Student-Performance-Evaluation-Classification-Regression.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Student-Performance-Evaluation using Classification-Regression\n",
    "___\n",
    "Here we would try to predict student performance in secondary education (high school). \n",
    "\n",
    "We would perform data analysis for 3 cases :\n",
    "\n",
    "[Case 1: Binary-Classification :-G3>10:-1-else-0](#CASE-1:-Binary-Classification-:-G3>10:-1-else-0)\n",
    "\n",
    "[Case 2: Multi-Class-Classification](#CASE-2:-Multi-Class-Classification-:)\n",
    "\n",
    "[Case 3: Regression](#Case-3-:-Regression)\n",
    "\n",
    "#### Data Set Information:\n",
    "\n",
    "This data approaches student achievement in secondary education of two Portuguese schools. The data attributes include student grades, demographic, social and school related features) and it was collected by using school reports and questionnaires. Two datasets are provided regarding the performance in two distinct subjects: Mathematics (mat) and Portuguese language (por). In [Cortez and Silva, 2008], the two datasets were modeled under binary/five-level classification and regression tasks. Important note: the target attribute G3 has a strong correlation with attributes G2 and G1. This occurs because G3 is the final year grade (issued at the 3rd period), while G1 and G2 correspond to the 1st and 2nd period grades. It is more difficult to predict G3 without G2 and G1, but such prediction is much more useful.\n",
    "\n",
    "Dataset : http://archive.ics.uci.edu/ml/datasets/Student+Performance\n",
    "\n",
    "### Attributes for both student-mat.csv (Math course) and student-por.csv (Portuguese language course) datasets: \n",
    "\n",
    "1. school - student's school (binary: 'GP' - Gabriel Pereira or 'MS' - Mousinho da Silveira) \n",
    "2. sex - student's sex (binary: 'F' - female or 'M' - male) \n",
    "3. age - student's age (numeric: from 15 to 22) \n",
    "4. address - student's home address type (binary: 'U' - urban or 'R' - rural) \n",
    "5. famsize - family size (binary: 'LE3' - less or equal to 3 or 'GT3' - greater than 3) \n",
    "6. Pstatus - parent's cohabitation status (binary: 'T' - living together or 'A' - apart) \n",
    "7. Medu - mother's education (numeric: 0 - none, 1 - primary education (4th grade), 2 - 5th to 9th grade, 3 - secondary education or 4 - higher education) \n",
    "8. Fedu - father's education (numeric: 0 - none, 1 - primary education (4th grade), 2 - 5th to 9th grade, 3 - secondary education or 4 - higher education) \n",
    "9. Mjob - mother's job (nominal: 'teacher', 'health' care related, civil 'services' (e.g. administrative or police), 'at_home' or 'other') \n",
    "10. Fjob - father's job (nominal: 'teacher', 'health' care related, civil 'services' (e.g. administrative or police), 'at_home' or 'other') \n",
    "11. reason - reason to choose this school (nominal: close to 'home', school 'reputation', 'course' preference or 'other') \n",
    "12. guardian - student's guardian (nominal: 'mother', 'father' or 'other') \n",
    "13. traveltime - home to school travel time (numeric: 1 - <15 min., 2 - 15 to 30 min., 3 - 30 min. to 1 hour, or 4 - \\>1 hour) \n",
    "14. studytime - weekly study time (numeric: 1 - <2 hours, 2 - 2 to 5 hours, 3 - 5 to 10 hours, or 4 - >10 hours) \n",
    "15. failures - number of past class failures (numeric: n if 1<=n\\<3, else 4) \n",
    "16. schoolsup - extra educational support (binary: yes or no) \n",
    "17. famsup - family educational support (binary: yes or no) \n",
    "18. paid - extra paid classes within the course subject (Math or Portuguese) (binary: yes or no) \n",
    "19. activities - extra-curricular activities (binary: yes or no) \n",
    "20. nursery - attended nursery school (binary: yes or no) \n",
    "21. higher - wants to take higher education (binary: yes or no) \n",
    "22. internet - Internet access at home (binary: yes or no) \n",
    "23. romantic - with a romantic relationship (binary: yes or no) \n",
    "24. famrel - quality of family relationships (numeric: from 1 - very bad to 5 - excellent) \n",
    "25. freetime - free time after school (numeric: from 1 - very low to 5 - very high) \n",
    "26. goout - going out with friends (numeric: from 1 - very low to 5 - very high) \n",
    "27. Dalc - workday alcohol consumption (numeric: from 1 - very low to 5 - very high) \n",
    "28. Walc - weekend alcohol consumption (numeric: from 1 - very low to 5 - very high) \n",
    "29. health - current health status (numeric: from 1 - very bad to 5 - very good) \n",
    "30. absences - number of school absences (numeric: from 0 to 93) \n",
    "\n",
    "31. G1 - first period grade (numeric: from 0 to 20) \n",
    "31. G2 - second period grade (numeric: from 0 to 20) \n",
    "32. G3 - final grade (numeric: from 0 to 20, output target)\n",
    "\n",
    "these grades are related with the course subject, Math or Portuguese: \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import os\n",
    "from sklearn.tree import DecisionTreeClassifier, export_graphviz\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "from sklearn.cross_validation import train_test_split\n",
    "from sklearn import cross_validation, metrics\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.naive_bayes import BernoulliNB\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.svm import SVC\n",
    "from time import time\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.metrics import roc_auc_score , classification_report\n",
    "from sklearn.grid_search import GridSearchCV\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.metrics import precision_score, recall_score, accuracy_score, classification_report\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# read .csv from provided dataset\n",
    "csv_filename=\"student/student-mat.csv\"\n",
    "\n",
    "# df=pd.read_csv(csv_filename,index_col=0)\n",
    "df=pd.read_csv(csv_filename, sep=\";\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>school</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>address</th>\n",
       "      <th>famsize</th>\n",
       "      <th>Pstatus</th>\n",
       "      <th>Medu</th>\n",
       "      <th>Fedu</th>\n",
       "      <th>Mjob</th>\n",
       "      <th>Fjob</th>\n",
       "      <th>...</th>\n",
       "      <th>famrel</th>\n",
       "      <th>freetime</th>\n",
       "      <th>goout</th>\n",
       "      <th>Dalc</th>\n",
       "      <th>Walc</th>\n",
       "      <th>health</th>\n",
       "      <th>absences</th>\n",
       "      <th>G1</th>\n",
       "      <th>G2</th>\n",
       "      <th>G3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>18</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>A</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>at_home</td>\n",
       "      <td>teacher</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>17</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>T</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>at_home</td>\n",
       "      <td>other</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>15</td>\n",
       "      <td>U</td>\n",
       "      <td>LE3</td>\n",
       "      <td>T</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>at_home</td>\n",
       "      <td>other</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>15</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>T</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>health</td>\n",
       "      <td>services</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>15</td>\n",
       "      <td>14</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>16</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>T</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>other</td>\n",
       "      <td>other</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>6</td>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 33 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "  school sex  age address famsize Pstatus  Medu  Fedu     Mjob      Fjob ...  \\\n",
       "0     GP   F   18       U     GT3       A     4     4  at_home   teacher ...   \n",
       "1     GP   F   17       U     GT3       T     1     1  at_home     other ...   \n",
       "2     GP   F   15       U     LE3       T     1     1  at_home     other ...   \n",
       "3     GP   F   15       U     GT3       T     4     2   health  services ...   \n",
       "4     GP   F   16       U     GT3       T     3     3    other     other ...   \n",
       "\n",
       "  famrel freetime  goout  Dalc  Walc health absences  G1  G2  G3  \n",
       "0      4        3      4     1     1      3        6   5   6   6  \n",
       "1      5        3      3     1     1      3        4   5   5   6  \n",
       "2      4        3      2     2     3      3       10   7   8  10  \n",
       "3      3        2      2     1     1      5        2  15  14  15  \n",
       "4      4        3      2     1     2      5        4   6  10  10  \n",
       "\n",
       "[5 rows x 33 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>age</th>\n",
       "      <th>Medu</th>\n",
       "      <th>Fedu</th>\n",
       "      <th>traveltime</th>\n",
       "      <th>studytime</th>\n",
       "      <th>failures</th>\n",
       "      <th>famrel</th>\n",
       "      <th>freetime</th>\n",
       "      <th>goout</th>\n",
       "      <th>Dalc</th>\n",
       "      <th>Walc</th>\n",
       "      <th>health</th>\n",
       "      <th>absences</th>\n",
       "      <th>G1</th>\n",
       "      <th>G2</th>\n",
       "      <th>G3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>16.696203</td>\n",
       "      <td>2.749367</td>\n",
       "      <td>2.521519</td>\n",
       "      <td>1.448101</td>\n",
       "      <td>2.035443</td>\n",
       "      <td>0.334177</td>\n",
       "      <td>3.944304</td>\n",
       "      <td>3.235443</td>\n",
       "      <td>3.108861</td>\n",
       "      <td>1.481013</td>\n",
       "      <td>2.291139</td>\n",
       "      <td>3.554430</td>\n",
       "      <td>5.708861</td>\n",
       "      <td>10.908861</td>\n",
       "      <td>10.713924</td>\n",
       "      <td>10.415190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.276043</td>\n",
       "      <td>1.094735</td>\n",
       "      <td>1.088201</td>\n",
       "      <td>0.697505</td>\n",
       "      <td>0.839240</td>\n",
       "      <td>0.743651</td>\n",
       "      <td>0.896659</td>\n",
       "      <td>0.998862</td>\n",
       "      <td>1.113278</td>\n",
       "      <td>0.890741</td>\n",
       "      <td>1.287897</td>\n",
       "      <td>1.390303</td>\n",
       "      <td>8.003096</td>\n",
       "      <td>3.319195</td>\n",
       "      <td>3.761505</td>\n",
       "      <td>4.581443</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>15.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>16.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>8.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>11.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>18.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>14.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>22.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>75.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>20.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              age        Medu        Fedu  traveltime   studytime    failures  \\\n",
       "count  395.000000  395.000000  395.000000  395.000000  395.000000  395.000000   \n",
       "mean    16.696203    2.749367    2.521519    1.448101    2.035443    0.334177   \n",
       "std      1.276043    1.094735    1.088201    0.697505    0.839240    0.743651   \n",
       "min     15.000000    0.000000    0.000000    1.000000    1.000000    0.000000   \n",
       "25%     16.000000    2.000000    2.000000    1.000000    1.000000    0.000000   \n",
       "50%     17.000000    3.000000    2.000000    1.000000    2.000000    0.000000   \n",
       "75%     18.000000    4.000000    3.000000    2.000000    2.000000    0.000000   \n",
       "max     22.000000    4.000000    4.000000    4.000000    4.000000    3.000000   \n",
       "\n",
       "           famrel    freetime       goout        Dalc        Walc      health  \\\n",
       "count  395.000000  395.000000  395.000000  395.000000  395.000000  395.000000   \n",
       "mean     3.944304    3.235443    3.108861    1.481013    2.291139    3.554430   \n",
       "std      0.896659    0.998862    1.113278    0.890741    1.287897    1.390303   \n",
       "min      1.000000    1.000000    1.000000    1.000000    1.000000    1.000000   \n",
       "25%      4.000000    3.000000    2.000000    1.000000    1.000000    3.000000   \n",
       "50%      4.000000    3.000000    3.000000    1.000000    2.000000    4.000000   \n",
       "75%      5.000000    4.000000    4.000000    2.000000    3.000000    5.000000   \n",
       "max      5.000000    5.000000    5.000000    5.000000    5.000000    5.000000   \n",
       "\n",
       "         absences          G1          G2          G3  \n",
       "count  395.000000  395.000000  395.000000  395.000000  \n",
       "mean     5.708861   10.908861   10.713924   10.415190  \n",
       "std      8.003096    3.319195    3.761505    4.581443  \n",
       "min      0.000000    3.000000    0.000000    0.000000  \n",
       "25%      0.000000    8.000000    9.000000    8.000000  \n",
       "50%      4.000000   11.000000   11.000000   11.000000  \n",
       "75%      8.000000   13.000000   13.000000   14.000000  \n",
       "max     75.000000   19.000000   19.000000   20.000000  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# CASE 1: Binary Classification : G3>10: 1 else 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    395.000000\n",
       "mean      10.415190\n",
       "std        4.581443\n",
       "min        0.000000\n",
       "25%        8.000000\n",
       "50%       11.000000\n",
       "75%       14.000000\n",
       "max       20.000000\n",
       "Name: G3, dtype: float64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.G3.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# handle G3 attrubte to binary\n",
    "high = df.G3 >= 10\n",
    "low = df.G3 < 10\n",
    "df.loc[high,'G3'] = 1\n",
    "df.loc[low,'G3'] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>school</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>address</th>\n",
       "      <th>famsize</th>\n",
       "      <th>Pstatus</th>\n",
       "      <th>Medu</th>\n",
       "      <th>Fedu</th>\n",
       "      <th>Mjob</th>\n",
       "      <th>Fjob</th>\n",
       "      <th>...</th>\n",
       "      <th>famrel</th>\n",
       "      <th>freetime</th>\n",
       "      <th>goout</th>\n",
       "      <th>Dalc</th>\n",
       "      <th>Walc</th>\n",
       "      <th>health</th>\n",
       "      <th>absences</th>\n",
       "      <th>G1</th>\n",
       "      <th>G2</th>\n",
       "      <th>G3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>18</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>A</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>at_home</td>\n",
       "      <td>teacher</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>17</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>T</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>at_home</td>\n",
       "      <td>other</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>15</td>\n",
       "      <td>U</td>\n",
       "      <td>LE3</td>\n",
       "      <td>T</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>at_home</td>\n",
       "      <td>other</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>15</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>T</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>health</td>\n",
       "      <td>services</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>15</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>16</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>T</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>other</td>\n",
       "      <td>other</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>6</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 33 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "  school sex  age address famsize Pstatus  Medu  Fedu     Mjob      Fjob ...  \\\n",
       "0     GP   F   18       U     GT3       A     4     4  at_home   teacher ...   \n",
       "1     GP   F   17       U     GT3       T     1     1  at_home     other ...   \n",
       "2     GP   F   15       U     LE3       T     1     1  at_home     other ...   \n",
       "3     GP   F   15       U     GT3       T     4     2   health  services ...   \n",
       "4     GP   F   16       U     GT3       T     3     3    other     other ...   \n",
       "\n",
       "  famrel freetime  goout  Dalc  Walc health absences  G1  G2 G3  \n",
       "0      4        3      4     1     1      3        6   5   6  0  \n",
       "1      5        3      3     1     1      3        4   5   5  0  \n",
       "2      4        3      2     2     3      3       10   7   8  1  \n",
       "3      3        2      2     1     1      5        2  15  14  1  \n",
       "4      4        3      2     1     2      5        4   6  10  1  \n",
       "\n",
       "[5 rows x 33 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    395.000000\n",
       "mean       0.670886\n",
       "std        0.470487\n",
       "min        0.000000\n",
       "25%        0.000000\n",
       "50%        1.000000\n",
       "75%        1.000000\n",
       "max        1.000000\n",
       "Name: G3, dtype: float64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.G3.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "cols = list(df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['school',\n",
       " 'sex',\n",
       " 'address',\n",
       " 'famsize',\n",
       " 'Pstatus',\n",
       " 'Mjob',\n",
       " 'Fjob',\n",
       " 'reason',\n",
       " 'guardian',\n",
       " 'schoolsup',\n",
       " 'famsup',\n",
       " 'paid',\n",
       " 'activities',\n",
       " 'nursery',\n",
       " 'higher',\n",
       " 'internet',\n",
       " 'romantic']"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "categorical_features = []\n",
    "for f in cols:\n",
    "    if df[f].dtype != 'int64':\n",
    "        categorical_features.append(f)\n",
    "categorical_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "for f in categorical_features:\n",
    "\n",
    "    #Get binarized columns\n",
    "    df[f] = pd.get_dummies(df[f])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>school</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>address</th>\n",
       "      <th>famsize</th>\n",
       "      <th>Pstatus</th>\n",
       "      <th>Medu</th>\n",
       "      <th>Fedu</th>\n",
       "      <th>Mjob</th>\n",
       "      <th>Fjob</th>\n",
       "      <th>...</th>\n",
       "      <th>famrel</th>\n",
       "      <th>freetime</th>\n",
       "      <th>goout</th>\n",
       "      <th>Dalc</th>\n",
       "      <th>Walc</th>\n",
       "      <th>health</th>\n",
       "      <th>absences</th>\n",
       "      <th>G1</th>\n",
       "      <th>G2</th>\n",
       "      <th>G3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>18</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>17</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>15</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>15</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>15</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>16</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>6</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 33 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   school  sex  age  address  famsize  Pstatus  Medu  Fedu  Mjob  Fjob ...  \\\n",
       "0       1    1   18        0        1        1     4     4     1     0 ...   \n",
       "1       1    1   17        0        1        0     1     1     1     0 ...   \n",
       "2       1    1   15        0        0        0     1     1     1     0 ...   \n",
       "3       1    1   15        0        1        0     4     2     0     0 ...   \n",
       "4       1    1   16        0        1        0     3     3     0     0 ...   \n",
       "\n",
       "   famrel  freetime  goout  Dalc  Walc  health  absences  G1  G2  G3  \n",
       "0       4         3      4     1     1       3         6   5   6   0  \n",
       "1       5         3      3     1     1       3         4   5   5   0  \n",
       "2       4         3      2     2     3       3        10   7   8   1  \n",
       "3       3         2      2     1     1       5         2  15  14   1  \n",
       "4       4         3      2     1     2       5         4   6  10   1  \n",
       "\n",
       "[5 rows x 33 columns]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "features=list(df.columns[:-1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "X = df[features]\n",
    "y = df['G3']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# split dataset to 60% training and 40% testing\n",
    "X_train, X_test, y_train, y_test = cross_validation.train_test_split(X,y, test_size=0.4, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(237, 32) (237,)\n"
     ]
    }
   ],
   "source": [
    "print (X_train.shape, y_train.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature importances with forests of trees\n",
    "\n",
    "This examples shows the use of forests of trees to evaluate the importance of features on an artificial classification task. The red bars are the feature importances of the forest, along with their inter-trees variability."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Feature ranking:\n",
      "1. feature 31 - G2 (0.240635) \n",
      "2. feature 30 - G1 (0.180379) \n",
      "3. feature 14 - failures (0.058718) \n",
      "4. feature 29 - absences (0.032248) \n",
      "5. feature 25 - goout (0.031282) \n",
      "6. feature 2 - age (0.031181) \n",
      "7. feature 6 - Medu (0.023217) \n",
      "8. feature 7 - Fedu (0.023036) \n",
