notebooks/08_kmeans_scratch.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K-means using basic python libraries "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "from sklearn.preprocessing import MinMaxScaler"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Considering normalized feature vectors. Let's take a sample dataset of 3 groups with 3 features. \n",
    "\n",
    "1. height (154-177 cm, 167-187 cm)\n",
    "2. gender (0: female, 1: male)\n",
    "3. weight (40-60 kg, 60-90 kg) \n",
    "\n",
    "Let's make random uniform data of 200 points for all the features, 100 for each group. (0-99: female, 100-199: male)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# index 0-99 female, 100-199 male\n",
    "\n",
    "ht = np.concatenate((np.random.uniform(154, 177, 100), np.random.uniform(167, 187, 100))) \n",
    "ge = np.concatenate((np.repeat(0,100), np.repeat(1,100)))\n",
    "wt = np.concatenate((np.random.randint(40, 60, 100), np.random.randint(60, 90, 100))) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((200,), (200,), (200,))"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ht.shape, ge.shape, wt.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# combined array with all 3 features which we would like to cluster\n",
    "\n",
    "array_to_clust = np.concatenate((ht[:,None], ge[:,None], wt[:,None]), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(200, 3)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "array_to_clust.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's visualize out data first as we know first 100 points represent 1 group and next 100 2nd group"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1103a6358>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "ax = fig.add_subplot(111, projection='3d')\n",
    "\n",
    "# plot 200 points of all 3 features\n",
    "\n",
    "ax.scatter(array_to_clust[:100,0], array_to_clust[:100,1], array_to_clust[:100,2], c = 'r')\n",
    "ax.scatter(array_to_clust[100:,0], array_to_clust[100:,1], array_to_clust[100:,2], c = 'b')\n",
    "\n",
    "ax.set_xlabel('X: Height')\n",
    "ax.set_ylabel('Y: Gender')\n",
    "ax.set_zlabel('Z: Weigth')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Clearly we have 2 clusters in our data. Let's now try to write k-means and see how it performs on the same data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Algo:**\n",
    "* Step 0: Normalize data. (Important step)\n",
    "* Step 1: Randomly choose k centroids where k is argument of function\n",
    "* Step 2: Find distance of all other points from chosen k points\n",
    "* Step 3: Assign each point one of k cluster ids based on which centroid is closest to that point\n",
    "* Step 4: Find centroid of each k group now and repeat from step 2\n",
    "* Step 5: Stop when either maximum iterations are reached (argument) or when the clusters are not changing \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "class kmeans():\n",
    "    def __init__(self, k=2, max_iter = 300, tol =  0.001):\n",
    "        \"\"\"\n",
    "        k: number of clusters\n",
    "        max_iter: maximum number of iterations\n",
    "        tol: tolerance value\n",
    "\n",
    "        \"\"\"\n",
    "\n",
    "        self.k = k\n",
    "        self.max_iter = max_iter\n",
    "        self.tol = tol\n",
    "    \n",
    "    def fit(self, data):\n",
    "        \n",
    "        self.centroids = {} # empty dictionary for storing centroids\n",
    "        \n",
    "        # making k random initial centroids from data\n",
    "        perm_data = np.random.permutation(data)\n",
    "        \n",
    "        for i in range(self.k):\n",
    "            self.centroids[i] = perm_data[i]\n",
    "            \n",
    "        # now looping until it reaches maximum iterations\n",
    "        for i in range(self.max_iter):\n",
    "            print(\"Iteration: \", i)\n",
    "            \n",
    "            self.clusters = {} # assign new group on each iteration\n",
    "            \n",
