BMD-KNN-GitHub.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# KNN - BMD Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn import preprocessing, neighbors\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Importing CSV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "bmd = pd.read_csv('../../bmd/bmd.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "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>id</th>\n",
       "      <th>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>fracture</th>\n",
       "      <th>weight_kg</th>\n",
       "      <th>height_cm</th>\n",
       "      <th>medication</th>\n",
       "      <th>waiting_time</th>\n",
       "      <th>bmd</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>469</td>\n",
       "      <td>57.052768</td>\n",
       "      <td>F</td>\n",
       "      <td>no fracture</td>\n",
       "      <td>64.0</td>\n",
       "      <td>155.5</td>\n",
       "      <td>Anticonvulsant</td>\n",
       "      <td>18</td>\n",
       "      <td>0.8793</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8724</td>\n",
       "      <td>75.741225</td>\n",
       "      <td>F</td>\n",
       "      <td>no fracture</td>\n",
       "      <td>78.0</td>\n",
       "      <td>162.0</td>\n",
       "      <td>No medication</td>\n",
       "      <td>56</td>\n",
       "      <td>0.7946</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6736</td>\n",
       "      <td>70.778900</td>\n",
       "      <td>M</td>\n",
       "      <td>no fracture</td>\n",
       "      <td>73.0</td>\n",
       "      <td>170.5</td>\n",
       "      <td>No medication</td>\n",
       "      <td>10</td>\n",
       "      <td>0.9067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>24180</td>\n",
       "      <td>78.247175</td>\n",
       "      <td>F</td>\n",
       "      <td>no fracture</td>\n",
       "      <td>60.0</td>\n",
       "      <td>148.0</td>\n",
       "      <td>No medication</td>\n",
       "      <td>14</td>\n",
       "      <td>0.7112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>17072</td>\n",
       "      <td>54.191877</td>\n",
       "      <td>M</td>\n",
       "      <td>no fracture</td>\n",
       "      <td>55.0</td>\n",
       "      <td>161.0</td>\n",
       "      <td>No medication</td>\n",
       "      <td>20</td>\n",
       "      <td>0.7909</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id        age sex     fracture  weight_kg  height_cm      medication  \\\n",
       "0    469  57.052768   F  no fracture       64.0      155.5  Anticonvulsant   \n",
       "1   8724  75.741225   F  no fracture       78.0      162.0   No medication   \n",
       "2   6736  70.778900   M  no fracture       73.0      170.5   No medication   \n",
       "3  24180  78.247175   F  no fracture       60.0      148.0   No medication   \n",
       "4  17072  54.191877   M  no fracture       55.0      161.0   No medication   \n",
       "\n",
       "   waiting_time     bmd  \n",
       "0            18  0.8793  \n",
       "1            56  0.7946  \n",
       "2            10  0.9067  \n",
       "3            14  0.7112  \n",
       "4            20  0.7909  "
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bmd.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Analyzing numeric columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "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>id</th>\n",
       "      <th>age</th>\n",
       "      <th>weight_kg</th>\n",
       "      <th>height_cm</th>\n",
       "      <th>waiting_time</th>\n",
       "      <th>bmd</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>169.000000</td>\n",
       "      <td>169.000000</td>\n",
       "      <td>169.000000</td>\n",
       "      <td>169.000000</td>\n",
       "      <td>169.000000</td>\n",
       "      <td>169.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>9102.556213</td>\n",
       "      <td>63.631531</td>\n",
       "      <td>64.665680</td>\n",
       "      <td>160.254438</td>\n",
       "      <td>19.739645</td>\n",
       "      <td>0.783104</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>8744.623598</td>\n",
       "      <td>12.356936</td>\n",
       "      <td>11.537171</td>\n",
       "      <td>7.928272</td>\n",
       "      <td>15.800570</td>\n",
       "      <td>0.166529</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>35.000000</td>\n",
       "      <td>35.814058</td>\n",
       "      <td>36.000000</td>\n",
       "      <td>142.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>0.407600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>2018.000000</td>\n",
       "      <td>54.424211</td>\n",
       "      <td>56.000000</td>\n",