      "9. feature 28 - health (0.022799) \n",
      "10. feature 13 - studytime (0.022735) \n",
      "11. feature 23 - famrel (0.021943) \n",
      "12. feature 24 - freetime (0.021443) \n",
      "13. feature 27 - Walc (0.020626) \n",
      "14. feature 15 - schoolsup (0.018678) \n",
      "15. feature 22 - romantic (0.017761) \n",
      "16. feature 10 - reason (0.016906) \n",
      "17. feature 18 - activities (0.016887) \n",
      "18. feature 26 - Dalc (0.016689) \n",
      "19. feature 3 - address (0.016217) \n",
      "20. feature 1 - sex (0.016042) \n",
      "21. feature 12 - traveltime (0.015887) \n",
      "22. feature 4 - famsize (0.015885) \n",
      "23. feature 17 - paid (0.015443) \n",
      "24. feature 16 - famsup (0.014640) \n",
      "25. feature 11 - guardian (0.014492) \n",
      "26. feature 21 - internet (0.013992) \n",
      "27. feature 8 - Mjob (0.013320) \n",
      "28. feature 19 - nursery (0.011929) \n",
      "29. feature 20 - higher (0.010266) \n",
      "30. feature 5 - Pstatus (0.008672) \n",
      "31. feature 9 - Fjob (0.008311) \n",
      "32. feature 0 - school (0.007740) \n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAzUAAAJZCAYAAACQrRt5AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzs3XuUnHWdJ/535QICEUwAc+nEhEsujWAIkMC48TfNaKIw\nwyUKGkRxMRodljOi7gBnx5EwzqjBUReBnc0yA4qriZ7VTVCTME6WZiC7sRVwUEBRGBLSkHBJ5JIo\nIen6/RHTJtCdrnRXJf0Nr9c5dU5V9/O869NPV3c/736qnqpUq9VqAAAACjVgXw8AAADQF0oNAABQ\nNKUGAAAomlIDAAAUTakBAACKptQAAABFU2oA2Cv+/M//PH/3d3+3r8cAYD9U8T41AP3buHHj8uST\nT2bQoEGpVqupVCp56KGHMmLEiF5n3nHHHXnf+96Xxx57rI6TluPiiy/OmDFj8jd/8zf7ehQA6sCR\nGoB+rlKp5Ac/+EGee+65PP/883nuuef6VGiSdJaj3tq2bVuf7n9f6ujo2NcjAFBnSg1AAbo7qL5q\n1ar8h//wHzJ06NBMmTIld9xxR+fnvvrVr+a4447LoYcemmOPPTb/43/8jyTJ5s2bc+aZZ+bxxx/P\na1/72hx66KFZt25dLr744nz605/uXP+OO+7ImDFjOm8fddRRueaaazJ58uQMGTIkHR0deeKJJ3Le\neefl9a9/fY455phcd9113X4NO+fvyP7CF76Q4cOHp6mpKUuWLMmyZcsyceLEHHHEEfnc5z7Xue7V\nV1+d888/P7Nnz86hhx6aU045Jffdd1/n53/xi1/k9NNPz9ChQ3PCCSfke9/73i73e8kll+RP//RP\n89rXvjb/9E//lG984xu55pprcuihh+acc85JksyfPz/HHntsDj300Bx//PFZvHhxZ8bXvva1vOUt\nb8lf/uVfZtiwYTnmmGOyfPnyzs9v3LgxH/zgB9PU1JTDDz8873znOzs/9/3vfz9TpkzJ0KFDM336\n9PzsZz/r/Nz8+fMzevToHHrooWlubs7tt9/e7fYDoHtKDUChHn/88fzZn/1ZPv3pT2fjxo35+7//\n+7zrXe/KM888kyQZPnx4li5dmueeey4333xzPv7xj+enP/1pDj744CxbtiyjRo3q8cjPy4/mLFq0\nKMuWLctvfvObVCqVnHXWWZkyZUqeeOKJrFixItdee21++MMf1jT/unXrsmXLljz++OO5+uqr8+EP\nfzjf+MY3cu+99+Zf//Vf85nPfCarV6/uXP7WW2/Ne97znmzcuDEXXHBBzj333Gzbti1bt27NWWed\nlXe84x156qmn8pWvfCUXXnhhfvWrX3Wuu3Dhwvz1X/91nn/++Vx00UW58MILc/nll+e5557LkiVL\nkiTHHntsVq5cmeeeey5XXXVV3ve+92X9+vWdGW1tbWlubs4zzzyTv/zLv8ycOXM6P/e+970vv/3t\nb/Pggw/mySefzMc//vEkyb333ps5c+bkxhtvzIYNG/KRj3wkZ599dl566aU89NBDueGGG3L33Xfn\nueeey2233ZZx48bVtO0A2JVSA1CAc889N8OGDcuwYcM6jwL8z//5P/Onf/qnefvb354keetb35pT\nTjklS5cuTZKcccYZnTvJb3nLWzJz5szceeedfZrjYx/7WEaNGpUDDzwwP/7xj/P000/nr/7qrzJw\n4MCMGzcuH/rQh7Jo0aKasg444ID8l//yXzJw4MDMnj07Tz/9dC677LIcfPDBOe6443Lcccfl3/7t\n3zqXP/nkkzNr1qwMHDgwn/jEJ/Liiy9m1apVWbVqVTZt2pQrrrgigwYNyumnn54/+7M/y8KFCzvX\nPeecc3LaaaclSQ488MAu53nXu96V4cOHJ0nOP//8jB8/Pm1tbZ2fHzt2bD74wQ+mUqnkAx/4QJ54\n4ok8+eSTWbduXW677bYsWLAghx56aAYOHJi3vOUtSZIbb7wxH/3oR3PKKaekUqnk/e9/fw488MCs\nWrUqAwcOzJYtW/Lzn/88W7duzRve8IYcddRRe/YNASCJUgNQhCVLlmTDhg3ZsGFDvvvd7yZJVq9e\nnW9/+9udZWfo0KFZuXJlnnjiiSTJsmXL8kd/9Ec5/PDDM3To0CxbtixPP/10n+YYPXp05/XVq1en\nvb19l/v/3Oc+lyeffLKmrMMPP7zzSNBBBx2UJHn961/f+fmDDjooL7zwQuftnZ8KV6lU0tTUlMcf\nfzyPP/74Lp9LtheQ9vb2Ltftzi233NL5NLGhQ4fm/vvv32V77Xw0a8e8L7zwQh577LEMGzYshx56\n6CsyV69enS9+8Yu7bKO1a9fm8ccfzzHHHJP/+l//a+bNm5fhw4fnve99b+f3DoA9o9QAFKCr19SM\nGTMmF110UWfZ2bhxY55//vlcfvnl2bJlS84777xcfvnleeqpp7Jx48acccYZnTldnSTgkEMOyebN\nmztvd7WDvfN6Y8aMydFHH73L/T/77LO7vJ6lnnY+U1u1Ws3atWszatSojBo1KmvWrNll2TVr1qSp\nqanLubu6vWbNmsydOzf/7b/9t2zcuDEbN27MG9/4xm5fy7SzMWPGZMOGDXnuuee6/Nxf/dVf7bKN\nXnjhhbznPe9JksyePTt33nln59Psrrzyyh7vD4BXUmoACvW+970v3/ve9/LP//zP6ejoyO9+97vc\ncccdefzxx7Nly5Zs2bIlRxxxRAYMGJBly5bln//5nzvXHT58eJ555plddsRPPPHELF26NBs3bsy6\ndety7bXX7vb+p02blte+9rW55ppr8rvf/S7btm3L/fffn5/85CcN+XrvvvvuLF68ONu2bcuXv/zl\nvOY1r8lpp52WU089NYccckiuueaabN26Na2trfn+97+fCy64oNus4cOH55FHHum8vWnTpgwYMCBH\nHHFEOjo6cvPNN+fnP/95TXONGDEiZ5xxRi655JL85je/ydatWzuf5vfhD384//2///fOp7Ft2rQp\nS5cuzaZNm/LQQw/l9ttvz5YtW3LAAQfkoIMOyoAB/iwD9IbfngD9XHenXh49enSWLFmSz372szny\nyCMzduzY/P3f/306OjoyZMiQfOUrX8n555+fYcOGZdGiRZ1n+UqSiRMn5oILLsjRRx+dYcOGZd26\ndXn/+9+fN73pTRk3blze8Y53ZPbs2budY8CAAfn+97+fn/70pznqqKPy+te/Ph/+8Ie7PGLRm6/z\n5bfPOeecfOtb38rQoUPzjW98I//7f//vDBw4MIMHD873vve9LF26NEcccUQuvfTSfP3rX8/48eO7\n3X5z5szJ/fff3/kapebm5nziE5/IaaedlhEjRuT+++/P9OnTa57361//egYNGpRJkyZl+PDhnYXw\n5JNPzo033phLL700w4YNy4QJE/K1r30tSfLiiy/myiuvzJFHHplRo0blqaee2uWMbwDUrqY331y+\nfHkuu+yydHR0ZM6cObniiiu6XO7HP/5x3vzmN+db3/pW5wtZa10XALpz9dVX5+GHH84tt9yyr0cB\noB/q8UhNR0dHLr300tx22225//77s3DhwvziF7/ocrkrr7yy8yw8e7IuAABAb/VYatra2jJ+/PiM\nHTs2gwcPzuzZszvP6b+z6667rvMN2PZ0XQAAgN4a1NMC7e3tu5wKc/To0buctz/Z/gZwixcvzu23\n377L52pZFwB6ctVVV+3rEQDox+pyooDLLrss8+fPr0cUAADAHunxSE1TU9Mu5/9fu3btLuf+T5Kf\n/OQnmT17dqrVap5++uksW7YsgwYNqmndHbo7uw8AAMAOXZ7nrNqDrVu3Vo855pjqo48+Wn3xxRer\nkydPrj7wwAPdLv8f/+N/rH7nO9/Z43VrGKXXrrrqqoZll55f8uyNzi959kbnlzx7o/NLnr3R+SXP\n3uj8kmdvdH7Jszc6v+TZG51f8uyNzi959mq1+87Q45GagQMH5vrrr8/MmTM7T8vc3NycBQsWpFKp\nZO7cubssv/MRl+7WBQAAqJceS02SvOMd78gvf/nLXT72kY98pMtlb7rpph7XBQAAqJeB8+bNm7ev\nh0i2v7FaI0cZN25cw7JLzy959kbnlzx7o/NLnr3R+SXP3uj8kmdvdH7Jszc6v+TZG51f8uyNzi95\n9kbnlzx7d52h8vvnpu1zlUql6xf9AAAApPvOUJdTOgMAAOwrSg0AAFA0pQYAACiaUgMAABRNqQEA\nAIqm1AAAAEVTagAAgKIpNQAAQNGUGgAAoGhKDQAAUDSlBgAAKJpSAwAAFE2pAQAAiqbUAAAARVNq\nAACAoik1AABA0ZQaAACgaEoNAABQNKUGAAAo2qB9PUCjtLZuv+y43tKy/XpLyx+uAwAA5atUq9Xq\nvh4iSSqVSho1SqWS9I+vEgAA6K3uOoOnnwEAAEVTagAAgKIpNQAAQNGUGgAAoGhKDQAAUDSlBgAA\nKJpSAwAAFE2pAQAAiqbUAAAARVNqAACAoik1AABA0ZQaAACgaEoNAABQNKUGAAAomlIDAAAUTakB\nAACKptQAAABFU2oAAICiKTUAAEDRlBoAAKBoSg0AAFA0pQYAACiaUgMAABRNqQEAAIqm1AAAAEVT\nagAAgKIpNQAAQNGUGgAAoGhKDQAAUDSlBgAAKJpSAwAAFE2pAQAAiqbUAAAARVNqAACAoik1AABA\n0ZQaAACgaEoNAABQNKUGAAAoWk2lZvny5Zk0aVImTJiQ+fPnv+Lzt956ayZPnpwpU6Zk2rRpWbly\nZefnxo0bt8vnAAAA6qlSrVaru1ugo6MjEyZMyIoVKzJq1KhMnTo1ixYtyqRJkzqX2bx5cw4++OAk\nyc9+9rO8+93vzoMPPpgkOfroo3P33Xdn6NChux+kUkkPo/RapZI0KBoAANhLuusMPR6paWtry/jx\n4zN27NgMHjw4s2fPzpIlS3ZZZkehSZIXXnghAwb8IbZaraajo6MvswMAAHSrx1LT3t6eMWPGdN4e\nPXp02tvbX7Hc4sWL09zcnLPOOis33XRT58crlUpmzJiRqVOn5sYbb6zT2AAAANvV7UQB5557bh58\n8MEsXrw4n/rUpzo/vnLlytxzzz1ZunRpbrjhhtx11131uksAAIAM6mmBpqamrFmzpvP22rVr09TU\n1O3y06dPzyOPPJINGzZk2LBhGTlyZJLkyCOPzKxZs9LW1pbp06d3ue68efM6r7e0tKSlpaXGLwMA\nANjftLa2prW1tcflejxRwLZt2zJx4sSsWLEiI0eOzLRp07Jw4cI0Nzd3LvPwww/nmGOOSZLcc889\nOeecc/LYY49l8+bN6ejoyJAhQ7Jp06bMnDkzV111VWbOnPnKQZwoAAAA2I3uOkOPR2oGDhyY66+/\nPjNnzkxHR0fmzJmT5ubmLFiwIJVKJXPnzs13vvOd3HLLLTnggANy0EEH5dvf/naSZP369Zk1a1Yq\nlUq2bt2aCy+8sMtCAwAA0Fs9HqnZWxypAQAAdqfXp3QGAADoz5QaAACgaEoNAABQNKUGAAAomlID\nAAAUTakBAACKptQAAABFU2oAAICiKTUAAEDRlBoAAKBoSg0AAFA0pQYAACiaUgMAABRNqQEAAIqm\n1AAAAEVTagAAgKIpNQAAQNGUGgAAoGhKDQAAUDSlBgAAKJpSAwAAFE2pAQAAiqbUAAAARVNqAACA\noik1AABA0ZQaAACgaEoNAABQNKUGAAAomlIDAAAUTakBAACKptQAAABFU2oAAICiKTUAAEDRlBoA\nAKBoSg0AAFA0pQYAACiaUgMAABRNqQEAAIqm1AAAAEVTagAAgKIpNQAAQNGUGgAAoGhKDQAAUDSl\nBgAAKJpSAwAAFE2pAQAAiqbUAAAARVNqAACAoik1AABA0ZQaAACgaEoNAABQNKUGAAAomlIDAAAU\nTakBAACKptQAAABFU2oAAICiKTUAAEDRlBoAAKBoSg0AAFA0pQYAACiaUgMAABStplKzfPnyTJo0\nKRMmTMj8+fNf8flbb701kydPzpQpUzJt2rSsXLmy5nUBAAD6olKtVqu7W6CjoyMTJkzIihUrMmrU\nqEydOjWLFi3KpEmTOpfZvHlzDj744CTJz372s7z73e/Ogw8+WNO6nYNUKulhlF6rVJIGRQMAAHtJ\nd52hxyM1bW1tGT9+fMaOHZvBgwdn9uzZWbJkyS7L7Cg0SfLCCy9kwIABNa8LAADQFz2Wmvb29owZ\nM6bz9ujRo9Pe3v6K5RYvXpzm5uacddZZuemmm/ZoXQAAgN4aVK+gc889N+eee27uuuuufOpTn8oP\nf/jDPc6YN29e5/WWlpa0tLTUazwAAKAwra2taW1t7XG5HktNU1NT1qxZ03l77dq1aWpq6nb56dOn\n55FHHsmGDRv2eN2dSw0AAPDq9vIDHVdffXWXy/X49LOpU6fm17/+dVavXp0tW7Zk0aJFOfvss3dZ\n5uGHH+68fs8992TLli0ZNmxYTesCAAD0RY9HagYOHJjrr78+M2fOTEdHR+bMmZPm5uYsWLAglUol\nc+fOzXe+853ccsstOeCAA3LQQQfl29/+9m7XBQAAqJceT+m8tzilMwAAsDu9PqUzAABAf6bUAAAA\nRavbKZ1fbVpbt192XN9xUoaWlj9cBwAAGs9ragrIBwAAvKYGAADYTyk1AABA0ZQaAACgaEoNAABQ\nNKUGAAAomlIDAAAUTakBAACKptQAAABFU2oAAICiKTUAAEDRlBoAAKBoSg0AAFA0pQYAACiaUgMA\nABRNqQEAAIqm1AAAAEVTagAAgKIpNQAAQNGUGgAAoGhKDQAAUDSlBgAAKJpSAwAAFE2pAQAAiqbU\nAAAARVNqAACAoik1AABA0ZQaAACgaEoNAABQNKUGAAAomlIDAAAUTakBAACKptQAAABFU2oAAICi\nKTUAAEDRlBoAAKBoSg0AAFA0pQYAACiaUgMAABRNqQEAAIqm1AAAAEVTagAAgKIpNQAAQNGUGgAA\noGhKDQAAUDSlBgAAKJpSAwAAFE2pAQAAiqbUAAAARVNqAACAoik1AABA0ZQaAACgaEoNAABQNKUG\nAAAomlIDAAAUTakBAACKptQAAABFq6nULF++PJMmTcqECRMyf/78V3z+m9/8ZiZPnpzJkydn+vTp\nue+++zo/N27cuEyePDlTpkzJtGnT6jc5AABAkkq1Wq3uboGOjo5MmDAhK1asyKhRozJ16tQsWrQo\nkyZN6lxm1apVaW5uzmGHHZbly5dn3rx5WbVqVZLk6KOPzt13352hQ4fufpBKJT2M0muVStKg6L2S\nDwAAdN8ZejxS09bWlvHjx2fs2LEZPHhwZs+enSVLluyyzGmnnZbDDjus83p7e3vn56rVajo6Ovo6\nPwAAQJd6LDXt7e0ZM2ZM5+3Ro0fvUlpe7h//8R9zxhlndN6uVCqZMWNGpk6dmhtvvLGP4wIAAOxq\nUD3Dbr/99tx888256667Oj+2cuXKjBw5Mk899VRmzJiR5ubmTJ8+vZ53CwAAvIr1WGqampqyZs2a\nzttr165NU1PTK5a77777Mnfu3CxfvnyX18+MHDkySXLkkUdm1qxZaWtr67bUzJs3r/N6S0tLWlpa\nav06AACA/Uxra2taW1t7XK7HEwVs27YtEydOzIoVKzJy5MhMmzYtCxcuTHNzc+cya9asyVvf+tZ8\n/etfz2mnndb58c2bN6ejoyNDhgzJpk2bMnPmzFx11VWZOXPmKwdxogAAAGA3uusMPR6pGThwYK6/\n/vrMnDkzHR0dmTNnTpqbm7NgwYJUKpXMnTs3n/nMZ7Jhw4ZccsklqVarGTx4cNra2rJ+/frMmjUr\nlUolW7duzYUXXthloQEAAOitHo/U7C2O1AAAALvT61M6AwAA9GdKDQAAUDSlBgAAKJpSAwAAFE2p\nAQAAiqbUAAAARVNqAACAoik1AABA0ZQaAACgaMWXmnEjRqRSqez2kqTHZSqVSsaNGLGPvxoAAGBP\nVarVanVfD5FsLx29GaVSqaSntSqppppKz1lJL2dI+sdWBACA/Vd3naH4IzUAAMCrm1IDAAAUTakB\nAACKptQAAABFU2oAAICiKTUAAEDRlBoAAKBoSg0AAFA0pQYAACiaUgMAABRNqQEAAIqm1AAAAEVT\nagAAgKIpNQAAQNGUGgAAoGhKDQAAUDSlBgAAKJpSAwAAFE2pAQAAiqbUAAAARVNqAACAoik1AABA\n0ZQaAACgaEoNAABQNKUGAAAomlIDAAAUTakBAACKptQAAABFU2oAAICiKTUAAEDRlBoAAKBoSg0A\nAFA0pQYAACiaUgMAABRNqQEAAIqm1AAAAEVTagAAgKIpNQAAQNGUGgAAoGhKDQAAUDSlBgAAKJpS\nAwAAFE2pAQAAiqbUAAAARVNqAACAoik1AABA0ZQaAACgaEoNAABQNKUGAAAomlIDAAAUraZSs3z5\n8kyaNCkTJkzI/PnzX/H5b37zm5k8eXImT56c6dOn57777qt5XQAAgL6oVKvV6u4W6OjoyIQJE7Ji\nxYqMGjUqU6dOzaJFizJp0qTOZVatWpXm5uYcdthhWb58eebNm5dVq1bVtG7nIJVKehil6y+gUklP\na1VSTTWVnrOSXs6Q9GI1AABgD3TXGXo8UtPW1pbx48dn7NixGTx4cGbPnp0lS5bsssxpp52Www47\nrPN6e3t7zesCAAD0RY+lpr29PWPGjOm8PXr06M7S0pV//Md/zBlnnNGrdQEAAPbUoHqG3X777bn5\n5ptz11131TMWAACgWz2WmqampqxZs6bz9tq1a9PU1PSK5e67777MnTs3y5cvz9ChQ/do3R3mzZvX\neb2lpSUtLS21fA0AAMB+qLW1Na2trT0u1+OJArZt25aJEydmxYoVGTlyZKZNm5aFCxemubm5c5k1\na9bkrW99a77+9a/ntNNO26N1OwdxogAAAGA3uusMPR6pGThwYK6//vrMnDkzHR0dmTNnTpqbm7Ng\nwYJUKpXMnTs3n/nMZ7Jhw4ZccsklqVarGTx4cNra2rpdFwAAoF56PFKztzhSAwAA7E6vT+kMAADQ\nnyk1AABA0ZQaAACgaEoNAABQNKUGAAAomlIDAAAUTakBAACKptQAAABFU2oAAICiKTUAAEDRlBoA\nAKBoSg0AAFA0pQYAACiaUgMAABRNqQEAAIqm1AAAAEVTagAAgKIpNQAAQNGUGgAAoGhKDQAAUDSl\nBgAAKJpSAwAAFE2pAQAAiqbU7Ma4ESNSqVR6vCTpcZlxI0bs468GAAD2T5VqtVrd10Mk20tBb0ap\nVCrpaa1Kqqmm0nNWsssMtWTXmv/ybAAAYM901xkcqQEAAIqm1AAAAEVTagAAgKIpNQAAQNGUGgAA\noGhKDQAAUDSlBgAAKJpSAwAAFE2pAQAAiqbUAAAARVNqAACAoik1AABA0ZQaAACgaEoNAABQNKUG\nAAAomlIDAAAUTakBAACKptQAAABFU2oAAICiKTUAAEDRlBoAAKBoSg0AAFA0pQYAACiaUgMAABRN\nqQEAAIqm1AAAAEVTagAAgKIpNQAAQNGUGgAAoGhKDQAAUDSlBgAAKJpSAwAAFE2pAQAAiqbUAAAA\nRVNqAACAoik1AABA0WoqNcuXL8+kSZMyYcKEzJ8//xWf/+Uvf5k3v/nNec1rXpMvfelLu3xu3Lhx\nmTx5cqZMmZJp06bVZ2oAAIDfG9TTAh0dHbn00kuzYsWKjBo1KlOnTs0555yTSZMmdS5z+OGH57rr\nrsvixYtfsf6AAQPS2tqaoUOH1ndyAACA1HCkpq2tLePHj8/YsWMzePDgzJ49O0uWLNllmSOOOCIn\nn3xyBg16ZUeqVqvp6Oio38QAAAA76bHUtLe3Z8yYMZ23R48enfb29prvoFKpZMaMGZk6dWpuvPHG\n3k0JAADQjR6fftZXK1euzMiRI/PUU09lxowZaW5uzvTp0xt9twAAwKtEj6Wmqakpa9as6by9du3a\nNDU11XwHI0eOTJIceeSRmTVrVtra2rotNfPmzeu83tLSkpaWlprvBwAA2L+0tramtbW1x+Uq1Wq1\nursFtm3blokTJ2bFihUZOXJkpk2bloULF6a5ufkVy1599dUZMmRIPvnJTyZJNm/enI6OjgwZMiSb\nNm3KzJkzc9VVV2XmzJmvHKRSSQ+jdP0FVCrpaa1Kqqmm0nNWsssMtWTXmv/ybAAAYM901xl6PFIz\ncODAXH/99Zk5c2Y6OjoyZ86cNDc3Z8GCBalUKpk7d27Wr1+fU045Jc8//3wGDBiQa6+9Ng888ECe\neuqpzJo1K5VKJVu3bs2FF17YZaEBAADorR6P1OwtjtQAAAC7011nqOnNNwEAAPorpQYAACiaUgMA\nABRNqQEAAIqm1AAAAEVTagAAgKIpNQAAQNGUGgAAoGhKDQAAUDSlBgAAKJpSAwAAFE2pAQAAiqbU\nAAAARVNqAACAoik1AABA0ZQaAACgaEoNAABQNKUGAAAomlIDAAAUTakBAACKptQAAABFU2oAAICi\nKTUAAEDRlBoAAKBoSg0AAFA0pQYAACiaUgMAABRNqQEAAIqm1AAAAEVTagAAgKIpNQAAQNGUGgAA\noGhKDQAAUDSlBgAAKJpSAwAAFE2pAQAAiqbUAAAARVNqAACAoik1AABA0ZQaAACgaEoNAABQNKUG\nAAAomlIDAAAUTakBAACKptQAAABFU2oAAICiKTUAAEDRlBoAAKBoSg0AAFA0pQYAACiaUgMAABRN\nqQEAAIqm1AAAAEVTagAAgKIpNQAAQNGUGgAAoGhKDQAAUDSlBgAAKJpSAwAAFE2pAQAAiqbUAAAA\nRaup1CxfvjyTJk3KhAkTMn/+/Fd8/pe//GXe/OY35zWveU2+9KUv7dG6AAAAfVGpVqvV3S3Q0dGR\nCRMmZMWKFRk1alSmTp2aRYsWZdKkSZ3LPP3001m9enUWL16coUOH5hOf+ETN63YOUqmkh1G6/gIq\nlfS0ViXVVFPpOSvZZYZasmvNf3k2AACwZ7rrDD0eqWlra8v48eMzduzYDB48OLNnz86SJUt2WeaI\nI47IySefnEGDBu3xugAAAH3RY6lpb2/PmDFjOm+PHj067e3tNYX3ZV0AAIBaDOp5kb1n3rx5nddb\nWlrS0tJn04qVAAAd00lEQVSyz2YBAAD2rdbW1rS2tva4XI+lpqmpKWvWrOm8vXbt2jQ1NdU0xJ6u\nu3OpAQAAXt1efqDj6quv7nK5HkvN1KlT8+tf/zqrV6/OyJEjs2jRoixcuLDb5Xd+4c6erltPrfnj\ntKYlSfLHac28XJUkaUlrWnLHXpkBAABovB7PfpZsPy3zxz72sXR0dGTOnDm58sors2DBglQqlcyd\nOzfr16/PKaeckueffz4DBgzIkCFD8sADD2TIkCFdrtvlIA08+1nNWXH2MwAA6K+66ww1lZq9QakB\nAAB2p7vO0K9OFMAftLZuv+y4vuOphC0tf7gOAAA4UrNrVvrnkZpKJekf3yUAANh3ev3mmwAAAP2Z\nUgMAABRNqQEAAIqm1AAAAEVTagAAgKIpNQAAQNGUGgAAoGhKDQAAUDSlBgAAKJpSAwAAFE2pAQAA\niqbUAAAARVNqAACAoik1AABA0ZQaAACgaEoNAABQNKUGAAAomlIDAAAUTakBAACKptQAAABFU2oA\nAICiKTUAAEDRlBoAAKBoSg0AAFA0pQYAACiaUgMAABRNqQEAAIqm1AAAAEVTagAAgKIpNQAAQNGU\nGgAAoGhKDQAAUDSlBgAAKJpSAwAAFE2pAQAAiqbUAAAARVNqAACAoik1AABA0ZQaAACgaEoNAABQ\nNKUGAAAomlIDAAAUrVKtVqv7eogkqVQq6c0olUol9foCKskuM+wuuzV/nNa0/P56S1rSmiRpSWta\nckeP2Xs0VyXpH98lAADYd7rrDErNzlmpvdT0NXuP1lVqAACg287g6WcAAEDRlBoAAKBoSg0AAFA0\npQYAACiaUgMAABRNqQEAAIqm1AAAAEVTavahcSNGpFKp9HhJ0uMy40aM2MdfDQAA7BvefHPnrOzd\nN9+sNb+Saqqp7HE+AADsT7z5JgAAsF9SagAAgKIpNQAAQNGUGgAAoGhKDQAAULSaSs3y5cszadKk\nTJgwIfPnz+9ymb/4i7/I+PHjc+KJJ+bee+/t/Pi4ceMyefLkTJkyJdOmTavP1AAAAL83qKcFOjo6\ncumll2bFihUZNWpUpk6dmnPOOSeTJk3qXGbZsmV5+OGH86tf/So/+tGP8ud//udZtWpVkmTAgAFp\nbW3N0KFDG/dVAAAAr1o9Hqlpa2vL+PHjM3bs2AwePDizZ8/OkiVLdllmyZIlueiii5Ikp556ap59\n9tmsX78+yfb3Tuno6GjA6AAAADWUmvb29owZM6bz9ujRo9Pe3r7bZZqamjqXqVQqmTFjRqZOnZob\nb7yxXnMDAAAkqeHpZ321cuXKjBw5Mk899VRmzJiR5ubmTJ8+vdF3CwAAvEr0WGqampqyZs2azttr\n165NU1PTK5Z57LHHulxm5MiRSZIjjzwys2bNSltbW7elZt68eZ3XW1pa0tLSUvMXAgAA7F9aW1vT\n2tra43KVarVa3d0C27Zty8SJE7NixYqMHDky06ZNy8KFC9Pc3Ny5zNKlS3PDDTfkBz/4QVatWpXL\nLrssq1atyubNm9PR0ZEhQ4Zk06ZNmTlzZq666qrMnDnzlYNUKulhlK6/gEole75WN1nJLjM0MntP\n8iuppprKHud3p7V1+2XH9R3dsaXlD9cBAKC/6a4z9HikZuDAgbn++uszc+bMdHR0ZM6cOWlubs6C\nBQtSqVQyd+7cnHnmmVm6dGmOPfbYHHLIIbn55puTJOvXr8+sWbNSqVSydevWXHjhhV0WGvaunctL\npfKHggMAACXq8UjN3uJIze7Wre+Rml1nSOr9CHAkCACARuiuMyg1O2dFqak3pQkAgHpRamrJilJT\nb6XnAwDQf3TXGXp8nxoAAID+TKkBAACKptQAAABFU2oAAICiKTUAAEDRlBoAAKBoSg0AAFA0pQYA\nACiaUgMAABRNqQEAAIqm1AAAAEVTagAAgKIpNQAAQNEG7esBgDK0tm6/7Lje0rL9ekvLH64DAOwL\nlWq1Wt3XQyRJpVJJb0apVCqp1xdQSXaZoZHZe5JfSTXVVPY4v6a5KkkjHwGl5Ze+47635m/09xUA\noCvddQalZuesKDX1VnJ+I7L3ZmmybQCA/Y1SU0tWlJp6Kzm/5NkbnV/i7EoTAJRPqaklK0pNvZWc\nX/Lsjc4vefa9kQ8ANEZ3ncHZzwAAgKIpNfuxcSNGpFKp7PaSpMdlKpVKxo0YsY+/GgAA6Jqnn+2c\nlf3r6We15NeS3VX+uBEjsnr9+h7XS6q/X7t7Y4cPz6Pr1u3ysUbn16L0p0B5+tney/d6HQDYO7ym\nppasKDW15veH2fuSX4vSdqz3Zn7Jszc63+t1AKBxvKYGAADYLw3a1wMAsHue3gYAu+fpZztnxdPP\nas3vD7P3Jb8WJT8FqtH5Jc/e6PwSZ1eaACiF19TUkhWlptb8/jB7X/JrUeLO6d7KL3n2RueXPPve\nyK+3RheykgtfybMDdEepqSUrSk2t+f1h9r7k16L0nUc77vsmv+TZG5G/N3esbZvulVZWAbqj1NSS\nFaWm1vz+MHtf8mtR2g7S3swvefZG55c8e6PzS5690fklzw6wN3XXGZwoAADYY4080uSpc8CecqRm\n56w4UlNrfn+YvS/5tSj9P6f+q7xv8kuevdH5Jc/e6PySZ290vpNjADvz9LNasqLU1JrfH2bvS34t\nSt4JaHR+ybM3Or/k2RudX/Lsjc4vefZG55c8O1B/nn4GOxk3YkRWr1/fw1LVVCo9F76xw4fn0XXr\n6jMYAMUr+al5jmJRKkdqds6KIzW15veH2Rud39ttU1thSpLq79fuXleFqX/k95zdXX4tSv7PbMmz\nNzq/5NkbnV/y7I3OL3n2RueXOLvSRF95+lktWVFqas3vD7M3Ot+26Vt2d/n1Kk19OUJmJ2Pf5Jc8\ne6PzS5690fklz97o/JJn3xv57J+Umlqy0n9KTWv+OK1p+f31lrSkNUnSkta05I4+5/9hPTvufcnu\nKr8/zN7o/L6Umnrl974wJY0sTaXvBNgB2zf5Jc/e6PySZ290fsmz74189k9KTS1Z6T+lZm/l23Hv\nW3ZX+f1h9kbn99dS0+ht46l5/Te79PySZ290fsmzNzq/5Nkbke+pba8OSk0tWVFqas3vDzvWjc63\nbfqW3ej8/XHb1KK0nYy9lV16fsmzNzq/5NkbnV/y7I3OdxRo/+XsZwAA0EuOBPVvjtTsnBVHamrN\n7w//EW90vm3Tt+xG5++P28ZJFF69+SXP3uj8kmdvdH7Jszc6v+TZ2b3uOsOAfTALAF1YvX59qslu\nL+nh89Xf57zcuBEjUqlUerwk6XGZcSNGNHIzAMAeU2oAXgVqKUyNLk1Jz4Wpu9JUr3yFDGD/pNQA\n0Gf1OsrUXWkq+SiWwgfQeF5Ts3NWXh2vqdnT98DpKn9fvnbBe/js+3yvqelbdqPzbZv6Zzc6v79u\nm0a/91OjT5O+v5+GvZHZpeeXOLsTEdSmu86g1OyclVdHqalHfn8oHbVSauqf3193wGybvmU3Ot+2\n6Vt2o/Ntmz3L92a//Tu/5Nkbkd/owrQ3C5lSU0tWlJpa80uevdZ8paZv2Y3Ot23qn93ofNumb9mN\nzrdt9iy/0dvGUay+Ka107M38kmffnu/sZwAARSjldWpd5TvbIvuCN98EAKBudhSmnlSSHper1PQU\nP3CkBgCAgjT6jIKUyZEaAACKUcuRoFqOAiWOBO1PlBrYyc5nbvvjtGZerkqy+9Nd96d8AKBvaj2J\nwo4jQrvT25MosOec/WznrLz6zvDV2/ySZ290/r6cvb+/h0898/vTmZrq8d5Ptc5f2rbZ0+xG59s2\n9c9udH5/2DZd5ds23efbNrvPr4Wzn+0u3ymde85K/9k57e/5Jc/e6PySZ99dfqN33Pfm+xs1egev\nFnZO902+bVP/7Ebn94dt01W+bdN9vm2z+/xaKDW7y1dqes7Kq2PntB75Jc/e6PySZy8pv78exarV\nq7Hw9YedDNum/tmNzu8P26arfNum+3zbpvv8Rr9p6/7+/kPb85WanrNi57TW/JJnb3R+ybOXlF/y\n7I3Ot3Na/+xG59s2fcveF/m2Tff5tk33+aVvm/5RmpSanrNi57TW/JJnb3R+ybOXlF/y7I3O709H\nsbzeqJwdMNtmz/Jtm+7zbZvu822b3efXQqmpJSt2TmvNL3n2RueXPHtJ+SXP3uj8kmff03xPzdt3\npebVdGKSrvI9brrPt226z7dtdp9fC6WmlqzYOa01v+TZG51f8uwl5Zc8e6PzS5690fl23OufXav+\ndISvHvkeN93n23HvPt+22X1+LZSaWrJi57TW/JJnb3R+ybOXlF/y7I3OL3n2Ruf3p9n3p6fm1WJ/\ny/e46T7fjnv3+bbN7vNrodTUkpX+8QurhPySZ290fsmzl5Rf8uyNzi959kbnlzz7nuaXfMa/EvJL\nnn13+Y5iKTX7Mr8WfSo1y5cvz2WXXZaOjo7MmTMnV1xxxSuW+Yu/+IssW7YshxxySL761a/mxBNP\nrHnd3Q3YE7+w9k1+ybM3Or/k2UvKL3n2RueXPHuj80uevdH5Jc++L/JLnr3R+SUXst3l77qeUtOI\n/Fp02xmqPdi2bVv1mGOOqT766KPVLVu2VCdPnlx98MEHd1lm6dKl1TPPPLNarVarq1atqp566qk1\nr7tDDaN0v14Pl9trWKaavGKGWrJrze/q6+sP+baNbbM3t009820b28a2sW32Zb5t8+reNsntDd02\nteTv7dn3Rn4tultvQE9tqK2tLePHj8/YsWMzePDgzJ49O0uWLNllmSVLluSiiy5Kkpx66ql59tln\ns379+prW3Rta5e+T7NLzG5lden4js0vPb2R26fmNzC49v5HZpec3Mrv0/EZml57fyOy9cw+NzG9k\n9t7I79qgnhZob2/PmDFjOm+PHj06bW1tPS7T3t5e07oAANDf7fz0trF5NPNyVZLdP72tv+Q3evb+\noMdS0xvbjwwBAMD+oSV37FIA5hWU3+jZ+0Vp6ul5a//v//2/6tvf/vbO25/73Oeqn//853dZ5iMf\n+Uh10aJFnbcnTpxYXbduXU3r7vz8OBcXFxcXFxcXFxcXl91dutLjkZqpU6fm17/+dVavXp2RI0dm\n0aJFWbhw4S7LnH322bnhhhvynve8J6tWrcrrXve6DB8+PEcccUSP6+7g6A4AANAbPZaagQMH5vrr\nr8/MmTM7T8vc3NycBQsWpFKpZO7cuTnzzDOzdOnSHHvssTnkkENy880373ZdAACAeuk3b74JAADQ\nGz2e0rk0L774Yk499dRMmTIlJ5xwQq6++uokyf/6X/8rxx9/fAYOHJh77rmnrtkbN27MzJkzM3Hi\nxLz97W/Ps88+26v8OXPmZPjw4XnTm970is998YtfzIABA7Jhw4ZeZSfJ2rVr8yd/8id54xvfmBNO\nOCFf+cpXkiT/9m//lje/+c2ZPHlyzjnnnLzwwgt9zr7uuuuSJFdffXVGjx6dk046KSeddFKWL1/e\n6/l7+jrq6dlnn83555+f5ubmvPGNb8yPfvSjumU/9NBDmTJlSk466aRMmTIlhx12WJ++ht19X//o\nj/4oU6ZMybRp0/KTn/ykV/ldPS4//elPZ/LkyZkyZUre8Y53ZN26dXWZfcfjplH5L9/Off256m7+\n2bNndz7mjzrqqJx00km9yu9q29frZ6q72S+//PI0NzfnxBNPzLve9a4899xzdZu9Xr8ru8uv12N+\nd4+b6667Ls3NzTnhhBNy5ZVX9nr+Hbr7u1Ivu/u7Uq+8ej1mutPR0ZGTTjopZ599dp+zupq/kT+v\nSf0eM13l12PfJun+MV+v/Jf78pe/nOOPPz5vetObcuGFF2bLli19yutq29x333193rdJut829fx9\ntrNx48Z1/v2bNm1aXTJ3du211+aEE05o2P7T8uXLM2nSpEyYMCHz58+ve/5u9XSigBJt2rSpWq1W\nq1u3bq2eeuqp1R/96EfVX/ziF9WHHnqoevrpp1fvvvvuumZffvnl1fnz51er1Wr185//fPWKK67o\nVfadd95Zvffee6snnHDCLh9/7LHHqm9/+9ur48aNqz7zzDO9nv2JJ56o3nvvvdVqtVp9/vnnqxMn\nTqw+8MAD1alTp1bvvPPOarVard58883Vv/7rv+5z9oQJE6oPPvhgdd68edUvfvGLvZ55T+6rnj7w\ngQ9Ub7rppmq1Wq2+9NJL1Weffbau+Tts27atOnLkyOqaNWt6ndHd93XmzJnV2267rVqtbn+D3JaW\nll7ld/W4fP755zuvf+UrX6l+9KMfrcvsO76Xjc6vVuvzc1XLY/GTn/xk9TOf+Uyv8rva9vX6mepu\n9h/+8IfVbdu2VavVavWKK66oXnnllXWbvV6/K7vLr9djvrttc/vtt1dnzJhRfemll6rVarX61FNP\n9Xr+nXX1d6Veuvu7Us+8ej1muvOlL32peuGFF1bPOuusPmf1tD3q/fNaz8dMV/n12rfp7jFfr/yd\ntbe3V4866qjqiy++WK1Wq9V3v/vd1a997Wt9yuxq29Rj36Za7X7b1PP32c6OOuqo6oYNG+qS9XI/\n//nPqyeccEL1d7/7XXXr1q3VGTNmVB9++OG65W/btq16zDHHVB999NHqli1bqpMnT677/tnu7HdH\napLk4IMPTrL9P2Bbt25NpVLJxIkTM378+D6fkKCr7CVLluQDH/hAkuQDH/hAFi9e3Kvs6dOnZ+jQ\noa/4+Mc//vF84Qtf6P3QvzdixIiceOKJSZIhQ4Zk0qRJaW9vz69+9atMnz49SfK2t70t3/nOd/qc\n3dzcnPb29iT1PwnE7u6rHp577rnceeedufjii5MkgwYNyqGHHlq3/J39y7/8S4455phd3s9pT3X1\nfX388cczYMCAzv8c/eY3v0lTU1Ov8rt6XA4ZMqTz+qZNmzJgQO9+lXT3vWx0flKfn6taHovf/va3\nc8EFF/Qqv7vfCfX4mepu9re97W2d2/u0007L2rVre5Xf1ez1+l3ZXX69HvPdbZt/+Id/yJVXXplB\ng7a/HPWII47o9fw76+rvSr109xiqZ169HjNdWbt2bZYuXZoPfehDdcnraXvU++e1no+ZrvLrtW/T\n3WO+Xvkvt23btmzatClbt27N5s2bM2rUqD7ldbVt6rFvk3S9bdauXVvX32c7q1ar6ejoqEvWyz34\n4IM59dRTc+CBB2bgwIH5//6//y/f/e5365bf1taW8ePHZ+zYsRk8eHBmz56dJUuW1C2/J/tlqeno\n6MiUKVMyYsSIzJgxI1OnTm1o9vr16zN8+PAk2x/8Tz75ZN3u79Zbb82YMWNywgkn1C0zSR599NH8\n9Kc/zWmnnZY3vvGNufXWW5Ns/4Xe1z9IO7JPPfXUJMn111+fE088MR/60Ifqdni2u/uqh3//93/P\nEUcckYsvvjgnnXRS5s6dm9/+9rd1y9/Zt771rV7/Ae3Kztvjy1/+cv7zf/7PecMb3pDLL788n/vc\n5+p2P0nyqU99Km94wxvyzW9+M3/zN3/T57yXfy8bmd+In6uuHot33nlnRowYkWOOOaZu95PU/2eq\nu5+jm266KWeccUaf83d48sknG/a7MklDHvM7b5uHHnoo//qv/5rTTjstp59+eq+f3vZyjfybtbfV\n+zGz458P9Sx63WnEz2ujHjON1Ii/qzsbNWpUPvnJT+YNb3hDmpqa8rrXvS5ve9vb6n4/9d63SXbd\nd2rUvl+lUun8PXDjjTfWJXOH448/PnfeeWc2btyYzZs3Z+nSpXnsscfqlt/e3r7LP2lHjx5d1386\n92S/LDUDBgzIvffem7Vr1+ZHP/pRHnjggYZkt7W15f7773/FL9t6/fL97W9/m89+9rO7PMe6Hv8t\neeGFF3Leeefl2muvzZAhQ/JP//RPueGGGzJ16tRs2rQpBxxwQN2yL7nkkjzyyCP56U9/mhEjRuQT\nn/hEn+fv7r7qZevWrbnnnnvyn/7Tf8o999yTgw8+OJ///Ofrlr/DSy+9lFtvvTXnn39+XfJevj3+\n4R/+Iddee23WrFmTL3/5y/ngBz9Yl/vZ4W//9m+zZs2aXHjhhZ2vx+itrr6XjcofOHBg3X+uunss\nLly4sK6lNUndf6a6m/3v/u7vMnjw4Lz3ve/t68jdqveOar0f8y/fNlu3bs3GjRuzatWqXHPNNXn3\nu99dl7kb+Tdrb6r3Y+YHP/hBhg8fnhNPPDHVarXhb/3QiJ/XRj1mGqVRf1d39pvf/CZLlizJ6tWr\n8/jjj+eFF17IN7/5zbrfz0033VS3fZvkldumUft+K1euzD333JOlS5fmhhtuyF133VWX3CSZNGlS\nrrjiisyYMSNnnnlmpkyZkoEDB9Ytf1/bL0vNDoceemhOP/30urw4vavslpaWLF++PMOHD8/69euT\nJOvWrcvrX//6utzHww8/nEcffTSTJ0/OUUcdlbVr1+bkk0/u038Dtm7dmvPOOy/vf//7c8455yTZ\nfvj6tttuy49//OPMnj271/+l6ir7yCOP7PxB//CHP5wf//jHvZ69p/uql9GjR2fMmDE55ZRTkiTn\nnXdeXV8gucOyZcty8skn58gjj+xzVlfb42tf+1rOPffcJNu/hra2tj7fT1fe+9739vqwftLz97Le\n+fX+uepu/m3btuW73/1u3vOe9/R69q7U82equ9m/+tWvZunSpXXf0WjU78od6vmY72rbjBkzJu98\n5zuTbH8PtwEDBuSZZ57p++C/18i/WY3WiMfMypUrc+utt+boo4/OBRdckNtvvz0XXXRR3fJ31qif\n10Y/ZuqpkX9Xd/Yv//IvOfroozNs2LAMHDgw73znO/N//+//rfv9TJgwoS77NknX26ZRv89GjhyZ\nZPvv+lmzZtX9b/fFF1+cn/zkJ2ltbc3rXve6TJgwoW7ZTU1NWbNmTefttWvX9vppwL2x35Wap59+\nuvPpGL/97W/zwx/+MJMmTdplmd7+t6er7Obm5px99tn56le/mmT7H9W+/DLY+b9Rxx9/fNatW5dH\nHnkk//7v/57Ro0fn3nvv7dMPzgc/+MEcd9xx+djHPtb5saeeeirJ9qdA/O3f/m0++tGP1i1757NW\nffe7383xxx/fy8l7vq96GT58eMaMGZOHHnooSbJixYocd9xxdb+fev5XsKvt0dTUlDvuuCPJ9q+h\nL7+4Xv5f0l//+ted1xcvXtyn95/qavZG5tf756q7x+KO3w99fa74y7d9PX+mupp9+fLl+cIXvpBb\nb701Bx54YK+zk1fOXs/flV3l1/Mx39W2Offcc/N//s//SbL9aUUvvfRSDj/88F7fR1Lb36y+qvdR\njpfn1fMxs7PPfvazWbNmTR555JEsWrQof/Inf5Jbbrmlz7ldbY9G/bzW+zGzu+9lX7/HPf1drddj\n6A1veENWrVqV3/3ud6lWq1mxYkVd3sPw5dumXvs2/397d+ySThjHcfxoCP8Ah4IKBLkC73juUMGl\nKYKmQpyECtf+BLeGoNmx/6DNTcFFh1ojmiWoIVrKCGw4h0+TEeHZD338wf1+79cf8LkvX5+7e74o\nj44zuTe2n2eO4zgfHx9fp7QNh0On0+lY2zeNjfvy+PjoNJtNq9/GF4tFp9/vOw8PD04URc7l5aWV\nUwv/2GLPIfj77u7uFIahjDHyfV9nZ2eSpGazqbW1NaVSKa2srGhvb89a9svLi3Z2duS6rnZ3dzUY\nDGaqvVqtanV1VcvLy1pfX/86fWssk8nMdfrZ1dWVlpaWZIxREAQKw1DtdluNRkOu62pzc1P1et1q\n9tHRkXzflzFGBwcHen5+nrn+365l0+3trQqFgowxKpfLent7s5o/HA6VTqf1/v4+d1ZcP66vr5XP\n5xUEgUqlkm5ubmbKn7QuK5WKPM+TMUb7+/t6enqyWvui87+b576all+r1XRxcTFT7tik3tu6pybV\n3mq1lM1mtbGxoTAMFYahTk5OrNX++vpq5VkZl29rzcd9rlEU6fDwUJ7nKZ/Pq9frzVz/WNx7xZbf\n3is28mytmWl6vZ6V08/i+rGo+3U0GllbM5PybextpPg1byv/p9PTU21tbcn3fR0fHyuKornyJvXG\nxt5Giu+Nrb3fd/f391/X8TxP5+fnc2f+tL29rVwupyAI1O12ree32225rqtsNruQ+qfhzzcBAAAA\nJNo/9/MzAAAAAP8XhhoAAAAAicZQAwAAACDRGGoAAAAAJBpDDQAAAIBEY6gBAAAAkGgMNQAAAAAS\njaEGAAAAQKJ9Ask9hYhHDpkeAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x99bbfd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.ensemble import ExtraTreesClassifier\n",