    "            # make k empty lists. Each list will have all rows corresponding to that cluster\n",
    "            for i in range(self.k):\n",
    "                self.clusters[i] = [] \n",
    "            \n",
    "            for row in data:\n",
    "                distances = [np.linalg.norm(row - centroid) for centroid in self.centroids] # distance from all centroics\n",
    "                cluster = distances.index(min(distances)) # index of whichever distance is minimum\n",
    "                self.clusters[cluster].append(row) # now append this row into cluster number\n",
    "                \n",
    "            old_centroids = self.centroids\n",
    "            \n",
    "            # calculating new centroids for all clusters\n",
    "            for clus_num in self.clusters:\n",
    "                self.centroids[clus_num] = np.average(self.clusters[clus_num], axis=0)\n",
    "            \n",
    "            # now we need to check stopping condition. i.e. if algo is optimized even befoer reaching\n",
    "            # max iterations, then break from the loop. \n",
    "            \n",
    "            optimized = True\n",
    "            \n",
    "            for k in self.centroids:\n",
    "                orig_cent = old_centroids[k]\n",
    "                new_cent = self.centroids[k]\n",
    "                \n",
    "                if np.sum((new_cent - orig_cent)/(orig_cent+0.0001)*100.) > self.tol:\n",
    "                    optimized=False\n",
    "                    \n",
    "            if optimized:\n",
    "                break\n",
    "    \n",
    "    def predict(self, data):\n",
    "        \"\"\"\n",
    "        gives cluster id of new data\n",
    "        data: data to get predictions for\n",
    "        \n",
    "        \"\"\"\n",
    "        \n",
    "        distances  = [np.linalg.norm(data - centroid) for centroid in self.centroids] \n",
    "        cluster = distances.index(min(distances))\n",
    "        return cluster           "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Fitting our data on k-means model now"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "MinMaxScaler(copy=True, feature_range=(0, 1))"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mm  = MinMaxScaler()\n",
    "mm.fit(array_to_clust)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = mm.transform(array_to_clust)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration:  0\n"
     ]
    }
   ],
   "source": [
    "model =  kmeans(max_iter=4,  tol=0.0001)\n",
    "model.fit(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{0: array([0.29733414, 0.        , 0.16816327]),\n",
       " 1: array([0.71676432, 1.        , 0.7244898 ])}"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.centroids"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{0: [array([0.30684682, 0.        , 0.2244898 ]),\n",
       "  array([0.41316884, 0.        , 0.26530612]),\n",
       "  array([0.38319582, 0.        , 0.18367347]),\n",
       "  array([0.57891641, 0.        , 0.24489796]),\n",
       "  array([0.07091727, 0.        , 0.        ]),\n",
       "  array([0.68296402, 0.        , 0.08163265]),\n",
       "  array([0.09156618, 0.        , 0.16326531]),\n",
       "  array([0.33007772, 0.        , 0.26530612]),\n",
       "  array([0.21463748, 0.        , 0.32653061]),\n",
       "  array([0.38300994, 0.        , 0.        ]),\n",
       "  array([0.05930343, 0.        , 0.04081633]),\n",
       "  array([0.40842859, 0.        , 0.28571429]),\n",
       "  array([0.15109567, 0.        , 0.20408163]),\n",
       "  array([0.45000508, 0.        , 0.36734694]),\n",
       "  array([0.59160442, 0.        , 0.24489796]),\n",
       "  array([0.23727694, 0.        , 0.        ]),\n",
       "  array([0.40769614, 0.        , 0.16326531]),\n",
       "  array([0.01351082, 0.        , 0.04081633]),\n",
       "  array([0.40608048, 0.        , 0.16326531]),\n",
       "  array([0.47931537, 0.        , 0.06122449]),\n",
       "  array([0.03180454, 0.        , 0.02040816]),\n",