       "      <td>154.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>0.670800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>6702.000000</td>\n",
       "      <td>63.487837</td>\n",
       "      <td>64.500000</td>\n",
       "      <td>160.500000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>0.786100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>17100.000000</td>\n",
       "      <td>72.080558</td>\n",
       "      <td>73.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>0.888800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>24208.000000</td>\n",
       "      <td>88.753795</td>\n",
       "      <td>96.000000</td>\n",
       "      <td>177.000000</td>\n",
       "      <td>96.000000</td>\n",
       "      <td>1.362400</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 id         age   weight_kg   height_cm  waiting_time  \\\n",
       "count    169.000000  169.000000  169.000000  169.000000    169.000000   \n",
       "mean    9102.556213   63.631531   64.665680  160.254438     19.739645   \n",
       "std     8744.623598   12.356936   11.537171    7.928272     15.800570   \n",
       "min       35.000000   35.814058   36.000000  142.000000      5.000000   \n",
       "25%     2018.000000   54.424211   56.000000  154.000000      9.000000   \n",
       "50%     6702.000000   63.487837   64.500000  160.500000     14.000000   \n",
       "75%    17100.000000   72.080558   73.000000  166.000000     24.000000   \n",
       "max    24208.000000   88.753795   96.000000  177.000000     96.000000   \n",
       "\n",
       "              bmd  \n",
       "count  169.000000  \n",
       "mean     0.783104  \n",
       "std      0.166529  \n",
       "min      0.407600  \n",
       "25%      0.670800  \n",
       "50%      0.786100  \n",
       "75%      0.888800  \n",
       "max      1.362400  "
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bmd.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "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>id</th>\n",
       "      <th>age</th>\n",
       "      <th>weight_kg</th>\n",
       "      <th>height_cm</th>\n",
       "      <th>waiting_time</th>\n",
       "      <th>bmd</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fracture</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>fracture</th>\n",
       "      <td>7869.040000</td>\n",
       "      <td>69.771320</td>\n",
       "      <td>58.410000</td>\n",
       "      <td>159.290000</td>\n",
       "      <td>23.300000</td>\n",
       "      <td>0.623308</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>no fracture</th>\n",
       "      <td>9620.840336</td>\n",
       "      <td>61.051787</td>\n",
       "      <td>67.294118</td>\n",
       "      <td>160.659664</td>\n",
       "      <td>18.243697</td>\n",
       "      <td>0.850245</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      id        age  weight_kg   height_cm  waiting_time  \\\n",
       "fracture                                                                   \n",
       "fracture     7869.040000  69.771320  58.410000  159.290000     23.300000   \n",
       "no fracture  9620.840336  61.051787  67.294118  160.659664     18.243697   \n",
       "\n",
       "                  bmd  \n",
       "fracture               \n",
       "fracture     0.623308  \n",
       "no fracture  0.850245  "
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bmd.groupby('fracture').mean()\n",
    "\n",
    "# id: To be removed\n",
    "# Age: The oldest you get, the biggest the probability to fracture.\n",
    "# Weight: A higher weight helps preventing a fracture.\n",
    "# Height: It does not really impact.\n",
    "# Waiting time (time the patient had to wait for the densitometry (in minutes): the longest the time, the more probability to fracture.\n",
    "# Bdm (bone mineral density): The most dense your bones ares, the less likely they are to fracture."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Analyzing string columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "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>sex</th>\n",
       "      <th>F</th>\n",
       "      <th>M</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fracture</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>fracture</th>\n",
       "      <td>25</td>\n",
       "      <td>25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>no fracture</th>\n",
       "      <td>58</td>\n",
       "      <td>61</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "sex           F   M\n",
       "fracture           \n",
       "fracture     25  25\n",
       "no fracture  58  61"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cross_sex = pd.crosstab(bmd['fracture'], bmd['sex'])\n",
