├── README.md
├── cover_fifa_22.jpg
├── fifa.ipynb
└── players_22.csv


/README.md:
--------------------------------------------------------------------------------
 1 | # K-Means Clustering
 2 | 
 3 | # ![fifa2022](https://i0.wp.com/gnova.com.ar/wp-content/uploads/2021/10/ea-fifa-22-cover-kylian-mbappe_1qeaco87s803l13iu0tnr84jhq.jpg)
 4 |   
 5 | 
 6 | Building a K-means clustering algorithm from scratch and using it to cluster the FIFA22 data.
 7 | 
 8 | Clustering is an unsupervised machine learning technique that can find patterns in your data. K-means is one of the most popular forms of clustering.
 9 | 
10 | We'll create our algorithm using python and pandas. We'll then compare it to the reference implementation from scikit-learn.
11 | 
12 | Project Steps
13 | 
14 | * Write out pseudocode for the algorithm
15 | * Code the k-means algorithm
16 | * Plot the clusters from the algorithm
17 | * Compare performance to the scikit-learn algorithm
18 | 
19 | # K-means overview
20 | 
21 | K-means is an unsupervised machine learning technique that allow us to cluster data points. This enables us to find patterns in the data that can help us analyze it more effectively. K-means is an iterative algorithm, which means that it will converge to the optimal clustering over time.
22 | 
23 | To run a k-means clustering:
24 | 
25 | * Specify the number of clusters you want (usually referred to as `k`).
26 | * Randomly initialize the centroid for each cluster. The centroid is the data point that is in the center of the cluster.
27 | * Determine which data points belong to which cluster by finding the closest `centroid` to each data point.
28 | * Update the centroids based on the geometric mean of all the data points in the cluster.
29 | * Run 3 and 4 until the `centroids` stop changing. Each run is referred to as an iteration.
30 | 
31 | # Code
32 | 
33 | You can find the code for this project [here](https://github.com/taureanjoe/kmeans-clustering).
34 | 
35 | ## Data
36 | 
37 | We'll be using data from FIFA, which you can download [here](https://www.kaggle.com/datasets/stefanoleone992/fifa-22-complete-player-dataset?select=players_22.csv). We'll use the file `players_22.csv`.
38 | 


--------------------------------------------------------------------------------
/cover_fifa_22.jpg:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/taureanjoe/kmeans-clustering/6317826da46702012622970ee99312d9da507125/cover_fifa_22.jpg


--------------------------------------------------------------------------------
/fifa.ipynb:
--------------------------------------------------------------------------------
   1 | {
   2 |  "cells": [
   3 |   {
   4 |    "cell_type": "markdown",
   5 |    "metadata": {},
   6 |    "source": [
   7 |     "We are going to cluster the FIFA data dataset using our own K-means clustering algorithm.\n",
   8 |     "\n",
   9 |     "# ![fifa2022](cover_fifa_22.jpg)\n",
  10 |     "\n",
  11 |     "The datasets provided include the players data for the Career Mode from FIFA 15 to FIFA 22 (\"players_22.csv\"). The data allows multiple comparisons for the same players across the last 8 version of the videogame.\n",
  12 |     "\n",
  13 |     "### Content\n",
  14 |     "\n",
  15 |     "* Every player available in FIFA 15, 16, 17, 18, 19, 20, 21, and also FIFA 22\n",
  16 |     "* 100+ attributes\n",
  17 |     "* URL of the scraped players\n",
  18 |     "* URL of the uploaded player faces, club and nation logos\n",
  19 |     "* Player positions, with the role in the club and in the national team\n",
  20 |     "* Player attributes with statistics as Attacking, Skills, Defense, Mentality, GK Skills, etc.\n",
  21 |     "* Player personal data like Nationality, Club, DateOfBirth, Wage, Salary, etc.\n",
  22 |     "\n",
  23 |     "Data has been scraped from the publicly available website [sofifa.com](https://sofifa.com/)."
  24 |    ]
  25 |   },
  26 |   {
  27 |    "cell_type": "code",
  28 |    "execution_count": 2,
  29 |    "metadata": {},
  30 |    "outputs": [],
  31 |    "source": [
  32 |     "import pandas as pd\n",
  33 |     "import numpy as np"
  34 |    ]
  35 |   },
  36 |   {
  37 |    "cell_type": "code",
  38 |    "execution_count": 18,
  39 |    "metadata": {},
  40 |    "outputs": [
  41 |     {
  42 |      "name": "stderr",
  43 |      "output_type": "stream",
  44 |      "text": [
  45 |       "/var/folders/tm/5w0c8x0d2rzck8tf0hvstwf40000gn/T/ipykernel_5324/2061566770.py:1: DtypeWarning: Columns (25,108) have mixed types. Specify dtype option on import or set low_memory=False.\n",
  46 |       "  players = pd.read_csv(\"players_22.csv\")\n"
  47 |      ]
  48 |     }
  49 |    ],
  50 |    "source": [
  51 |     "players = pd.read_csv(\"players_22.csv\")"
  52 |    ]
  53 |   },
  54 |   {
  55 |    "cell_type": "code",
  56 |    "execution_count": 19,
  57 |    "metadata": {},
  58 |    "outputs": [
  59 |     {
  60 |      "data": {
  61 |       "text/html": [
  62 |        "<div>\n",
  63 |        "<style scoped>\n",
  64 |        "    .dataframe tbody tr th:only-of-type {\n",
  65 |        "        vertical-align: middle;\n",
  66 |        "    }\n",
  67 |        "\n",
  68 |        "    .dataframe tbody tr th {\n",
  69 |        "        vertical-align: top;\n",
  70 |        "    }\n",
  71 |        "\n",
  72 |        "    .dataframe thead th {\n",
  73 |        "        text-align: right;\n",
  74 |        "    }\n",
  75 |        "</style>\n",
  76 |        "<table border=\"1\" class=\"dataframe\">\n",
  77 |        "  <thead>\n",
  78 |        "    <tr style=\"text-align: right;\">\n",
  79 |        "      <th></th>\n",
  80 |        "      <th>sofifa_id</th>\n",
  81 |        "      <th>player_url</th>\n",
  82 |        "      <th>short_name</th>\n",
  83 |        "      <th>long_name</th>\n",
  84 |        "      <th>player_positions</th>\n",
  85 |        "      <th>overall</th>\n",
  86 |        "      <th>potential</th>\n",
  87 |        "      <th>value_eur</th>\n",
  88 |        "      <th>wage_eur</th>\n",
  89 |        "      <th>age</th>\n",
  90 |        "      <th>...</th>\n",
  91 |        "      <th>lcb</th>\n",
  92 |        "      <th>cb</th>\n",
  93 |        "      <th>rcb</th>\n",
  94 |        "      <th>rb</th>\n",
  95 |        "      <th>gk</th>\n",
  96 |        "      <th>player_face_url</th>\n",
  97 |        "      <th>club_logo_url</th>\n",
  98 |        "      <th>club_flag_url</th>\n",
  99 |        "      <th>nation_logo_url</th>\n",
 100 |        "      <th>nation_flag_url</th>\n",
 101 |        "    </tr>\n",
 102 |        "  </thead>\n",
 103 |        "  <tbody>\n",
 104 |        "    <tr>\n",
 105 |        "      <th>0</th>\n",
 106 |        "      <td>158023</td>\n",
 107 |        "      <td>https://sofifa.com/player/158023/lionel-messi/...</td>\n",
 108 |        "      <td>L. Messi</td>\n",
 109 |        "      <td>Lionel Andrés Messi Cuccittini</td>\n",
 110 |        "      <td>RW, ST, CF</td>\n",
 111 |        "      <td>93</td>\n",
 112 |        "      <td>93</td>\n",
 113 |        "      <td>78000000.0</td>\n",
 114 |        "      <td>320000.0</td>\n",
 115 |        "      <td>34</td>\n",
 116 |        "      <td>...</td>\n",
 117 |        "      <td>50+3</td>\n",
 118 |        "      <td>50+3</td>\n",
 119 |        "      <td>50+3</td>\n",
 120 |        "      <td>61+3</td>\n",
 121 |        "      <td>19+3</td>\n",
 122 |        "      <td>https://cdn.sofifa.net/players/158/023/22_120.png</td>\n",
 123 |        "      <td>https://cdn.sofifa.net/teams/73/60.png</td>\n",
 124 |        "      <td>https://cdn.sofifa.net/flags/fr.png</td>\n",
 125 |        "      <td>https://cdn.sofifa.net/teams/1369/60.png</td>\n",
 126 |        "      <td>https://cdn.sofifa.net/flags/ar.png</td>\n",
 127 |        "    </tr>\n",
 128 |        "    <tr>\n",
 129 |        "      <th>1</th>\n",
 130 |        "      <td>188545</td>\n",
 131 |        "      <td>https://sofifa.com/player/188545/robert-lewand...</td>\n",
 132 |        "      <td>R. Lewandowski</td>\n",
 133 |        "      <td>Robert Lewandowski</td>\n",
 134 |        "      <td>ST</td>\n",
 135 |        "      <td>92</td>\n",
 136 |        "      <td>92</td>\n",
 137 |        "      <td>119500000.0</td>\n",
 138 |        "      <td>270000.0</td>\n",
 139 |        "      <td>32</td>\n",
 140 |        "      <td>...</td>\n",
 141 |        "      <td>60+3</td>\n",
