Notebooks/PMF_SU_2018_Vjezbe3_00_Odabir_znacajki.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Strojno učenje - PMF\n",
    "\n",
    "Akademska godina 2017./2018.\n",
    "[https://web.math.pmf.unizg.hr/nastava/su/](https://web.math.pmf.unizg.hr/nastava/su/)\n",
    "\n",
    "Matija Piškorec\n",
    "\n",
    "kernel: Python 3.6\n",
    "\n",
    "# 3. Odabir značajki\n",
    "\n",
    "Glavni cilj odabira značajki je odabrati podskup \"važnih\" značajki kako bi se 1) poboljšala prediktivna moć, 2) smanjili računalni zahtjevi potrebni za predikciju, 3) poboljšala interpretabilnost prediktivnog modela. Ugrubo, odabir značajki može se podijeliti na sljedeći način:\n",
    "\n",
    "1. *Metode neovisne o modelu* - Koje ne ovise o prediktivnom modelu koji evaluira važnost značajki nego o raznim statističkim testovima kao što su Pearsonov koeficijent korelacije, $\\chi^2$-statistika, uzajamna informacija (engl. mutual information), ili ANOVA F-vrijednost. Ove metode se razlikuju prema tome jesu li nadzirane ili nenadzirane, odnosno koriste li ciljnu varijablu u odabiru značajki ili ne.\n",
    "    \n",
    "2. *Metode ovisne o modelu* - Koje ovise o konkretnom prediktivnom modelu koji evaluira važnost značajki. Ako prediktivni model pruža važnost značajki kao jedan od međukoraka pri izgradnji modela, onda se te važnosti mogu koristiti direktno. Ovo se naziva *transparent model* (ili *embedded*) pristup. Drugi pristup je evaluacija važnosti značajki preko njihovog doprinosa u predikciji tako što se one uključe ili isključe prilikom izgradnje modela. Ovo se naziva *wrapper* pristup.\n",
    "\n",
    "Primjetite razliku između *odabira značajki*, gdje odabiremo podskup značajki iz originalnog skupa podataka, i *smanjenja dimenzionalnosti*, gdje se podaci projiciraju u drugačiji prostor značajki. Metode za smanjenje dimenzionalnosti su pokrivene u sljedećim vježbama.\n",
    "\n",
    "Većina sadržaja ove vježbe preuzeta je s jedne od vježbi za [First Int'l Summer School on Data Science](https://sites.google.com/site/ssdatascience2016/lecture-notes) održanoj u Splitu u rujnu 2016., koje su dostupne na [Github repozitoriju](https://github.com/SSDS2016/Day-3). Od ostalih materijala, scikit-learn [korisnička dokumentacija](http://scikit-learn.org/stable/user_guide.html) ima poglavlje o odabiru značajki [[1]](#scikit-feature-selection). Neki od sljedećih primjera su iz četvrtog poglavlja knjige \"Python Machine Learning\" [[2]](#pyml-ch4) i petog poglavlja knjige \"Introduction to Machine Learning with Python\" [[3]](#intro-ml-py-ch5). Također, dobar pregled odabira značajki je niz blog postova \"Selecting good features\" s [\"Diving into data\"](http://blog.datadive.net/) bloga [[4]](#diving-into-data)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Preduvjeti\n",
    "--------------\n",
    "\n",
    "U ovoj vježbi koristit ćemo \"Wisconsin Breast Cancer\" dataset koji sadrži 569 primjera i 30 atributa, dok je ciljna varijabla ili 1 (maligni tumor) ili 0 (benigni tumor). Dataset je dostupan lokalno, inače se može dohvatiti preko scikit-learn funkcije [`sklearn.datasets.load_breast_cancer`](http://scikit-learn.org/dev/modules/generated/sklearn.datasets.load_breast_cancer.html#sklearn.datasets.load_breast_cancer) (verzija 0.17.1).\n",
    "\n",
    "Za ostale datasetove dostupne unutar scikit-learna pogledajte službenu dokumentaciju na [`sklearn.datasets`](http://scikit-learn.org/dev/modules/classes.html#module-sklearn.datasets)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean radius</th>\n",
       "      <th>mean texture</th>\n",
       "      <th>mean perimeter</th>\n",
       "      <th>mean area</th>\n",
       "      <th>mean smoothness</th>\n",
       "      <th>mean compactness</th>\n",
       "      <th>mean concavity</th>\n",
       "      <th>mean concave points</th>\n",
       "      <th>mean symmetry</th>\n",
       "      <th>mean fractal dimension</th>\n",
       "      <th>...</th>\n",
       "      <th>worst texture</th>\n",
       "      <th>worst perimeter</th>\n",
       "      <th>worst area</th>\n",
       "      <th>worst smoothness</th>\n",
       "      <th>worst compactness</th>\n",
       "      <th>worst concavity</th>\n",
       "      <th>worst concave points</th>\n",
       "      <th>worst symmetry</th>\n",
       "      <th>worst fractal dimension</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>17.99</td>\n",
       "      <td>10.38</td>\n",
       "      <td>122.80</td>\n",
       "      <td>1001.0</td>\n",
       "      <td>0.11840</td>\n",
       "      <td>0.27760</td>\n",
       "      <td>0.3001</td>\n",
       "      <td>0.14710</td>\n",
       "      <td>0.2419</td>\n",
       "      <td>0.07871</td>\n",
       "      <td>...</td>\n",
       "      <td>17.33</td>\n",
       "      <td>184.60</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>0.1622</td>\n",
       "      <td>0.6656</td>\n",
       "      <td>0.7119</td>\n",
       "      <td>0.2654</td>\n",
       "      <td>0.4601</td>\n",
       "      <td>0.11890</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>20.57</td>\n",
       "      <td>17.77</td>\n",
       "      <td>132.90</td>\n",
       "      <td>1326.0</td>\n",
       "      <td>0.08474</td>\n",
       "      <td>0.07864</td>\n",
       "      <td>0.0869</td>\n",
       "      <td>0.07017</td>\n",
       "      <td>0.1812</td>\n",
       "      <td>0.05667</td>\n",
       "      <td>...</td>\n",
       "      <td>23.41</td>\n",
       "      <td>158.80</td>\n",
       "      <td>1956.0</td>\n",
       "      <td>0.1238</td>\n",
       "      <td>0.1866</td>\n",
       "      <td>0.2416</td>\n",
       "      <td>0.1860</td>\n",
       "      <td>0.2750</td>\n",
       "      <td>0.08902</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>19.69</td>\n",
       "      <td>21.25</td>\n",
       "      <td>130.00</td>\n",
       "      <td>1203.0</td>\n",
       "      <td>0.10960</td>\n",
       "      <td>0.15990</td>\n",
       "      <td>0.1974</td>\n",
       "      <td>0.12790</td>\n",
       "      <td>0.2069</td>\n",
       "      <td>0.05999</td>\n",
       "      <td>...</td>\n",
       "      <td>25.53</td>\n",
       "      <td>152.50</td>\n",
       "      <td>1709.0</td>\n",
       "      <td>0.1444</td>\n",
       "      <td>0.4245</td>\n",
       "      <td>0.4504</td>\n",
       "      <td>0.2430</td>\n",
       "      <td>0.3613</td>\n",