    "cross_sex\n",
    "\n",
    "# There is not a lot of difference between Men and Women"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.30120481927710846"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Female_Fracture = cross_sex['F']['fracture'] / (cross_sex['F']['fracture']  + cross_sex['F']['no fracture'])\n",
    "Female_Fracture"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.29069767441860467"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Male_Fracture = cross_sex['M']['fracture'] / (cross_sex['M']['fracture']  + cross_sex['M']['no fracture'])\n",
    "Male_Fracture"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "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>medication</th>\n",
       "      <th>Anticonvulsant</th>\n",
       "      <th>Glucocorticoids</th>\n",
       "      <th>No medication</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fracture</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>fracture</th>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>no fracture</th>\n",
       "      <td>5</td>\n",
       "      <td>19</td>\n",
       "      <td>95</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "medication   Anticonvulsant  Glucocorticoids  No medication\n",
       "fracture                                                   \n",
       "fracture                  4                5             41\n",
       "no fracture               5               19             95"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cross_med = pd.crosstab(bmd['fracture'], bmd['medication'])\n",
    "cross_med\n",
    "\n",
    "# Medication prevents fractures, especially for Gluco. For Anticonvulsant there are not many cases."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.4444444444444444"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Anticonvulsant_fracture = cross_med['Anticonvulsant']['fracture'] / (cross_med['Anticonvulsant']['fracture']  \n",
    "+ cross_med['Anticonvulsant']['no fracture'])\n",
    "\n",
    "Anticonvulsant_fracture"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.20833333333333334"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Glucocorticoids_fracture = cross_med['Glucocorticoids']['fracture'] / (cross_med['Glucocorticoids']['fracture']  \n",
    "+ cross_med['Glucocorticoids']['no fracture'])\n",
    "\n",
    "Glucocorticoids_fracture"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.3014705882352941"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "No_medication_fracture = cross_med['No medication']['fracture'] / (cross_med['No medication']['fracture']  \n",
    "+ cross_med['No medication']['no fracture'])\n",
    "\n",
    "No_medication_fracture"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Removing columns \"id\" and sending \"fracture\" column to the end."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['id',\n",
       " 'age',\n",
       " 'sex',\n",
       " 'fracture',\n",
       " 'weight_kg',\n",
       " 'height_cm',\n",
       " 'medication',\n",
       " 'waiting_time',\n",
       " 'bmd']"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cols = bmd.columns.values.tolist()\n",
    "cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_new = ['age',\n",
    " 'sex',\n",
    " 'weight_kg',\n",
    " 'height_cm',\n",
    " 'medication',\n",
    " 'waiting_time',\n",
    " 'bmd',\n",
    " 'fracture']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "bmd = bmd[cols_new]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "dum_sex = pd.get_dummies(bmd['sex'], prefix = 'sex')\n",
    "new_bmd = bmd.join(dum_sex)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "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>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>weight_kg</th>\n",
       "      <th>height_cm</th>\n",
       "      <th>medication</th>\n",
       "      <th>waiting_time</th>\n",
       "      <th>bmd</th>\n",
       "      <th>fracture</th>\n",
       "      <th>sex_F</th>\n",
       "      <th>sex_M</th>\n",
       "      <th>medication_Anticonvulsant</th>\n",
       "      <th>medication_Glucocorticoids</th>\n",
       "      <th>medication_No medication</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>57.052768</td>\n",
       "      <td>F</td>\n",
       "      <td>64.0</td>\n",
       "      <td>155.5</td>\n",
       "      <td>Anticonvulsant</td>\n",
       "      <td>18</td>\n",
       "      <td>0.8793</td>\n",
       "      <td>no fracture</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>75.741225</td>\n",
       "      <td>F</td>\n",
       "      <td>78.0</td>\n",
       "      <td>162.0</td>\n",
       "      <td>No medication</td>\n",
