{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Causal Structures - applied on the California School data set"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pydataset\n",
    "import pylab as pl\n",
    "\n",
    "from halerium import CausalStructure\n",
    "from halerium import Evaluator"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The California Test Score Data Set"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this example we apply the `CausalStructure` class to the California Test Score Data Set. The preferred data format to use with `CausalStructure` is the pandas `DataFrame`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>distcod</th>\n",
       "      <th>county</th>\n",
       "      <th>district</th>\n",
       "      <th>grspan</th>\n",
       "      <th>enrltot</th>\n",
       "      <th>teachers</th>\n",
       "      <th>calwpct</th>\n",
       "      <th>mealpct</th>\n",
       "      <th>computer</th>\n",
       "      <th>testscr</th>\n",
       "      <th>compstu</th>\n",
       "      <th>expnstu</th>\n",
       "      <th>str</th>\n",
       "      <th>avginc</th>\n",
       "      <th>elpct</th>\n",
       "      <th>readscr</th>\n",
       "      <th>mathscr</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>75119</td>\n",
       "      <td>Alameda</td>\n",
       "      <td>Sunol Glen Unified</td>\n",
       "      <td>KK-08</td>\n",
       "      <td>195</td>\n",
       "      <td>10.900000</td>\n",
       "      <td>0.510200</td>\n",
       "      <td>2.040800</td>\n",
       "      <td>67</td>\n",
       "      <td>690.799988</td>\n",
       "      <td>0.343590</td>\n",
       "      <td>6384.911133</td>\n",
       "      <td>17.889910</td>\n",
       "      <td>22.690001</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>691.599976</td>\n",
       "      <td>690.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>61499</td>\n",
       "      <td>Butte</td>\n",
       "      <td>Manzanita Elementary</td>\n",
       "      <td>KK-08</td>\n",
       "      <td>240</td>\n",
       "      <td>11.150000</td>\n",
       "      <td>15.416700</td>\n",
       "      <td>47.916698</td>\n",
       "      <td>101</td>\n",
       "      <td>661.200012</td>\n",
       "      <td>0.420833</td>\n",
       "      <td>5099.380859</td>\n",
       "      <td>21.524664</td>\n",
       "      <td>9.824000</td>\n",
       "      <td>4.583333</td>\n",
       "      <td>660.500000</td>\n",
       "      <td>661.900024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>61549</td>\n",
       "      <td>Butte</td>\n",
       "      <td>Thermalito Union Elementary</td>\n",
       "      <td>KK-08</td>\n",
       "      <td>1550</td>\n",
       "      <td>82.900002</td>\n",
       "      <td>55.032299</td>\n",
       "      <td>76.322601</td>\n",
       "      <td>169</td>\n",
       "      <td>643.599976</td>\n",
       "      <td>0.109032</td>\n",
       "      <td>5501.954590</td>\n",
       "      <td>18.697226</td>\n",
       "      <td>8.978000</td>\n",
       "      <td>30.000002</td>\n",
       "      <td>636.299988</td>\n",
       "      <td>650.900024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>61457</td>\n",
       "      <td>Butte</td>\n",
       "      <td>Golden Feather Union Elementary</td>\n",
       "      <td>KK-08</td>\n",
       "      <td>243</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>36.475399</td>\n",
       "      <td>77.049202</td>\n",
       "      <td>85</td>\n",
       "      <td>647.700012</td>\n",
       "      <td>0.349794</td>\n",
       "      <td>7101.831055</td>\n",
       "      <td>17.357143</td>\n",
       "      <td>8.978000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>651.900024</td>\n",
       "      <td>643.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>61523</td>\n",
       "      <td>Butte</td>\n",
       "      <td>Palermo Union Elementary</td>\n",
       "      <td>KK-08</td>\n",
       "      <td>1335</td>\n",
       "      <td>71.500000</td>\n",
       "      <td>33.108601</td>\n",
       "      <td>78.427002</td>\n",
       "      <td>171</td>\n",
       "      <td>640.849976</td>\n",
       "      <td>0.128090</td>\n",
       "      <td>5235.987793</td>\n",
       "      <td>18.671329</td>\n",
       "      <td>9.080333</td>\n",
       "      <td>13.857677</td>\n",
       "      <td>641.799988</td>\n",
       "      <td>639.900024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>416</th>\n",
       "      <td>68957</td>\n",
       "      <td>San Mateo</td>\n",
       "      <td>Las Lomitas Elementary</td>\n",
       "      <td>KK-08</td>\n",
       "      <td>984</td>\n",
       "      <td>59.730000</td>\n",
       "      <td>0.101600</td>\n",
       "      <td>3.556900</td>\n",
       "      <td>195</td>\n",
       "      <td>704.300049</td>\n",
       "      <td>0.198171</td>\n",
       "      <td>7290.338867</td>\n",
       "      <td>16.474134</td>\n",
       "      <td>28.716999</td>\n",
       "      <td>5.995935</td>\n",
       "      <td>700.900024</td>\n",
       "      <td>707.700012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>417</th>\n",
       "      <td>69518</td>\n",
       "      <td>Santa Clara</td>\n",
       "      <td>Los Altos Elementary</td>\n",
       "      <td>KK-08</td>\n",
       "      <td>3724</td>\n",
       "      <td>208.479996</td>\n",
       "      <td>1.074100</td>\n",
       "      <td>1.503800</td>\n",
       "      <td>721</td>\n",
       "      <td>706.750000</td>\n",
       "      <td>0.193609</td>\n",
       "      <td>5741.462891</td>\n",
       "      <td>17.862625</td>\n",
       "      <td>41.734108</td>\n",
       "      <td>4.726101</td>\n",
       "      <td>704.000000</td>\n",
       "      <td>709.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>418</th>\n",
       "      <td>72611</td>\n",
       "      <td>Ventura</td>\n",
       "      <td>Somis Union Elementary</td>\n",
       "      <td>KK-08</td>\n",
       "      <td>441</td>\n",
       "      <td>20.150000</td>\n",
       "      <td>3.563500</td>\n",
       "      <td>37.193802</td>\n",
       "      <td>45</td>\n",
       "      <td>645.000000</td>\n",
       "      <td>0.102041</td>\n",
       "      <td>4402.831543</td>\n",
       "      <td>21.885857</td>\n",
       "      <td>23.733000</td>\n",
       "      <td>24.263039</td>\n",
       "      <td>648.299988</td>\n",
       "      <td>641.700012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>419</th>\n",
       "      <td>72744</td>\n",
       "      <td>Yuba</td>\n",
       "      <td>Plumas Elementary</td>\n",
       "      <td>KK-08</td>\n",
       "      <td>101</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>11.881200</td>\n",
       "      <td>59.405899</td>\n",
       "      <td>14</td>\n",
       "      <td>672.200012</td>\n",
       "      <td>0.138614</td>\n",
       "      <td>4776.336426</td>\n",
       "      <td>20.200001</td>\n",
       "      <td>9.952000</td>\n",
       "      <td>2.970297</td>\n",
       "      <td>667.900024</td>\n",
       "      <td>676.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>420</th>\n",
       "      <td>72751</td>\n",
       "      <td>Yuba</td>\n",
       "      <td>Wheatland Elementary</td>\n",
       "      <td>KK-08</td>\n",
       "      <td>1778</td>\n",
       "      <td>93.400002</td>\n",
       "      <td>6.923500</td>\n",
       "      <td>47.571201</td>\n",
       "      <td>313</td>\n",
       "      <td>655.750000</td>\n",
       "      <td>0.176041</td>\n",
       "      <td>5993.392578</td>\n",
       "      <td>19.036402</td>\n",
       "      <td>12.502000</td>\n",
       "      <td>5.005624</td>\n",
       "      <td>660.500000</td>\n",
       "      <td>651.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>420 rows × 17 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     distcod       county                         district grspan  enrltot  \\\n",
       "1      75119      Alameda               Sunol Glen Unified  KK-08      195   \n",
       "2      61499        Butte             Manzanita Elementary  KK-08      240   \n",
       "3      61549        Butte      Thermalito Union Elementary  KK-08     1550   \n",
       "4      61457        Butte  Golden Feather Union Elementary  KK-08      243   \n",
       "5      61523        Butte         Palermo Union Elementary  KK-08     1335   \n",
       "..       ...          ...                              ...    ...      ...   \n",
       "416    68957    San Mateo           Las Lomitas Elementary  KK-08      984   \n",
       "417    69518  Santa Clara             Los Altos Elementary  KK-08     3724   \n",
       "418    72611      Ventura           Somis Union Elementary  KK-08      441   \n",
       "419    72744         Yuba                Plumas Elementary  KK-08      101   \n",
       "420    72751         Yuba             Wheatland Elementary  KK-08     1778   \n",
       "\n",
       "       teachers    calwpct    mealpct  computer     testscr   compstu  \\\n",
       "1     10.900000   0.510200   2.040800        67  690.799988  0.343590   \n",
       "2     11.150000  15.416700  47.916698       101  661.200012  0.420833   \n",
       "3     82.900002  55.032299  76.322601       169  643.599976  0.109032   \n",
       "4     14.000000  36.475399  77.049202        85  647.700012  0.349794   \n",
       "5     71.500000  33.108601  78.427002       171  640.849976  0.128090   \n",
       "..          ...        ...        ...       ...         ...       ...   \n",
       "416   59.730000   0.101600   3.556900       195  704.300049  0.198171   \n",
       "417  208.479996   1.074100   1.503800       721  706.750000  0.193609   \n",
       "418   20.150000   3.563500  37.193802        45  645.000000  0.102041   \n",
       "419    5.000000  11.881200  59.405899        14  672.200012  0.138614   \n",
       "420   93.400002   6.923500  47.571201       313  655.750000  0.176041   \n",
       "\n",
       "         expnstu        str     avginc      elpct     readscr     mathscr  \n",
       "1    6384.911133  17.889910  22.690001   0.000000  691.599976  690.000000  \n",
       "2    5099.380859  21.524664   9.824000   4.583333  660.500000  661.900024  \n",
       "3    5501.954590  18.697226   8.978000  30.000002  636.299988  650.900024  \n",
       "4    7101.831055  17.357143   8.978000   0.000000  651.900024  643.500000  \n",
       "5    5235.987793  18.671329   9.080333  13.857677  641.799988  639.900024  \n",
       "..           ...        ...        ...        ...         ...         ...  \n",
       "416  7290.338867  16.474134  28.716999   5.995935  700.900024  707.700012  \n",
       "417  5741.462891  17.862625  41.734108   4.726101  704.000000  709.500000  \n",
       "418  4402.831543  21.885857  23.733000  24.263039  648.299988  641.700012  \n",
       "419  4776.336426  20.200001   9.952000   2.970297  667.900024  676.500000  \n",
       "420  5993.392578  19.036402  12.502000   5.005624  660.500000  651.000000  \n",
       "\n",
       "[420 rows x 17 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pydataset.data(\"Caschool\")\n",
    "data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The data set relates average test scores in California schools with various data about the schools, such as the amount of students and teachers etc."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Before we start, we split the data into a training and a test set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(123)\n",
    "random_indices = np.random.choice([True, False], size=len(data), p=[0.75, 0.25])\n",
    "data_train = data.iloc[random_indices]\n",
    "data_test = data.iloc[~random_indices]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Simple structure - only inputs and outputs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The columns all represent different data about schools in california. We want to predict the average reading score 'readscr', average math score 'mathscr' and average score 'testscr'.\n",
    "We do not care for now about the details of the other columns and simply treat all other numerical columns as inputs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "outputs = {'readscr', 'mathscr', 'testscr'}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'avginc',\n",
       " 'calwpct',\n",
       " 'compstu',\n",
       " 'computer',\n",
       " 'elpct',\n",
       " 'enrltot',\n",
       " 'expnstu',\n",
       " 'mealpct',\n",
       " 'str',\n",
       " 'teachers'}"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inputs = set(data.columns[4:]) - outputs\n",
    "inputs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we define a causal structure."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "causal_structure_1 = CausalStructure([[inputs, outputs]])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We provided the training data as scaling data. The CausalStructure will use these in order to apply the correct locations and scales when building the Graph. Alternatively, we could have standardized the data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we simply train the causal_structure with the training data and test the predictive power on the test data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "causal_structure_1.train(data_train, method=\"MGVI\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>teachers</th>\n",
       "      <th>compstu</th>\n",
       "      <th>mealpct</th>\n",
       "      <th>calwpct</th>\n",
       "      <th>str</th>\n",
       "      <th>avginc</th>\n",
       "      <th>readscr</th>\n",
       "      <th>enrltot</th>\n",
       "      <th>mathscr</th>\n",
       "      <th>testscr</th>\n",
       "      <th>computer</th>\n",
       "      <th>expnstu</th>\n",
       "      <th>elpct</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>10.000000</td>\n",
       "      <td>0.143590</td>\n",
       "      <td>94.623703</td>\n",
       "      <td>12.903200</td>\n",
       "      <td>19.500000</td>\n",
       "      <td>6.577000</td>\n",
       "      <td>617.612475</td>\n",
       "      <td>195.0</td>\n",
       "      <td>625.262781</td>\n",
       "      <td>621.376790</td>\n",
       "      <td>28.0</td>\n",
       "      <td>5253.331055</td>\n",
       "      <td>68.717949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>21.000000</td>\n",
       "      <td>0.111579</td>\n",
       "      <td>91.546402</td>\n",
       "      <td>21.649500</td>\n",
       "      <td>22.619047</td>\n",
       "      <td>9.630000</td>\n",
       "      <td>627.282873</td>\n",
       "      <td>475.0</td>\n",
       "      <td>631.942011</td>\n",
       "      <td>629.668145</td>\n",
       "      <td>53.0</td>\n",
       "      <td>4542.104980</td>\n",
       "      <td>16.210526</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>17.500000</td>\n",
       "      <td>0.063380</td>\n",
       "      <td>94.932404</td>\n",
       "      <td>14.527000</td>\n",
       "      <td>16.228571</td>\n",
       "      <td>14.558000</td>\n",
       "      <td>628.064876</td>\n",
       "      <td>284.0</td>\n",
       "      <td>631.237212</td>\n",
       "      <td>629.626784</td>\n",
       "      <td>18.0</td>\n",
       "      <td>6516.533203</td>\n",
       "      <td>32.394367</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>280.000000</td>\n",
       "      <td>0.104655</td>\n",
       "      <td>81.117302</td>\n",
       "      <td>19.571699</td>\n",
       "      <td>19.178572</td>\n",
       "      <td>22.059999</td>\n",
       "      <td>632.454008</td>\n",
       "      <td>5370.0</td>\n",
       "      <td>639.352250</td>\n",
       "      <td>635.985587</td>\n",
       "      <td>562.0</td>\n",
       "      <td>4559.176758</td>\n",
       "      <td>65.512100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>221.000000</td>\n",
       "      <td>0.076092</td>\n",
       "      <td>92.743103</td>\n",
       "      <td>24.659500</td>\n",
       "      <td>19.266968</td>\n",