       "      <td>0.08758</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>11.42</td>\n",
       "      <td>20.38</td>\n",
       "      <td>77.58</td>\n",
       "      <td>386.1</td>\n",
       "      <td>0.14250</td>\n",
       "      <td>0.28390</td>\n",
       "      <td>0.2414</td>\n",
       "      <td>0.10520</td>\n",
       "      <td>0.2597</td>\n",
       "      <td>0.09744</td>\n",
       "      <td>...</td>\n",
       "      <td>26.50</td>\n",
       "      <td>98.87</td>\n",
       "      <td>567.7</td>\n",
       "      <td>0.2098</td>\n",
       "      <td>0.8663</td>\n",
       "      <td>0.6869</td>\n",
       "      <td>0.2575</td>\n",
       "      <td>0.6638</td>\n",
       "      <td>0.17300</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>20.29</td>\n",
       "      <td>14.34</td>\n",
       "      <td>135.10</td>\n",
       "      <td>1297.0</td>\n",
       "      <td>0.10030</td>\n",
       "      <td>0.13280</td>\n",
       "      <td>0.1980</td>\n",
       "      <td>0.10430</td>\n",
       "      <td>0.1809</td>\n",
       "      <td>0.05883</td>\n",
       "      <td>...</td>\n",
       "      <td>16.67</td>\n",
       "      <td>152.20</td>\n",
       "      <td>1575.0</td>\n",
       "      <td>0.1374</td>\n",
       "      <td>0.2050</td>\n",
       "      <td>0.4000</td>\n",
       "      <td>0.1625</td>\n",
       "      <td>0.2364</td>\n",
       "      <td>0.07678</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 31 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   mean radius  mean texture  mean perimeter  mean area  mean smoothness  \\\n",
       "0        17.99         10.38          122.80     1001.0          0.11840   \n",
       "1        20.57         17.77          132.90     1326.0          0.08474   \n",
       "2        19.69         21.25          130.00     1203.0          0.10960   \n",
       "3        11.42         20.38           77.58      386.1          0.14250   \n",
       "4        20.29         14.34          135.10     1297.0          0.10030   \n",
       "\n",
       "   mean compactness  mean concavity  mean concave points  mean symmetry  \\\n",
       "0           0.27760          0.3001              0.14710         0.2419   \n",
       "1           0.07864          0.0869              0.07017         0.1812   \n",
       "2           0.15990          0.1974              0.12790         0.2069   \n",
       "3           0.28390          0.2414              0.10520         0.2597   \n",
       "4           0.13280          0.1980              0.10430         0.1809   \n",
       "\n",
       "   mean fractal dimension   ...    worst texture  worst perimeter  worst area  \\\n",
       "0                 0.07871   ...            17.33           184.60      2019.0   \n",
       "1                 0.05667   ...            23.41           158.80      1956.0   \n",
       "2                 0.05999   ...            25.53           152.50      1709.0   \n",
       "3                 0.09744   ...            26.50            98.87       567.7   \n",
       "4                 0.05883   ...            16.67           152.20      1575.0   \n",
       "\n",
       "   worst smoothness  worst compactness  worst concavity  worst concave points  \\\n",
       "0            0.1622             0.6656           0.7119                0.2654   \n",
       "1            0.1238             0.1866           0.2416                0.1860   \n",
       "2            0.1444             0.4245           0.4504                0.2430   \n",
       "3            0.2098             0.8663           0.6869                0.2575   \n",
       "4            0.1374             0.2050           0.4000                0.1625   \n",
       "\n",
       "   worst symmetry  worst fractal dimension  target  \n",
       "0          0.4601                  0.11890     0.0  \n",
       "1          0.2750                  0.08902     0.0  \n",
       "2          0.3613                  0.08758     0.0  \n",
       "3          0.6638                  0.17300     0.0  \n",
       "4          0.2364                  0.07678     0.0  \n",
       "\n",
       "[5 rows x 31 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "breast_cancer = pd.read_csv('../assets/datasets/csv/breast_cancer.csv',index_col=0)\n",
    "breast_cancer.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 569 entries, 0 to 568\n",
      "Data columns (total 31 columns):\n",
      "mean radius                569 non-null float64\n",
      "mean texture               569 non-null float64\n",
      "mean perimeter             569 non-null float64\n",
      "mean area                  569 non-null float64\n",
      "mean smoothness            569 non-null float64\n",
      "mean compactness           569 non-null float64\n",
      "mean concavity             569 non-null float64\n",
      "mean concave points        569 non-null float64\n",
      "mean symmetry              569 non-null float64\n",
      "mean fractal dimension     569 non-null float64\n",
      "radius error               569 non-null float64\n",
      "texture error              569 non-null float64\n",
      "perimeter error            569 non-null float64\n",
      "area error                 569 non-null float64\n",
      "smoothness error           569 non-null float64\n",
      "compactness error          569 non-null float64\n",
      "concavity error            569 non-null float64\n",
      "concave points error       569 non-null float64\n",
      "symmetry error             569 non-null float64\n",
      "fractal dimension error    569 non-null float64\n",
      "worst radius               569 non-null float64\n",
      "worst texture              569 non-null float64\n",
      "worst perimeter            569 non-null float64\n",
      "worst area                 569 non-null float64\n",
      "worst smoothness           569 non-null float64\n",
      "worst compactness          569 non-null float64\n",
      "worst concavity            569 non-null float64\n",
      "worst concave points       569 non-null float64\n",
      "worst symmetry             569 non-null float64\n",
      "worst fractal dimension    569 non-null float64\n",
      "target                     569 non-null float64\n",
      "dtypes: float64(31)\n",
      "memory usage: 142.2 KB\n"
     ]
    }
   ],
   "source": [