       "      <td>56</td>\n",
       "      <td>0.7946</td>\n",
       "      <td>no fracture</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>70.778900</td>\n",
       "      <td>M</td>\n",
       "      <td>73.0</td>\n",
       "      <td>170.5</td>\n",
       "      <td>No medication</td>\n",
       "      <td>10</td>\n",
       "      <td>0.9067</td>\n",
       "      <td>no fracture</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         age sex  weight_kg  height_cm      medication  waiting_time     bmd  \\\n",
       "0  57.052768   F       64.0      155.5  Anticonvulsant            18  0.8793   \n",
       "1  75.741225   F       78.0      162.0   No medication            56  0.7946   \n",
       "2  70.778900   M       73.0      170.5   No medication            10  0.9067   \n",
       "\n",
       "      fracture  sex_F  sex_M  medication_Anticonvulsant  \\\n",
       "0  no fracture      1      0                          1   \n",
       "1  no fracture      1      0                          0   \n",
       "2  no fracture      0      1                          0   \n",
       "\n",
       "   medication_Glucocorticoids  medication_No medication  \n",
       "0                           0                         0  \n",
       "1                           0                         1  \n",
       "2                           0                         1  "
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dum_med = pd.get_dummies(new_bmd['medication'], prefix = 'medication')\n",
    "new_bmd = new_bmd.join(dum_med)\n",
    "new_bmd.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['age',\n",
       " 'sex',\n",
       " 'weight_kg',\n",
       " 'height_cm',\n",
       " 'medication',\n",
       " 'waiting_time',\n",
       " 'bmd',\n",
       " 'fracture',\n",
       " 'sex_F',\n",
       " 'sex_M',\n",
       " 'medication_Anticonvulsant',\n",
       " 'medication_Glucocorticoids',\n",
       " 'medication_No medication']"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cols = new_bmd.columns.values.tolist()\n",
    "cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_new = ['age',\n",
    " 'weight_kg',\n",
    " 'height_cm',\n",
    " 'waiting_time',\n",
    " 'bmd',\n",
    " 'fracture',\n",
    " 'sex_F',\n",
    " 'sex_M',\n",
    " 'medication_Anticonvulsant',\n",
    " 'medication_Glucocorticoids',\n",
    " 'medication_No medication']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "new_bmd = new_bmd[cols_new]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>age</th>\n",
       "      <th>weight_kg</th>\n",
       "      <th>height_cm</th>\n",
       "      <th>waiting_time</th>\n",
       "      <th>bmd</th>\n",
       "      <th>fracture</th>\n",
       "      <th>sex_F</th>\n",
       "      <th>sex_M</th>\n",
       "      <th>medication_Anticonvulsant</th>\n",
       "      <th>medication_Glucocorticoids</th>\n",
       "      <th>medication_No medication</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>57.052768</td>\n",
       "      <td>64.0</td>\n",
       "      <td>155.5</td>\n",
       "      <td>18</td>\n",
       "      <td>0.8793</td>\n",
       "      <td>no fracture</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>75.741225</td>\n",
       "      <td>78.0</td>\n",
       "      <td>162.0</td>\n",
       "      <td>56</td>\n",
       "      <td>0.7946</td>\n",
       "      <td>no fracture</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>70.778900</td>\n",
       "      <td>73.0</td>\n",
       "      <td>170.5</td>\n",
       "      <td>10</td>\n",
       "      <td>0.9067</td>\n",
       "      <td>no fracture</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>78.247175</td>\n",
       "      <td>60.0</td>\n",
       "      <td>148.0</td>\n",
       "      <td>14</td>\n",
       "      <td>0.7112</td>\n",
       "      <td>no fracture</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>54.191877</td>\n",
       "      <td>55.0</td>\n",
       "      <td>161.0</td>\n",
       "      <td>20</td>\n",
       "      <td>0.7909</td>\n",
       "      <td>no fracture</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         age  weight_kg  height_cm  waiting_time     bmd     fracture  sex_F  \\\n",
       "0  57.052768       64.0      155.5            18  0.8793  no fracture      1   \n",
       "1  75.741225       78.0      162.0            56  0.7946  no fracture      1   \n",
       "2  70.778900       73.0      170.5            10  0.9067  no fracture      0   \n",
       "3  78.247175       60.0      148.0            14  0.7112  no fracture      1   \n",
       "4  54.191877       55.0      161.0            20  0.7909  no fracture      0   \n",
       "\n",