       "      <td>10.050500</td>\n",
       "      <td>622.697385</td>\n",
       "      <td>4258.0</td>\n",
       "      <td>630.101735</td>\n",
       "      <td>626.458898</td>\n",
       "      <td>324.0</td>\n",
       "      <td>5092.917480</td>\n",
       "      <td>36.801315</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>39.000000</td>\n",
       "      <td>0.104423</td>\n",
       "      <td>67.444702</td>\n",
       "      <td>19.778900</td>\n",
       "      <td>20.871796</td>\n",
       "      <td>15.177000</td>\n",
       "      <td>643.666348</td>\n",
       "      <td>814.0</td>\n",
       "      <td>643.284873</td>\n",
       "      <td>643.517021</td>\n",
       "      <td>85.0</td>\n",
       "      <td>4518.016113</td>\n",
       "      <td>13.759213</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>74.500000</td>\n",
       "      <td>0.151250</td>\n",
       "      <td>79.725502</td>\n",
       "      <td>27.261400</td>\n",
       "      <td>21.476511</td>\n",
       "      <td>10.472000</td>\n",
       "      <td>630.008711</td>\n",
       "      <td>1600.0</td>\n",
       "      <td>632.956701</td>\n",
       "      <td>631.485806</td>\n",
       "      <td>242.0</td>\n",
       "      <td>4720.086426</td>\n",
       "      <td>35.250000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>126.120003</td>\n",
       "      <td>0.054150</td>\n",
       "      <td>52.396400</td>\n",
       "      <td>15.459500</td>\n",
       "      <td>22.256580</td>\n",
       "      <td>12.997000</td>\n",
       "      <td>642.490168</td>\n",
       "      <td>2807.0</td>\n",
       "      <td>644.200948</td>\n",
       "      <td>643.344747</td>\n",
       "      <td>152.0</td>\n",
       "      <td>4353.019531</td>\n",
       "      <td>33.487709</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65</th>\n",
       "      <td>50.000000</td>\n",
       "      <td>0.155695</td>\n",
       "      <td>82.131699</td>\n",
       "      <td>27.168200</td>\n",
       "      <td>19.139999</td>\n",
       "      <td>9.082000</td>\n",
       "      <td>633.654592</td>\n",
       "      <td>957.0</td>\n",
       "      <td>634.547127</td>\n",
       "      <td>634.078879</td>\n",
       "      <td>149.0</td>\n",
       "      <td>5306.132812</td>\n",
       "      <td>17.659353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>209.839996</td>\n",
       "      <td>0.065438</td>\n",
       "      <td>93.922699</td>\n",
       "      <td>17.702600</td>\n",
       "      <td>20.682425</td>\n",
       "      <td>14.601625</td>\n",
       "      <td>625.472844</td>\n",
       "      <td>4340.0</td>\n",
       "      <td>632.377607</td>\n",
       "      <td>629.086134</td>\n",
       "      <td>284.0</td>\n",
       "      <td>5181.579590</td>\n",
       "      <td>31.198156</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>141.210007</td>\n",
       "      <td>0.091396</td>\n",
       "      <td>68.396400</td>\n",
       "      <td>21.258801</td>\n",
       "      <td>21.152891</td>\n",
       "      <td>13.730599</td>\n",
       "      <td>641.871966</td>\n",
       "      <td>2987.0</td>\n",
       "      <td>641.980173</td>\n",
       "      <td>642.028016</td>\n",
       "      <td>273.0</td>\n",
       "      <td>4826.666016</td>\n",
       "      <td>14.429193</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>85</th>\n",
       "      <td>26.500000</td>\n",
       "      <td>0.075949</td>\n",
       "      <td>75.738403</td>\n",
       "      <td>21.518999</td>\n",
       "      <td>17.886793</td>\n",
       "      <td>8.258000</td>\n",
       "      <td>632.714856</td>\n",
       "      <td>474.0</td>\n",
       "      <td>633.982099</td>\n",
       "      <td>633.323229</td>\n",
       "      <td>36.0</td>\n",
       "      <td>5329.002441</td>\n",
       "      <td>28.691982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>86</th>\n",
       "      <td>57.700001</td>\n",
       "      <td>0.248654</td>\n",
       "      <td>62.212799</td>\n",
       "      <td>29.446800</td>\n",
       "      <td>19.306759</td>\n",
       "      <td>8.896000</td>\n",
       "      <td>650.578864</td>\n",
       "      <td>1114.0</td>\n",
       "      <td>645.014199</td>\n",
       "      <td>647.825519</td>\n",
       "      <td>277.0</td>\n",
       "      <td>5929.809082</td>\n",
       "      <td>1.795332</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>15.000000</td>\n",
       "      <td>0.183333</td>\n",
       "      <td>72.185402</td>\n",
       "      <td>48.344398</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>11.553000</td>\n",
       "      <td>647.135993</td>\n",
       "      <td>300.0</td>\n",
       "      <td>640.133688</td>\n",
       "      <td>643.631130</td>\n",
       "      <td>55.0</td>\n",
       "      <td>5869.389160</td>\n",
       "      <td>0.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>8.000000</td>\n",
       "      <td>0.232877</td>\n",
       "      <td>54.109600</td>\n",
       "      <td>25.342501</td>\n",
       "      <td>18.250000</td>\n",
       "      <td>7.105000</td>\n",
       "      <td>654.489627</td>\n",
       "      <td>146.0</td>\n",
       "      <td>647.243790</td>\n",
       "      <td>650.944615</td>\n",
       "      <td>34.0</td>\n",
       "      <td>6231.601562</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>846.380005</td>\n",
       "      <td>0.060922</td>\n",
       "      <td>51.242100</td>\n",
       "      <td>20.523701</td>\n",
       "      <td>22.923510</td>\n",
       "      <td>14.127667</td>\n",
       "      <td>647.845877</td>\n",
       "      <td>19402.0</td>\n",
       "      <td>646.546757</td>\n",
       "      <td>647.165516</td>\n",
       "      <td>1182.0</td>\n",
       "      <td>4906.229980</td>\n",
       "      <td>18.436245</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>136.830002</td>\n",
       "      <td>0.151469</td>\n",
       "      <td>64.097702</td>\n",
       "      <td>25.829800</td>\n",
       "      <td>19.155155</td>\n",
       "      <td>13.390000</td>\n",
       "      <td>650.950101</td>\n",
       "      <td>2621.0</td>\n",
       "      <td>646.368921</td>\n",
       "      <td>648.613525</td>\n",
       "      <td>397.0</td>\n",
       "      <td>5718.551270</td>\n",
       "      <td>2.022129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>114</th>\n",
       "      <td>332.100006</td>\n",
       "      <td>0.076557</td>\n",
       "      <td>71.023804</td>\n",
       "      <td>12.279100</td>\n",
       "      <td>19.626617</td>\n",
       "      <td>14.242901</td>\n",
       "      <td>634.948600</td>\n",
       "      <td>6518.0</td>\n",
       "      <td>639.780808</td>\n",
       "      <td>637.422017</td>\n",
       "      <td>499.0</td>\n",
       "      <td>5172.019531</td>\n",
       "      <td>43.494938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>120</th>\n",
       "      <td>26.610001</td>\n",
       "      <td>0.202532</td>\n",
       "      <td>82.067497</td>\n",
       "      <td>20.675100</td>\n",
       "      <td>17.812851</td>\n",
       "      <td>11.238000</td>\n",
       "      <td>630.479794</td>\n",
       "      <td>474.0</td>\n",
       "      <td>634.182173</td>\n",
       "      <td>632.480049</td>\n",
       "      <td>96.0</td>\n",
       "      <td>4945.031738</td>\n",
       "      <td>35.443039</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>30.000000</td>\n",
       "      <td>0.266544</td>\n",
       "      <td>74.862396</td>\n",
       "      <td>20.550501</td>\n",
       "      <td>18.133333</td>\n",
       "      <td>10.056000</td>\n",
       "      <td>641.267530</td>\n",
       "      <td>544.0</td>\n",
       "      <td>641.698759</td>\n",
       "      <td>641.549939</td>\n",
       "      <td>145.0</td>\n",
       "      <td>5223.025391</td>\n",
       "      <td>8.639706</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>114.500000</td>\n",
       "      <td>0.041667</td>\n",
       "      <td>53.210602</td>\n",
       "      <td>14.414000</td>\n",
       "      <td>19.283842</td>\n",
       "      <td>9.630000</td>\n",
       "      <td>647.832903</td>\n",
       "      <td>2208.0</td>\n",
       "      <td>647.167778</td>\n",
       "      <td>647.609700</td>\n",
       "      <td>92.0</td>\n",
       "      <td>4960.515625</td>\n",
       "      <td>9.873188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>126</th>\n",
       "      <td>55.000000</td>\n",
       "      <td>0.131474</td>\n",
       "      <td>44.259701</td>\n",
       "      <td>11.322200</td>\n",
       "      <td>22.818182</td>\n",
       "      <td>15.274000</td>\n",
       "      <td>653.823708</td>\n",
       "      <td>1255.0</td>\n",
       "      <td>651.019501</td>\n",
       "      <td>652.490324</td>\n",
       "      <td>165.0</td>\n",
       "      <td>4880.040527</td>\n",
       "      <td>16.095617</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>129</th>\n",
       "      <td>98.000000</td>\n",
       "      <td>0.086646</td>\n",
       "      <td>67.751801</td>\n",
       "      <td>24.415100</td>\n",
       "      <td>20.020409</td>\n",
       "      <td>9.485000</td>\n",
       "      <td>639.821536</td>\n",
       "      <td>1962.0</td>\n",
       "      <td>639.065021</td>\n",
       "      <td>639.379896</td>\n",
       "      <td>170.0</td>\n",
       "      <td>5204.687500</td>\n",
       "      <td>18.195719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>134</th>\n",
       "      <td>39.500000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>34.078899</td>\n",
       "      <td>11.710500</td>\n",
       "      <td>19.037975</td>\n",
       "      <td>14.076000</td>\n",
       "      <td>662.788831</td>\n",
       "      <td>752.0</td>\n",
       "      <td>656.968347</td>\n",
       "      <td>659.849665</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5500.107910</td>\n",
       "      <td>0.132979</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>135</th>\n",
       "      <td>537.880005</td>\n",
       "      <td>0.184284</td>\n",
       "      <td>39.036800</td>\n",
       "      <td>7.487500</td>\n",
       "      <td>17.342157</td>\n",
       "      <td>25.487333</td>\n",
       "      <td>658.662077</td>\n",
       "      <td>9328.0</td>\n",
       "      <td>663.079096</td>\n",
       "      <td>661.096120</td>\n",
       "      <td>1719.0</td>\n",
       "      <td>5360.517090</td>\n",
       "      <td>50.857632</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143</th>\n",
       "      <td>391.420013</td>\n",
       "      <td>0.158389</td>\n",
       "      <td>71.912102</td>\n",
       "      <td>12.399000</td>\n",
       "      <td>21.501200</td>\n",
       "      <td>12.669900</td>\n",
       "      <td>633.088379</td>\n",
       "      <td>8416.0</td>\n",
       "      <td>639.744562</td>\n",
       "      <td>636.421265</td>\n",
       "      <td>1333.0</td>\n",
       "      <td>5065.911133</td>\n",
       "      <td>43.750000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>232.970001</td>\n",
       "      <td>0.130312</td>\n",
       "      <td>65.915001</td>\n",
       "      <td>17.476101</td>\n",
       "      <td>19.796539</td>\n",
       "      <td>12.827000</td>\n",
       "      <td>642.932220</td>\n",
       "      <td>4612.0</td>\n",
       "      <td>643.894790</td>\n",
       "      <td>643.420662</td>\n",
       "      <td>601.0</td>\n",
       "      <td>5124.836426</td>\n",
       "      <td>16.652212</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>150</th>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>35.820900</td>\n",
       "      <td>11.194000</td>\n",
       "      <td>22.166666</td>\n",
       "      <td>11.426000</td>\n",
       "      <td>659.636291</td>\n",
       "      <td>133.0</td>\n",
       "      <td>654.006918</td>\n",
       "      <td>656.901619</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5213.089355</td>\n",
       "      <td>0.751880</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>152</th>\n",
       "      <td>11.660000</td>\n",
       "      <td>0.135135</td>\n",
       "      <td>40.625000</td>\n",
       "      <td>7.589300</td>\n",
       "      <td>19.039452</td>\n",
       "      <td>14.578000</td>\n",
       "      <td>656.280216</td>\n",
       "      <td>222.0</td>\n",
       "      <td>653.613198</td>\n",
       "      <td>655.002750</td>\n",
       "      <td>30.0</td>\n",
       "      <td>4886.983887</td>\n",
       "      <td>13.513513</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>158</th>\n",
       "      <td>11.500000</td>\n",
       "      <td>0.154472</td>\n",
       "      <td>43.902401</td>\n",
       "      <td>9.756100</td>\n",
       "      <td>21.391304</td>\n",
       "      <td>11.081000</td>\n",
       "      <td>652.815764</td>\n",
       "      <td>246.0</td>\n",
       "      <td>649.519740</td>\n",
       "      <td>651.190263</td>\n",
       "      <td>38.0</td>\n",
       "      <td>5161.202148</td>\n",
       "      <td>16.666668</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>162</th>\n",
       "      <td>39.900002</td>\n",
       "      <td>0.068776</td>\n",
       "      <td>32.187099</td>\n",
       "      <td>10.178800</td>\n",
       "      <td>18.220551</td>\n",
       "      <td>9.972000</td>\n",
       "      <td>655.948312</td>\n",
       "      <td>727.0</td>\n",
       "      <td>653.918119</td>\n",
       "      <td>654.969426</td>\n",
       "      <td>50.0</td>\n",
       "      <td>4842.607910</td>\n",
       "      <td>13.067400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>166</th>\n",
       "      <td>38.000000</td>\n",
       "      <td>0.075282</td>\n",
       "      <td>35.759102</td>\n",
       "      <td>7.026300</td>\n",
       "      <td>20.973684</td>\n",
       "      <td>16.292999</td>\n",
       "      <td>661.393048</td>\n",
       "      <td>797.0</td>\n",
       "      <td>658.033962</td>\n",
       "      <td>659.672199</td>\n",
       "      <td>60.0</td>\n",
       "      <td>4674.298828</td>\n",
       "      <td>3.262233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>168</th>\n",
       "      <td>13.700000</td>\n",
       "      <td>0.217021</td>\n",
       "      <td>67.234001</td>\n",
       "      <td>6.808500</td>\n",
       "      <td>17.153284</td>\n",
       "      <td>16.622999</td>\n",
       "      <td>641.824404</td>\n",
       "      <td>235.0</td>\n",
       "      <td>642.706149</td>\n",
       "      <td>642.305920</td>\n",
       "      <td>51.0</td>\n",
       "      <td>5621.687988</td>\n",
       "      <td>39.574467</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>371.100006</td>\n",
       "      <td>0.156137</td>\n",
       "      <td>51.567402</td>\n",
       "      <td>23.197500</td>\n",
       "      <td>22.349771</td>\n",
       "      <td>12.549883</td>\n",
       "      <td>649.462177</td>\n",
       "      <td>8294.0</td>\n",
       "      <td>649.298686</td>\n",
       "      <td>649.430883</td>\n",
       "      <td>1295.0</td>\n",
       "      <td>4948.868164</td>\n",
       "      <td>10.127803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>171</th>\n",
       "      <td>8.250000</td>\n",
       "      <td>0.106667</td>\n",
       "      <td>23.225800</td>\n",
       "      <td>3.225800</td>\n",
       "      <td>18.181818</td>\n",
       "      <td>18.326000</td>\n",
       "      <td>665.447342</td>\n",
       "      <td>150.0</td>\n",
       "      <td>661.930925</td>\n",
       "      <td>663.703699</td>\n",
       "      <td>16.0</td>\n",
       "      <td>5132.788574</td>\n",
       "      <td>14.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>173</th>\n",
       "      <td>117.800003</td>\n",
       "      <td>0.148323</td>\n",
       "      <td>52.252300</td>\n",
       "      <td>17.975100</td>\n",
       "      <td>19.745331</td>\n",
       "      <td>11.426000</td>\n",
       "      <td>652.022155</td>\n",
       "      <td>2326.0</td>\n",