    "breast_cancer.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Iako nije nužno potrebno u našem slučaju, slijedit ćemo standardnu praksu strojnog učenja i podijeliti naš dataset u skup za treniranje i skup za testiranje s funkcijom [`sklearn.cross_validation.train_test_split`](http://scikit-learn.org/stable/modules/generated/sklearn.cross_validation.train_test_split.html), te provesti svu analizu samo na skupu za treniranje."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Broj primjera za treniranje: 455\n",
      "Broj primjera za testiranje: 114\n"
     ]
    }
   ],
   "source": [
    "breast_cancer_train, breast_cancer_test = train_test_split(breast_cancer, test_size=0.20, random_state=42)\n",
    "print('Broj primjera za treniranje: ' + str(len(breast_cancer_train)))\n",
    "print('Broj primjera za testiranje: ' + str(len(breast_cancer_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "3.1. Univarijantni odabir značajki\n",
    "-----------------------------------\n",
    "\n",
    "Vjerojatno najjednostavnija nenadzirani način za univarijantni odabir značajki je preko njihove varijance, pri čemu se zadržavaju samo značajke čija je varijanca iznad nekog predefiniranog praga. Ova metoda je implementirana u [`sklearn.feature_selection.VarianceThreshold`](http://scikit-learn.org/stable/modules/generated/sklearn.feature_selection.VarianceThreshold.html#sklearn.feature_selection.VarianceThreshold). Također, ako imamo dvije ili više međusobno korelirane značajke možemo zadržati samo jednu od njih, obično onu koja je odabrana kao važnija nekom drugom metodom.\n",
    "\n",
    "U sljedećem primjeru primjenit ćemo nadzirani univarijantni odabir značajki na breast cancer skupu podataka koristeći ANOVA F-vrijednost kao mjeru povezanosti značajke s ciljnom varijablom. Ova metoda je implementirana u  [`sklearn.feature_selection.f_classif`](http://scikit-learn.org/stable/modules/generated/sklearn.feature_selection.f_classif.html#sklearn.feature_selection.f_classif). Za bolju demonstraciju dodat ćemo umjetne značajke koje ne nose nikakvu informaciju o ciljnoj varijabli (šum). Očekujemo da ćemo takve značajke lako prepoznati s jednostavnim univarijantnim metodama odabira značajki. Jednom kada odaberemo mjeru za važnost značajki možemo koristiti [`sklearn.feature_selection.SelectKBest`](http://scikit-learn.org/stable/modules/generated/sklearn.feature_selection.SelectKBest.html#sklearn.feature_selection.SelectKBest) ili [`sklearn.feature_selection.SelectPercentile`](http://scikit-learn.org/stable/modules/generated/sklearn.feature_selection.SelectPercentile.html#sklearn.feature_selection.SelectPercentile) za odabir najboljih značajki, ovisno o tome želimo li odabrati odreženi broj ili postotak najvažnijih značajki. Pošto znamo da imamo 30 *stvarnih* i 10 *lažnih* značajki, odabrat ćemo točno 30 značajki da provjerimo možemo li vidjeti razliku. Za popis svih metoda odabira značajki u scikit-learnu pogledajte [`sklearn.feature_selection`](http://scikit-learn.org/stable/modules/classes.html#module-sklearn.feature_selection). Sličan primjer je dostupan i u korisničkoj dokumentaciji scikit-learna [\"Univariate feature selection\"](http://scikit-learn.org/stable/auto_examples/feature_selection/plot_feature_selection.html)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.feature_selection import SelectKBest, f_classif"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean radius</th>\n",
       "      <th>mean texture</th>\n",
       "      <th>mean perimeter</th>\n",
       "      <th>mean area</th>\n",
       "      <th>mean smoothness</th>\n",
       "      <th>mean compactness</th>\n",
       "      <th>mean concavity</th>\n",
       "      <th>mean concave points</th>\n",
       "      <th>mean symmetry</th>\n",
       "      <th>mean fractal dimension</th>\n",
       "      <th>...</th>\n",
       "      <th>noise 0</th>\n",
       "      <th>noise 1</th>\n",
       "      <th>noise 2</th>\n",
       "      <th>noise 3</th>\n",
       "      <th>noise 4</th>\n",
       "      <th>noise 5</th>\n",
       "      <th>noise 6</th>\n",
       "      <th>noise 7</th>\n",
       "      <th>noise 8</th>\n",
       "      <th>noise 9</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>17.99</td>\n",
       "      <td>10.38</td>\n",
       "      <td>122.80</td>\n",
       "      <td>1001.0</td>\n",
       "      <td>0.11840</td>\n",
       "      <td>0.27760</td>\n",
       "      <td>0.3001</td>\n",
       "      <td>0.14710</td>\n",
       "      <td>0.2419</td>\n",
       "      <td>0.07871</td>\n",
       "      <td>...</td>\n",
       "      <td>0.253753</td>\n",
       "      <td>0.237089</td>\n",
       "      <td>0.113005</td>\n",
       "      <td>-0.590136</td>\n",
       "      <td>1.001166</td>\n",
       "      <td>-0.067253</td>\n",
       "      <td>0.127288</td>\n",
       "      <td>-1.663485</td>\n",
       "      <td>-2.006971</td>\n",
       "      <td>-0.010836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>20.57</td>\n",
       "      <td>17.77</td>\n",
       "      <td>132.90</td>\n",
       "      <td>1326.0</td>\n",
       "      <td>0.08474</td>\n",
       "      <td>0.07864</td>\n",
       "      <td>0.0869</td>\n",
       "      <td>0.07017</td>\n",
       "      <td>0.1812</td>\n",
       "      <td>0.05667</td>\n",
       "      <td>...</td>\n",
       "      <td>1.315888</td>\n",
       "      <td>1.545658</td>\n",
       "      <td>0.170740</td>\n",
       "      <td>1.677040</td>\n",
       "      <td>0.491498</td>\n",
       "      <td>1.237345</td>\n",
       "      <td>0.362460</td>\n",
       "      <td>-0.148685</td>\n",
       "      <td>1.227565</td>\n",
       "      <td>-1.239274</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>19.69</td>\n",
       "      <td>21.25</td>\n",
       "      <td>130.00</td>\n",
       "      <td>1203.0</td>\n",
       "      <td>0.10960</td>\n",
       "      <td>0.15990</td>\n",
       "      <td>0.1974</td>\n",
       "      <td>0.12790</td>\n",
       "      <td>0.2069</td>\n",
       "      <td>0.05999</td>\n",
       "      <td>...</td>\n",
       "      <td>0.381878</td>\n",
       "      <td>-1.323544</td>\n",
       "      <td>-1.312740</td>\n",
       "      <td>-0.143689</td>\n",
       "      <td>-0.557137</td>\n",
       "      <td>0.345895</td>\n",
       "      <td>0.553536</td>\n",
       "      <td>-0.207683</td>\n",
       "      <td>-0.442661</td>\n",
       "      <td>-1.344949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>11.42</td>\n",