       "   sex_M  medication_Anticonvulsant  medication_Glucocorticoids  \\\n",
       "0      0                          1                           0   \n",
       "1      0                          0                           0   \n",
       "2      1                          0                           0   \n",
       "3      0                          0                           0   \n",
       "4      1                          0                           0   \n",
       "\n",
       "   medication_No medication  \n",
       "0                         0  \n",
       "1                         1  \n",
       "2                         1  \n",
       "3                         1  \n",
       "4                         1  "
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_bmd.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "# We change the Fracture columns from string to numeric. True=1, False=0.\n",
    "\n",
    "new_bmd['fracture'] = (new_bmd['fracture'] == 'fracture').astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "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>age</th>\n",
       "      <th>weight_kg</th>\n",
       "      <th>height_cm</th>\n",
       "      <th>waiting_time</th>\n",
       "      <th>bmd</th>\n",
       "      <th>fracture</th>\n",
       "      <th>sex_F</th>\n",
       "      <th>sex_M</th>\n",
       "      <th>medication_Anticonvulsant</th>\n",
       "      <th>medication_Glucocorticoids</th>\n",
       "      <th>medication_No medication</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>57.052768</td>\n",
       "      <td>64.0</td>\n",
       "      <td>155.5</td>\n",
       "      <td>18</td>\n",
       "      <td>0.8793</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>75.741225</td>\n",
       "      <td>78.0</td>\n",
       "      <td>162.0</td>\n",
       "      <td>56</td>\n",
       "      <td>0.7946</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>70.778900</td>\n",
       "      <td>73.0</td>\n",
       "      <td>170.5</td>\n",
       "      <td>10</td>\n",
       "      <td>0.9067</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>78.247175</td>\n",
       "      <td>60.0</td>\n",
       "      <td>148.0</td>\n",
       "      <td>14</td>\n",
       "      <td>0.7112</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>54.191877</td>\n",
       "      <td>55.0</td>\n",
       "      <td>161.0</td>\n",
       "      <td>20</td>\n",
       "      <td>0.7909</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         age  weight_kg  height_cm  waiting_time     bmd  fracture  sex_F  \\\n",
       "0  57.052768       64.0      155.5            18  0.8793         0      1   \n",
       "1  75.741225       78.0      162.0            56  0.7946         0      1   \n",
       "2  70.778900       73.0      170.5            10  0.9067         0      0   \n",
       "3  78.247175       60.0      148.0            14  0.7112         0      1   \n",
       "4  54.191877       55.0      161.0            20  0.7909         0      0   \n",
       "\n",
       "   sex_M  medication_Anticonvulsant  medication_Glucocorticoids  \\\n",
       "0      0                          1                           0   \n",
       "1      0                          0                           0   \n",
       "2      1                          0                           0   \n",
       "3      0                          0                           0   \n",
       "4      1                          0                           0   \n",
       "\n",
       "   medication_No medication  \n",
       "0                         0  \n",
       "1                         1  \n",
       "2                         1  \n",
       "3                         1  \n",
       "4                         1  "
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_bmd.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    119\n",
       "1     50\n",
       "Name: fracture, dtype: int64"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_bmd.fracture.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "# Select columns to build the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['age',\n",
       " 'weight_kg',\n",
       " 'height_cm',\n",
       " 'waiting_time',\n",
       " 'bmd',\n",
       " 'fracture',\n",
       " 'sex_F',\n",
       " 'sex_M',\n",
       " 'medication_Anticonvulsant',\n",
       " 'medication_Glucocorticoids',\n",
       " 'medication_No medication']"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_bmd.columns.values.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = ['age',\n",
    " 'weight_kg',\n",
    " 'height_cm',\n",
    " 'waiting_time',\n",
    " 'bmd',\n",
    " 'sex_F',\n",