    "\n",
    "# Build a classification task using 3 informative features\n",
    "\n",
    "\n",
    "# Build a forest and compute the feature importances\n",
    "forest = ExtraTreesClassifier(n_estimators=250,\n",
    "                              random_state=0)\n",
    "\n",
    "forest.fit(X, y)\n",
    "importances = forest.feature_importances_\n",
    "std = np.std([tree.feature_importances_ for tree in forest.estimators_],\n",
    "             axis=0)\n",
    "indices = np.argsort(importances)[::-1]\n",
    "\n",
    "# Print the feature ranking\n",
    "print(\"Feature ranking:\")\n",
    "\n",
    "for f in range(X.shape[1]):\n",
    "    print(\"%d. feature %d - %s (%f) \" % (f + 1, indices[f], features[indices[f]], importances[indices[f]]))\n",
    "\n",
    "# Plot the feature importances of the forest\n",
    "plt.figure(num=None, figsize=(14, 10), dpi=80, facecolor='w', edgecolor='k')\n",
    "plt.title(\"Feature importances\")\n",
    "plt.bar(range(X.shape[1]), importances[indices],\n",
    "       color=\"r\", yerr=std[indices], align=\"center\")\n",
    "plt.xticks(range(X.shape[1]), indices)\n",
    "plt.xlim([-1, X.shape[1]])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.24063511,  0.18037904,  0.05871839,  0.03224845,  0.03128223])"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "importances[indices[:5]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1. feature 31 - G2 (0.240635)\n",
      "2. feature 30 - G1 (0.180379)\n",
      "3. feature 14 - failures (0.058718)\n",
      "4. feature 29 - absences (0.032248)\n",
      "5. feature 25 - goout (0.031282)\n"
     ]
    }
   ],
   "source": [
    "for f in range(5):\n",
    "    print(\"%d. feature %d - %s (%f)\" % (f + 1, indices[f], features[indices[f]] ,importances[indices[f]]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "best_features = []\n",
    "for i in indices[:5]:\n",
    "    best_features.append(features[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAecAAAF6CAYAAAAj9ZDJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nsAzhDACAZQhnAAAsQzgDAGCZ/wdX8OppbhO1+wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xa81f550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the top 5 feature importances of the forest\n",
    "plt.figure(num=None, figsize=(8, 6), dpi=80, facecolor='w', edgecolor='k')\n",
    "plt.title(\"Feature importances\")\n",
    "plt.bar(range(5), importances[indices][:5], \n",
    "       color=\"r\",  yerr=std[indices][:5], align=\"center\")\n",
    "plt.xticks(range(5), best_features)\n",
    "plt.xlim([-1, 5])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Decision Tree accuracy and time elapsed caculation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "DecisionTree\n",
      "Acurracy:  0.886075949367\n",
      "time elapsed:  0.007000446319580078\n"
     ]
    }
   ],
   "source": [
    "t0=time()\n",
    "print (\"DecisionTree\")\n",
    "\n",
    "dt = DecisionTreeClassifier(min_samples_split=20,random_state=99)\n",
    "# dt = DecisionTreeClassifier(min_samples_split=20,max_depth=5,random_state=99)\n",
    "\n",
    "clf_dt=dt.fit(X_train,y_train)\n",
    "\n",
    "print (\"Acurracy: \", clf_dt.score(X_test,y_test))\n",
    "t1=time()\n",
    "print (\"time elapsed: \", t1-t0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## cross validation for DT"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cross result========\n",
      "[ 0.89873418  0.86075949  0.84810127  0.82278481  0.86075949]\n",
      "0.858227848101\n",
      "time elapsed:  0.039002180099487305\n"
     ]
    }
   ],
   "source": [
    "tt0=time()\n",
    "print (\"cross result========\")\n",
    "scores = cross_validation.cross_val_score(dt, X,y, cv=5)\n",
    "print (scores)\n",
    "print (scores.mean())\n",
    "tt1=time()\n",
    "print (\"time elapsed: \", tt1-tt0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Tuning our hyperparameters using GridSearch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done  34 tasks      | elapsed:   19.6s\n",
      "[Parallel(n_jobs=-1)]: Done  81 out of  81 | elapsed:   20.2s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 27 candidates, totalling 81 fits\n",
      "Best score: 0.933\n",
      "Best parameters set:\n",
      "\tclf__max_depth: 50\n",
      "\tclf__min_samples_leaf: 2\n",
      "\tclf__min_samples_split: 5\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.90      0.73      0.80        59\n",
      "          1       0.85      0.95      0.90        99\n",
      "\n",
      "avg / total       0.87      0.87      0.86       158\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "\n",
    "pipeline = Pipeline([\n",
    "    ('clf', DecisionTreeClassifier(criterion='entropy'))\n",
    "])\n",
    "\n",
    "parameters = {\n",
    "    'clf__max_depth': (5, 25 , 50),\n",
    "    'clf__min_samples_split': (1, 5, 10),\n",
    "    'clf__min_samples_leaf': (1, 2, 3)\n",
    "}\n",
    "\n",
    "grid_search = GridSearchCV(pipeline, parameters, n_jobs=-1, verbose=1, scoring='f1')\n",
    "grid_search.fit(X_train, y_train)\n",
    "\n",
    "print 'Best score: %0.3f' % grid_search.best_score_\n",
    "print 'Best parameters set:'\n",
    "\n",
    "best_parameters = grid_search.best_estimator_.get_params()\n",
    "for param_name in sorted(parameters.keys()):\n",
    "    print '\\t%s: %r' % (param_name, best_parameters[param_name])\n",
    "\n",
    "predictions = grid_search.predict(X_test)\n",
    "\n",
    "print classification_report(y_test, predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Random Forest accuracy and time elapsed caculation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RandomForest\n",
      "Acurracy:  0.917721518987\n",
      "time elapsed:  0.5780332088470459\n"
     ]
    }
   ],
   "source": [
    "t2=time()\n",
    "print (\"RandomForest\")\n",
    "rf = RandomForestClassifier(n_estimators=100,n_jobs=-1)\n",
    "clf_rf = rf.fit(X_train,y_train)\n",
    "print (\"Acurracy: \", clf_rf.score(X_test,y_test))\n",
    "t3=time()\n",
    "print (\"time elapsed: \", t3-t2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## cross validation for RF"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cross result========\n",
      "[ 0.88607595  0.93670886  0.84810127  0.88607595  0.92405063]\n",
      "0.896202531646\n",
      "time elapsed:  2.511143684387207\n"
     ]
    }
   ],
   "source": [
    "tt0=time()\n",
    "print (\"cross result========\")\n",
    "scores = cross_validation.cross_val_score(rf, X,y, cv=5)\n",
    "print (scores)\n",
    "print (scores.mean())\n",
    "tt1=time()\n",
    "print (\"time elapsed: \", tt1-tt0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Receiver Operating Characteristic (ROC) curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.90010272213662046"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "roc_auc_score(y_test,rf.predict(X_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiAAAAGHCAYAAACJeOnXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x9bfc208>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import roc_curve, auc\n",
    "\n",
    "predictions = rf.predict_proba(X_test)\n",
    "\n",
    "false_positive_rate, recall, thresholds = roc_curve(y_test, predictions[:, 1])\n",
    "roc_auc = auc(false_positive_rate, recall)\n",
    "plt.title('Receiver Operating Characteristic')\n",
    "plt.plot(false_positive_rate, recall, 'b', label='AUC = %0.2f' % roc_auc)\n",
    "plt.legend(loc='lower right')\n",
    "\n",
    "plt.plot([0, 1], [0, 1], 'r--')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.0])\n",
    "plt.ylabel('Recall')\n",
    "plt.xlabel('Fall-out')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Tuning Models using GridSearch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done  34 tasks      | elapsed:   21.5s\n",
      "[Parallel(n_jobs=-1)]: Done 184 tasks      | elapsed:   36.3s\n",
      "[Parallel(n_jobs=-1)]: Done 324 out of 324 | elapsed:   58.4s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 108 candidates, totalling 324 fits\n",
      "Best score: 0.916\n",
      "Best parameters set:\n",
      "\tclf__max_depth: 5\n",
      "\tclf__min_samples_leaf: 3\n",
      "\tclf__min_samples_split: 1\n",
      "\tclf__n_estimators: 50\n",
      "Accuracy: 0.905063291139\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.94      0.80      0.86        59\n",
      "          1       0.89      0.97      0.93        99\n",
      "\n",
      "avg / total       0.91      0.91      0.90       158\n",
      "\n"
     ]
    }
   ],
   "source": [
    "\n",
    "pipeline2 = Pipeline([\n",
    "('clf', RandomForestClassifier(criterion='entropy'))\n",
    "])\n",
    "\n",
    "parameters = {\n",
    "    'clf__n_estimators': (5, 25, 50, 100),\n",
    "    'clf__max_depth': (5, 25 , 50),\n",
    "    'clf__min_samples_split': (1, 5, 10),\n",
    "    'clf__min_samples_leaf': (1, 2, 3)\n",
    "}\n",
    "\n",
    "grid_search = GridSearchCV(pipeline2, parameters, n_jobs=-1, verbose=1, scoring='accuracy', cv=3)\n",
    "\n",
    "grid_search.fit(X_train, y_train)\n",
    "\n",
    "print 'Best score: %0.3f' % grid_search.best_score_\n",
    "\n",
    "print 'Best parameters set:'\n",
    "best_parameters = grid_search.best_estimator_.get_params()\n",
    "\n",
    "for param_name in sorted(parameters.keys()):\n",
    "    print '\\t%s: %r' % (param_name, best_parameters[param_name])\n",
    "\n",
    "predictions = grid_search.predict(X_test)\n",
    "print 'Accuracy:', accuracy_score(y_test, predictions)\n",
    "print classification_report(y_test, predictions)\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Naive Bayes accuracy and time elapsed caculation\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "NaiveBayes\n",
      "Acurracy:  0.70253164557\n",
      "time elapsed:  0.2800161838531494\n"
     ]
    }
   ],
   "source": [
    "t4=time()\n",
    "print (\"NaiveBayes\")\n",
    "nb = BernoulliNB()\n",
    "clf_nb=nb.fit(X_train,y_train)\n",
    "print (\"Acurracy: \", clf_nb.score(X_test,y_test))\n",
    "t5=time()\n",
    "print (\"time elapsed: \", t5-t4)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## cross-validation for NB"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cross result========\n",
      "[ 0.67088608  0.83544304  0.67088608  0.72151899  0.65822785]\n",
      "0.711392405063\n",
      "time elapsed:  0.04200243949890137\n"
     ]
    }
   ],
   "source": [
    "tt0=time()\n",
    "print (\"cross result========\")\n",
    "scores = cross_validation.cross_val_score(nb, X,y, cv=5)\n",
    "print (scores)\n",
    "print (scores.mean())\n",
    "tt1=time()\n",
    "print (\"time elapsed: \", tt1-tt0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# KNN accuracy and time elapsed caculation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KNN\n",
      "Acurracy:  0.892405063291\n",
      "time elapsed:  0.008000373840332031\n"
     ]
    }
   ],
   "source": [
    "t6=time()\n",
    "print (\"KNN\")\n",
    "# knn = KNeighborsClassifier(n_neighbors=3)\n",
    "knn = KNeighborsClassifier()\n",
    "clf_knn=knn.fit(X_train, y_train)\n",
    "print (\"Acurracy: \", clf_knn.score(X_test,y_test) )\n",
    "t7=time()\n",
    "print (\"time elapsed: \", t7-t6)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## cross validation for KNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cross result========\n",
      "[ 0.83544304  0.87341772  0.88607595  0.92405063  0.87341772]\n",
      "0.878481012658\n",
      "time elapsed:  0.04500269889831543\n"
     ]
    }
   ],
   "source": [
    "tt0=time()\n",
    "print (\"cross result========\")\n",
    "scores = cross_validation.cross_val_score(knn, X,y, cv=5)\n",
    "print (scores)\n",
    "print (scores.mean())\n",
    "tt1=time()\n",
    "print (\"time elapsed: \", tt1-tt0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# SVM accuracy and time elapsed caculation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVM\n",
      "Acurracy:  0.867088607595\n",
      "time elapsed:  0.012000799179077148\n"
     ]
    }
   ],
   "source": [
    "t7=time()\n",
    "print (\"SVM\")\n",
    "\n",
    "svc = SVC()\n",
    "clf_svc=svc.fit(X_train, y_train)\n",
    "print (\"Acurracy: \", clf_svc.score(X_test,y_test) )\n",
    "t8=time()\n",
    "print (\"time elapsed: \", t8-t7)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## cross validation for SVM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cross result========\n",
      "[ 0.87341772  0.88607595  0.89873418  0.91139241  0.88607595]\n",
      "0.891139240506\n",
      "time elapsed:  0.09800553321838379\n"
     ]
    }
   ],
   "source": [
    "tt0=time()\n",
    "print (\"cross result========\")\n",
    "scores = cross_validation.cross_val_score(svc, X,y, cv=5)\n",
    "print (scores)\n",
    "print (scores.mean())\n",
    "tt1=time()\n",
    "print (\"time elapsed: \", tt1-tt0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 4 candidates, totalling 12 fits\n",
      "Best score: 0.907\n",
      "Best parameters set:\n",
      "\tC: 10\n",
      "\tkernel: 'rbf'\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.92      0.76      0.83        59\n",
      "          1       0.87      0.96      0.91        99\n",
      "\n",
      "avg / total       0.89      0.89      0.88       158\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done  12 out of  12 | elapsed:   18.2s finished\n"
     ]
    }
   ],
   "source": [
    "from sklearn.svm import SVC\n",
    "from sklearn.cross_validation import cross_val_score\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn import grid_search\n",
    "\n",
    "svc = SVC()\n",
    "\n",
    "parameters = {'kernel':('linear', 'rbf'), 'C':[1, 10]}\n",
    "\n",
    "grid = grid_search.GridSearchCV(svc, parameters, n_jobs=-1, verbose=1, scoring='accuracy')\n",
    "\n",
    "\n",
    "grid.fit(X_train, y_train)\n",
    "\n",
    "print 'Best score: %0.3f' % grid.best_score_\n",
    "\n",
    "print 'Best parameters set:'\n",
    "best_parameters = grid.best_estimator_.get_params()\n",
    "\n",
    "for param_name in sorted(parameters.keys()):\n",
    "    print '\\t%s: %r' % (param_name, best_parameters[param_name])\n",
    "    \n",
    "predictions = grid.predict(X_test)\n",
    "print classification_report(y_test, predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done  34 tasks      | elapsed:   19.4s\n",
      "[Parallel(n_jobs=-1)]: Done  90 out of  90 | elapsed:   20.2s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 30 candidates, totalling 90 fits\n",
      "Best score: 0.907\n",
      "Best parameters set:\n",
      "\tclf__C: 3\n",
      "\tclf__gamma: 0.01\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.92      0.83      0.88        59\n",
      "          1       0.90      0.96      0.93        99\n",
      "\n",
      "avg / total       0.91      0.91      0.91       158\n",
      "\n"
     ]
    }
   ],
   "source": [
    "pipeline = Pipeline([\n",
    "    ('clf', SVC(kernel='rbf', gamma=0.01, C=100))\n",
    "])\n",
    "\n",
    "parameters = {\n",
    "    'clf__gamma': (0.01, 0.03, 0.1, 0.3, 1),\n",
    "    'clf__C': (0.1, 0.3, 1, 3, 10, 30),\n",
    "}\n",
    "\n",
    "grid_search = GridSearchCV(pipeline, parameters, n_jobs=-1, verbose=1, scoring='accuracy')\n",
    "\n",
    "grid_search.fit(X_train, y_train)\n",
    "\n",
    "print 'Best score: %0.3f' % grid_search.best_score_\n",
    "\n",
    "print 'Best parameters set:'\n",
    "best_parameters = grid_search.best_estimator_.get_params()\n",
    "\n",
    "for param_name in sorted(parameters.keys()):\n",
    "    print '\\t%s: %r' % (param_name, best_parameters[param_name])\n",
    "    \n",
    "predictions = grid_search.predict(X_test)\n",
    "print classification_report(y_test, predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "___"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# read .csv from provided dataset\n",
    "csv_filename=\"student/student-mat.csv\"\n",
    "\n",
    "# df=pd.read_csv(csv_filename,index_col=0)\n",
    "df=pd.read_csv(csv_filename, sep=\";\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>school</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>address</th>\n",
       "      <th>famsize</th>\n",
       "      <th>Pstatus</th>\n",
       "      <th>Medu</th>\n",
       "      <th>Fedu</th>\n",
       "      <th>Mjob</th>\n",
       "      <th>Fjob</th>\n",
       "      <th>...</th>\n",
       "      <th>famrel</th>\n",
       "      <th>freetime</th>\n",
       "      <th>goout</th>\n",
       "      <th>Dalc</th>\n",
       "      <th>Walc</th>\n",
       "      <th>health</th>\n",
       "      <th>absences</th>\n",
       "      <th>G1</th>\n",
       "      <th>G2</th>\n",
       "      <th>G3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>18</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>A</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>at_home</td>\n",
       "      <td>teacher</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>17</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>T</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>at_home</td>\n",
       "      <td>other</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>15</td>\n",
       "      <td>U</td>\n",
       "      <td>LE3</td>\n",
       "      <td>T</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>at_home</td>\n",
       "      <td>other</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>15</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>T</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>health</td>\n",
       "      <td>services</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>15</td>\n",
       "      <td>14</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>16</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>T</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>other</td>\n",
       "      <td>other</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>6</td>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 33 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "  school sex  age address famsize Pstatus  Medu  Fedu     Mjob      Fjob ...  \\\n",
       "0     GP   F   18       U     GT3       A     4     4  at_home   teacher ...   \n",
       "1     GP   F   17       U     GT3       T     1     1  at_home     other ...   \n",
       "2     GP   F   15       U     LE3       T     1     1  at_home     other ...   \n",
       "3     GP   F   15       U     GT3       T     4     2   health  services ...   \n",
       "4     GP   F   16       U     GT3       T     3     3    other     other ...   \n",
       "\n",
       "  famrel freetime  goout  Dalc  Walc health absences  G1  G2  G3  \n",
       "0      4        3      4     1     1      3        6   5   6   6  \n",
       "1      5        3      3     1     1      3        4   5   5   6  \n",
       "2      4        3      2     2     3      3       10   7   8  10  \n",
       "3      3        2      2     1     1      5        2  15  14  15  \n",
       "4      4        3      2     1     2      5        4   6  10  10  \n",
       "\n",
       "[5 rows x 33 columns]"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>age</th>\n",
       "      <th>Medu</th>\n",
       "      <th>Fedu</th>\n",
       "      <th>traveltime</th>\n",
       "      <th>studytime</th>\n",
       "      <th>failures</th>\n",
       "      <th>famrel</th>\n",
       "      <th>freetime</th>\n",
       "      <th>goout</th>\n",
       "      <th>Dalc</th>\n",
       "      <th>Walc</th>\n",
       "      <th>health</th>\n",
       "      <th>absences</th>\n",
       "      <th>G1</th>\n",
       "      <th>G2</th>\n",
       "      <th>G3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "      <td>395.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>16.696203</td>\n",
       "      <td>2.749367</td>\n",
       "      <td>2.521519</td>\n",
       "      <td>1.448101</td>\n",
       "      <td>2.035443</td>\n",
       "      <td>0.334177</td>\n",
       "      <td>3.944304</td>\n",
       "      <td>3.235443</td>\n",
       "      <td>3.108861</td>\n",
       "      <td>1.481013</td>\n",
       "      <td>2.291139</td>\n",
       "      <td>3.554430</td>\n",
       "      <td>5.708861</td>\n",
       "      <td>10.908861</td>\n",
       "      <td>10.713924</td>\n",
       "      <td>10.415190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.276043</td>\n",
       "      <td>1.094735</td>\n",
       "      <td>1.088201</td>\n",
       "      <td>0.697505</td>\n",
       "      <td>0.839240</td>\n",
       "      <td>0.743651</td>\n",
       "      <td>0.896659</td>\n",
       "      <td>0.998862</td>\n",
       "      <td>1.113278</td>\n",
       "      <td>0.890741</td>\n",
       "      <td>1.287897</td>\n",
       "      <td>1.390303</td>\n",
       "      <td>8.003096</td>\n",
       "      <td>3.319195</td>\n",
       "      <td>3.761505</td>\n",
       "      <td>4.581443</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>15.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>16.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>8.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>11.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>18.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>14.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>22.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>75.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>20.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              age        Medu        Fedu  traveltime   studytime    failures  \\\n",