       "  array([0.50025896, 0.        , 0.20408163]),\n",
       "  array([0.26929456, 0.        , 0.36734694]),\n",
       "  array([0.        , 0.        , 0.12244898]),\n",
       "  array([0.48565334, 0.        , 0.18367347]),\n",
       "  array([0.06847605, 0.        , 0.26530612]),\n",
       "  array([0.22002328, 0.        , 0.32653061]),\n",
       "  array([0.4550724 , 0.        , 0.12244898]),\n",
       "  array([0.53873408, 0.        , 0.32653061]),\n",
       "  array([0.67810116, 0.        , 0.32653061]),\n",
       "  array([0.59254043, 0.        , 0.10204082]),\n",
       "  array([0.51221069, 0.        , 0.20408163]),\n",
       "  array([0.29664859, 0.        , 0.18367347]),\n",
       "  array([0.45886659, 0.        , 0.28571429]),\n",
       "  array([0.12149928, 0.        , 0.36734694]),\n",
       "  array([0.33213405, 0.        , 0.08163265]),\n",
       "  array([0.10554298, 0.        , 0.34693878]),\n",
       "  array([0.17005901, 0.        , 0.        ]),\n",
       "  array([0.13926498, 0.        , 0.02040816]),\n",
       "  array([0.08629528, 0.        , 0.10204082]),\n",
       "  array([0.29940158, 0.        , 0.12244898]),\n",
       "  array([0.1111051 , 0.        , 0.16326531]),\n",
       "  array([0.02725887, 0.        , 0.08163265]),\n",
       "  array([0.63500517, 0.        , 0.        ]),\n",
       "  array([0.06565304, 0.        , 0.08163265]),\n",
       "  array([0.20679535, 0.        , 0.02040816]),\n",
       "  array([0.54851621, 0.        , 0.36734694]),\n",
       "  array([0.33969638, 0.        , 0.20408163]),\n",
       "  array([0.28228753, 0.        , 0.16326531]),\n",
       "  array([0.23712824, 0.        , 0.06122449]),\n",
       "  array([0.42703213, 0.        , 0.08163265]),\n",
       "  array([0.08670237, 0.        , 0.04081633]),\n",
       "  array([0.2384819 , 0.        , 0.06122449]),\n",
       "  array([0.5343123 , 0.        , 0.12244898]),\n",
       "  array([0.59690152, 0.        , 0.10204082]),\n",
       "  array([0.22145271, 0.        , 0.        ]),\n",
       "  array([0.6666873 , 0.        , 0.36734694]),\n",
       "  array([0.68684836, 0.        , 0.12244898]),\n",
       "  array([0.04941797, 0.        , 0.28571429]),\n",
       "  array([0.32760941, 0.        , 0.10204082]),\n",
       "  array([0.23514047, 0.        , 0.32653061]),\n",
       "  array([0.26408937, 0.        , 0.08163265]),\n",
       "  array([0.23886212, 0.        , 0.10204082]),\n",
       "  array([0.50446046, 0.        , 0.14285714]),\n",
       "  array([0.53256486, 0.        , 0.18367347]),\n",
       "  array([0.41216193, 0.        , 0.28571429]),\n",
       "  array([0.04205639, 0.        , 0.14285714]),\n",
       "  array([0.16611601, 0.        , 0.2244898 ]),\n",
       "  array([0.16370816, 0.        , 0.14285714]),\n",
       "  array([0.2814297 , 0.        , 0.14285714]),\n",
       "  array([0.23594843, 0.        , 0.        ]),\n",
       "  array([0.24957743, 0.        , 0.08163265]),\n",
       "  array([0.51223064, 0.        , 0.12244898]),\n",
       "  array([0.03473748, 0.        , 0.26530612]),\n",
       "  array([0.61094057, 0.        , 0.30612245]),\n",
       "  array([0.11173553, 0.        , 0.20408163]),\n",
       "  array([0.27778325, 0.        , 0.14285714]),\n",
       "  array([0.54861173, 0.        , 0.        ]),\n",
       "  array([0.59547183, 0.        , 0.10204082]),\n",
       "  array([0.07464442, 0.        , 0.16326531]),\n",
       "  array([0.11985135, 0.        , 0.16326531]),\n",
       "  array([0.06928607, 0.        , 0.24489796]),\n",
       "  array([0.09496457, 0.        , 0.14285714]),\n",
       "  array([0.24548606, 0.        , 0.04081633]),\n",
       "  array([0.29270357, 0.        , 0.30612245]),\n",
       "  array([0.00802881, 0.        , 0.28571429]),\n",
       "  array([0.24093077, 0.        , 0.36734694]),\n",
       "  array([0.6071877 , 0.        , 0.18367347]),\n",
       "  array([0.14968599, 0.        , 0.08163265]),\n",
       "  array([0.07478353, 0.        , 0.16326531]),\n",