       "      <td>20.38</td>\n",
       "      <td>77.58</td>\n",
       "      <td>386.1</td>\n",
       "      <td>0.14250</td>\n",
       "      <td>0.28390</td>\n",
       "      <td>0.2414</td>\n",
       "      <td>0.10520</td>\n",
       "      <td>0.2597</td>\n",
       "      <td>0.09744</td>\n",
       "      <td>...</td>\n",
       "      <td>1.157201</td>\n",
       "      <td>-1.588217</td>\n",
       "      <td>1.789478</td>\n",
       "      <td>1.097915</td>\n",
       "      <td>0.449084</td>\n",
       "      <td>-0.184956</td>\n",
       "      <td>0.388488</td>\n",
       "      <td>0.375909</td>\n",
       "      <td>1.199361</td>\n",
       "      <td>0.857071</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>20.29</td>\n",
       "      <td>14.34</td>\n",
       "      <td>135.10</td>\n",
       "      <td>1297.0</td>\n",
       "      <td>0.10030</td>\n",
       "      <td>0.13280</td>\n",
       "      <td>0.1980</td>\n",
       "      <td>0.10430</td>\n",
       "      <td>0.1809</td>\n",
       "      <td>0.05883</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.364740</td>\n",
       "      <td>-1.488392</td>\n",
       "      <td>-0.074576</td>\n",
       "      <td>0.696764</td>\n",
       "      <td>0.018299</td>\n",
       "      <td>0.192341</td>\n",
       "      <td>0.063116</td>\n",
       "      <td>-1.052365</td>\n",
       "      <td>-0.315774</td>\n",
       "      <td>3.146814</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 41 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   mean radius  mean texture  mean perimeter  mean area  mean smoothness  \\\n",
       "0        17.99         10.38          122.80     1001.0          0.11840   \n",
       "1        20.57         17.77          132.90     1326.0          0.08474   \n",
       "2        19.69         21.25          130.00     1203.0          0.10960   \n",
       "3        11.42         20.38           77.58      386.1          0.14250   \n",
       "4        20.29         14.34          135.10     1297.0          0.10030   \n",
       "\n",
       "   mean compactness  mean concavity  mean concave points  mean symmetry  \\\n",
       "0           0.27760          0.3001              0.14710         0.2419   \n",
       "1           0.07864          0.0869              0.07017         0.1812   \n",
       "2           0.15990          0.1974              0.12790         0.2069   \n",
       "3           0.28390          0.2414              0.10520         0.2597   \n",
       "4           0.13280          0.1980              0.10430         0.1809   \n",
       "\n",
       "   mean fractal dimension    ...      noise 0   noise 1   noise 2   noise 3  \\\n",
       "0                 0.07871    ...     0.253753  0.237089  0.113005 -0.590136   \n",
       "1                 0.05667    ...     1.315888  1.545658  0.170740  1.677040   \n",
       "2                 0.05999    ...     0.381878 -1.323544 -1.312740 -0.143689   \n",
       "3                 0.09744    ...     1.157201 -1.588217  1.789478  1.097915   \n",
       "4                 0.05883    ...    -0.364740 -1.488392 -0.074576  0.696764   \n",
       "\n",
       "    noise 4   noise 5   noise 6   noise 7   noise 8   noise 9  \n",
       "0  1.001166 -0.067253  0.127288 -1.663485 -2.006971 -0.010836  \n",
       "1  0.491498  1.237345  0.362460 -0.148685  1.227565 -1.239274  \n",
       "2 -0.557137  0.345895  0.553536 -0.207683 -0.442661 -1.344949  \n",
       "3  0.449084 -0.184956  0.388488  0.375909  1.199361  0.857071  \n",
       "4  0.018299  0.192341  0.063116 -1.052365 -0.315774  3.146814  \n",
       "\n",
       "[5 rows x 41 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "breast_cancer = pd.read_csv('../assets/datasets/csv/breast_cancer.csv',index_col=0)\n",
    "\n",
    "N_noise_features = 10\n",
    "noise_features = pd.DataFrame(np.random.randn(breast_cancer.shape[0],N_noise_features),\n",
    "                              index=breast_cancer.index,\n",
    "                              columns=['noise ' + str(i) for i in range(N_noise_features)])\n",
    "\n",
    "breast_cancer_extended = pd.concat([breast_cancer, noise_features], axis=1)\n",
    "\n",
    "breast_cancer_extended.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "breast_cancer_train, breast_cancer_test = train_test_split(breast_cancer_extended, test_size=0.20, random_state=42)\n",
    "\n",
    "X_train = breast_cancer_train.drop('target',axis=1)\n",
    "y_train = breast_cancer_train['target']\n",
    "\n",
    "k_best = 30\n",
    "selector = SelectKBest(f_classif, k=k_best)\n",
    "selector.fit(X_train, y_train)\n",
    "scores = -np.log10(selector.pvalues_)\n",
    "indices = np.argsort(scores)[::-1] # Vrati indekse koji sortiraju polje (obratno)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfAAAAFgCAYAAABEyiulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xa028a20>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(7,5))\n",
    "\n",
    "plt.plot(range(X_train.shape[1]), scores[indices], 'o', color='gray')\n",
    "\n",
    "plt.title('Univarijantni odabir znacajki za najvaznijih 30 znacajki')\n",
    "plt.ylabel('log ANOVA p-vrijednosti')\n",
    "plt.xlim([-1, X_train.shape[1]])\n",
    "plt.xticks(range(X_train.shape[1]), breast_cancer_train.drop('target',axis=1).columns.values[indices], rotation=90)\n",
    "plt.vlines(k_best-0.5, 0, np.max(scores), color='red', linestyle='--')\n",
    "plt.grid(True,linestyle='--')\n",
    "plt.tight_layout()\n",
    "\n",
    "# Hack za bojanje noise znacajki u crveno\n",
    "[ i.set_color(\"red\") for i in plt.gca().get_xticklabels() if i.get_text()[:5] == 'noise' ]\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Vidimo da su neke od naših stvarnih značajki rangirane lošije od lažnih značajki. Ovo nije iznenađujuće jer je riječ o značajkama koje su slabo informativne, bilo zato što su dosta opskurne (mean fractal dimension) ili zapravo kodiraju grešku neke druge značajke. Ovo još uvijek nije garancija da su navedene značajke beskorisne, ali može poslužiti kao jednostavni prvi korak u procesu odabira značajki."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Vježba 3.1.** Umjesto odabira određenog broja ili određenog postotka najvažnijih značajki možemo odabrati gornju granicu na udio *lažnih otkrića* (engl. false discoveries) - značajki za koje vjerujemo da su statistički povezane s ciljnom varijablom a da to zapravo nisu. Funkcija [`sklearn.feature_selection.SelectFdr`](http://scikit-learn.org/stable/modules/generated/sklearn.feature_selection.SelectFdr.html#sklearn.feature_selection.SelectFdr) odabire najvažnije značajke koristeći Benjamini-Hochberg proceduru uz očekivani udio lažnih otkrića `alpha` (engl. false discovery rate, FDR). Iskoristite ovu funkciju za odabir značajki uz FDR `alpha=0.05`. Napomena: Možete dobiti logičku masku za odabrane značajke pomoću funkcije `get_support()`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.feature_selection import SelectFdr, f_classif"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "breast_cancer = pd.read_csv('../assets/datasets/csv/breast_cancer.csv',index_col=0)\n",
