Random_Sampling.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import datasets\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.cluster import KMeans"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = datasets.load_iris()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "X=np.array(data.data)\n",
    "y=np.array(data.target)\n",
    "X=np.delete(X, np.array([1,2]), axis = 1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(7,6))\n",
    "plt.scatter(X[:,0],X[:,1],c=y)\n",
    "plt.xlabel('Sepal Length')\n",
    "plt.ylabel('Petal Width')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "km = KMeans(n_clusters=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KMeans(algorithm='auto', copy_x=True, init='k-means++', max_iter=300,\n",
       "    n_clusters=3, n_init=10, n_jobs=None, precompute_distances='auto',\n",
       "    random_state=None, tol=0.0001, verbose=0)"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "km.fit(X,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_ = km.predict(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "32.726526159026164"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "km_obj=km.inertia_\n",
    "km_obj"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAb0AAAF3CAYAAAAxXZWWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAIABJREFUeJzs3Xd8FHX6wPHPd/tuEgKBUESkKBZQkSKIKPaO9dBTz64nng37nZ56lp96eujZT/Hs/VT08MSCXSwgoKioIMVCLwFSt87z+2PCJpvdhGySzSbZ5/165eXOzHe/80yC++zMfL/PGBFBKaWUygWObAeglFJKtRZNekoppXKGJj2llFI5Q5OeUkqpnKFJTymlVM7QpKeUUipnaNJTSimVMzTpKaWUyhma9JRSSuUMTXpKKaVyhivbAaSrW7du0q9fv2yHoZRSqg2ZM2fOOhEp3lK7dpf0+vXrx+zZs7MdhlJKqTbEGPNLY9rp5U2llFI5Q5OeUkqpnKFJTymlVM7QpKeUUipnaNJTSimVMzTpKaWUyhma9JRSSuUMTXpKKaVyhiY9pZRSOUOTnlJKqZyRsaRnjOljjPnAGPO9MWa+MWZiijb7GmM2GWO+rv65PlPxKKWSiUSwyh/EWjMWa/VIrI1/RmKrsx1WnATfw1p3FNbq4VjrT0bCc7IdUtokPAtr/e/tY1h3LBL6MNsh5bRM1t6MApeLyFxjTAEwxxgzXUS+r9PuExEZl8E4lFL1kI2XQOgTIGivCE5Fwp9At7cxjoKsxmZVvgal19fEFpmNlJwJRY9iPLtnNbbGktBnyIbziB9DdD6y4WKk8A4c/kOzGluuytiZnoisFJG51a/LgB+A3pnan1IqPRJdAqGPiX8gAxADqxypfClbYQEgIlB+B4mxAQSRsknZCKlJpOx2Uh0DZbdlIxxFK93TM8b0A4YCM1NsHm2MmWeMedMYM7g14lFKAZHvwaS62BOEyFetHk4CKQdrY+pt0QWtG0tzRBenXm+tRiTcurEooBWSnjEmH3gFuERESutsngv0FZEhwH3Aa/X0ca4xZrYxZvbatWszG7BSucLZB5AUGzzgGtDa0SQyATDe1NscPVs3luZwdk+93hQA7lYNRdkymvSMMW7shPesiEypu11ESkWkvPr1NMBtjOmWot1kERkhIiOKi7f4jEClVGO4dwVnX5Ju7RsXJnBSVkKKh2CckHcW4K+zwY/JvygrMTVJ3oWkOgbyzsUYk5WQcl0mR28a4FHgBxG5q542PavbYYwZWR3P+kzFpJSqYYzBFD0B3n2wzzpc4BqI6fIkxpn9symTd4Gd+EwA8IDpDPl/wfiPyHZojeYIHAcFl4PphH0MeZA3AZN3TrZDy1lGJNXljRbo2Ji9gE+AbwGrevU1wDYAIvKQMeZC4E/YIz2rgMtE5LOG+h0xYoTok9OValkiVSBhjKMw26EkEYmAlIEptM8A2yGRGMgmMJ0wKe+jquYyxswRkRFbapex376IzAAaPH8XkfuB+zMVg1KqcYzx25fd2iBj3GCKsh1GsxjjbPfH0FFoRRallFI5Q5OeUkqpnKFJTymlVM7QO6pKNZGI8PUH3zH9qY+IRS32P3kvRh42tN0NRZfwPKTqJbDKML5DwHdwmxlsIbFVSOULEP0J3LthAie0ycE2qv3I2OjNTNHRm6qteOjyJ3hj8rsEK0IA+PK87HXcKK564sJ2k/isiseh7J9AGLDs6QHuIZguj2Y98UlkPlLyB5BodXw+MAFMtykY51ZZjU21PY0dvamXN5Vqgl9+WMbrD02PJzyAYEWIGa/M5IeZP2UxssYTqwTK7sSuDVk9q0gqITIPQu9mMzQ7lE3X2PGwuVxXEGQjUvaPbIal2jlNeko1wZy35yGWlbQ+WBVi5hvt5PE34VlgUpTCkkok+Fbrx1M7BKsSogtTbLEg9FGrx6M6Dk16SjWBL9+H05U8UdrldhEoaJvz3ZKYAKmn0jqqK4hkkXFR78eT8bVqKKpj0aSnVBPsddxIUt0Odzgd7HfSXq0fUFN49gBSVTjxYALHt3Y0CYzxgPcAkosy+8Cf3bqgqn3TpKdUE3QqKuBvr1yBP99HoJOfQCc/3oCHqx6/gO59kmqmt0nGeDBFj4IpBJNv14XECwWXY9y7ZDs8TOHN4NoRu2BzHuAD756Y/AlZjky1Zzp6U6lmCFaG+Oq9b7FiFkMP2KX9XNqsRSQM4S9AqsAzEuPoku2Q4kQEot9B9Ddwb49xbZftkFQblfXam0rlAl/Ay+gjt/j/WZtmX0ocm+0wUjLGgHsX+0epFqCXN5VSSuUMTXpKKaVyhiY9pZRSOUPv6SnVQYnE7AEq1hq7tJhrQPP6szZB6BMwTvDsjXHkt1CkDexTBCJzIPYruHbEuAfV0y4C4c/AKgH3CIyrT8p2VnQtVP4bJAx5p+Jo5u8kl1jB9yH0HjgHQOB0HI72mT7aZ9RKqQZJdBlScor9tG4EJIb4DsUU3o4x6V/gsSpfg9LrqieNA2Ihne/C4TugZQOvRawSpOQ0iC2r2adnOKbLQxjjrWkXXWS3kyrixxo4AVNwbUINVKv8ISi/q2YHVc9ieY/C0WVSxo6hI7CsIKw7GKxVNSvLJ2EVvYjDs2v2AmsivbypVAckGy+yP6Skorp+ZQiC7yCVL6XfV3SZnfAIVfdXAVTBxkvt+p0ZIpuug+gSO36pBIIQno2U31/TRgTZMAGs9YnHWvUyhN6Jt7OiKxIT3mahqVjBDzN2DB3CxssTEx4AMdhwelbCaS5Nekp1MBJbCdFFxItIx1VB1bPp9xd8I0VfAAaC76RY33wiYQh9AETrbKlOaJtFF9gJjzrzjaUKqXyuZrniofp3Vt7ANgXhD1OvlwqsSPsorl6bJj2lOhoJUu//2hJsQn+VJCcfgBhIKMX6FiBRkhJZfFu41usgqeuHUh13itdJqtIMLtfE6t8k5a0XRgvRpKdUR+PsC46CFBs84Ds87e6Md3/Am2oLePdJu79G7dMRANdOKbY4wbtfzaJ7EKk/xnzgG1ezGPhD/TvzH9fEKHNEyr8DgBNcQ1o1lJagSU+pDsYYB6ZwEnbNSk/1ygA4e2Pyzkm/P88Q8B8Nxo99VmXsvvPOxLj6tVjcSfstvNWuCRpPuH5wFGEKrqxpYzyYwtsBHzXFqQPg2g4T+H28ncMzFNx7Ju/E0Rv8p2boCDqIzveSsjB5wXU4HO0vhWjtTaU6KImtQCr/A7HlGO9o8B2RMOoxrb5EIPxF9f09J8Z/DMYztGUDTrXf2Hqk6mX7HqV7iL3fFFMlJPoLUvUSxNZgvGPBd7BdXq0Oq2oKlD8GRMB/LATObZcf3K3Nskqg9HYIfwnOraDgqjY3crOxtTc16SmllGr3Gpv09CuOUkqpnKFJTymlVM7QpKeUUipnaNJTSjWaxNZvsQqLiCCxVYjV/uZwqfqJxJDYSsRqaM5j26e1N5VSWySRhcimyyG6FBDEvROm8C6Ma5vEdqFPkE1/BWsDYCHesXa9T0enrMStWoZV+QqU/b26GIEg/qMxna5POUK2rdMzPaVUg8QqQ0pOtkt+EQYiEPkOKTnRLhe2uV1kIbLhguo6jSG7Xehjuzamarck9BGU3lhdvDyIXQpuKlJ6Q5YjaxpNekqphgXfAInUWWnZTzUIvRdfI5WPYyfF2iIQmY9EF2c6SpUhUv4gdrKrLQhVr7fLS9ia9JRSDZLob6SsTykhiK2oWY7+QsrC1MYNsZWZCk9lWu2/cQKH/fzCdkaTnlKqQcYzxC5jlrwB3LvULHtGEi97VpuEwLVjxuJTGebejZSpwjjB2bPVw2kuTXpKqYZ59wfn1iQmNK9diNi9e3yNCZwKJo/EOo1+CJyAcXZrpWBVSzMFE8H4SHyahR/yL9eBLEqpjscYF6boBcg7HRw9wbEV5J2LKXoi4cnkxtkV0+01uzi1oxs4B0DBNZiC67IYvWou49oOU/QSeA8AR1dwDcZ0/geOvAaeXNGGae1NpZRS7Z7W3lRKKaXq0KSnlFIqZ2jSU0oplTM06SmllMoZWntTKbVFYpUjFQ9D1euAAwLHYfL+2OQnsTd6v6FPkfL7IPYbuHbCFFyCce+c2EYEgq8hFY/aNT89YzAFEzHO3nXahZGKx6HqJbvCjO9wTP75GEdB02KL/oqU3wPhL8DRFZN3bvXT6c2W39xGSGyt/fsNfWBPNwmchgmciDEd93xIR28qpRokEkXWHwfRJdSUGfOCexdM0bMZ+5C3qt6ETX8msQSWD1P0tD1hfnO7srug4klqqsY4wBRguv0P4+xR067kbAh/Was/Nzj7YLpNTXu+mcRWIOuOBKkgXoXG+CHvXBz5F6R3oFkiVimy7rDq4uDR6rV+8B+Bo/DWbIbWJDp6UynVMkIfQOxXEutqhiD6PUQy8wVURKDsVlLVfJSyO2raWZug4nESy6RZIJX2Wd3mdpFvq2Ot3V/ELo4dfCf9+MoftmuP1i67JlVQ/nC7qUcplS+AVUpNwgOosmtq1lt6rP3TpKeUapCEvwZJ8Qw1CUPkmwzttAKs9am3RefXev2TXQ4tSQTCs2otfgOprmpJJRJuQuIOf0lisqhmXBD7Of3+siE8C/tpGHUYN0S+b/VwWosmPaVUg4xra/vSXdIGLzi3ytBO/fUkM8DRvdbrnnbyTe4AnH1qtdvKrhWZxAfObVKs34Lafdcm4cT42jJXX1IP64iBs1drR9NqNOkppRrmGwe466x02PUYvQdkZJfGOMF/CuCrs8UPeX+qaefaGjwjSC507cXkn1NrcW8wBSR95BknJnBM+vHln5siNg9498I420fSM4FTSf67usDZD1yDshBR69Ckp5RqkHEUYIqeBedA7OTiBtcgTNELGS04bAouhcCJgK/6zC8PCi7G+BOTlOl8H3j3q47NB45umM53YWo9AcIYF6br8+De1Y4fDzgH2INiHEXpx+YZDoW3gymy94kHfAdiCu9q+gG3MuPqh+nyEDh6AV7ADZ49MEWPt6sRqOnS0ZtKqUaT2FrAgXF2bb19SpU9wtDRrcEkK1YZSDk4ejQ45F6sEpBoi5yRicTAWg2mE8aR3+z+skFEqo/Bj3EUZjucJmvs6E2dp6eUajTjLG79fRo/OFPcU6zbzlEAbHnOXVPO7Ortyzgzd1+zlRhj2uVz8ZpKL28qpZTKGZr0lFJK5QxNekoppXJGxu7pGWP6AE8BPQABJovIPXXaGOAe4HCgEjhDROZmKialNqzeyJuPvc+yBSsYvOcOHHDKWHyB5PqRyxau4K3H3mfT2lJGHjGcPY8agdOVap6Xqs2yglB+D4TeBVMIBRNxePfOdliqmUSCUPUGEp4Frj4Y//EJJd7S7i+2HKl8CaxVGM8Y8B2S0ZHAtWVs9KYxphfQS0TmGmMKgDnAMSLyfa02hwMXYSe9UcA9IjKqoX519KZqqoVzFnPF/jcSi0QJByP48rx06lbAA7P+TufimlFrH/3nM/5x5gNEozFikRi+fB8Dh/bn9unX4fbUndekNrOsCli7tz2Csjb/6TgK/5qdoFSzibUJWT8erDXVpdc8YFyYLo9hPMPS7y/0KbLhfOyKNhEgYCfSohcxjkCT48x67U0RWbn5rE1EyoAfgN51mh0NPCW2L4DO1clSqRZ3xxkPUFVWRTgYASBYEaJkxQae/NuL8TahqhB3nvMvQlVhYpGY3a48yE9zl/DeM59kJe52o/Tm5IQHUPUklrWx9eNRLULKH4LYyuqEBxC2y7dtupJ0T5pEYsimy7FrpUaq11ZC9Gek8ukWjLp+rXJPzxjTDxgKzKyzqTfwW63lZSQnRqWabdO6Upb/tDJpfTQS49NXa2o0/vDFTxhH8sTcYEWI95+fkdEY273Qe/Vvq5zSenGolhV8k8Ri49Via8BK/n+qQdFFIHWLiAOEIPi/pkSXtownPWNMPvAKcImIlDaxj3ONMbONMbPXrl3bsgGqnODy1H/72u2ruWTp8bnr/fbqy8vss+PaPdPAEIF2OnFbYddYTUmwK7mk2ZdY9WyrW9YtMzKa9IwxbuyE96yIpPqqtxyoXbl16+p1CURksoiMEJERxcWtPzlWtX95nQLsOnYQTlfiP3mv38MRfzwwvrzjqIH485P/5/PleTni3IMyHme7Fji5ng0O8B3XqqGoFhQ4ieQ6ow5wD067Mo9x9QPn1kCdqynGj6n330/LyljSqx6Z+Sjwg4jUV5BuKnCase0BbBKRNM+XlWqcq568kJ79e+Av8OHL8+L1exiy72COv+KoeBuHw8Et/7uGgqJ8AgV+fPk+3D434847mJGHDc1i9O1A4AJw7VxnpYHCSTgcWvypvTKBU8A7FjvxVddAdW6F6Xx30/rr8iA4utn9mADgBd9h4Du6JcOuf/8ZHL25F/AJ8C01T1q8BtgGQEQeqk6M9wOHYk9ZOFNEGhyaqaM3VXNYlsW8D+ez+pd1DBzWn22H9EvZLhyKMPvtrykrKWfIvoPp2a99VM5vC6zQbKiaAs5u9pPE9dJmhyCRhRD51i5Z5tnDLsHW1L4kCuFPIbYOPMMwrv7Njq+xoze14LRSSql2L+tTFpRSSqm2RpOeUkqpnKFJTymlVM7QIVVKNUNpSRmv3TuNaDTGMRceRlHPLtkOKW0iEQjPAqkAz0iMo3PqdlYFhL8AHOAdjWmleVUqkT0IZBZIGXh2b9HnA+YCTXpKNdELf3+VR695Lr78/K2vMv7yI5nwj9OyGFV6JPIdUnI2dsUNAxJBCv6MI++UhHZW1duw6SqIj9gT6Hw/xjumtUPOaRJZgGw4s7okWPXfK38ijvxzsh1au6GXN5VqguWLVyYkvM1evvN1fpi5MAsRpU8kgpScBbLBPsuTciAEZXcgke9q2sVWwqYrgCq7jZSDVCAbzkesTVmLP9eIxJANZ4G1LvHvVX4fEtYR7Y2lSU+pJnjmplfq3fb0TS+3YiTNEP6cmqK/CRuQyv/El6TqdeySU3UYIPhOhoJTSSJzQSpTbAgilc+3ejjtlSY9pZqgYlNFk7a1KVJfnBbULpMr5aRMjhJL/VQFlRlWOUnluwAQsJpU1jgnadJTqgkOO3v/ercdfPq+rRdIc3hGgaRIZiaA8R1Ss+jdl+TaiwAG9AGxrcczop6/lx/jP6z142mnNOkp1QSjj9w9ZQmzHv2KOezsA1o/oCYwjiIouBQ7oVWfQRg/uHYFb63i2u6h4DvI3hZ/sx8CJ2Jc27VmyDnNOAqg4Brsv9fmj24/uHYE37gsRta+aBkypZrIsixemjSV1//1DpZlcdBp+3L6jSfgcLSv75IS/hqpehGscozvUPAdgqnzmCARgdAHSHAq4MT4jwPPntjlc1Vrksi3SOWLYG2wz8h9h2KMJ9thZZ3W3lRKKZUztPamUkopVYcmPaWUUjlDk55SSqmcoUlPqVay5te1lJaUNdhGRCjbUE4knGrSeI1YLEZpSRmxWKzBdpFwhLIN5Wzp3n1pSRlrfl3bYBuw62+KlWqCdK02IohVikh4i/01hkgUsTYiYm25cRslYlUfQ3QL7SLV7drXWIv2RGtvKpVh7zz5AXf/6REiQTuRFfXszF2f3ETvbXsltJvx6kwemPg4G9dswuF0cNhZ+zPhztNwe9zxNiLCy3e9zrO3vEKoMow34OEPf/0d4y87MmEkZSQc4eHLn+LNx97Hill07l7IBfecyV7HjkrY5/LFK7ls7+spWbURALfPzSX/+iMHn75fQjuJLkU2/Rmqy5OJZwSm8HaMM/EYJDwL2XQtxJYBDsR3CKbTjZgmPD1dxEIqHoCKx0DCYPKRgstwBH6fdl/ZZFU8A+X32NVUjBfJOxeTNyHh7yUSRcr+AZUvAFFwdEYKrsHhPyJ7gXdQOnpTqQz68ctFXDTq6qT1Xr+HqWVPx6c3fPvJD1x96P8RqgontNn3pDFc8e/z4+um/uttJl/5NKHKUE27gJcJk07lyPNqJpRPOvtBPnzh06T+bnvrWnbZeyfAnnJxZMGphKuSz8gemHUb24+w5+CJVYGs3R9kIzXlyJzgKMYUv4cxdlKW6BJk/bHVxZA384BnBI6iJxr1+6rNKr8fyh8BavfnxxTehvEfnnZ/2WBVvgqlN1D3GMi/GEf+2TXtSm+CypeBYK12PkyXf2lR70bS0ZtKtQEPX/ZEyvWhqjDvPv1RfPnZW15JSFCb27z/7AzKN9aUC3v25pcTEh5AqDLEMzfX1AIt31jB+8/NSNnfs7fUtHvnyQ9TJjyAf11aK+7gGyBBEutvxuxH24Q+iK+RisftM7IEYQjPQaI/p9xPfURiUPEoickCoAopvy+tvrKq4j5SHQMVD8UvYYpVCZUvkZjwAIJI+f2tEGRu0aSnVAatXLqm3m2Lvloaf738p5Up27g8Ttav3BBf3rA69VMNNqzeGH+9fkUJLo8zZbsVi1bV2v/P9cZWO26J/kLyBzcgIYj9VrMcXQykuMdoPBBbXu++UpIqu/9UrNS/qzYpVs/fX8qI1zO1SsDU81Fc+/erWoQmPaUyaLth/evdNvzg3eKvd9h9O4wjubqJFRN69C2OL281sGfKvnoPrLm31qNfd6xY8m0Lh8Oww+7bxpd3P3S3pDabDRw+IP7aeHYGAsmNjAdcO9Use4YBKSqDSAjSLVdm8sBRmHqba/v0+som17ap1zt61FRRcXYn9UexAdfgTEWWszTpKZVBF913dspkVtSrM6MOHxZfPvX68Xj93oQ2voCXE/9yDL5AzfrzJp2O15+YWLx+T8KDa30BL7//89EJ7wPw+L2cct34+PKow4fRpWfyU9KNw3DhvWfV2sGB4OwBuGu18oBzAHhG17wvcBqY2nUhAXzgPwrj7JG0n4YYYyD/KpILXfswBVem1Vc2mYLUx0DBn2vaGA/kXwT4k9qZgokZjjD3aNJTKoN69O3O/V/cSs/+3cHYZ1u77b8zT/6UeF+q76A+3D3jZoYfPIRApwC9B/biwvvO5g9//V1Cuz3GDeeGV69i++EDCHTys/3wAdzw6lXsMW54QrtTrh3PhfedTe+BvQh0CjDi4CHcPeNm+g7qk9DuqUX3sdv+O+NwGDDQs78db4++3eNtjHFjuv4H/MeD6QKOrhA4BVP0TMIIROPsjuk6xS5WbQrAsRXkX4LpdHOTfneOwLGYzneCawcw+eAeiil6FOPZvUn9ZYPxjsF0mQzuIfYxuHbCdL4HR52BOI68s6DTzfYXCZMPnj0wXZ/BuAdlKfKOS0dvKqWUavd09KZSSilVhyY9pZRSOUOTnlJKqZyhSU8ppVTO0NqbHdjyRSt5+qaXmf/pj/ToW8zJ1xzHsAN3zXZYSjWahOcg5Q9CbCm4dsbkX4RxD8x2WKod06TXQS1buILzd/8LoYogliWsWrqGH2ct4uIHz+Hg0/bNdnhKbZEE30c2XkK8PFdsBRL6CLo+h3HrpG3VNHp5s4N64voXCVYnvM1ClSEeuuxJYtGGH0ejVLaJCFJ6E4n1KC2gCim7I0tRqY5Ak14H9d2MHxAreQ5mOBhh3fKSLESkVBqkAqx66lZGvmndWFSHokmvgyrq1SXleitmUVCU/rPNlGpVxke9d18cRa0aiupYNOl1UCddfRzeurUXfW72OX40gYK6Nf6UaluMcUHgRJLrVvohcG42QlIdhCa9Dmrv40Zxxs0n4sv34S/w4fa62fOYkVzysH5gqPbBFFwJ/iMBj/3UBXyQdzYmcEK2Q1PtmNbe7ODCwTArl66hS49COhUVZDscpdImVql9f8/ZG2P0KoVKrbG1N3XKQgfn8Xnou9PW2Q5DqSYzjk7g6JTtMFQHoZc3lVJK5QxNekoppXKGJj2llFI5Q+/pKdVGhINhnrj+ReZMn0f3Pt045/ZTWuV+7IrFq3j3mY+pLKti9LgR7LrPoIQnoqvmEQlC1TQkMh9c22H8R2IcOlc2W3T0plJtwIY1Gzml/wWEq8IJ6y+dPIHDzzkwY/t995mP+OeEycSiMaxoDG/Ay6gjhnPNcxNxOPRCUHNJbA2yfjxYpUAl4Afjw3R9CePaJtvhdSj65HSl2pFbT7o7KeEB3POnR7AsKyP7rNhUwT8nTCZcFSYWiSECwYoQM6fN5cs3v8rIPnONlP0drLXYCQ+gCmQTUnpdNsPKaZr0lGoDvpvxY8r1Vszi6w++y8g+5773HS63M2l9sDzI+8/PyMg+c07oPaBugXcLwrMQiWYjopynSU+pNsA0cCnR6/NkZJ+pEh6AMeD2ujOyz9yT+ncMpvpHtTZNekq1AaOOGJZyvdvrZvCYHTOyz2EH7kKqW/oev5eDT983I/vMOb4jgbpfIFzg3R9j6kuIKpM06SnVBlz9zMV07lGYsM4Yww2vXpGxfXr9Xm545Qp8eV78+T48fg8en5tjLz6MXccOyth+c4kpuBJcA8EEAK9dQ9TZB1N4U7ZDy1k6elOpNuTdZz7is6mz6dmvO6dcP55AfuZrTVaUVvLZf78kWB5kxCG70WtAj4zvM5eICIRnQXQBuPqDZwzG6PlGS2vs6E1Nekoppdo9nbKglFJK1aFJTymlVM7QpKeUUipnZCzpGWMeM8asMcaknFlrjNnXGLPJGPN19c/1mYpFtW+lJWXMevMrFsxeTEvcg/7l+9+YOW0u65avb4HoGqeqvIrZ78xj/mcLGqyw8vP833j65pf56D+fNXufIsLieT8z682v2LBmU73totEoU//1Ns/dNoWSVRvqbRcOhpn73rfM+3A+0Uj9E6vFKkVCHyPheS3y95LoEiT0IRJb2fy+JIqEv0RCn9o1MVXOydhAFmPMWKAceEpEdk6xfV/gChEZl06/OpAltzx7yys8d8sruL1urJhF195F3P72tXTfpjjtvso3VnDtkX9n0VdLcLldhIMRDjp1LBMfOjejdSbfevx97r/oUZwuJyJCoFOA26ZdQ/9d+sbbWJbFRXtczcLZS+LrPH4398/8O/13Tr9G44bVG7n6sFtY/tNKnC4n4VCEYy48lD/efmpCMelPpnzBzSfchVg1nwMHn7kvVz56QUJ/M6fN5daT744vO5wObnz1qqSpDVbF41B2FxgPEANHV0yXx5tUZ1KsCmTjnyD8NRg3SBhGokXxAAAgAElEQVR8h2AK/44x6dfKl/DXyIYJQBh7YrgFne7A4T847b5U29NiozeNMcXAH4F+1Hoqg4ic1Ygg+gH/06SnmuLLt77ixvF3EqoMxdc5HIZtBm3NI9/clXZ/N/zuH8x8Yy7RcM1Zijfg5ezbTubYiw5vkZjrWjzvZybu+VdCdepqdu5eyAvLHsbpsico33fhv5n64NtJ7w8UBvjvhifT3u9l+/6N7z9bQCxaUwLLl+fl0ocnsP/JewP2mdu4vFNSno1d+8Kl7HPCngCsW1HCGQMvSjoGf76P5397iLzCPAAkPAsp+SNQVauVseeldZue9pMbrI1XQvBN7CQVPwrIPx9H/nlp9SVShawZA1JeZ4sP020axpX5p1mozGrJ0Zv/BQqBd4E3av20hNHGmHnGmDeNMYNbqE/VQUy5d1pCwgOwLGHlkjX88sOytPqqKq9KSngAocoQr907rdmx1ueNydOJhJMvBYarwgk1Nd9+4oOU76/cVMmir5emtc/1Kzfw48yfEhIe2MWkp9xT87/ulLvfqPfy4zM3vxx//cHzM7Cs5HYi8MmUWTXLFU+TmPAABKx1EJ2f1jGIhCE4jcSEBxCEymfT6st+2/t2LEliSNWr6fen2q3GXCMIiMifM7DvuUBfESk3xhwOvAYMTNXQGHMucC7ANtvo4zhyRen6spTrnS4H5Rsq0uorWBmmvhONik11P6hbzqZ1ZVix5Ht4AgnHEA3XLUpcY/2KDWy3W/9G77NiUyVOl4NIKHlbWUnNPjes3lhvH+WbatqVlpQTCUWS2sQiUco31Dpzsuq7H+gAq/57iilJBKjn3mfS2Vpj+isFSfU7jjQQt+qIGnOm97/qpNSiRKRUxP7XKyLTALcxpls9bSeLyAgRGVFcnP69HNU+jTlmJB5fcuFjyxK2G9ovrb46F3ei61ZFSesdTge7H7ZbU0Pcoj2P2h1fnjdpfTQSZdd9au6H9d+lT8r3GwNDD9wlrX32HtgTjz+5SLXL42L0UTVXfw45c796+9jzqN3jr0ccPCTlMTicDoYftGvNCt9BgC+5M4mCe0jjgq9mHHngTJXoHeAZk1ZfAHhGk/JMzwQwvn3T70+1W/UmPWNMmTGmFJiInfiqjDGltdY3izGmp6m+yG+MGVkdS+sNp1Nt3tEXHEpxn254qz/AjcPgDXi48N6z8PqTP4QbYozh8n//CW/Ai9Nl/7P3+Nzkd8njzJtPbPHYN9vnhNH032WbhKThC3g5+epj6dKjc3zdX56ZiMORfCp64p+PxeNJ74kHTqeTyx/5E96AB4ez+lj9HjoXd+LEvxwTbzdg137stl/yXQV/vo8/3nFKfHnXsYMYftCuiceQ52X/k/dOGIxjAieAqw81ic/Yrwv+3KQnhZvC/wPjp+aClAdMPqbgqvT7cvWDwIlA7bJufnAPB89eafen2q9Mjt58HtgX6AasBv5GdblxEXnIGHMh8Ccgin0j4DIR2eI4bR3Ikluqyqt489H3+eJ/c+i6VReOufAwdth9uyb399uC5bx6zzR+W7CCXcbuxFHnH0Ln4sItv7EZwqEI7z/7CR+++Cl5hQGO/NMh7LZf0tguVv+yhrvPe4QfZ/5EYXEnzrrlJMaOH93k/S755hdevXcaq5auYfjBQxg34SDyO+cltXv5rtd55e7/EaoMM/qoEZx/95nkdQoktInFYnzy8hdMf+ojHC4Hh511AKOPGpE0OEWkCqmcAqF37JGbgVMxnqFNPgaJ/opUPAnRn8AzFBM4BeNs2tUeuwbmR0jlf0BCGP/R4Du8SSNBVdvTkqM33xORA7a0rrVo0lNKKVVXY5NevV9xjDE+IA/oZozpQs0TDzsBvVskSqWUUqoVNXRePwG4BNgKe6TlZqXA/ZkMSimllMqEepOeiNwD3GOMuUhE7mvFmJRSSqmMaOjy5nHVL5fXeh0nIlMyFpVq92KxGGIJLrcOEsiEWCxGWUk5BUX5OJ3ObIfTJPZ4ggjgbrBaS+PbWUAUY5KnazQtvgjgwJj2+ftVqTX0iXRk9X+LgTHA+9XL+wGfAZr0VJLSkjLuu+DfzJgyEysmDNpzey6dfB7b7Ki3gVvKBaP+wsIvF8eXt999Wx6Y+fcsRpQ+q+ptKLsVrNVg8pC8czB5ExKeKC4iSMVjUPEQSBk4ipGCq3D4j0zoSySMlN0OlS8BYcS1LabTjRjPFsc0pCTRxcim6yAyF3AivoMwnW7AODpv8b2q7WvM6M3pwGkisrJ6uRfwhIgc0grxJdHRm22XiHDesCv59Yfl8XJfxkBe5zyeXHgfnboWZDnC9m/iXn/l+88WJq0ftOf23DPjlixElD4JzUA2nA/UfsqBH/LOwlEwMb7GKn8Eyu8nsbSZD9P5LozvwJp2Gy6G0IdJ/Zlur2Bc6U1vEWsDsvYgO8nGJ7O7wbUtput/064fqlpPS9be3Hpzwqu2GtBaYCrJ/E9/ZOXi1Qn1LUUgEozw1uOpa0uq9KRKeA2tb4uk/B4SExRAFVQ+btfcpPpSZcVDJNfyDCLlNU97kNgqCL2for8QUv5I+rFVTrGf5pBQvSUCsV8hMift/lTb05gbLu8ZY94Gnq9e/j128WmlEixbuDJlAeNQVZifv/s1CxGpNin6S+r1ErNrdDqLQSpA6qmJGluW+Np4qxNVbRZEm/BFILqQ5ASKnQOjP0MTL5mqtmOLZ3oiciHwMDCk+meyiFyU6cBU+9N/l9QXALwBLzvsvm0rR6PaLFfKuvL2M/gcXapf54Gp53K4c0Ct1/1BUlTWxgXuXVOs3wL3ziSWKttMwL19+v2pNqdRT84UkSkicmn1jz6HQ6W0/Yht2W5Y/4Qi0Q6ng0CBjwNP3SeLkXUco8YNS2t9W2QKLiWpMLXxQ/6F8ZJgxjgg/1KSE5APU3B5zducXcF/TIr+PJi8s9OPzX8sOPJI/Gj0gnswuNIr/K3apoYKTs+o/m9ZdaHp0pYsOK06HmMMt715LeMmHEx+lzx8AS9jjh3J/bP+nlTLUTXN/029mjHHjkxYN+bYkfzf1KuzFFH6jGcEpssj4BoMeMHZGwquwwTOSGjnyDsROt0Mzm3sdq4dMV3+hfEmPmXBdLoR8s8HRze7nWcPTNELTXpau3HkY7q+At6D7URsOkHgREzRozqIpYPIWMHpTNHRm0oppepq9uhNY8zdxpgTjDFbtWxoSimlVHY0NHpzEXAMcEf1af1n1T+fAvPELn+glFJKtRsN1d68n+rC0tVne3tW/1wCdMd+2oJSSinVbjQ4T6/6yea7YCe7McAg7DPApzMfmlJKKdWyGio4PR37bO5r4AvgVhH5obUCUyodc9/7hnvP/zdrfl1Lp26dOOe2kznwlORpEo9d+zwv3/k6kXCEvE4BLp08gX2O3zOp3c2/v4tPX51JLGZR3LuIm6b+he12698ah7JFP8//jeduncLSb35hwJC+nHT1cfQb3Cep3byP5vOff/yXNb+uY+j+u3DCVUfTbauihDYiwkf/+Yz/PvAWlWVV7HP8aI656HACBYlTBcKhCG9Mns70pz7C6XRw2NkHcMhZ+zW52LVElyEVD0N4Lrj62nU3PUOa1JdS6ah39KYx5mFgV+w6QF8AnwOfi8i61gsvmY7eVHV9MuULbhp/Z9L6M285iZOvrnlAyI3jJzFjysykdpdMPpcjzjkovnzKgPNZ/fPapHaPfHMn/XbObgW+7z9fwFUH3UwkGMayBIfD4PZ5uGP6dQwavUO83dtPfMB9Fz5KqNKeuO1yO/F38vPwV5Mo3rprvN0DEx/jrcfeJ1hht/P43PTs350HZ9+O1+8FwLIsrtj/RhbOXkSo0q584gt4GX7wrtww5aq0j0GiS5D146srrsSwn0/txXT+J8Z3QNN+MSrnNXv0pohMEJHR2INZPgSGA88YY+YYY55ssUiVaqa7Jzyccv1Tf3sRy7LHW8VisZQJD+CBix+Pv14wZ1HKhAdww+8mNTPS5ntg4mOEKkNYlv1l1bKEUGWIByY+Fm8TjUT516VPxBOevS5G5aYqnv2/V+Lr1vy2jmmPvBtPeADhYIQ1v67j/edmxNfNmf4NP81dEk94AMHKEHOmf8OCLxelfQxSdhdIJXbCA7vGVxAp/VvKMnZKtaTGVGQJAZXYZ3whYGug/ZR/UB1e6frylOtjUYuSVRsB+G7Gj/W+PxKMxF//78F36m23csnqJkbYcn6au3SL65cvWoUVSx5cHYvGmPvuN/Hl7z9bkPJ5h8GKEF++9XV8+ZsP5xMsT65HGY1E+ebjJtzxCH8JpBj8bW0Ea336/SmVhobm6f3TGDMTWAncCBQADwE7iIjW41FthsNZ/3e3TkX5APTs371RfW0zaOt6t7lrlVfLlvoq2+QV1qzv1LWAaCSWsl1Rz5pnwnXp0Rkh+czK6XJS3KfmEmhRry54/MkPZnV73An9Ndrm+popt+Wl359SaWjoTG8pcD5QLCIHiMi1IvKmiGxspdiUapQ9j9k95fq+g/vg8dkf1j22KcaX70vZbvCYmnthx11yhH2LKYXjLxvXvEBbwDEXH4Y3kJiAvAEPx1x0WHy5S/dChh6wM25P4lmcL+DlhCuPji/vMnYnCrrkYxyJB+zyOBk3oeYe5/4n74UzxRcLp8tZ7+++QXnnkFxT0wu+IzAmVbFnpVpOQ/f07hWROSKS+iujUm3EdS9exsDhAxLWFW/TjXs+/b+Edf/+7i7c3sSztW69i7jzwxvjy06nk1um/TUpEQzZbzCn33hiC0eevlOuHc8Bf9gbj89NXmEAj8/NAX8YyynXjk9od82zE9l1n0F4fG4Cnfx4A15OveEE9jy6Jkk5HA4mvX8DfQdtjTfgwV/gp6BLHtc8dwl9dqh50n1ht07cOu0ainp1wZ/vw5fnpUe/Yv7x/t/w56X+ItEQ4/8d5J0OeMHkAx7w7oMpvKGJvxWlGk9rb6oOY