       "      <td>649.161762</td>\n",
       "      <td>650.526578</td>\n",
       "      <td>345.0</td>\n",
       "      <td>5149.186523</td>\n",
       "      <td>6.405847</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>174</th>\n",
       "      <td>30.500000</td>\n",
       "      <td>0.325349</td>\n",
       "      <td>54.761902</td>\n",
       "      <td>20.634899</td>\n",
       "      <td>16.426229</td>\n",
       "      <td>8.830000</td>\n",
       "      <td>652.584348</td>\n",
       "      <td>501.0</td>\n",
       "      <td>650.315801</td>\n",
       "      <td>651.378747</td>\n",
       "      <td>163.0</td>\n",
       "      <td>5373.206543</td>\n",
       "      <td>2.395210</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>175</th>\n",
       "      <td>28.270000</td>\n",
       "      <td>0.102128</td>\n",
       "      <td>89.699600</td>\n",
       "      <td>31.330500</td>\n",
       "      <td>16.625397</td>\n",
       "      <td>11.176000</td>\n",
       "      <td>636.501320</td>\n",
       "      <td>470.0</td>\n",
       "      <td>634.375519</td>\n",
       "      <td>635.416954</td>\n",
       "      <td>48.0</td>\n",
       "      <td>6485.166016</td>\n",
       "      <td>5.957447</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>184</th>\n",
       "      <td>41.150002</td>\n",
       "      <td>0.096154</td>\n",
       "      <td>50.763401</td>\n",
       "      <td>16.539400</td>\n",
       "      <td>18.955042</td>\n",
       "      <td>14.197000</td>\n",
       "      <td>655.280541</td>\n",
       "      <td>780.0</td>\n",
       "      <td>650.376220</td>\n",
       "      <td>652.903479</td>\n",
       "      <td>75.0</td>\n",
       "      <td>5261.370605</td>\n",
       "      <td>3.846154</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>186</th>\n",
       "      <td>6.750000</td>\n",
       "      <td>0.178571</td>\n",
       "      <td>45.714298</td>\n",
       "      <td>7.857100</td>\n",
       "      <td>20.740740</td>\n",
       "      <td>10.639000</td>\n",
       "      <td>652.446694</td>\n",
       "      <td>140.0</td>\n",
       "      <td>651.186929</td>\n",
       "      <td>651.756556</td>\n",
       "      <td>25.0</td>\n",
       "      <td>4566.270020</td>\n",
       "      <td>10.714286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>189</th>\n",
       "      <td>5.000000</td>\n",
       "      <td>0.231481</td>\n",
       "      <td>32.110100</td>\n",
       "      <td>3.669700</td>\n",
       "      <td>21.600000</td>\n",
       "      <td>9.665000</td>\n",
       "      <td>659.411046</td>\n",
       "      <td>108.0</td>\n",
       "      <td>658.471727</td>\n",
       "      <td>658.905142</td>\n",
       "      <td>25.0</td>\n",
       "      <td>4432.478027</td>\n",
       "      <td>4.629630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>204</th>\n",
       "      <td>15.850000</td>\n",
       "      <td>0.117264</td>\n",
       "      <td>28.664499</td>\n",
       "      <td>5.537500</td>\n",
       "      <td>19.369085</td>\n",
       "      <td>14.578000</td>\n",
       "      <td>662.314650</td>\n",
       "      <td>307.0</td>\n",
       "      <td>659.310833</td>\n",
       "      <td>660.853105</td>\n",
       "      <td>36.0</td>\n",
       "      <td>4718.163086</td>\n",
       "      <td>7.491857</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>205</th>\n",
       "      <td>17.500000</td>\n",
       "      <td>0.161383</td>\n",
       "      <td>53.602299</td>\n",
       "      <td>18.443800</td>\n",
       "      <td>19.828571</td>\n",
       "      <td>10.202000</td>\n",
       "      <td>649.437663</td>\n",
       "      <td>347.0</td>\n",
       "      <td>646.738300</td>\n",
       "      <td>648.038495</td>\n",
       "      <td>56.0</td>\n",
       "      <td>4751.299805</td>\n",
       "      <td>9.798271</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>211</th>\n",
       "      <td>61.990002</td>\n",
       "      <td>0.145261</td>\n",
       "      <td>62.267502</td>\n",
       "      <td>34.189499</td>\n",
       "      <td>18.212614</td>\n",
       "      <td>11.291000</td>\n",
       "      <td>651.567296</td>\n",
       "      <td>1129.0</td>\n",
       "      <td>644.711060</td>\n",
       "      <td>648.104441</td>\n",
       "      <td>164.0</td>\n",
       "      <td>5805.636719</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>214</th>\n",
       "      <td>229.199997</td>\n",
       "      <td>0.125406</td>\n",
       "      <td>25.888500</td>\n",
       "      <td>8.542000</td>\n",
       "      <td>21.500874</td>\n",
       "      <td>14.623000</td>\n",
       "      <td>663.465537</td>\n",
       "      <td>4928.0</td>\n",
       "      <td>660.573657</td>\n",
       "      <td>661.936855</td>\n",
       "      <td>618.0</td>\n",
       "      <td>5139.937500</td>\n",
       "      <td>5.925324</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>219</th>\n",
       "      <td>43.500000</td>\n",
       "      <td>0.110983</td>\n",
       "      <td>12.574100</td>\n",
       "      <td>3.084200</td>\n",
       "      <td>19.885057</td>\n",
       "      <td>14.578000</td>\n",
       "      <td>669.613275</td>\n",
       "      <td>865.0</td>\n",
       "      <td>666.541670</td>\n",
       "      <td>668.121803</td>\n",
       "      <td>96.0</td>\n",
       "      <td>4797.503418</td>\n",
       "      <td>3.352601</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>222</th>\n",
       "      <td>329.119995</td>\n",
       "      <td>0.097913</td>\n",
       "      <td>36.889999</td>\n",
       "      <td>12.678500</td>\n",
       "      <td>19.363758</td>\n",
       "      <td>14.597667</td>\n",
       "      <td>658.688447</td>\n",
       "      <td>6373.0</td>\n",
       "      <td>658.211347</td>\n",
       "      <td>658.471972</td>\n",
       "      <td>624.0</td>\n",
       "      <td>4733.746582</td>\n",
       "      <td>6.260787</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>224</th>\n",
       "      <td>138.119995</td>\n",
       "      <td>0.066138</td>\n",
       "      <td>34.447102</td>\n",
       "      <td>11.126400</td>\n",
       "      <td>21.017956</td>\n",
       "      <td>16.271999</td>\n",
       "      <td>663.085644</td>\n",
       "      <td>2903.0</td>\n",
       "      <td>657.967585</td>\n",
       "      <td>660.539652</td>\n",
       "      <td>192.0</td>\n",
       "      <td>5431.177246</td>\n",
       "      <td>3.169135</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>225</th>\n",
       "      <td>29.650000</td>\n",
       "      <td>0.106195</td>\n",
       "      <td>36.637199</td>\n",
       "      <td>6.725700</td>\n",
       "      <td>19.055649</td>\n",
       "      <td>13.630000</td>\n",
       "      <td>657.091211</td>\n",
       "      <td>565.0</td>\n",
       "      <td>653.529834</td>\n",
       "      <td>655.365003</td>\n",
       "      <td>60.0</td>\n",
       "      <td>5382.088867</td>\n",
       "      <td>16.991150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>234</th>\n",
       "      <td>188.199997</td>\n",
       "      <td>0.119102</td>\n",
       "      <td>49.426899</td>\n",
       "      <td>9.770800</td>\n",
       "      <td>18.692881</td>\n",
       "      <td>22.473000</td>\n",
       "      <td>653.031523</td>\n",
       "      <td>3518.0</td>\n",
       "      <td>652.258787</td>\n",
       "      <td>652.659949</td>\n",
       "      <td>419.0</td>\n",
       "      <td>5642.832031</td>\n",
       "      <td>40.108017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>235</th>\n",
       "      <td>924.570007</td>\n",
       "      <td>0.124443</td>\n",
       "      <td>50.970798</td>\n",
       "      <td>18.931900</td>\n",
       "      <td>20.868078</td>\n",
       "      <td>15.684654</td>\n",
       "      <td>652.773767</td>\n",
       "      <td>19294.0</td>\n",
       "      <td>653.725935</td>\n",
       "      <td>653.493526</td>\n",
       "      <td>2401.0</td>\n",
       "      <td>5280.023438</td>\n",
       "      <td>16.663212</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>238</th>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.358974</td>\n",
       "      <td>36.134499</td>\n",
       "      <td>8.403400</td>\n",
       "      <td>19.500000</td>\n",
       "      <td>16.356001</td>\n",
       "      <td>668.170509</td>\n",
       "      <td>117.0</td>\n",
       "      <td>662.806404</td>\n",
       "      <td>665.457846</td>\n",
       "      <td>42.0</td>\n",
       "      <td>6039.985840</td>\n",
       "      <td>0.854701</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>243</th>\n",
       "      <td>242.500000</td>\n",
       "      <td>0.091066</td>\n",
       "      <td>37.291100</td>\n",
       "      <td>8.318900</td>\n",
       "      <td>21.463917</td>\n",
       "      <td>15.749917</td>\n",
       "      <td>653.984485</td>\n",
       "      <td>5205.0</td>\n",
       "      <td>653.846936</td>\n",
       "      <td>653.818800</td>\n",
       "      <td>474.0</td>\n",
       "      <td>4954.466309</td>\n",
       "      <td>22.939482</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>247</th>\n",
       "      <td>153.199997</td>\n",
       "      <td>0.103080</td>\n",
       "      <td>26.943300</td>\n",
       "      <td>3.530500</td>\n",
       "      <td>20.770235</td>\n",
       "      <td>12.900000</td>\n",
       "      <td>660.471034</td>\n",
       "      <td>3182.0</td>\n",
       "      <td>659.690492</td>\n",
       "      <td>659.956172</td>\n",
       "      <td>328.0</td>\n",
       "      <td>4777.512695</td>\n",
       "      <td>8.956632</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>249</th>\n",
       "      <td>588.840027</td>\n",
       "      <td>0.127119</td>\n",
       "      <td>32.765499</td>\n",
       "      <td>9.015800</td>\n",
       "      <td>20.132803</td>\n",
       "      <td>17.435200</td>\n",
       "      <td>659.251062</td>\n",
       "      <td>11855.0</td>\n",
       "      <td>659.739467</td>\n",
       "      <td>659.553594</td>\n",
       "      <td>1507.0</td>\n",
       "      <td>5486.029785</td>\n",
       "      <td>20.725431</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>250</th>\n",
       "      <td>51.669998</td>\n",
       "      <td>0.120787</td>\n",
       "      <td>36.797798</td>\n",
       "      <td>12.827700</td>\n",
       "      <td>20.669636</td>\n",
       "      <td>14.760000</td>\n",
       "      <td>663.553877</td>\n",
       "      <td>1068.0</td>\n",
       "      <td>656.782451</td>\n",
       "      <td>659.998883</td>\n",
       "      <td>129.0</td>\n",
       "      <td>5486.382324</td>\n",
       "      <td>0.187266</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>256</th>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.193277</td>\n",
       "      <td>33.599998</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>18.326000</td>\n",
       "      <td>660.448138</td>\n",
       "      <td>119.0</td>\n",
       "      <td>656.326883</td>\n",
       "      <td>658.311479</td>\n",
       "      <td>23.0</td>\n",
       "      <td>5990.794434</td>\n",
       "      <td>28.571430</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>258</th>\n",
       "      <td>27.600000</td>\n",
       "      <td>0.109890</td>\n",
       "      <td>54.761902</td>\n",
       "      <td>13.919400</td>\n",
       "      <td>19.782608</td>\n",
       "      <td>10.098000</td>\n",
       "      <td>646.835857</td>\n",
       "      <td>546.0</td>\n",
       "      <td>645.633578</td>\n",
       "      <td>646.294001</td>\n",
       "      <td>60.0</td>\n",
       "      <td>4777.769043</td>\n",
       "      <td>13.736264</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>263</th>\n",
       "      <td>115.300003</td>\n",
       "      <td>0.129892</td>\n",
       "      <td>15.655900</td>\n",
       "      <td>2.881700</td>\n",
       "      <td>20.164787</td>\n",
       "      <td>21.110500</td>\n",
       "      <td>667.167368</td>\n",
       "      <td>2325.0</td>\n",
       "      <td>666.148685</td>\n",
       "      <td>666.576950</td>\n",
       "      <td>302.0</td>\n",
       "      <td>4890.986816</td>\n",
       "      <td>22.709679</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>281</th>\n",
       "      <td>46.650002</td>\n",
       "      <td>0.127737</td>\n",
       "      <td>25.605499</td>\n",
       "      <td>8.073800</td>\n",
       "      <td>17.620579</td>\n",
       "      <td>14.163000</td>\n",
       "      <td>667.893071</td>\n",
       "      <td>822.0</td>\n",
       "      <td>661.581156</td>\n",
       "      <td>664.708583</td>\n",
       "      <td>105.0</td>\n",
       "      <td>5784.625000</td>\n",
       "      <td>0.851582</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>284</th>\n",
       "      <td>27.350000</td>\n",
       "      <td>0.112621</td>\n",
       "      <td>26.213600</td>\n",
       "      <td>16.116501</td>\n",
       "      <td>18.829981</td>\n",
       "      <td>14.209000</td>\n",
       "      <td>666.968963</td>\n",
       "      <td>515.0</td>\n",
       "      <td>659.846349</td>\n",
       "      <td>663.479007</td>\n",
       "      <td>58.0</td>\n",
       "      <td>5190.310059</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>287</th>\n",
       "      <td>46.000000</td>\n",
       "      <td>0.154313</td>\n",
       "      <td>55.285500</td>\n",
       "      <td>13.851800</td>\n",
       "      <td>17.891304</td>\n",
       "      <td>10.551000</td>\n",
       "      <td>652.062665</td>\n",
       "      <td>823.0</td>\n",
       "      <td>648.806330</td>\n",
       "      <td>650.402278</td>\n",
       "      <td>127.0</td>\n",
       "      <td>5331.920898</td>\n",
       "      <td>2.187120</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>288</th>\n",
       "      <td>114.300003</td>\n",
       "      <td>0.115643</td>\n",
       "      <td>13.369200</td>\n",
       "      <td>2.916100</td>\n",
       "      <td>19.518810</td>\n",
       "      <td>16.955999</td>\n",
       "      <td>669.367531</td>\n",
       "      <td>2231.0</td>\n",
       "      <td>665.806105</td>\n",
       "      <td>667.439416</td>\n",
       "      <td>258.0</td>\n",
       "      <td>5599.607910</td>\n",
       "      <td>13.626176</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>290</th>\n",
       "      <td>15.500000</td>\n",
       "      <td>0.275081</td>\n",
       "      <td>32.686100</td>\n",
       "      <td>4.207100</td>\n",
       "      <td>19.935484</td>\n",
       "      <td>16.271999</td>\n",
       "      <td>665.115102</td>\n",
       "      <td>309.0</td>\n",
       "      <td>662.837178</td>\n",
       "      <td>664.027088</td>\n",
       "      <td>85.0</td>\n",
       "      <td>4700.432129</td>\n",
       "      <td>1.941748</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>292</th>\n",
       "      <td>47.959999</td>\n",
       "      <td>0.118012</td>\n",
       "      <td>47.305401</td>\n",
       "      <td>20.958099</td>\n",
       "      <td>20.141785</td>\n",
       "      <td>11.834000</td>\n",
       "      <td>656.883658</td>\n",
       "      <td>966.0</td>\n",
       "      <td>650.485988</td>\n",
       "      <td>653.693831</td>\n",
       "      <td>114.0</td>\n",
       "      <td>5318.240723</td>\n",
       "      <td>0.621118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>294</th>\n",
       "      <td>35.500000</td>\n",
       "      <td>0.070866</td>\n",
       "      <td>30.373199</td>\n",
       "      <td>10.810800</td>\n",
       "      <td>21.464788</td>\n",
       "      <td>12.431000</td>\n",
       "      <td>663.358136</td>\n",
       "      <td>762.0</td>\n",
       "      <td>657.475610</td>\n",
       "      <td>660.403959</td>\n",
       "      <td>54.0</td>\n",
       "      <td>5143.190430</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>296</th>\n",
       "      <td>489.299988</td>\n",