       "  array([0.05339734, 0.        , 0.18367347]),\n",
       "  array([0.1674454 , 0.        , 0.02040816]),\n",
       "  array([0.12137084, 0.        , 0.20408163]),\n",
       "  array([0.12234319, 0.        , 0.14285714]),\n",
       "  array([0.01066447, 0.        , 0.34693878]),\n",
       "  array([0.0587577 , 0.        , 0.28571429]),\n",
       "  array([0.49907717, 0.        , 0.10204082]),\n",
       "  array([0.13191559, 0.        , 0.16326531]),\n",
       "  array([0.63752463, 0.        , 0.36734694]),\n",
       "  array([0.62534765, 0.        , 0.02040816])],\n",
       " 1: [array([0.65512273, 1.        , 0.65306122]),\n",
       "  array([0.88731776, 1.        , 0.87755102]),\n",
       "  array([0.62479058, 1.        , 0.95918367]),\n",
       "  array([0.70323499, 1.        , 0.67346939]),\n",
       "  array([0.42351884, 1.        , 0.7755102 ]),\n",
       "  array([0.7199592 , 1.        , 0.40816327]),\n",
       "  array([0.92016535, 1.        , 0.44897959]),\n",
       "  array([0.90950535, 1.        , 0.73469388]),\n",
       "  array([0.80960901, 1.        , 0.7755102 ]),\n",
       "  array([0.67640663, 1.        , 0.85714286]),\n",
       "  array([0.97882574, 1.        , 0.89795918]),\n",
       "  array([0.45154704, 1.        , 0.87755102]),\n",
       "  array([0.92749868, 1.        , 0.69387755]),\n",
       "  array([0.94441822, 1.        , 0.97959184]),\n",
       "  array([0.91950034, 1.        , 0.93877551]),\n",
       "  array([0.95916945, 1.        , 0.46938776]),\n",
       "  array([0.59078401, 1.        , 0.97959184]),\n",
       "  array([0.77491981, 1.        , 0.69387755]),\n",
       "  array([0.89646121, 1.        , 0.81632653]),\n",
       "  array([0.99135467, 1.        , 0.97959184]),\n",
       "  array([0.58446549, 1.        , 0.79591837]),\n",
       "  array([0.84611717, 1.        , 0.75510204]),\n",
       "  array([0.71748289, 1.        , 0.63265306]),\n",
       "  array([0.65017356, 1.        , 0.83673469]),\n",
       "  array([0.76814442, 1.        , 0.93877551]),\n",
       "  array([0.92707349, 1.        , 0.81632653]),\n",
       "  array([0.84399706, 1.        , 0.59183673]),\n",
       "  array([0.8426264, 1.       , 0.7755102]),\n",
       "  array([0.53810844, 1.        , 0.59183673]),\n",
       "  array([0.94922299, 1.        , 0.91836735]),\n",
       "  array([0.71029853, 1.        , 0.7755102 ]),\n",
       "  array([0.73135033, 1.        , 0.63265306]),\n",
       "  array([0.66203336, 1.        , 0.65306122]),\n",
       "  array([0.63464068, 1.        , 0.69387755]),\n",
       "  array([0.87274035, 1.        , 0.40816327]),\n",
       "  array([0.70456952, 1.        , 0.69387755]),\n",
       "  array([0.92955556, 1.        , 0.40816327]),\n",
       "  array([0.46166263, 1.        , 1.        ]),\n",
       "  array([0.71389712, 1.        , 0.57142857]),\n",
       "  array([0.85313797, 1.        , 1.        ]),\n",
       "  array([0.55543003, 1.        , 0.51020408]),\n",
       "  array([0.89618262, 1.        , 0.67346939]),\n",
       "  array([1.        , 1.        , 0.89795918]),\n",
       "  array([0.40659175, 1.        , 0.65306122]),\n",
       "  array([0.66341441, 1.        , 0.6122449 ]),\n",
       "  array([0.48953126, 1.        , 0.42857143]),\n",
       "  array([0.7892286 , 1.        , 0.75510204]),\n",
       "  array([0.53271977, 1.        , 0.55102041]),\n",
       "  array([0.4779494 , 1.        , 0.59183673]),\n",
       "  array([0.53918504, 1.        , 0.57142857]),\n",
       "  array([0.56165204, 1.        , 0.97959184]),\n",
       "  array([0.72320436, 1.        , 1.        ]),\n",
       "  array([0.55949185, 1.        , 0.69387755]),\n",
       "  array([0.46619788, 1.        , 0.87755102]),\n",
       "  array([0.76963696, 1.        , 0.93877551]),\n",
       "  array([0.99353096, 1.        , 0.93877551]),\n",
       "  array([0.91263358, 1.        , 0.59183673]),\n",
       "  array([0.73521461, 1.        , 0.97959184]),\n",
       "  array([0.8689549 , 1.        , 0.75510204]),\n",