 142 |        "      <td>60+3</td>\n",
 143 |        "      <td>60+3</td>\n",
 144 |        "      <td>61+3</td>\n",
 145 |        "      <td>19+3</td>\n",
 146 |        "      <td>https://cdn.sofifa.net/players/188/545/22_120.png</td>\n",
 147 |        "      <td>https://cdn.sofifa.net/teams/21/60.png</td>\n",
 148 |        "      <td>https://cdn.sofifa.net/flags/de.png</td>\n",
 149 |        "      <td>https://cdn.sofifa.net/teams/1353/60.png</td>\n",
 150 |        "      <td>https://cdn.sofifa.net/flags/pl.png</td>\n",
 151 |        "    </tr>\n",
 152 |        "    <tr>\n",
 153 |        "      <th>2</th>\n",
 154 |        "      <td>20801</td>\n",
 155 |        "      <td>https://sofifa.com/player/20801/c-ronaldo-dos-...</td>\n",
 156 |        "      <td>Cristiano Ronaldo</td>\n",
 157 |        "      <td>Cristiano Ronaldo dos Santos Aveiro</td>\n",
 158 |        "      <td>ST, LW</td>\n",
 159 |        "      <td>91</td>\n",
 160 |        "      <td>91</td>\n",
 161 |        "      <td>45000000.0</td>\n",
 162 |        "      <td>270000.0</td>\n",
 163 |        "      <td>36</td>\n",
 164 |        "      <td>...</td>\n",
 165 |        "      <td>53+3</td>\n",
 166 |        "      <td>53+3</td>\n",
 167 |        "      <td>53+3</td>\n",
 168 |        "      <td>60+3</td>\n",
 169 |        "      <td>20+3</td>\n",
 170 |        "      <td>https://cdn.sofifa.net/players/020/801/22_120.png</td>\n",
 171 |        "      <td>https://cdn.sofifa.net/teams/11/60.png</td>\n",
 172 |        "      <td>https://cdn.sofifa.net/flags/gb-eng.png</td>\n",
 173 |        "      <td>https://cdn.sofifa.net/teams/1354/60.png</td>\n",
 174 |        "      <td>https://cdn.sofifa.net/flags/pt.png</td>\n",
 175 |        "    </tr>\n",
 176 |        "    <tr>\n",
 177 |        "      <th>3</th>\n",
 178 |        "      <td>190871</td>\n",
 179 |        "      <td>https://sofifa.com/player/190871/neymar-da-sil...</td>\n",
 180 |        "      <td>Neymar Jr</td>\n",
 181 |        "      <td>Neymar da Silva Santos Júnior</td>\n",
 182 |        "      <td>LW, CAM</td>\n",
 183 |        "      <td>91</td>\n",
 184 |        "      <td>91</td>\n",
 185 |        "      <td>129000000.0</td>\n",
 186 |        "      <td>270000.0</td>\n",
 187 |        "      <td>29</td>\n",
 188 |        "      <td>...</td>\n",
 189 |        "      <td>50+3</td>\n",
 190 |        "      <td>50+3</td>\n",
 191 |        "      <td>50+3</td>\n",
 192 |        "      <td>62+3</td>\n",
 193 |        "      <td>20+3</td>\n",
 194 |        "      <td>https://cdn.sofifa.net/players/190/871/22_120.png</td>\n",
 195 |        "      <td>https://cdn.sofifa.net/teams/73/60.png</td>\n",
 196 |        "      <td>https://cdn.sofifa.net/flags/fr.png</td>\n",
 197 |        "      <td>NaN</td>\n",
 198 |        "      <td>https://cdn.sofifa.net/flags/br.png</td>\n",
 199 |        "    </tr>\n",
 200 |        "    <tr>\n",
 201 |        "      <th>4</th>\n",
 202 |        "      <td>192985</td>\n",
 203 |        "      <td>https://sofifa.com/player/192985/kevin-de-bruy...</td>\n",
 204 |        "      <td>K. De Bruyne</td>\n",
 205 |        "      <td>Kevin De Bruyne</td>\n",
 206 |        "      <td>CM, CAM</td>\n",
 207 |        "      <td>91</td>\n",
 208 |        "      <td>91</td>\n",
 209 |        "      <td>125500000.0</td>\n",
 210 |        "      <td>350000.0</td>\n",
 211 |        "      <td>30</td>\n",
 212 |        "      <td>...</td>\n",
 213 |        "      <td>69+3</td>\n",
 214 |        "      <td>69+3</td>\n",
 215 |        "      <td>69+3</td>\n",
 216 |        "      <td>75+3</td>\n",
 217 |        "      <td>21+3</td>\n",
 218 |        "      <td>https://cdn.sofifa.net/players/192/985/22_120.png</td>\n",
 219 |        "      <td>https://cdn.sofifa.net/teams/10/60.png</td>\n",
 220 |        "      <td>https://cdn.sofifa.net/flags/gb-eng.png</td>\n",
 221 |        "      <td>https://cdn.sofifa.net/teams/1325/60.png</td>\n",
 222 |        "      <td>https://cdn.sofifa.net/flags/be.png</td>\n",
 223 |        "    </tr>\n",
 224 |        "    <tr>\n",
 225 |        "      <th>...</th>\n",
 226 |        "      <td>...</td>\n",
 227 |        "      <td>...</td>\n",
 228 |        "      <td>...</td>\n",
 229 |        "      <td>...</td>\n",
 230 |        "      <td>...</td>\n",
 231 |        "      <td>...</td>\n",
 232 |        "      <td>...</td>\n",
 233 |        "      <td>...</td>\n",
 234 |        "      <td>...</td>\n",
 235 |        "      <td>...</td>\n",
 236 |        "      <td>...</td>\n",
 237 |        "      <td>...</td>\n",
 238 |        "      <td>...</td>\n",
 239 |        "      <td>...</td>\n",
 240 |        "      <td>...</td>\n",
 241 |        "      <td>...</td>\n",
 242 |        "      <td>...</td>\n",
 243 |        "      <td>...</td>\n",
 244 |        "      <td>...</td>\n",
 245 |        "      <td>...</td>\n",
 246 |        "      <td>...</td>\n",
 247 |        "    </tr>\n",
 248 |        "    <tr>\n",
 249 |        "      <th>19234</th>\n",
 250 |        "      <td>261962</td>\n",
 251 |        "      <td>https://sofifa.com/player/261962/defu-song/220002</td>\n",
 252 |        "      <td>Song Defu</td>\n",
 253 |        "      <td>宋德福</td>\n",
 254 |        "      <td>CDM</td>\n",
 255 |        "      <td>47</td>\n",
 256 |        "      <td>52</td>\n",
 257 |        "      <td>70000.0</td>\n",
 258 |        "      <td>1000.0</td>\n",
 259 |        "      <td>22</td>\n",
 260 |        "      <td>...</td>\n",
 261 |        "      <td>46+2</td>\n",
 262 |        "      <td>46+2</td>\n",
 263 |        "      <td>46+2</td>\n",
 264 |        "      <td>48+2</td>\n",
 265 |        "      <td>15+2</td>\n",
 266 |        "      <td>https://cdn.sofifa.net/players/261/962/22_120.png</td>\n",
 267 |        "      <td>https://cdn.sofifa.net/teams/112541/60.png</td>\n",
 268 |        "      <td>https://cdn.sofifa.net/flags/cn.png</td>\n",
 269 |        "      <td>NaN</td>\n",
 270 |        "      <td>https://cdn.sofifa.net/flags/cn.png</td>\n",
 271 |        "    </tr>\n",
 272 |        "    <tr>\n",
 273 |        "      <th>19235</th>\n",
 274 |        "      <td>262040</td>\n",
 275 |        "      <td>https://sofifa.com/player/262040/caoimhin-port...</td>\n",
 276 |        "      <td>C. Porter</td>\n",
 277 |        "      <td>Caoimhin Porter</td>\n",
 278 |        "      <td>CM</td>\n",
 279 |        "      <td>47</td>\n",
 280 |        "      <td>59</td>\n",
 281 |        "      <td>110000.0</td>\n",
 282 |        "      <td>500.0</td>\n",
 283 |        "      <td>19</td>\n",
 284 |        "      <td>...</td>\n",
 285 |        "      <td>44+2</td>\n",
 286 |        "      <td>44+2</td>\n",
 287 |        "      <td>44+2</td>\n",
 288 |        "      <td>48+2</td>\n",
 289 |        "      <td>14+2</td>\n",
 290 |        "      <td>https://cdn.sofifa.net/players/262/040/22_120.png</td>\n",
 291 |        "      <td>https://cdn.sofifa.net/teams/445/60.png</td>\n",
 292 |        "      <td>https://cdn.sofifa.net/flags/ie.png</td>\n",
 293 |        "      <td>NaN</td>\n",
 294 |        "      <td>https://cdn.sofifa.net/flags/ie.png</td>\n",
 295 |        "    </tr>\n",
 296 |        "    <tr>\n",
 297 |        "      <th>19236</th>\n",
 298 |        "      <td>262760</td>\n",
 299 |        "      <td>https://sofifa.com/player/262760/nathan-logue/...</td>\n",
 300 |        "      <td>N. Logue</td>\n",
 301 |        "      <td>Nathan Logue-Cunningham</td>\n",
 302 |        "      <td>CM</td>\n",
 303 |        "      <td>47</td>\n",
 304 |        "      <td>55</td>\n",
 305 |        "      <td>100000.0</td>\n",
 306 |        "      <td>500.0</td>\n",
 307 |        "      <td>21</td>\n",
 308 |        "      <td>...</td>\n",
 309 |        "      <td>45+2</td>\n",
 310 |        "      <td>45+2</td>\n",
 311 |        "      <td>45+2</td>\n",
 312 |        "      <td>47+2</td>\n",
 313 |        "      <td>12+2</td>\n",
 314 |        "      <td>https://cdn.sofifa.net/players/262/760/22_120.png</td>\n",
 315 |        "      <td>https://cdn.sofifa.net/teams/111131/60.png</td>\n",
 316 |        "      <td>https://cdn.sofifa.net/flags/ie.png</td>\n",
 317 |        "      <td>NaN</td>\n",
 318 |        "      <td>https://cdn.sofifa.net/flags/ie.png</td>\n",
 319 |        "    </tr>\n",
 320 |        "    <tr>\n",
 321 |        "      <th>19237</th>\n",
 322 |        "      <td>262820</td>\n",