/Uva5j30Q9sP3xAyiH8m331/jd8+8mP7HPCnvTdqf7LmR+8OIPlC1dy5AWHUljUtp76Xrq+jJVLVtNrQI8Gn0i/dtl6SlZtZJudejeYoJb9tJJgeZD+u2yD05V6GoJlWfz83W84nA76Dtq62QWYxSqH2FJw9MQ4i5vVl1KNHb2pSU8ppVS719ik19Dk9KL6tgGISElTAlNKKaWypaEyZHOwJ9CkuoYhwIAU65VSSqk2q6GC022j5pJSSinVQhosOL2ZMaYLMBCI3wkXkY8zFZRS6YpGonz++hwWf72UrbbtydjjR+MLeJvc34Y1m/jwhU8pLSlj+IG7MnjMjikHbsx97xtemjSVWNTi2IsPY/SRTRjCn6ZwKMKMKTP59YdlbLPT1ux13Cg83uzPIVSqPdjiQBZjzDnAROxKLF8De2DX4Nw/8+El04Esqq7SkjImjrmW9ctLqCoP4sv34gv4uPezW+g1oEfa/c2ZPo8bjv0HlgjhYLi6zuQQrvvPZQkFlm86fhKfvJJY2mzIvoOZ9P4NzT2keq1fuYGLR19DWUk5VeVB/Pk+CoryuffzW+naq4FJ30p1cM2uvVnLRGB34BcR2Q8YCugEddVmPHbN86xaspqq6lJZwfIQpetKmXT2g2n3FQlHuPmEuwhWhghXhUHsslxz3pnHR//5PN7u+88XJCU8gHkfzmfGq6lrfLaE+y96lPUrSuLHWlUeZP2KEh64+NGM7VOpjqQxSS8oIkEAY4xXRH4EdtjCe5RqNR+/9HlS2S3LEuZ/+iOhqlA970rt+88Xpix6HKwIMf2pD+PLL//zf/X28eq909LaZzq++N8cYtHEupWxqMXnr8/J2D6V6kgac09vmTGmM/AaMN0YswH4JbNhKZWG5s2RTuBw1P89sPY9vYYmZjd30rZSKnO2eKYnIseKyEYRuQG4DngUOLrhdynVevY7aS/c3sTvbw6ng133GRx/Jlxj7bTHwJQVSXx5Xg45s+Y29u+vPKrePo6/ov5tzbXXsSNxuhPjc7qd7HXsyIztU6mOZItJzxjz9ObXIvKRiEwFHmvgLUq1qrNuOYk+O/TGn+/D6XLgL/BR1LMzVzx2ftp9udwubphyJb7qGpNOtxNvwMuYY0ex9+9GxdttP2I7Djot+cnso44YxqjDM/fkrQvuPYsefYvxF9Qca4++xVxw71kZ26dSHUljRm/OFZFhtZadwLciMijTwaWiozdVKpZlMeedeSye9wtbbduD0UeNwO1p+jD+8o0VfPzyF5SVlDPswF0YOCx1LYYfv1zEi39/jVgsxvjLxrHr2MFN3mdjxaIxZr4x156yMGhrRh0+rN56mUrlimbX3jTGXA1cg/0MkEpq7pyEgckicnULxZoWTXpKKaXqavaUBRG5TUQKgH+ISCcRKaj+6ZqthKeUUko1R2OmLPzVGHOKMeY6AGNMH2OM3jVXSinV7jQm6T0AjAZOrl4ur16nlFJKtSuNSXqjROQCIAggIhsAT0ajUm3Wkg0lzFq+jLJQepO+W8OGNZuY99F81vy6tsF2q35ew7yP5rNpXWkrRaaUaisaMzk9Uj1iUwCMMcWA1fBbVEeztrKCP77+GgvXr8PtcBCxLCaOHM2EEdm/0m1ZFvdd8G/efvJDPF43kVCE4QcN4ZrnL0koOl1ZVsVNx9/Jtx9/j9vrJhyMcOR5B3HeXWfohHKlckRjzvTuBV4FuhtjbgFmALdmNCrV5vzpjal8v2Y1wWiUsnCYYDTKvbM+54Ofl2Q7NKbc/QbTn/6YSDBCxaZKwsEIc6bP48GJidNJ/znhYb75aD7h6naRUIRpj7zHG5OnZylypVRra0xFlmeBq4DbgJXAMSLyUqYDU23HstJNzF+zmmid6S1V0SiPzs3+9JEpd79BqDLxcms4GOG9Zz8hGokCUFUR5NNXZxIJRRPaBStDvNJAHU2lVMdS7+VNY4wPOA/YDvgWeFhEovW1Vx3XxmAQt8NJKBZL2ra+qioLESUq31SZcn0sGiMcjOByuwiWB6mvSGdZSXkGo1NKtSUNnek9CYzATniHAZNaJSLV5gws6oqQXMTA43SyX7/+WYgo0a5jd0p5T67XgB4ECvwAdO5eSOfunZLaOByGoQfskvEYlVJtQ0NJb5CInCIiDwPjgbGtFJNqY7wuF9eP3Q+/yxU/V/I6nXTx+Tln2BYLIGTcuf84DX+BD1d1IWaH04E34OWShybE2xhjuPThCXgDXhwO+yhcHieBwgBn3XJyyn6VUh1PQ2XI6tbcTFjOFi1Dlj2zVyzn0a9ms6q8nH379ef0IUPp7PNnOywA1vy6lpfufJ0fZ/5En516c8IVR9NvcJ+kdovn/cxLk6ay/KeVDN5rR8ZfOo5uvbtmIWKlVEtqidqbMaBi8yKJNThFRJKvFbUCTXpKKaXqamzSq3cgi4ho2XallFIdSmPm6SmllFIdgiY9pZRSOUOTnlJKqZzRmNqbTWKMeQwYB6wRkZ1TbDfAPcDh2ANkzhCRuZmKJxfN+PUXbv74AxaVrKeL38+E4btzztARSXParpr+FlN+/B5LBKcxnDFkKH8du1+Wom6az6Z+yeQrn2bF4lUU9ezMKdcfzxF/PLBJNTXDoQhPXPc8bzz8LsHKEDvtsT0X3X822w7pl9CudH0ZD0x8jE9emYmIMOqIYVx439l026qoScfwzcff8+Alj7Pkm18o6JLP7y4dx4l/OQaHQ7+bKtVS6h292eyOjRmL/Riip+pJeocDF2EnvVHAPSIyakv96ujNxpmzcjmnvvoywWhNER2/y8U5w0Zw6R5j4usufXsa/13wQ9L7zxu+O1eNaR9TM2e9+RU3jZ9EqCocX+cLeDnrtpM59qLD0+7vxvGTmPXmV4Rr9ecv8PHIN3fRo28xALFYjHN2voxVS1YTjdiVahxOB0W9OvPkwvvw+NJ7EMnCOYu5bJ/rCVXW7NMb8DJuwkGcd+fpaR+DUrmm2U9Oby4R+RgoaaDJ0dgJUUTkC6CzMaZXpuLJNf/8/LOEhAd2rcx/z51DqHq9ZVlMTZHwAB79ak7GY2wpj13zXELCA7um5tM3voRlpfdAkFU/r2HWtLkJCQ8gEowk1Oic/dbXrF9REk94AFbMomJjJR+//EXax/DMzS8n7TNUGeL1f71DZVn2S70p1VFk87pJb+C3WsvLqtepFrCoZH2929ZW2tMvKyLhFMXFbJE0k0U2LV+8KuX6qrIqqsqDafW1bOFK3F530vpoJMZPXy2NL//6w3LCwUjyPsuD/Pzdr2ntE2Dpt7+S6qKL0+1g7W/r0u5PKZVau7hZYIw51xgz2xgze+3ahh8QqmzbFdVXZUQoDuQBkOf21FOCGdzt6D5S7217plzvy/fhz/el1dfW2/ciEkpOZi63k4FDa+qMbrNTbzy+5OToy/fRN0UlmC3pt/M2pLr9GIvEKO7TLe3+lFKpZfOTbTlQ+9Nh6+p1SURksoiMEJERxcXFrRJce3fJHnvicyWOU/K7XJw9dATe6vUOh4Ojdtgp5fvPHjo84zG2lDNvOQmvP/Eemi/g5dS/HZ/2IJCe/boz8rBheOr05/a6Oe6SI+LLIw7dja69iuL1PsG+p5dfGGCf40enfQynXj8+aZ/egJcjzzs4XjRbKdV82Ux6U4HTjG0PYJOIrMxiPB3KiK16M3ncMQws6ooBivx+LtljTy7dY8+Edv885HB+t9NgHNWnGU5jOGu3Ye1mEAvAqMOHcfWzE9lqu54Yh6GoVxcm3HlakwaxAFz93ESOOv8QAgV+jMMweMyO3PXxTfTs1z3exul0cveMm9n7d3vg9rpwuZ3sMW44931xa9qDWAC2H74tt7xxDQOG9MU4DJ265nPyNcdx7qTTmnQMSqnUMjl683lgX6AbsBr4G+AGEJGHqqcs3A8cij1l4UwR2eKwTB29qZRSqq5m195sLhE5aQvbBbggU/tXSiml6mo/oxWUUkqpZtKkp5RSKmdo0lNKKZUzNOkppZTKGZr0lFJK5QxNekoppXKGJj2llFI5Q5OeUkqpnKFJTymlVM7QpKeUUipnaNJTSimVMzTpKaWUyhma9JRSSuUMTXpKKaVyhiY9pZRSOUOTnlJKqZyhSU8ppVTO0KSnlFIqZ2jSU0oplTM06SmllMoZmvSUUkrlDE16SimlcoYmPaWUUjlDk55SSqmcoUlPKaVUztCkp5RSKmdo0lNKKZUzNOkppZTKGZr0lFJK5QxNekoppXKGJj2llFI5Q5OeUkqpnKFJTymlVM7QpKeUUipnaNJTSimVMzTpKaWUyhma9JRSSuUMTXpKKaVyhiY9pZRSOUOTnlJKqZzhynYAKvveWLiAu774lBVlpfTr3IU/jxnLvv36ZzsspZRqcXqml+Om/DCfq959i6UbNxCKxViwfh3nT5vKBz8vyXZoSinV4jTp5TAR4Y7PPqEqGk1YH4xGuePTT7IUlVJKZY4mvRwWjsVYV1mZctvSjRtaORqllMo8TXo5zON0Uuj1pdy2VX5BK0ejlFKZp0kvhxljuGjkHvhdieOZ/C4Xl+4xJktRKaVU5ujozRx3+pChANw36ws2hYJ09Qe4cs+9OHKHHbMcmVJKtTxNejnOGMMZuw3j9CFDCcdieJxOjDHZDksppTJCk54C7OTndek/B6VUx6b39JRSSuUMTXpKKaVyhiY9pZRSOSOjSc8Yc6gxZoExZpEx5i8ptp9hjFlrjPm6+uecTMajmmf+mtVcNf0tTnn1Jf49dzZloVC2Q1JKqbRkbOSCMcYJPAAcBCwDvjTGTBWR7+s0fVFELsxUHKplvL7gR/783tuEYzEsEeauWMFT33zF6yeeSqEv9QR3pZRqazJ5pjcSWCQiS0QkDLwAHJ3B/akMCcdiXPvBdILRKJYIAMFYlDUVFTz29ZwsR6eUUo2XyaTXG/it1vKy6nV1/c4Y840x5mVjTJ8MxqOaaOH6dfFkV1s4FuOdxYuyEJFSSjVNtgeyvA70E5FdgenAk6kaGWPONcbMNsbMXrt2basGqKCT10vMSk56AJ310qZSqh3JZNJbDtQ+c9u6el2ciKwXkc2jIf4NDE/VkYhMFpERIjKiuLg4I8Gq+m1T2JntunbFWadSi9/l4ozdhmUpKqWUSl8mk96XwEBjTH9jjAc4EZhau4ExpletxaOAHzIYj2qGyeOOZkCXIgJuNwUeD16nk3OGjeDgAdtlOzSllGq0jI3eFJGoMeZC4G3ACTwmIvONMTcBs0VkKnCxMeYoIAqUAGdkKh7VPD3zC3jrD6czf+0a1lZWMKRHT4r8gWyHpZRSaTGSYoBCWzZixAiZPXt2tsNQSinVhhhj5ojIiC21y/ZAFqWUUqrVaNJTSimVMzTpKaWUyhn6ADVFzLL49Ldf+XnjBnbo2o2RvbdO+SDZ0mCQ+2Z9wcryMg4fuAOHD9y+VeJ74us5vPz993TPy+PWAw6iZ35BxvdZGYnw3tLFbAwG2XPrPmxb1DXj+1RKZZ4OZMlx6ysr+f3LL7C6opyoZeF0OBjQpYjnjjuBfI8n3u7txT9x/htTqf2vpVd+AR+cdhaeDD18NhyLMezhB6iMRhLWXzF6DOfvvkdG9gnw9aqVnP7ay1gi9qR8A7/bcTA37XeAPlVeqTZKB7KoRrn2g+n8WrqJikiEUCxGZSTCwnXruOPTj+NtLMviojf/R92vRyvLy7j2w/cyFtvZU6ckJTyASZ9/SiwWy8g+Y5bFOa+/Slk4TEUkQjAWJRiN8uqC73l3yeKM7FMp1Xo06eWwmGXx3tIlRC0rYX3YivHfBTV1Aj78ZWlSm82mLVyQsfi+WPZbvdsenvtlRvb51aqVhKLJCbUyEuGF+d9kZJ9KqdajSS+HCVDf5e3aSS6YIglsZiWd/7WchnoORqIZ2WfUsqjvCmYkljrxK6XaD016OczlcDCy99Y46nzKO43hwFrlxQ7edjvqu5O1V5++GYtvcLf666yeO2JkRvY5tGevlNk24HJzzI47ZUtFORgAABaASURBVGSfSqnWo0kvx926/8EUen34qwejBFxuivPy+Ove+8TbuBwO/rr3vknvDbjdTDro0IzF9sQx45MSMsD4nQYnDLJpSV6Xi38ecjg+lwu3w/7fI+B2M3yrrThqB016SrV3OnpTURYKMXXBD/xUsp7B3XswbuAO+N3upHY/rl3L3z/9mDUV5ezXbwATR43O2MjNzarCYa567x0+/+1XCjxe/rbPfuzbf0BG9wmwvKyUV3/4ng1VVezTrz97bdM3ZQJWSrUNjR29qUlPKaVUu6dTFpRSSqk6NOkppZTKGZr0lFJK5QxNei1gY7CKRSXrCUUbnju2pqKcJRtKiNUz0TsdMctiyYYS1lSUN7svsAezLCpZT2UkuQJKbYvWr+eDpYupDIdbZL9t2dqKihb7eyml2gYtON0MwWiEK6e/zfQli+LD2y8fvRdn7DYsod2ainIumPY6361ZjcM4CLjd3HHQIezXr2mjEN9bupi/vPs2VdEoUctiSI+e3H/YkRTn5aXdV9SyuPnjD/jP/G9xORzERDhrt2H8f3t3Hh5VdTdw/PubyWSDEJAEZBUQBAIIAkZQcIPigoIoCqhVW33UWrV7H2sXa98ur/V9H9tatVpfrFsRRWndQKTWpW4YVlnDomwJSwKShEySmcl5/5ibyczkTpJJMsxM5vd5njzc5cy95+SE/HLPvfd3fjB5Skieyf0VFcxe/BxH3O7AtnkFo/nd9Iva1IZEVl5dzd3LX2d1aQlpDgcZzjR+c+F0Lh56YhJsK6ViR6/02uEnK1ewctcO6nw+jns8HPd4ePCjD1ixc3ugjDGGry9dwroDpdT6fLi9Hsrd1dz55mvsOFIe9TmLy8u4a9nrlLvdVHs81Pl8rCkt4YZ/LImYXaU5f/jkI17avJFaqw01Xi9PrVvD0+vXhpSb9cKzIQEPYPHmjSxcuzrqcya6b776Cp+V7KfOykV6tMbND1YsY/PhQ/GumlKqnTTotVFFbS3Ldm6nNizxsdvr5bGiVYH1DYcOsr+yAl9YQKrz+XgmLLC0xjPr11IXdk6fMeytOMamKH8pG2N4ev0aasKGZd1eL0+sbsxt+fnBAxytqbE9xqOffRrVORNdcXkZO46UN8k1WuvzdcoAr1Sq0aDXRl/VuHGK/bfvYFXjfbZDVVW2LzX7jGFfRUXU591bUUG9zRWdQyTkvK3hqa/HHeE+5JGaxqu6HUePRDxGladz3ds7eLyKNEfTfq03hn0Vx+JQI6VUR9Kg10Z9c7rhcjb99jlEOLNf/8D6mN698dhMg5OZlsY5A6PPWzllwEAybbKg1Pl8jOndO6pjpTudDMzNtd03Or9XYHnqwEERjzGke4+ozpnoRuX3anIlDZDhdHJODPOMKqVODA16bZTmcPDTKecFclaCP+Blu1x8d9LZgW0nd81h3qgxIeVcDgcnZWZxTcHoqM87b/Tp9MjMCjw4A5CVlsa1o8fSq0vXqI/3y/OmhQRRsY73s3MvCGzLy87mgkGDm3xWgAc62YMsJ2Vlc9O48WSlNaZhczkc5GZk8vWx4+JYM6VUR9A0ZO30we4vebToU0oqK5nYtx93FU5iUNjVjzGGpVs389S6NVTV1TFjyFBun1hIj6ysNp3ziLuax4pWsXLXTnIyMvjG2PFcMWJkm2f1Xltawp9WfczOo0coyOvFd86azMigK70GD374Ac9sWEuN18ug7j34/fSLOKNP3zadM5EZY3iteCsL163hK7ebaYNP5fYzC8nPjv7pWKXUiaG5N5VSSqUMzb2plFJKhdGgp5RSKmVo0FNKKZUyNOgppZRKGZp7M4KfvfM2L2z8nHoMAswaPpKHLrq0Tceqqqvj8kXPsvvYV4D/Hb2HLrqUi04dFlKutLKSW157ha1lZQCMyMvnycvn0CcnJ6Tc8u3FfH/FMmp8/hfLB+f24J8LrqdrenpIuaVbN3P/e+9QUVuLy+Fg/ujTuf/8aU3qd/Fzf6M4KCXa+JP7suSaBSFlfD4fV720iA2HDgL+x/h/OvV8bhh7Rki5ytoabnntH6wuLcEYwym53Xl05ixG5OWHlDvirubBj/7Dip3bSXM4mFswmrsKJ5GZFjpj+/u7v+BHby/ncHU1ThFmnDqMhy+eiSPsBfJnN6zl9x9+wHGPh3Snk5vHTeBH50xt0taOVO3x8IdPPmLp1s3UG8PMYafxg8lTyM3MjOl5lVJtp09v2vjx28tZsmVTk+3TThnMX2dfGfXxCh79Y5NUXwAvXz2fM/r0A6DO6+X0x//c5MXoDKeT9bfdSbr1Ll3R/v1c8/ILTY6VlZbGpju+E1h/vXgrdy9/o0m56YOH8MTlcwLrUxc+zn6bTC5je/Vm6fzrA+uTnnyMQ9XVTcr979cuYc7IgsD6+Mcf4ava0JRlDhE+ufl28rKzAX+i7hnP/Y2DVVV4rHRfGU4nY3ufzKKr5gVevVhTWsLclxY1OefIvDzeuPbGwPrT69Zw//v/blJu/qgx/HbajCbbO0K9McxZ/DzbyssCfeZyOBiQm8uya2/E5XTG5LxKKXv69GY7vGwT8AD+tfuLqI/1z22bbQMewPfeWhZY/vNnn9hmAqn1+XhsdWMuz++veNP2WG6vl+XbiwPr97/3jm25lV/sosqaFsjj8dgGPID11hUdwLayMtuAB3DfuysDy69u29Ik4IE/QPw6KCi9sb2YI253IOCBv50bDx9i7YHSwLZ733nb9pxbysr44ujRwPqDH//HttziTZ9TH6NpgT7et4ddR4+E9Jmnvp4DVVW8vWtnTM6plGo/DXo2mrv2rYpyHrnlO7ZH3FdaVRlYLiopiViuqGR/YPlA0GfCvbFjW2D5qNs+QTTAtrLDAJS0ci6+5TuKI+6rCpp/74M9uyOWW3fwQOPygVLbefvqjQmZyWCPNRxs58O9jeeKNAegAQ4ePx7xGO2x+fChJsnGG+qyMegPBqVUYtGgF6Xw+2YtGde7T8R93TIyAstDekTOYTk4KMNLTtBnwgVnR8lyRb5de4p1vL6tTFtWGJRLNFx60DDeyLD7dsEG5XYPLA/p0cM2f6hTHAzo1pgLtGdWdsTjjQrKGOOySRDdeIy2Zb1pyYBu3cmwGcLMSnNxSoR8pkqp+NOgZ2PCyfaptYb2OCnqY902sdB2lgWA+869MLD8w7OnYFdKgB9OnhJY/8V5F9iU8t83++a4CYH1b008y7bc8J55gXtrLpeLrAj3nvIyG4PF5AEDyYxQ7rbxhYHlG8aeETEA3Tv1vMDylSNGke50hrTXKULP7CymBCXhvifCgyh5WdkhAf76MfY5MSf16x+4F9rRLhw8hJyMDJxBfSv4H1K67LQRMTmnUqr9NOjZeOmaBYzomReybVBud5Zde0Objvfa/OubXBXcNuFMLhve+Muxe2YWz195NV1djVeSXdPTWXTlPLoFPQ04e3gBN58xIeRYGU4nbywIrdsdZ57FDaePwxEUWkbn92LpvOtCyq279dtkOELrluNyserWO0K2vXfjLeSEXeXOGj6C701uTK6d5nDw2oKvc1LQ1VWmM42HL57JaUHfz9zMTF6cO5/RvXqT5nCQ5nBw9oCBvDh3Ps6goDnztBH86OwpIYHllNzuLL+u8SEWgJ+fdwGzh48MCaKFffvzzJyriZV0p5MlVy+gsF//QBvGndyHl66eT5coRwOUUieOPr3ZDHddHduOlDP0pJ5RD2vaOVBVyVG32zaZc7CDVVU4BPJbGH7ccvgQPbKyOLlrTsQy9fX17K04Rn52F7KbaUNZdTWf7d/L2f0GkJsdeVixvPo4JZWVFOTl42zmCcUj1dW4vV76devWbBsqa2txOhxku1wRy9TX17O/soLumZnkZER+HcBbX8/+imP07ppjO3waK8fr6jBEP/StlOo4mnBaKaVUytBXFpRSSqkwGvSUUkqlDA16SimlUkbK5d50ezy8snUz7+/+kr45OVw/ZiynntQz3tWKSp3Px5vbi1mxczvdMzNZMGYsY3r1blKutLKS+9/7F2sOlNKnSw4/mXouk/oPjHn9PivZx4ubPsft8TLztOHMGDI05KlMpZSKl5R6kKWytpYrFj/Pgaoq3F4PThFcTid/ungm04cM7eCaxkadz8e1Ly9ma3kZ1R4PDhEynE7unXo+140ZGyhXXF7GpX9/hvqw/v3Z1PP5ZtgrDx3p4VUf85eiVdR4vRgg2+WisF9/nrx8TsT3FZVSqr30QRYbC9etpqSyArfXn7bKZww1Xi8/XvkW3hjlaOxor27bwpayw4HUW/XG4PZ6+fX771JZWxsod/ey15sEPIDf/ee9mOWjPFBVySOffYrbCnjgT8u1av8+3v0y+rylSinV0VIq6C3bXmybL9Hj87G9vCwONYrem9uLcdsksHY5HRSVNubo3HH0iO3nfcawPigPZkf6cO8e0myGMas9HlbsjJyDVCmlTpSUCnpdI+St9BmTNFk0cjMybNOVYaBLUDYXZzNDiSfFKB9lF1c6YlM7p0hInlGllIqXlAp6N449g6ywSUodIgzp3oOBQQmRE9mCMWNts41ku1xMCMpHGekeZW5GRiDhdEc7f9AgHDax1uV0MrdgdEzOqZRS0UipoHfZsOHMGzWGdKeTLq50urhc9M/pxuOXXRHvqrVaYb/+3F04mQynk66udLq40umZlc1TV1wV8oTkQzMuCZmxAPz5IhddNS9mdctMc7Fw9pV0y8iga3o6XV3pZDjTuO/cC0NybyqlVLyk1NObDUorK1l7oJT8LtlM7NMvMFN3MimvrmZVyT66ZWRwVr8BtvfSAD7Zt4eVu3ZyWs885o4cheMEvDpQ5/Px8d491Pi8TO4/UIc2lVIxp7k3lVJKpQx9ZUEppZQKo0FPKaVUytCgp5RSKmXENOiJyMUisk1EdojIPTb7M0RksbX/UxEZFMv6KKWUSm0xC3oi4gQeAS4BCoAFIlIQVuxm4KgxZijwEPBArOqjlFJKxfJKrxDYYYzZZYypA14AZoeVmQ08bS0vAaZJMr4/oJRSKinEMuj1A/YGre+zttmWMcZ4gWNAcs3zo5RSKmkkxYMsInKriBSJSNHhw4fjXR2llFJJKpZBbz8wIGi9v7XNtoyIpAG5QHn4gYwxTxhjJhpjJubn58eoukoppTq7WAa9z4BhIjJYRNKB+cCrYWVeBW60lucC75hkSxGjlFIqaTRN199BjDFeEbkTeAtwAguNMZtE5FdAkTHmVeD/gGdFZAdwBH9gVEoppWIi6XJvishhYHe862EjD0iOmWgj6wxtgM7RDm1DYtA2JIbWtOEUY0yL97+SLuglKhEpak2y00TWGdoAnaMd2obEoG1IDB3ZhqR4elMppZTqCBr0lFJKpQwNeh3niXhXoAN0hjZA52iHtiExaBsSQ4e1Qe/pKaWUShl6paeUUiplaNBrAxFxishaEXndZt9NInJYRNZZX7fEo47NEZEvReRzq35FNvtFRP5kTfm0QUTGx6OezWlFG84XkWNB/fCLeNSzOSLSXUSWiMhWEdkiIpPD9id8P0Cr2pHQfSEiw4Pqtk5EKkTku2FlErovWtmGhO4HABH5nohsEpGNIrJIRDLD9rd7OrqYvZzeyX0H2AJ0i7B/sTHmzhNYn7a4wBgT6b2XS4Bh1tdZwGPWv4mmuTYAfGCMueyE1SZ6fwSWG2PmWlmLssP2J0s/tNQOSOC+MMZsA8ZBYEq0/cDSsGIJ3RetbAMkcD+ISD/gbqDAGOMWkRfxJyz5W1CxwHR0IjIf/3R086I5j17pRUlE+gMzgSfjXZcYmg08Y/w+AbqLSJ94V6ozEZFc4Fz8WYkwxtQZY74KK5bw/dDKdiSTacBOY0x4AoyE74sgkdqQDNKALCsXczZQEra/3dPRadCL3h+AHwP1zZS5yhoCWSIiA5opFy8GWCEiq0XkVpv9rZkWKt5aagPAZBFZLyLLRGTUiaxcKwwGDgNPWUPlT4pIl7AyydAPrWkHJHZfBJsPLLLZngx90SBSGyCB+8EYsx/4H2APUAocM8asCCvW7unoNOhFQUQuAw4ZY1Y3U+w1YJAx5nTgbRr/KkkkU4wx4/EP2XxbRM6Nd4XaoKU2rMGflmgs8DDwjxNdwRakAeOBx4wxZwDHgXviW6U2aU07Er0vALCGZmcBL8W7Lm3VQhsSuh9EpAf+K7nBQF+gi4hc39Hn0aAXnXOAWSLyJf6Z4C8UkeeCCxhjyo0xtdbqk8CEE1vFlll/UWGMOYR/3L8wrEhrpoWKq5baYIypMMZUWctvAi4RyTvhFY1sH7DPGPOptb4Ef/AIlvD9QCvakQR90eASYI0x5qDNvmToC2imDUnQD9OBL4wxh40xHuAV4OywMq2ajq45GvSiYIz5iTGmvzFmEP4hhHeMMSF/iYSN88/C/8BLwhCRLiKS07AMzAA2hhV7FbjBemJtEv5hhtITXNWIWtMGETm5YaxfRArx/6xH9Z8jlowxB4C9IjLc2jQN2BxWLKH7AVrXjkTviyALiDwsmPB9YYnYhiTohz3AJBHJtuo5jaa/P9s9HZ0+vdkBJHS6pLtFZBbgxT9d0k3xrJuN3sBS62c/Dfi7MWa5iNwOYIz5C/AmcCmwA6gGvhGnukbSmjbMBb4lIl7ADcxPwLka7wKet4akdgHfSLJ+aNBSOxK+L6w/nr4G3Ba0Lan6ohVtSOh+MMZ8KiJL8A/DeoG1wBPSwdPRaUYWpZRSKUOHN5VSSqUMDXpKKaVShgY9pZRSKUODnlJKqZShQU8ppVTK0KCnVDuIyE+trPAbrMz1HZqEWPyZ8e1m87Dd3sHnvjdoeZCIhL/PqVTS0aCnVBuJfwqdy4DxVtq56YTmZ0x297ZcRKnkokFPqbbrA5Q1pJ0zxpQZY0oARGSCiLxnJcR+qyFTj4i8KyJ/tK4KN1qZMRCRQhH52Era