online_kmeans.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pylab as plt\n",
    "from sklearn.decomposition import PCA\n",
    "from IPython import display\n",
    "import matplotlib.cm as cm\n",
    "from matplotlib.pyplot import savefig\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualizing the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Day</th>\n",
       "      <th>Weigth</th>\n",
       "      <th>Steps</th>\n",
       "      <th>Walk (min)</th>\n",
       "      <th>Swimming (min)</th>\n",
       "      <th>other sports (min)</th>\n",
       "      <th>Spent kcal</th>\n",
       "      <th>GPS country</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>20/05/2015</th>\n",
       "      <td>W</td>\n",
       "      <td>68,4</td>\n",
       "      <td>6620.0</td>\n",
       "      <td>75</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>380</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21/05/2015</th>\n",
       "      <td>Th</td>\n",
       "      <td>68,3</td>\n",
       "      <td>8111.0</td>\n",
       "      <td>70</td>\n",
       "      <td>0</td>\n",
       "      <td>11</td>\n",
       "      <td>452</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22/05/2015</th>\n",
       "      <td>F</td>\n",
       "      <td>68,7</td>\n",
       "      <td>9283.0</td>\n",
       "      <td>85</td>\n",
       "      <td>40</td>\n",
       "      <td>0</td>\n",
       "      <td>686</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23/05/2015</th>\n",
       "      <td>Sa</td>\n",
       "      <td>68,6</td>\n",
       "      <td>5995.0</td>\n",
       "      <td>53</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>237</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24/05/2015</th>\n",
       "      <td>Su</td>\n",
       "      <td>68,4</td>\n",
       "      <td>1419.0</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>71</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Day Weigth   Steps  Walk (min)  Swimming (min)  other sports (min)  \\\n",
       "Date                                                                            \n",
       "20/05/2015   W   68,4  6620.0          75               0                   6   \n",
       "21/05/2015  Th   68,3  8111.0          70               0                  11   \n",
       "22/05/2015   F   68,7  9283.0          85              40                   0   \n",
       "23/05/2015  Sa   68,6  5995.0          53               0                   0   \n",
       "24/05/2015  Su   68,4  1419.0          12               0                   0   \n",
       "\n",
       "            Spent kcal  GPS country  \n",
       "Date                                 \n",
       "20/05/2015         380           33  \n",
       "21/05/2015         452           33  \n",
       "22/05/2015         686           33  \n",
       "23/05/2015         237           33  \n",
       "24/05/2015          71           33  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('shealth.csv', sep=';')\n",
    "df = df.set_index('Date')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Day</th>\n",
       "      <th>Weigth</th>\n",
       "      <th>Steps</th>\n",
       "      <th>Walk (min)</th>\n",
       "      <th>Swimming (min)</th>\n",
       "      <th>other sports (min)</th>\n",
       "      <th>Spent kcal</th>\n",
       "      <th>GPS country</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>20/05/2015</th>\n",
       "      <td>W</td>\n",
       "      <td>68.4</td>\n",
       "      <td>6620.0</td>\n",
       "      <td>75</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>380</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21/05/2015</th>\n",
       "      <td>Th</td>\n",
       "      <td>68.3</td>\n",
       "      <td>8111.0</td>\n",
       "      <td>70</td>\n",
       "      <td>0</td>\n",
       "      <td>11</td>\n",
       "      <td>452</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22/05/2015</th>\n",
       "      <td>F</td>\n",
       "      <td>68.7</td>\n",
       "      <td>9283.0</td>\n",
       "      <td>85</td>\n",
       "      <td>40</td>\n",
       "      <td>0</td>\n",
       "      <td>686</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23/05/2015</th>\n",
       "      <td>Sa</td>\n",
       "      <td>68.6</td>\n",
       "      <td>5995.0</td>\n",
       "      <td>53</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>237</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24/05/2015</th>\n",
       "      <td>Su</td>\n",
       "      <td>68.4</td>\n",
       "      <td>1419.0</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>71</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Day  Weigth   Steps  Walk (min)  Swimming (min)  \\\n",
       "Date                                                         \n",