 323 |        "      <td>https://sofifa.com/player/262820/luke-rudden/2...</td>\n",
 324 |        "      <td>L. Rudden</td>\n",
 325 |        "      <td>Luke Rudden</td>\n",
 326 |        "      <td>ST</td>\n",
 327 |        "      <td>47</td>\n",
 328 |        "      <td>60</td>\n",
 329 |        "      <td>110000.0</td>\n",
 330 |        "      <td>500.0</td>\n",
 331 |        "      <td>19</td>\n",
 332 |        "      <td>...</td>\n",
 333 |        "      <td>26+2</td>\n",
 334 |        "      <td>26+2</td>\n",
 335 |        "      <td>26+2</td>\n",
 336 |        "      <td>32+2</td>\n",
 337 |        "      <td>15+2</td>\n",
 338 |        "      <td>https://cdn.sofifa.net/players/262/820/22_120.png</td>\n",
 339 |        "      <td>https://cdn.sofifa.net/teams/111131/60.png</td>\n",
 340 |        "      <td>https://cdn.sofifa.net/flags/ie.png</td>\n",
 341 |        "      <td>NaN</td>\n",
 342 |        "      <td>https://cdn.sofifa.net/flags/ie.png</td>\n",
 343 |        "    </tr>\n",
 344 |        "    <tr>\n",
 345 |        "      <th>19238</th>\n",
 346 |        "      <td>264540</td>\n",
 347 |        "      <td>https://sofifa.com/player/264540/emanuel-lalch...</td>\n",
 348 |        "      <td>E. Lalchhanchhuaha</td>\n",
 349 |        "      <td>Emanuel Lalchhanchhuaha</td>\n",
 350 |        "      <td>CAM</td>\n",
 351 |        "      <td>47</td>\n",
 352 |        "      <td>60</td>\n",
 353 |        "      <td>110000.0</td>\n",
 354 |        "      <td>500.0</td>\n",
 355 |        "      <td>19</td>\n",
 356 |        "      <td>...</td>\n",
 357 |        "      <td>41+2</td>\n",
 358 |        "      <td>41+2</td>\n",
 359 |        "      <td>41+2</td>\n",
 360 |        "      <td>45+2</td>\n",
 361 |        "      <td>16+2</td>\n",
 362 |        "      <td>https://cdn.sofifa.net/players/264/540/22_120.png</td>\n",
 363 |        "      <td>https://cdn.sofifa.net/teams/113040/60.png</td>\n",
 364 |        "      <td>https://cdn.sofifa.net/flags/in.png</td>\n",
 365 |        "      <td>NaN</td>\n",
 366 |        "      <td>https://cdn.sofifa.net/flags/in.png</td>\n",
 367 |        "    </tr>\n",
 368 |        "  </tbody>\n",
 369 |        "</table>\n",
 370 |        "<p>19239 rows × 110 columns</p>\n",
 371 |        "</div>"
 372 |       ],
 373 |       "text/plain": [
 374 |        "       sofifa_id                                         player_url  \\\n",
 375 |        "0         158023  https://sofifa.com/player/158023/lionel-messi/...   \n",
 376 |        "1         188545  https://sofifa.com/player/188545/robert-lewand...   \n",
 377 |        "2          20801  https://sofifa.com/player/20801/c-ronaldo-dos-...   \n",
 378 |        "3         190871  https://sofifa.com/player/190871/neymar-da-sil...   \n",
 379 |        "4         192985  https://sofifa.com/player/192985/kevin-de-bruy...   \n",
 380 |        "...          ...                                                ...   \n",
 381 |        "19234     261962  https://sofifa.com/player/261962/defu-song/220002   \n",
 382 |        "19235     262040  https://sofifa.com/player/262040/caoimhin-port...   \n",
 383 |        "19236     262760  https://sofifa.com/player/262760/nathan-logue/...   \n",
 384 |        "19237     262820  https://sofifa.com/player/262820/luke-rudden/2...   \n",
 385 |        "19238     264540  https://sofifa.com/player/264540/emanuel-lalch...   \n",
 386 |        "\n",
 387 |        "               short_name                            long_name  \\\n",
 388 |        "0                L. Messi       Lionel Andrés Messi Cuccittini   \n",
 389 |        "1          R. Lewandowski                   Robert Lewandowski   \n",
 390 |        "2       Cristiano Ronaldo  Cristiano Ronaldo dos Santos Aveiro   \n",
 391 |        "3               Neymar Jr        Neymar da Silva Santos Júnior   \n",
 392 |        "4            K. De Bruyne                      Kevin De Bruyne   \n",
 393 |        "...                   ...                                  ...   \n",
 394 |        "19234           Song Defu                                  宋德福   \n",
 395 |        "19235           C. Porter                      Caoimhin Porter   \n",
 396 |        "19236            N. Logue              Nathan Logue-Cunningham   \n",
 397 |        "19237           L. Rudden                          Luke Rudden   \n",
 398 |        "19238  E. Lalchhanchhuaha              Emanuel Lalchhanchhuaha   \n",
 399 |        "\n",
 400 |        "      player_positions  overall  potential    value_eur  wage_eur  age  ...  \\\n",
 401 |        "0           RW, ST, CF       93         93   78000000.0  320000.0   34  ...   \n",
 402 |        "1                   ST       92         92  119500000.0  270000.0   32  ...   \n",
 403 |        "2               ST, LW       91         91   45000000.0  270000.0   36  ...   \n",
 404 |        "3              LW, CAM       91         91  129000000.0  270000.0   29  ...   \n",
 405 |        "4              CM, CAM       91         91  125500000.0  350000.0   30  ...   \n",
 406 |        "...                ...      ...        ...          ...       ...  ...  ...   \n",
 407 |        "19234              CDM       47         52      70000.0    1000.0   22  ...   \n",
 408 |        "19235               CM       47         59     110000.0     500.0   19  ...   \n",
 409 |        "19236               CM       47         55     100000.0     500.0   21  ...   \n",
 410 |        "19237               ST       47         60     110000.0     500.0   19  ...   \n",
 411 |        "19238              CAM       47         60     110000.0     500.0   19  ...   \n",
 412 |        "\n",
 413 |        "        lcb    cb   rcb    rb    gk  \\\n",
 414 |        "0      50+3  50+3  50+3  61+3  19+3   \n",
 415 |        "1      60+3  60+3  60+3  61+3  19+3   \n",
 416 |        "2      53+3  53+3  53+3  60+3  20+3   \n",
 417 |        "3      50+3  50+3  50+3  62+3  20+3   \n",
 418 |        "4      69+3  69+3  69+3  75+3  21+3   \n",
 419 |        "...     ...   ...   ...   ...   ...   \n",
 420 |        "19234  46+2  46+2  46+2  48+2  15+2   \n",
 421 |        "19235  44+2  44+2  44+2  48+2  14+2   \n",
 422 |        "19236  45+2  45+2  45+2  47+2  12+2   \n",
 423 |        "19237  26+2  26+2  26+2  32+2  15+2   \n",
 424 |        "19238  41+2  41+2  41+2  45+2  16+2   \n",
 425 |        "\n",
 426 |        "                                         player_face_url  \\\n",
 427 |        "0      https://cdn.sofifa.net/players/158/023/22_120.png   \n",
 428 |        "1      https://cdn.sofifa.net/players/188/545/22_120.png   \n",
 429 |        "2      https://cdn.sofifa.net/players/020/801/22_120.png   \n",
 430 |        "3      https://cdn.sofifa.net/players/190/871/22_120.png   \n",
 431 |        "4      https://cdn.sofifa.net/players/192/985/22_120.png   \n",
 432 |        "...                                                  ...   \n",
 433 |        "19234  https://cdn.sofifa.net/players/261/962/22_120.png   \n",
 434 |        "19235  https://cdn.sofifa.net/players/262/040/22_120.png   \n",
 435 |        "19236  https://cdn.sofifa.net/players/262/760/22_120.png   \n",
 436 |        "19237  https://cdn.sofifa.net/players/262/820/22_120.png   \n",
 437 |        "19238  https://cdn.sofifa.net/players/264/540/22_120.png   \n",
 438 |        "\n",
 439 |        "                                    club_logo_url  \\\n",
 440 |        "0          https://cdn.sofifa.net/teams/73/60.png   \n",
 441 |        "1          https://cdn.sofifa.net/teams/21/60.png   \n",
 442 |        "2          https://cdn.sofifa.net/teams/11/60.png   \n",
 443 |        "3          https://cdn.sofifa.net/teams/73/60.png   \n",
 444 |        "4          https://cdn.sofifa.net/teams/10/60.png   \n",
 445 |        "...                                           ...   \n",
 446 |        "19234  https://cdn.sofifa.net/teams/112541/60.png   \n",
 447 |        "19235     https://cdn.sofifa.net/teams/445/60.png   \n",
 448 |        "19236  https://cdn.sofifa.net/teams/111131/60.png   \n",
 449 |        "19237  https://cdn.sofifa.net/teams/111131/60.png   \n",
 450 |        "19238  https://cdn.sofifa.net/teams/113040/60.png   \n",
 451 |        "\n",
 452 |        "                                 club_flag_url  \\\n",
 453 |        "0          https://cdn.sofifa.net/flags/fr.png   \n",
 454 |        "1          https://cdn.sofifa.net/flags/de.png   \n",
