household_segmentation.model_def2(1).ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h1 id=\"weather_station_clustering\" align=\"center\"> Clustering Ghanaian households based on their expenditure patterns</h1>\n",
    "<hr>\n",
    "\n",
    "#### 1.4 Model Definition - Part 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### SECOND PART  - DEEP LEARNING ALGORITHMS\n",
    "\n",
    "In this second part (using deep learning frameworks), we will consider the following algorithms and compare their results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "a. kmeans clustering algorithm\n",
    "\n",
    "b. autoencoder + kmeans\n",
    "\n",
    "c. DEC ( Deep Embedded Clustering) with autoencoder and kmeans"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Importing necessary libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Populating the interactive namespace from numpy and matplotlib\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/jupyterlab/conda/lib/python3.6/site-packages/IPython/core/magics/pylab.py:160: UserWarning: pylab import has clobbered these variables: ['imread', 'time', 'seed']\n",
      "`%matplotlib` prevents importing * from pylab and numpy\n",
      "  \"\\n`%matplotlib` prevents importing * from pylab and numpy\"\n"
     ]
    }
   ],
   "source": [
    "%pylab inline\n",
    "import os\n",
    "import keras\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import keras.backend as K\n",
    "\n",
    "from time import time\n",
    "\n",
    "from keras import callbacks\n",
    "from keras.models import Model\n",
    "from keras.optimizers import SGD\n",
    "from keras.layers import Dense, Input\n",
    "from keras.initializers import VarianceScaling\n",
    "from keras.engine.topology import Layer, InputSpec\n",
    "\n",
    "from scipy.misc import imread\n",
    "from sklearn.cluster import KMeans\n",
    "from sklearn.metrics import accuracy_score, normalized_mutual_info_score\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "#!pip install metrics\n",
    "import metrics"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Also. lets import the DEC algorithm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "fatal: destination path 'DEC-keras' already exists and is not an empty directory.\n"
     ]
    }
   ],
   "source": [
    "!git clone https://github.com/XifengGuo/DEC-keras\n",
    "!cd DEC-keras"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "# To stop potential randomness\n",
    "seed = 128\n",
    "rng = np.random.RandomState(seed)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "df = pd.read_csv('df_norm_dataframe.csv')\n",
    "df.drop(['Unnamed: 0'], axis = 1, inplace = True)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Renaming the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "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>totfood</th>\n",
       "      <th>totalch</th>\n",
       "      <th>totclth</th>\n",
       "      <th>tothous</th>\n",
       "      <th>totfurn</th>\n",
       "      <th>tothlth</th>\n",
       "      <th>tottrsp</th>\n",
       "      <th>totcmnq</th>\n",
       "      <th>totrcre</th>\n",
       "      <th>toteduc</th>\n",
       "      <th>totmisc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.869768</td>\n",
       "      <td>-0.282063</td>\n",
       "      <td>0.813326</td>\n",
       "      <td>0.597423</td>\n",
       "      <td>0.930850</td>\n",
       "      <td>0.270093</td>\n",
       "      <td>0.381865</td>\n",
       "      <td>-0.274590</td>\n",
       "      <td>-0.216681</td>\n",
       "      <td>-0.115215</td>\n",
       "      <td>0.970906</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.333146</td>\n",
       "      <td>-0.282063</td>\n",
       "      <td>-0.095515</td>\n",
       "      <td>3.937930</td>\n",
       "      <td>1.783318</td>\n",
       "      <td>-0.204459</td>\n",
       "      <td>-0.270768</td>\n",
       "      <td>0.137049</td>\n",
       "      <td>-0.224311</td>\n",
       "      <td>0.806321</td>\n",
       "      <td>-0.077166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.524523</td>\n",
       "      <td>-0.282063</td>\n",
       "      <td>-0.488894</td>\n",
       "      <td>0.058295</td>\n",
       "      <td>-0.199684</td>\n",
       "      <td>-0.182889</td>\n",
       "      <td>-0.331876</td>\n",
       "      <td>-0.261965</td>\n",
       "      <td>-0.265152</td>\n",
       "      <td>-0.282872</td>\n",
       "      <td>0.781011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.234851</td>\n",
       "      <td>-0.282063</td>\n",
       "      <td>1.200771</td>\n",
       "      <td>0.042361</td>\n",
       "      <td>0.104170</td>\n",
       "      <td>-0.182889</td>\n",
       "      <td>0.486971</td>\n",
       "      <td>0.487875</td>\n",
       "      <td>-0.278814</td>\n",
       "      <td>-0.187837</td>\n",