       "20/05/2015   W    68.4  6620.0          75               0   \n",
       "21/05/2015  Th    68.3  8111.0          70               0   \n",
       "22/05/2015   F    68.7  9283.0          85              40   \n",
       "23/05/2015  Sa    68.6  5995.0          53               0   \n",
       "24/05/2015  Su    68.4  1419.0          12               0   \n",
       "\n",
       "            other sports (min)  Spent kcal  GPS country  \n",
       "Date                                                     \n",
       "20/05/2015                   6         380           33  \n",
       "21/05/2015                  11         452           33  \n",
       "22/05/2015                   0         686           33  \n",
       "23/05/2015                   0         237           33  \n",
       "24/05/2015                   0          71           33  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.fillna(np.mean(df.Steps), inplace=True)\n",
    "df['Weigth'] = df.Weigth.apply(lambda s: s.replace(',', '.')).astype(float)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Weigth</th>\n",
       "      <th>Steps</th>\n",
       "      <th>Walk (min)</th>\n",
       "      <th>Swimming (min)</th>\n",
       "      <th>other sports (min)</th>\n",
       "      <th>Spent kcal</th>\n",
       "      <th>GPS country</th>\n",
       "      <th>F</th>\n",
       "      <th>M</th>\n",
       "      <th>Sa</th>\n",
       "      <th>Su</th>\n",
       "      <th>Th</th>\n",
       "      <th>Tu</th>\n",
       "      <th>W</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>20/05/2015</th>\n",
       "      <td>68.4</td>\n",
       "      <td>6620.0</td>\n",
       "      <td>75</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>380</td>\n",
       "      <td>33</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21/05/2015</th>\n",
       "      <td>68.3</td>\n",
       "      <td>8111.0</td>\n",
       "      <td>70</td>\n",
       "      <td>0</td>\n",
       "      <td>11</td>\n",
       "      <td>452</td>\n",
       "      <td>33</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22/05/2015</th>\n",
       "      <td>68.7</td>\n",
       "      <td>9283.0</td>\n",
       "      <td>85</td>\n",
       "      <td>40</td>\n",
       "      <td>0</td>\n",
       "      <td>686</td>\n",
       "      <td>33</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23/05/2015</th>\n",
       "      <td>68.6</td>\n",
       "      <td>5995.0</td>\n",
       "      <td>53</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>237</td>\n",
       "      <td>33</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24/05/2015</th>\n",
       "      <td>68.4</td>\n",
       "      <td>1419.0</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>71</td>\n",
       "      <td>33</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            Weigth   Steps  Walk (min)  Swimming (min)  other sports (min)  \\\n",
       "Date                                                                         \n",
       "20/05/2015    68.4  6620.0          75               0                   6   \n",
       "21/05/2015    68.3  8111.0          70               0                  11   \n",
       "22/05/2015    68.7  9283.0          85              40                   0   \n",
       "23/05/2015    68.6  5995.0          53               0                   0   \n",
       "24/05/2015    68.4  1419.0          12               0                   0   \n",
       "\n",
       "            Spent kcal  GPS country  F  M  Sa  Su  Th  Tu  W  \n",
       "Date                                                          \n",
       "20/05/2015         380           33  0  0   0   0   0   0  1  \n",
       "21/05/2015         452           33  0  0   0   0   1   0  0  \n",
       "22/05/2015         686           33  1  0   0   0   0   0  0  \n",
       "23/05/2015         237           33  0  0   1   0   0   0  0  \n",
       "24/05/2015          71           33  0  0   0   1   0   0  0  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# classic one hot encoding\n",
    "X_ohe = pd.concat([df.drop(['Day'], axis=1), pd.get_dummies(df['Day'])], axis=1)\n",
    "X_ohe.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Day</th>\n",
       "      <th>Weigth</th>\n",
       "      <th>Steps</th>\n",
       "      <th>Walk (min)</th>\n",
       "      <th>Swimming (min)</th>\n",
       "      <th>other sports (min)</th>\n",
       "      <th>Spent kcal</th>\n",
       "      <th>GPS country</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>20/05/2015</th>\n",
       "      <td>0</td>\n",
       "      <td>68.4</td>\n",
       "      <td>6620.0</td>\n",
       "      <td>75</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>380</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21/05/2015</th>\n",
