTi7yuOXdfVlR9vbe3O6y9NUzX3fdxFjTnJg/EjrJrDv1VYvs5pByGLBU5dYQH40CHYMJg4mW9/cGv/dUgqFVwOjPdGbNFgPFaZ07fVkn56gvj7L2D7QrQ2Ui67YdaHJ4T6P0cviKXVlcwDnQIJvwEvKy3/DwAWFxM8SoUgNkM5L51BdOXS7+MhLMO69U7jfj0fHtA56m7duYGaz+Hr8ihIzwYBzoEk4D/5uzbtw9TpkxBcnIyduzY4e/i9PoHPypeba+8BmiAMEUvbwBAEACL0DkUs8MMrHzXiO9qzJJ1OjoLiB1mYNFbV4JqjaTkhEgUrdDiD88MRtEKba8tL3b4ygMvAOQtoNNctbW1WL9+PT788EOEh4dj5syZmDhxIkaOHOm3MvWVY+6eLjj6j1YU7G7ut8VisQJz1v6AcLU0KSdnS+UDQGtH57/BlKfuL30YbOm+QCeXlJszwbacjrsCOpiUl5fj7rvvxqBBgwAAKSkpKC0txW9+8xu/lquvP/iulVfmvQMxdqQGc/N/QIe57/c0WQCTRZpOx+5L5XcXKnlqdvjKkxwHOvCiI8DTXHV1ddBqtfbHOp0OtbW1Ds8xGo2oqqpy+KmpqfF52VzNMd+oV2HpLMemu6KPFBjg23H+zpbK7y5U8tS8wRa5oq9JsaEkoFsmVqsVii41ryAIDo8BYPv27SgsLJS6aG7p3pI5/VU78t4zIkwBdJg6+0+68mVl7mzEjEbdOQ+ma5pNbleGvsAOX3IFR5l1CuhgotfrUVFRYX9sMBig0+kcnjN37lxkZmY6bKupqUF2drYkZexL9xyr7Q/PFly+qjLhd29egaXbqsILM6N89kfqrAIFgD/9bgha24WQygfLaY4FyRcvOjoFdDC555578Ic//AENDQ2IjIzEoUOHsGrVKofnxMTEICYmxk8l7F1/OdbB0WGIGRDW44pngEaBUcPDnb1lD550CHqyhH0wk3OHL8kDLzo6BXQNMXToUDz//POYM2cOTCYTHnnkEdxxxx3+Lla/XO3YdXbFY7E6v+LpHjjEdAiyAnUkxw5fkhd+ZwI8mABARkYGMjIy/F0Mt7iaY3X1iqd74FiYGYUNe5rdHoXUW9qNiPoX6t+ZgA8mgcidHGt/VzzOWjnrPmiGWqUA4HqHIIc2EpEYoRtG/cjdmbx9DTN2NnxVrYJbHYIc2khEYrFl4idic6y2lJSzG2xZrMDzWdHYsKfJpQ7BQBra6OqgglCfjUwkNQYTP+ha0d02Qu3267unpDLujsC+T9t6pKjuuzPCpQo1UIY2upqKY8qOSHoMJhITW9E56yPZ92kbtuX8NA8E6Ly/vKvBKhCGNro6Ao5LoBD5B4OJhLxR0fWWkmptF3DbCLXHwcrZ/T5sAclZ2aROI7maiguklJ2vMMVH/sBgIiFvVNuZshYAABrFSURBVHR9paTEBivb0Mb+ApI/0kiupuICJWXnK0zxkb/wskVC3qjouo4EG6BRIFwFLMyMxuDoMKcju5RhcGthwv5Gdvlr5JerI+BC+Z4XHJVH/sSWiYS80TfR2GRFvFaFX6cNxB//0gK1CtiwpwkDIxQYP0rTI1hdawe+umRyuaO/v9aTP9NIro6AC9XZyEzxkT8xmEhMTEVnS2EowxS41t5ZYdjuKZ/3nhFFK7RYmBmFtbuaHV63YU8Txo4Md2mhxv5aT/5OI7k6y9jbs5EDoR/C358NhTZ5fiuCXF+TEHu7RWzXFIYtkHRluwIdNTwcAzSOqS5BAObm/4BnNzX2e9vd3tJEQGeHPICQSyMdrGjFjBUGl86fP4Vyio/8jy0TGemr89RZCqOrrleglm63SbTdxbHD7Ngpb3vf7lfb3VtPn33VjhkrDA7lKlqhlf2VeleetiwCbahxqKb4yP8YTGSiv0qrt/uMDNB0znjvegW6MDPqx/W5AJMZCAsD2rvcE0WlVGDPiWvYfrilzyXwB0eH9VquohVajyZc+oOYEU7+6ocQk1YL9QUHyT8YTGSiv0rLWef9wsxojBqudqhwDla0YsOezoUezRYBT6cPxNYDLQ77MpkFbDvUgg5z/1fbgd6pK7Zl4Y9+CA7vpUDEYCITrlRa7qwgbFsxeOuBFizMjEbB7iZYfuyGsVg7hwx31VuACPROXbHBUOrVAQItrUZkw2AiE65WWn2lMHqrOOOGKKFSwiGYWLpNPegtQATCUit98UYwlLIfItBbghS6GEz84LsaM85Vds79GDQwzF5Jia20eqs4e6MMAyLCHQOEs1x9IHfqeisYStUPEegtQQpdDCYSe33XVXxwvM3+uHuFnpwQ6VGlZQsCCzN7Lj0/dLDSoQPeRgEg75c/w6h4db/LqARyp24gBcNAbwlS6GIwkdB3NWaHQAJ0ppta2sTlxp3dtnfU8HB7xXmu0gS18qcJjjYqFRAzoO9RW8GQq/dkZJQ/JykGUvAjsmEwkZBt0l9vPMmNOwsCG/Y0Y1vOEPuaXHGxSoSFAegWTKxW2NMn3srVy22muCcjoz48fg0bPuxs3Vms/hlNFcgtQQpNDCYS6m9ehie58d4mM85Z+wPC1T9VoEsej8ErO4ww/xhQlGHA0lk/pU/cydX3FjDkNqTVk9bWh8dbkP/jcjTdJ3mycqdA58uLPQYTCd2oV+GRn0dg91977zPxRqe7rX/EZHGcZLjvFS2+qur8pa2fxMbVXH1vAUOOaTJ3W1uNTVas/6C5x3ZlGEdTyYXcWr6BxNcXewwmEnvx0Z/h4Z8PdDqay5MvR/cgYPrxatq2hArwUwV62wg17v53Ta/v5c48lu4B43KDxeW5K73xdkXh7sioyw2WzlUDuqUDOZpKHuTW8g0kUlzsMZj4wY16FW7U/3TqxX6YXYNApEaBX7z+g8Pv3akMPZnHcrnBgi8vmXCt3fH57uzXFxWFuyOj4mKVPebfAMDzWdG8CvYzObZ8A4kU85cYTIJE1yDgq6GlvV3pR2oU+P2eph7PX5gZ5dJ+xVQU/bVm3BkZ1TX4KMM61zX77cNRyLx3QL/HQL7FyZziSDF/icEkCPlqaGlvV/qt7UKPL/oAjQKjhoe79L6eVhSutmbcGRnFYbnyxMmc4kgxf4nBJEh0v0L31dBSZ5VtY5O1xxfdYnXti97YZIXxWs/X91dR+DLtwWG58sPJnOL5+kKJwSQISN0x2b2y9fSL7lhuQKUENGrXXs+0R+hhq1E8X14oMZgEOG9eoYsZTeXuF91ZucNVjsu79IVpj9DEVqN88VMJcLYr9K5sV+jusN2a9jeFDZi2zIA9x1v6f5EIzsqtVinsy7v0x9Ya6nqL2oWZ0bjcYOlxy2Mi8j22TGTEk5aBN67QHe+D0mntrmYACpdHMrmbavP20vBfXuroscAl5yCQv4XSJMvgProAYmsZPLupETNWGHCootWl1zm7Qne3Y9LZhEMAKNjd5NJVftdg1NImoN3UmWrr67XeKLftfeJilfj9nma39k/ka55+pwMVWyYyILbfwxv3QTGZe263WIGvqkx9zpoHPO8M91aHKjvjSW5CcZJlcB5VgPFGv8fg6DDcNqL/juveXjvzfs8n5olJWYkptzf2T+QL3urLDCQMJjIgh8pw1v0D0e1vH8qwzgUh++OtlJU7GpusOFdpQmOT1ef777ovIlfI4TstNcnTXGfOnMGaNWtgMpkwaNAgvPrqqxg2bBiMRiNefPFFXLp0CbGxsdiwYQO0Wi06OjqwZMkSnD17FhERESgoKMDNN98MQRCQn5+Pjz/+GGFhYVi1ahXGjRsn9eF4hRwmZA2ODsPy2TFYvdOIsLDOe510XaK+P1LOAeits98X++figuQJOXynpSZ5MMnJycHmzZtx6623Yvfu3Vi9ejW2bNmCDRs2ICEhAVu3bkVRURHy8vKwYcMGvPPOO4iMjMSBAwdw+vRpLFq0CLt27cLBgwdx4cIFlJSUoLKyEvPmzUNJSQlUqsDsBpLDhCyxZZBiDkB/uWhv7j8U897kPXL4TktJ0qPr6OjAggULcOuttwIARo0ahcuXLwMAjh49ioyMDADA1KlT8cknn8BkMuHo0aOYNm0aAGD8+PFoaGhAdXU1jh07hilTpiAsLAw33ngj4uLi8Pnnn0t5OF7njf4DMaQYxig2ZSRlLjoU897kXf7+TktJ0sv48PBwTJ8+HQBgtVpRWFiIBx98EABQV1cHrVbbWSiVClFRUWhoaHDYDgBarRY1NTWoq6uDTqfrsb07o9EIo9HosM3Z80KdFOkcZ/tw98pNylx0KOa9iTzls2By4MABrFmzxmHbTTfdhG3btqGjowO5ubkwm82YN2+e09cLgoCwsDAIggCFQtFju9Vqdbq9u+3bt6OwsNBLRyV/nrQupEjnONvHyneNUCnhVgCTMhcdinlvIk/5LJikpaUhLS2tx/aWlhbMnz8fgwYNwpYtW6BWd44W0ul0qK+vh16vh9lsRktLCwYNGoShQ4eirq4ON9xwAwCgvr4eOp0Oer0edXV19ve1be9u7ty5yMzMdNhWU1OD7Oxsbx6uLHjaupBinoazfVisnT/uBjApc9Ghlvcm8pTk34ycnByMGDECGzZsQHj4T/e7SEpKQlFREQCgpKQECQkJUKvVSEpKQnFxMQCgoqICGo0G119/PRITE7Fv3z5YLBZUVlbi4sWLuP3223vsLyYmBvHx8Q4/er1emoOVkCez0G2kSOc420d37vRHSJmLDqW8N5GnJO0zOXfuHMrKyjBy5Eh7a0Gn0+GNN97AggULkJubi/T0dERHR6OgoAAAMHv2bCxbtgzp6ekIDw9Hfn4+ACA1NRVffPGFvXM+Ly8PERERUh6OrIhpXUiRznF2r3qrAJi7xA72RxAFLoUgCH1fLgahqqoqTJ48GWVlZYiPj/d3cbyiscmKGSsMDos1atRA0QqtW30nUozmsu3j9FftPQIY53AQyVdfdWdgTsqgHrzRupBinkjXfbA/gih4MJgEkUCsnHmzI6LgwGASZFg5E5E/sNYhIiLRGEwooHAFXyJ5YpqLAgZX8CWSL7ZMKCCImZRJRL7HYEIBIdRX8GV6j+SOaS5ySooJjO4I5RV8md6jQOD/WoJk52BFK2asMODZTY2YscKAQxWt/i6SX24NLAdM71GgYMuEHMj57oK2SZlfVXWuGePK/ekDnRQrOntCbi1X8j8GE3Ig18rL5rOv2kMq5RMXq0Rbh2N6r63Dv+k9pt3IGf/XDiQrcu6bCNWUj0LR92MphepnQP1jMCEHcu2baGyyovxcO5TdihHsI7ouN1igUTtGD43af8cc6qPqqHdMc1EPclsw0pZWUYYpcK3d8XdyaTX5itxainIrD8kHWybklFzuLtg1rXKt/adKbIAGsmk1+ZLcWopyKw/JB1smJGvOBgQM0CjwwiPRuOc2/48wk4LcWopyKw/JA4MJyZqztIrFKoRMILGR260F5FYe8j/+NZCsMa1CFBjYMiGv8dVENqZViOSPwYS8wtcT2ZhWIZI3fjtJNE5kIyIGExKNE9mIiMGERONENiJiMCHROOKKiNgBT14hlxFXXBqdyD8YTMhr/D3iypsjyhiUiNzDYEJBwZs39eL9Oojcx0suCgreGlHGYc5EnmEwoaDgrRFlHOZM5BkGEwoK3UeUhauAXzw00O334TBnIs8wmFDQSE6IRNEKLbIfGACFAnj3yDXMWGHAoYpWl9+Dw5yJPMMOeAo62w+3iOqIl8swZ6JAwmBCQcXZzbRsfR7uBAV/D3MmCjR++7acO3cOY8aMsT/u6OhATk4O0tLSkJmZiQsXLgAABEHA2rVrkZqaiilTpuDMmTP21/zP//wPUlNTkZKSgkOHDkl+DCQ/7PMg8g+/BJPW1lasWrUKJpPJvu2dd95BZGQkDhw4gMWLF2PRokUAgIMHD+LChQsoKSnBpk2bsGjRIpjNZnzxxRfYu3cviouLsXPnTuTn5+PKlSv+OBySEfZ5EPmHX9Jcr732GubOnYu//e1v9m1Hjx7FggULAADjx49HQ0MDqqurcezYMUyZMgVhYWG48cYbERcXh88//xynTp3CQw89BI1GA41GgwkTJuDo0aOYMWOGPw6JZIR9HkTSkzyYlJWVoa2tDampqQ7b6+rqoNVq7Y+1Wi1qampQV1cHnU7ndPvtt9/eY3t3RqMRRqPRYZuz51FwYZ8HkbR8FkwOHDiANWvWOGy76aab0NzcjG3btvV4viAIUCgUDo/DwsJgtVp73d5dWFjPymP79u0oLCwUcSRERNQfnwWTtLQ0pKWlOWx7//338cc//hHZ2dn2bdOnT8eOHTswdOhQ1NXV4YYbbgAA1NfXQ6fTQa/Xo66uzv78rtsNBoN9u8FgwI033tijHHPnzkVmZqbDtpqaGocyEBGROJLmAR599FF89NFHKC4uRnFxMQCguLgYUVFRSEpKsm+rqKiARqPB9ddfj8TEROzbtw8WiwWVlZW4ePEibr/9diQmJuLQoUNobW1FQ0MDPv30U0yaNKnHPmNiYhAfH+/wo9frpTxsIqKgJ5t5JrNnz8ayZcuQnp6O8PBw5OfnAwBSU1PxxRdfYNq0aQCAvLw8RERE4I477sC0adPwyCOPwGw247nnnsPQoUP9eQhERCFLIQiC0P/TgktVVRUmT56MsrIyxMfH+7s4REQBoa+6k8NdiIhINAYTIiISjcGEvKaxyYpzlSbeSIooBMmmA54CG29124n3jqdQxWBConnz/uuBjAGVQlnofNPJZ3irW947nojBhETjsu8MqEQMJiQal31nQCVinwl5Ragv+24LqHnd+kxC7TxQ6GIwIa8J9WXfQz2gUmhjMCHyolAPqBS6+FdPRESiMZgQEZFoDCZERCQagwkREYnGYEJERKIxmBARkWgMJkREJBqDCRERicZgQkREooXkDHiLpXMl15qaGj+XhIgocNjqTFsd2lVIBhODwQAAyM7O9nNJiIgCj8FgwIgRIxy2KQRBEHp5ftBqa2vD2bNnodVqoVR6d4nwmpoaZGdnY8eOHdDr9V59b29g+cSRc/nkXDaA5RNDLmWzWCwwGAwYM2YMIiIiHH4Xki2TiIgIJCQk+HQfer0e8fHxPt2HGCyfOHIun5zLBrB8YsihbN1bJDbsgCciItEYTIiISDQGEyIiEk25YsWKFf4uRLDRaDSYOHEiNBqNv4viFMsnjpzLJ+eyASyfGHIuGxCio7mIiMi7mOYiIiLRGEyIiEg0BhM3zJ49G+np6Zg+fTqmT5+Of/zjH9i3bx+mTJmC5ORk7Nixw/7c8vJyZGRkIDk5GevXr7dvP3/+PLKyspCSkoIlS5bAbDaLKlNzczOmTp2Kqqoqj/ZbXV2N7OxspKamYv78+WhpaQEAGI1GPPXUU0hLS0N2drZ91QCx5Vu0aBGSk5Pt5/Dw4cNeLbc7CgsLkZ6ejvT0dOTn