    "\n",
    "N_noise_features = 10\n",
    "noise_features = pd.DataFrame(np.random.randn(breast_cancer.shape[0],N_noise_features),\n",
    "                              index=breast_cancer.index,\n",
    "                              columns=['noise ' + str(i) for i in range(N_noise_features)])\n",
    "\n",
    "breast_cancer_extended = pd.concat([breast_cancer, noise_features], axis=1)\n",
    "\n",
    "breast_cancer_train, breast_cancer_test = train_test_split(breast_cancer_extended, test_size=0.20, random_state=42)\n",
    "\n",
    "X_train = breast_cancer_train.drop('target',axis=1)\n",
    "y_train = breast_cancer_train['target']\n",
    "\n",
    "k_best = 30\n",
    "selector = SelectFdr(f_classif, alpha=0.05)\n",
    "selector.fit(X_train, y_train)\n",
    "scores = -np.log10(selector.pvalues_)\n",
    "indices = np.argsort(scores)[::-1] # Vrati indekse koji sortiraju polje (obratno)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfAAAAFgCAYAAABEyiulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x911c710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(7,5))\n",
    "\n",
    "plt.plot(range(X_train.shape[1]), scores[indices], 'o', color='gray')\n",
    "\n",
    "plt.title('Odabir znacajki uz FDR (alpha = 0.5)')\n",
    "plt.ylabel('log ANOVA p-vrijednosti')\n",
    "plt.xlim([-1, X_train.shape[1]])\n",
    "plt.xticks(range(X_train.shape[1]), breast_cancer_train.drop('target',axis=1).columns.values[indices], rotation=90)\n",
    "plt.vlines(np.sum(selector.get_support())-0.5, 0, np.max(scores), color='red', linestyle='--')\n",
    "plt.grid(True,linestyle='--')\n",
    "plt.tight_layout()\n",
    "\n",
    "# Hack za bojanje noise znacajki u crveno\n",
    "[ i.set_color(\"red\") for i in plt.gca().get_xticklabels() if i.get_text()[:5] == 'noise' ]\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Vježba 3.2.** U prošlom primjeru koristili smo podatke s kategorijskom ciljnom varijablom. Odaberite skup podataka s numeričkom ciljnom varijablom i provedite univarijantni odabir značajki koristeći neku od regresijskih evaluacijskih mjera, primjerice [`sklearn.feature_selection.f_regression`](http://scikit-learn.org/stable/modules/generated/sklearn.feature_selection.f_regression.html#sklearn.feature_selection.f_regression). Za skup podataka možete koristiti neki od skupova dostupnih u [`sklearn.datasets`](http://scikit-learn.org/dev/modules/classes.html#module-sklearn.datasets), primjerice [`sklearn.datasets.load_diabetes`](http://scikit-learn.org/dev/modules/generated/sklearn.datasets.load_diabetes.html#sklearn.datasets.load_diabetes) ili [`sklearn.datasets.load_boston`](http://scikit-learn.org/dev/modules/generated/sklearn.datasets.load_boston.html#sklearn.datasets.load_boston)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "3.2. Odabir značajki uz pomoć prediktivnog modela\n",
    "------------------------------------------------------\n",
    "\n",
    "U sljedećem primjeru izračunat ćemo važnost značajki izravno koristeći metodu slučajnih šuma (engl. random forest classifier), implementiranu u funkciji [`sklearn.ensemble.RandomForestClassifier`](http://scikit-learn.org/stable/modules/generated/sklearn.ensemble.RandomForestClassifier.html#sklearn.ensemble.RandomForestClassifier) (regresijska verzija je [`sklearn.ensemble.RandomForestRegressor`](http://scikit-learn.org/stable/modules/generated/sklearn.ensemble.RandomForestRegressor.html#sklearn.ensemble.RandomForestRegressor)), koja evaluira važnost značajki preko pozicije u stablu gdje je značajka korištena kao split node. Značajke koje se nalaze bliže korjenu stabla imaju veći utjecaj na predikciju i prema tome se mogu smatrati važnijima."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "breast_cancer = pd.read_csv('../assets/datasets/csv/breast_cancer.csv',index_col=0)\n",
    "breast_cancer_train, breast_cancer_test = train_test_split(breast_cancer, test_size=0.20, random_state=42)\n",
    "\n",
    "forest = RandomForestClassifier(n_estimators=1000,random_state=0,n_jobs=-1) # Promjeni n_jobs za paralelizaciju!\n",
    "\n",
    "X_train = breast_cancer_train.drop('target',axis=1)\n",
    "y_train = breast_cancer_train['target']\n",
    "\n",
    "forest.fit(X_train, y_train)\n",
    "importances = forest.feature_importances_\n",
    "indices = np.argsort(importances)[::-1] # Vrati indekse koji sortiraju polje (obratno)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfAAAAFgCAYAAABEyiulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nKETTgTdo0KBBgwajEP8fsbx0DzkcxjQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xb08fef0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(7,5))\n",
    "\n",
    "plt.plot(range(X_train.shape[1]), importances[indices], 'o', color='gray')\n",
    "\n",
    "plt.title('Odabir znacajki uz pomoc slucajnih suma')\n",
    "plt.ylabel('vaznost znacajke')\n",
    "plt.xlim([-1, X_train.shape[1]])\n",
    "plt.xticks(range(X_train.shape[1]), breast_cancer_train.drop('target',axis=1).columns.values[indices], rotation=90)\n",
    "plt.tight_layout()\n",
    "plt.grid(True,linestyle='--')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Vježba 3.3.** Svako stablo u ansamblu slučajnih šuma evaluira važnost svake značajke zasebno. U prijašnjem primjeru prikazali smo samo srednju vrijednost važnosti dobivenu od svih stabala. Izračunajte i prikažite standardnu devijaciju svih evaluacija. Napomena: Pojedinačna stabla su dostupna u varijabli `forest.estimators_`, i svako stablo zasebno sprema važnost svakog atributa u varijabli `feature_importances_`.\n",
    "\n",
    "```Python\n",
    "importances_std = np.std([tree.feature_importances_ for tree in forest.estimators_],axis=0)\n",
    "```\n",
    "\n",