       "      <td>0.112995</td>\n",
       "      <td>30.643700</td>\n",
       "      <td>3.335000</td>\n",
       "      <td>20.130800</td>\n",
       "      <td>20.875750</td>\n",
       "      <td>661.465628</td>\n",
       "      <td>9850.0</td>\n",
       "      <td>663.110579</td>\n",
       "      <td>662.196270</td>\n",
       "      <td>1113.0</td>\n",
       "      <td>5081.455566</td>\n",
       "      <td>18.558376</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>297</th>\n",
       "      <td>5.000000</td>\n",
       "      <td>0.077519</td>\n",
       "      <td>50.387600</td>\n",
       "      <td>9.302300</td>\n",
       "      <td>25.799999</td>\n",
       "      <td>10.639000</td>\n",
       "      <td>648.618780</td>\n",
       "      <td>129.0</td>\n",
       "      <td>648.549217</td>\n",
       "      <td>648.603046</td>\n",
       "      <td>10.0</td>\n",
       "      <td>4016.416260</td>\n",
       "      <td>6.201550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>298</th>\n",
       "      <td>565.510010</td>\n",
       "      <td>0.108296</td>\n",
       "      <td>28.376600</td>\n",
       "      <td>2.538400</td>\n",
       "      <td>18.777740</td>\n",
       "      <td>25.029619</td>\n",
       "      <td>666.036055</td>\n",
       "      <td>10619.0</td>\n",
       "      <td>667.101409</td>\n",
       "      <td>666.739172</td>\n",
       "      <td>1150.0</td>\n",
       "      <td>5374.050293</td>\n",
       "      <td>18.457481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>306</th>\n",
       "      <td>33.700001</td>\n",
       "      <td>0.128889</td>\n",
       "      <td>31.851900</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>20.029673</td>\n",
       "      <td>21.957001</td>\n",
       "      <td>663.041307</td>\n",
       "      <td>675.0</td>\n",
       "      <td>662.365592</td>\n",
       "      <td>662.682382</td>\n",
       "      <td>87.0</td>\n",
       "      <td>4543.305176</td>\n",
       "      <td>14.962962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>316</th>\n",
       "      <td>10.880000</td>\n",
       "      <td>0.237500</td>\n",
       "      <td>53.503201</td>\n",
       "      <td>13.375800</td>\n",
       "      <td>14.705882</td>\n",
       "      <td>11.826000</td>\n",
       "      <td>657.115402</td>\n",
       "      <td>160.0</td>\n",
       "      <td>650.725460</td>\n",
       "      <td>653.880704</td>\n",
       "      <td>38.0</td>\n",
       "      <td>6870.346191</td>\n",
       "      <td>2.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>318</th>\n",
       "      <td>108.750000</td>\n",
       "      <td>0.135123</td>\n",
       "      <td>16.065100</td>\n",
       "      <td>4.049300</td>\n",
       "      <td>20.891954</td>\n",
       "      <td>17.332001</td>\n",
       "      <td>671.475950</td>\n",
       "      <td>2272.0</td>\n",
       "      <td>667.167409</td>\n",
       "      <td>669.243356</td>\n",
       "      <td>307.0</td>\n",
       "      <td>5182.365723</td>\n",
       "      <td>1.320423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>323</th>\n",
       "      <td>245.179993</td>\n",
       "      <td>0.137522</td>\n",
       "      <td>23.642000</td>\n",
       "      <td>7.037000</td>\n",
       "      <td>18.892242</td>\n",
       "      <td>20.469999</td>\n",
       "      <td>667.193966</td>\n",
       "      <td>4632.0</td>\n",
       "      <td>664.397837</td>\n",
       "      <td>665.754992</td>\n",
       "      <td>637.0</td>\n",
       "      <td>5666.713867</td>\n",
       "      <td>17.854059</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>328</th>\n",
       "      <td>5.500000</td>\n",
       "      <td>0.084746</td>\n",
       "      <td>51.666698</td>\n",
       "      <td>8.333300</td>\n",
       "      <td>21.454546</td>\n",
       "      <td>8.934000</td>\n",
       "      <td>648.312381</td>\n",
       "      <td>118.0</td>\n",
       "      <td>647.954079</td>\n",
       "      <td>648.068150</td>\n",
       "      <td>10.0</td>\n",
       "      <td>4451.256836</td>\n",
       "      <td>7.627119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>330</th>\n",
       "      <td>232.750000</td>\n",
       "      <td>0.123785</td>\n",
       "      <td>19.372499</td>\n",
       "      <td>2.607800</td>\n",
       "      <td>20.339420</td>\n",
       "      <td>18.827230</td>\n",
       "      <td>666.844189</td>\n",
       "      <td>4734.0</td>\n",
       "      <td>666.145109</td>\n",
       "      <td>666.526671</td>\n",
       "      <td>586.0</td>\n",
       "      <td>4922.939941</td>\n",
       "      <td>10.857626</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>332</th>\n",
       "      <td>286.920013</td>\n",
       "      <td>0.089182</td>\n",
       "      <td>30.085100</td>\n",
       "      <td>5.613400</td>\n",
       "      <td>21.103443</td>\n",
       "      <td>23.667376</td>\n",
       "      <td>663.895376</td>\n",
       "      <td>6055.0</td>\n",
       "      <td>663.783643</td>\n",
       "      <td>663.791609</td>\n",
       "      <td>540.0</td>\n",
       "      <td>4829.717285</td>\n",
       "      <td>14.236169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>334</th>\n",
       "      <td>139.300003</td>\n",
       "      <td>0.219564</td>\n",
       "      <td>23.098900</td>\n",
       "      <td>5.033900</td>\n",
       "      <td>20.107681</td>\n",
       "      <td>20.546000</td>\n",
       "      <td>671.557943</td>\n",
       "      <td>2801.0</td>\n",
       "      <td>668.080271</td>\n",
       "      <td>669.752223</td>\n",
       "      <td>615.0</td>\n",
       "      <td>5205.941895</td>\n",
       "      <td>2.070689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>338</th>\n",
       "      <td>65.900002</td>\n",
       "      <td>0.095710</td>\n",
       "      <td>46.122101</td>\n",
       "      <td>22.194700</td>\n",
       "      <td>18.391502</td>\n",
       "      <td>10.551000</td>\n",
       "      <td>654.703879</td>\n",
       "      <td>1212.0</td>\n",
       "      <td>650.066687</td>\n",
       "      <td>652.415548</td>\n",
       "      <td>116.0</td>\n",
       "      <td>5090.587891</td>\n",
       "      <td>3.960396</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>341</th>\n",
       "      <td>288.630005</td>\n",
       "      <td>0.148634</td>\n",
       "      <td>15.512800</td>\n",
       "      <td>2.163500</td>\n",
       "      <td>21.678272</td>\n",
       "      <td>21.095751</td>\n",
       "      <td>669.614810</td>\n",
       "      <td>6257.0</td>\n",
       "      <td>669.478912</td>\n",
       "      <td>669.555534</td>\n",
       "      <td>930.0</td>\n",
       "      <td>4889.479980</td>\n",
       "      <td>8.726227</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>342</th>\n",
       "      <td>45.000000</td>\n",
       "      <td>0.140553</td>\n",
       "      <td>19.009199</td>\n",
       "      <td>4.723500</td>\n",
       "      <td>19.288889</td>\n",
       "      <td>15.167000</td>\n",
       "      <td>669.900481</td>\n",
       "      <td>868.0</td>\n",
       "      <td>665.261388</td>\n",
       "      <td>667.561157</td>\n",
       "      <td>122.0</td>\n",
       "      <td>5118.130371</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>349</th>\n",
       "      <td>70.500000</td>\n",
       "      <td>0.126033</td>\n",
       "      <td>10.414200</td>\n",
       "      <td>2.426000</td>\n",
       "      <td>20.595745</td>\n",
       "      <td>17.656000</td>\n",
       "      <td>673.992356</td>\n",
       "      <td>1452.0</td>\n",
       "      <td>669.820784</td>\n",
       "      <td>672.024306</td>\n",
       "      <td>183.0</td>\n",
       "      <td>5028.561523</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>350</th>\n",
       "      <td>8.000000</td>\n",
       "      <td>0.206452</td>\n",
       "      <td>12.258100</td>\n",
       "      <td>0.645200</td>\n",
       "      <td>19.375000</td>\n",
       "      <td>17.709000</td>\n",
       "      <td>674.467598</td>\n",
       "      <td>155.0</td>\n",
       "      <td>668.955346</td>\n",
       "      <td>671.693862</td>\n",
       "      <td>32.0</td>\n",
       "      <td>5695.466797</td>\n",
       "      <td>4.516129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>352</th>\n",
       "      <td>30.080000</td>\n",
       "      <td>0.181658</td>\n",
       "      <td>29.021000</td>\n",
       "      <td>4.895100</td>\n",
       "      <td>18.849733</td>\n",
       "      <td>18.593000</td>\n",
       "      <td>668.890252</td>\n",
       "      <td>567.0</td>\n",
       "      <td>663.800812</td>\n",
       "      <td>666.315450</td>\n",
       "      <td>103.0</td>\n",
       "      <td>5156.437012</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>371</th>\n",
       "      <td>25.500000</td>\n",
       "      <td>0.111359</td>\n",
       "      <td>48.484798</td>\n",
       "      <td>12.727300</td>\n",
       "      <td>17.607843</td>\n",
       "      <td>12.934000</td>\n",
       "      <td>658.155302</td>\n",
       "      <td>449.0</td>\n",
       "      <td>651.847919</td>\n",
       "      <td>654.937895</td>\n",
       "      <td>50.0</td>\n",
       "      <td>5884.432129</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>375</th>\n",
       "      <td>5.100000</td>\n",
       "      <td>0.172840</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.614500</td>\n",
       "      <td>15.882353</td>\n",
       "      <td>22.528999</td>\n",
       "      <td>687.413953</td>\n",
       "      <td>81.0</td>\n",
       "      <td>677.710248</td>\n",
       "      <td>682.690134</td>\n",
       "      <td>14.0</td>\n",
       "      <td>7667.571777</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>377</th>\n",
       "      <td>108.959999</td>\n",
       "      <td>0.113776</td>\n",
       "      <td>3.730200</td>\n",
       "      <td>1.941700</td>\n",
       "      <td>17.988253</td>\n",
       "      <td>24.603001</td>\n",
       "      <td>681.105457</td>\n",
       "      <td>1960.0</td>\n",
       "      <td>677.414093</td>\n",
       "      <td>679.339899</td>\n",
       "      <td>223.0</td>\n",
       "      <td>5368.502930</td>\n",
       "      <td>1.479592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>379</th>\n",
       "      <td>49.169998</td>\n",
       "      <td>0.327696</td>\n",
       "      <td>6.535300</td>\n",
       "      <td>2.178400</td>\n",
       "      <td>19.239374</td>\n",
       "      <td>33.455002</td>\n",
       "      <td>687.972117</td>\n",
       "      <td>946.0</td>\n",
       "      <td>685.793505</td>\n",
       "      <td>686.758283</td>\n",
       "      <td>310.0</td>\n",
       "      <td>5387.458984</td>\n",
       "      <td>1.479915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>382</th>\n",
       "      <td>46.000000</td>\n",
       "      <td>0.228571</td>\n",
       "      <td>10.651500</td>\n",
       "      <td>4.239900</td>\n",
       "      <td>20.543478</td>\n",
       "      <td>19.346500</td>\n",
       "      <td>675.490470</td>\n",
       "      <td>945.0</td>\n",
       "      <td>671.170728</td>\n",
       "      <td>673.307374</td>\n",
       "      <td>216.0</td>\n",
       "      <td>5088.708008</td>\n",
       "      <td>2.539683</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>384</th>\n",
       "      <td>10.510000</td>\n",
       "      <td>0.213415</td>\n",
       "      <td>12.941200</td>\n",
       "      <td>2.352900</td>\n",
       "      <td>15.604186</td>\n",
       "      <td>22.863335</td>\n",
       "      <td>676.657773</td>\n",
       "      <td>164.0</td>\n",
       "      <td>670.484640</td>\n",
       "      <td>673.555717</td>\n",
       "      <td>35.0</td>\n",
       "      <td>6588.430664</td>\n",
       "      <td>11.585366</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>388</th>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.201493</td>\n",
       "      <td>3.731300</td>\n",
       "      <td>0.746300</td>\n",
       "      <td>22.333334</td>\n",
       "      <td>21.957001</td>\n",
       "      <td>679.589037</td>\n",
       "      <td>134.0</td>\n",
       "      <td>675.039429</td>\n",
       "      <td>677.465932</td>\n",
       "      <td>27.0</td>\n",
       "      <td>5094.958984</td>\n",
       "      <td>2.238806</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>393</th>\n",
       "      <td>28.020000</td>\n",
       "      <td>0.062738</td>\n",
       "      <td>11.406800</td>\n",
       "      <td>3.992400</td>\n",
       "      <td>18.772305</td>\n",
       "      <td>13.567000</td>\n",
       "      <td>671.865756</td>\n",
       "      <td>526.0</td>\n",
       "      <td>666.245290</td>\n",
       "      <td>669.097893</td>\n",
       "      <td>33.0</td>\n",
       "      <td>5644.286133</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>400</th>\n",
       "      <td>4.850000</td>\n",
       "      <td>0.259259</td>\n",
       "      <td>20.481899</td>\n",
       "      <td>13.253000</td>\n",
       "      <td>16.701031</td>\n",
       "      <td>30.840000</td>\n",
       "      <td>683.225780</td>\n",
       "      <td>81.0</td>\n",
       "      <td>673.540505</td>\n",
       "      <td>678.458755</td>\n",
       "      <td>21.0</td>\n",
       "      <td>7614.379395</td>\n",
       "      <td>12.345679</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>403</th>\n",
       "      <td>146.300003</td>\n",
       "      <td>0.158930</td>\n",
       "      <td>9.677400</td>\n",
       "      <td>0.826100</td>\n",
       "      <td>17.375256</td>\n",
       "      <td>27.947750</td>\n",
       "      <td>681.855225</td>\n",
       "      <td>2542.0</td>\n",
       "      <td>676.931333</td>\n",
       "      <td>679.498758</td>\n",
       "      <td>404.0</td>\n",
       "      <td>6604.063965</td>\n",
       "      <td>10.385523</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>405</th>\n",
       "      <td>65.120003</td>\n",
       "      <td>0.256846</td>\n",
       "      <td>1.416400</td>\n",
       "      <td>0.849900</td>\n",
       "      <td>16.262285</td>\n",
       "      <td>55.327999</td>\n",
       "      <td>705.686334</td>\n",
       "      <td>1059.0</td>\n",
       "      <td>708.851129</td>\n",
       "      <td>707.182838</td>\n",
       "      <td>272.0</td>\n",
       "      <td>6460.657227</td>\n",
       "      <td>2.266289</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>413</th>\n",
       "      <td>12.330000</td>\n",
       "      <td>0.100000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.454500</td>\n",
       "      <td>17.842661</td>\n",
       "      <td>43.230000</td>\n",
       "      <td>697.051592</td>\n",
       "      <td>220.0</td>\n",
       "      <td>693.210634</td>\n",
       "      <td>695.000096</td>\n",
       "      <td>22.0</td>\n",
       "      <td>6500.449707</td>\n",
       "      <td>1.363636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>420</th>\n",
       "      <td>93.400002</td>\n",
       "      <td>0.176041</td>\n",
       "      <td>47.571201</td>\n",
       "      <td>6.923500</td>\n",
       "      <td>19.036402</td>\n",
       "      <td>12.502000</td>\n",
       "      <td>657.326803</td>\n",
       "      <td>1778.0</td>\n",
       "      <td>652.903342</td>\n",
       "      <td>655.114523</td>\n",
       "      <td>313.0</td>\n",
       "      <td>5993.392578</td>\n",
       "      <td>5.005624</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       teachers   compstu    mealpct    calwpct        str     avginc  \\\n",
       "7     10.000000  0.143590  94.623703  12.903200  19.500000   6.577000   \n",
       "22    21.000000  0.111579  91.546402  21.649500  22.619047   9.630000   \n",
       "38    17.500000  0.063380  94.932404  14.527000  16.228571  14.558000   \n",
       "39   280.000000  0.104655  81.117302  19.571699  19.178572  22.059999   \n",
       "45   221.000000  0.076092  92.743103  24.659500  19.266968  10.050500   \n",