    " 'sex_M',\n",
    " 'medication_Anticonvulsant',\n",
    " 'medication_Glucocorticoids',\n",
    " 'medication_No medication']\n",
    "\n",
    "Y = ['fracture']\n",
    "\n",
    "# We put under X the columns to be used as predictors.\n",
    "# Y is the column we want to predict"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# We evaluate dtypes and assign numeric type to them"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "age                           float64\n",
       "weight_kg                     float64\n",
       "height_cm                     float64\n",
       "waiting_time                    int64\n",
       "bmd                           float64\n",
       "fracture                        int32\n",
       "sex_F                           uint8\n",
       "sex_M                           uint8\n",
       "medication_Anticonvulsant       uint8\n",
       "medication_Glucocorticoids      uint8\n",
       "medication_No medication        uint8\n",
       "dtype: object"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_bmd.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Usuario\\anaconda3\\lib\\site-packages\\pandas\\core\\ops\\array_ops.py:253: FutureWarning: elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison\n",
      "  res_values = method(rvalues)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "age                           float64\n",
       "weight_kg                     float64\n",
       "height_cm                     float64\n",
       "waiting_time                    int64\n",
       "bmd                           float64\n",
       "fracture                        int32\n",
       "sex_F                           int32\n",
       "sex_M                           int32\n",
       "medication_Anticonvulsant       int32\n",
       "medication_Glucocorticoids      int32\n",
       "medication_No medication        int32\n",
       "dtype: object"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_bmd['sex_F'] = (new_bmd['sex_F'] == 'sex_F').astype(int)\n",
    "new_bmd['sex_M'] = (new_bmd['sex_M'] == 'sex_M').astype(int)\n",
    "new_bmd['medication_Anticonvulsant'] = (new_bmd['medication_Anticonvulsant'] == 'medication_Anticonvulsant').astype(int)\n",
    "new_bmd['medication_Glucocorticoids'] = (new_bmd['medication_Glucocorticoids'] == 'medication_Glucocorticoids').astype(int)\n",
    "new_bmd['medication_No medication'] = (new_bmd['medication_No medication'] == 'medication_No medication').astype(int) \n",
    "new_bmd.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Splitting DataSet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, Y_train, Y_test = train_test_split(new_bmd[X], new_bmd[Y], test_size=0.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "clf = neighbors.KNeighborsClassifier()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Usuario\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:1: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  \"\"\"Entry point for launching an IPython kernel.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='minkowski',\n",
       "                     metric_params=None, n_jobs=None, n_neighbors=5, p=2,\n",
       "                     weights='uniform')"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.fit(X_train, Y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [],
   "source": [
    "accuracy = clf.score(X_test, Y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7450980392156863"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Fine Tunning our model (using 1000 iterations)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# We will finetune our model using two methods.\n",
    "# We first will choose the best combination of columns to be used as predictors.\n",
    "# Then we will change the parameters contained inside KNeighborsClassifier (algorithm, leaf_size, metric, metric_params, \n",
    "# n_neighbors, p, weights)\n",
    "# We will always run 1000 times each option and calculate the mean score."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Choose the right columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "A = ['age', 'weight_kg', 'height_cm', 'waiting_time', 'bmd', 'sex_F', 'sex_M', 'medication_Anticonvulsant', 'medication_Glucocorticoids', 'medication_No medication']\n",
    "B = ['age', 'weight_kg', 'height_cm', 'waiting_time', 'bmd', 'medication_Anticonvulsant', 'medication_Glucocorticoids', 'medication_No medication']\n",
    "C = ['age', 'weight_kg', 'height_cm', 'bmd']\n",
    "D = ['age', 'bmd']\n",
    "list = A,B,C,D"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Columns: ['age', 'weight_kg', 'height_cm', 'waiting_time', 'bmd', 'sex_F', 'sex_M', 'medication_Anticonvulsant', 'medication_Glucocorticoids', 'medication_No medication'] - Result: 0.7321764705882339\n",