 455 |        "2      https://cdn.sofifa.net/flags/gb-eng.png   \n",
 456 |        "3          https://cdn.sofifa.net/flags/fr.png   \n",
 457 |        "4      https://cdn.sofifa.net/flags/gb-eng.png   \n",
 458 |        "...                                        ...   \n",
 459 |        "19234      https://cdn.sofifa.net/flags/cn.png   \n",
 460 |        "19235      https://cdn.sofifa.net/flags/ie.png   \n",
 461 |        "19236      https://cdn.sofifa.net/flags/ie.png   \n",
 462 |        "19237      https://cdn.sofifa.net/flags/ie.png   \n",
 463 |        "19238      https://cdn.sofifa.net/flags/in.png   \n",
 464 |        "\n",
 465 |        "                                nation_logo_url  \\\n",
 466 |        "0      https://cdn.sofifa.net/teams/1369/60.png   \n",
 467 |        "1      https://cdn.sofifa.net/teams/1353/60.png   \n",
 468 |        "2      https://cdn.sofifa.net/teams/1354/60.png   \n",
 469 |        "3                                           NaN   \n",
 470 |        "4      https://cdn.sofifa.net/teams/1325/60.png   \n",
 471 |        "...                                         ...   \n",
 472 |        "19234                                       NaN   \n",
 473 |        "19235                                       NaN   \n",
 474 |        "19236                                       NaN   \n",
 475 |        "19237                                       NaN   \n",
 476 |        "19238                                       NaN   \n",
 477 |        "\n",
 478 |        "                           nation_flag_url  \n",
 479 |        "0      https://cdn.sofifa.net/flags/ar.png  \n",
 480 |        "1      https://cdn.sofifa.net/flags/pl.png  \n",
 481 |        "2      https://cdn.sofifa.net/flags/pt.png  \n",
 482 |        "3      https://cdn.sofifa.net/flags/br.png  \n",
 483 |        "4      https://cdn.sofifa.net/flags/be.png  \n",
 484 |        "...                                    ...  \n",
 485 |        "19234  https://cdn.sofifa.net/flags/cn.png  \n",
 486 |        "19235  https://cdn.sofifa.net/flags/ie.png  \n",
 487 |        "19236  https://cdn.sofifa.net/flags/ie.png  \n",
 488 |        "19237  https://cdn.sofifa.net/flags/ie.png  \n",
 489 |        "19238  https://cdn.sofifa.net/flags/in.png  \n",
 490 |        "\n",
 491 |        "[19239 rows x 110 columns]"
 492 |       ]
 493 |      },
 494 |      "execution_count": 19,
 495 |      "metadata": {},
 496 |      "output_type": "execute_result"
 497 |     }
 498 |    ],
 499 |    "source": [
 500 |     "players"
 501 |    ]
 502 |   },
 503 |   {
 504 |    "cell_type": "markdown",
 505 |    "metadata": {},
 506 |    "source": [
 507 |     "Let's list the features based on which we will cluster our players\n",
 508 |     "\n",
 509 |     "* `overall` - overall rating of the player\n",
 510 |     "* `potential` - potential rating of the player\n",
 511 |     "* `value_eur` - total value of the player to the Club in Euros\n",
 512 |     "* `wage_eur` - wage of the player in Euros\n",
 513 |     "* `age` - age of the player"
 514 |    ]
 515 |   },
 516 |   {
 517 |    "cell_type": "code",
 518 |    "execution_count": 20,
 519 |    "metadata": {},
 520 |    "outputs": [],
 521 |    "source": [
 522 |     "features = [\"overall\", \"potential\", \"wage_eur\", \"value_eur\", \"age\"]"
 523 |    ]
 524 |   },
 525 |   {
 526 |    "cell_type": "code",
 527 |    "execution_count": 21,
 528 |    "metadata": {},
 529 |    "outputs": [],
 530 |    "source": [
 531 |     "# dropping null values from columns mentioned in features\n",
 532 |     "players = players.dropna(subset=features)\n",
 533 |     "# this will make sure that there will be no missing values in the columns that we are clustering"
 534 |    ]
 535 |   },
 536 |   {
 537 |    "cell_type": "code",
 538 |    "execution_count": 22,
 539 |    "metadata": {},
 540 |    "outputs": [],
 541 |    "source": [
 542 |     "data = players[features].copy()"
 543 |    ]
 544 |   },
 545 |   {
 546 |    "cell_type": "code",
 547 |    "execution_count": 23,
 548 |    "metadata": {},
 549 |    "outputs": [
 550 |     {
 551 |      "data": {
 552 |       "text/html": [
 553 |        "<div>\n",
 554 |        "<style scoped>\n",
 555 |        "    .dataframe tbody tr th:only-of-type {\n",
 556 |        "        vertical-align: middle;\n",
 557 |        "    }\n",
 558 |        "\n",
 559 |        "    .dataframe tbody tr th {\n",
 560 |        "        vertical-align: top;\n",
 561 |        "    }\n",
 562 |        "\n",
 563 |        "    .dataframe thead th {\n",
 564 |        "        text-align: right;\n",
 565 |        "    }\n",
 566 |        "</style>\n",
 567 |        "<table border=\"1\" class=\"dataframe\">\n",
 568 |        "  <thead>\n",
 569 |        "    <tr style=\"text-align: right;\">\n",
 570 |        "      <th></th>\n",
 571 |        "      <th>overall</th>\n",
 572 |        "      <th>potential</th>\n",
 573 |        "      <th>wage_eur</th>\n",
 574 |        "      <th>value_eur</th>\n",
 575 |        "      <th>age</th>\n",
 576 |        "    </tr>\n",
 577 |        "  </thead>\n",
 578 |        "  <tbody>\n",
 579 |        "    <tr>\n",
 580 |        "      <th>0</th>\n",
 581 |        "      <td>93</td>\n",
 582 |        "      <td>93</td>\n",
 583 |        "      <td>320000.0</td>\n",
 584 |        "      <td>78000000.0</td>\n",
 585 |        "      <td>34</td>\n",
 586 |        "    </tr>\n",
 587 |        "    <tr>\n",
 588 |        "      <th>1</th>\n",
 589 |        "      <td>92</td>\n",
 590 |        "      <td>92</td>\n",
 591 |        "      <td>270000.0</td>\n",
 592 |        "      <td>119500000.0</td>\n",
 593 |        "      <td>32</td>\n",
 594 |        "    </tr>\n",
 595 |        "    <tr>\n",
 596 |        "      <th>2</th>\n",
 597 |        "      <td>91</td>\n",
 598 |        "      <td>91</td>\n",
 599 |        "      <td>270000.0</td>\n",
 600 |        "      <td>45000000.0</td>\n",
 601 |        "      <td>36</td>\n",
 602 |        "    </tr>\n",
 603 |        "    <tr>\n",
 604 |        "      <th>3</th>\n",
 605 |        "      <td>91</td>\n",
 606 |        "      <td>91</td>\n",
 607 |        "      <td>270000.0</td>\n",
 608 |        "      <td>129000000.0</td>\n",
 609 |        "      <td>29</td>\n",
 610 |        "    </tr>\n",
 611 |        "    <tr>\n",
 612 |        "      <th>4</th>\n",
 613 |        "      <td>91</td>\n",
 614 |        "      <td>91</td>\n",
 615 |        "      <td>350000.0</td>\n",
 616 |        "      <td>125500000.0</td>\n",
 617 |        "      <td>30</td>\n",
 618 |        "    </tr>\n",
 619 |        "    <tr>\n",
 620 |        "      <th>...</th>\n",
 621 |        "      <td>...</td>\n",
 622 |        "      <td>...</td>\n",
 623 |        "      <td>...</td>\n",
 624 |        "      <td>...</td>\n",
 625 |        "      <td>...</td>\n",
 626 |        "    </tr>\n",
 627 |        "    <tr>\n",
 628 |        "      <th>19234</th>\n",
 629 |        "      <td>47</td>\n",
 630 |        "      <td>52</td>\n",
 631 |        "      <td>1000.0</td>\n",
 632 |        "      <td>70000.0</td>\n",
 633 |        "      <td>22</td>\n",
 634 |        "    </tr>\n",
 635 |        "    <tr>\n",
 636 |        "      <th>19235</th>\n",
 637 |        "      <td>47</td>\n",
 638 |        "      <td>59</td>\n",
 639 |        "      <td>500.0</td>\n",
 640 |        "      <td>110000.0</td>\n",
 641 |        "      <td>19</td>\n",
 642 |        "    </tr>\n",
 643 |        "    <tr>\n",
 644 |        "      <th>19236</th>\n",
 645 |        "      <td>47</td>\n",
 646 |        "      <td>55</td>\n",
 647 |        "      <td>500.0</td>\n",
 648 |        "      <td>100000.0</td>\n",
 649 |        "      <td>21</td>\n",
 650 |        "    </tr>\n",
 651 |        "    <tr>\n",
 652 |        "      <th>19237</th>\n",
 653 |        "      <td>47</td>\n",
 654 |        "      <td>60</td>\n",
 655 |        "      <td>500.0</td>\n",
 656 |        "      <td>110000.0</td>\n",
 657 |        "      <td>19</td>\n",
 658 |        "    </tr>\n",
 659 |        "    <tr>\n",
 660 |        "      <th>19238</th>\n",
 661 |        "      <td>47</td>\n",
 662 |        "      <td>60</td>\n",
 663 |        "      <td>500.0</td>\n",
 664 |        "      <td>110000.0</td>\n",
 665 |        "      <td>19</td>\n",
 666 |        "    </tr>\n",
 667 |        "  </tbody>\n",
 668 |        "</table>\n",
 669 |        "<p>19165 rows × 5 columns</p>\n",