       "count  395.000000  395.000000  395.000000  395.000000  395.000000  395.000000   \n",
       "mean    16.696203    2.749367    2.521519    1.448101    2.035443    0.334177   \n",
       "std      1.276043    1.094735    1.088201    0.697505    0.839240    0.743651   \n",
       "min     15.000000    0.000000    0.000000    1.000000    1.000000    0.000000   \n",
       "25%     16.000000    2.000000    2.000000    1.000000    1.000000    0.000000   \n",
       "50%     17.000000    3.000000    2.000000    1.000000    2.000000    0.000000   \n",
       "75%     18.000000    4.000000    3.000000    2.000000    2.000000    0.000000   \n",
       "max     22.000000    4.000000    4.000000    4.000000    4.000000    3.000000   \n",
       "\n",
       "           famrel    freetime       goout        Dalc        Walc      health  \\\n",
       "count  395.000000  395.000000  395.000000  395.000000  395.000000  395.000000   \n",
       "mean     3.944304    3.235443    3.108861    1.481013    2.291139    3.554430   \n",
       "std      0.896659    0.998862    1.113278    0.890741    1.287897    1.390303   \n",
       "min      1.000000    1.000000    1.000000    1.000000    1.000000    1.000000   \n",
       "25%      4.000000    3.000000    2.000000    1.000000    1.000000    3.000000   \n",
       "50%      4.000000    3.000000    3.000000    1.000000    2.000000    4.000000   \n",
       "75%      5.000000    4.000000    4.000000    2.000000    3.000000    5.000000   \n",
       "max      5.000000    5.000000    5.000000    5.000000    5.000000    5.000000   \n",
       "\n",
       "         absences          G1          G2          G3  \n",
       "count  395.000000  395.000000  395.000000  395.000000  \n",
       "mean     5.708861   10.908861   10.713924   10.415190  \n",
       "std      8.003096    3.319195    3.761505    4.581443  \n",
       "min      0.000000    3.000000    0.000000    0.000000  \n",
       "25%      0.000000    8.000000    9.000000    8.000000  \n",
       "50%      4.000000   11.000000   11.000000   11.000000  \n",
       "75%      8.000000   13.000000   13.000000   14.000000  \n",
       "max     75.000000   19.000000   19.000000   20.000000  "
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# CASE 2: Multi Class Classification : \n",
    "<table>\n",
    "<tr>\n",
    "<th>Class</th><th>G3</th><th>Label</th>   \n",
    "</tr>\n",
    "<tr>\n",
    "<td>I   (excellent/very good)</td><td>16-20</td><td>A</td>\n",
    "</tr>\n",
    "<tr>\n",
    "<td>II  (good)</td><td>14-15</td><td>B</td>\n",
    "</tr>\n",
    "<tr>\n",
    "<td>III (satisfactory)</td><td>12-13</td><td>C</td> \n",
    "</tr>\n",
    "<tr>\n",
    "<td>IV  (sufficient)</td><td>10-11</td><td>D</td>\n",
    "</tr>\n",
    "<tr>\n",
    "<td>V   (fail)</td><td>0-9</td><td>E</td>         \n",
    "</tr>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    395.000000\n",
       "mean      10.415190\n",
       "std        4.581443\n",
       "min        0.000000\n",
       "25%        8.000000\n",
       "50%       11.000000\n",
       "75%       14.000000\n",
       "max       20.000000\n",
       "Name: G3, dtype: float64"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.G3.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "for i in range(len(df.G3)):\n",
    "    if df.G3.loc[i] < 10:\n",
    "        df.G3.loc[i] = 5\n",
    "    elif df.G3.loc[i] < 12:\n",
    "        df.G3.loc[i] = 4\n",
    "    elif df.G3.loc[i] < 14:\n",
    "        df.G3.loc[i] = 3\n",
    "    elif df.G3.loc[i] < 16:\n",
    "        df.G3.loc[i] = 2\n",
    "    elif df.G3.loc[i] < 21:\n",
    "        df.G3.loc[i] = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([5, 4, 2, 1, 3], dtype=int64)"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.G3.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>school</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>address</th>\n",
       "      <th>famsize</th>\n",
       "      <th>Pstatus</th>\n",
       "      <th>Medu</th>\n",
       "      <th>Fedu</th>\n",
       "      <th>Mjob</th>\n",
       "      <th>Fjob</th>\n",
       "      <th>...</th>\n",
       "      <th>famrel</th>\n",
       "      <th>freetime</th>\n",
       "      <th>goout</th>\n",
       "      <th>Dalc</th>\n",
       "      <th>Walc</th>\n",
       "      <th>health</th>\n",
       "      <th>absences</th>\n",
       "      <th>G1</th>\n",
       "      <th>G2</th>\n",
       "      <th>G3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>18</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>A</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>at_home</td>\n",
       "      <td>teacher</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>17</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>T</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>at_home</td>\n",
       "      <td>other</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>15</td>\n",
       "      <td>U</td>\n",
       "      <td>LE3</td>\n",
       "      <td>T</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>at_home</td>\n",
       "      <td>other</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>15</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>T</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>health</td>\n",
       "      <td>services</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>15</td>\n",
       "      <td>14</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>16</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>T</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>other</td>\n",
       "      <td>other</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>6</td>\n",
       "      <td>10</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 33 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "  school sex  age address famsize Pstatus  Medu  Fedu     Mjob      Fjob ...  \\\n",
       "0     GP   F   18       U     GT3       A     4     4  at_home   teacher ...   \n",
       "1     GP   F   17       U     GT3       T     1     1  at_home     other ...   \n",
       "2     GP   F   15       U     LE3       T     1     1  at_home     other ...   \n",
       "3     GP   F   15       U     GT3       T     4     2   health  services ...   \n",
       "4     GP   F   16       U     GT3       T     3     3    other     other ...   \n",
       "\n",
       "  famrel freetime  goout  Dalc  Walc health absences  G1  G2 G3  \n",
       "0      4        3      4     1     1      3        6   5   6  5  \n",
       "1      5        3      3     1     1      3        4   5   5  5  \n",
       "2      4        3      2     2     3      3       10   7   8  4  \n",
       "3      3        2      2     1     1      5        2  15  14  2  \n",
       "4      4        3      2     1     2      5        4   6  10  4  \n",
       "\n",
       "[5 rows x 33 columns]"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    395.000000\n",
       "mean       3.564557\n",
       "std        1.349096\n",
       "min        1.000000\n",
       "25%        2.000000\n",
       "50%        4.000000\n",
       "75%        5.000000\n",
       "max        5.000000\n",
       "Name: G3, dtype: float64"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.G3.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "cols = list(df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['school',\n",
       " 'sex',\n",
       " 'address',\n",
       " 'famsize',\n",
       " 'Pstatus',\n",
       " 'Mjob',\n",
       " 'Fjob',\n",
       " 'reason',\n",
       " 'guardian',\n",
       " 'schoolsup',\n",
       " 'famsup',\n",
       " 'paid',\n",
       " 'activities',\n",
       " 'nursery',\n",
       " 'higher',\n",
       " 'internet',\n",
       " 'romantic']"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "categorical_features = []\n",
    "for f in cols:\n",
    "    if df[f].dtype != 'int64':\n",
    "        categorical_features.append(f)\n",
    "categorical_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "for f in categorical_features:\n",
    "\n",
    "    #Get binarized columns\n",
    "    df[f] = pd.get_dummies(df[f])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>school</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>address</th>\n",
       "      <th>famsize</th>\n",
       "      <th>Pstatus</th>\n",
       "      <th>Medu</th>\n",
       "      <th>Fedu</th>\n",
       "      <th>Mjob</th>\n",
       "      <th>Fjob</th>\n",
       "      <th>...</th>\n",
       "      <th>famrel</th>\n",
       "      <th>freetime</th>\n",
       "      <th>goout</th>\n",
       "      <th>Dalc</th>\n",
       "      <th>Walc</th>\n",
       "      <th>health</th>\n",
       "      <th>absences</th>\n",
       "      <th>G1</th>\n",
       "      <th>G2</th>\n",
       "      <th>G3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>18</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>17</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>15</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>15</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>15</td>\n",
       "      <td>14</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>16</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>6</td>\n",
       "      <td>10</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 33 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   school  sex  age  address  famsize  Pstatus  Medu  Fedu  Mjob  Fjob ...  \\\n",
       "0       1    1   18        0        1        1     4     4     1     0 ...   \n",
       "1       1    1   17        0        1        0     1     1     1     0 ...   \n",
       "2       1    1   15        0        0        0     1     1     1     0 ...   \n",
       "3       1    1   15        0        1        0     4     2     0     0 ...   \n",
       "4       1    1   16        0        1        0     3     3     0     0 ...   \n",
       "\n",
       "   famrel  freetime  goout  Dalc  Walc  health  absences  G1  G2  G3  \n",
       "0       4         3      4     1     1       3         6   5   6   5  \n",
       "1       5         3      3     1     1       3         4   5   5   5  \n",
       "2       4         3      2     2     3       3        10   7   8   4  \n",
       "3       3         2      2     1     1       5         2  15  14   2  \n",
       "4       4         3      2     1     2       5         4   6  10   4  \n",
       "\n",
       "[5 rows x 33 columns]"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "features=list(df.columns[:-1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "X = df[features]\n",
    "y = df['G3']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# split dataset to 60% training and 40% testing\n",
    "X_train, X_test, y_train, y_test = cross_validation.train_test_split(X,y, test_size=0.4, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(237, 32) (237,)\n"
     ]
    }
   ],
   "source": [
    "print (X_train.shape, y_train.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature importances with forests of trees\n",
    "\n",
    "This examples shows the use of forests of trees to evaluate the importance of features on an artificial classification task. The red bars are the feature importances of the forest, along with their inter-trees variability."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Feature ranking:\n",
      "1. feature 31 - G2 (0.195782) \n",
      "2. feature 30 - G1 (0.145836) \n",
      "3. feature 29 - absences (0.035497) \n",
      "4. feature 25 - goout (0.032551) \n",
      "5. feature 2 - age (0.032141) \n",
      "6. feature 6 - Medu (0.031506) \n",
      "7. feature 28 - health (0.030933) \n",
      "8. feature 24 - freetime (0.029983) \n",
      "9. feature 27 - Walc (0.029495) \n",
      "10. feature 13 - studytime (0.029247) \n",
      "11. feature 23 - famrel (0.029207) \n",
      "12. feature 14 - failures (0.028804) \n",
      "13. feature 7 - Fedu (0.028431) \n",
      "14. feature 12 - traveltime (0.022976) \n",
      "15. feature 26 - Dalc (0.022761) \n",
      "16. feature 18 - activities (0.021505) \n",
      "17. feature 1 - sex (0.021448) \n",
      "18. feature 16 - famsup (0.021390) \n",
      "19. feature 17 - paid (0.020667) \n",
      "20. feature 10 - reason (0.020605) \n",
      "21. feature 4 - famsize (0.020506) \n",
      "22. feature 22 - romantic (0.019582) \n",
      "23. feature 19 - nursery (0.018253) \n",
      "24. feature 11 - guardian (0.017402) \n",
      "25. feature 3 - address (0.016361) \n",
      "26. feature 15 - schoolsup (0.014624) \n",
      "27. feature 21 - internet (0.014290) \n",
      "28. feature 5 - Pstatus (0.012596) \n",
      "29. feature 8 - Mjob (0.012404) \n",
      "30. feature 0 - school (0.010073) \n",
      "31. feature 9 - Fjob (0.007019) \n",
      "32. feature 20 - higher (0.006123) \n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAzUAAAJZCAYAAACQrRt5AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABA0pQaAAAgaUoNAACQNKUGAABI2v8Hg+haVm24u/kAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xa82eeb8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.ensemble import ExtraTreesClassifier\n",
    "\n",
    "# Build a classification task using 3 informative features\n",
    "\n",
    "\n",
    "# Build a forest and compute the feature importances\n",
    "forest = ExtraTreesClassifier(n_estimators=250,\n",
    "                              random_state=0)\n",
    "\n",
    "forest.fit(X, y)\n",
    "importances = forest.feature_importances_\n",
    "std = np.std([tree.feature_importances_ for tree in forest.estimators_],\n",
    "             axis=0)\n",
    "indices = np.argsort(importances)[::-1]\n",
    "\n",
    "# Print the feature ranking\n",
    "print(\"Feature ranking:\")\n",
    "\n",
    "for f in range(X.shape[1]):\n",
    "    print(\"%d. feature %d - %s (%f) \" % (f + 1, indices[f], features[indices[f]], importances[indices[f]]))\n",
    "\n",
    "# Plot the feature importances of the forest\n",
    "plt.figure(num=None, figsize=(14, 10), dpi=80, facecolor='w', edgecolor='k')\n",
    "plt.title(\"Feature importances\")\n",
    "plt.bar(range(X.shape[1]), importances[indices],\n",
    "       color=\"r\", yerr=std[indices], align=\"center\")\n",
    "plt.xticks(range(X.shape[1]), indices)\n",
    "plt.xlim([-1, X.shape[1]])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.19578233,  0.14583578,  0.03549731,  0.03255128,  0.03214145])"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "importances[indices[:5]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1. feature 31 - G2 (0.195782)\n",
      "2. feature 30 - G1 (0.145836)\n",
      "3. feature 29 - absences (0.035497)\n",
      "4. feature 25 - goout (0.032551)\n",
      "5. feature 2 - age (0.032141)\n"
     ]
    }
   ],
   "source": [
    "for f in range(5):\n",
    "    print(\"%d. feature %d - %s (%f)\" % (f + 1, indices[f], features[indices[f]] ,importances[indices[f]]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "best_features = []\n",
    "for i in indices[:5]:\n",
    "    best_features.append(features[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAecAAAF6CAYAAAAj9ZDJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xa5d4748>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the top 5 feature importances of the forest\n",
    "plt.figure(num=None, figsize=(8, 6), dpi=80, facecolor='w', edgecolor='k')\n",
    "plt.title(\"Feature importances\")\n",
    "plt.bar(range(5), importances[indices][:5], \n",
    "       color=\"r\",  yerr=std[indices][:5], align=\"center\")\n",
    "plt.xticks(range(5), best_features)\n",
    "plt.xlim([-1, 5])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Decision Tree accuracy and time elapsed caculation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "DecisionTree\n",
      "Acurracy:  0.708860759494\n",
      "time elapsed:  0.0060002803802490234\n"
     ]
    }
   ],
   "source": [
    "t0=time()\n",
    "print (\"DecisionTree\")\n",
    "\n",
    "dt = DecisionTreeClassifier(min_samples_split=20,random_state=99)\n",
    "# dt = DecisionTreeClassifier(min_samples_split=20,max_depth=5,random_state=99)\n",
    "\n",
    "clf_dt=dt.fit(X_train,y_train)\n",
    "\n",
    "print (\"Acurracy: \", clf_dt.score(X_test,y_test))\n",
    "t1=time()\n",
    "print (\"time elapsed: \", t1-t0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## cross validation for DT"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cross result========\n",
      "[ 0.725       0.7625      0.64556962  0.56410256  0.66666667]\n",
      "0.672767770204\n",
      "time elapsed:  0.049002885818481445\n"
     ]
    }
   ],
   "source": [
    "tt0=time()\n",
    "print (\"cross result========\")\n",
    "scores = cross_validation.cross_val_score(dt, X,y, cv=5)\n",
    "print (scores)\n",
    "print (scores.mean())\n",
    "tt1=time()\n",
    "print (\"time elapsed: \", tt1-tt0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Tuning our hyperparameters using GridSearch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done  34 tasks      | elapsed:   22.4s\n",
      "[Parallel(n_jobs=-1)]: Done  81 out of  81 | elapsed:   23.2s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 27 candidates, totalling 81 fits\n",
      "Best score: 0.718\n",
      "Best parameters set:\n",
      "\tclf__max_depth: 50\n",
      "\tclf__min_samples_leaf: 3\n",
      "\tclf__min_samples_split: 10\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          1       0.75      0.94      0.83        16\n",
      "          2       0.68      0.71      0.69        24\n",
      "          3       0.65      0.46      0.54        24\n",
      "          4       0.56      0.71      0.63        35\n",
      "          5       0.90      0.78      0.84        59\n",
      "\n",
      "avg / total       0.74      0.72      0.72       158\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "\n",
    "pipeline = Pipeline([\n",
    "    ('clf', DecisionTreeClassifier(criterion='entropy'))\n",
    "])\n",
    "\n",
    "parameters = {\n",
    "    'clf__max_depth': (5, 25 , 50),\n",
    "    'clf__min_samples_split': (1, 5, 10),\n",
    "    'clf__min_samples_leaf': (1, 2, 3)\n",
    "}\n",
    "\n",
    "grid_search = GridSearchCV(pipeline, parameters, n_jobs=-1, verbose=1, scoring='f1')\n",
    "grid_search.fit(X_train, y_train)\n",
    "\n",
    "print 'Best score: %0.3f' % grid_search.best_score_\n",
    "print 'Best parameters set:'\n",
    "\n",
    "best_parameters = grid_search.best_estimator_.get_params()\n",
    "for param_name in sorted(parameters.keys()):\n",
    "    print '\\t%s: %r' % (param_name, best_parameters[param_name])\n",
    "\n",
    "predictions = grid_search.predict(X_test)\n",
    "\n",
    "print classification_report(y_test, predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Random Forest accuracy and time elapsed caculation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RandomForest\n",
      "Acurracy:  0.670886075949\n",
      "time elapsed:  0.5430312156677246\n"
     ]
    }
   ],
   "source": [
    "t2=time()\n",
    "print (\"RandomForest\")\n",
    "rf = RandomForestClassifier(n_estimators=100,n_jobs=-1)\n",
    "clf_rf = rf.fit(X_train,y_train)\n",
    "print (\"Acurracy: \", clf_rf.score(X_test,y_test))\n",
    "t3=time()\n",
    "print (\"time elapsed: \", t3-t2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## cross validation for RF"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cross result========\n",
      "[ 0.65        0.6875      0.67088608  0.71794872  0.74358974]\n",
      "0.693984907498\n",
      "time elapsed:  2.495142698287964\n"
     ]
    }
   ],
   "source": [
    "tt0=time()\n",
    "print (\"cross result========\")\n",
    "scores = cross_validation.cross_val_score(rf, X,y, cv=5)\n",
    "print (scores)\n",
    "print (scores.mean())\n",
    "tt1=time()\n",
    "print (\"time elapsed: \", tt1-tt0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Tuning Models using GridSearch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done  34 tasks      | elapsed:   23.1s\n",
      "[Parallel(n_jobs=-1)]: Done 184 tasks      | elapsed:   37.6s\n",
      "[Parallel(n_jobs=-1)]: Done 324 out of 324 | elapsed:  1.1min finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 108 candidates, totalling 324 fits\n",
      "Best score: 0.713\n",
      "Best parameters set:\n",
      "\tclf__max_depth: 50\n",
      "\tclf__min_samples_leaf: 3\n",
      "\tclf__min_samples_split: 5\n",
      "\tclf__n_estimators: 100\n",
      "Accuracy: 0.753164556962\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          1       0.80      0.50      0.62        16\n",
      "          2       0.61      0.83      0.70        24\n",
      "          3       0.65      0.46      0.54        24\n",
      "          4       0.67      0.74      0.70        35\n",
      "          5       0.92      0.92      0.92        59\n",
      "\n",
      "avg / total       0.76      0.75      0.75       158\n",
      "\n"
     ]
    }
   ],
   "source": [
    "\n",
    "pipeline2 = Pipeline([\n",
    "('clf', RandomForestClassifier(criterion='entropy'))\n",
    "])\n",
    "\n",
    "parameters = {\n",
    "    'clf__n_estimators': (5, 25, 50, 100),\n",
    "    'clf__max_depth': (5, 25 , 50),\n",
    "    'clf__min_samples_split': (1, 5, 10),\n",
    "    'clf__min_samples_leaf': (1, 2, 3)\n",
    "}\n",
    "\n",
    "grid_search = GridSearchCV(pipeline2, parameters, n_jobs=-1, verbose=1, scoring='accuracy', cv=3)\n",
    "\n",
    "grid_search.fit(X_train, y_train)\n",
    "\n",
    "print 'Best score: %0.3f' % grid_search.best_score_\n",
    "\n",
    "print 'Best parameters set:'\n",
    "best_parameters = grid_search.best_estimator_.get_params()\n",
    "\n",
    "for param_name in sorted(parameters.keys()):\n",