    "Možete koristiti [`matplotlib.pyplot.errorbar`](http://matplotlib.org/api/pyplot_api.html?highlight=errorbar#matplotlib.pyplot.errorbar) za prikaz standardne devijacije.\n",
    "\n",
    "```Python\n",
    "plt.errorbar(range(X_train.shape[1]), importances[indices], yerr=importances_std[indices], fmt='o')\n",
    "```\n",
    "\n",
    "Također, pogledajte primjer na scikit-learn stranici [\"Feature importances with forests of trees\"](http://scikit-learn.org/stable/auto_examples/ensemble/plot_forest_importances.html#example-ensemble-plot-forest-importances-py)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Vježba 3.4.** Iskoristite neki drugi klasifikator za evaluaciju važnosti atributa, primjerice SVM klasifikator implementiran u [`sklearn.svm.SVC`](http://scikit-learn.org/stable/modules/generated/sklearn.svm.SVC.html). SVM računa $k-1$ zasebnih važnosti značajki, gdje je $k$ broj klasa, koje su spremljene u varijabli `coef_` (u našem skupu imamo samo dvije klase pa ćemo dobiti samo jednu procjenu važnosti za svaku od značajki). Napomena: U gornjoj implementaciji SVM-a važnosti atributa su dostupne samo za linearni kernel. Također, funkcija [`sklearn.svm.SVC`](http://scikit-learn.org/stable/modules/generated/sklearn.svm.SVC.html) koristi [LIBSVM](http://www.csie.ntu.edu.tw/~cjlin/libsvm/) implementaciju SVM-a. Kao alternativu možete koristiti funkciju [`sklearn.svm.LinearSVC`](http://scikit-learn.org/stable/modules/generated/sklearn.svm.LinearSVC.html#sklearn.svm.LinearSVC) koja koristi [LIBLINEAR](http://www.csie.ntu.edu.tw/~cjlin/liblinear/) implementaciju. Ta implementacija pruža samo linearni kernel ali nudi bolje performanse na velikom broju primjera."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.svm import SVC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "svc = SVC(kernel='linear')\n",
    "\n",
    "X_train = breast_cancer_train.drop('target',axis=1).as_matrix()\n",
    "y_train = breast_cancer_train['target'].as_matrix()\n",
    "\n",
    "svc.fit(X_train, y_train);\n",
    "\n",
    "importances_svc = svc.coef_.flatten() # Posto imamo samo dvije klase imamo samo jedan skup vaznosti znacajki\n",
    "indices_svc = np.argsort(importances_svc)[::-1] # Vrati indekse koji sortiraju polje (obratno)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfAAAAFgCAYAAABEyiulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\naydkAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xb28a780>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(7,5))\n",
    "\n",
    "plt.plot(range(X_train.shape[1]), importances_svc[indices_svc], 'o', color='gray')\n",
    "\n",
    "plt.title('Vaznosti znacajki uz pomoc SVM klasifikatora')\n",
    "plt.ylabel('vaznosti znacajki')\n",
    "plt.xlim([-1, X_train.shape[1]])\n",
    "plt.xticks(range(X_train.shape[1]), breast_cancer.columns.values, rotation=90)\n",
    "plt.tight_layout()\n",
    "plt.grid(True,linestyle='--')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "3.3. Rekurzivna eliminacija značajki\n",
    "--------------------------------------------\n",
    "\n",
    "Procedura za rekurzivnu eliminaciju značajki (engl. recursive feature elimination), također znana i kao selekcija unazad (engl. backward selection), je sljedeća. Prvo, evaluiraj važnost značajki koristeći dataset sa svim značajkama, potom iterativno eliminiraj najlošije rangirane značajke i ponovi evaluaciju. Procedura se ponavlja dok ne ostane željeni broj značajki. Primjetite da je ovo jednostavna pohlepna strategija koja ne garantira nalaženje najboljeg mogućeg skupa značajki određene veličine.\n",
    "\n",
    "Slična procedura je i selekcija unaprijed (engl. forward selection), s iznimkom da se počinje s praznim skupom značajki nakon čega se značajke dodaju jedna po jedna dok se ne distigne željeni broj. Nažalost, ova procedura nije implementirana u scikit-learnu.\n",
    "\n",
    "Koristit ćemo funkciju [`sklearn.feature_selection.RFE`](http://scikit-learn.org/stable/modules/generated/sklearn.feature_selection.RFE.html#sklearn.feature_selection.RFE) za rekurzivnu eliminaciju značajki. Sljedeći primjer je inspiriran poglavljem [\"Recursive feature elimination\"](http://scikit-learn.org/stable/auto_examples/feature_selection/plot_rfe_digits.html#example-feature-selection-plot-rfe-digits-py) iz scikit-learn korisničke dokumentacije. Rekurzivna eliminacija značajki radi na bilo kojem prediktivnom modelu koji dodjeljuje važnosti značajkama. U sljedećem primjeru koristit ćemo klasifikator slučajnih šuma implementiranom u [`sklearn.ensemble.RandomForestClassifier`](http://scikit-learn.org/stable/modules/generated/sklearn.ensemble.RandomForestClassifier.html#sklearn.ensemble.RandomForestClassifier)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.feature_selection import RFE\n",
    "from sklearn.ensemble import RandomForestClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "breast_cancer = pd.read_csv('../assets/datasets/csv/breast_cancer.csv',index_col=0)\n",
    "breast_cancer_train, breast_cancer_test = train_test_split(breast_cancer, test_size=0.20, random_state=42)\n",
    "\n",
    "X_train = breast_cancer_train.drop('target',axis=1)\n",
    "y_train = breast_cancer_train['target']\n",
    "\n",
    "rfe_rf = RFE(RandomForestClassifier(n_estimators=100,random_state=0,n_jobs=-1), n_features_to_select=10, step=1)\n",
    "rfe_rf.fit(X_train, y_train);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Znacajke odabrane s RFE (koristeci slucajne sume):\n",
      " - mean texture\n",
      " - mean area\n",
      " - mean concavity\n",
      " - mean concave points\n",
      " - worst radius\n",
      " - worst texture\n",
      " - worst perimeter\n",
      " - worst area\n",
      " - worst concavity\n",
      " - worst concave points\n"
     ]
    }
   ],
   "source": [
    "print('Znacajke odabrane s RFE (koristeci slucajne sume):')\n",
    "for i in list(breast_cancer_train.columns.values[:-1][rfe_rf.get_support()]):\n",
    "    print(' - ' + str(i))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Vježba 3.5.** Umjesto slučajnih šuma, iskoristite logističku regresiju implementiranu u [`sklearn.linear_model.LogisticRegression`](http://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html#sklearn.linear_model.LogisticRegression) i stroj s potpornim vektorima implementiranom u [`sklearn.svm.SVC`](http://scikit-learn.org/stable/modules/generated/sklearn.svm.SVC.html#sklearn.svm.SVC) za rekurzivnu eliminaciju značajki kako biste odabrali deset najvažnijih značajki. Usporedite tako dobivene značajke s onima dobivenima u prošlom primjeru."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.feature_selection import RFE\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.linear_model import LogisticRegression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "breast_cancer = pd.read_csv('../assets/datasets/csv/breast_cancer.csv',index_col=0)\n",