 670 |        "</div>"
 671 |       ],
 672 |       "text/plain": [
 673 |        "       overall  potential  wage_eur    value_eur  age\n",
 674 |        "0           93         93  320000.0   78000000.0   34\n",
 675 |        "1           92         92  270000.0  119500000.0   32\n",
 676 |        "2           91         91  270000.0   45000000.0   36\n",
 677 |        "3           91         91  270000.0  129000000.0   29\n",
 678 |        "4           91         91  350000.0  125500000.0   30\n",
 679 |        "...        ...        ...       ...          ...  ...\n",
 680 |        "19234       47         52    1000.0      70000.0   22\n",
 681 |        "19235       47         59     500.0     110000.0   19\n",
 682 |        "19236       47         55     500.0     100000.0   21\n",
 683 |        "19237       47         60     500.0     110000.0   19\n",
 684 |        "19238       47         60     500.0     110000.0   19\n",
 685 |        "\n",
 686 |        "[19165 rows x 5 columns]"
 687 |       ]
 688 |      },
 689 |      "execution_count": 23,
 690 |      "metadata": {},
 691 |      "output_type": "execute_result"
 692 |     }
 693 |    ],
 694 |    "source": [
 695 |     "data"
 696 |    ]
 697 |   },
 698 |   {
 699 |    "cell_type": "markdown",
 700 |    "metadata": {},
 701 |    "source": [
 702 |     "1. Scale the data\n",
 703 |     "2. Initialize random centroids\n",
 704 |     "3. Label each data points\n",
 705 |     "4. Update centroids\n",
 706 |     "5. Repeat steps 3 and 4 until centroids stop changing"
 707 |    ]
 708 |   },
 709 |   {
 710 |    "cell_type": "markdown",
 711 |    "metadata": {},
 712 |    "source": [
 713 |     "## 1. Scaling the Data"
 714 |    ]
 715 |   },
 716 |   {
 717 |    "cell_type": "code",
 718 |    "execution_count": 26,
 719 |    "metadata": {},
 720 |    "outputs": [],
 721 |    "source": [
 722 |     "data = ((data - data.min()) / (data.max() - data.min())) * 9 + 1"
 723 |    ]
 724 |   },
 725 |   {
 726 |    "cell_type": "code",
 727 |    "execution_count": 27,
 728 |    "metadata": {},
 729 |    "outputs": [
 730 |     {
 731 |      "data": {
 732 |       "text/html": [
 733 |        "<div>\n",
 734 |        "<style scoped>\n",
 735 |        "    .dataframe tbody tr th:only-of-type {\n",
 736 |        "        vertical-align: middle;\n",
 737 |        "    }\n",
 738 |        "\n",
 739 |        "    .dataframe tbody tr th {\n",
 740 |        "        vertical-align: top;\n",
 741 |        "    }\n",
 742 |        "\n",
 743 |        "    .dataframe thead th {\n",
 744 |        "        text-align: right;\n",
 745 |        "    }\n",
 746 |        "</style>\n",
 747 |        "<table border=\"1\" class=\"dataframe\">\n",
 748 |        "  <thead>\n",
 749 |        "    <tr style=\"text-align: right;\">\n",
 750 |        "      <th></th>\n",
 751 |        "      <th>overall</th>\n",
 752 |        "      <th>potential</th>\n",
 753 |        "      <th>wage_eur</th>\n",
 754 |        "      <th>value_eur</th>\n",
 755 |        "      <th>age</th>\n",
 756 |        "    </tr>\n",
 757 |        "  </thead>\n",
 758 |        "  <tbody>\n",
 759 |        "    <tr>\n",
 760 |        "      <th>count</th>\n",
 761 |        "      <td>19165.000000</td>\n",
 762 |        "      <td>19165.000000</td>\n",
 763 |        "      <td>19165.000000</td>\n",
 764 |        "      <td>19165.000000</td>\n",
 765 |        "      <td>19165.000000</td>\n",
 766 |        "    </tr>\n",
 767 |        "    <tr>\n",
 768 |        "      <th>mean</th>\n",
 769 |        "      <td>4.670472</td>\n",
 770 |        "      <td>5.319998</td>\n",
 771 |        "      <td>1.219443</td>\n",
 772 |        "      <td>1.131826</td>\n",
 773 |        "      <td>4.063345</td>\n",
 774 |        "    </tr>\n",
 775 |        "    <tr>\n",
 776 |        "      <th>std</th>\n",
 777 |        "      <td>1.346635</td>\n",
 778 |        "      <td>1.191076</td>\n",
 779 |        "      <td>0.501528</td>\n",
 780 |        "      <td>0.353229</td>\n",
 781 |        "      <td>1.575838</td>\n",
 782 |        "    </tr>\n",
 783 |        "    <tr>\n",
 784 |        "      <th>min</th>\n",
 785 |        "      <td>1.000000</td>\n",
 786 |        "      <td>1.000000</td>\n",
 787 |        "      <td>1.000000</td>\n",
 788 |        "      <td>1.000000</td>\n",
 789 |        "      <td>1.000000</td>\n",
 790 |        "    </tr>\n",
 791 |        "    <tr>\n",
 792 |        "      <th>25%</th>\n",
 793 |        "      <td>3.739130</td>\n",
 794 |        "      <td>4.521739</td>\n",
 795 |        "      <td>1.012876</td>\n",
 796 |        "      <td>1.021620</td>\n",
 797 |        "      <td>2.666667</td>\n",
 798 |        "    </tr>\n",
 799 |        "    <tr>\n",
 800 |        "      <th>50%</th>\n",
 801 |        "      <td>4.717391</td>\n",
 802 |        "      <td>5.304348</td>\n",
 803 |        "      <td>1.064378</td>\n",
 804 |        "      <td>1.044817</td>\n",
 805 |        "      <td>4.000000</td>\n",
 806 |        "    </tr>\n",
 807 |        "    <tr>\n",
 808 |        "      <th>75%</th>\n",
 809 |        "      <td>5.500000</td>\n",
 810 |        "      <td>6.086957</td>\n",
 811 |        "      <td>1.193133</td>\n",
 812 |        "      <td>1.092370</td>\n",
 813 |        "      <td>5.333333</td>\n",
 814 |        "    </tr>\n",
 815 |        "    <tr>\n",
 816 |        "      <th>max</th>\n",
 817 |        "      <td>10.000000</td>\n",
 818 |        "      <td>10.000000</td>\n",
 819 |        "      <td>10.000000</td>\n",
 820 |        "      <td>10.000000</td>\n",
 821 |        "      <td>10.000000</td>\n",
 822 |        "    </tr>\n",
 823 |        "  </tbody>\n",
 824 |        "</table>\n",
 825 |        "</div>"
 826 |       ],
 827 |       "text/plain": [
 828 |        "            overall     potential      wage_eur     value_eur           age\n",
 829 |        "count  19165.000000  19165.000000  19165.000000  19165.000000  19165.000000\n",
 830 |        "mean       4.670472      5.319998      1.219443      1.131826      4.063345\n",
 831 |        "std        1.346635      1.191076      0.501528      0.353229      1.575838\n",
 832 |        "min        1.000000      1.000000      1.000000      1.000000      1.000000\n",
 833 |        "25%        3.739130      4.521739      1.012876      1.021620      2.666667\n",
 834 |        "50%        4.717391      5.304348      1.064378      1.044817      4.000000\n",
 835 |        "75%        5.500000      6.086957      1.193133      1.092370      5.333333\n",
 836 |        "max       10.000000     10.000000     10.000000     10.000000     10.000000"
 837 |       ]
 838 |      },
 839 |      "execution_count": 27,
 840 |      "metadata": {},
 841 |      "output_type": "execute_result"
 842 |     }
 843 |    ],
 844 |    "source": [
 845 |     "data.describe()"
 846 |    ]
 847 |   },
 848 |   {
 849 |    "cell_type": "code",
 850 |    "execution_count": 28,
 851 |    "metadata": {},
 852 |    "outputs": [
 853 |     {
 854 |      "data": {
 855 |       "text/html": [
 856 |        "<div>\n",
 857 |        "<style scoped>\n",
 858 |        "    .dataframe tbody tr th:only-of-type {\n",
 859 |        "        vertical-align: middle;\n",
 860 |        "    }\n",
 861 |        "\n",
 862 |        "    .dataframe tbody tr th {\n",
 863 |        "        vertical-align: top;\n",
 864 |        "    }\n",
 865 |        "\n",
 866 |        "    .dataframe thead th {\n",
 867 |        "        text-align: right;\n",
 868 |        "    }\n",
 869 |        "</style>\n",
 870 |        "<table border=\"1\" class=\"dataframe\">\n",
 871 |        "  <thead>\n",
 872 |        "    <tr style=\"text-align: right;\">\n",
 873 |        "      <th></th>\n",
 874 |        "      <th>overall</th>\n",
 875 |        "      <th>potential</th>\n",
 876 |        "      <th>wage_eur</th>\n",
 877 |        "      <th>value_eur</th>\n",
 878 |        "      <th>age</th>\n",
 879 |        "    </tr>\n",
 880 |        "  </thead>\n",
 881 |        "  <tbody>\n",
 882 |        "    <tr>\n",
 883 |        "      <th>0</th>\n",
 884 |        "      <td>10.000000</td>\n",
 885 |        "      <td>9.608696</td>\n",
 886 |        "      <td>9.227468</td>\n",
 887 |        "      <td>4.618307</td>\n",
 888 |        "      <td>7.000000</td>\n",
 889 |        "    </tr>\n",
 890 |        "    <tr>\n",
 891 |        "      <th>1</th>\n",