      "Columns: ['age', 'weight_kg', 'height_cm', 'waiting_time', 'bmd', 'medication_Anticonvulsant', 'medication_Glucocorticoids', 'medication_No medication'] - Result: 0.7325686274509816\n",
      "Columns: ['age', 'weight_kg', 'height_cm', 'bmd'] - Result: 0.7330522875816999\n",
      "Columns: ['age', 'bmd'] - Result: 0.7302352941176258\n"
     ]
    }
   ],
   "source": [
    "accuracyList = []\n",
    "for i in list:\n",
    "    X = new_bmd[i]\n",
    "    Y = new_bmd['fracture']\n",
    "    for r in range(1,1001):\n",
    "        X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.3)\n",
    "        clf = neighbors.KNeighborsClassifier()\n",
    "        clf.fit(X_train, Y_train)\n",
    "        accuracy = clf.score(X_test, Y_test)\n",
    "        accuracyList.append(accuracy)\n",
    "        result = sum(accuracyList)/len(accuracyList)\n",
    "    print('Columns: ' + str(i) + \" - Result: \" + str(result))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The best score is obtained using the variables 'age', 'weight_kg', 'height_cm', 'bmd' as predictors."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Choosing the best value for \"n_neighbors\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = new_bmd[['age', 'weight_kg', 'height_cm', 'bmd']]\n",
    "Y = new_bmd['fracture']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "A = 5\n",
    "B = 11\n",
    "C = 19\n",
    "D = 25\n",
    "E = 35\n",
    "F = 45\n",
    "list = A,B,C,D,E,F"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "N_Neighbors: 5 - Result: 0.7353921568627452\n",
      "N_Neighbors: 11 - Result: 0.7530784313725433\n",
      "N_Neighbors: 19 - Result: 0.762444444444441\n",
      "N_Neighbors: 25 - Result: 0.7661617647058919\n",
      "N_Neighbors: 35 - Result: 0.7655137254902098\n",
      "N_Neighbors: 45 - Result: 0.7597973856209491\n"
     ]
    }
   ],
   "source": [
    "accuracyList = []\n",
    "for i in list:\n",
    "    for r in range(1,1001):\n",
    "        X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.3)\n",
    "        clf = neighbors.KNeighborsClassifier(n_neighbors = i)\n",
    "        clf.fit(X_train, Y_train)\n",
    "        accuracy = clf.score(X_test, Y_test)\n",
    "        accuracyList.append(accuracy)\n",
    "        result = sum(accuracyList)/len(accuracyList)\n",
    "    print('N_Neighbors: ' + str(i) + \" - Result: \" + str(result))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The best score is obtained with the value \"25\" for the n_neighbors parameter"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Choosing the best value for \"algorithm\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "A = \"auto\"\n",
    "B = \"ball_tree\"\n",
    "C = \"kd_tree\"\n",
    "D = \"brute\"\n",
    "list = A,B,C,D"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Algorithm: auto - Result: 0.7774117647058819\n",
      "Algorithm: ball_tree - Result: 0.7774117647058755\n",
      "Algorithm: kd_tree - Result: 0.7779411764705864\n",
      "Algorithm: brute - Result: 0.7773676470588345\n"
     ]
    }
   ],
   "source": [
    "accuracyList = []\n",
    "for i in list:\n",
    "    for r in range(1,1001):\n",
    "        X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.3)\n",
    "        clf = neighbors.KNeighborsClassifier(algorithm = i, n_neighbors = 25)\n",
    "        clf.fit(X_train, Y_train)\n",
    "        accuracy = clf.score(X_test, Y_test)\n",
    "        accuracyList.append(accuracy)\n",
    "        result = sum(accuracyList)/len(accuracyList)\n",
    "    print('Algorithm: ' + str(i) + \" - Result: \" + str(result))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "# None of the results is much higher than the others. We will keep the default option (auto)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Other parameters, like 'algorithm','leaf_size','metric','metric_params','n_job','weight' and 'p' do not improve the model. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Our final model "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Final model: brute - Result: 0.7774869281046202\n"
     ]
    }
   ],
   "source": [
    "for r in range(1,1001):\n",
    "    X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.3)\n",
    "    clf = neighbors.KNeighborsClassifier(n_neighbors = 25)\n",
    "    clf.fit(X_train, Y_train)\n",
    "    accuracy = clf.score(X_test, Y_test)\n",
    "    accuracyList.append(accuracy)\n",
    "    result = sum(accuracyList)/len(accuracyList)\n",
    "print('Final model: ' + str(i) + \" - Result: \" + str(result))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}