    "breast_cancer_train, breast_cancer_test = train_test_split(breast_cancer, test_size=0.20, random_state=42)\n",
    "\n",
    "X_train = breast_cancer_train.drop('target',axis=1)\n",
    "y_train = breast_cancer_train['target']\n",
    "\n",
    "rfe_svc = RFE(SVC(kernel='linear'), n_features_to_select=10, step=1)\n",
    "rfe_svc.fit(X_train, y_train);\n",
    "\n",
    "rfe_lr = RFE(LogisticRegression(), n_features_to_select=10, step=1)\n",
    "rfe_lr.fit(X_train, y_train);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Znacajke odabrane s RFE (koristeci SVC):\n",
      " - mean radius\n",
      " - mean concavity\n",
      " - mean concave points\n",
      " - texture error\n",
      " - worst radius\n",
      " - worst smoothness\n",
      " - worst compactness\n",
      " - worst concavity\n",
      " - worst concave points\n",
      " - worst symmetry\n",
      "\n",
      "Znacajke odabrane s RFE (koristeci logisticku regresiju):\n",
      " - mean radius\n",
      " - mean concavity\n",
      " - mean concave points\n",
      " - radius error\n",
      " - perimeter error\n",
      " - worst radius\n",
      " - worst perimeter\n",
      " - worst compactness\n",
      " - worst concavity\n",
      " - worst concave points\n"
     ]
    }
   ],
   "source": [
    "print('Znacajke odabrane s RFE (koristeci SVC):')\n",
    "for i in list(breast_cancer.columns.values[:-1][rfe_svc.get_support()]):\n",
    "    print(' - ' + str(i))\n",
    "    \n",
    "print('\\nZnacajke odabrane s RFE (koristeci logisticku regresiju):')\n",
    "for i in list(breast_cancer.columns.values[:-1][rfe_lr.get_support()]):\n",
    "    print(' - ' + str(i))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "3.4. Rekurzivna eliminacija zančajki s unakrsnom validacijom\n",
    "--------------------------------------------\n",
    "\n",
    "Prilikom rekurzivne eliminacije značajki trebali smo unaprijed definirati broj ili udio značajki koje želimo odabrati. Ali koliko značajki je optimalno? Bilo bi prikladnije kada bi algoritam mogao odabrati broj značajki za nas. Jedan od načina kako se to može napraviti je pomoću unakrsne validacije (engl. cross-validation). Važnost značajki se i dalje koristi za odabir značajki, ali optimalni broj značajki se bira unakrsnom validacijom. Za to ćemo koristiti funkciju [`sklearn.feature_selection.RFECV`](http://scikit-learn.org/stable/modules/generated/sklearn.feature_selection.RFECV.html#sklearn.feature_selection.RFECV) sa stratificiranom 2-fold unakrsnom validacijom implementiranom u [`sklearn.cross_validation.StratifiedKFold`](http://scikit-learn.org/stable/modules/generated/sklearn.cross_validation.StratifiedKFold.html#sklearn.cross_validation.StratifiedKFold). Za ostale vrste unakrsne validacije dostupne u scikit-learnu pogledajte [`module-sklearn.cross_validation`](http://scikit-learn.org/stable/modules/classes.html#module-sklearn.cross_validation). Za sličan primjer pogledajte [\"Recursive feature elimination with cross-validation\"](http://scikit-learn.org/stable/auto_examples/feature_selection/plot_rfe_with_cross_validation.html#example-feature-selection-plot-rfe-with-cross-validation-py) u scikit-learn korisničkoj dokumentaciji."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.model_selection import StratifiedKFold\n",
    "from sklearn.feature_selection import RFECV\n",
    "from sklearn.ensemble import RandomForestClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "breast_cancer = pd.read_csv('../assets/datasets/csv/breast_cancer.csv',index_col=0)\n",
    "breast_cancer_train, breast_cancer_test = train_test_split(breast_cancer, test_size=0.20, random_state=42)\n",
    "\n",
    "X_train = breast_cancer_train.drop('target',axis=1)\n",
    "y_train = breast_cancer_train['target']\n",
    "\n",
    "rfecv_rf = RFECV(RandomForestClassifier(n_estimators=100,random_state=0,n_jobs=-1), \n",
    "                 step=1, \n",
    "                 cv=StratifiedKFold(n_splits=2),\n",
    "                 scoring='accuracy')\n",
    "rfecv_rf.fit(X_train, y_train);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Odabrane znacajke s RFECV (koristeci slucajne sume):\n",
      " - mean radius\n",
      " - mean texture\n",
      " - mean perimeter\n",
      " - mean area\n",
      " - mean smoothness\n",
      " - mean compactness\n",
      " - mean concavity\n",
      " - mean concave points\n",
      " - radius error\n",
      " - perimeter error\n",
      " - area error\n",
      " - compactness error\n",
      " - concavity error\n",
      " - symmetry error\n",
      " - fractal dimension error\n",
      " - worst radius\n",
      " - worst texture\n",
      " - worst perimeter\n",
      " - worst area\n",
      " - worst smoothness\n",
      " - worst compactness\n",
      " - worst concavity\n",
      " - worst concave points\n",
      " - worst symmetry\n",
      " - worst fractal dimension\n"
     ]
    }
   ],
   "source": [
    "print('Odabrane znacajke s RFECV (koristeci slucajne sume):')\n",
    "for i in list(breast_cancer.columns.values[:-1][rfecv_rf.get_support()]):\n",
    "    print(' - ' + str(i))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAcMAAAFNCAYAAAB8PAR2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ns0zZF5lg9QAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xb378198>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(7,5))\n",
    "\n",
    "plt.plot(range(1, len(rfecv_rf.grid_scores_) + 1), rfecv_rf.grid_scores_, '-o',color='gray')\n",
    "plt.xlabel('Broj odabranih znacajki')\n",
    "plt.ylabel('tocnost na unakrsnoj validaciji')\n",
    "plt.title('Rekurzivna eliminacija znacajki s unakrsnom validacijom')\n",
    "plt.vlines(rfecv_rf.n_features_, \n",
    "           np.min(rfecv_rf.grid_scores_), \n",
    "           rfecv_rf.grid_scores_[rfecv_rf.n_features_-1], \n",