 892 |        "      <td>9.804348</td>\n",
 893 |        "      <td>9.413043</td>\n",
 894 |        "      <td>7.939914</td>\n",
 895 |        "      <td>6.543654</td>\n",
 896 |        "      <td>6.333333</td>\n",
 897 |        "    </tr>\n",
 898 |        "    <tr>\n",
 899 |        "      <th>2</th>\n",
 900 |        "      <td>9.608696</td>\n",
 901 |        "      <td>9.217391</td>\n",
 902 |        "      <td>7.939914</td>\n",
 903 |        "      <td>3.087308</td>\n",
 904 |        "      <td>7.666667</td>\n",
 905 |        "    </tr>\n",
 906 |        "    <tr>\n",
 907 |        "      <th>3</th>\n",
 908 |        "      <td>9.608696</td>\n",
 909 |        "      <td>9.217391</td>\n",
 910 |        "      <td>7.939914</td>\n",
 911 |        "      <td>6.984396</td>\n",
 912 |        "      <td>5.333333</td>\n",
 913 |        "    </tr>\n",
 914 |        "    <tr>\n",
 915 |        "      <th>4</th>\n",
 916 |        "      <td>9.608696</td>\n",
 917 |        "      <td>9.217391</td>\n",
 918 |        "      <td>10.000000</td>\n",
 919 |        "      <td>6.822018</td>\n",
 920 |        "      <td>5.666667</td>\n",
 921 |        "    </tr>\n",
 922 |        "  </tbody>\n",
 923 |        "</table>\n",
 924 |        "</div>"
 925 |       ],
 926 |       "text/plain": [
 927 |        "     overall  potential   wage_eur  value_eur       age\n",
 928 |        "0  10.000000   9.608696   9.227468   4.618307  7.000000\n",
 929 |        "1   9.804348   9.413043   7.939914   6.543654  6.333333\n",
 930 |        "2   9.608696   9.217391   7.939914   3.087308  7.666667\n",
 931 |        "3   9.608696   9.217391   7.939914   6.984396  5.333333\n",
 932 |        "4   9.608696   9.217391  10.000000   6.822018  5.666667"
 933 |       ]
 934 |      },
 935 |      "execution_count": 28,
 936 |      "metadata": {},
 937 |      "output_type": "execute_result"
 938 |     }
 939 |    ],
 940 |    "source": [
 941 |     "data.head()"
 942 |    ]
 943 |   },
 944 |   {
 945 |    "cell_type": "markdown",
 946 |    "metadata": {},
 947 |    "source": [
 948 |     "## 2. Intializing Random Centroids"
 949 |    ]
 950 |   },
 951 |   {
 952 |    "cell_type": "code",
 953 |    "execution_count": 32,
 954 |    "metadata": {},
 955 |    "outputs": [],
 956 |    "source": [
 957 |     "def random_centroid(data, k):\n",
 958 |     "    centroids = []\n",
 959 |     "    for i in range(k):\n",
 960 |     "        centroid = data.apply(lambda x: float(x.sample()))\n",
 961 |     "        centroids.append(centroid)\n",
 962 |     "    return pd.concat(centroids, axis=1)"
 963 |    ]
 964 |   },
 965 |   {
 966 |    "cell_type": "code",
 967 |    "execution_count": 33,
 968 |    "metadata": {},
 969 |    "outputs": [],
 970 |    "source": [
 971 |     "centroids = random_centroid(data, 5)"
 972 |    ]
 973 |   },
 974 |   {
 975 |    "cell_type": "code",
 976 |    "execution_count": 34,
 977 |    "metadata": {},
 978 |    "outputs": [
 979 |     {
 980 |      "data": {
 981 |       "text/html": [
 982 |        "<div>\n",
 983 |        "<style scoped>\n",
 984 |        "    .dataframe tbody tr th:only-of-type {\n",
 985 |        "        vertical-align: middle;\n",
 986 |        "    }\n",
 987 |        "\n",
 988 |        "    .dataframe tbody tr th {\n",
 989 |        "        vertical-align: top;\n",
 990 |        "    }\n",
 991 |        "\n",
 992 |        "    .dataframe thead th {\n",
 993 |        "        text-align: right;\n",
 994 |        "    }\n",
 995 |        "</style>\n",
 996 |        "<table border=\"1\" class=\"dataframe\">\n",
 997 |        "  <thead>\n",
 998 |        "    <tr style=\"text-align: right;\">\n",
 999 |        "      <th></th>\n",
1000 |        "      <th>0</th>\n",
1001 |        "      <th>1</th>\n",
1002 |        "      <th>2</th>\n",
1003 |        "      <th>3</th>\n",
1004 |        "      <th>4</th>\n",
1005 |        "    </tr>\n",
1006 |        "  </thead>\n",
1007 |        "  <tbody>\n",
1008 |        "    <tr>\n",
1009 |        "      <th>overall</th>\n",
1010 |        "      <td>6.282609</td>\n",
1011 |        "      <td>5.108696</td>\n",
1012 |        "      <td>6.282609</td>\n",
1013 |        "      <td>5.500000</td>\n",
1014 |        "      <td>4.521739</td>\n",
1015 |        "    </tr>\n",
1016 |        "    <tr>\n",
1017 |        "      <th>potential</th>\n",
1018 |        "      <td>6.478261</td>\n",
1019 |        "      <td>4.717391</td>\n",
1020 |        "      <td>5.108696</td>\n",
1021 |        "      <td>4.130435</td>\n",
1022 |        "      <td>5.695652</td>\n",
1023 |        "    </tr>\n",
1024 |        "    <tr>\n",
1025 |        "      <th>wage_eur</th>\n",
1026 |        "      <td>1.115880</td>\n",
1027 |        "      <td>1.012876</td>\n",
1028 |        "      <td>1.373391</td>\n",
1029 |        "      <td>1.270386</td>\n",
1030 |        "      <td>1.064378</td>\n",
1031 |        "    </tr>\n",
1032 |        "    <tr>\n",
1033 |        "      <th>value_eur</th>\n",
1034 |        "      <td>1.097010</td>\n",
1035 |        "      <td>1.025099</td>\n",
1036 |        "      <td>1.175879</td>\n",
1037 |        "      <td>1.087731</td>\n",
1038 |        "      <td>1.055255</td>\n",
1039 |        "    </tr>\n",
1040 |        "    <tr>\n",
1041 |        "      <th>age</th>\n",
1042 |        "      <td>7.000000</td>\n",
1043 |        "      <td>5.666667</td>\n",
1044 |        "      <td>2.666667</td>\n",
1045 |        "      <td>2.000000</td>\n",
1046 |        "      <td>2.666667</td>\n",
1047 |        "    </tr>\n",
1048 |        "  </tbody>\n",
1049 |        "</table>\n",
1050 |        "</div>"
1051 |       ],
1052 |       "text/plain": [
1053 |        "                  0         1         2         3         4\n",
1054 |        "overall    6.282609  5.108696  6.282609  5.500000  4.521739\n",
1055 |        "potential  6.478261  4.717391  5.108696  4.130435  5.695652\n",
1056 |        "wage_eur   1.115880  1.012876  1.373391  1.270386  1.064378\n",
1057 |        "value_eur  1.097010  1.025099  1.175879  1.087731  1.055255\n",
1058 |        "age        7.000000  5.666667  2.666667  2.000000  2.666667"
1059 |       ]
1060 |      },
1061 |      "execution_count": 34,
1062 |      "metadata": {},
1063 |      "output_type": "execute_result"
1064 |     }
1065 |    ],
1066 |    "source": [
1067 |     "centroids"
1068 |    ]
1069 |   },
1070 |   {
1071 |    "cell_type": "code",
1072 |    "execution_count": 41,
1073 |    "metadata": {},
1074 |    "outputs": [],
1075 |    "source": [
1076 |     "def get_labels(data, centroids):\n",
1077 |     "    distances = centroids.apply(lambda x: np.sqrt(((data - x) ** 2).sum(axis=1)))\n",
1078 |     "    return distances.idxmin(axis=1)"
1079 |    ]
1080 |   },
1081 |   {
1082 |    "cell_type": "code",
1083 |    "execution_count": 43,
1084 |    "metadata": {},
1085 |    "outputs": [],
1086 |    "source": [
1087 |     "labels = get_labels(data, centroids)"
1088 |    ]
1089 |   },
1090 |   {
1091 |    "cell_type": "code",
1092 |    "execution_count": 45,
1093 |    "metadata": {},
1094 |    "outputs": [
1095 |     {
1096 |      "data": {
1097 |       "text/plain": [
1098 |        "4    9064\n",
1099 |        "1    6605\n",
1100 |        "0    1713\n",
1101 |        "2    1701\n",
1102 |        "3      82\n",
1103 |        "dtype: int64"
1104 |       ]
1105 |      },
1106 |      "execution_count": 45,
1107 |      "metadata": {},
1108 |      "output_type": "execute_result"
1109 |     }
1110 |    ],
1111 |    "source": [
1112 |     "labels.value_counts()"
1113 |    ]
1114 |   },
1115 |   {
1116 |    "cell_type": "code",
1117 |    "execution_count": 48,
1118 |    "metadata": {},
1119 |    "outputs": [],
1120 |    "source": [
1121 |     "def new_centroids(data, labels, k):\n",
1122 |     "    return data.groupby(labels).apply(lambda x: np.exp(np.log(x).mean())).T\n"
1123 |    ]
1124 |   },
1125 |   {
1126 |    "cell_type": "code",
1127 |    "execution_count": 49,
1128 |    "metadata": {},
1129 |    "outputs": [],
1130 |    "source": [
1131 |     "from sklearn.decomposition import PCA\n",
1132 |     "import matplotlib.pyplot as plt\n",
1133 |     "from IPython.display import clear_output"
1134 |    ]
1135 |   },
1136 |   {
1137 |    "cell_type": "code",
1138 |    "execution_count": 54,
1139 |    "metadata": {},
1140 |    "outputs": [],
1141 |    "source": [
1142 |     "def plot_clusters(data, labels, centroids, iteration):\n",