       "      <td>0</td>\n",
       "      <td>68.3</td>\n",
       "      <td>8111.0</td>\n",
       "      <td>70</td>\n",
       "      <td>0</td>\n",
       "      <td>11</td>\n",
       "      <td>452</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22/05/2015</th>\n",
       "      <td>0</td>\n",
       "      <td>68.7</td>\n",
       "      <td>9283.0</td>\n",
       "      <td>85</td>\n",
       "      <td>40</td>\n",
       "      <td>0</td>\n",
       "      <td>686</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23/05/2015</th>\n",
       "      <td>1</td>\n",
       "      <td>68.6</td>\n",
       "      <td>5995.0</td>\n",
       "      <td>53</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>237</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24/05/2015</th>\n",
       "      <td>1</td>\n",
       "      <td>68.4</td>\n",
       "      <td>1419.0</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>71</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            Day  Weigth   Steps  Walk (min)  Swimming (min)  \\\n",
       "Date                                                          \n",
       "20/05/2015    0    68.4  6620.0          75               0   \n",
       "21/05/2015    0    68.3  8111.0          70               0   \n",
       "22/05/2015    0    68.7  9283.0          85              40   \n",
       "23/05/2015    1    68.6  5995.0          53               0   \n",
       "24/05/2015    1    68.4  1419.0          12               0   \n",
       "\n",
       "            other sports (min)  Spent kcal  GPS country  \n",
       "Date                                                     \n",
       "20/05/2015                   6         380           33  \n",
       "21/05/2015                  11         452           33  \n",
       "22/05/2015                   0         686           33  \n",
       "23/05/2015                   0         237           33  \n",
       "24/05/2015                   0          71           33  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# weekend encoding\n",
    "X_weekend = df\n",
    "X_weekend['Day'] = ((df['Day']=='Su') | (df['Day'] == 'Sa')).astype(int)\n",
    "X_weekend.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Parser function for the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def parse_dataset(df, enc='weekend'):\n",
    "    \"\"\"\n",
    "    ohe: normal one hot enconding for the 'Day' variable\n",
    "    weekend: 1 if 'Day' is Sunday or Saturday, 0 otherwise\n",
    "    \"\"\"\n",
    "\n",
    "    if enc not in ['ohe', 'weekend']:\n",
    "        print(\"enc parameter should be either 'ohe' or 'weekend'\")\n",
    "\n",
    "    df = pd.read_csv('shealth.csv', sep=';')\n",
    "    df = df.set_index('Date')\n",
    "\n",
    "    df = df.fillna(np.mean(df.Steps))\n",
    "    df['Weigth'] = df.Weigth.apply(lambda s: s.replace(',', '.')).astype(float)\n",
    "\n",
    "    if enc == 'ohe':\n",
    "        X = pd.concat([df.drop(['Day'], axis=1),\n",
    "                       pd.get_dummies(df['Day'])], axis=1)\n",
    "    else:\n",
    "        X = df\n",
    "        X['Day'] = ((df['Day'] == 'Su') | (df['Day'] == 'Sa')).astype(int)\n",
    "    return X"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### KMeans algorithm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def KMeans(X, K, maxIters=10):\n",
    "    \"\"\"\"\n",
    "    X: data, numpy array\n",
    "    K: number of cluster, integer\n",
    "    maxIters: number maximum of iterations, integer\n",
    "    \"\"\"\n",
    "    centroids = X[np.random.choice(np.arange(len(X)), K), :]\n",
    "    for i in range(maxIters):\n",
    "        # Cluster Assignment step\n",
    "        C = np.array([np.argmin([np.dot(x_i - y_k, x_i - y_k)\n",
    "                                 for y_k in centroids]) for x_i in X])\n",
    "        # Move centroids step\n",
    "        centroids = [X[C == k].mean(axis=0) for k in range(K)]\n",
    "    return np.array(centroids), C"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x26c2873bc18>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAD8CAYAAAB6paOMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n3P8DP88lonPsqJgAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x26c286eea58>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# test KMeans algorithm\n",
    "X = parse_dataset(df, enc='weekend').values\n",
    "\n",
    "centroids, C = KMeans(X, 3, 100)\n",
    "\n",
    "pca = PCA(2)\n",
    "X_pca = pca.fit_transform(X)\n",
    "X_pca.shape\n",
    "\n",
    "plt.scatter(X_pca[:,0], X_pca[:, 1], c=C)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Online KMeans algorithm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "class OnlineKMeans:\n",