       "48    39.000000  0.104423  67.444702  19.778900  20.871796  15.177000   \n",
       "52    74.500000  0.151250  79.725502  27.261400  21.476511  10.472000   \n",
       "59   126.120003  0.054150  52.396400  15.459500  22.256580  12.997000   \n",
       "65    50.000000  0.155695  82.131699  27.168200  19.139999   9.082000   \n",
       "67   209.839996  0.065438  93.922699  17.702600  20.682425  14.601625   \n",
       "72   141.210007  0.091396  68.396400  21.258801  21.152891  13.730599   \n",
       "85    26.500000  0.075949  75.738403  21.518999  17.886793   8.258000   \n",
       "86    57.700001  0.248654  62.212799  29.446800  19.306759   8.896000   \n",
       "92    15.000000  0.183333  72.185402  48.344398  20.000000  11.553000   \n",
       "94     8.000000  0.232877  54.109600  25.342501  18.250000   7.105000   \n",
       "107  846.380005  0.060922  51.242100  20.523701  22.923510  14.127667   \n",
       "109  136.830002  0.151469  64.097702  25.829800  19.155155  13.390000   \n",
       "114  332.100006  0.076557  71.023804  12.279100  19.626617  14.242901   \n",
       "120   26.610001  0.202532  82.067497  20.675100  17.812851  11.238000   \n",
       "121   30.000000  0.266544  74.862396  20.550501  18.133333  10.056000   \n",
       "125  114.500000  0.041667  53.210602  14.414000  19.283842   9.630000   \n",
       "126   55.000000  0.131474  44.259701  11.322200  22.818182  15.274000   \n",
       "129   98.000000  0.086646  67.751801  24.415100  20.020409   9.485000   \n",
       "134   39.500000  0.000000  34.078899  11.710500  19.037975  14.076000   \n",
       "135  537.880005  0.184284  39.036800   7.487500  17.342157  25.487333   \n",
       "143  391.420013  0.158389  71.912102  12.399000  21.501200  12.669900   \n",
       "146  232.970001  0.130312  65.915001  17.476101  19.796539  12.827000   \n",
       "150    6.000000  0.000000  35.820900  11.194000  22.166666  11.426000   \n",
       "152   11.660000  0.135135  40.625000   7.589300  19.039452  14.578000   \n",
       "158   11.500000  0.154472  43.902401   9.756100  21.391304  11.081000   \n",
       "162   39.900002  0.068776  32.187099  10.178800  18.220551   9.972000   \n",
       "166   38.000000  0.075282  35.759102   7.026300  20.973684  16.292999   \n",
       "168   13.700000  0.217021  67.234001   6.808500  17.153284  16.622999   \n",
       "169  371.100006  0.156137  51.567402  23.197500  22.349771  12.549883   \n",
       "171    8.250000  0.106667  23.225800   3.225800  18.181818  18.326000   \n",
       "173  117.800003  0.148323  52.252300  17.975100  19.745331  11.426000   \n",
       "174   30.500000  0.325349  54.761902  20.634899  16.426229   8.830000   \n",
       "175   28.270000  0.102128  89.699600  31.330500  16.625397  11.176000   \n",
       "184   41.150002  0.096154  50.763401  16.539400  18.955042  14.197000   \n",
       "186    6.750000  0.178571  45.714298   7.857100  20.740740  10.639000   \n",
       "189    5.000000  0.231481  32.110100   3.669700  21.600000   9.665000   \n",
       "204   15.850000  0.117264  28.664499   5.537500  19.369085  14.578000   \n",
       "205   17.500000  0.161383  53.602299  18.443800  19.828571  10.202000   \n",
       "211   61.990002  0.145261  62.267502  34.189499  18.212614  11.291000   \n",
       "214  229.199997  0.125406  25.888500   8.542000  21.500874  14.623000   \n",
       "219   43.500000  0.110983  12.574100   3.084200  19.885057  14.578000   \n",
       "222  329.119995  0.097913  36.889999  12.678500  19.363758  14.597667   \n",
       "224  138.119995  0.066138  34.447102  11.126400  21.017956  16.271999   \n",
       "225   29.650000  0.106195  36.637199   6.725700  19.055649  13.630000   \n",
       "234  188.199997  0.119102  49.426899   9.770800  18.692881  22.473000   \n",
       "235  924.570007  0.124443  50.970798  18.931900  20.868078  15.684654   \n",
       "238    6.000000  0.358974  36.134499   8.403400  19.500000  16.356001   \n",
       "243  242.500000  0.091066  37.291100   8.318900  21.463917  15.749917   \n",
       "247  153.199997  0.103080  26.943300   3.530500  20.770235  12.900000   \n",
       "249  588.840027  0.127119  32.765499   9.015800  20.132803  17.435200   \n",
       "250   51.669998  0.120787  36.797798  12.827700  20.669636  14.760000   \n",
       "256    7.000000  0.193277  33.599998   4.000000  17.000000  18.326000   \n",
       "258   27.600000  0.109890  54.761902  13.919400  19.782608  10.098000   \n",
       "263  115.300003  0.129892  15.655900   2.881700  20.164787  21.110500   \n",
       "281   46.650002  0.127737  25.605499   8.073800  17.620579  14.163000   \n",
       "284   27.350000  0.112621  26.213600  16.116501  18.829981  14.209000   \n",
       "287   46.000000  0.154313  55.285500  13.851800  17.891304  10.551000   \n",
       "288  114.300003  0.115643  13.369200   2.916100  19.518810  16.955999   \n",
       "290   15.500000  0.275081  32.686100   4.207100  19.935484  16.271999   \n",
       "292   47.959999  0.118012  47.305401  20.958099  20.141785  11.834000   \n",
       "294   35.500000  0.070866  30.373199  10.810800  21.464788  12.431000   \n",
       "296  489.299988  0.112995  30.643700   3.335000  20.130800  20.875750   \n",
       "297    5.000000  0.077519  50.387600   9.302300  25.799999  10.639000   \n",
       "298  565.510010  0.108296  28.376600   2.538400  18.777740  25.029619   \n",
       "306   33.700001  0.128889  31.851900   0.000000  20.029673  21.957001   \n",
       "316   10.880000  0.237500  53.503201  13.375800  14.705882  11.826000   \n",
       "318  108.750000  0.135123  16.065100   4.049300  20.891954  17.332001   \n",
       "323  245.179993  0.137522  23.642000   7.037000  18.892242  20.469999   \n",
       "328    5.500000  0.084746  51.666698   8.333300  21.454546   8.934000   \n",
       "330  232.750000  0.123785  19.372499   2.607800  20.339420  18.827230   \n",
       "332  286.920013  0.089182  30.085100   5.613400  21.103443  23.667376   \n",
       "334  139.300003  0.219564  23.098900   5.033900  20.107681  20.546000   \n",
       "338   65.900002  0.095710  46.122101  22.194700  18.391502  10.551000   \n",
       "341  288.630005  0.148634  15.512800   2.163500  21.678272  21.095751   \n",
       "342   45.000000  0.140553  19.009199   4.723500  19.288889  15.167000   \n",
       "349   70.500000  0.126033  10.414200   2.426000  20.595745  17.656000   \n",
       "350    8.000000  0.206452  12.258100   0.645200  19.375000  17.709000   \n",
       "352   30.080000  0.181658  29.021000   4.895100  18.849733  18.593000   \n",
       "371   25.500000  0.111359  48.484798  12.727300  17.607843  12.934000   \n",
       "375    5.100000  0.172840   0.000000   3.614500  15.882353  22.528999   \n",
       "377  108.959999  0.113776   3.730200   1.941700  17.988253  24.603001   \n",
       "379   49.169998  0.327696   6.535300   2.178400  19.239374  33.455002   \n",
       "382   46.000000  0.228571  10.651500   4.239900  20.543478  19.346500   \n",
       "384   10.510000  0.213415  12.941200   2.352900  15.604186  22.863335   \n",
       "388    6.000000  0.201493   3.731300   0.746300  22.333334  21.957001   \n",
       "393   28.020000  0.062738  11.406800   3.992400  18.772305  13.567000   \n",
       "400    4.850000  0.259259  20.481899  13.253000  16.701031  30.840000   \n",
       "403  146.300003  0.158930   9.677400   0.826100  17.375256  27.947750   \n",
       "405   65.120003  0.256846   1.416400   0.849900  16.262285  55.327999   \n",
       "413   12.330000  0.100000   0.000000   0.454500  17.842661  43.230000   \n",
       "420   93.400002  0.176041  47.571201   6.923500  19.036402  12.502000   \n",
       "\n",
       "        readscr  enrltot     mathscr     testscr  computer      expnstu  \\\n",
       "7    617.612475    195.0  625.262781  621.376790      28.0  5253.331055   \n",
       "22   627.282873    475.0  631.942011  629.668145      53.0  4542.104980   \n",
       "38   628.064876    284.0  631.237212  629.626784      18.0  6516.533203   \n",
       "39   632.454008   5370.0  639.352250  635.985587     562.0  4559.176758   \n",
       "45   622.697385   4258.0  630.101735  626.458898     324.0  5092.917480   \n",
       "48   643.666348    814.0  643.284873  643.517021      85.0  4518.016113   \n",
       "52   630.008711   1600.0  632.956701  631.485806     242.0  4720.086426   \n",
       "59   642.490168   2807.0  644.200948  643.344747     152.0  4353.019531   \n",
       "65   633.654592    957.0  634.547127  634.078879     149.0  5306.132812   \n",
       "67   625.472844   4340.0  632.377607  629.086134     284.0  5181.579590   \n",
       "72   641.871966   2987.0  641.980173  642.028016     273.0  4826.666016   \n",
       "85   632.714856    474.0  633.982099  633.323229      36.0  5329.002441   \n",
       "86   650.578864   1114.0  645.014199  647.825519     277.0  5929.809082   \n",
       "92   647.135993    300.0  640.133688  643.631130      55.0  5869.389160   \n",
       "94   654.489627    146.0  647.243790  650.944615      34.0  6231.601562   \n",
       "107  647.845877  19402.0  646.546757  647.165516    1182.0  4906.229980   \n",
       "109  650.950101   2621.0  646.368921  648.613525     397.0  5718.551270   \n",
       "114  634.948600   6518.0  639.780808  637.422017     499.0  5172.019531   \n",
       "120  630.479794    474.0  634.182173  632.480049      96.0  4945.031738   \n",
       "121  641.267530    544.0  641.698759  641.549939     145.0  5223.025391   \n",
       "125  647.832903   2208.0  647.167778  647.609700      92.0  4960.515625   \n",
       "126  653.823708   1255.0  651.019501  652.490324     165.0  4880.040527   \n",
       "129  639.821536   1962.0  639.065021  639.379896     170.0  5204.687500   \n",
       "134  662.788831    752.0  656.968347  659.849665       0.0  5500.107910   \n",
       "135  658.662077   9328.0  663.079096  661.096120    1719.0  5360.517090   \n",
       "143  633.088379   8416.0  639.744562  636.421265    1333.0  5065.911133   \n",
       "146  642.932220   4612.0  643.894790  643.420662     601.0  5124.836426   \n",
       "150  659.636291    133.0  654.006918  656.901619       0.0  5213.089355   \n",
       "152  656.280216    222.0  653.613198  655.002750      30.0  4886.983887   \n",
       "158  652.815764    246.0  649.519740  651.190263      38.0  5161.202148   \n",
       "162  655.948312    727.0  653.918119  654.969426      50.0  4842.607910   \n",
       "166  661.393048    797.0  658.033962  659.672199      60.0  4674.298828   \n",
       "168  641.824404    235.0  642.706149  642.305920      51.0  5621.687988   \n",
       "169  649.462177   8294.0  649.298686  649.430883    1295.0  4948.868164   \n",
       "171  665.447342    150.0  661.930925  663.703699      16.0  5132.788574   \n",
       "173  652.022155   2326.0  649.161762  650.526578     345.0  5149.186523   \n",
       "174  652.584348    501.0  650.315801  651.378747     163.0  5373.206543   \n",
       "175  636.501320    470.0  634.375519  635.416954      48.0  6485.166016   \n",
       "184  655.280541    780.0  650.376220  652.903479      75.0  5261.370605   \n",
       "186  652.446694    140.0  651.186929  651.756556      25.0  4566.270020   \n",
       "189  659.411046    108.0  658.471727  658.905142      25.0  4432.478027   \n",
       "204  662.314650    307.0  659.310833  660.853105      36.0  4718.163086   \n",
       "205  649.437663    347.0  646.738300  648.038495      56.0  4751.299805   \n",
       "211  651.567296   1129.0  644.711060  648.104441     164.0  5805.636719   \n",
       "214  663.465537   4928.0  660.573657  661.936855     618.0  5139.937500   \n",
       "219  669.613275    865.0  666.541670  668.121803      96.0  4797.503418   \n",
       "222  658.688447   6373.0  658.211347  658.471972     624.0  4733.746582   \n",
       "224  663.085644   2903.0  657.967585  660.539652     192.0  5431.177246   \n",
       "225  657.091211    565.0  653.529834  655.365003      60.0  5382.088867   \n",
       "234  653.031523   3518.0  652.258787  652.659949     419.0  5642.832031   \n",
       "235  652.773767  19294.0  653.725935  653.493526    2401.0  5280.023438   \n",
       "238  668.170509    117.0  662.806404  665.457846      42.0  6039.985840   \n",
       "243  653.984485   5205.0  653.846936  653.818800     474.0  4954.466309   \n",
       "247  660.471034   3182.0  659.690492  659.956172     328.0  4777.512695   \n",
       "249  659.251062  11855.0  659.739467  659.553594    1507.0  5486.029785   \n",
       "250  663.553877   1068.0  656.782451  659.998883     129.0  5486.382324   \n",
       "256  660.448138    119.0  656.326883  658.311479      23.0  5990.794434   \n",
       "258  646.835857    546.0  645.633578  646.294001      60.0  4777.769043   \n",
       "263  667.167368   2325.0  666.148685  666.576950     302.0  4890.986816   \n",
       "281  667.893071    822.0  661.581156  664.708583     105.0  5784.625000   \n",
       "284  666.968963    515.0  659.846349  663.479007      58.0  5190.310059   \n",
       "287  652.062665    823.0  648.806330  650.402278     127.0  5331.920898   \n",
       "288  669.367531   2231.0  665.806105  667.439416     258.0  5599.607910   \n",
       "290  665.115102    309.0  662.837178  664.027088      85.0  4700.432129   \n",
       "292  656.883658    966.0  650.485988  653.693831     114.0  5318.240723   \n",
       "294  663.358136    762.0  657.475610  660.403959      54.0  5143.190430   \n",
       "296  661.465628   9850.0  663.110579  662.196270    1113.0  5081.455566   \n",
       "297  648.618780    129.0  648.549217  648.603046      10.0  4016.416260   \n",
       "298  666.036055  10619.0  667.101409  666.739172    1150.0  5374.050293   \n",
       "306  663.041307    675.0  662.365592  662.682382      87.0  4543.305176   \n",
       "316  657.115402    160.0  650.725460  653.880704      38.0  6870.346191   \n",
       "318  671.475950   2272.0  667.167409  669.243356     307.0  5182.365723   \n",
       "323  667.193966   4632.0  664.397837  665.754992     637.0  5666.713867   \n",
       "328  648.312381    118.0  647.954079  648.068150      10.0  4451.256836   \n",
       "330  666.844189   4734.0  666.145109  666.526671     586.0  4922.939941   \n",
       "332  663.895376   6055.0  663.783643  663.791609     540.0  4829.717285   \n",
       "334  671.557943   2801.0  668.080271  669.752223     615.0  5205.941895   \n",
       "338  654.703879   1212.0  650.066687  652.415548     116.0  5090.587891   \n",
       "341  669.614810   6257.0  669.478912  669.555534     930.0  4889.479980   \n",
       "342  669.900481    868.0  665.261388  667.561157     122.0  5118.130371   \n",
       "349  673.992356   1452.0  669.820784  672.024306     183.0  5028.561523   \n",
       "350  674.467598    155.0  668.955346  671.693862      32.0  5695.466797   \n",
       "352  668.890252    567.0  663.800812  666.315450     103.0  5156.437012   \n",
       "371  658.155302    449.0  651.847919  654.937895      50.0  5884.432129   \n",
       "375  687.413953     81.0  677.710248  682.690134      14.0  7667.571777   \n",
       "377  681.105457   1960.0  677.414093  679.339899     223.0  5368.502930   \n",
       "379  687.972117    946.0  685.793505  686.758283     310.0  5387.458984   \n",
       "382  675.490470    945.0  671.170728  673.307374     216.0  5088.708008   \n",
       "384  676.657773    164.0  670.484640  673.555717      35.0  6588.430664   \n",
       "388  679.589037    134.0  675.039429  677.465932      27.0  5094.958984   \n",
       "393  671.865756    526.0  666.245290  669.097893      33.0  5644.286133   \n",
       "400  683.225780     81.0  673.540505  678.458755      21.0  7614.379395   \n",
       "403  681.855225   2542.0  676.931333  679.498758     404.0  6604.063965   \n",
       "405  705.686334   1059.0  708.851129  707.182838     272.0  6460.657227   \n",
       "413  697.051592    220.0  693.210634  695.000096      22.0  6500.449707   \n",