       "  array([0.49841884, 1.        , 0.42857143]),\n",
       "  array([0.49887984, 1.        , 0.97959184]),\n",
       "  array([0.61180272, 1.        , 0.97959184]),\n",
       "  array([0.85725835, 1.        , 0.83673469]),\n",
       "  array([0.99279022, 1.        , 0.55102041]),\n",
       "  array([0.53249356, 1.        , 0.67346939]),\n",
       "  array([0.86845768, 1.        , 0.73469388]),\n",
       "  array([0.48250297, 1.        , 0.97959184]),\n",
       "  array([0.64156291, 1.        , 0.79591837]),\n",
       "  array([0.54214073, 1.        , 0.53061224]),\n",
       "  array([0.78806966, 1.        , 0.89795918]),\n",
       "  array([0.88430455, 1.        , 0.46938776]),\n",
       "  array([0.57940129, 1.        , 0.87755102]),\n",
       "  array([0.56835783, 1.        , 0.40816327]),\n",
       "  array([0.46131711, 1.        , 0.71428571]),\n",
       "  array([0.63033661, 1.        , 0.67346939]),\n",
       "  array([0.58425993, 1.        , 0.7755102 ]),\n",
       "  array([0.62790229, 1.        , 0.71428571]),\n",
       "  array([0.77378685, 1.        , 0.51020408]),\n",
       "  array([0.69818659, 1.        , 0.81632653]),\n",
       "  array([0.94070007, 1.        , 0.51020408]),\n",
       "  array([0.64300939, 1.        , 0.79591837]),\n",
       "  array([0.68628829, 1.        , 0.85714286]),\n",
       "  array([0.42035646, 1.        , 0.75510204]),\n",
       "  array([0.54921874, 1.        , 0.69387755]),\n",
       "  array([0.41871299, 1.        , 0.73469388]),\n",
       "  array([0.8107787 , 1.        , 0.65306122]),\n",
       "  array([0.86617071, 1.        , 0.6122449 ]),\n",
       "  array([0.91668165, 1.        , 0.67346939]),\n",
       "  array([0.84956031, 1.        , 0.71428571]),\n",
       "  array([0.64898725, 1.        , 0.48979592]),\n",
       "  array([0.76126625, 1.        , 0.71428571]),\n",
       "  array([0.4334674 , 1.        , 0.59183673]),\n",
       "  array([0.55619931, 1.        , 0.51020408]),\n",
       "  array([0.78706917, 1.        , 0.67346939]),\n",
       "  array([0.81607274, 1.        , 0.6122449 ]),\n",
       "  array([0.78107804, 1.        , 0.42857143]),\n",
       "  array([0.63726949, 1.        , 0.57142857]),\n",
       "  array([0.99668661, 1.        , 0.81632653]),\n",
       "  array([0.98179732, 1.        , 0.87755102]),\n",
       "  array([0.4066726 , 1.        , 0.83673469])]}"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.clusters"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see from above array that all the males fell in one cluser and females in another cluster. Let's plot it "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.29733414, 0.        , 0.16816327])"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.centroids[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a1bcb67f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "ax = fig.add_subplot(111, projection='3d')\n",
    "\n",
    "# plot 200 points of all 3 features\n",
    "\n",
    "# 0th cluster\n",
    "ax.scatter([m[0] for m in model.clusters[0]],\n",
    "           [m[1] for m in model.clusters[0]], \n",
    "           [m[2] for m in model.clusters[0]], c = 'r')\n",
    "\n",
    "# 1st cluster\n",
    "ax.scatter([m[0] for m in model.clusters[1]], \n",
    "           [m[1] for m in model.clusters[1]], \n",
    "           [m[2] for m in model.clusters[1]], c = 'b')\n",
    "\n",
    "# centroids\n",
    "ax.scatter(model.centroids[0][0], \n",
    "           model.centroids[0][1], \n",
    "           model.centroids[0][2], c = 'k', s = 300)\n",
    "\n",
    "ax.scatter(model.centroids[1][0], \n",
    "           model.centroids[1][1], \n",
    "           model.centroids[1][2], c = 'k', s = 300)\n",
    "\n",
    "\n",
    "ax.set_xlabel('X: Height')\n",
    "ax.set_ylabel('Y: Gender')\n",
    "ax.set_zlabel('Z: Weigth')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### End "
   ]
  }
 ],
 "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": "3.6.4"
  },
  "toc": {
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}