    "           color='red', linestyle='--')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Vježba 3.6.** Umjesto slučajnih šuma i unakrsne validacije iskoristite neki drugi klasifikator i tip unakrsne validacije. Primjerice, za klasifikator možete iskoristiti SVM klasifikator implementiran u funkciji [`sklearn.svm.SVC`](http://scikit-learn.org/stable/modules/generated/sklearn.svm.SVC.html#sklearn.svm.SVC) ili logističku regresiju implementiranu u funkciji [`sklearn.linear_model.LogisticRegression`](http://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html#sklearn.linear_model.LogisticRegression). Za unakrsnu validaciju možete iskoristiti leave-one-out unakrsnu validaciju implementiranu u [`sklearn.cross_validation.LeaveOneOut`](http://scikit-learn.org/stable/modules/generated/sklearn.cross_validation.LeaveOneOut.html#sklearn.cross_validation.LeaveOneOut) ili shuffle-split validaciju implementiranu u [`sklearn.cross_validation.ShuffleSplit`](http://scikit-learn.org/stable/modules/generated/sklearn.cross_validation.ShuffleSplit.html#sklearn.cross_validation.ShuffleSplit). Pogledajte [`module-sklearn.cross_validation`](http://scikit-learn.org/stable/modules/classes.html#module-sklearn.cross_validation) za sve vrste unakrsne validacije dostupne u scikit-learnu."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import StratifiedKFold\n",
    "from sklearn.feature_selection import RFECV\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.linear_model import LogisticRegression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "breast_cancer = pd.read_csv('../assets/datasets/csv/breast_cancer.csv',index_col=0)\n",
    "breast_cancer_train, breast_cancer_test = train_test_split(breast_cancer, test_size=0.20, random_state=42)\n",
    "\n",
    "X_train = breast_cancer_train.drop('target',axis=1)\n",
    "y_train = breast_cancer_train['target']\n",
    "\n",
    "rfecv_svc = RFECV(SVC(kernel='linear'), \n",
    "                 step=1, \n",
    "                 cv=StratifiedKFold(n_splits=2),\n",
    "                 scoring='accuracy')\n",
    "rfecv_svc.fit(X_train, y_train);\n",
    "\n",
    "rfecv_lr = RFECV(LogisticRegression(), \n",
    "                 step=1, \n",
    "                 cv=StratifiedKFold(n_splits=2),\n",
    "                 scoring='accuracy')\n",
    "rfecv_lr.fit(X_train, y_train);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Odabrane znacajke s RFECV (koristeci SVC):\n",
      " - mean radius\n",
      " - mean concavity\n",
      " - mean concave points\n",
      " - texture error\n",
      " - worst radius\n",
      " - worst smoothness\n",
      " - worst compactness\n",
      " - worst concavity\n",
      " - worst concave points\n",
      " - worst symmetry\n",
      "\n",
      "Odabrane znacajke s RFECV (koristeci logisticku regresiju):\n",
      " - mean radius\n",
      " - mean texture\n",
      " - mean perimeter\n",
      " - mean area\n",
      " - mean smoothness\n",
      " - mean compactness\n",
      " - mean concavity\n",
      " - mean concave points\n",
      " - mean symmetry\n",
      " - mean fractal dimension\n",
      " - radius error\n",
      " - texture error\n",
      " - perimeter error\n",
      " - area error\n",
      " - smoothness error\n",
      " - compactness error\n",
      " - concavity error\n",
      " - concave points error\n",
      " - symmetry error\n",
      " - fractal dimension error\n",
      " - worst radius\n",
      " - worst texture\n",
      " - worst perimeter\n",
      " - worst area\n",
      " - worst smoothness\n",
      " - worst compactness\n",
      " - worst concavity\n",
      " - worst concave points\n",
      " - worst symmetry\n",
      " - worst fractal dimension\n"
     ]
    }
   ],
   "source": [
    "print('\\nOdabrane znacajke s RFECV (koristeci SVC):')\n",
    "for i in list(breast_cancer.columns.values[:-1][rfecv_svc.get_support()]):\n",
    "    print(' - ' + str(i))\n",
    "    \n",
    "print('\\nOdabrane znacajke s RFECV (koristeci logisticku regresiju):')\n",
    "for i in list(breast_cancer.columns.values[:-1][rfecv_lr.get_support()]):\n",
    "    print(' - ' + str(i))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA4IAAAFNCAYAAABVKNEpAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nVjL/BQNkke44T4u3AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xb510278>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(1, 2, figsize=(15, 5))\n",
    "\n",
    "ax[0].plot(range(1, len(rfecv_svc.grid_scores_) + 1), rfecv_svc.grid_scores_, '-o', color='gray')\n",
    "ax[0].set(xlabel='Broj odabranih znacajki',ylabel='tocnost na unakrsnoj validaciji',title='SVC')\n",
    "ax[0].vlines(rfecv_svc.n_features_, \n",
    "             np.min(rfecv_svc.grid_scores_), \n",
    "             rfecv_svc.grid_scores_[rfecv_svc.n_features_-1], \n",
    "             color='red', linestyle='--')\n",
    "\n",
    "ax[1].plot(range(1, len(rfecv_lr.grid_scores_) + 1), rfecv_lr.grid_scores_, '-o', color='gray')\n",
    "ax[1].set(xlabel='Broj odabranih znacajki',ylabel='tocnost na unakrsnoj validaciji',title='logisticka regresija')\n",
    "ax[1].vlines(rfecv_lr.n_features_, \n",
    "             np.min(rfecv_lr.grid_scores_), \n",
    "             rfecv_lr.grid_scores_[rfecv_lr.n_features_-1], \n",
    "             color='red', linestyle='--')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Literatura\n",
    "-----------\n",
    "\n",
    "<a name=\"scikit-feature-selection\">[1]</a> *\"scikit-learn documentation - 1.13. Feature selection\"*, [link](http://scikit-learn.org/stable/modules/feature_selection.html#feature-selection)\n",
    "\n",
    "<a name=\"pyml-ch4\">[2]</a> *\"Python Machine Learning Code examples - Chapter 4 - Building Good Training\n",
    "Sets – Data Preprocessing\"*, [link](https://github.com/rasbt/python-machine-learning-book/blob/master/code/ch04/ch04.ipynb)\n",
    "\n",
    "<a name=\"intro-ml-py-ch5\">[3]</a> *\"Introduction to Machine Learning with Python - Chapter 5 - Representing data and engineering features\"* \n",
    "\n",
    "<a name=\"diving-into-data\">[4]</a> *\"Diving into data - Selecting good features\"*, [part 1](http://blog.datadive.net/selecting-good-features-part-i-univariate-selection/), [part 2](http://blog.datadive.net/selecting-good-features-part-ii-linear-models-and-regularization/), [part 3](http://blog.datadive.net/selecting-good-features-part-iii-random-forests/), [part 4](http://blog.datadive.net/selecting-good-features-part-iv-stability-selection-rfe-and-everything-side-by-side/) "
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}