    "\n",
    "    def __init__(self, X, K=2, window_size=np.inf, plot_progress=False, add_remove=False,\n",
    "                 threshold_create=np.inf, threshold_remove=np.inf, cold_start=False):\n",
    "        self.X = X\n",
    "        self.K = K\n",
    "        self.window_size = window_size\n",
    "        self.plot_progress = plot_progress\n",
    "        self._nsamples = X.shape[0]\n",
    "        self._dim = X.shape[1]\n",
    "        self._eps = 1e-3\n",
    "        self.threshold_create = threshold_create\n",
    "        self.threshold_remove = threshold_remove\n",
    "        self.add_remove = add_remove\n",
    "        self.cold_start = cold_start\n",
    "        plt.figure(figsize=(15, 4))\n",
    "\n",
    "        if self._nsamples < K:\n",
    "            print(\"Error\")\n",
    "\n",
    "    def _plot_progress(self, c, centroids):\n",
    "        pca = PCA(2)\n",
    "        pca.fit(self._X_norm)\n",
    "        X_pca = pca.transform(self._X_norm)\n",
    "\n",
    "        centroids_ = [centroids[i]\n",
    "                      for i in range(len(centroids)) if i not in self._infidx]\n",
    "        centroids_pca = pca.transform(centroids_)\n",
    "\n",
    "        plt.clf()\n",
    "\n",
    "        plt.subplot(121)\n",
    "        plt.title('Online clustering progress')\n",
    "\n",
    "        colors = cm.rainbow([0.2 * i for i in range(self.K)]).tolist()\n",
    "        map_colors = [colors[k] for k in c]\n",
    "        plt.scatter(X_pca[:, 0], X_pca[:, 1], c=map_colors)\n",
    "        plt.scatter(centroids_pca[:, 0],\n",
    "                    centroids_pca[:, 1], color='black', marker='x')\n",
    "\n",
    "        plt.subplot(122)\n",
    "        plt.title('Average distance from the centroid')\n",
    "        plt.ylim(ymax=12)\n",
    "\n",
    "        max_dim = np.max([len(self._mean_dist[i])\n",
    "                          for i in range(len(self._mean_dist))])\n",
    "\n",
    "        for i in range(len(self._mean_dist)):\n",
    "            self._mean_dist[i].extend(\n",
    "                [0] * (max_dim - len(self._mean_dist[i])))\n",
    "\n",
    "        for k in range(self.K):\n",
    "            plt.plot(np.array(self._mean_dist)[\n",
    "                     :, k], color=colors[k], label='cluster ' + str(k))\n",
    "\n",
    "        plt.legend()\n",
    "\n",
    "        display.display(plt.gcf())\n",
    "        display.clear_output(wait=True)\n",
    "        plt.savefig('imagen.pdf')\n",
    "\n",
    "    def data_stream(self):\n",
    "        \"\"\"\n",
    "        Simulate data streaming\n",
    "        \"\"\"\n",
    "        yield from self.X\n",
    "\n",
    "    def distance(self, x, y, method='euclidean'):\n",
    "        if method is 'euclidean':\n",
    "            return np.sqrt(np.dot(x - y, x - y))\n",
    "        else:\n",
    "            print('%s distance unknown' % method)\n",
    "\n",
    "    def update_centroids(self, gamma, data_point, centroid):\n",
    "        return (centroid + gamma * (data_point - centroid))\n",
    "\n",
    "    def add_cluster(self):\n",
    "        self.K += 1\n",
    "\n",
    "    def run(self, max_iter=100):\n",
    "        \n",
    "        stream = self.data_stream()\n",
    "        C = list(range(self.K))\n",
    "        \n",
    "        if self.cold_start:\n",
    "            centroids = np.random.multivariate_normal(np.zeros(self._dim), np.eye(self._dim), self.K)            \n",
    "            init = 0\n",
    "            \n",
    "        else:\n",
    "            centroids = np.array([next(stream) for _ in range(self.K)])\n",
    "            init = self.K\n",
    "        \n",
    "        \n",
    "        # To normalize data on the fly\n",
    "        mean = np.mean(centroids, axis=0)\n",
    "        sigma = np.std(centroids, axis=0)\n",
    "        centroids = (centroids - mean) / (sigma + self._eps)\n",
    "\n",
    "        # auxiliary vectors only for plotting purposes\n",
    "        self._X_norm = list(centroids)\n",
    "        self._mean_dist = [[0] * self.K]\n",
    "        self._infidx = []\n",
    "\n",
    "        for t in range(init, min(self._nsamples, max_iter)):\n",
    "            p = next(stream)\n",
    "\n",
    "            mean_new = (t * mean + p) / (t + 1)\n",
    "            sigma_new = np.sqrt(\n",
    "                (sigma**2 * t + (p - mean) * (p - mean_new)) / (t + 1))\n",
    "\n",
    "            normalized_p = (p - mean_new) / (sigma_new + self._eps)\n",
    "            centroids = (centroids * sigma + mean - mean_new) / (sigma_new + self._eps)\n",
    "\n",
    "            if self.add_remove:\n",
    "                idx = np.argmin([self.distance(normalized_p, c)\n",
    "                                 for c in centroids])\n",