53u1HGLPn7Oyyenc/f73v8eUKVOQnp6Ot99+W1bnrrfyyen8AcDatWuRm5vr1XPU0dGBnJwcpKWlITMzExcuXPCobB4RyCVWq1W49957BZPJZN9WU1Mj3H///UJjY6PQ0tIiZGRkCN98843Q2toqJCUlCf/3f/8nmEwm4cknnxSOHj0qCIIgpKenC59//rkgCIKwaNEiYceOHR6X6e9//7swdepUYfTo0cKlS5c82u9TTz0l7N+/XxAEQSgsLBTy8/MFQRCElStXCn/84x8FQRCEPXv2CAsWLBBdPkEQhKlTpwq1tbUOz/NmuV114sQJ4bHHHhPa29uFjo4OYc6cOcK+fftkcf6cle3QoUOyOXenTp0SZs6cKZhMJqG1tVW4//77hfPnz8vi3PVWvgsXLsjm/AmCIJSXlwsTJ04Ufve733m0r97O0Ztvvim8/PLLgiAIwmeffSY8+uijbpfNU2yZuOjbb78FADz55JOYNm0a3n33XZSXl+Puu+/GoEGDMGDAAKSkpKC0tBRffPEFRowYgeHDh0OlUiEjIwOlpaX4/vvv0dbWhrFjxwIAsrKyUFpa6nGZdu3aheXLl0On0wGA2/s1mUw4ffo0UlJSepTn6NGjyMjIAABMnToVn3zyCUwmk6jytba2orq6GosXL0ZGRgY2btwIq9Xq1XK7SqvVIjc3F+Hh4VCr1bj55ptx8eJFWZw/Z2Wrrq6WzbmbMGEC/vSnP0GlUuGHH36AxWKB0WiUxbnrrXwRERGyOX9XrlzB+vXr8fTTTwOAV8/R0aNHMW3aNADA+PHj0dDQgOrqarfK5ykGExcZjUZMmjQJmzZtwrZt2/C///u/qK6uhlartT9Hp9OhtrYWdXV1Lm3XarWora31uEx5eXkOy8K4u9/GxkZERUVBpVL1KE/X16hUKkRFRaGhoUFU+err63H33Xfj1Vdfxa5du1BRUYHdu3d7tdyuuuWWW+xf3osXL+LAgQNQKBSyOH/Oyvbzn/9cNucOANRqNTZu3Ij09HRMmjRJdn973ctnNptlc/6WLVuG559/HjExMT2OV+w5cvZeUq2OzmDiorvuugv5+fmIjo5GbGwsHnnkEWzcuBEKhcL+HEEQoFAoYLVa3druLe7u19n+eyuPIAgICxP35zJ8+HBs2rQJOp0OkZGRmD17No4dO+bTcvfnm2++wZNPPomXXnoJw4cPl9X561q2m266SXbn7rnnnsPJkydx+fJlXLx4UVbnrnv5Tp48KYvz9/777yMuLg6TJk2yb/PmOer+Gm98b13FYOKiiooKnDx50v5YEAQMGzbMoXPQYDBAp9NBr9e7tL2+vt6eAvIGd/cbGxuLpqYm+70JbM8HOq/Q6uvrAQBmsxktLS0YNGiQqPJ99dVXOHjwoP2xIAhQqVReLbc7zpw5g1/84hd44YUXkJmZKavz171scjp3Fy5cwPnz5wEAkZGRSE5OxqlTp2Rz7pyVr6SkRBbnr6SkBCdOnMD06dOxceNGHDlyBLt37/baORo6dCjq6up6vJcUGExc1NTUhPz8fLS3t6O5uRl79uzB66+/jpMnT6KhoQGtra04dOgQEhMTceedd+K7775DZWUlLBYL9u/fj8TERAwbNgwajQZnzpwBABQXFyMxMdFrZXR3v2q1GgkJCSgpKQEAFBUV2cuTlJSEoqIiAJ1fgISEBKjValHlEwQBr776Kq5evQqTyYQ///nPeOihh7xablddvnwZzzzzDAoKCpCeni6r8+esbHI6d1VVVVi6dCk6OjrQ0dGBsrIyzJw5UxbnrrfyjR8/Xhbn7+2338b+/ftRXFyM5557Dg888ADWrFnjtXOUlJSE4uJiAJ0XwBqNBtdff73L5RPFxx38QWX9+vVCamqqkJycLGzbtk0QBEHYu3evkJ6eLiQnJwtbt261P7e8vFzIyMgQkpOThby8PMFqtQqCIAjnz58XHn74YSElJUX47W9/K7S3t4su1/33328fLeXufquqqoQnnnhCSEtLE5588knhypUrgiAIQmNjozBv3jxhypQpwmOPPWZ/f7Hle/fdd4W0tDThoYceEl5//XX7c7xVbletWrVKGDt2rDBt2jT7z86dO2Vx/norm1zOnSAIwsaNG4W0tDRh6tSpwsaNG71aDm/87Tkrn5zOnyAIwgcffGAfzeWtc9TW1ia89NJLwpQpU4QZM2YIZ8+e9ahsnuByKkREJBrTXEREJBqDCRERicZgQkREojGYEBGRaAwmREQkGoMJUQC566677CswE8kJgwkREYmm8ncBiILVqVOnUFBQgOuvvx7ffvstIiIi8Nprr+GNN97AlStXcOnSJdx3331YsGABCgoKcPr0aVgsFtx2221YunQpoqKiUFFRgVWrVkGhUOD222+H1WoFALS0tGDRokWorKxEWFgYRo8ejVdeeUWydZiIuuNfHpEPnT17FrNnz8a+ffuQlZWFnJwcAEBbWxv+8pe/ICcnB1u3boVSqcSHH36IvXv3QqfToaCgAB0dHViwYAFyc3NRVFSEiRMnoq2tDQBw+PBhtLS0oLi4GLt37wYAXLp0yW/HScRgQuRDt956q30Z/ocffhjnz5/HlStXMG7cOPtzjh49iiNHjmDGjBmYPn06PvroI1y4cAFff/01VCqVfYXZqVOnYuDAgQCAcePG4V//+hdmz56NrVu3Yu7cuRgxYoT0B0j0I6a5iHxIqVT22BYWFoYBAwbYH1utVixevBhJSUkAOlNY7e3tqK6uRvfVjmz3thg+fDgOHz6MU6dO4dNPP8Uvf/lLvPLKK3jggQd8eDREvWPLhMiHvvzyS3z55ZcAgD//+c+466677DdFsrn33nuxY8cOdHR0wGq14uWXX8a6deswatQoCIKAY8eOAQDKyspw9epVAMDOnTuxaNEi3HvvvcjJycG9996Lc+fOSXtwRF0wmBD50HXXXYcNGzYgIyMDH330EfLz83s857/+678wbNgwZGZmYsqUKRAEAbm5uVCr1di0aRN+//vfY/r06Th8+DCGDBkCAJgxYwYsFgumTJmCrKwsNDU1Yfbs2VIfHpEdVw0m8pFTp05h1apV2L9/v7+LQuRzbJkQEZFobJkQEZFobJkQEZFoDCZERCQagwkREYnGYEJERKIxmBARkWgMJkREJNr/B6Vzst7Clvx2AAAAAElFTkSuQmCC\n",
+      "text/plain": [
+       "<Figure size 432x432 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plt.rcParams['figure.figsize'] = (6.0,6.0)\n",
+    "\n",
+    "preds = pd.DataFrame({'preds':model_l1.predict(X_train), 'true':y_train})\n",
+    "preds['residuals'] = preds['true'] - preds['preds']\n",
+    "preds.plot(x='preds', y='residuals', kind='scatter', color=c)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "常用的评测方法：\n",
+    "<br>\n",
+    "MSE均方误差（越小越好）\n",