1143 |     "    pca = PCA(n_components=2)\n",
1144 |     "    data_2d = pca.fit_transform(data)\n",
1145 |     "    centroids_2d = pca.transform(centroids.T)\n",
1146 |     "    clear_output(wait=True)\n",
1147 |     "    plt.title(f'Iteration {iteration}')\n",
1148 |     "    plt.scatter(x=data_2d[:, 0], y=data_2d[:,1], c=labels)\n",
1149 |     "    plt.show()"
1150 |    ]
1151 |   },
1152 |   {
1153 |    "cell_type": "code",
1154 |    "execution_count": 55,
1155 |    "metadata": {},
1156 |    "outputs": [
1157 |     {
1158 |      "data": {
1159 |       "image/png": "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",
1160 |       "text/plain": [
1161 |        "<Figure size 432x288 with 1 Axes>"
1162 |       ]
1163 |      },
1164 |      "metadata": {
1165 |       "needs_background": "light"
1166 |      },
1167 |      "output_type": "display_data"
1168 |     }
1169 |    ],
1170 |    "source": [
1171 |     "max_iterations = 100\n",
1172 |     "k = 3 # number of clusters\n",
1173 |     "\n",
1174 |     "centroids = random_centroid(data, k)\n",
1175 |     "old_centroids = pd.DataFrame()\n",
1176 |     "iteration = 1\n",
1177 |     "\n",
1178 |     "while iteration < max_iterations and not centroids.equals(old_centroids):\n",
1179 |     "    old_centroids = centroids\n",
1180 |     "\n",
1181 |     "    labels = get_labels(data, centroids)\n",
1182 |     "    centroids = new_centroids(data, labels, k)\n",
1183 |     "    plot_clusters(data, labels, centroids, iteration)\n",
1184 |     "    iteration += 1"
1185 |    ]
1186 |   },
1187 |   {
1188 |    "cell_type": "code",
1189 |    "execution_count": 56,
1190 |    "metadata": {},
1191 |    "outputs": [
1192 |     {
1193 |      "data": {
1194 |       "text/html": [
1195 |        "<div>\n",
1196 |        "<style scoped>\n",
1197 |        "    .dataframe tbody tr th:only-of-type {\n",
1198 |        "        vertical-align: middle;\n",
1199 |        "    }\n",
1200 |        "\n",
1201 |        "    .dataframe tbody tr th {\n",
1202 |        "        vertical-align: top;\n",
1203 |        "    }\n",
1204 |        "\n",
1205 |        "    .dataframe thead th {\n",
1206 |        "        text-align: right;\n",
1207 |        "    }\n",
1208 |        "</style>\n",
1209 |        "<table border=\"1\" class=\"dataframe\">\n",
1210 |        "  <thead>\n",
1211 |        "    <tr style=\"text-align: right;\">\n",
1212 |        "      <th></th>\n",
1213 |        "      <th>0</th>\n",
1214 |        "      <th>1</th>\n",
1215 |        "      <th>2</th>\n",
1216 |        "    </tr>\n",
1217 |        "  </thead>\n",
1218 |        "  <tbody>\n",
1219 |        "    <tr>\n",
1220 |        "      <th>overall</th>\n",
1221 |        "      <td>5.807503</td>\n",
1222 |        "      <td>4.781960</td>\n",
1223 |        "      <td>3.205672</td>\n",
1224 |        "    </tr>\n",
1225 |        "    <tr>\n",
1226 |        "      <th>potential</th>\n",
1227 |        "      <td>6.497870</td>\n",
1228 |        "      <td>4.506813</td>\n",
1229 |        "      <td>4.930905</td>\n",
1230 |        "    </tr>\n",
1231 |        "    <tr>\n",
1232 |        "      <th>wage_eur</th>\n",
1233 |        "      <td>1.420500</td>\n",
1234 |        "      <td>1.118498</td>\n",
1235 |        "      <td>1.028564</td>\n",
1236 |        "    </tr>\n",
1237 |        "    <tr>\n",
1238 |        "      <th>value_eur</th>\n",
1239 |        "      <td>1.285685</td>\n",
1240 |        "      <td>1.044909</td>\n",
1241 |        "      <td>1.026655</td>\n",
1242 |        "    </tr>\n",
1243 |        "    <tr>\n",
1244 |        "      <th>age</th>\n",
1245 |        "      <td>3.598215</td>\n",
1246 |        "      <td>5.467648</td>\n",
1247 |        "      <td>2.514741</td>\n",
1248 |        "    </tr>\n",
1249 |        "  </tbody>\n",
1250 |        "</table>\n",
1251 |        "</div>"
1252 |       ],
1253 |       "text/plain": [
1254 |        "                  0         1         2\n",
1255 |        "overall    5.807503  4.781960  3.205672\n",
1256 |        "potential  6.497870  4.506813  4.930905\n",
1257 |        "wage_eur   1.420500  1.118498  1.028564\n",
1258 |        "value_eur  1.285685  1.044909  1.026655\n",
1259 |        "age        3.598215  5.467648  2.514741"
1260 |       ]
1261 |      },
1262 |      "execution_count": 56,
1263 |      "metadata": {},
1264 |      "output_type": "execute_result"
1265 |     }
1266 |    ],
1267 |    "source": [
1268 |     "centroids"
1269 |    ]
1270 |   },
1271 |   {
1272 |    "cell_type": "code",
1273 |    "execution_count": 58,
1274 |    "metadata": {},
1275 |    "outputs": [
1276 |     {
1277 |      "data": {
1278 |       "text/html": [
1279 |        "<div>\n",
1280 |        "<style scoped>\n",
1281 |        "    .dataframe tbody tr th:only-of-type {\n",
1282 |        "        vertical-align: middle;\n",
1283 |        "    }\n",
1284 |        "\n",
1285 |        "    .dataframe tbody tr th {\n",
1286 |        "        vertical-align: top;\n",
1287 |        "    }\n",
1288 |        "\n",
1289 |        "    .dataframe thead th {\n",
1290 |        "        text-align: right;\n",
1291 |        "    }\n",
1292 |        "</style>\n",
1293 |        "<table border=\"1\" class=\"dataframe\">\n",
1294 |        "  <thead>\n",
1295 |        "    <tr style=\"text-align: right;\">\n",
1296 |        "      <th></th>\n",
1297 |        "      <th>short_name</th>\n",
1298 |        "      <th>overall</th>\n",
1299 |        "      <th>potential</th>\n",
1300 |        "      <th>wage_eur</th>\n",
1301 |        "      <th>value_eur</th>\n",
1302 |        "      <th>age</th>\n",
1303 |        "    </tr>\n",
1304 |        "  </thead>\n",
1305 |        "  <tbody>\n",
1306 |        "    <tr>\n",
1307 |        "      <th>199</th>\n",
1308 |        "      <td>Pepe</td>\n",
1309 |        "      <td>82</td>\n",
1310 |        "      <td>82</td>\n",
1311 |        "      <td>14000.0</td>\n",
1312 |        "      <td>5500000.0</td>\n",
1313 |        "      <td>38</td>\n",
1314 |        "    </tr>\n",
1315 |        "    <tr>\n",
1316 |        "      <th>284</th>\n",
1317 |        "      <td>Joaquín</td>\n",
1318 |        "      <td>81</td>\n",
1319 |        "      <td>81</td>\n",
1320 |        "      <td>23000.0</td>\n",
1321 |        "      <td>8500000.0</td>\n",
1322 |        "      <td>39</td>\n",
1323 |        "    </tr>\n",
1324 |        "    <tr>\n",
1325 |        "      <th>292</th>\n",
1326 |        "      <td>José Fonte</td>\n",
1327 |        "      <td>81</td>\n",
1328 |        "      <td>81</td>\n",
1329 |        "      <td>30000.0</td>\n",
1330 |        "      <td>4600000.0</td>\n",
1331 |        "      <td>37</td>\n",
1332 |        "    </tr>\n",
1333 |        "    <tr>\n",
1334 |        "      <th>388</th>\n",
1335 |        "      <td>G. Buffon</td>\n",
1336 |        "      <td>80</td>\n",
1337 |        "      <td>80</td>\n",
1338 |        "      <td>18000.0</td>\n",
1339 |        "      <td>2300000.0</td>\n",
1340 |        "      <td>43</td>\n",
1341 |        "    </tr>\n",
1342 |        "    <tr>\n",
1343 |        "      <th>509</th>\n",
1344 |        "      <td>Iniesta</td>\n",
1345 |        "      <td>79</td>\n",
1346 |        "      <td>79</td>\n",
1347 |        "      <td>10000.0</td>\n",
1348 |        "      <td>5500000.0</td>\n",
1349 |        "      <td>37</td>\n",
1350 |        "    </tr>\n",
1351 |        "    <tr>\n",
1352 |        "      <th>...</th>\n",
1353 |        "      <td>...</td>\n",
1354 |        "      <td>...</td>\n",
1355 |        "      <td>...</td>\n",
1356 |        "      <td>...</td>\n",
1357 |        "      <td>...</td>\n",
1358 |        "      <td>...</td>\n",
1359 |        "    </tr>\n",
1360 |        "    <tr>\n",
1361 |        "      <th>18890</th>\n",
1362 |        "      <td>S. Haokip</td>\n",
1363 |        "      <td>51</td>\n",
1364 |        "      <td>51</td>\n",
1365 |        "      <td>500.0</td>\n",
1366 |        "      <td>60000.0</td>\n",
1367 |        "      <td>28</td>\n",
1368 |        "    </tr>\n",
1369 |        "    <tr>\n",
1370 |        "      <th>18971</th>\n",
1371 |        "      <td>Lalkhawpuimawia</td>\n",
1372 |        "      <td>51</td>\n",
1373 |        "      <td>51</td>\n",
1374 |        "      <td>500.0</td>\n",
1375 |        "      <td>60000.0</td>\n",
1376 |        "      <td>29</td>\n",
1377 |        "    </tr>\n",
1378 |        "    <tr>\n",
1379 |        "      <th>19032</th>\n",
1380 |        "      <td>Song Yue</td>\n",