    "    print '\\t%s: %r' % (param_name, best_parameters[param_name])\n",
    "\n",
    "predictions = grid_search.predict(X_test)\n",
    "print 'Accuracy:', accuracy_score(y_test, predictions)\n",
    "print classification_report(y_test, predictions)\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Naive Bayes accuracy and time elapsed caculation\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "NaiveBayes\n",
      "Acurracy:  0.379746835443\n",
      "time elapsed:  0.0370020866394043\n"
     ]
    }
   ],
   "source": [
    "t4=time()\n",
    "print (\"NaiveBayes\")\n",
    "nb = BernoulliNB()\n",
    "clf_nb=nb.fit(X_train,y_train)\n",
    "print (\"Acurracy: \", clf_nb.score(X_test,y_test))\n",
    "t5=time()\n",
    "print (\"time elapsed: \", t5-t4)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## cross-validation for NB"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cross result========\n",
      "[ 0.3125      0.375       0.27848101  0.41025641  0.35897436]\n",
      "0.347042356378\n",
      "time elapsed:  0.0740041732788086\n"
     ]
    }
   ],
   "source": [
    "tt0=time()\n",
    "print (\"cross result========\")\n",
    "scores = cross_validation.cross_val_score(nb, X,y, cv=5)\n",
    "print (scores)\n",
    "print (scores.mean())\n",
    "tt1=time()\n",
    "print (\"time elapsed: \", tt1-tt0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# KNN accuracy and time elapsed caculation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KNN\n",
      "Acurracy:  0.626582278481\n",
      "time elapsed:  0.01500082015991211\n"
     ]
    }
   ],
   "source": [
    "t6=time()\n",
    "print (\"KNN\")\n",
    "# knn = KNeighborsClassifier(n_neighbors=3)\n",
    "knn = KNeighborsClassifier()\n",
    "clf_knn=knn.fit(X_train, y_train)\n",
    "print (\"Acurracy: \", clf_knn.score(X_test,y_test) )\n",
    "t7=time()\n",
    "print (\"time elapsed: \", t7-t6)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## cross validation for KNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cross result========\n",
      "[ 0.6125      0.6625      0.64556962  0.64102564  0.66666667]\n",
      "0.645652385589\n",
      "time elapsed:  0.05500316619873047\n"
     ]
    }
   ],
   "source": [
    "tt0=time()\n",
    "print (\"cross result========\")\n",
    "scores = cross_validation.cross_val_score(knn, X,y, cv=5)\n",
    "print (scores)\n",
    "print (scores.mean())\n",
    "tt1=time()\n",
    "print (\"time elapsed: \", tt1-tt0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# SVM accuracy and time elapsed caculation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVM\n",
      "Acurracy:  0.721518987342\n",
      "time elapsed:  0.017000913619995117\n"
     ]
    }
   ],
   "source": [
    "t7=time()\n",
    "print (\"SVM\")\n",
    "\n",
    "svc = SVC()\n",
    "clf_svc=svc.fit(X_train, y_train)\n",
    "print (\"Acurracy: \", clf_svc.score(X_test,y_test) )\n",
    "t8=time()\n",
    "print (\"time elapsed: \", t8-t7)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## cross validation for SVM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cross result========\n",
      "[ 0.6375      0.7         0.69620253  0.73076923  0.78205128]\n",
      "0.709304608893\n",
      "time elapsed:  0.128007173538208\n"
     ]
    }
   ],
   "source": [
    "tt0=time()\n",
    "print (\"cross result========\")\n",
    "scores = cross_validation.cross_val_score(svc, X,y, cv=5)\n",
    "print (scores)\n",
    "print (scores.mean())\n",
    "tt1=time()\n",
    "print (\"time elapsed: \", tt1-tt0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 4 candidates, totalling 12 fits\n",
      "Best score: 0.679\n",
      "Best parameters set:\n",
      "\tC: 1\n",
      "\tkernel: 'linear'\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          1       0.67      0.62      0.65        16\n",
      "          2       0.59      0.54      0.57        24\n",
      "          3       0.52      0.62      0.57        24\n",
      "          4       0.67      0.63      0.65        35\n",
      "          5       0.92      0.92      0.92        59\n",
      "\n",
      "avg / total       0.73      0.72      0.72       158\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done  12 out of  12 | elapsed:   17.0s finished\n"
     ]
    }
   ],
   "source": [
    "from sklearn.svm import SVC\n",
    "from sklearn.cross_validation import cross_val_score\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn import grid_search\n",
    "\n",
    "svc = SVC()\n",
    "\n",
    "parameters = {'kernel':('linear', 'rbf'), 'C':[1, 10]}\n",
    "\n",
    "grid = grid_search.GridSearchCV(svc, parameters, n_jobs=-1, verbose=1, scoring='accuracy')\n",
    "\n",
    "\n",
    "grid.fit(X_train, y_train)\n",
    "\n",
    "print 'Best score: %0.3f' % grid.best_score_\n",
    "\n",
    "print 'Best parameters set:'\n",
    "best_parameters = grid.best_estimator_.get_params()\n",
    "\n",
    "for param_name in sorted(parameters.keys()):\n",
    "    print '\\t%s: %r' % (param_name, best_parameters[param_name])\n",
    "    \n",
    "predictions = grid.predict(X_test)\n",
    "print classification_report(y_test, predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done  34 tasks      | elapsed:   19.1s\n",
      "[Parallel(n_jobs=-1)]: Done  90 out of  90 | elapsed:   19.8s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 30 candidates, totalling 90 fits\n",
      "Best score: 0.700\n",
      "Best parameters set:\n",
      "\tclf__C: 3\n",
      "\tclf__gamma: 0.01\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          1       0.79      0.69      0.73        16\n",
      "          2       0.73      0.79      0.76        24\n",
      "          3       0.62      0.67      0.64        24\n",
      "          4       0.61      0.66      0.63        35\n",
      "          5       0.93      0.85      0.88        59\n",
      "\n",
      "avg / total       0.76      0.75      0.76       158\n",
      "\n"
     ]
    }
   ],
   "source": [
    "pipeline = Pipeline([\n",
    "    ('clf', SVC(kernel='rbf', gamma=0.01, C=100))\n",
    "])\n",
    "\n",
    "parameters = {\n",
    "    'clf__gamma': (0.01, 0.03, 0.1, 0.3, 1),\n",
    "    'clf__C': (0.1, 0.3, 1, 3, 10, 30),\n",
    "}\n",
    "\n",
    "grid_search = GridSearchCV(pipeline, parameters, n_jobs=-1, verbose=1, scoring='accuracy')\n",
    "\n",
    "grid_search.fit(X_train, y_train)\n",
    "\n",
    "print 'Best score: %0.3f' % grid_search.best_score_\n",
    "\n",
    "print 'Best parameters set:'\n",
    "best_parameters = grid_search.best_estimator_.get_params()\n",
    "\n",
    "for param_name in sorted(parameters.keys()):\n",
    "    print '\\t%s: %r' % (param_name, best_parameters[param_name])\n",
    "    \n",
    "predictions = grid_search.predict(X_test)\n",
    "print classification_report(y_test, predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "___\n",
    "# Case 3 : Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.linear_model import SGDRegressor\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "from sklearn.cross_validation import train_test_split\n",
    "from sklearn. cross_validation import cross_val_score\n",
    "\n",
    "from sklearn.feature_selection import *\n",
    "from sklearn import metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# read .csv from provided dataset\n",
    "csv_filename=\"student/student-mat.csv\"\n",
    "\n",
    "# df=pd.read_csv(csv_filename,index_col=0)\n",
    "df=pd.read_csv(csv_filename,sep=\";\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>school</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>address</th>\n",
       "      <th>famsize</th>\n",
       "      <th>Pstatus</th>\n",
       "      <th>Medu</th>\n",
       "      <th>Fedu</th>\n",
       "      <th>Mjob</th>\n",
       "      <th>Fjob</th>\n",
       "      <th>...</th>\n",
       "      <th>famrel</th>\n",
       "      <th>freetime</th>\n",
       "      <th>goout</th>\n",
       "      <th>Dalc</th>\n",
       "      <th>Walc</th>\n",
       "      <th>health</th>\n",
       "      <th>absences</th>\n",
       "      <th>G1</th>\n",
       "      <th>G2</th>\n",
       "      <th>G3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>18</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>A</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>at_home</td>\n",
       "      <td>teacher</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>17</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>T</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>at_home</td>\n",
       "      <td>other</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>15</td>\n",
       "      <td>U</td>\n",
       "      <td>LE3</td>\n",
       "      <td>T</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>at_home</td>\n",
       "      <td>other</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>15</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>T</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>health</td>\n",
       "      <td>services</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>15</td>\n",
       "      <td>14</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>16</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>T</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>other</td>\n",
       "      <td>other</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>6</td>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 33 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "  school sex  age address famsize Pstatus  Medu  Fedu     Mjob      Fjob ...  \\\n",
       "0     GP   F   18       U     GT3       A     4     4  at_home   teacher ...   \n",
       "1     GP   F   17       U     GT3       T     1     1  at_home     other ...   \n",
       "2     GP   F   15       U     LE3       T     1     1  at_home     other ...   \n",
       "3     GP   F   15       U     GT3       T     4     2   health  services ...   \n",
       "4     GP   F   16       U     GT3       T     3     3    other     other ...   \n",
       "\n",
       "  famrel freetime  goout  Dalc  Walc health absences  G1  G2  G3  \n",
       "0      4        3      4     1     1      3        6   5   6   6  \n",
       "1      5        3      3     1     1      3        4   5   5   6  \n",
       "2      4        3      2     2     3      3       10   7   8  10  \n",
       "3      3        2      2     1     1      5        2  15  14  15  \n",
       "4      4        3      2     1     2      5        4   6  10  10  \n",
       "\n",
       "[5 rows x 33 columns]"
      ]
     },
     "execution_count": 137,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "cols = list(df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['school',\n",
       " 'sex',\n",
       " 'address',\n",
       " 'famsize',\n",
       " 'Pstatus',\n",
       " 'Mjob',\n",
       " 'Fjob',\n",
       " 'reason',\n",
       " 'guardian',\n",
       " 'schoolsup',\n",
       " 'famsup',\n",
       " 'paid',\n",
       " 'activities',\n",
       " 'nursery',\n",
       " 'higher',\n",
       " 'internet',\n",
       " 'romantic']"
      ]
     },
     "execution_count": 143,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "categorical_features = []\n",
    "for f in cols:\n",
    "    if df[f].dtype != 'int64':\n",
    "        categorical_features.append(f)\n",
    "categorical_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "for f in categorical_features:\n",
    "\n",
    "    #Get binarized columns\n",
    "    df[f] = pd.get_dummies(df[f])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>school</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>address</th>\n",
       "      <th>famsize</th>\n",
       "      <th>Pstatus</th>\n",
       "      <th>Medu</th>\n",
       "      <th>Fedu</th>\n",
       "      <th>Mjob</th>\n",
       "      <th>Fjob</th>\n",
       "      <th>...</th>\n",
       "      <th>famrel</th>\n",
       "      <th>freetime</th>\n",
       "      <th>goout</th>\n",
       "      <th>Dalc</th>\n",
       "      <th>Walc</th>\n",
       "      <th>health</th>\n",
       "      <th>absences</th>\n",
       "      <th>G1</th>\n",
       "      <th>G2</th>\n",
       "      <th>G3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>18</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>17</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>15</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>15</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>15</td>\n",
       "      <td>14</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>6</td>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 33 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   school  sex  age  address  famsize  Pstatus  Medu  Fedu  Mjob  Fjob ...  \\\n",
       "0     1.0  1.0   18      0.0      1.0      1.0     4     4   1.0   0.0 ...   \n",
       "1     1.0  1.0   17      0.0      1.0      0.0     1     1   1.0   0.0 ...   \n",
       "2     1.0  1.0   15      0.0      0.0      0.0     1     1   1.0   0.0 ...   \n",
       "3     1.0  1.0   15      0.0      1.0      0.0     4     2   0.0   0.0 ...   \n",
       "4     1.0  1.0   16      0.0      1.0      0.0     3     3   0.0   0.0 ...   \n",
       "\n",
       "   famrel  freetime  goout  Dalc  Walc  health  absences  G1  G2  G3  \n",
       "0       4         3      4     1     1       3         6   5   6   6  \n",
       "1       5         3      3     1     1       3         4   5   5   6  \n",
       "2       4         3      2     2     3       3        10   7   8  10  \n",
       "3       3         2      2     1     1       5         2  15  14  15  \n",
       "4       4         3      2     1     2       5         4   6  10  10  \n",
       "\n",
       "[5 rows x 33 columns]"
      ]
     },
     "execution_count": 155,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "features=list(df.columns[:-1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "X = df[features]\n",
    "y = df['G3']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[(False, 'school'), (False, 'sex'), (False, 'age'), (False, 'address'), (False, 'famsize'), (False, 'Pstatus'), (True, 'Medu'), (True, 'Fedu'), (False, 'Mjob'), (False, 'Fjob'), (False, 'reason'), (False, 'guardian'), (False, 'traveltime'), (False, 'studytime'), (True, 'failures'), (False, 'schoolsup'), (False, 'famsup'), (False, 'paid'), (False, 'activities'), (False, 'nursery'), (False, 'higher'), (False, 'internet'), (False, 'romantic'), (False, 'famrel'), (False, 'freetime'), (False, 'goout'), (False, 'Dalc'), (False, 'Walc'), (False, 'health'), (False, 'absences'), (True, 'G1'), (True, 'G2')]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAs4AAAIiCAYAAADYak7VAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzs3X18VOWd9/HvTKwNwWobo6KQCguLD03xJQlU2iAxkCpQ\nkFRJxF2yCmprwLLWlRBAqFgX0T5YFFJ8YNlQEYgrqFW2RGkwrHdFxruB3Ig8VGjAgkA2rZZEkDn3\nH4mQkadrwnXNzAmf9+vVVznhzHd+Xuc31/wYDknA8zxPAAAAAE4qGO8CAAAAAD9gcAYAAAAMMDgD\nAAAABhicAQAAAAMMzgAAAIABBmcAAADAAIMzAAAAYIDB+RRyc3N11VVXqXfv3rr66qvVu3dv7d27\n97Qy165dqwEDBliq0Nxjjz2mb33rW7rmmmv0s5/9LObP3561lz55++23VVRUpKysLA0cODCmz32m\naC+98uyzz2rYsGHq3bu3Bg0apGeffTamz9/etZc+WbBggQYNGqTMzExde+21euSRRxQOh2NaQ3vX\nXnrlc4cOHdLgwYOVk5MTl+c/lbPiXYAfzJs3T9dcc421PM/zFAgE2vz4w4cPKykpKarHLF68WKtW\nrdIrr7wiSbr99tuVnp6uwsLCNteBSO2hTzp06KCbb75Zn376qX7961+3+blxcu2hVyTp0Ucf1WWX\nXaYdO3Zo7NixuvjiizVkyJA214FI7aFPBg4cqPz8fJ133nn629/+pnvuuUfl5eW67bbb2lwHjtUe\neuVzzzzzjNLS0lRXV9fm53eJT5wNnOiHK/7xj3/ULbfcoj59+mjEiBFau3btkd978cUXNWTIEPXu\n3Vt5eXlasmSJJKmxsVF33XWXPvroo4g/GZaWlupXv/rVkcd/8U97ubm5evrppzV8+HBdffXVCofD\n+uijj/SjH/1I/fr106BBg7Rw4cIT/jcsX75cY8aM0YUXXqgLL7xQY8aM0bJly053adBKe+iTXr16\nafjw4erSpcvpLgdOoj30ytixY3XFFVcoGAyqW7duys3N1bvvvnu6S4NW2kOfpKen67zzzpPUPEwF\ng0H9+c9/Pq11wbHaQ69IUl1dnX7729/qrrvuOp3lcMvDSV133XXeW2+9dczXd+/e7fXt29d78803\nPc/zvLfeesvr27evV19f73me51VVVXl1dXWe53neO++841111VXexo0bPc/zvLffftsbMGBARN6k\nSZO8xx9//MjxF8+57rrrvBEjRni7d+/2Pv30Uy8cDnv5+fne3Llzvc8++8yrq6vzBg0a5K1Zs+a4\n/x2ZmZleTU3NkePa2lqvd+/e0S8Ijqu99Mnn3nrrLS83NzfqdcCptbde+dyIESO8xYsXG68DTq49\n9ckrr7zi9e7d27vsssu8fv36eZs2bWrTmuD42lOv/OAHP/Bef/314z5/ouATZwPjxo1T37591bdv\nX40fP16S9PLLLysnJ0f9+/eXJPXr108ZGRlavXq1JGnAgAFHPrXLysrSd77zHa1bt+606igqKtJF\nF12ks88+Wxs2bFBDQ4PuvvtuJSUlqUuXLho5cqReffXV4z72wIED+spXvnLk+JxzztGBAwdOqx5E\nag99gthob70ye/ZseZ6n73//+6dVDyK1lz753ve+p1AopJUrV+qWW25RWlraadWDY7WHXqmsrFQ4\nHE74f1/DPc4G5s6de8y9Qx9++KFWrFih3//+95Ka/5rks88+O3Le6tWrNXfuXG3fvl3hcFhNTU26\n7LLLTquOTp06Hfn1rl27tGfPHvXt2/fI84fDYfXp0+e4j01JSdEnn3xy5Pjjjz9WSkrKadWDSO2h\nTxAb7alXfvOb3+jll1/WokWL9KUvfem06kGk9tQnkvT1r39dPXr00E9+8hM98cQTp1UTIvm9Vxob\nG/Wzn/1MTz/99JFzExWDs4HjXcCLL75YI0aM0IwZM475vYMHD2rChAl67LHHNHDgQAWDQY0bN+5I\nzvFuuO/QoYOampqOHJ/qX8RefPHF6tKli373u98Z/Tf06NFDmzZt0je/+U1J0nvvvad//Md/NHos\nzLSHPkFstJdeeeGFF/TMM8/oueee04UXXmj8OJhpL33S2qFDhxL2H335md97ZceOHdq1a5duvfVW\nSc198vHHHys7O1tLly7VJZdccsqMWOFWjTYaPny4Vq1apTVr1igcDuvTTz/V2rVrtWfPHh06dEiH\nDh3S1772NQWDQa1evVr/8z//c+Sx559/vhoaGiI+Ab7iiiu0evVq/fWvf9XevXtVXl5+0ufv1auX\nOnbsqKefflqffvqpDh8+rC1btmjDhg3HPX/EiBFasGCB9uzZoz179mjBggX8tWoM+K1PPM/TwYMH\ndfDgwSO/PnTokJ3FwEn5rVdefvllPf7445o/f746d+5sZxFwSn7rk4qKCtXX10uStm7dqqefflr9\n+vWzsBI4FT/1Ss+ePbV69Wq99NJLeumll/TTn/5UaWlpevnll3XxxRfbWxQLGJxP4UTfjqVTp06a\nO3eu5s2bp379+um6667T/Pnz5XmeOnbsqClTpmjChAnq27evXnvttYh7dv7hH/5BQ4cO1cCBA9W3\nb1/t3btXN954oy677DLl5ubqjjvuOOZbOn2xjmAwqHnz5mnTpk0aOHCgvv3tb+uBBx6IaPLWbrnl\nFl133XUaPny4hg8frtzcXBUUFJzm6uBz7aVP3nnnHfXq1Us//OEP9Ze//EVXXXWVxo4de5qrg9ba\nS6/86le/UkNDg26++eYj//L+Jz/5yektDo5oL33y7rvvatiwYbr66qv1gx/8QDk5Obr33ntPc3XQ\nWnvolWAwqPPPP//I/8477zwFg0Glpqae1rfFcyHgJfKNJAAAAECC4BNnAAAAwACDMwAAAGCAwRkA\nAAAwwOAMAAAAGPDN93EOhULxLgEWZWZmOsmlT9ofegUmXPWJRK+0N+wpMHXcXonBj/W2Yt26dU7O\nbQuX+WdCdqLU0ZbzyY5tNrWQHe86os1PlDUh2875LrMTqRayzc/nVg0AAADAAIMzAAAAYIDBGQAA\nADDA4AwAAAAYYHAGAAAADDj9dnS7d+/WxIkTtX//fgWDQY0cOVJFRUX661//qnvvvVe7du1Sly5d\n9Pjjj+srX/mKy1KQwOgTmKJXYIpegQn6BNFy+olzUlKSSktL9eqrr2rx4sV67rnntG3bNj311FPq\n16+ffve73+lb3/qW5s2b57IMJDj6BKboFZiiV2CCPkG0nA7OF1xwga644gpJUseOHdW9e3ft2bNH\nb7zxhvLz8yVJ+fn5ev31112WAYf2729QYeHzKirarMLCRaqvb4g6gz5JDDaupWvx6BU/rAuOxb4C\nE/QJohWznxy4c+dObdq0SVdddZX279+vtLQ0Sc1NW19fH6syYFlx8QotXXqLpIA2bvQkLdaSJaPa\nnEefxI/ta+larHrFb+uCY7GvwAR9AhMxGZz//ve/60c/+pEmT56sjh07KhAIRPz+F49PJJofZ+n6\nR1+6zPdTdm2tJH1+/QKqrW37c8SjT9pyfnvNtnktT8ZGZix7xU/rQvax2tv7D9lusnn/IduYjR9j\neDKHDh3yxowZ4y1YsODI12644QZv7969nud53kcffeTdcMMNp8xJlB956jrfb9kFBc95UtiTPE8K\newUFi9pURzz6pC3nt+fstlzLaNm4PrHulURcF7LNzm1v7z9kn352IuwpbT2f7Nhmx+1Hbk+ePFk9\nevTQv/zLvxz5Wm5url588UVJ0rJlyzRw4EDXZcCRsrIhKihYrCuvfF4FBYtVVja4TTn0SfzZupau\nxbpX/LIuOBb7CkzQJ4iG01s1QqGQXnnlFfXs2VMjRoxQIBDQvffeqzvvvFP/+q//qv/6r/9S586d\n9fjjj7ssAw6lpn5VS5aMUigUUmZmZpsy6JPEYONauhaPXvHDuuBY7CswQZ8gWk4H58zMTL333nvH\n/b0FCxa4fGr4CH0CU/QKTNErMEGfIFr85EAAAADAAIMzAAAAYIDBGQAAADDA4AwAAAAYYHAGAAAA\nDDA4AwAAAAYYnAEAAAADDM4AAACAAQZnAAAAwACDMwAAAGCAwRkAAAAwwOAMAAAAGGBwBgAAAAww\nOAMAAAAGGJwBAAAAAwzOAAAAgAEGZwAAAMAAgzNOy5YtO5Se/oiyszcoPX2mtm3bEe+SkID83Cd+\nrh0AYNdZ8S4A/pab+7x27iyRFNDOnZ5ycmaprm5SvMtCgvFzn/i5dgCAXXzijNNSX3+JpEDLUaDl\nGIjk5z7xc+0AALsYnHFaUlN3SfJajjylpn4Yz3KQoPzcJ36uHQBgF7dq4LRUVd2qnJxZ2revk9LS\ndquqalS8S0IC8nOf+Ll2AIBdDM44Ld27X6q6ukkKhULKzLwt3uUgQfm5T/xcOwDALm7VAAAAAAww\nOAMAAAAGGJwBAAAAAwzOAAAAgAEGZwAAAMAAgzMAAABggMEZAAAAMMDgDAAAABhgcAYAAAAMMDgD\nAAAABhicAQAAAAMMzgAAAIABBmcAAADAAIMzAAAAYIDBGQAAADDA4AwAAAAYYHAGAAAADDA4nwG2\nbNmh9PRHlJ29QenpM7Vt2w5r2RUVKxQITFJWVo0CgUlavnyFtWwca//+BhUWPq+ios0qLFyk+voG\na9ku+2TlyjVKSpqsrKwaJSWVatWqNdayXZs/v6JVj5do4cIKa9kurycAwL6z4l0A3MvNfV47d5ZI\nCmjnTk85ObNUVzfJSnZBwWpJMyUFJHnKzy+V5w22ko1jFRev0NKlt0gKaONGT9JiLVkyykq2yz4Z\nPPg1hcMPSwooHPaUlzdFhw9nW8l2bezYkFr3eFFRqUaPHmkl2+X1BADYxyfOZ4D6+kvU/KYvSYGW\nY1t6RmQ3H8OVDz44R63Xu/nYDpd9Eg73iMhuPvYLdz3u8noCAOxjcD4DpKbukuS1HHlKTf3QYvrm\niOzmY7jSrdvHar3e3bp9Yi3bZZ8Eg1sisoPBrday3Xtfrnrc5fUEANjHrRpngKqqW5WTM0v79nVS\nWtpuVVXZ+6vgZcsGKD+/VM2fwm3WsmUDrGXjWGVlQyQtVm2tlJEhlZXZuy3GZZ9UVg5VXt4UhcM9\nFAxuVWXlEGvZrpWXZ6mo6GiPl5dnWst2eT0BAPYxOJ8Bune/VHV1kxQKhZSZeZvV7BEjBsvzBrdk\nj7GajWOlpn5VS5aMallvewOc5LZPcnOzdfhwti/7ZPTokRo9eqST2l1eTwCAfdyqAQAAABhgcAYA\nAAAMMDgDAAAABhicAQAAAAMMzgAAAIABBmcAAADAAIMzAAAAYIDBGQAAADDA4AwAAAAYYHAGAAAA\nDDA4AwAAAAYYnAEAAAADDM4AAACAAQZnAAAAwACDMwAAAGCAwRkAAAAwwOAMAAAAGGBwThBbtuxQ\nevojys7eoPT0mdq2bccZn43YcnktV65co6SkycrKqlFSUqlWrVpjLdu1+fMrFAhMUlZWjQKBEi1c\nWBHvkgAAcXJWvAtAs9zc57VzZ4mkgHbu9JSTM0t1dZPO6GzElstrOXjwawqHH5YUUDjsKS9vig4f\nzraS7drYsSFJMyUFJHkqKirV6NEj41wVACAe+MQ5QdTXX6LmN2ZJCrQcn9nZiC2X1zIc7hGR3Xzs\nFz3VuvbmYwDAmYjBOUGkpu6S5LUceUpN/fCMz0ZsubyWweCWiOxgcKu1bPfeV+vapc1xrAUAEE/c\nqpEgqqpuVU7OLO3b10lpabtVVTXqjM9GbLm8lpWVQ5WXN0XhcA8Fg1tVWTnEWrZr5eVZKioqVfMn\nzZtVXp4Z75IAAHHC4Jwgune/VHV1kxQKhZSZeRvZiDmX1zI3N1uHD2e3ZI+xmu3a6NEjNXr0SF/W\nDgCwi1s1AAAAAAMMzgAAAIABBmcAAADAAIMzAAAAYIDBGQAAADDA4AwAAAAYYHAGAAAADDA4AwAA\nAAYYnAEAAAADDM4AAACAAQZnAAAAwACDMwAAAGCAwRkAAAAwwOAMAAAAGGBwBgAAAAwwOAMAAAAG\nGJwBAAAAAwzOUdi/v0GFhc+rqGizCgsXqb6+wRfZa9euV0rKNPXpU6OUlKkKhdZby549e4ECgUnK\nyqpRIFCiuXMXWMv2K5frvWXLDqWnP6Ls7A1KT5+pbdt2WMteuXKNkpImKyurRklJpVq1ao21bJdr\n4tptt02M6PE777wv3iUBAOLkrHgX4CfFxSu0dOktkgLauNGTtFhLloxK+OycnBfU2PigpIAaGz31\n7z9dBw70spI9YcImSTMlBSR5GjeuVMXFVqJ9y+V65+Y+r507SyQFtHOnp5ycWaqrm2Qle/Dg1xQO\nPywpoHDYU17eFB0+nG0l2+WauPaf/xlU6x5/5plSPf10nIsCAMQFnzhH4YMPzlHzm6ckBVqOEz+7\nqalrRHbzsS09I7Kbj89sLte7vv6SiOzmYzvC4R4R2c3HdrjtQdfocQBAMwbnKHTr9rEkr+XIU7du\nn/giOzn5TxHZycnbrWVL70dkS5stZvuTy/VOTd0VkZ2a+qG17GBwS0R2MLjVWrbbHnSNHgcANONW\njSiUlQ2RtFi1tVJGhlRWNtgX2dXVBerff7qamroqOXm7qqtvtpY9Z84VGjeuVM2fwm3WnDmXW8v2\nK5frXVV1q3JyZmnfvk5KS9utqio7t/NIUmXlUOXlTVE43EPB4FZVVg6xlu1yTVy7447P9MwzR3v8\njjsOxbskAECcMDhHITX1q1qyZJRCoZAyMzN9k52Z2UsHDvRqyR5jNbu4+DYVF8tJtl+5XO/u3S9V\nXd2kluzbrGbn5mbr8OFsJ3W7XBPXnn7653r6aXocAMCtGgAAAIARBmcAAADAAIMzAAAAYIDBGQAA\nADDA4AwAAAAYYHAGAAAADDA4AwAAAAYYnAEAAAADDM4AAACAAQZnAAAAwACDMwAAAGCAwRkAAAAw\nwOAMAAAAGGBwBgAAAAw4HZwnT56sb3/72xo2bNiRrz355JO69tprlZ+fr/z8fL355psuS4BP0Csw\nQZ/AFL0CE/QJonWWy/Dvf//7Gj16tCZOnBjx9dtvv1233367y6eGz9ArMEGfwBS9AhP0CaLl9BPn\nrKwsnXvuucd83fM8l08LH6JXYII+gSl6BSboE0QrLvc4/+Y3v9GNN96oKVOm6OOPP45HCW2yf3+D\nCgufV1HRZhUWLlJ9fYO17JUr1ygpabKysmqUlFSqVavWWMuePXuBAoFJysqqUSBQorlzF1jLnjr1\n5xHZ06f/zFq25M9eWbt2vVJSpqlPnxqlpExVKLTeWvbMmXMj1vvRR+day3ZZt8selNz2yVVXDY2o\nPTPzeqv5rrjcr/zMj3sKYo8+wYnEfHC+9dZb9cYbb+ill15SWlqaZs6cGesS2qy4eIWWLr1FGzeO\n0tKlo3T33SusZQ8e/JrC4YcljVE4/O/Ky3vNWvaECZskzZQ0RtIjGjduk7Xshx/eG5E9Y8Y+a9l+\n7ZWcnBfU2PigPG+MGhsfUv/+L1jLnjz5z2q93iUlf7aW7bJulz3ouk/Wr/+mWtf+7rtXW813xeV+\n5Vd+3VMQW/QJTsbpPc7Hk5qaeuTXBQUF+uEPf2j82FAo5ORcU7W1khRoOQqottbe84TDPSKyw+Ee\nFv8bekZkSz19kd3WXon2+W33SlNTV7Vek6amrr5Yb7/W7X5Pcfn6ibYWcy73q9ZcZLrK9vP7D9mx\ny45Vn7TlfLLjn+18cP7ifUJ79+7VBRdcIEmqrKxUz549jbMyMzONzguFQsbnRiMj431t3Oip+c3I\nU0aGeU2nEgy+oHD4aHYwuFWZmWOsZEtLJR3NljbHNftEjWurV6K5Ji56JTl5mRobj65JcvJ2X1zL\nRKz7eL0S+z3F5eunmYs+dLlffc7VXhtttus9RYr/+w/ZdrITY085WkuirAvZxz//eJwOzvfdd5/e\nfvttNTQ0KCcnR/fcc4/efvttvffeewoGg+rcubNmzJjhsgSrysqGSFqs2lopI0MqKxtsLbuycqjy\n8qYoHO6hYHCrKiuHWMueM+cKjRtXquZPzjZrzpzLrWVPm3aBZsw4mj1tWlqbctpTr1RXF6h//+lq\nauqq5OTtqq6+2Vr2rFmXqqTk6HrPmvV1a9ku67bVg/Hok969/6h33z1ae+/e/9dqvisu9ys/aE97\nCtyhTxA1zyfWrVvn5Ny2cJl/JmQnSh1tOZ/s2GZTC9nxriPa/ERZE7LtnO8yO5FqIdv8fH5yIAAA\nAGCAwRkAAAAwwOAMAAAAGGBwBgAAAAwwOAMAAAAGGJwBAAAAAwzOAAAAgAEGZwAAAMAAgzMAAABg\ngMEZAAAAMMDgDAAAABhgcAYAAAAMMDgDAAAABhicAQAAAAMMzgAAAIABBmcAAADAAIMzAAAAYIDB\nOQr79zeosPB5FRVtVmHhItXXN1jLnj+/QoHAJGVl1SgQKNHChRXWsrds2aH09EeUnb1B6ekztW3b\nDmvZd9/9QETd99wz2Vq2X61cuUZJSZOVlVWjpKRSrVq1xlq2yz5Zu3a9UlKmqU+fGqWkTFUotN5a\n9uzZCyLqnjt3gbVs17Kyboyo/ZprhlrLdrmnAADsOyveBfhJcfEKLV16i6SANm70JC3WkiWjrGSP\nHRuSNFNSQJKnoqJSjR490kp2bu7z2rmzRFJAO3d6ysmZpbq6SVayf/3rQ2pd95NPluqJJ6xE+9bg\nwa8pHH5YUkDhsKe8vCk6fDjbSrbLPsnJeUGNjQ9KCqix0VP//tN14EAvK9kTJmxS67rHjStVcbGV\naOdCoSvUuva33y61lu1yTwEA2McnzlH44INz1PzmKUmBlmNbekZkNx/bUV9/SUR287Et7ur2q3C4\nh1qvSfOxLe7Wu6mpa0R287Etfu4Td7W73VMAALYxOEehW7ePJXktR566dfvEYvr7EdnSZmvJqam7\nIrJTUz+0lu2ybr8KBreo9ZoEg1stprtb7+TkP0VkJydvt5bt7z5xV7vbPQUAYBu3akShrGyIpMWq\nrZUyMqSyssHWssvLs1RUVKrmT7M2q7w801p2VdWtysmZpX37Oiktbbeqquz9VfD48Ul68smjdY8f\nz5/FKiuHKi9visLhHgoGt6qycoi1bJd9Ul1doP79p6upqauSk7eruvpma9lz5lyhceOO1j1nzuXW\nsl371rdqW27PaK79W9/aYC3b5Z4CALCPwTkKqalf1ZIloxQKhZSZaW9gkaTRo0dq9OiRLdljrGZ3\n736p6uomtWTfZjX7iSf+XU88ISd1+1VubrYOH852siYu+yQzs5cOHOjlJLu4+DYVF/uzT/7wh1cl\nuand5Z4CALCPjwcBAAAAAwzOAAAAgAEGZwAAAMAAgzMAAABggMEZAAAAMMDgDAAAABhgcAYAAAAM\nMDgDAAAABhicAQAAAAMMzgAAAIABBmcAAADAAIMzAAAAYIDBGQAAADDA4AwAAAAYYHAGAAAADDA4\nAwAAAAYYnAEAAAADDM5RWLlyjZKSJisrq0ZJSaVatWqNtez58ysUCExSVlaNAoESLVxYYS27omJF\nq+xJWr58hbXsmTPnRtT96KNzrWX71dq165WSMk19+tQoJWWqQqH11rJdrrfLPpk9e0FE3XPnLrCW\n7Vq/fjdF1N6//whr2S73lP37G1RY+LyKijarsHCR6usbrGUDwJnqrHgX4CeDB7+mcPhhSQGFw57y\n8qbo8OFsK9ljx4YkzZQUkOSpqKhUo0ePtJJdULA6Ijs/v1SeN9hK9uTJf47ILikp1cSJVqJ9Kyfn\nBTU2PigpoMZGT/37T9eBA72sZLtcb5d9MmHCpojsceNKVVxsJdq5P/zhH9W69jVrSq1lu9xTiotX\naOnSWyQFtHGjJ2mxliwZZSUbAM5UfOIchXC4h5rfPKXmN7oeFtN7RmQ3H5/p2f7U1NRVrdek+dgW\nv15LP/eJu9pd7ikffHBORHbzMQDgdDA4RyEY3CLJaznyFAxutZj+fkS2tNli9maH2S7r9qfk5D+p\n9ZokJ2+3mE6fxJ672l3uKd26fRyR3a3bJ9ayAeBMxa0aUaisHKq8vCkKh3soGNyqysoh1rLLy7NU\nVFSq5k+zNqu8PNNa9rJlA5SffzR72bIB1rJnzbpUJSVHs2fN+rq1bL+qri5Q//7T1dTUVcnJ21Vd\nfbO1bJfr7bJP5sy5QuPGHc2eM+dya9muZWe/33J7RnPt2dmbrGW73FPKyoZIWqzaWikjQyors3Pb\nDQCcyRico5Cbm63Dh7MVCoWUmTnGavbo0SM1evRIJ9kjRgyW5w12kj1xYrEmTpSTbL/KzOylAwd6\n+W69XfZJcfFtKi72Z59UVy+X5KZ2l3tKaupXtWTJqJZse38QB4AzGbdqAAAAAAYYnAEAAAADDM4A\nAACAAQZnAAAAwACDMwAAAGCAwRkAAAAwwOAMAAAAGGBwBgAAAAwwOAMAAAAGGJwBAAAAAwzOAAAA\ngAEGZwAAAMAAgzMAAABggMEZAAAAMMDgDAAAABhgcAYAAAAMMDgDAAAABhico7By5RolJU1WVlaN\nkpJKtWrVGmvZU6f+XIHAJGVl1SgQKNH06T+zlj1/fkVE9sKFFb7I9iuXazJ79oKI7LlzF1jLrqhY\n0Sp7kpYvX2Et22Xdrl111dCI2jMzr7eW7XLNAQD2MThHYfDg1xQOPyxpjMLhf1de3mvWsh9+eK+k\nmZLGSHpEM2bss5Y9dmwoIruoKOSLbL9yuSYTJmyKyB43bpO17IKC1a2yZyo/f7W1bJd1u7Z+/TfV\nuvZ3373aWrbLNQcA2MfgHIVwuIekQMtRoOXYlp4R2c3HZ3q2X/l1vf2a7RrrAgBoxuAchWBwiySv\n5chTMLjVYvr7EdnSZrJ9y6/rvdlhtp/7xK9rDgCwjcE5CpWVQxUMTpE0X8HgFFVWDrGWPW3aBZJK\nJc2XVKpp09KsZZeXZ0Vkl5dn+iLbr1yuyZw5V0Rkz5lzubXsZcsGRGQ3H9vhsm7Xevf+o1rX3rv3\n/7WW7XLNAQD2nRXvAvwkNzdbhw9nKxQKKTNzjNXsBx/8Nz34oJxkjx49UqNHj/Rdtl+5XJPi4ttU\nXOymT0aRedIGAAAgAElEQVSMGCzPG+y7ul0LhX7X8v/+WnMAgH184gwAAAAYYHAGAAAADDA4AwAA\nAAYYnAEAAAADDM4AAACAAQZnAAAAwACDMwAAAGCAwRkAAAAwwOAMAAAAGGBwBgAAAAwwOAMAAAAG\nGJwBAAAAAwzOAAAAgAEGZwAAAMAAgzMAAABggMEZAAAAMMDgDAAAABhgcI7C7NkLFAhMUlZWjQKB\nEs2du8Ba9t13PxCRfc89k61l//jHD0dkT5z4kLVsl3X71fz5FRFrsnBhhbXsESPujsguKLjLWvbU\nqT+PyJ4+/WfWsl32oGtnn50RUfuXv9zNWnZFxYpW2ZO0fPkKa9kAAPsYnKMwYcImSTMljZH0iMaN\n22Qt+9e/PhSR/eSTYWvZv/zlxxHZjz32d2vZLuv2q7FjQ2q9JkVFIWvZL710XkR2RUWqteyHH94b\nkT1jxj5r2S570LVDh76n1rUfPFhoLbugYHWr7JnKz19tLRsAYB+Dc1R6Sgq0/DrQcky2u2y/8ut6\n+zXbNdYFANCMwTkq70vyWn7tSdpMttNsv/Lrevs12zWXtW92mA0AsI3BOQpz5lwhqVTSfEmlmjPn\ncmvZ48cnRWSPH2/v0tx//zkR2fff39Fatsu6/aq8PEut16S8PNNa9siR/xuRPXJkvbXsadMuiMie\nNi3NWrbLHnTt7LOXqnXtZ5+9xFr2smUDIrKbjwEACcvziXXr1jk5ty1c5p8J2YlSR1vOJzu22dRC\ndrzriDY/UdaEbDvnu8xOpFrINj+fjwcBAAAAAwzOAAAAgAEGZwAAAMAAgzMAAABgwGhwbmho0Ftv\nvSVJmjdvnn70ox9p69atTguD/9AnMEWvwAR9AlP0CmLFaHC+77779Kc//UlvvfWW/vu//1u5ubma\nPn2669rgM/QJTNErMEGfwBS9glgxGpz/+te/6p//+Z/1xhtvKD8/XyNGjFBjY6Pr2uAz9AlM0Ssw\nQZ/AFL2CWDEanMPhsGpra/X666/ruuuu03vvvafDhw+7rg0+Q5/AFL0CE/QJTNEriJWzTE66//77\n9eijj2rMmDFKT09XQUGBSktLXdcGn6FPYIpegQn6BKboFcSK0eDcr18/9erVS3V1dfI8TwsWLFBK\nSorr2uAz9AlM0SswQZ/AFL2CWDG6VeP//J//oxEjRqi4uFh79+7VwIEDtWbNGte1wWfoE5iiV2CC\nPoEpegWxYjQ4/+IXv9CiRYt07rnn6sILL9TChQv16KOPuq4NPkOfwBS9AhP0CUzRK4gV438ceMEF\nFxw57tGjh7OC4F/0CUzRKzBBn8AUvYJYMbrHuVOnTvr973+vQCCgv/3tb3ruued0ySWXuK4NPkOf\nwBS9AhP0CUzRK4gVo0+cZ8yYoVdeeUV/+ctfNGjQIL333nuaMWOG69rgM/QJTNErMEGfwBS9glgx\n+sT5/PPP1y9+8YuowydPnqyqqiqdf/75euWVVyQ1f5Pye++9V7t27VKXLl30+OOP6ytf+UrU2Ug8\nbe0TiV4507CnwAR7CkyxpyBWTjo45+bmKhAInPD333jjjZOGf//739fo0aM1ceLEI1976qmn1K9f\nP91555166qmnNG/ePP3bv/1blGUjkeTm5urgwYP68pe/fNzfP1WfSPTKmWLChAkn7BOJPQXN2FNg\nij0FsXbSWzUWLlyo//zP/1Tfvn1100036bnnntPixYv1T//0TxowYMApw7OysnTuuedGfO3zH4cp\nSfn5+Xr99ddPo/zY+vGPH1YgMElZWTUKBEo0ceJD1rKnTv15RPb06T+zln3bbRMjsu+88z5r2VOn\n/lzV1Rl6++2b9P/+X1Bf+9pFUfeJ1L565e67H4hY73vumWwt22WfDBkyNiL7xhtvs5ZdUbFCgcAk\nvf32TXrzzSvVqdMlvtlTUlJ6RaxLx47/aC175co1SkqarKysGiUllWrVKnvfPmv//gYVFj6voqLN\nKixcpPr6BmvZa9euV0rKNPXpU6OUlKkKhdZby5aa33umTp3a5vceqX3tKTixqVOnMqcgtjwD+fn5\nRl87np07d3rf+973jhz36dMn4ve/eHwi69atMzov2nOjIZV4UtiTvJb/LyG7VfbXv55/TPbx+uRE\n18dGr0R77V30Snu4lq6zv/71vsecY9orsd5TXK5LMFgakR0MllrLLihYFJFdULDIWnaHDg9EZHfo\n8IC17M+tW7fO+L3H5Z5ysvzTPTdaZJ84+3R6JR590pbzyY5t9onON7rHWZL+8Ic/6JprrpEkrV69\nWklJSVYG95PdCvJFoVDIybnmekr6vN6ApJ4Wn6c9ZEsdOnz5SPYf//hHNTU1WXsu016J9vns90p7\nuJaus7+s8vJyfeMb35Bkt1fs7ynu1iUc7hGRHQ73sJZdW6uI7Npae73e1NQ1IrupqauTPffvf/+7\nsz6R/Pb+Q/bJsl32iqs+acv5ZMc/22hw/ulPf6qSkhLt3btXnuepc+fObf7G4ueff7727duntLQ0\n7d27V6mpqcaPzczMNDovFAoZnxudpZI8Nb9heJI2KzNzDNkt2Xv2PKROnW7TuHG/O9Ins2fPPub7\naZo2blt7JZpr76ZX/H8t7WdXRGTv2XOlKioqNHfu3NPuFfd7irt1CQZfUDh8NDsY3GotOyPjfW3c\neDQ7IyO618bJJCcvU2Pj0ezk5O0We6VZKBTSL3/5S5WUlFjpE8nv7z9knyzbZq/Eok9a1+4C2aef\nfaJeMRqcr7zySr3yyiv63//9XwUCAX31q181fmLP8yKOc3Nz9eKLL+quu+7SsmXLNHDgQOOseLv/\n/nP02GOlav4EarPuv7+jtexp0y7QjBlHs6dNS7OWfccdn+mZZ45m33HHIWvZrev+9NPN+pd/KdS/\n/uvYqPtEaj+9Mn58kp588uh6jx9v9F0fjbjsk+HD9+jll49mDx++21r2smUDlJ9/NHvx4hs1YsRg\nX+wpKSkv6sCBwJHaU1L+S9IjVrIrK4cqL2+KwuEeCga3qrJyiJVcSSorGyJpsWprpYwMqaxssLXs\n6uoC9e8/XU1NXZWcvF3V1Tdby27tdN57pPazp+DUmFMQKwHvix1zHOvWrdOzzz6rAwcOyPM8hcNh\nffjhh1q1atVJH3fffffp7bffVkNDg9LS0nTPPfdo0KBBmjBhgv7yl7+oc+fOevzxx4+5Mf94ovmT\ngss/sbjO93O253lGfXK8Omz1Slv+ROnX9fZztume8sVa4rWntOX8aJB9/OxE2FNOlG/j3GiRfeJs\n9hSybWef6HyjT5ynTp2qO++8U8uWLdPo0aP15ptv6sorrzzl437+858f9+sLFiwweVr4TFv7RKJX\nzjTsKTDBngJT7CmIFaO/Q05OTtZNN92kvn376txzz9VPf/pTvfPOO65rg8/QJzBFr8AEfQJT9Api\nxWhw/vKXv6yGhgZ169ZNNTU1CgQCOnDggOva4DP0CUzRKzBBn8AUvYJYMRqcb7/9dt1777267rrr\ntHz5cg0dOlQZGRmua4PP0CcwRa/ABH0CU/QKYsXoHufk5GTNnz9fgUBAL774orZv367LL7/cdW3w\nGfoEpugVmKBPYIpeQawYfeL82GOPHfkG4CkpKbryyisVDNr7FltoH+gTmKJXYII+gSl6BbFi9Ilz\nenq6SktLddVVVyk5OfnI10eMGOGsMPgPfQJT9ApM0CcwRa8gVowG56997WuSpJqamoiv05BojT6B\nKXoFJugTmKJXECunHJwXLVqk3Nxc5eXl6eabb1Z9fb3OOussPf3007GoDz5RWVlJn8AIewpMsKfA\nFHsKYumkNwDNmzdPK1euPPKz3g8ePKiFCxeqqKhI8+bNi0mBSHzz5s3T2rVr6ROc0ksvvcSeglNi\nT4Ep9hTE2kkH5+XLl2vOnDnq1q1b88nBoDp37qxbb731mL8OwZlr+fLluu++++gTnFJ1dTV7Ck6J\nPQWm2FMQaycdnJOSktSxY8cjx3fffXfzg4JBnX322W4rg28kJSVF/GMM+gQnEgwG2VNwSuwpMMWe\nglg76eAcDof1ySefHDm+/vrrJUkff/yx26rgK+FwWI2NjUeO6ROciOd57Ck4JfYUmGJPQayddHAe\nNmyYSkpKIpry73//uyZPnqzhw4c7Lw7+MGzYMJWVldEnOKXvfOc77Ck4JfYUmGJPQayd9Ltq3HXX\nXfrJT36i/v37q3v37goEAtq6datuvPFG3X777bGqEQnurrvuUk1NDX2CUxo+fLhefvllegUnxZ4C\nU+wpiLWTDs5JSUl66KGHNH78eK1fv16S9I1vfEOXXHJJTIqDPyQlJenOO+/Ugw8+SJ/gpILBIHsK\nTok9BabYUxBrRj8A5aKLLlJeXp7rWuBz9AlM0SswQZ/AFL2CWGlXP8h9//4GFRY+r6KizSosXKT6\n+gar+Vu27FB6+iPKzt6g9PSZ2rZth7XsiooVCgQmKSurRoHAJC1fvsJa9vz5Fa2yS7RwYYW17LVr\n1yslZZr69KlRSspUhULrrWX7lcs1mT17QcS1nDt3gbVslz3o8rXj2t13PxCx5vfcM9latst1cdmH\nrvdaAEhYnk+sW7fulOcUFCzypLAneZ4U9goKFlmtoUuXmRH5XbrMtJYtlURkSyW+yO7Q4YGI7A4d\nHjjlY0yuZVtFm+2ilrasiSm/9klbXzuJ0CuJuC4mXPah673W86K79i77JNr8ROhZsu2d7zI7kWoh\n2/z8dvWJ8wcfnCMp0HIUaDm2p77+koj85mNbekZkNx8nfnZTU9eI7ObjM5vbNfFnn7h97bjmz3Vx\n2Yeu91oASFTtanDu1u1jSV7Lkadu3T452elRS03dFZGfmvqhxfTNEdnNx7a87yw7OflPEdnJydut\nZfuV2zVxdy1d9qDb145r7tbc5bq47EPXey0AJCqjfxzoF2VlQyQtVm2tlJEhlZUNtppfVXWrcnJm\nad++TkpL262qqlHWspctG6D8/FI1f5q1WcuWDbCWXV6epaKio9nl5ZnWsqurC9S//3Q1NXVVcvJ2\nVVffbC3br1yuyZw5V2jcuKPXcs6cy61lu+xBl68d18aPT9KTTx5dl/Hj7X3e4HJdXPah670WABKW\njftGYiFR7jFznX8mZCdKHW05n+zYZlML2fGuI9r8RFkTsu2c7zI7kWoh2/z8dnWrBgAAAOAKgzMA\nAABggMEZAAAAMMDgDAAAABhgcAYAAAAMMDgDAAAABhicAQAAAAMMzgAAAIABBmcAAADAAIMzAAAA\nYIDBGQAAADDA4AwAAAAYYHAGAAAADDA4AwAAAAYYnAEAAAADDM4AAACAAQZnAAAAwACDMwAAAGCA\nwTkKFRUrFAhMUlZWjQKBSVq+fMUZnz179oJW2SWaO3eBtWy/mj+/ImJNFi6ssJa9du16paRMU58+\nNUpJmapQaL217B//+OGIuidOfMhatsu6Xbv77gci1uWeeyZby3b5+tmyZYfS0x9RdvYGpafP1LZt\nO3yRDQAJzfOJdevWOTk3GlKJJ4U9yWv5/xKy25Dt6vq0JdtFLS7Xu0OHByKyO3R4wFp2ItadCL2S\naK8fU126zIzI7tJlpi+yP5cIe35b8hOhZ8m2d77L7ESqhWzz8/nEOSo9JQVafh1oOSbbXbZfuVuT\npqauEdnNx7b4tW7X/Pn6qa+/JCK7+TjxswEgkTE4R2WzJK/l117L8Zme/b7DbL9ytybJyX+KyE5O\n3m4t2791u+ayx91lp6buishOTf3QF9kAkMgYnKOwbNkASaWS5ksqbTk+s7PnzLkiInvOnMutZftV\neXmWWq9JeXmmtezq6gJ16DBdgcB8degwXdXVN1vLvv/+c9S67vvv72gt22Xdro0fn6TW6zJ+vL1t\n0+Xrp6rqVnXpMkvJyQvUpcssVVWN8kU2ACQ0G/eNxEKi3GPmOv9MyE6UOtpyPtmxzaYWsuNdR7T5\nibImZNs532V2ItVCtvn5fOIMAAAAGGBwBgAAAAwwOAMAAAAGGJwBAAAAAwzOAAAAgAEGZwAAAMAA\ngzMAAABggMEZAAAAMMDgDAAAABhgcAYAAAAMMDgDAAAABhicAQAAAAMMzgAAAIABBmcAAADAAIMz\nAAAAYIDBGQAAADDA4AwAAAAYYHCOwv79DSosfF5FRZtVWLhI9fUNvsjesmWH0tMfUXb2BqWnz9S2\nbTusZbus26/Wrl2vlJRp6tOnRikpUxUKrbeW7dce9DO/vn78mg0AieyseBfgJ8XFK7R06S2SAtq4\n0ZO0WEuWjEr47Nzc57VzZ4mkgHbu9JSTM0t1dZOsZLus269ycl5QY+ODkgJqbPTUv/90HTjQy0q2\nX3vQz/z6+vFrNgAkMj5xjsIHH5wjKdByFGg5Tvzs+vpLIrKbj+1wWbdfNTV1Ves1aT62w6896Gd+\nff34NRsAEhmDcxS6dftYktdy5Klbt098kZ2auisiOzX1Q2vZLuv2q+TkP6n1miQnb7eW7dce9DO/\nvn78mg0AiYxbNaJQVjZE0mLV1koZGVJZ2WBfZFdV3aqcnFnat6+T0tJ2q6rK3l+puqzbr6qrC9S/\n/3Q1NXVVcvJ2VVffbC3brz3oZ359/fg1GwASGYNzFFJTv6olS0YpFAopMzPTN9ndu1+qurpJLdm3\nWc12WbdfZWb20oEDvVrWZIzVbL/2oJ/59fXj12wASGTcqgEAAAAYYHAGAAAADDA4AwAAAAYYnAEA\nAAADDM4AAACAAQZnAAAAwACDMwAAAGCAwRkAAAAwwOAMAAAAGGBwBgAAAAwwOAMAAAAGGJwBAAAA\nAwzOAAAAgAEGZwAAAMAAgzMAAABggMEZAAAAMMDgDAAAABhgcI7C/v0NKix8XkVFm1VYuEj19Q3W\nsleuXKOkpMnKyqpRUlKpVq1a44vsLVt2KD39EWVnb1B6+kxt27bDWrZfrV27Xikp09SnT41SUqYq\nFFpvLduv19LPfXL33Q8oEJikrKwaBQIluueeyfEuCQAQJ2fFuwA/KS5eoaVLb5EU0MaNnqTFWrJk\nlJXswYNfUzj8sKSAwmFPeXlTdPhwdsJn5+Y+r507SyQFtHOnp5ycWaqrm2Ql269ycl5QY+ODkgJq\nbPTUv/90HTjQy0q2X6+ln/vk178+JGmmpIAkT08+WaonnohzUQCAuOAT5yh88ME5an7zlKRAy7Ed\n4XCPiOzm48TPrq+/JCK7+fjM1tTUVa3XpPnYDr9eS3/3SU+1rr35GABwJmJwjkK3bh9L8lqOPHXr\n9om17GBwS0R2MLjVF9mpqbsislNTP7SW7VfJyX9S6zVJTt5uLduv19LfffK+WtcubY5jLQCAeOJW\njSiUlQ2RtFi1tVJGhlRWNthadmXlUOXlTVE43EPB4FZVVg7xRXZV1a3KyZmlffs6KS1tt6qq7Ny6\n4mfV1QXq33+6mpq6Kjl5u6qrb7aW7ddr6ec+GT8+SU8+WarmT5o3a/x4Pm8AgDMVg3MUUlO/qiVL\nRikUCikzM9Nqdm5utg4fzm7JHuOb7O7dL1Vd3aSW7NusZvtVZmYvHTjQi2sZo2zXnnji3/XEE3Ky\n5gAAf+GjEwAAAMAAgzMAAABggMEZAAAAMMDgDAAAABhgcAYAAAAMMDgDAAAABhicAQAAAAMMzgAA\nAIABBmcAAADAAIMzAAAAYIDBGQAAADDA4AwAAAAYYHAGAAAADDA4AwAAAAbOitcT5+bm6pxzzlEw\nGNRZZ52lF154IV6lIMHRKzBBn8AUvQJT9Aq+KG6DcyAQ0MKFC3XeeefFqwT4BL0CE/QJTNErMEWv\n4IvidquG53kKh8Pxenr4CL0CE/QJTNErMEWv4IviNjgHAgGNGTNGN910k5YuXRqvMqKyZcsOpac/\nouzsDUpPn6lt23bEuyQja9euV0rKNPXpU6OUlKkKhdZby96/v0GFhc+rqGizCgsXqb6+wVr25/zW\nKy7XxK/X0s994vJ17zLb79fTJb/tKX7l9z6R6BUchxcne/bs8TzP8/bv3+8NHz7ce+edd056/rp1\n64yzozk3Gl26zPSksCd5nhT2unSZaf05XNTeocMDEXV36PCAteyCgkUR2QUFi075mGj/G6PplWiz\nXax3W9bEVKJdS9fZ0VwfV3uKy9e9y+xEvJ7RcLnn+/H9x4/ZidYnbTnfb+8/ZNvLPtH5cbvH+cIL\nL5QkpaamKi8vTxs2bFBWVtZJHxMKhYzzoznX1L59nSQFWo4C2revk5PnsZ3Z1NRVretuaupq7Tlq\naxWRXVtrv/5oeyXa57ddr8s18eu1TMQ+kcxqcPm6d5nt9+spucmU/Pn+48dsv/eJ5L/3H7LdZ8dl\ncG5sbFQ4HFbHjh114MABrVmzRuPHjz/l4zIzM43yQ6GQ8bnRSEtbqZ07PTVvBJ7S0nYrM/M2q8/h\novbk5GVqbDxad3LydmVmjrGSnZHxvjZuPJqdkXHq6xRN47alV6JZPxfr3ZY1MZVo19J1tmmvuNxT\nXL7uXWYn4vWMRjSvTdd7ihT/9x8/Zidan3x+vik/vv+QbS/7RL0Sl8F53759Gj9+vAKBgA4fPqxh\nw4YpOzs7HqVEparqVuXkzNK+fZ2UlrZbVVWj4l2SkerqAvXvP11NTV2VnLxd1dU3W8suKxsiabFq\na6WMDKmsbLC1bMmfveJyTfx6Lf3cJy5f9y6z/Xw9XfLjnuJXfu4TiV7BCdi4byQWEuUeM9f5Z0J2\notTRlvPJjm02tZAd7zqizU+UNSHbzvkusxOpFrLNz+cnBwIAAAAGGJwBAAAAAwzOAAAAgAEGZwAA\nAMAAgzMAAABggMEZAAAAMMDgDAAAABhgcAYAAAAMMDgDAAAABhicAQAAAAMMzgAAAIABBmcAAADA\nAIMzAAAAYIDBGQAAADDA4AwAAAAYYHAGAAAADDA4AwAAAAYYnAEAAAADDM4AAACAAQZnAAAAwACD\nMwAAAGCAwRkAAAAwwOAMAAAAGGBwBgAAAAwwOAMAAAAGGJwBAAAAAwzOAAAAgAEGZwAAAMAAgzMA\nAABggMEZAAAAMMDgDAAAABhgcAYAAAAMMDgDAAAABhicAQAAAAMMzgAAAIABBmcAAADAAIMzAAAA\nYIDBGQAAADDA4AwAAAAYYHAGAAAADDA4AwAAAAYYnAEAAAADDM4AAACAAQZnAAAAwACDMwAAAGCA\nwRkAAAAwwOAMAAAAGGBwBgAAAAwwOAMAAAAGGJwBAAAAAwzOAAAAgAEGZwAAAMAAgzMAAABggMEZ\nAAAAMMDgDAAAABhgcAYAAAAMMDgDAAAABhicAQAAAAMMzgAAAIABBmcAAADAAIMzAAAAYIDBGQAA\nADDA4AwAAAAYYHAGAAAADDA4AwAAAAYYnAEAAAADDM4AAACAAQZnAAAAwACDMwAAAGCAwRkAAAAw\nwOAMAAAAGGBwBgAAAAwwOAMAAAAGGJwBAAAAAwzOAAAAgAEGZwAAAMAAgzMAAABggMEZAAAAMMDg\nDAAAABhgcAYAAAAMMDgDAAAABhicAQAAAAMMzgAAAIABBmcAAADAAIMzAAAAYIDBGQAAADDA4AwA\nAAAYYHAGAAAADDA4AwAAAAYYnAEAAAADDM4AAACAAQZnAAAAwACDMwAAAGCAwRkAAAAwwOAMAAAA\nGGBwBgAAAAwwOAMAAAAGGJwBAAAAAwzOAAAAgAEGZwAAAMAAgzMAAABggMEZAAAAMMDgDAAAABhg\ncAYAAAAMMDgDAAAABhicAQAAAAMMzgAAAICBuA3Ob775pm644QZdf/31euqpp+JVBnyAXoEJ+gSm\n6BWYoE9wPHEZnMPhsB566CE9++yz+u1vf6tXX31V27Zti0cpSHD0CkzQJzBFr8AEfYITicvgvH79\nel166aXq3LmzvvSlL2no0KF644034lFKVFauXKOkpMnKyqpRUlKpVq1aYy175sy5CgQmKSurRoFA\niR59dK61bJd1r127Xikp09SnT41SUqYqFFpvLVvyZ6/Mn18RcS0XLqzwRbbLa+myByW3ffLjHz8c\nseYTJz5kJVeSKipWtMqepOXLV1jLdr3mfuXHPQXH2r+/QYWFz6uoaLMKCxepvr7Bar7LPnH5fo9j\n2e6VsyzVFZU9e/bo4osvPnJ80UUXacOGDfEoJSqDB7+mcPhhSQGFw57y8qbo8OFsK9mTJ/9Z0kxJ\nAUmeSkpKNXGilWindefkvKDGxgclBdTY6Kl//+k6cKCXlWzJn70ydmxIra9lUVGpRo8emfDZLq+l\nyx6U3PbJL3/5sVqv+WOPlerRR61Eq6BgdUR2fn6pPG+wlWzXa+5XftxTcKzi4hVauvQWSQFt3OhJ\nWqwlS0ZZy3fZJy7f73Es270Sl8G5rUKhkJNzTYXDPdTc6FLzm1EPi8/TMyJb6mkt22XdTU1dI7Kb\nmro6WftoRPv89ut1dy1dZru8lm5fO21nVoM/r2es1tzldUyEHvlcvN9/yI5UWyu17u/a2sTol/jv\nKdHW0v6zbfdKXAbniy66SB9++OGR4z179ujCCy885eMyMzON8kOhkPG50QgGX1A47OnzPyUGg1uV\nmTnGUvpSSUezpc3Wsl3WnZy8TI2NR7OTk7efMjuahm1Lr0Rz7d30irtr6TK7LdfSVFt70LRX3O4p\nLq9nhbNst/tVM1d7bbTZrvcUKf7vP2RHysh4v+XTw+b+zsgwu0btf09p5rfr6TLbdq/E5R7nb37z\nm/rzn/+sXbt26eDBg3r11Vc1cODAeJQSlcrKoQoGp0iar2Bwiiorh1jLnjXrUkmlkuZLKtWsWV+3\nlu2y7urqAnXoMF2BwHx16DBd1dU3W8uW/Nkr5eVZan0ty8vtbQIus11eS5c9KLntk/vvP0et1/z+\n+ztayZWkZcsGRGQ3H9vhes39yo97Co5VVjZEBQWLdeWVz6ugYLHKyuzc4vQ5l33i8v0ex7LeK16c\nrF692vvud7/r5eXlefPmzTvl+evWrTPOjubctnCZfyZkR1tHNL0SbXairAnZp3++yz2lLeeTHbts\nlzpZkN0AAAe+SURBVHuK61qiQfbpZ7OnkH2658ftHudrr71W1157bbyeHj5Cr8AEfQJT9ApM0Cc4\nHn5yIAAAAGCAwRkAAAAwwOAMAAAAGGBwBgAAAAwwOAMAAAAGGJwBAAAAAwzOAAAAgIGA53levIsw\nkQg/gx72uPxxnWhf6BWYcNUnEr3S3rCnwNTxesU3gzMAAAAQT9yqAQAAABhgcAYAAAAMMDgDAAAA\nBhicAQAAAAMMzgAAAICBdjc4v/nmm7rhhht0/fXX66mnnrKaPXnyZH3729/WsGHDrObu3r1bRUVF\nGjp0qIYNG6by8nJr2QcPHtTIkSM1YsQIDRs2TE8++aS17M+Fw2Hl5+frhz/8ofVsV/zYJxK9Eg9+\n7BX6JD5c9Qp7yon5sVfYUyL5rk+8duTw4cPeoEGDvJ07d3oHDx70hg8f7m3dutVa/jvvvONt3LjR\n+973vmct0/M876OPPvI2btzoeZ7nffLJJ953v/tdq3UfOHDA8zzP++yzz7yRI0d6NTU11rI9z/P+\n4z/+w7vvvvu8H/zgB1ZzXfFrn3gevRJrfu0V+iT2XPYKe8qJ+a1X2FOOz0990q4+cV6/fr0uvfRS\nde7cWV/60pc0dOhQvfHGG9bys7KydO6551rL+9wFF1ygK664QpLUsWNHde/eXR/9//buL6SpPo7j\n+OfkA1GGEI3G6LI/IEg3FQqCVoyK0Jw5QQbdzKCLSIiY/UG7CcUKgiCoq8KLoG7KiG5DCxYTDAQv\n+iOEuQlTJKnG0s52ngtB6GlPz692tjo+79eVCn7Ppm8O321nOjvr2vx169ZJWn5UZ9u2a3Ol5Ueh\nIyMjam9vd3VuKXm1E4lWys2rrdBJ+ZWyFc4phXmxFc4phXmpk1W1OKfTaQUCgZXP/X6/q7/Yckgm\nk3r16pV27tzp2sx8Pq9QKKT6+nrV19e7Oru/v1/d3d2yLMu1maW2GjqRaKUcVkMrdFIetFIYrXyL\nTgrzUieranH2ukwmo66uLl24cEGVlZWuzV2zZo2Ghob07NkzjY+Pa3Jy0pW5w8PD8vl8qq6ulsM/\noCwrWoEJOoEpWoEJOpH+cmXKH8Lv92tmZmbl83Q6rc2bN//GW2TOtm11dXWppaVFwWCwJMfYsGGD\namtr9fz5c23btq3oeS9fvtTTp081MjKixcVFZTIZdXd368qVKy7c2tLxcicSrZSTl1uhk/KilR+j\nlWV08mOe6KToq6T/ILZtr1x0v7i46PpF947jONPT0yV5g0YsFnP6+/tdnzs/P+98/PjRcRzHyWaz\nTiQScYaHh10/TiKR8MybM7zciePQSjl5uRU6Ka9St8I55d95qRXOKd/zWier6hnniooK9fb2KhqN\nynEchcNhbd261bX5Z86cUSKR0MLCgvbu3atTp06pra2t6LljY2N6/PixduzYoVAoJMuydPr0aTU0\nNBQ9e25uTufOnVM+n1c+n9fhw4fV2NhY9Fwv82onEq2Um1dboZPyK2UrnFNWD84p3/NaJ5bjcHEQ\nAAAA8F94cyAAAABggMUZAAAAMMDiDAAAABhgcQYAAAAMsDgDAAAABlicAQAAAAOr6u84/06pVEoH\nDx7U9u3bJUmO48iyLN26dUt+v994TjKZ1M2bN9XX11eqm4rfiE5gilZggk5gilbcweLsIr/fr4cP\nHxY1I5VKaXp62qVbhD8RncAUrcAEncAUrRSPSzVKbH5+XidPnlRbW5va29v14sULScv/n/748ePq\n6OjQ/v37de3aNUlSX1+fJiYmdOnSJY2OjurYsWMrs86fP6+hoSGlUikdOnRIkUhE0WhU+XxeAwMD\nOnr0qEKhkAYHB3/LfcWvoxOYohWYoBOYopWfwzPOLkqn02ptbV15+aO5uVkTExMKh8Pat2+f5ubm\nFIlE9OjRIz158kRNTU0KhUL6/PmzGhsbFY1G1dPToxs3bqi3t1ejo6OyLKvgsaampnTnzh0FAgHd\nu3dPlmXpwYMHWlpaUmdnp2pqarRr164y/wRggk5gilZggk5gilaKx+LsokIvgdTV1endu3e6fv26\nJCmXy+n9+/eKRqNKJBK6ffu23r59K9u2lc1mjY+1adMmBQIBSVI8Htfr169XHiVms1m9efPGk0H+\nH9AJTNEKTNAJTNFK8VicS8xxHA0ODqqqqkqSNDs7K5/Pp4GBAaVSKTU3NysYDCoej8txnG++17Ks\nb7729evXlY/Xrl278nE+n1csFlMwGJQkffjwQZWVlaW8W3AZncAUrcAEncAUrfwcrnF20T+DkqTa\n2lrdvXtXkjQ5OamWlhZ9+fJF8XhcnZ2dOnDggGZmZjQ7O6tcLqeKigrZti1J2rhxo5LJpJaWlrSw\nsKCxsbGCx6qrq9P9+/dl27YymYwikYjGx8dLfG/xq+gEpmgFJugEpmileDzj7KJC1/n09PTo4sWL\nOnLkiCTp6tWrWr9+vU6cOKFYLKaqqir5fD7V1NQomUyqurpanz590tmzZ3X58mU1NDSoqalJW7Zs\n0e7duwseq6OjQ1NTU2ptbVUul1M4HNaePXtKf4fxS+gEpmgFJugEpmileJZT6OEHAAAAgG9wqQYA\nAABggMUZAAAAMMDiDAAAABhgcQYAAAAMsDgDAAAABlicAQAAAAMszgAAAIABFmcAAADAwN+UR6OS\nhPqaqQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2ce9d898>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.feature_selection import *\n",
    "fs=SelectKBest(score_func=f_regression,k=5)\n",
    "X_new=fs.fit_transform(X_train,y_train)\n",
    "z =  zip(fs.get_support(),features)\n",
    "print z\n",
    "\n",
    "x_min, x_max = X_new[:,0].min() - .5, X_new[:, 0].max() + .5\n",
    "y_min, y_max = y_train.min() - .5, y_train.max() + .5\n",
    "#fig=plt.figure()\n",
    "#fig.subplots_adjust(left=0, right=1, bottom=0, top=1, hspace=0.05, wspace=0.05)\n",
    "\n",
    "# Two subplots, unpack the axes array immediately\n",
    "fig, axes = plt.subplots(1,5)\n",
    "fig.set_size_inches(12,12)\n",
    "\n",
    "for i in range(5):\n",
    "    axes[i].set_aspect('equal')\n",
    "    axes[i].set_title('Feature {}'.format(i))\n",
    "    axes[i].set_xlabel('Feature')\n",
    "    axes[i].set_ylabel('Grades')\n",
    "    axes[i].set_xlim(x_min, x_max)\n",
    "    axes[i].set_ylim(y_min, y_max)\n",
    "    plt.sca(axes[i])\n",
    "    plt.scatter(X_new[:,i],y_train)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "best_features = []\n",
    "for bool,feature in z:\n",
    "    if bool:\n",
    "        best_features.append(feature)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "correlated = best_features + ['G3']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Medu', 'Fedu', 'failures', 'G1', 'G2', 'G3']"
      ]
     },
     "execution_count": 177,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "correlated"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA1UAAANWCAYAAAAWa6ZoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x25a4fa20>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "sns.set(style='whitegrid', context='notebook')\n",
    "\n",
    "sns.pairplot(df[correlated], size=2.0);\n",
    "plt.tight_layout()\n",
    "# plt.savefig('./figures/scatter.png', dpi=300)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAcQAAAF0CAYAAAC9lUlqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/wG/g+gcqVtwkgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x24cd1c88>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "cm = np.corrcoef(df[correlated].values.T)\n",
    "sns.set(font_scale=1.5)\n",
    "hm = sns.heatmap(cm, \n",
    "            cbar=True,\n",
    "            annot=True, \n",
    "            square=True,\n",
    "            fmt='.2f',\n",
    "            annot_kws={'size': 15},\n",
    "            yticklabels=correlated,\n",
    "            xticklabels=correlated)\n",
    "\n",
    "plt.tight_layout()\n",
    "# plt.savefig('./figures/corr_mat.png', dpi=300)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZMAAAEpCAYAAABWTc9yAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRBJTMhERkcSUTEREJDElExERSez/Ax2YzFkSo0uiAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x272232e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "plt.scatter(df['failures'], df['G3'])\n",
    "plt.xlabel('Failures')\n",
    "plt.ylabel('G3')\n",
    "plt.title('Failures Against G3')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn.cross_validation import train_test_split\n",
    "\n",
    "X = df[features].values\n",
    "y = df['G3'].values\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "slr = LinearRegression()\n",
    "\n",
    "slr.fit(X_train, y_train)\n",
    "y_train_pred = slr.predict(X_train)\n",