       "420  657.326803   1778.0  652.903342  655.114523     313.0  5993.392578   \n",
       "\n",
       "         elpct  \n",
       "7    68.717949  \n",
       "22   16.210526  \n",
       "38   32.394367  \n",
       "39   65.512100  \n",
       "45   36.801315  \n",
       "48   13.759213  \n",
       "52   35.250000  \n",
       "59   33.487709  \n",
       "65   17.659353  \n",
       "67   31.198156  \n",
       "72   14.429193  \n",
       "85   28.691982  \n",
       "86    1.795332  \n",
       "92    0.666667  \n",
       "94    0.000000  \n",
       "107  18.436245  \n",
       "109   2.022129  \n",
       "114  43.494938  \n",
       "120  35.443039  \n",
       "121   8.639706  \n",
       "125   9.873188  \n",
       "126  16.095617  \n",
       "129  18.195719  \n",
       "134   0.132979  \n",
       "135  50.857632  \n",
       "143  43.750000  \n",
       "146  16.652212  \n",
       "150   0.751880  \n",
       "152  13.513513  \n",
       "158  16.666668  \n",
       "162  13.067400  \n",
       "166   3.262233  \n",
       "168  39.574467  \n",
       "169  10.127803  \n",
       "171  14.000000  \n",
       "173   6.405847  \n",
       "174   2.395210  \n",
       "175   5.957447  \n",
       "184   3.846154  \n",
       "186  10.714286  \n",
       "189   4.629630  \n",
       "204   7.491857  \n",
       "205   9.798271  \n",
       "211   0.000000  \n",
       "214   5.925324  \n",
       "219   3.352601  \n",
       "222   6.260787  \n",
       "224   3.169135  \n",
       "225  16.991150  \n",
       "234  40.108017  \n",
       "235  16.663212  \n",
       "238   0.854701  \n",
       "243  22.939482  \n",
       "247   8.956632  \n",
       "249  20.725431  \n",
       "250   0.187266  \n",
       "256  28.571430  \n",
       "258  13.736264  \n",
       "263  22.709679  \n",
       "281   0.851582  \n",
       "284   0.000000  \n",
       "287   2.187120  \n",
       "288  13.626176  \n",
       "290   1.941748  \n",
       "292   0.621118  \n",
       "294   0.000000  \n",
       "296  18.558376  \n",
       "297   6.201550  \n",
       "298  18.457481  \n",
       "306  14.962962  \n",
       "316   2.500000  \n",
       "318   1.320423  \n",
       "323  17.854059  \n",
       "328   7.627119  \n",
       "330  10.857626  \n",
       "332  14.236169  \n",
       "334   2.070689  \n",
       "338   3.960396  \n",
       "341   8.726227  \n",
       "342   0.000000  \n",
       "349   0.000000  \n",
       "350   4.516129  \n",
       "352   0.000000  \n",
       "371   0.000000  \n",
       "375   0.000000  \n",
       "377   1.479592  \n",
       "379   1.479915  \n",
       "382   2.539683  \n",
       "384  11.585366  \n",
       "388   2.238806  \n",
       "393   0.000000  \n",
       "400  12.345679  \n",
       "403  10.385523  \n",
       "405   2.266289  \n",
       "413   1.363636  \n",
       "420   5.005624  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_data_inputs = data_test[inputs]\n",
    "test_data_predictions_1 = causal_structure_1.predict(test_data_inputs)\n",
    "test_data_predictions_1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that the `predict` method completes the provided data frame filling in the missing columns with estimates.\n",
    "Let us check the predictions by looking at a scatter plot if the true test data values vs. the predictions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pl.figure(figsize=(10,10))\n",
    "for i, col in [(1, \"mathscr\"), (2, \"readscr\"), (3, \"testscr\")]:\n",
    "    ax = pl.subplot(2,2,i)\n",
    "    ax.scatter(data_test[col], test_data_predictions_1[col])\n",
    "    minval, maxval = data_test[col].min(), data_test[col].max()\n",
    "    ax.plot([minval, maxval], [minval, maxval], ls=\"--\", c=\"k\")\n",
    "    ax.set_xlabel(\"real value\")\n",
    "    ax.set_ylabel(\"prediction\")\n",
    "    ax.set_title(col)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that the predictions seem to follow the real test values rather ok."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Apart from making predictions we can also look at objectives. For example we can evaluate the performance using the `Evaluator`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'teachers': None,\n",
       " 'compstu': None,\n",
       " 'mealpct': None,\n",
       " 'calwpct': None,\n",
       " 'str': None,\n",
       " 'avginc': None,\n",
       " 'readscr': 0.8020755684153762,\n",
       " 'enrltot': None,\n",
       " 'mathscr': 0.6574450121854509,\n",
       " 'testscr': 0.7577271442716229,\n",
       " 'computer': None,\n",
       " 'expnstu': None,\n",
       " 'elpct': None}"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "performance_1 = causal_structure_1.evaluate_objective(Evaluator, data=data_test,\n",
    "                                                      inputs=inputs, outputs=outputs,\n",
    "                                                      metric=\"r2\")\n",
    "performance_1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that the R2-score for the inputs columns is returned as `None`. The values for the outputs match with the visual impression of a decent fit."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To judge whether this makes sense we have to know what these columns mean."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Caschool\n",
      "\n",
      "PyDataset Documentation (adopted from R Documentation. The displayed examples are in R)\n",
      "\n",
      "## The California Test Score Data Set\n",
      "\n",
      "### Description\n",
      "\n",
      "a cross-section from 1998-1999\n",
      "\n",
      "_number of observations_ : 420\n",
      "\n",
      "_observation_ : schools\n",
      "\n",
      "_country_ : United States\n",
      "\n",
      "### Usage\n",
      "\n",
      "    data(Caschool)\n",
      "\n",
      "### Format\n",
      "\n",
      "A dataframe containing :\n",
      "\n",
      "distcod\n",
      "\n",
      "disctric code\n",
      "\n",
      "county\n",
      "\n",
      "county\n",
      "\n",
      "district\n",
      "\n",
      "district\n",
      "\n",
      "grspan\n",
      "\n",
      "grade span of district\n",
      "\n",
      "enrltot\n",
      "\n",
      "total enrollment\n",
      "\n",
      "teachers\n",
      "\n",
      "number of teachers\n",
      "\n",
      "calwpct\n",
      "\n",
      "percent qualifying for CalWorks\n",
      "\n",
      "mealpct\n",
      "\n",
      "percent qualifying for reduced-price lunch\n",
      "\n",
      "computer\n",
      "\n",
      "number of computers\n",
      "\n",
      "testscr\n",
      "\n",
      "average test score (read.scr+math.scr)/2\n",
      "\n",
      "compstu\n",
      "\n",
      "computer per student\n",
      "\n",
      "expnstu\n",
      "\n",
      "expenditure per student\n",
      "\n",
      "str\n",
      "\n",
      "student teacher ratio\n",
      "\n",
      "avginc\n",
      "\n",
      "district average income\n",
      "\n",
      "elpct\n",
      "\n",
      "percent of English learners\n",
      "\n",
      "readscr\n",
      "\n",
      "average reading score\n",
      "\n",
      "mathscr\n",
      "\n",
      "average math score\n",
      "\n",
      "### Source\n",
      "\n",
      "California Department of Education http://www.cde.ca.gov.\n",
      "\n",
      "### References\n",
      "\n",
      "Stock, James H. and Mark W. Watson (2003) _Introduction to Econometrics_,\n",
      "Addison-Wesley Educational Publishers,\n",
      "http://wps.aw.com/aw_stockwatsn_economtrcs_1, chapter 4–7.\n",
      "\n",
      "### See Also\n",
      "\n",
      "`Index.Source`, `Index.Economics`, `Index.Econometrics`, `Index.Observations`\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "pydataset.data(\"Caschool\", show_doc=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that in the fully connected model (all inputs influence all outputs) the biggest influences seem to be the total enrollment 'enrltot', the amount of teachers 'teachers' adn the percentage of students that get subsidized meals 'mealpct'. The influence of the other tests on 'testscr' is zero.\n",
    "\n",
    "We might now ask the question whether it makes sense that these quantities are the direct main influences or if they are only confounders. If we take a look at the correlations of the inputs, we see that the inputs have significant correlations that can lead to confounding and the explain-away-effect."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "#T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col0,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col6,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col1,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col2,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col3,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col4,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col5,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col0,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col6,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col7,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col8,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col9{\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col1,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col0{\n",
       "            background-color:  #bfd3f6;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col2,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col0{\n",
       "            background-color:  #ead5c9;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col3,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col0,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col6,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col3{\n",
       "            background-color:  #edd1c2;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col4,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col0{\n",
       "            background-color:  #f5c1a9;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col5,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col0{\n",
       "            background-color:  #e0dbd8;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col7,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col7,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col0,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col6{\n",
       "            background-color:  #c0282f;\n",
       "            color:  #f1f1f1;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col8,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col0{\n",
       "            background-color:  #d2dbe8;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col9,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col0{\n",
       "            background-color:  #f7b497;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col2,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col1{\n",
       "            background-color:  #c5d6f2;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col3,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col6,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col1,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col1{\n",
       "            background-color:  #bed2f6;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col4,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col1{\n",
       "            background-color:  #afcafc;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col5,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col1{\n",
       "            background-color:  #f1cdba;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col7,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col1{\n",
       "            background-color:  #d7dce3;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col8,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col1{\n",
       "            background-color:  #f5c0a7;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col9,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col1{\n",
       "            background-color:  #b5cdfa;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col3,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col2{\n",
       "            background-color:  #df634e;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col4,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col2{\n",
       "            background-color:  #dfdbd9;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col5,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col2{\n",
       "            background-color:  #8badfd;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col6,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col2{\n",
       "            background-color:  #e9d5cb;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col7,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col7,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col7,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col2,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col3,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col5{\n",
       "            background-color:  #e6d7cf;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col8,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col2{\n",
       "            background-color:  #e7d7ce;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col9,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col2{\n",
       "            background-color:  #f7b599;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col4,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col3{\n",
       "            background-color:  #eed0c0;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col5,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col3{\n",
       "            background-color:  #688aef;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col8,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col3{\n",
       "            background-color:  #d8dce2;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col9,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col3{\n",
       "            background-color:  #e67259;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col5,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col4{\n",
       "            background-color:  #bcd2f7;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col6,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col4{\n",
       "            background-color:  #f7bca1;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col7,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col4{\n",
       "            background-color:  #f4c5ad;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col8,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col4{\n",
       "            background-color:  #7da0f9;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col9,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col4{\n",
       "            background-color:  #f3c8b2;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col6,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col5{\n",
       "            background-color:  #dedcdb;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col8,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col5{\n",
       "            background-color:  #f6bda2;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col9,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col5{\n",
       "            background-color:  #a5c3fe;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col8,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col6{\n",
       "            background-color:  #cfdaea;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col9,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col6{\n",
       "            background-color:  #f7b396;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col8,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col7{\n",
       "            background-color:  #d5dbe5;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col9,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col7{\n",