1381 |        "      <td>50</td>\n",
1382 |        "      <td>50</td>\n",
1383 |        "      <td>2000.0</td>\n",
1384 |        "      <td>40000.0</td>\n",
1385 |        "      <td>29</td>\n",
1386 |        "    </tr>\n",
1387 |        "    <tr>\n",
1388 |        "      <th>19100</th>\n",
1389 |        "      <td>J. Russell</td>\n",
1390 |        "      <td>49</td>\n",
1391 |        "      <td>49</td>\n",
1392 |        "      <td>500.0</td>\n",
1393 |        "      <td>15000.0</td>\n",
1394 |        "      <td>36</td>\n",
1395 |        "    </tr>\n",
1396 |        "    <tr>\n",
1397 |        "      <th>19118</th>\n",
1398 |        "      <td>Gao Xiang</td>\n",
1399 |        "      <td>49</td>\n",
1400 |        "      <td>49</td>\n",
1401 |        "      <td>2000.0</td>\n",
1402 |        "      <td>35000.0</td>\n",
1403 |        "      <td>32</td>\n",
1404 |        "    </tr>\n",
1405 |        "  </tbody>\n",
1406 |        "</table>\n",
1407 |        "<p>7191 rows × 6 columns</p>\n",
1408 |        "</div>"
1409 |       ],
1410 |       "text/plain": [
1411 |        "            short_name  overall  potential  wage_eur  value_eur  age\n",
1412 |        "199               Pepe       82         82   14000.0  5500000.0   38\n",
1413 |        "284            Joaquín       81         81   23000.0  8500000.0   39\n",
1414 |        "292         José Fonte       81         81   30000.0  4600000.0   37\n",
1415 |        "388          G. Buffon       80         80   18000.0  2300000.0   43\n",
1416 |        "509            Iniesta       79         79   10000.0  5500000.0   37\n",
1417 |        "...                ...      ...        ...       ...        ...  ...\n",
1418 |        "18890        S. Haokip       51         51     500.0    60000.0   28\n",
1419 |        "18971  Lalkhawpuimawia       51         51     500.0    60000.0   29\n",
1420 |        "19032         Song Yue       50         50    2000.0    40000.0   29\n",
1421 |        "19100       J. Russell       49         49     500.0    15000.0   36\n",
1422 |        "19118        Gao Xiang       49         49    2000.0    35000.0   32\n",
1423 |        "\n",
1424 |        "[7191 rows x 6 columns]"
1425 |       ]
1426 |      },
1427 |      "execution_count": 58,
1428 |      "metadata": {},
1429 |      "output_type": "execute_result"
1430 |     }
1431 |    ],
1432 |    "source": [
1433 |     "players[labels == 1][[\"short_name\"] + features]"
1434 |    ]
1435 |   },
1436 |   {
1437 |    "cell_type": "markdown",
1438 |    "metadata": {},
1439 |    "source": [
1440 |     "## Conclusion\n",
1441 |     "Cluster 0 represets star players\n",
1442 |     "\n",
1443 |     "Cluster 1 represents older players who have hit their potential.\n",
1444 |     "\n",
1445 |     "Cluster 2 represents young players\n",
1446 |     "\n",
1447 |     "We were able to confirm this hypothesis by looking at the name of players. \n",
1448 |     "\n",
1449 |     "K-means allows you to find patterns in  your data that you didn't know were there. This allowed us to categorize players automatically into different groups. "
1450 |    ]
1451 |   },
1452 |   {
1453 |    "cell_type": "code",
1454 |    "execution_count": 60,
1455 |    "metadata": {},
1456 |    "outputs": [],
1457 |    "source": [
1458 |     "from sklearn.cluster import KMeans"
1459 |    ]
1460 |   },
1461 |   {
1462 |    "cell_type": "code",
1463 |    "execution_count": 61,
1464 |    "metadata": {},
1465 |    "outputs": [
1466 |     {
1467 |      "data": {
1468 |       "text/html": [
1469 |        "<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>KMeans(n_clusters=3)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" checked><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">KMeans</label><div class=\"sk-toggleable__content\"><pre>KMeans(n_clusters=3)</pre></div></div></div></div></div>"
1470 |       ],
1471 |       "text/plain": [
1472 |        "KMeans(n_clusters=3)"
1473 |       ]
1474 |      },
1475 |      "execution_count": 61,
1476 |      "metadata": {},
1477 |      "output_type": "execute_result"
1478 |     }
1479 |    ],
1480 |    "source": [
1481 |     "kmeans = KMeans(3)\n",
1482 |     "kmeans.fit(data)"
1483 |    ]
1484 |   },
1485 |   {
1486 |    "cell_type": "code",
1487 |    "execution_count": 62,
1488 |    "metadata": {},
1489 |    "outputs": [],
1490 |    "source": [
1491 |     "centroids = kmeans.cluster_centers_"
1492 |    ]
1493 |   },
1494 |   {
1495 |    "cell_type": "code",
1496 |    "execution_count": 63,
1497 |    "metadata": {},
1498 |    "outputs": [
1499 |     {
1500 |      "data": {
1501 |       "text/html": [
1502 |        "<div>\n",
1503 |        "<style scoped>\n",
1504 |        "    .dataframe tbody tr th:only-of-type {\n",
1505 |        "        vertical-align: middle;\n",
1506 |        "    }\n",
1507 |        "\n",
1508 |        "    .dataframe tbody tr th {\n",
1509 |        "        vertical-align: top;\n",
1510 |        "    }\n",
1511 |        "\n",
1512 |        "    .dataframe thead th {\n",
1513 |        "        text-align: right;\n",
1514 |        "    }\n",
1515 |        "</style>\n",
1516 |        "<table border=\"1\" class=\"dataframe\">\n",
1517 |        "  <thead>\n",
1518 |        "    <tr style=\"text-align: right;\">\n",
1519 |        "      <th></th>\n",
1520 |        "      <th>0</th>\n",
1521 |        "      <th>1</th>\n",
1522 |        "      <th>2</th>\n",
1523 |        "    </tr>\n",
1524 |        "  </thead>\n",
1525 |        "  <tbody>\n",
1526 |        "    <tr>\n",
1527 |        "      <th>overall</th>\n",
1528 |        "      <td>6.232940</td>\n",
1529 |        "      <td>4.800860</td>\n",
1530 |        "      <td>3.601045</td>\n",
1531 |        "    </tr>\n",
1532 |        "    <tr>\n",
1533 |        "      <th>potential</th>\n",
1534 |        "      <td>6.623509</td>\n",
1535 |        "      <td>4.503855</td>\n",
1536 |        "      <td>5.207475</td>\n",
1537 |        "    </tr>\n",
1538 |        "    <tr>\n",
1539 |        "      <th>wage_eur</th>\n",
1540 |        "      <td>1.657900</td>\n",
1541 |        "      <td>1.112885</td>\n",
1542 |        "      <td>1.040009</td>\n",
1543 |        "    </tr>\n",
1544 |        "    <tr>\n",
1545 |        "      <th>value_eur</th>\n",
1546 |        "      <td>1.414134</td>\n",
1547 |        "      <td>1.040007</td>\n",
1548 |        "      <td>1.035859</td>\n",
1549 |        "    </tr>\n",
1550 |        "    <tr>\n",
1551 |        "      <th>age</th>\n",
1552 |        "      <td>4.142102</td>\n",
1553 |        "      <td>5.608923</td>\n",
1554 |        "      <td>2.712137</td>\n",
1555 |        "    </tr>\n",
1556 |        "  </tbody>\n",
1557 |        "</table>\n",
1558 |        "</div>"
1559 |       ],
1560 |       "text/plain": [
1561 |        "                  0         1         2\n",
1562 |        "overall    6.232940  4.800860  3.601045\n",
1563 |        "potential  6.623509  4.503855  5.207475\n",
1564 |        "wage_eur   1.657900  1.112885  1.040009\n",
1565 |        "value_eur  1.414134  1.040007  1.035859\n",
1566 |        "age        4.142102  5.608923  2.712137"
1567 |       ]
1568 |      },
1569 |      "execution_count": 63,
1570 |      "metadata": {},
1571 |      "output_type": "execute_result"
1572 |     }
1573 |    ],
1574 |    "source": [
1575 |     "pd.DataFrame(centroids, columns=features).T"
1576 |    ]
1577 |   },
1578 |   {
1579 |    "cell_type": "code",
1580 |    "execution_count": null,
1581 |    "metadata": {},
1582 |    "outputs": [],
1583 |    "source": []
1584 |   }
1585 |  ],
1586 |  "metadata": {
1587 |   "kernelspec": {
1588 |    "display_name": "Python 3.9.13 64-bit",
1589 |    "language": "python",
1590 |    "name": "python3"
1591 |   },
1592 |   "language_info": {
1593 |    "codemirror_mode": {
1594 |     "name": "ipython",
1595 |     "version": 3
1596 |    },
1597 |    "file_extension": ".py",
1598 |    "mimetype": "text/x-python",
1599 |    "name": "python",
1600 |    "nbconvert_exporter": "python",
1601 |    "pygments_lexer": "ipython3",
1602 |    "version": "3.9.13"
1603 |   },
1604 |   "orig_nbformat": 4,
1605 |   "vscode": {
1606 |    "interpreter": {
1607 |     "hash": "aee8b7b246df8f9039afb4144a1f6fd8d2ca17a180786b69acc140d282b71a49"
1608 |    }
1609 |   }
1610 |  },
1611 |  "nbformat": 4,
1612 |  "nbformat_minor": 2
1613 | }
1614 | 


--------------------------------------------------------------------------------