    "y_test_pred = slr.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAacAAAEYCAYAAAD4czk4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzsnWd4VNXWgN9J6BBICAkhIkkIsMECXBtiQ0CxIYhcBFQu\neAUUyxWwxF4RNKBiQxAEERWxUQRRNDTlYrl+omLZQIAAoUlCKEpN5vtxZpKZyfSSORnW+zw8ZE7Z\ns86ec/Y6a+2117JYrVYEQRAEwUzERVsAQRAEQXBFlJMgCIJgOkQ5CYIgCKZDlJMgCIJgOkQ5CYIg\nCKZDlJMgCIJgOmpEWwCl1GQgTms93GFbD+BZQAHrgPu11p9FSURBEAShiomq5aSUehIY7rLtFGA+\nMAfoCCwA5iml2lW9hIIgCEI0iIrlpJTKAt4ATgUKXHb/B1ittX7G9vlRpdQFwF3ArVUnpSAIghAt\nomU5nQdsAU4HNrvsuxBY7rJtuW27IAiCcAIQFctJa/0O8A6AUsp1d3Og0GXbduDkyEsmCIIgmAEz\nRuvVAw67bDsC1ImCLIIgCEIUMKNyOgTUdtlWG/grCrIIgiAIUSDqoeRu2Ao0c9mWTmVXXyWsVqvV\nYrFERChBEAQhYIIekM2onL4GugBPO2zrCqz0daLFYuHPPw9ESi7BRkpKgvRzhJE+jjzSx5EnJSUh\n6HPNqJxeBv6nlHocmA3cAJyDhJELgiCcMJhhzsmp2qHWei3QB+gL/Aj0BHpqrXUUZBMEQRCigCXG\nKuFaxUyPPOIOiTzSx5FH+jjypKQkBD3nZAbLSRAEQRCcEOUkCIIgmA5RToIgCILpEOUkCIIgmA5R\nToIgCILpEOUkCIIgmA4zLsIVhGpFaWkpmzdvdNqWmdmS+Pj4KEkkCNUfUU6CECKbN2/kzW2vk5yR\nDEBRQRFDGE52dusoSyYI1RdRToIQAsXFJTz88Jd0GpdMaqvUih2l0ZNJEGIBUU7ViLFjn2Dx4oVY\nLBbcZfZIS0vngw/mB932n3/u5oUXXvXr+H79enH11dfwr3/9O6jvC4aRI28jNbUpDz74mF/Hb968\niR07Cunc+YKIyZSTs4y8vN504q2IfYcgnIiIcqpGjBx5DyNG3AnArl07GT58CM888zzt2p0CQFxc\n8PEtI0feQ1mZ/6mspk17izp1zF3/8f777+ayy66IqHIqKGgIWCgqKCrfVlRQZNRzFgQhaEQ5VSPq\n1atPvXr1AThy5AhWq5WEhASSkhqHpe1AaNQoMeTvjDyRzxuZkbGPNWtaMqXfaMBK9+7zGTOmL5mZ\nLSP+3YIQy4hyikH69evFxRd3Z9WqlRw4cIDnn3+ZhISGvPLKRNas+YGDBw+SkpJKnz79uP76QYCz\nW+/HH3/g7rv/wxNPPM3kya+wY8d2MjOzGDnyXtq371j+HXa33vTpr/Prr2s5/fT2zJ37IQcPHuDM\nM88mJ+dhkpObALB16xaef/5Z1q79mcTExgwdegtjxz7Biy++RseOZ1S6BqvVyvTpr7NgwVwOHz5E\nz57XUFZW5nTM8uV5vPPOTDZuzMdisdC6teKuu+6hbdt23HnnLRQWbmP69Nf59NOFfPDBfHbs2M6r\nr77osQ/c4SsSLze3G/AuBQUNycjYT27uQJKSqoPiFgRzI+ucgsBqtXL//XM577w8LrnkUxYs+D7a\nIlVi3ryPyMl5hNzcF2jVqg05OaM4fvwYL7/8Ou+88yGXX34Vr732Ehs2rHd7/vHjx5gxYxr33/8I\nb775LvXrN2DcuCc9ft+PP/6P/PwNvPTSZCZOnMS6dZpp06YAcPjwYUaOvI06deowdepb5OQ8xLRp\nU9zOm9mZOfMNPvxwDqNH5zB16kwOHNjPjz/+UL7/jz9+47HHHuSqq3rxzjsf8corUwErubljAHj6\n6fGkpaUzcOCNvPGGMR90//2jA+oDqIjE+yT+Iz6J/4g3t73upKySkhKZOrUPS5Z0Z+rUPqKYBCFM\niHIKgqlT85gx42o2bLiGn3/uz6OPHmHv3uJoi+XEBRdcRIcOHWnX7lSOHj3KlVdezT33PEhWVktO\nOqk5gwffTFxcHBs3bnB7vtVq5dZb7+D00zvQokUm/ftfT2HhNvbtK/F4/EMPPU5GRiann96B7t0v\n5ddffwYgL28Jf/11kEceeYrMzCzOOuscRo2616v8c+d+wMCBN9KlS1datMjkvvseIiWlIhouPr4G\nd999P9dc80/S0tJo27YdPXteQ36+cT0NGzYkPj6OunXr0bBhI44cORJwH9hJzjAi8VJbpZaHiwuC\nEFlM69ZTStUDngWuBeoBq4G7tda/R1UwoKDgOFZrUvnn7dvbUlCwPSxzP+EiPf2k8r9r167Ntdde\nR17eEn7//Ve2bt3Khg3rsFqtlJa6j3m2WCw0b35y+ef69RsAcOzYcbfHN26c7BQgUb9+A44dOwbA\n+vWajIws6tWrV76/ffuOHi2nkpISiouLadOmbfm2GjVq0KaNKv/cunUbEhISmDXrTTZv3si2bVtZ\nv36dxzaD6QNBEKKHmS2nl4BuGBVxzwUOA4uVUrWiKhVw5pmNqFOnwrXTps0PtGqVGT2B3FC7du3y\nvw8dOsTw4YOZPfttkpIa07v3tUyf/jYWi/c6YLVquetq94N/zZqVj7Xrifj4+ErzRd6wi+WqaGrU\nqFn+9w8/fM/11/+TDRvWccopp3LrrXdw1113e2wz2D4AI/pu94bd7N6w2ykqTxCEyGFaywnoDTyu\ntf4GQCn1EPArcAqwJpqCXXttZ/bsWcqXX/5IvXrHGDWqDQ0aNIimSF757rvV5Odv4NNPl5bLuWXL\n5oAURihkZ7dm4cL5/P33X+VRgb/++otHxdCoUSIpKamsXfsznTufDxiKav16TYcO/wDg/fffpVOn\nc3niibHl53377WqXliraD7YPMjNbMoThFYtqmyOReIJQBZhZOf0J9FdKvQ/sA4YCRcBGr2dVEcOH\nd2P48GhL4R+pqU0B+PzzRZx/fhe2bdvCK69MxGKxcPToUbfnuHOPeQtg8MYll1zGG29M4amnHmPY\nsBGUlOzlhRfGA3hUUAMG3Mi0aZNp0SKDdu1O5YMP3mPnzh3lyik1NY3Vq7/m11/X0rhxY1atWskH\nH8wG4NixY9SsWZN69eqxdesW9uzZE1QfgGH1SRoiQah6zOzWGw60AHYBfwE3A1dprfdHVSoT4dkl\n5by9XbtTue22//DOO28xaFA/Jk6cwGWXXck//nEWf/zxm99tO2+z+OUSA8M9OGHCSxw8eIBhw/7F\nM888Re/e1wLOrjpHrrtuIEOH3sLrr0/i5ptv5PDhQ1x0Udfy/UOH3kKbNm25++47GTp0EF99tYKH\nH34CgN9//xWAAQNuYPXqVQwZMpB27U5lxIjA+kAQhOhhCfZtONIopW7CUFCPA8XAPcB5QCet9XYP\np1n//PNA1Qh4ApOSkkAg/bxz504KC7dy5plnl29bu/YXbrvtZj76aKFTFJ5gEGgfC4EjfRx5UlIS\n/HuDdYMpLSelVCbwOvAfrfXnWuvvgRswgiJGRVM2IXCOHDnM6NF3MHfuh+zcuYPfflvLq6++QMeO\nZ0RVMRUXlzBs2Fx69Mhj2LCP2bvXfZi8IAhVjyktJ6XUP4E5QE2tdZnD9vdt2/p4ONV8FyMAsHjx\nYiZPnkxBQQH16tWja9eu3HvvvSQmRm/Rav/+s3n//QEYblAr1133HnPmDIyaPIIQgwRtOZk1IGKb\n7f/2OEfmnQJ86u1EMdMjTzDukLPOuoBp05wTsB47Ft3fa926ulQ8OxbWratrmvtHXE6RR/o48qSk\nJAR9rlmV03fAt8CbSqnbgT0Y7ryTgZejKZhQ/Th69CirVq102nb++RfZkrZasVtOGRkSayMIZsGU\nyklrXaaU6gmMA2YDDYD/ARdorbdGVTih2rFq1UqWxn3uVKmWVfakrbMckrZ29d6QIAhVhimVE4DW\nuhi4JdpyCLGBPT9eOZsqkrYKgmA+TBmtJwiCIJzYmNZyEoRgcFd/qaystHKlWkEQTI0oJyGmsNdf\ncpxfuuGkm+i24zLYVHHc+edfFCUJBUHwB1FO1YixY59g8eKFWCwWt3nu0tLS+eCD+SF/z8aN+eze\nvYtzzz3P73McK+P6wy+//ITFEsdpp50erJgecZ1fqllak65dLwn79wiCEDlkzqkaMXLkPSxY8Dnz\n53/G1KkzsVgsPPvsCyxY8DkLFnzOtGkzw/I9OTmjWbfuj7C05YnbbhvK9u2FYWvPnu3hpptcM5ML\nglAdEcupGlGvXv3ykhNHjhzBarWSkJAQgSKH1S/RRk7OMubPHwSsp6jg+fLtRQVF0Dx6cgmCEByi\nnGKQo0ePMnnyy+TlLeHw4cMo1Y7bbvsPbdueAkBR0R6ee+5Z1qz5P44dO8Zpp53OHXeMIju7Fbfd\nNpSdO3cwdeprfPrpQt577+NK7R85coQJE55h6dIviIuzcMMNQyod89FH7zNv3ocUFm6jRo2anH56\nB+65536aNUvn2muvwmKxMGbMoyxe/AkvvPAq69b9wZQpk/j11585cuQI6eknMXjwUHr0uNyvay4o\naIixmDabKf1G07btF8yY0VnqLwlCNUXcekFQWlpKfv56p39mKvX9xBMP8dtvv/L00xOYNm0WHTr8\ngzvuGF7uRhs/3ijQN2XKDN54YxZ16tTh0UfvB+CZZ54nJSWVG28cwuuvz3Db/qOPPsq33/6XMWOe\n5aWXJvP999+wa9fO8v15eUuYPPllhg69ldmzPyY39wUKC7cxadJLAEyf/g5Wq5XRo+/jqaee5e+/\n/2b06Dtp1iydN954m5kz3+P00zuQmzuG/fv3+XXNGRn7MCy+eKA1SiWQnd2a7OzWxMfHB9eRgiBE\nDbGcgsBdRNgQhpuiKN2WLZtZuXI57777ESef3AKAm2++hZ9++pE5c95h1Kj7KCzcximnnEbTpmnU\nqlWL++57mIICI5StYcOGxMfHU7duXRo2bFSp/QMHDrBw4UIeeeQpzjjjLAAeeeQp+vbtWX5M48bJ\nPPDAY3Tp0g2Apk3T6NbtElasWApQnuy1Xr0GNGjQgL17i7n++kH06zeQmjWN+k433DCYTz/9hK1b\nt3LqqZXlcEWyPQhCbCHKKUgqZRwwieG0bp0G4OabBzlF9B0/foy4OMNQHjJkGGPHPs7SpV/QseMZ\nnHvu+Vx6qb/uM6O0uVJty7clJSWRnn5S+ed//ONM8vM3MH366xQUbGbr1gI2bswnLS3dbZtJSY3p\n3ftaFi2az/r169i2bWv5dZSV+dexku1BEGILUU4xRs2aNbFYLEybNpP4eOeft3bt2gB0734pnTp1\nZvXqr/n++2+ZMWMqs2bN4K233nNrLTlir37rGspes2bFdy1atIAJE8ZxxRU9OeOMs7juuuv5+usV\nLFuW57bNP//czS233ESzZumcd94FXHDBRSQlNWbYsMHlxxQXlzBq1CJWrz4AJNO5cykTJ15OUlL0\nSm4IghA5RDkFSaWMAyaJCMvKMib/i4uL6djxjPLt48ePpXVrRc+evXnttZe58sqrufTSy7n00svZ\ns2cPffpcwZo1P3LRRRfjrQRLZmYmNWrUYO3an8vdhgcPHmTbtm3lx8yePYu+fftzxx0jnbZ5igJc\nsmQxx44d49VXp5Zv++9/v3Zaz5WTs4zFixtipFu0sHixlVq1Zom1JAgxiiinIMjMbMkQhle48kwU\nEdaiRSYXXdSVZ54Zw+jR93HSSc1ZsGAuixYt4MUXJ1OjRg20/p3fflvLXXfdTcOGjVi8eCE1a9ak\nTRvDVVevXj22bClgz549NGnSxKn9+vUbcN111zF58iskJibStGkzpk2bzLFjR8uPSU1tyk8//R/5\n+RuoVasWS5YsZuXKZTRpklJ+TL169di0KZ+9e/eSmtqUgwcPsGLFUpQ6hXXr/uDFFycAcOzYMcAe\njRePY/0lY5sgCLGIKKcgiI+PN0Xwg93F5spDDz3Oa6+9xNNPP8Zff/1FZmZLxo2bQIcOHQF44omx\nvPTSc9x770j+/vsvsrNb8+yzL5CWlgbAgAE38NJLz/P9998wf/7nldp/8MEHgRqMHfskx44do1ev\nPijVrnz/6NE5jB8/lltv/Td169bllFNO5Z57HuC5554pV3g33jiEWbPe5H//+46pU2fyxx+/89xz\nz3L48GGaNz+ZYcNGMH366/z++6+ceebZttpLNTCsL6m/JAixjinLtIeAVSpbRp5IVBB1l7A1M7Nl\neRj43r0ljBy5kNWrD2LMOZUxceJlMTvnJFVaI4/0ceRJSUmIuTLtwgmGr/D8pKREZs68MZoiCoJQ\nhZhaOSmlhgL3YpRn/w24V2u9LLpSCZHCrOH5giBUPabNEKGUGgy8AowFTgNWAAuUUi2iKpgQNuyZ\nNn744f8YPXo2RQVFlJWWRVssQRBMgGmVE/A4ME5rPVNrvRG4B1gP+F/HQTA1dlfeV8lL6PpiLQD0\ncs3uDbulIKAgnOCY0q2nlFJABvC+fZvW2gqc4fEkwfS4Bj1s2VJAUoskJ1fe2HMuom3bDcyY0dc0\n4fmCIFQ9plROQBuMmOEkpVQehlvvD+B+rbUU7KmmVAp6iCuCbZCm0hyOykSpXaYI1RcEIXqY1a1n\nr3/wJvA6cBmwFlhqs6qEaoo96CG1VSrJGcmUbC9h94bd5a687t3nS9JWQRBMazkds/0/Rms9x/b3\n7UqpC4ERwEj3pxlrF4TIE0w/793bAFwqYNykbiKrUZbxoT1kf54tJS5syL0ceaSPzYtZlVMhhltv\nrcv234EsbyfKorrIE+zixeLigxRtc85J2Kh5CklJzRyO+TssMlZ3ZIFo5JE+jjyhKH+zKqf/A/4G\nzrb9becU4IuoSCSEjJlzEgqCYC5MqZy01oeUUi8ATyuldgO/ALcDLYHXoiqc4BNvqYgk0EEQBH8w\npXIC0Fo/qpT6C3gBSAXWAJdqrddHVzLBF2auFCwIQvXAtMoJQGv9LPBstOUQAkdSEQmCEApmDSUX\nBEEQTmBMbTkJ1RezVgoWBKF6IMpJCDsSlScIQqiIchLCjkTlCYIQKjLnJAiCIJgOUU6CIAiC6RDl\nJPhFcXEJw4bNpUePPPr3f5e9e0uiLZIgCDGMzDkJfpGTs4z58wcBFtassXLkyCymTu0TbbEEQYhR\nxHIS/KKgwF7FBMBi+ywIghAZRDkJfpGRsQ8jUTyAlYyM/dEURxCEGEfceoJf5OZ2A2ZRUNCQNm0O\n8dRTUhBQEITIIcpJ8IukpMTyOSapgyMIQqQRt54gCIJgOsRyOoHxVndJEAQhmohyOoFxV3dpw8NZ\nTJp0PUlJiVGWThCEExnTu/WUUucqpY4ppS6KtiyxiL3uUmqrVJIzksnLu4b77lsWbbEEQTjBCchy\nUkoNAvZorRcrpdoDs4CTgQ+AO7XWR8MpnFKqnu07TK9EqwN2N15JyQEmTPiWTZsOMuA916NkDZMg\nCNHHb+WklLobyAUeBRYDr2GUT38TGALsBe4Ps3wvAFsAqbcQBhzdeJ3GQZPV+RQVVLjvjBpMsoZJ\nEIToE4jlNBTI1Vo/rZTKBDoDt2utX1NK/QE8RBiVk1LqSuAK279fwtXuiY5j+fSy0jKe6XwxrVv/\nQosWZRQWptOr12pyc7tHWUpBEE50AlFOWRgWE8BVGOkCPrF9/gPDigoLSqkmwDRgMCAZRiNEXHwc\nkMkpp+ySPHmCIJiKQOZy/qRCAV0J/KG13mb73B7YGUa5JgPztNZfhLFNAcN1t3vDbnZv2E1RQRHd\nu88nN1eyPQiCYC4CsZw+AZ5VSl2C4Wp7CEApNRpjHuqNcAiklBoMdMRQeFCRbVQIkOLiEnJyllFQ\n0JCMjH2MG9elcvn0t2VdkyAI5sNitVp9HwUopeoALwIXAcuAu7TWx2zzTd8Bt2itD4UqkFJqKXAe\ncMwuI1APOATM1Frf5uV0/y7mBKF//9m8//4AjC60ct117zFnzsBoiyUIwolD0MaF35aT1vowcIub\nXR201keCFcANNwB1HT43A74Cbga+9HXyiZrzzV22B61r4VjmYt26umHpH8mtF3mkj0PDn+wn0seR\nJyUlIehzvSonpdR5/jSilAJAa/3foCWxobXe4dK2XfFt11rvCbX9WMVdtocmTbIwjEnDcpIQceFE\nwd3zMIThZGe3jrJkgr/4spy+prKrzOJlW6QmL8Rd5weOYeIAfe49lwYNZtnmnPZL4INwQuH6PJTP\ntQrVAl/KKeqjmda6kMgpvZimUaMECREXBKFa4lU5aa1XVJUgQugYGR4c/m4eRWEEIcrI81C9CTS3\n3j+BLoDjTHscUB84T2udEV7xBEe8TfJmZrasHCaeKVmfhBMTeR6qP4Hk1nsEeALYZzvvmO1fClAG\nTI2EgEIF3iZ54+PjZbJXEGzI81D9CSRDxBDgLaAxMBH4RGvdFDgbKAJ+Dbt0AmBYTPn569mypaB8\nW5OsJuVKShAEIdYIRDk1B97RWluB/8NYKIvW+gfgaYzEsEIEsFtMP2X9j+SMZIoKitizSaLqBUGI\nXQKZc/oLw30HsAHIUkrVtWWFWIORGFaIEK5hseWTvTLJKwhCDBKI5fQ9MMj29zrgONDN9lkB4cwS\nIfigW9llDGk+XCZ5BUGISQKxnMYBS5RSSVrr3kqpt4G3lFJfYmQpnxcRCQWgclhsixYZMuErCELM\nEkhuveVKqXOB02yb7sBw850PfAiMDr94AkhYrCAIJx5+ZyWvJlglkWPkkYSZkUf6OPJIH0eelJSE\nyGclV0pd7+sYrfW7wQoiCIIgCHYCmXN628N2K4bD6TggykkQqgn+lJUQhGgRiHJyFyreALgQuB+4\nJiwSCYKJiOUBPBxlJWK5f4ToEkhARIGHXb8qpWoBL2MoKkGIGWK9LlCoZSVivX+E6BFQ4lcv/Aw8\nE6a2Ypri4hJycpbZaiztIze3G0lJidEWS/CC1AXyjvSPEAlCVk5KqZoYJdR3hS5OeZupwHjgUoyS\n7d8Cd2utq33+vpycZcyfPwiwsGaNFZglNZcEn0TqpUbKSghmJZBovfVUrkgbD6QC9YB7wiGQUsqC\nsaDXClyNkTbpCSBPKdVOa703HN8TLQoKGlJRbcRi+yyYGTMM4JF4qQnX+jkz9I8QewRiOa2isnKy\nAvuBhVrrL8MkUwegE9BOa70OQCk1CCgGrsJz1GC1ICNjn21wMSrbZ2Tsj7ZI1ZpIT8ibZQF0JF5q\nwlFWIpj+Ede24A+BBEQMiaAcjmwBetoVkw17wtmkKpIhYuTmdgNm2R7M/eTmdo22SNWaSE/IBzOA\nR2LwjeZLjbfrCaZ/xLUt+INX5aSUOi+QxrTW/w1NHNBaFwOLXTbfBdQBloTafrRJSkqUBzHMmG1C\nPhKDbzRfasJ9PbHi2hYLMLL4spy+psKVZ8HZrWe/uxy3hX1xg1KqFzAWeE5rrcPdviCEm0gMvtF8\nqQn39cSKa1sswMjiSzk5vp5lAK8D04H3gR1AMtALuBUYHm7hlFJDbN/5rtY6J9ztC7GB2SbkY2Xw\ntRPu64kV13asWIBmxe/Er0qpZcA3WusH3Ox7DLhSa90pXIIppR4CngJe0lqP9PO0mMpiK/jGKGGf\n77QtOzs7qhkKiotLGDFiMZs2NSAr6yCvvXYFjRtXX3dPrF1PuOjf/13ef38gdqV93XXvMWfOwGiL\nZTaCTvwaiHL6G+jlLipPKXUpMF9rXS9YQVzauw+jftQjWuuxAZwqWcmrAH+yOUtam9CQjNmRJ9Q+\n3ru3hPvuW+ZkAVbVnFN1me+qkqzkwFbgcsBdyHhfjNLtIaOUag88jeE+fEMp1dRh9wGt9d/h+B4h\nskhaGyHWieY84Ikw3xWIcnoeeE0plQYsAP4EmgL9MJK+DgiTTP0xysf/2/bPkUcwgiOEaoDZougE\nIVY4Eea7AlnnNMWWquhhwLG201bgBq31B+EQSGv9EPBQONoShFjEbC4ds8lzIhBrQTfuCCi3ntb6\nFeAVpVRbjAWxe7TW6yMimVDtiWYUXXFxCSNHfs4338QBe+jcuQETJ15d7QfN4uISunWbxfbtpwEH\nWbOmF/BJ1Fw6FfI8QCy7mMxGrEQ8eiOoxK9a6z/CLYgQW0Q77U9OzjI+++xm7G+WixfPplatZVEZ\nNMNpWeTkLCtXBEZw6ntRdekY8pxGKC6mWLS8In1NJ8Jifl8ZIo4C52utv1dKHcN7qLZVa107rNIJ\n1ZZw5G0LBVefPCRQUFB1k16O0YoPP/wleXm9gVasWRNHKJZF5euqT0ZGSRgk9o27CMzNm+sDBzGG\nhuBcTLE4uR+L11TV+LKcngYKHf6WdURCtcDVJw8HyMg4XmXf7xit2GkctCp4gSn9hgJnhWTpuF5X\nevpacnMHeTz+6NGjrFq10mnb+edfRK1atQL+bncRmE2aZGGsv38PqO9THjuOlsXmzTuBfUAisTK5\nfyIELEQar8pJa/2Ew9+PR1waQQgCdy6U3NxuHD36BqtXxwFFdO7cgNzcnlUqV6VoRVYCZ/ptWXi6\nLue5hkFe3UWrVq1kadznTgqFVdC16yVhuaY+955Lgwaf2OQp8SmPHUfLwnh5mI0RZ+W/5WVmd+CJ\nELAQaQKac1JKZQF1tNa/K6UaAU8CJwMfaK1nR0JAITYJ9zyMOxfKzJn9wypzqNSufZzLL5/l9+S1\np+sKxD20f/9fJHdwUZKbApO7uLiEUaMW8fXXG7kzr6bTvkaNEvyWx9laOo6jZZGYeJjMzHkBTe5H\n03Xma5F5dQ9YMIPiD6TY4BXAfOAljMKCUzAW3/4CvK2Uqqu1nh4RKYWYI5wDixlcKK4P84gRmRQd\ncI5WvOCCRgFdYziua/r0n+j6YuAuPEdycpaxeHFD4GaKCl4o3x5oBKaztfQOjvNUXbrUYOrU7m7P\n86QIovW7l5aWsnLlskoWqeMi8+oesBDK8+n4e6WmntnGarWu83GKWwKxnB4BPgeeUEolAn2AcVrr\nR5VSY4CRGFkdBMEn4RxYzOBCcX2YrdaZ3HbbQMY/8C2FhfU56aRGvPzydQG1Gch1OQ4IJSUHmDDh\nW/bsyWYyhLfsAAAgAElEQVTDhlLau4b0+4FjKP7+/Tswwi4vY0q/0QC0bfsFM2b0DSgC0/k3v4rE\nxAlkZrb2aVl4yjZS1b+7vY+3bCng451zyO6cHbOLzEN5Ph1/rwe/e1A7NBQQgSinDhi59Q4opQba\nzv3Qtu8L4O5gBBBOTIIdWPybh6l6F4rrw7xlSyJnnHEGs2efEXSbgVyX0wCeDK1uLSKv3zXAQab0\na4q9nzt3zuORDy7y+r3FxSVcfPFMdu58COc5oc+xzwsp9Y3PaEzXtWZ16hwCetvabESXLk09Wkuu\nuMs2UtW/u72PiYPEdHPMbYWKJ6s0lOfz4Ye/pNM41/nWwAlEOR2iol7TZcAurfXPts9pQNXEswox\nQbADSzjmYSJBOJWt3bfvj2vIPrisXfubEewGNMlq4nBEbxITpzpYKP/xGamXk7OMnTvb4xqKHxe3\nnYYNp9O5cxm5uZd5lQecQ+iNjGTTSE8fR2rqKX795vn5BfTtu4Ciohrc/VXl/b76J9zJh7ds2U5y\nVoX1ZrZSLcHgySoN5fnMy+tNJ94KWbZAlNMq4F6lVGPgn8CbAEqpM4HHADe3jyC4J1ifvFnml0aN\nWsTq1QeAZDp3LuXxxzsTTmXrL+WDS4dkkkl247bzz0KpHNpdhuOcEBzg6qvr+pTNfQj9aKANkEZq\nagpLlvhnLfXtu8C24Hg9RQXPl2/3VxG4G3j77BvIGWcEbs0WF5cwaNCn3PN1EgBlpWXExcdRVFBE\nyfYSrk3rX6WLzEPBUWnbX2qaZDUhLj7OdkDoz6fTfXh2cHIGopxGAYuAd4HfgDG27YuAA8D9wYkg\nCP5jlvklI0DgFsDC4sVWatXyX6l4i1oLRtm6urzsA0Pbtl+g1DceFaUvKwfewlBMhTRtepjcXP/q\niVYOoYdA1prZ+2fHjjSMvslmSr/R1K49j+XLLwko24irLOMf+DZgV6s9RdPRo80pKthCUvMk4uLj\nypVSi3My/LLIzBABZw/m+HjnHMM1mWgo2j2b9oTshgP789nS9kJiBcZ++nSQq2MDSfy6USl1CpCq\ntd7lsOtqYI3W+lhwIgiC/0Rzfsk+uCxZAnCUYJWKt6i1cCjbKf2uoXv3n1m27AavA6Y3K6dhw3ji\n4kox1og1ZuLEnkEPpAkJHxIfX9OvtWbOufpmY/RNPNCa5OTjIWcdKSys7/ex9t97xYrjlJRkYMzf\n9cX+W6WmTmHiT139dhOaIWvE5s0bWRr3OdmdswHjRcZuAdo/h+KeNJ7Pd8ufz/nzv7wx2LYCTfxq\nVUrtU0pdBKRjzJDuqA6KyQxvLULoVGWIruuchWFdjMZ4bPxTKu6i6FatSiWYqDXwL2S9e/efmTRp\noM9Bs6TkgEcrp2vXGkH3s+tczJIlff1WKs65+q4E3sNiOUSzZjuZO7dXyLKcdFIjv8+tvFD4LeD/\ngEPUqbOJTz75d0DWUjis5HDgztKe0q8PkEn37vPJfDt496Tr8zlv3r/2BttWoItwb8conZ6I8Wud\nDTyllKoN9NZa/xWsIJHGDG8tgmfMWDnXdc6i1a1F5OVtxJg/uYq4uHE0bJjmNUDAbRRdXjwVis3/\nqDUjiu5Ndu48GVjPmjWNOHJkLY895pJg923/XEyDB89n6MK6Ttt9uQJ9EWrCX2PAtufqSwQG0KvX\nLKZOvTMoWfrs8y+c33PeQMegkMZAT9LTx7Fs2fAgMmGE30r2hrtrKi0tqwhrs1GyvYSEhB/o1m0X\nubm+X2qqikAW4f4bYwHui8AnQJ5t1xvADOAJjMW5YUEpFYeRz28wkAB8Btyutd7tz/muP4zWBzAm\neeORXFfmIxyVc8Oh4Bzb2LKlgOQsV8vC7kBvxNVXN/frBaeyddIFeI+6dY/RowcBBU/s3PkwjuHd\n33xTGFAfVbiqdlFSckulRbUzZni3cuz9Y7cCCwvr06ZNPcaPv4SkpMSAE/7aI/L27m1OUtJW2rVr\nAPyLUHL1OXpG/Ann97So1sgbWKFMEhP/oEuXvX6laHJ2AQdnJYeKu2eqW9llFMU5W5Nb3juZb7+9\nniZNkiMqT6AEYjndCzyntb5PKVX+tGutP1ZKnYSxzilsyglD2Q0CbgSKgdcw1lV5X6Rhw/WH6fJU\nEX/02wAoJNeVOQm1cm44FJxjG0VxRSTj/MB27z6foqI2XgcXVwVXFFfkHA1FQ2AAPXr4Z727H+iM\n8G4IbECpeJNfCLQqX1QLeXTv3sinS2fNmp+Yd3B2+TxVUcFGpvQbhcXifzkSx/7p3fttdu++BWjF\noUNxlJWNoXdv37n6vLlcA/WMbN680e2iWiNvoOP85gCPSsmzPO8TrJUcjmkI12fqpKPNGRLvYtl+\nHF0PhScCUU5ZwBIP+37BWOsUFmwVd/8D3KG1XmrbNgDYpJQ6V2v9jT/tuP4w/gwsgn+E6+GxL9Rc\ntWo3d+b5Pt4bbudQAlBw7hYQus5ZjBnje/7ESUlmAQWgl2tD4RUUBew6q1Ao9gCBivDuzp3LvJ7r\naul8/TUYHoQDGFF5bTCym3/IpEmDvA5SFa5AV0swME+Eo4IburAuRQ6BGPv2ZfilVLy7XAOTp6Tk\ngNtFtYHkDXQvzxrs82bBWMmRmIaIdimbQAhEOW0DzgG+dLPvH7b94aIj0ABYYd+gtS5QSm0GLgT8\nUk6ujBlzSbX5YcxMUVH4qrFWFAUMbi0LVCjKpUvXVkpM6i8VUWJJdGIfYKz9KCooosOms2jRIsPr\n/ImjEnjyyYV0fdF5AO+w6SxalBptPL4ssDfVirVdxkAXF3eIhg2LbNV9vUe/VY7IK2JKvw22tmaT\nmHiYLl1q+HRV5ecX0LXrHA4fbgK4Ti0H5omYMOFbDxkErCQl+T+MVHaXWsv/D2QR9ODB8+nzauiV\nmyvLswI4i0CsZDuhruezv2i1urX6LhQORDlNBx5RSv2N4RMAqKuU6gU8hDEfFS7sXVjosn07RhZ0\nt/TvP5t16+p6jGKqTj+MmbnttsVhq8Za8RBm296el9K2banXvG2lpaXk56+nsNAYyCZO/C+rV98E\nHKSo4Ofy43z95o6umPvu+5Tt2/sBZRQVvOh0XIsWGR5falzzrSWmJ3L6SCvQxO823Mmzd28DiosP\nuqSSMQIErr56FlOn3uq1LTAUSu/eb7uxdBZRt24T25v8pX5ZvH37LuDw4Sdx9xKRmjqF3NxRXs93\nXLi8b99xOrnsr117HsnJx4OKyLPjj2fEXt9q+/bdjBnzLUVFDYB0yko3Oi2q3fLeyWR+7DuQw/5i\npPUBrnVJipCQcITs7MAyrdsJdT2fkalhtM2atNK9+3zGjAksF2K0CUQ5jQMygOds/8AoUGPB8Dc8\nHUa56gFlWmtXp8wRoI6nk95/fwCOiTcfeCB6ZcJjmU2bGhCuaqwVD6GxlgX+h1Le17Ns3ryRl394\njsT0RJIzkun6Yi3aF7zElH7ZTOk3CvsD7Sss1tGquPA5OKXgBab0u8umJPPo3v2Azwe6vI2sZLKz\nsikqKMISZwnqLdzJNbQPirYZGQ2CXdvVt+8C23yOayqZq+jR49uA3uT37m2O80vEh9SoUch556Wx\nfPkQnwru9ts/JC+vGOM3Xk1RQcXKzKKCIpYv9+wurXgB2M6ddy5l3740GjYs5NpJR5za8OZytb/Q\nLFr0CRvT1pHYPJF/Tq5FWelhpg5oz9QBtYCewDxSU3eyfPntflm3Fe63ykr7nHMaMXu2f9kwXAl1\nPZ/x0lcDw80JRUVtqp3XyGK1BrZ8VynVGuiGMRO7D0NB/QrcqrWeFA6hlFLXAh8ANbXWZQ7bvwa+\n11q7f02zWKRSryAIglmwWoPKSA5+WE5KqcuBIdhWoWmtFwPrHfZfCPwAtAfCopyArbb/m+Hs2kun\nsqtPEARBiDG8Kiel1A3ALIxcLUeA65RS/9Raz7UlgH0JGAgcp8LVFw5+wliJ1wUjlx9KqUwgE8NS\nc8uVVzzHjh0ty81gyQARGVJSEvjzzwNR+/78/PVM/C633K0HtrDx5v6HjQ8bNpf58ztxywcvOLWx\n4pHTUSrB7/snP399pfD1n59PYO/eTNq0qc/48d09hkIvXLiQL/9eWB4pVlZaxv5d+51Sy2yY3JLZ\ns0d4lcFxPmf//oaUldUGrgH2UbPmG9So0YqkpG3MnduLrKwMv69jSr/WwM1AIh07zvM7Yau7tgCH\nBLALqF37O37++Ra/QrO3bCngp6z/lc+b7d6wm7Hn/Iv09I9Ys8a/xbkDB75Gq1s3OgUt7N6wm/zV\n+Xx8Txxff/2gx77xxt69Jdx33zIn95s/903Hji87zdump4/z61o81e3KzDxoyqw3KSGc68tyGgl8\ni1Ei4zBGJvJHlFJrMaL2TsZYHDtSax1UtUN3aK2PKqUmAROUUkXAn8CrwDKt9XeezjPyfzULlxhC\niLgusJw7t3dQA4ArmZktubP0biMgwlZy/KSTTvY5p1g5lUzldT7e8tG5W+R78skZlTMi+LFuZPPm\njfwv8WuyOzjnOJtz100Y72CT6N69pc8Chfn5BVx88XscOZKIscywIkgFBnDlla38mluyl4NwDpzo\nCSwGBgQ8Ie8pGS1YqVPne1as8LxmCJzDzYviiqCgYp89CGPu3H/7LU9hYX2abHeeFy0qKELPqM//\n/d+DNG0aXNLTYNNpVczhAVhsn33jqW7XTz+1xt9wczNmY3GHL+XUBhimtd4PoJR6AiMj+TygNtBP\na/1RhGR72CbfLKAmxlNyR4S+S4gAFSUPLBw6ZKVPH//eDsH7AxQfH0+bNm1p06ZtQPJUTiXj3zof\ndxF5Sc2T2Lttb0AlGHyFmxsD+E/AMq64IpuZM2/y2l5xcQldu87hyJGnMAJoKwY7Y13NLL8n0u+8\nc2mlVEbBtOMJI0XOh1xwQSoTJ3pO/WO3Ar/44hfu+6/RP02ymqCXa6eQ/sd/CiwL+L59xcy5ayTG\nNLmVxo3nsGjRrTy+WFVqpyoG76SkrRw6VBGNF1oIPQQSbh6OxepVgS/l1ICK+R+AzRi9eRxo728q\noWCwRerda/vnF9nZ2RQX/x0pkYQACfbtEEJ/gNxFeB0/nuAgj2sqGc/rfMrf4h0i8uLi40jOSPar\nBIO/4eYl20uADVx4YQvefvtflHpZQGxfl3X4cGuMYNmjOC7Q7dEDn4X41qz5qTwF0e7dxfy5saII\noaEofbcD8MMPv9Cnz2KOHMmidu2NTJp0GkUJztGK16b1Z+I67xm8K3IHpgJNMWYSKO/rFqW+w/Ht\n7VRkE68NXAz0dil0+CRt2pzs1j3t7t7b8HAWkyZdHza32dy5venTZ5zNq7AtpBB6A+/h5j7Tcpmw\nxLwv5WTBWWx7MZaHI6mYgsVsZumJTihvhxB8OiPXmjV9XoWy0g1MHWBYSYGmkhk79isufM592Wlf\nJRic8rb5CDf/6G4rv/32IE2aJNO4sfO8nmumh+++28OBA2MxHmErRorL2cAB0tJ2k5s7xK0s9gHq\nm2++55u4ZZw9JotO8fsoKqjDlH5HMeaqvgKa07TpHHJzvVtvAH36LLatgbJw+LCVESMeZvnyyss4\nfD2fzrkDJ1BUsN2pfzyF43vPHm93cQ4kNfUUv+fNXO+9KXnduO8+/1M0+SIrK8NvL4IrrveNPxlH\nfKXlMiMBZSV3QCLmBJ+E/+3QPzzVrIGLiYsbR/v2/pUJt7NmTRwXumyzDxDeSjCUlpayfPmXzFo7\ng469OzoNdmWlZezdtpex5xzEYkkmJaUeK1eO8Jh8s3Kmh/1M6bcGIwOBBYjDYsmnW7dGTJrkft2R\nk0VwPiQWJBIXH1cuV3x8TazWL4iLS+bii628+upNXt1vdpfZ4cO1MdxliYCFo0ezg0pG65w78Cam\n9MsFGpGaupO33hrocU7RVyojwwEUaj7NrygoiP46SXdZ3/3NOOKocKtDiXl/lJO7tUOynqiaUxX1\nrUJ5O4TQHiD3E/IJpKXh9e3ZnkFg//6/mD79J3btqstff+2iqMDi1NYXOalkZTXwGrCwefNGZq2d\nQUJqQqV9e7fZy9wc4o8/3CsBx0wY27dvL7/+Jll2d6A9PY6V9PSdLFvmezGs+/kKg4svruczMtCO\n8/xdbwyr7XrASu3am/xqo3JbjrkDGwMdSEvbyIoVo4K4ropURhXZxP2fN6t079E8oJRIkXq2wpEb\nL5C0XHaiUQ/PH+X0slLK/qvYn9BJSilXZ61Va+2+qI1gOsxe3yrUukCulGwv8SvCa9WqlYYLroOR\neaKoYDdT+vViSr+lGMGpf3HRRTVZtWqkz7fVLVu2k5CaQGJ6YqXBbkq/elgsfzFnzjUeH3KnTBjn\nJ5NMslM7cIzExOl+5cbzhGMF1Hvv7eHxOF8laCyWfcB0atfexIIFV/j8XufIyZ0YlteVwFvExR2m\nYcP9ttyBg71eV0WyXuftCQlvkJ19ns9s4u7IzGzJhoezmJJ3DXaXdLNmH/is4msn1GcrUorA+d7x\nL6WWnWiMF76U00qMVxDHbJr2ZKzBZdgUTEGoiSUjTahviK7K4Nq0/j5LahcXlzBx4n8rRdFBDRIT\nTyUzs7FtsLvOZzs5OctYtOhH+j4HSc2TystlFBUU0Xbb6ezY4b2EOlRky/YUlp2aeoivvrohoIHL\nnZJMTT3Eq6/2pEOHDh7P81WCplev+j4HK0cFZ8wL9QZaYURN2i2vf9lyB17v81ock/W2KqhQnEbq\noJZ+pw5ypwwmTbreZf2SdyXpSKjPViSUW3iKQFbteOFVOWmtL464BEJUCDWxpJlwfRjHjevi9kH0\nZ0J+9erudOUrlz1WunSp4VfwhL0dY3D5m7LSQvZu21s+qAOcfXYnv+YIJkz4llZu8rtO6VeP7t3T\nWL78upAtgl69VjNtWl+v57krJQKQmjqF9PSLfM7fuYtWPHtMEq1urSiVkZh4mMzMwJKkGiXdHwDK\nbJnWF1G79nEuuMBzxVtP7bhTBsFaBqE+W5FSbqG87EVjvAg2IEKo5oSaWNIMuAsZXrOmEcG6HOyD\ngqt10blzHrm5/wm4HWjAB6Mfx1iNYaytWbDgdrKzW7k9zzUr+aZNB2m8tdjpGGMBqv9zQ47Ex8e7\nsQh8K1wjw3VvOrkkkG3cONuv6Dd3yXHt4eEGgSl/OxX9HI9hwWkuv7zE7wKOd9yxkHXr6toWZIfP\nKgj12Yq2cnNHNMYLUU4xjif/daAr2x3badPmb5566sKop0pxnpSvCBkO9mE0BoUrmNIPwErnznmM\nHHkZj3xwEbVq1fJ4nr1g4jffxAF7qF17P0aQQE/gU+LiDpCWVsTcuXd7zZDhmBWBfXDps0VM6TcY\nKAKWA4mkpNRh/vzBQV0fBJfRwJPS9hapaCc/v4Crr36L4Z+6D8SoXXsel1+e5tdg53ovN2t21GkQ\n97ekO7hbkB0+qyDYrBF2oq3c3BHqNQWDKKcYJxT/tac5gjVr4jhypGoDKFxrOAFovQ/HSflQQ4aN\nQWGOw6DwH78U8MiRn9sKJtqV5BsOCz6Pk5t7pc9JfSMrgi7PimAnNTXP5jq72K+8beGYTHdNO9Wu\nXQOgJ1P6DQVWkJBwhHPO8e06s2ewOHx4BK5lO+yK7oILGvlcLOzpHrziisn07u04iFcOCvGU7cHZ\nunBdkB1dL0K0lZtZEOUUozivHZmNEQmVGHQ5bWNtTcUcQSDthGPAdK3hBJUn5YMJGXYkWGtyyZJS\nnOtb1QhowWdOzjIWL24I3IrrIO6v68wujxEgYKSMCuRlxHEQ7937bVsdqFYcOhRHWdkYevd+1/b7\npfid3DQnZxmHD5+NO6vLn1B88H4P7tjRzGffeMr20KxZYwfrwv8F2dWBaFg5kUCUU4zi3uUVWAJP\nb+W0A2knXGGo7iLXnKufug8Zds2uUFhYnzZt6jF+/CUhuSYrrmsCjm4h2ENGRpzf7RiK3rD+gnGd\nOcqzfftpBDPf4DiID11YlyIHJbBvX0aQShuM4gItbW1Zgdfp3r0lq1YN9zsoJNR7sHK2h2u44oql\nXHfde7bK2f5ZF9FY63MiI8opRnGdFA0mgaeRmmef07a2bb+gffsfeOop/9uJZBjqmDGX+IxCcs2u\nUFSwkSn9RmGxBJeOprJVWt/2fwJwgKZN95Gb29/v9ow5gho4DuI1arzMeeelMXmy/1FnRr8eJNj5\nE/cLdANPO+W8qPYKjLqh9bFYvqFbt3QmTfIeiu+Ip3vQV7oe71jYsaMZn356TUClX8y+NjDWEOUU\no7hOinpL4OnJL3/SSQcrvcnPmNGXc889I6CHOtgJWsc31eRkTZNelUse+Js1ovLAG5ySdHWdVeS1\ng8TEP20LYj1n3XZHbm43jh5dyOrVLwDJdO5cxttvj6G01P/SHZmZLW393AvDSq7vNUDAtY0tWwog\ny/mY2rXnkZx8POC0UxUvI1cCi20vRiXk5o4I2NLwdA8GEhZdOdtDcPOSZl8bGGuIcopRApkU9ZQB\n/OWX+3Pnne+zorA+J530F/fe2yOoLA3BTtA6uyaP0806gWtvUQHVcPJMcAOUO9dZ3bq16NHjOLm5\nlwbl5klKSmTmzBudtrkmfnXE0+9l9PMntn4u8Zo1wrWN/J35JMZVHFtUUMTy5d6VgHclacXItTeA\nHj2CtzBCvQfti0/3FR9g/PhvKSxMp1ev1X6F0LsSzEuWuAKDR5RTjBLopKi7DOBNmiQHtJ4mXGHr\ndpzfVGtQXNyWrl2Dm7R2fXvu3n0+ubkDfZ7nek35+TVwdZ35U1Yi3Lj7vfzpZ8cFsY5lE8pKy/jH\nlnNoUWoLdfcjg4B3Jen7ZcSfukn+3oPe2rIrWF+lTXwRzEuWuAKDR5STEDbCFbZup0WLvWFZr+E2\ndcvbvjNG/PDDL1x11ULKytoCv7NmzUDS098FRuCP68yM2BUKcTiVTYiLjwso15qdYJWkoyzhKHpX\nFQX0QlkjZiCuwEAwpXJSSp0BPIuRcvlv4FPgPq31Xq8nCkETjhT6oTyITgtQbTLcdttALJbQ12sE\nkqfPsQhfXt4mYCQV+d+eJTm5DWef7Z/rLJKEmrE91DbCRbA1uyLdVriIpTRhVY3plJNSqhnwBUaI\nzx0YufMnA3MAz2mThaAJVwbwUAIfBg+ez9CFzoNLo9KEKneBbN68kXkHZ9NpXDKdaOwUUg3ptGz5\nd9TdMuH4vexlN4oKiuhWdhktWmQEXDZhxIhMig5ET8G5Bsy4ZiY3A7GyIDYamE45Af2BQ8AIrbUV\nQCl1O7BCKdVcax1YXKvgk3DUiAH/H0R3lUt37x6KkUYm+ngKqY6L+4PcXN/zH5GeBA9nxnbwv3SC\nq9vWap3JAw+EpiRDsd6cA2Y0rQpeCLqtSBErC2KjgRmV03zge7tismH/OwkQ5WRSgp1rMCqXbo2a\nm8n3G/gH1KlzmAULevpMQzRy5Ofk5e3i2LEHMeMkeChWl6vbdsuWxJCUZHjLOLRixSOnM2NG56Da\nEsyH6ZST1noT5cHC5eRglIZfW/USVX/MFM7qqfwCnFy+ADU1dYrXstzhlsd53VLlN/DPPuvBGWf4\njvQaNWoRn33WEKMooTknwQOxunwlWg11/iRUC9DZjRyHUgkBXZs9K3m0nwnBPVWunJRSGRjKx35X\nOXJYa13P5fhnMFbz9XaxpgQ/MUs4a4Ui6Fup/EL37gsd0hD5LssdLiqvWwr+DXz16gPALTiXG6++\nk+Cu903lRKvRnT8JZT7HLM+E4JloWE6FQFsP+8rsfyil4oBXgWHArVrrRf40npKSELKAscb27Uk4\nvslv354Ucj8Fc/4ddywsVwSuLrxJkwbTpk2bkGQKBqNvDlChTOJo374x554b+JoYiyWFiswI7wGH\naN58N9On30rjxoH3V7TvZdf7Zs+e5nz33dXRFMmJlJQE5s37V1DnRuKZEMJLlSsnrfVxYJ23Y5RS\ntTGi9XoAN2it5/jbfiBpdU4U0tOLcXyTT0/fG1I/paR4zl7gjXXr6uKaCLR79/mMGdOXhg1To/Lb\nGX3jnPLnqacGBSXLueceY/HiiswIaWljyMsbQmlpfMDtBdvH4STc942ZiOVrMxOhKHzTzTkppSzA\nh8DFQE+t9ZfRlaj6Y5Zw1orcb0Yi0PT0tUya5HmdkGNUnz2j+J492WRmHgzbHEEgKX98MXHi5dSq\n5djPQ6rEPRmpOUWz3DeRIDe3G7VrB5aVXKhaLFaruaZxbGHjLwM3Yyy+daTIZnl5wipvP5En2Lf6\nvXtLXEqEe68LlJ+/vtKqf8Piak3v3rE9RxBIHw8bNtepPEqs9024MIN1GuukpCS4xhX4jeksJ+B6\nDHt7msM2e/rnC4H/RkMoIXQCXfNRUnLAw5ojc0XABUq4LR1JkSPEIqZTTlrr86Mtg2AOjEJz7vZU\n3wg4CH+kmKTIEWIR0yknQbBTWFifooKKTBJFBUVhKDQXHRytpc2bd2IU0EskHJZOLM8NCScuopwE\n09KmTT2m9BuF3SLo3n0+y5bd4HcV1VAIt+vNOdWOFWMtlOHBDtXSkRQ5QiwiykkwLePHX4LF4hhA\nMdCtYvKnLlCghNv15jovlJh4mMzMeWLpCIIHRDkJpsXf4nkrVy7j451zSEw3LJuy0jL+za0hpcYJ\nd5CB67xQly41mDo1uMKJgnAiIMpJqNZs3ryRpXGfk905GzDmpeLi40JuN9xBBjIvJAiBIcpJqPa4\nhpu7loTwhbv5pXArE5kXEoTAEOUkxBwl20uCrgvkOL8kykQQoocoJ6Ha45pE9tq0/iHUBZJFrIJg\nBkQ5CdUaTwXrAonUk0WsgmA+RDkJ1ZpwlJiXYAVBMB+inIQTHglWEATzEXrMrSAIgiCEGVFOgiAI\ngukQt54QNiJV9E4QhBMPUU5C2Ah3PjpBEE5cTO/WU0rdq5Qqi7Ycgm9kvZAgCOHC1MpJKdUeeBKj\nxoBgcjIy9lHxU8l6IUEQgse0bj2lVE3gLYyy7BdHVxrBH2S9kCAI4cK0ygl4GtiGUZXt4uiKIviD\nrM51fx0AAA3ISURBVBcSBCFcmNKtp5S6CBgM3BxtWQRBEISqp8otJ6VUBrAJY3LC4rL7MNAUeBO4\nU2u9SylVtQIKgiAIUScabr1CoK2HfWXAS8D3Wuv3bdtcFZggCIIQ41isVnMFwtnCxg9hKCowFGht\n4CBwi9Z6tpfTzXUxgiAIJzZBGxdmDIho5fL5GmA80AHY7evkP/88EAmZBAdSUhKknyOM9HHkkT6O\nPCkpCUGfazrlpLXe6PhZKbXLtn1TdCQSBEEQqhpTRusJgiAIJzams5xc0Vq/A7wTbTkEQRCEqkMs\nJ0EQBMF0iHISBEEQTIcoJ0EQBMF0iHISBEEQTIcoJ0EQBMF0iHISBEEQTIcoJ0EQBMF0iHISBEEQ\nTIcoJ0EQBMF0iHISBEEQTIcoJ0EQBMF0iHISBEEQTIcoJ0EQBMF0iHISBEEQTIcoJ0EQBMF0mLKe\nk1KqJvAMcD1QH1gJ3KG13hxNuQRBEISqwayW0+vAP4EBwLlAXWBBVCUSBEEQqgzTKSelVBYwGPiX\n1nqF1vo3YASQoJRqGV3pBEEQhKrAjG69HsBurfUK+wat9TogK3oiCYIgCFWJGZVTG2CjUmogkAOk\nAKuAUVrrwqhKJgiCIFQJVa6clFIZwCbAClhcdh8G3gHaAaOBu4CjGMEReUqp9lrro1UoriAIghAF\nomE5FQJtPewrw1BKDYG+WustAEqpfwI7gCuBeVUhpCAIghA9LFarNdoyOKGUegjI0Vo3dNm+C3hG\na/1CdCQTBEEQqgrTResBXwH1lVLKvkEplQY0ATZETSpBEAShyjCd5QSglFoBJAC3AX8DE4FUoKPW\n+ng0ZRMEQRAijxktJ4Crgf8BCzEsqWKghygmQRCEEwNTWk6CIAjCiY1ZLSdBEAThBEaUkyAIgmA6\nzJghImCUUnHA0xg5+RKAz4Dbtda7oypYDKGUagf8ivPiaStwodb6v1ETLEZQSk0G4rTWwx229QCe\nBRSwDrhfa/1ZlESs9njo4++AsxwOswJvOB4jeEcplQqMBy7FSNL9LXC31vpX2/6g7uNYsZyeAAYB\nNwIXAs2BD6MqUexxOvAnkObwrxnGjSiEgFLqSWC4y7ZTgPnAHKAjRlb+ebaXBCFA3PWxjVOAgTjf\n06OrULRqjVLKgpEYoRVGIFtnYB9GRp+kUO7jam852Wo//Qej3tNS27YBwCal1Lla62+iKmDscBrw\nm9b6z2gLEivYMvC/AZwKFLjs/g+wWmv9jO3zo0qpCzBSet1adVJWb7z1sa3KQV3gG/GyBE0HoBPQ\nzpagG6XUIIwI66uACwjyPo4Fy6kj0ABwzGJeAGzGsKKE8HAa8Hu0hYgxzgO2YFilm132XQgsd9m2\nHLmnA8VbH58GHLKNF0JwbAF62hWTjTLb/0mEcB9Xe8sJw4UHRs4+R7YDJ1exLLHMaUAdpdRqIBNY\nCzyotf4+qlJVY7TW72AkOsYhIYqd5sg9HTI++vg0YJ9S6l2gC1AEzAAmaq1ljY0faK2LgcUum+8C\n6gBLgDEEeR/HguVUDyjTWpe6bD+C0UFCiCil6gAtMYJN7sHwLW8HVig3T7wQFuphZOl3RO7p8HIq\nUB9jcO0BvIIxf/1oNIWqziilegFjgee01poQ7uNYsJwOAXFKqTitdZnD9trAX1GSKabQWh9WSiUC\nR7TWxwCUUkOAMzFSTN0VRfFilUMY97Ajck+Hl0FAA631ftvnX233+YMYSkoIANuY8DrwrtY6x7Y5\n6Ps4Fiynrbb/m7lsT6eyOSkEidb6oF0x2T5bMULLxc0UGbYi93RE0VqXOSgmO78ACUqphu7OEdxj\nqyYxHZiktR7isCvo+zgWlNNPwEEMnzEASqlMjHmRldERKbZQSp2hlNqnlPqHw7Y4jGCUtdGTLKb5\nGod72kZX5J4OG0qp1UqpiS6bzwa2u1FaggeUUvcBTwIPa61HuuwO+j6u9m49rfVRpdQkYIJSqghj\nLc6rwDKt9XfRlS5m+AmjevEUpdQdGCZ5DpAMvBRNwWKYl4H/KaUeB2YDNwDnIGHk4eRj4Aml1A/A\nKoxB816MMH7BD5RS7TESIEwH3lBKNXXYfYAQ7uNYsJwAHsaIyJkF5GEMpP2iKlEMYQs2uQLQGIvo\nvsEoYXKh1npPNGWLIZyiw7TWa4E+QF/gR6AnRsiujoJssYJrH4/HmF96CMMDcC8wUms9IwqyVVf6\nY+iRf2MESTn+GxnKfSxZyQVBEATTESuWkyAIghBDiHISBEEQTIcoJ0EQBMF0iHISBEEQTIcoJ0EQ\nBMF0iHISBEEQTIcoJ0EQBMF0VPsMEULso5RaDlzksvkosA34CHhUa30kQt/9OPCQ1rqm7fMy4JjW\nuoef55+LkdalZxhkGYKxEr+51np7qO15+Z43gfO11q0j9R2C4AtRTkJ1wAp8B9wJWGzb6mDk7HoM\nI/ns9RH8bseV6iNcPvviZoxS4JGQJVJU1fcIgkdEOQnVhf1uChuuVEqdDPxbKTVKa70r0kJorf+I\n9HcIgiDKSaj+/B+GddIC2KWU2oTh6jsDOBd4Q2t9p1KqMfAs0AujaOIPQI7W+r/2hpRStYFxwECg\nAfA+sNvxy2wuxqN2t55SqiZGcbobgKbABuBprfX7SqkZwGDbcaXATVrrt2zFG58CBgBNgN+Bx7TW\nnzh8jwUj59sw2zFL8JHJWSmVB9TVWp/nsv1LIF5r3dWWTT4Hw9LMxiip/SOG63GFh3bLbPvHOmx7\nHAd3p21bH5vMpwLFGPkuH9ZaH7XtrwM8j1GsMgUjB+Y0rfVz3q5LODGRgAihumOvxJvvsO1OjFT9\nvYGZNqWzFCN5bQ5GEsq9QJ5S6kyH897BUHRjgH8CScBol+9zdXe9C4wEJmEktVwBzFZKXYlRRmAB\nsANDUS6ynTMXGIqhLHtjKId5SqmrHdodDzyCUbztGmAP8IyPvpgFdLJZkwAopVIx3J+zbJsmAA9g\nZO6/zCZHE+B9m/LwFyfXn1LqeoyXgp9t1/Q0cAu2Euk2XrR95yiMyrPzgFyl1KAAvlc4QRDLSagu\nWJRS8Q6fmwBXYgyA72utix325Wuty0ttK6WGAacBZ2utf7RtXqyU+h6jpPRlSqlTgWuB4Vrrabbz\nlmAUn2vjTiCl1GkYiu42rfVk2+ZlSqlsoKvW+lOl1J8YFYS/t51zKcYAfa3Wep7tnCVKqSQMhfSJ\nUqoRhoIdr7V+2nbMF0qpk2zneuIjDKUzwNYWGFmjjwEf2j6nAfc7yItS6oht/6kYFmUwPAMs0Fr/\n2+GatgLzlVKdtdarMYJavtBa22VZqZQ6iKF4BcEJUU5CdaE7xiDryHEMK+Q2l+1rXD53w6i8+bOD\ngrNgWDI5SqkawIUYlsAC+0laa6tS6kOMsgruuMB2zlzHjfr/2zu/EKnKMIz/BqEuLIR2wYuk6EJe\n9yK6EULoDwSFFVtXBmVFV5VQGtttLcFGSVFIEcFKhBhEBYK1VhZhQaRFGAqVb1BJCEaIpJul1Tpd\nPN+xs2dnplmvZtvnB8M5M+ecb97vwJxn3u97znkzb+3RjxtK3LsbYvsOcHtEXAaMoN/m241j36SH\nOGXmdETsZK44TVXF8zLzLoCIGEZZ50o0zAZwQY+4uxIRAaxAtZHqffoQuSpvBPYCe4AHS2b3LrCr\nJr7GzMLiZBYKnyMRaiFB+AM4nJmnO+z7W+P9EHL0NcWtGpoaRkN4oGKVdY72iOmSsvylxz5NhtDv\n7lSHbWdRCeuq3fnEUrEdmIqIlcAZYA0aZgMgIlajIcjVJYavgZ/K5hbnx1BZTgJbG9vaqE8Am1DZ\n7rtRkcoXI2IvsCEzD57nd5v/KRYns1CYrg3JzZcTwDfAPXS+AB/j36Gl5cDPtW1Dc3ef1S5ocv+c\nQJUhwqVdKjGfAH5FmWCnWBIZNlollh/7jKXiAyRq61DWchx4r8R1cVnfD4xUBd8i4mY0PNmLJY33\nF9XWq/PwCKoo2+QYQGb+hQwnT0fECpSxjSNBvaqPvplFhA0RZjHwCXA5cDQz91cvVKFzY2b+jQwT\nLeZWUL6tR7uflmNGG5+/gAwBADMdYlkGzDRiWYMMEG3gM5QZzicW4FzV4tfLvuuAN8pnAKuQwG1p\nVCK9pSy7XQ9OosyzzjW19W+RIF7R6NNx4BlgJCIujIhDETFW4jySmS+XWJttG+PMySwKXkUGg48i\n4ik0/zSK/uk/AZCZ30fEJLC5uNYOAPcCV3ZrNDMPRMQO4PmSlRxEgncdcqOBsqTlEbEWzYXtQuIz\nFRETwHfoQj8OvJaZvwOUbRMRcRr4mFLeus/+bgc2IqHbVA8ZCc3jmiZiBrkSKxPD0i7tTQHrI+JL\nZJW/D9nQq/NwNiIeA16KiDbKzobRDdLLgK8y80xEfAGMR8Sf5VytKm291We/zCLCmZNZKPT7xII5\nTzfIzFNIAPYBzyGBuAl4KDMnartuQP/0HwZ2oKdQPPkfsdyJ5nDGkIHhamA0M/eU7duAw8g2vT4z\n28Ba5I4bB95H90JNAPfXYt6MxPMOYCdy0jVt7R0pWcsh4IfM3Ff7/CTKqJYgQdgGXIrMINNl2amP\nY8iw8Ww5bhpZ8uvfuRXNJV1fzsMWJPDXZmY15PkA8ArwKLAb3RM1ic63MbNotdt+SokxxpjBwpmT\nMcaYgcPiZIwxZuCwOBljjBk4LE7GGGMGDouTMcaYgcPiZIwxZuCwOBljjBk4LE7GGGMGDouTMcaY\ngeMfmxWogP1xuUIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1df5f080>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(y_train_pred,  y_train_pred - y_train, c='blue', marker='o', label='Training data')\n",
    "plt.scatter(y_test_pred,  y_test_pred - y_test, c='lightgreen', marker='s', label='Test data')\n",
    "plt.xlabel('Predicted values')\n",
    "plt.ylabel('Residuals')\n",
    "plt.legend(loc='upper left')\n",
    "plt.hlines(y=0, xmin=0, xmax=20, lw=2, color='red')\n",
    "plt.xlim([0, 20])\n",
    "plt.tight_layout()\n",
    "\n",
    "# plt.savefig('./figures/slr_residuals.png', dpi=300)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE train: 2.486, test: 5.642\n",
      "R^2 train: 0.859, test: 0.798\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import r2_score\n",
    "from sklearn.metrics import mean_squared_error\n",
    "print('MSE train: %.3f, test: %.3f' % (\n",
    "        mean_squared_error(y_train, y_train_pred),\n",
    "        mean_squared_error(y_test, y_test_pred)))\n",
    "print('R^2 train: %.3f, test: %.3f' % (\n",
    "        r2_score(y_train, y_train_pred),\n",
    "        r2_score(y_test, y_test_pred)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Using regularized methods for regression\n",
    "\n",
    "A Lasso Regression model can be initialized as follows:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[-0.         -0.         -0.14097623 -0.         -0.          0.          0.\n",
      " -0.          0.         -0.          0.          0.          0.          0.\n",
      " -0.         -0.         -0.         -0.          0.          0.         -0.\n",
      "  0.          0.          0.02030755  0.          0.         -0.\n",
      "  0.05042061  0.01666828  0.03440757  0.09543018  0.98554785]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import Lasso\n",
    "lasso = Lasso(alpha=0.1)\n",
    "lasso.fit(X_train, y_train)\n",
    "y_train_pred = lasso.predict(X_train)\n",
    "y_test_pred = lasso.predict(X_test)\n",
    "print(lasso.coef_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE train: 2.666, test: 5.868\n",
      "R^2 train: 0.849, test: 0.790\n"
     ]
    }
   ],
   "source": [
    "print('MSE train: %.3f, test: %.3f' % (\n",
    "        mean_squared_error(y_train, y_train_pred),\n",
    "        mean_squared_error(y_test, y_test_pred)))\n",
    "print('R^2 train: %.3f, test: %.3f' % (\n",
    "        r2_score(y_train, y_train_pred),\n",
    "        r2_score(y_test, y_test_pred)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Similiarly Ridge regression can be used:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[-0.35325243 -0.20836897 -0.21409541 -0.14297151 -0.09381759  0.27131556\n",
      "  0.17368312 -0.12499116  0.21471346  0.15413121 -0.01400552  0.18185536\n",
      "  0.06190845  0.07503396  0.00234185 -0.44425912 -0.13290367 -0.10680339\n",
      "  0.30051497  0.32123389 -0.23413804  0.20521655  0.25409234  0.15749261\n",
      "  0.02279431  0.00097435 -0.19989632  0.1902062   0.07026242  0.04768856\n",
      "  0.13956252  0.95878554]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import Ridge\n",
    "ridge = Ridge(alpha=1.0)\n",
    "ridge.fit(X_train, y_train)\n",
    "y_train_pred = ridge.predict(X_train)\n",
    "y_test_pred = ridge.predict(X_test)\n",
    "print(ridge.coef_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lastly, the ElasticNet implementation allows us to vary the L1 to L2 ratio:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.         -0.         -0.         -0.         -0.          0.          0.\n",
      "  0.         -0.         -0.         -0.          0.          0.          0.\n",
      " -0.         -0.         -0.         -0.          0.          0.         -0.\n",
      " -0.          0.          0.          0.          0.         -0.          0.\n",
      "  0.          0.01991938  0.13861396  0.89133515]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import ElasticNet\n",
    "en = ElasticNet(alpha=1.0, l1_ratio=0.5)\n",
    "en.fit(X_train, y_train)\n",
    "y_train_pred = en.predict(X_train)\n",
    "y_test_pred = en.predict(X_test)\n",
    "print(en.coef_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For example, if we set l1_ratio to 1.0, the ElasticNet regressor would be\n",
    "equal to LASSO regression."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Decision tree regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X = df[features].values\n",
    "y = df['G3'].values\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE train: 2.519, test: 4.577\n",
      "R^2 train: 0.857, test: 0.836\n"
     ]
    }
   ],
   "source": [
    "from sklearn.tree import DecisionTreeRegressor\n",
    "\n",
    "\n",
    "tree = DecisionTreeRegressor(max_depth=3)\n",
    "tree.fit(X_train, y_train)\n",
    "y_train_pred = tree.predict(X_train)\n",
    "y_test_pred = tree.predict(X_test)\n",
    "\n",
    "print('MSE train: %.3f, test: %.3f' % (\n",
    "        mean_squared_error(y_train, y_train_pred),\n",
    "        mean_squared_error(y_test, y_test_pred)))\n",
    "print('R^2 train: %.3f, test: %.3f' % (\n",
    "        r2_score(y_train, y_train_pred),\n",
    "        r2_score(y_test, y_test_pred)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random forest regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "X = df[features].values\n",
    "y = df['G3'].values\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE train: 0.250, test: 3.806\n",
      "R^2 train: 0.986, test: 0.864\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "\n",
    "forest = RandomForestRegressor(n_estimators=1000, \n",
    "                               criterion='mse', \n",
    "                               random_state=1, \n",
    "                               n_jobs=-1)\n",
    "forest.fit(X_train, y_train)\n",
    "y_train_pred = forest.predict(X_train)\n",
    "y_test_pred = forest.predict(X_test)\n",
    "\n",
    "print('MSE train: %.3f, test: %.3f' % (\n",
    "        mean_squared_error(y_train, y_train_pred),\n",
    "        mean_squared_error(y_test, y_test_pred)))\n",
    "print('R^2 train: %.3f, test: %.3f' % (\n",
    "        r2_score(y_train, y_train_pred),\n",
    "        r2_score(y_test, y_test_pred)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAacAAAEYCAYAAAD4czk4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1df36358>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(y_train_pred,  \n",
    "            y_train_pred - y_train, \n",
    "            c='black', \n",
    "            marker='o', \n",
    "            s=35,\n",
    "            alpha=0.5,\n",
    "            label='Training data')\n",
    "plt.scatter(y_test_pred,  \n",
    "            y_test_pred - y_test, \n",
    "            c='lightgreen', \n",
    "            marker='s', \n",
    "            s=35,\n",
    "            alpha=0.7,\n",
    "            label='Test data')\n",
    "\n",
    "plt.xlabel('Predicted values')\n",
    "plt.ylabel('Residuals')\n",
    "plt.legend(loc='upper left')\n",
    "plt.hlines(y=0, xmin=0, xmax=22, lw=2, color='red')\n",
    "plt.xlim([0, 22])\n",
    "plt.tight_layout()\n",
    "\n",
    "# plt.savefig('./figures/slr_residuals.png', dpi=300)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "___"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Linear Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R-squared: 0.780914360184\n"
     ]
    }
   ],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y)\n",
    "regressor = LinearRegression()\n",
    "regressor.fit(X_train, y_train)\n",
    "y_predictions = regressor.predict(X_test)\n",
    "print 'R-squared:', regressor.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Cross Validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average of scores:  0.794475403313\n",
      "Cross validation scores:  [ 0.81813068  0.88591594  0.77938816  0.78516026  0.70378198]\n"
     ]
    }
   ],
   "source": [
    "scores = cross_val_score(regressor, X, y, cv=5)\n",
    "print \"Average of scores: \", scores.mean()\n",
    "print \"Cross validation scores: \", scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZEAAAEpCAYAAABSuB9PAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEREp2/8H30PSet3xDaoAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2a918f98>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(y_test,y_predictions)\n",
    "plt.xlabel('True Quality')\n",
    "plt.ylabel('Predicted Quality')\n",
    "plt.title('Predicted Quality Against True Quality')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fitting models with gradient descent"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# SGDRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Miniconda2\\lib\\site-packages\\sklearn\\utils\\validation.py:420: DataConversionWarning: Data with input dtype int64 was converted to float64 by StandardScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "C:\\Miniconda2\\lib\\site-packages\\sklearn\\preprocessing\\data.py:583: DeprecationWarning: Passing 1d arrays as data is deprecated in 0.17 and will raise ValueError in 0.19. Reshape your data either using X.reshape(-1, 1) if your data has a single feature or X.reshape(1, -1) if it contains a single sample.\n",
      "  warnings.warn(DEPRECATION_MSG_1D, DeprecationWarning)\n",
      "C:\\Miniconda2\\lib\\site-packages\\sklearn\\utils\\validation.py:420: DataConversionWarning: Data with input dtype int64 was converted to float64 by StandardScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "C:\\Miniconda2\\lib\\site-packages\\sklearn\\preprocessing\\data.py:646: DeprecationWarning: Passing 1d arrays as data is deprecated in 0.17 and will raise ValueError in 0.19. Reshape your data either using X.reshape(-1, 1) if your data has a single feature or X.reshape(1, -1) if it contains a single sample.\n",
      "  warnings.warn(DEPRECATION_MSG_1D, DeprecationWarning)\n",
      "C:\\Miniconda2\\lib\\site-packages\\sklearn\\utils\\validation.py:420: DataConversionWarning: Data with input dtype int64 was converted to float64 by StandardScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "C:\\Miniconda2\\lib\\site-packages\\sklearn\\preprocessing\\data.py:646: DeprecationWarning: Passing 1d arrays as data is deprecated in 0.17 and will raise ValueError in 0.19. Reshape your data either using X.reshape(-1, 1) if your data has a single feature or X.reshape(1, -1) if it contains a single sample.\n",
      "  warnings.warn(DEPRECATION_MSG_1D, DeprecationWarning)\n"
     ]
    }
   ],
   "source": [
    "# Scaling the features using StandardScaler:\n",
    "X_scaler = StandardScaler()\n",
    "y_scaler = StandardScaler()\n",
    "X_train = X_scaler.fit_transform(X_train)\n",
    "y_train = y_scaler.fit_transform(y_train)\n",
    "X_test = X_scaler.transform(X_test)\n",
    "y_test = y_scaler.transform(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 211,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cross validation r-squared scores: [ 0.75096581  0.86510753  0.76876374  0.7369485   0.7987154 ]\n",
      "Average cross validation r-squared score: 0.784100195438\n",
      "Test set r-squared score 0.773056308523\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Miniconda2\\lib\\site-packages\\sklearn\\utils\\__init__.py:93: DeprecationWarning: Function transform is deprecated; Support to use estimators as feature selectors will be removed in version 0.19. Use SelectFromModel instead.\n",
      "  warnings.warn(msg, category=DeprecationWarning)\n"
     ]
    }
   ],
   "source": [
    "regressor = SGDRegressor(loss='squared_loss')\n",
    "scores = cross_val_score(regressor, X_train, y_train, cv=5)\n",
    "print 'Cross validation r-squared scores:', scores\n",
    "print 'Average cross validation r-squared score:', np.mean(scores)\n",
    "regressor.fit_transform(X_train, y_train)\n",
    "print 'Test set r-squared score', regressor.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.cross_validation import *\n",
    "def train_and_evaluate(clf, X_train, y_train):\n",
    "    \n",
    "    clf.fit(X_train, y_train)\n",
    "    \n",
    "    print \"Coefficient of determination on training set:\",clf.score(X_train, y_train)\n",
    "    \n",
    "    # create a k-fold croos validation iterator of k=5 folds\n",
    "    cv = KFold(X_train.shape[0], 5, shuffle=True, random_state=33)\n",
    "    scores = cross_val_score(clf, X_train, y_train, cv=cv)\n",
    "    print \"Average coefficient of determination using 5-fold crossvalidation:\",np.mean(scores)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# Support Vector Machines for regression\n",
    "The regression version of SVM can be used instead to find the hyperplane (note how easy is to change the classification method in scikit-learn!). We will try a linear kernel, a polynomial kernel, and finally, a rbf kernel. For more information on kernels, see http://scikit-learn.org/stable/modules/svm.html#svm-kernels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Coefficient of determination on training set: 0.849157093784\n",
      "Average coefficient of determination using 5-fold crossvalidation: 0.842490490255\n"
     ]
    }
   ],
   "source": [
    "from sklearn import svm\n",
    "clf_svr= svm.SVR(kernel='linear')\n",
    "train_and_evaluate(clf_svr,X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Coefficient of determination on training set: 0.991247519011\n",
      "Average coefficient of determination using 5-fold crossvalidation: -92.8405695274\n"
     ]
    }
   ],
   "source": [
    "clf_svr_poly= svm.SVR(kernel='poly')\n",
    "train_and_evaluate(clf_svr_poly,X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Coefficient of determination on training set: 0.76854274345\n",
      "Average coefficient of determination using 5-fold crossvalidation: 0.637749259288\n"
     ]
    }
   ],
   "source": [
    "clf_svr_rbf= svm.SVR(kernel='rbf')\n",
    "train_and_evaluate(clf_svr_rbf,X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Coefficient of determination on training set: 0.875086659141\n",
      "Average coefficient of determination using 5-fold crossvalidation: 0.577096119973\n"
     ]
    }
   ],
   "source": [
    "clf_svr_poly2= svm.SVR(kernel='poly',degree=2)\n",
    "train_and_evaluate(clf_svr_poly2,X_train,y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Random Forests for Regression\n",
    "Finally, let's try again Random Forests, in their Extra Trees, and Regression version"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Coefficient of determination on training set: 1.0\n",
      "Average coefficient of determination using 5-fold crossvalidation: 0.834438823442\n"
     ]
    }
   ],
   "source": [
    "from sklearn import ensemble\n",
    "clf_et=ensemble.ExtraTreesRegressor(n_estimators=10,random_state=42)\n",
    "train_and_evaluate(clf_et,X_train,y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "An interesting side effect of random forest classification, is that you can measure how 'important' each feature is when predicting the final result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[['0.000414630355532' 'Dalc']\n",
      " ['0.000514007655848' 'Fedu']\n",
      " ['0.000616839340144' 'Fjob']\n",
      " ['0.000812859348316' 'G1']\n",
      " ['0.000894604750305' 'G2']\n",
      " ['0.000964541098088' 'Medu']\n",
      " ['0.0018054901535' 'Mjob']\n",
      " ['0.00181670492675' 'Pstatus']\n",
      " ['0.00185431089792' 'Walc']\n",
      " ['0.00216206896705' 'absences']\n",
      " ['0.00220641658743' 'activities']\n",
      " ['0.00239183938313' 'address']\n",
      " ['0.0024096646334' 'age']\n",
      " ['0.00249476017392' 'failures']\n",
      " ['0.00250597775474' 'famrel']\n",
      " ['0.00255391358532' 'famsize']\n",
      " ['0.00279086333248' 'famsup']\n",
      " ['0.0031383209147' 'freetime']\n",
      " ['0.00360150739838' 'goout']\n",
      " ['0.00431525126952' 'guardian']\n",
      " ['0.00445755164541' 'health']\n",
      " ['0.0048610587442' 'higher']\n",
      " ['0.00532054800371' 'internet']\n",
      " ['0.00710396827835' 'nursery']\n",
      " ['0.00841846425943' 'paid']\n",
      " ['0.00925093324961' 'reason']\n",
      " ['0.00956984349144' 'romantic']\n",
      " ['0.0100368287457' 'school']\n",
      " ['0.0144620072805' 'schoolsup']\n",
      " ['0.0435626263148' 'sex']\n",
      " ['0.330796342132' 'studytime']\n",
      " ['0.511895255329' 'traveltime']]\n"
     ]
    }
   ],
   "source": [
    "print np.sort(zip(clf_et.feature_importances_,features),axis=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, evaluate our classifiers on the testing set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 223,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Coefficient of determination:0.766 \n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn import metrics\n",
    "def measure_performance(X,y,clf, show_accuracy=True,\n",
    "                        show_classification_report=True,\n",
    "                        show_confusion_matrix=True,\n",
    "                        show_r2_score=False):\n",
    "    y_pred=clf.predict(X)   \n",
    "    if show_accuracy:\n",
    "        print \"Accuracy:{0:.3f}\".format(metrics.accuracy_score(y,y_pred)),\"\\n\"\n",
    "\n",
    "    if show_classification_report:\n",
    "        print \"Classification report\"\n",
    "        print metrics.classification_report(y,y_pred),\"\\n\"\n",
    "        \n",
    "    if show_confusion_matrix:\n",
    "        print \"Confusion matrix\"\n",
    "        print metrics.confusion_matrix(y,y_pred),\"\\n\"\n",
    "        \n",
    "    if show_r2_score:\n",
    "        print \"Coefficient of determination:{0:.3f}\".format(metrics.r2_score(y,y_pred)),\"\\n\"\n",
    "\n",
    "        \n",
    "measure_performance(X_test,y_test,clf_et,\n",
    "                    show_accuracy=False,\n",
    "                    show_classification_report=False,\n",
    "                    show_confusion_matrix=False,\n",
    "                    show_r2_score=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "___"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}