+    "<br>R2决定系数（0-1之间，越大表示模型拟合效果越好）"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 94,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "MSE: 4232851.141\n",
+      "R2: 0.871\n"
+     ]
+    }
+   ],
+   "source": [
+    "def MSE(y_true, y_pred):\n",
+    "    mse = mean_squared_error(y_true, y_pred)\n",
+    "    print(\"MSE: %2.3f\" % mse)\n",
+    "    return mse\n",
+    "\n",
+    "def R2(y_true, y_pred):\n",
+    "    r2 = r2_score(y_true, y_pred)\n",
+    "    print(\"R2: %2.3f\" % r2)\n",
+    "    return r2\n",
+    "\n",
+    "MSE(y_test, y_pred_l1);R2(y_test, y_pred_l1);"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 95,
+   "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>true</th>\n",
+       "      <th>predicted</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>9279</td>\n",
+       "      <td>8053.973904</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>35056</td>\n",
+       "      <td>32509.124398</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>17075</td>\n",
+       "      <td>19477.432690</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>7898</td>\n",
+       "      <td>6878.526602</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>8058</td>\n",
+       "      <td>7740.960545</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    true     predicted\n",
+       "0   9279   8053.973904\n",
+       "1  35056  32509.124398\n",
+       "2  17075  19477.432690\n",
+       "3   7898   6878.526602\n",
+       "4   8058   7740.960545"
+      ]
+     },
+     "execution_count": 95,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# 打印真实值和预测值\n",
+    "p = {'true':list(y_test),\n",
+    "    'predicted':pd.Series(y_pred_l1)\n",
+    "    }\n",
+    "pd.DataFrame(p).head()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
    "metadata": {},
-   "outputs": [],
    "source": []
   }
  ],