       "            background-color:  #f6bea4;\n",
       "            color:  #000000;\n",
       "        }#T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col9,#T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col8{\n",
       "            background-color:  #d6dce4;\n",
       "            color:  #000000;\n",
       "        }</style><table id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0\" ><thead>    <tr>        <th class=\"blank level0\" ></th>        <th class=\"col_heading level0 col0\" >teachers</th>        <th class=\"col_heading level0 col1\" >compstu</th>        <th class=\"col_heading level0 col2\" >calwpct</th>        <th class=\"col_heading level0 col3\" >mealpct</th>        <th class=\"col_heading level0 col4\" >str</th>        <th class=\"col_heading level0 col5\" >avginc</th>        <th class=\"col_heading level0 col6\" >enrltot</th>        <th class=\"col_heading level0 col7\" >computer</th>        <th class=\"col_heading level0 col8\" >expnstu</th>        <th class=\"col_heading level0 col9\" >elpct</th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0level0_row0\" class=\"row_heading level0 row0\" >teachers</th>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col0\" class=\"data row0 col0\" >1.000000</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col1\" class=\"data row0 col1\" >-0.205573</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col2\" class=\"data row0 col2\" >0.105457</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col3\" class=\"data row0 col3\" >0.142855</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col4\" class=\"data row0 col4\" >0.272616</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col5\" class=\"data row0 col5\" >0.023591</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col6\" class=\"data row0 col6\" >0.997424</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col7\" class=\"data row0 col7\" >0.936511</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col8\" class=\"data row0 col8\" >-0.082112</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row0_col9\" class=\"data row0 col9\" >0.357818</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0level0_row1\" class=\"row_heading level0 row1\" >compstu</th>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col0\" class=\"data row1 col0\" >-0.205573</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col1\" class=\"data row1 col1\" >1.000000</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col2\" class=\"data row1 col2\" >-0.171575</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col3\" class=\"data row1 col3\" >-0.217232</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col4\" class=\"data row1 col4\" >-0.301937</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col5\" class=\"data row1 col5\" >0.183565</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col6\" class=\"data row1 col6\" >-0.213354</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col7\" class=\"data row1 col7\" >-0.041737</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col8\" class=\"data row1 col8\" >0.274414</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row1_col9\" class=\"data row1 col9\" >-0.272306</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0level0_row2\" class=\"row_heading level0 row2\" >calwpct</th>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col0\" class=\"data row2 col0\" >0.105457</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col1\" class=\"data row2 col1\" >-0.171575</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col2\" class=\"data row2 col2\" >1.000000</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col3\" class=\"data row2 col3\" >0.735591</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col4\" class=\"data row2 col4\" >0.017576</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col5\" class=\"data row2 col5\" >-0.511519</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col6\" class=\"data row2 col6\" >0.099840</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col7\" class=\"data row2 col7\" >0.072180</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col8\" class=\"data row2 col8\" >0.081346</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row2_col9\" class=\"data row2 col9\" >0.350795</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0level0_row3\" class=\"row_heading level0 row3\" >mealpct</th>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col0\" class=\"data row3 col0\" >0.142855</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col1\" class=\"data row3 col1\" >-0.217232</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col2\" class=\"data row3 col2\" >0.735591</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col3\" class=\"data row3 col3\" >1.000000</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col4\" class=\"data row3 col4\" >0.155836</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col5\" class=\"data row3 col5\" >-0.708045</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col6\" class=\"data row3 col6\" >0.147787</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col7\" class=\"data row3 col7\" >0.076777</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col8\" class=\"data row3 col8\" >-0.031435</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row3_col9\" class=\"data row3 col9\" >0.679559</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0level0_row4\" class=\"row_heading level0 row4\" >str</th>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col0\" class=\"data row4 col0\" >0.272616</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col1\" class=\"data row4 col1\" >-0.301937</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col2\" class=\"data row4 col2\" >0.017576</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col3\" class=\"data row4 col3\" >0.155836</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col4\" class=\"data row4 col4\" >1.000000</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col5\" class=\"data row4 col5\" >-0.222932</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col6\" class=\"data row4 col6\" >0.305878</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col7\" class=\"data row4 col7\" >0.243671</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col8\" class=\"data row4 col8\" >-0.590435</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row4_col9\" class=\"data row4 col9\" >0.226483</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0level0_row5\" class=\"row_heading level0 row5\" >avginc</th>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col0\" class=\"data row5 col0\" >0.023591</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col1\" class=\"data row5 col1\" >0.183565</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col2\" class=\"data row5 col2\" >-0.511519</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col3\" class=\"data row5 col3\" >-0.708045</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col4\" class=\"data row5 col4\" >-0.222932</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col5\" class=\"data row5 col5\" >1.000000</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col6\" class=\"data row5 col6\" >0.010920</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col7\" class=\"data row5 col7\" >0.076274</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col8\" class=\"data row5 col8\" >0.298755</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row5_col9\" class=\"data row5 col9\" >-0.362238</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0level0_row6\" class=\"row_heading level0 row6\" >enrltot</th>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col0\" class=\"data row6 col0\" >0.997424</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col1\" class=\"data row6 col1\" >-0.213354</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col2\" class=\"data row6 col2\" >0.099840</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col3\" class=\"data row6 col3\" >0.147787</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col4\" class=\"data row6 col4\" >0.305878</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col5\" class=\"data row6 col5\" >0.010920</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col6\" class=\"data row6 col6\" >1.000000</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col7\" class=\"data row6 col7\" >0.929746</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col8\" class=\"data row6 col8\" >-0.100289</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row6_col9\" class=\"data row6 col9\" >0.365512</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0level0_row7\" class=\"row_heading level0 row7\" >computer</th>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col0\" class=\"data row7 col0\" >0.936511</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col1\" class=\"data row7 col1\" >-0.041737</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col2\" class=\"data row7 col2\" >0.072180</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col3\" class=\"data row7 col3\" >0.076777</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col4\" class=\"data row7 col4\" >0.243671</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col5\" class=\"data row7 col5\" >0.076274</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col6\" class=\"data row7 col6\" >0.929746</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col7\" class=\"data row7 col7\" >1.000000</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col8\" class=\"data row7 col8\" >-0.061414</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row7_col9\" class=\"data row7 col9\" >0.292201</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0level0_row8\" class=\"row_heading level0 row8\" >expnstu</th>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col0\" class=\"data row8 col0\" >-0.082112</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col1\" class=\"data row8 col1\" >0.274414</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col2\" class=\"data row8 col2\" >0.081346</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col3\" class=\"data row8 col3\" >-0.031435</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col4\" class=\"data row8 col4\" >-0.590435</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col5\" class=\"data row8 col5\" >0.298755</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col6\" class=\"data row8 col6\" >-0.100289</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col7\" class=\"data row8 col7\" >-0.061414</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col8\" class=\"data row8 col8\" >1.000000</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row8_col9\" class=\"data row8 col9\" >-0.049224</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0level0_row9\" class=\"row_heading level0 row9\" >elpct</th>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col0\" class=\"data row9 col0\" >0.357818</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col1\" class=\"data row9 col1\" >-0.272306</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col2\" class=\"data row9 col2\" >0.350795</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col3\" class=\"data row9 col3\" >0.679559</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col4\" class=\"data row9 col4\" >0.226483</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col5\" class=\"data row9 col5\" >-0.362238</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col6\" class=\"data row9 col6\" >0.365512</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col7\" class=\"data row9 col7\" >0.292201</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col8\" class=\"data row9 col8\" >-0.049224</td>\n",
       "                        <td id=\"T_eb604f28_49bc_11eb_958a_00155d2a16c0row9_col9\" class=\"data row9 col9\" >1.000000</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2142b1f9b88>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_train[inputs].corr().style.background_gradient(cmap='coolwarm', vmin=-1, vmax=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see for example that the total enrollment 'enrltot' and the amount of teachers 'teachers' are very strongly correlated (also with the amount of computers)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "#T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col0{\n",
       "            background-color:  #c6d6f1;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col1{\n",
       "            background-color:  #f5c0a7;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col2{\n",
       "            background-color:  #7699f6;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col3{\n",
       "            background-color:  #4c66d6;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col4{\n",
       "            background-color:  #bbd1f8;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col5{\n",
       "            background-color:  #e0654f;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col6{\n",
       "            background-color:  #c5d6f2;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col7{\n",
       "            background-color:  #d2dbe8;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col8{\n",
       "            background-color:  #eed0c0;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col9{\n",
       "            background-color:  #7093f3;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col0,#T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col4{\n",
       "            background-color:  #c1d4f4;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col1{\n",
       "            background-color:  #f6bfa6;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col2,#T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col2{\n",
       "            background-color:  #799cf8;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col3{\n",
       "            background-color:  #4b64d5;\n",
       "            color:  #f1f1f1;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col4{\n",
       "            background-color:  #b7cff9;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col5,#T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col5{\n",
       "            background-color:  #e16751;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col6{\n",
       "            background-color:  #c0d4f5;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col7{\n",
       "            background-color:  #ccd9ed;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col8{\n",
       "            background-color:  #f0cdbb;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col9{\n",
       "            background-color:  #688aef;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col0{\n",
       "            background-color:  #cdd9ec;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col1{\n",
       "            background-color:  #f5c2aa;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col3{\n",
       "            background-color:  #5470de;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col6{\n",
       "            background-color:  #cbd8ee;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col7{\n",
       "            background-color:  #d7dce3;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col8{\n",
       "            background-color:  #ead4c8;\n",
       "            color:  #000000;\n",
       "        }#T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col9{\n",
       "            background-color:  #7da0f9;\n",
       "            color:  #000000;\n",