    "\n",
    "                # Condition to add a new cluster\n",
    "                if self.distance(normalized_p, centroids[idx]) > self.threshold_create:\n",
    "                    C.append(self.K)\n",
    "                    self.add_cluster()\n",
    "                    centroids = np.append(centroids, [normalized_p], axis=0)\n",
    "\n",
    "                # Condition to remove cluster\n",
    "                else:\n",
    "                    while t - np.max(np.where(C == idx)) > self.threshold_remove:\n",
    "                        centroids[idx] = np.inf * np.ones(self._dim)\n",
    "                        self._infidx.append(idx)\n",
    "                        idx = np.argmin([self.distance(normalized_p, c)\n",
    "                                         for c in centroids])\n",
    "\n",
    "                    C.append(\n",
    "                        np.argmin([self.distance(normalized_p, c) for c in centroids]))\n",
    "\n",
    "                    if self.window_size < t:\n",
    "                        centroids[C[-1]] = self.update_centroids(1 / self.window_size,\n",
    "                                                                 normalized_p, centroids[C[-1]])\n",
    "\n",
    "                    else:\n",
    "                        centroids[C[-1]] = self.update_centroids(1 / (t+1),\n",
    "                                                                 normalized_p, centroids[C[-1]])\n",
    "\n",
    "            else:\n",
    "                C.append(\n",
    "                    np.argmin([self.distance(normalized_p, c) for c in centroids]))\n",
    "\n",
    "                if self.window_size < t:\n",
    "                    centroids[C[-1]] = self.update_centroids(1 / self.window_size,\n",
    "                                                             normalized_p, centroids[C[-1]])\n",
    "                else:\n",
    "                    centroids[C[-1]] = self.update_centroids(1 / (t+1),\n",
    "                                                             normalized_p, centroids[C[-1]])\n",
    "\n",
    "            # Update mean and sigma\n",
    "            mean = mean_new\n",
    "            sigma = sigma_new\n",
    "                    \n",
    "            # We save the normalized points and mean distance only for plotting purposes\n",
    "            self._X_norm.append(normalized_p)\n",
    "            self._mean_dist.append([np.mean([self.distance(\n",
    "                p, centroids[k]) for p in self._X_norm[C == k]], axis=0) for k in range(self.K)])\n",
    "\n",
    "            if self.plot_progress:\n",
    "                self._plot_progress(C, centroids)\n",
    "\n",
    "        return np.array(centroids), C"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA2wAAAEICAYAAAAjlnc0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/Lgtix4bpw8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x26c2c290cf8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X = parse_dataset(df, enc='weekend').values\n",
    "\n",
    "online_kmeans = OnlineKMeans(X, K=3, plot_progress=True,\n",
    "                             add_remove=True, cold_start=True)\n",
    "centroids, c = online_kmeans.run(max_iter=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA2wAAAEICAYAAAAjlnc0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nER6Y2pIqppYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x26c2ab7f780>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "online_kmeans = OnlineKMeans(X, K=3, plot_progress=True,\n",
    "                             add_remove=True)\n",
    "centroids, c = online_kmeans.run(max_iter=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA2wAAAEICAYAAAAjlnc0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nzMzMcur/QIPyoTpq3U8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x26c2c28dc50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X = parse_dataset(df, enc='weekend').values\n",
    "\n",
    "online_kmeans = OnlineKMeans(X, K=3, plot_progress=True,\n",
    "                             add_remove=True, threshold_create=7, threshold_remove=10, cold_start=True)\n",
    "centroids, c = online_kmeans.run(max_iter=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA2wAAAEICAYAAAAjlnc0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSpQSNhERERERkRKlhE1ERERERKRE/T8fyHFr4AwVUQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x26c2c28d5c0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X = parse_dataset(df, enc='weekend').values\n",
    "\n",
    "online_kmeans = OnlineKMeans(X, K=3, plot_progress=True,\n",
    "                             add_remove=True, threshold_create=7, threshold_remove=10, cold_start=False)\n",
    "centroids, c = online_kmeans.run(max_iter=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}