/FFQhpOotHDeB0RklYgUi8hUa3u2iLwoIptFZKn45yibKCL/jT/b/ToRed46vFNE/mpd2a4Qkax2ffeUigMNekq13QpggBVEHhWR8wBExIU/oe9cY8wEYCHwm6DPZRtjxgF3WPsAtgJTraTNvwB+G21lWnHeNGNMIfBd4D5r2x345ycrAH6OlSvWGHMP4DbGjDPGXGeVHQY8YowZBXwFXBVtHZWKN01DplQbGWOqRGQCMBW4AFgsIvcARcBo4G0rVZoT/1QpDRZZn39fRLqJSHcgB3haRIbhnzbJ1YYqDW/hvK9Y/64GBlnLU/BPAosxZqOIbGjm+F8YY9bZHEOppKFBT6l2MMb4gHeBd0Xkc/zJcFcDm4wxkyN9zGb9v4B/G2PmiMgg65jRkhbO2zD7h4+2/d+vDVr2ATq8qZKODm8q1UYiMty6MmswDtgNbAPyrQddEBGXhE7YOc/aPgV/tv5j+KdIaZiq5qY2Vqml89r5ELjGKl8AjAna57GGTJXqNDToKdV2XfEPSW62hgULgF8aY+rwZ7R/QETWA+sInResRkTWAn8Bbra2/R74nbW9tVdh00RkX8MX/vtxzZ3XzqP4A+Vm4NfAJvyzUQM8AWwIepBFqaSnsywodQKJyLvAD40xRfGuC4CIOAGXMaZGRE4FVgLDrcCtVKej9/SUSm3ZwL+tYUwB7tCApzozvdJTSimVMvSenlJKqZShQU8ppVTK0KCnlFIqZWjQU0oplTI06CmllEoZGvSUUkqljP8HDmCrJOMYo5kAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 504x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(7,6))\n",
    "plt.scatter(X[:,0],X[:,1],c=y_)\n",
    "plt.xlabel('Sepal Length')\n",
    "plt.ylabel('Petal Width')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random Sampling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_ = np.insert(X, axis=1,obj=len(X[0]), values=y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[5.1, 0.2, 0. ],\n",
       "       [4.9, 0.2, 0. ],\n",
       "       [4.7, 0.2, 0. ],\n",
       "       [4.6, 0.2, 0. ],\n",
       "       [5. , 0.2, 0. ],\n",
       "       [5.4, 0.4, 0. ],\n",
       "       [4.6, 0.3, 0. ],\n",
       "       [5. , 0.2, 0. ],\n",
       "       [4.4, 0.2, 0. ],\n",
       "       [4.9, 0.1, 0. ]])"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "def k_means_objective(X, centres):\n",
    "    obj = 0\n",
    "    for point in X:\n",
    "        if point not in centres:\n",
    "            distance = [sum((point-i)**2) for i in centres]\n",
    "            obj += min(distance)\n",
    "    return obj"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "def random_sampling(X,y, n_sampleratio=0.7, seed=0):\n",
    "    data_ = np.insert(X, axis=1,obj=len(X[0]), values=y)\n",
    "    np.random.seed(seed)\n",
    "    np.random.shuffle(data_)\n",
    "    sample_data = data_[0:int(len(data_)*n_sampleratio)]\n",
    "    return sample_data[:,:-1], sample_data[:,-1]\n",
    "def random_sampling1(X,y, n=0.7, seed=0):\n",
    "    data_ = np.insert(X, axis=1,obj=len(X[0]), values=y)\n",
    "    np.random.seed(seed)\n",
    "    sampling_idx = np.random.randint(0,len(X),size=int(n*len(X)))\n",
    "    new_data = []\n",
    "    for i in sampling_idx:\n",
    "        new_data.append(data_[i])\n",
    "    new_data=np.array(new_data)\n",
    "    return new_data[:,:-1], new_data[:,-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_,y_ = random_sampling(X,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "X__,y__ = random_sampling1(X,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(105, 2)\n",
      "(105, 2)\n",
      "(105,)\n",
      "(105,)\n"
     ]
    }
   ],
   "source": [
    "for i in [X_,X__,y_,y__]:\n",
    "    print(i.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KMeans(algorithm='auto', copy_x=True, init='k-means++', max_iter=300,\n",
       "    n_clusters=3, n_init=10, n_jobs=None, precompute_distances='auto',\n",
       "    random_state=None, tol=0.0001, verbose=0)"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "km1 = KMeans(n_clusters=3)\n",
    "km1.fit(X_,y_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [],
   "source": [
    "random_sample_centres = km1.cluster_centers_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [],
   "source": [
    "score = k_means_objective(X,random_sample_centres)/km_obj"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.010312591642149"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KMeans(algorithm='auto', copy_x=True, init='k-means++', max_iter=300,\n",
       "    n_clusters=3, n_init=10, n_jobs=None, precompute_distances='auto',\n",
       "    random_state=None, tol=0.0001, verbose=0)"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "km1.fit(X__,y__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0224544472913304"
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "k_means_objective(X, km1.cluster_centers_)/km_obj"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}