       "        }</style><table id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0\" ><thead>    <tr>        <th class=\"blank level0\" ></th>        <th class=\"col_heading level0 col0\" >teachers</th>        <th class=\"col_heading level0 col1\" >compstu</th>        <th class=\"col_heading level0 col2\" >calwpct</th>        <th class=\"col_heading level0 col3\" >mealpct</th>        <th class=\"col_heading level0 col4\" >str</th>        <th class=\"col_heading level0 col5\" >avginc</th>        <th class=\"col_heading level0 col6\" >enrltot</th>        <th class=\"col_heading level0 col7\" >computer</th>        <th class=\"col_heading level0 col8\" >expnstu</th>        <th class=\"col_heading level0 col9\" >elpct</th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0level0_row0\" class=\"row_heading level0 row0\" >testscr</th>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col0\" class=\"data row0 col0\" >-0.157794</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col1\" class=\"data row0 col1\" >0.278550</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col2\" class=\"data row0 col2\" >-0.628098</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col3\" class=\"data row0 col3\" >-0.877528</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col4\" class=\"data row0 col4\" >-0.227190</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col5\" class=\"data row0 col5\" >0.734062</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col6\" class=\"data row0 col6\" >-0.167217</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col7\" class=\"data row0 col7\" >-0.084573</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col8\" class=\"data row0 col8\" >0.150482</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row0_col9\" class=\"data row0 col9\" >-0.662471</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0level0_row1\" class=\"row_heading level0 row1\" >readscr</th>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col0\" class=\"data row1 col0\" >-0.189992</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col1\" class=\"data row1 col1\" >0.285396</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col2\" class=\"data row1 col2\" >-0.615917</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col3\" class=\"data row1 col3\" >-0.887392</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col4\" class=\"data row1 col4\" >-0.252007</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col5\" class=\"data row1 col5\" >0.720649</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col6\" class=\"data row1 col6\" >-0.199871</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col7\" class=\"data row1 col7\" >-0.117787</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col8\" class=\"data row1 col8\" >0.178198</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row1_col9\" class=\"data row1 col9\" >-0.704534</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0level0_row2\" class=\"row_heading level0 row2\" >mathscr</th>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col0\" class=\"data row2 col0\" >-0.116787</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col1\" class=\"data row2 col1\" >0.260359</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col2\" class=\"data row2 col2\" >-0.617021</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col3\" class=\"data row2 col3\" >-0.832906</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col4\" class=\"data row2 col4\" >-0.191503</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col5\" class=\"data row2 col5\" >0.720223</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col6\" class=\"data row2 col6\" >-0.125351</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col7\" class=\"data row2 col7\" >-0.045295</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col8\" class=\"data row2 col8\" >0.114616</td>\n",
       "                        <td id=\"T_eb651406_49bc_11eb_ac91_00155d2a16c0row2_col9\" class=\"data row2 col9\" >-0.591257</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2142b3fa508>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_train.corr().loc[inputs, outputs].T.style.background_gradient(cmap='coolwarm', vmin=-1, vmax=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Advanced structure"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "With the information we have about the data set we can now try to make a more realistic model of the involved columns.\n",
    "\n",
    "We start with a number of assumptions which result in dependencies."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "dependencies = []"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First of all we see in the data documentation that 'testscr' is just the average test score."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "dependencies += [\n",
    "    [[\"mathscr\", \"readscr\"], \"testscr\"]\n",
    "]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Assumption 1: We believe the student-teacher-ratio is what actually matters for the test performance. The student teacher ratio of course depends on the number of students and the number of teachers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "dependencies += [\n",
    "    [\"str\", [\"mathscr\", \"readscr\"]],\n",
    "    [[\"teachers\", \"enrltot\"], \"str\"]\n",
    "]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Assumption 2: We believe the amount of teachers depends on the number of students and the expenses per student."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "dependencies += [\n",
    "    [[\"expnstu\", \"enrltot\"], \"teachers\"]\n",
    "]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Assumption 3: The amount of computers per student might indicate the learning possibilities at the school and therefore influence the test results. The amount of computer per student of course is determined by the number of students and the number of computers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "dependencies += [\n",
    "    [\"compstu\", [\"mathscr\", \"readscr\"]],\n",
    "    [[\"computer\", \"enrltot\"], \"compstu\"]\n",
    "]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Assumption 4: The amount of computers probably depends on the amount of students and the funding per student."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "dependencies += [\n",
    "    [[\"expnstu\", \"enrltot\"], \"computer\"]\n",
    "]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Assumption 5: We assume poorer students achieve worse test results on average. In the data we have the percent qualifying for reduced-price lunch 'mealpct', the percent qualifying for CalWorks 'calwpct' and the average income in the school district 'avginc'. Since the latter is a bit indirect we assume 'mealpct' and 'calwpct' are direct influences on the test results while both of them are influenced by 'avginc'."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "dependencies += [\n",
    "    [[\"mealpct\", \"calwpct\"], [\"mathscr\", \"readscr\"]],\n",
    "    [\"avginc\", [\"mealpct\", \"calwpct\"]]\n",
    "]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Assumption 6: We assume that being a non-native speaker influences the test results directly as well as the economic situation of the student. The percent of English learners is 'elpct'."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "dependencies += [\n",
    "    [\"elpct\", [\"mathscr\", \"readscr\"]],\n",
    "    [\"elpct\", [\"mealpct\", \"calwpct\"]]\n",
    "]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we put in these assumptions into a causal structure."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "causal_structure_2 = CausalStructure(dependencies)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "and train it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "causal_structure_2.train(data_train, method=\"MGVI\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let us have a look at the performance on the test data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_data_inputs = data_test[inputs]\n",
    "test_data_predictions_2 = causal_structure_2.predict(test_data_inputs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First visually..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pylab as pl\n",
    "pl.figure(figsize=(10,10))\n",
    "for i, col in [(1, \"mathscr\"), (2, \"readscr\"), (3, \"testscr\")]:\n",
    "    ax = pl.subplot(2,2,i)\n",
    "    ax.scatter(data_test[col], test_data_predictions_2[col])\n",
    "    minval, maxval = data_test[col].min(), data_test[col].max()\n",
    "    ax.plot([minval, maxval], [minval, maxval], ls=\"--\", c=\"k\")\n",
    "    ax.set_xlabel(\"real value\")\n",
    "    ax.set_ylabel(\"prediction\")\n",
    "    ax.set_title(col)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "and by the R2 score."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "performance_2 = causal_structure_2.evaluate_objective(Evaluator, data=data_test,\n",
    "                                                      inputs=inputs, outputs=outputs,\n",
    "                                                      metric=\"r2\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'performance of simple model:'"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{'readscr': 0.8020755684153762,\n",
       " 'mathscr': 0.6574450121854509,\n",
       " 'testscr': 0.7577271442716229}"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'performance of structured model:'"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{'readscr': 0.7736833098673347,\n",
       " 'testscr': 0.733377201980135,\n",
       " 'mathscr': 0.6371819769563171}"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(\"performance of simple model:\")\n",
    "display({\n",
    "    key: value for key, value in performance_1.items() if key in outputs\n",
    "})\n",
    "display(\"performance of structured model:\")\n",
    "display({\n",
    "    key: value for key, value in performance_2.items() if key in outputs\n",
    "})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that the R2 scores are actually slightly worse than for the simple causal structure, which only had inputs and outputs. Maybe some of our assumptions were incomplete or even wrong.\n",
    "\n",
    "So what have we gained if the prediction performance is not better? We can see that once we start comparing predicions on partial data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Comparing the performance of the simple and the advanced structure."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we established the performance when predicting the outputs based on all inputs is the same in both models. When all input data are available and the training set and the test set are split randomly it does not matter that much whether the learned behavior is based on real causality or confounders.\n",
    "\n",
    "However, if we make predictions from incomplete data, this changes."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Example 1: predicting based on subsidized meals and CalWorks\n",
    "Let's try a prediction based on only the amount of subsidized meals and the percentage in the CalWorks program."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'prediction based on simple model:'"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{'readscr': 0.6342623417400537,\n",
       " 'mathscr': 0.5509301033970446,\n",
       " 'testscr': 0.590015453901932}"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'prediction based on structured model:'"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{'readscr': 0.7078249997685941,\n",
       " 'testscr': 0.6881614958989846,\n",
       " 'mathscr': 0.6110603334844447}"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(\"prediction based on simple model:\")\n",
    "display({\n",
    "    key: value for key, value in \n",
    "    causal_structure_1.evaluate_objective(Evaluator, data=data_test,\n",
    "                                          inputs=['calwpct', 'mealpct'], outputs=outputs,\n",
    "                                          metric=\"r2\").items()\n",
    "    if key in outputs\n",
    "})\n",
    "display(\"prediction based on structured model:\")\n",
    "display({\n",
    "    key: value for key, value in \n",
    "    causal_structure_2.evaluate_objective(Evaluator, data=data_test,\n",
    "                                          inputs=['calwpct', 'mealpct'], outputs=outputs,\n",
    "                                          metric=\"r2\").items()\n",
    "    if key in outputs\n",
    "})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that in this scenario the structured model outperforms the simple model, but both models still achieve decent scores."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Example 2: predicting based on CalWorks only\n",
    "Let's try a prediction based on only the percentage in the CalWorks program."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'prediction based on simple model:'"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{'readscr': -0.033142940725477965,\n",
       " 'mathscr': 0.11518897669104289,\n",
       " 'testscr': 0.029182778885538996}"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'prediction based on structured model:'"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{'readscr': 0.32351923326782916,\n",
       " 'testscr': 0.3688571477752969,\n",
       " 'mathscr': 0.37847951309032846}"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(\"prediction based on simple model:\")\n",
    "display({\n",
    "    key: value for key, value in \n",
    "    causal_structure_1.evaluate_objective(Evaluator, data=data_test,\n",
    "                                          inputs=['calwpct'], outputs=outputs,\n",
    "                                          metric=\"r2\").items()\n",
    "    if key in outputs\n",
    "})\n",
    "display(\"prediction based on structured model:\")\n",
    "display({\n",
    "    key: value for key, value in \n",
    "    causal_structure_2.evaluate_objective(Evaluator, data=data_test,\n",
    "                                          inputs=['calwpct'], outputs=outputs,\n",
    "                                          metric=\"r2\").items()\n",
    "    if key in outputs\n",
    "})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this scenario the structured model outperforms the simple model significantly."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Example 3: predicting based on district average income\n",
    "Let's try a prediction based on only the district average income."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'prediction based on simple model:'"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{'readscr': 0.2816587802736583,\n",
       " 'mathscr': 0.2002535289464773,\n",
       " 'testscr': 0.2320481281685418}"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'prediction based on structured model:'"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{'readscr': 0.34039792627046295,\n",
       " 'testscr': 0.3543316153774535,\n",
       " 'mathscr': 0.3358698650229551}"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(\"prediction based on simple model:\")\n",
    "display({\n",
    "    key: value for key, value in \n",
    "    causal_structure_1.evaluate_objective(Evaluator, data=data_test,\n",
    "                                          inputs=['avginc'], outputs=outputs,\n",
    "                                          metric=\"r2\").items()\n",
    "    if key in outputs\n",
    "})\n",
    "display(\"prediction based on structured model:\")\n",
    "display({\n",
    "    key: value for key, value in \n",
    "    causal_structure_2.evaluate_objective(Evaluator, data=data_test,\n",
    "                                          inputs=['avginc'], outputs=outputs,\n",
    "                                          metric=\"r2\").items()\n",
    "    if key in outputs\n",
    "})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this scenario the structured model outperforms the simple model significantly."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To summarize, a causal structure based on realistic dependency assumptions will not necessarily outperform a black-box model with only inputs and outputs. It is however, more robust when making predicions based on incomplete data."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}