       "      <td>0.737953</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2.104382</td>\n",
       "      <td>-0.282063</td>\n",
       "      <td>0.356362</td>\n",
       "      <td>0.286199</td>\n",
       "      <td>1.399336</td>\n",
       "      <td>6.288280</td>\n",
       "      <td>-0.007448</td>\n",
       "      <td>0.594741</td>\n",
       "      <td>-0.195824</td>\n",
       "      <td>0.020165</td>\n",
       "      <td>0.877056</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    totfood   totalch   totclth   tothous   totfurn   tothlth   tottrsp  \\\n",
       "0  0.869768 -0.282063  0.813326  0.597423  0.930850  0.270093  0.381865   \n",
       "1  0.333146 -0.282063 -0.095515  3.937930  1.783318 -0.204459 -0.270768   \n",
       "2 -0.524523 -0.282063 -0.488894  0.058295 -0.199684 -0.182889 -0.331876   \n",
       "3  0.234851 -0.282063  1.200771  0.042361  0.104170 -0.182889  0.486971   \n",
       "4  2.104382 -0.282063  0.356362  0.286199  1.399336  6.288280 -0.007448   \n",
       "\n",
       "    totcmnq   totrcre   toteduc   totmisc  \n",
       "0 -0.274590 -0.216681 -0.115215  0.970906  \n",
       "1  0.137049 -0.224311  0.806321 -0.077166  \n",
       "2 -0.261965 -0.265152 -0.282872  0.781011  \n",
       "3  0.487875 -0.278814 -0.187837  0.737953  \n",
       "4  0.594741 -0.195824  0.020165  0.877056  "
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train = df\n",
    "train.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "a. kmeans clustering algorithm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will first apply K-Means directly to our data and divide it into 3 clusters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "km = KMeans(n_clusters=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train = train.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "kmeans = km.fit(x_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 2, 0, 0, 2, 2, 2, 2, 0], dtype=int32)"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "kmeans.labels_[0:10] #first 10 cluster labels"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "at best we can find the optimal number of clusters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Silhouette Coefficient: 0.378\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import silhouette_samples, silhouette_score\n",
    "from sklearn import metrics\n",
    "print(\"Silhouette Coefficient: %0.3f\"\n",
    "      % metrics.silhouette_score(x_train, kmeans.labels_))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This silhouette score is worse than the results under HDBSCAN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "####  b. kmeans + autoencoder"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "Now instead of directly applying K-Means on the problem, we will first use an autoencoder to decrease the dimensionality of the data and extract useful information. This will then pass on the information to the K-Means algorithm.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "# this is our input placeholder\n",
    "input_img = Input(shape=(11,)) # 11 FEATURES\n",
    "\n",
    "# \"encoded\" is the encoded representation of the input  - extracting relevant information\n",
    "encoded = Dense(500, activation='relu')(input_img)\n",
    "encoded = Dense(500, activation='relu')(encoded)\n",
    "encoded = Dense(2000, activation='relu')(encoded)\n",
    "encoded = Dense(11, activation='softmax')(encoded)\n",
    "\n",
    "# \"decoded\" is the lossy reconstruction of the input - decoding relevant information\n",
    "decoded = Dense(2000, activation='relu')(encoded)\n",
    "decoded = Dense(500, activation='relu')(decoded)\n",
    "decoded = Dense(500, activation='softmax')(decoded)\n",
    "decoded = Dense(11)(decoded) # 11 features\n",
    "\n",
    "# this model maps an input to its reconstruction\n",
    "autoencoder = Model(input_img, decoded)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_2 (InputLayer)         (None, 11)                0         \n",
      "_________________________________________________________________\n",
      "dense_9 (Dense)              (None, 500)               6000      \n",
      "_________________________________________________________________\n",
      "dense_10 (Dense)             (None, 500)               250500    \n",
      "_________________________________________________________________\n",
      "dense_11 (Dense)             (None, 2000)              1002000   \n",
      "_________________________________________________________________\n",
      "dense_12 (Dense)             (None, 11)                22011     \n",
      "_________________________________________________________________\n",
      "dense_13 (Dense)             (None, 2000)              24000     \n",
      "_________________________________________________________________\n",
      "dense_14 (Dense)             (None, 500)               1000500   \n",
      "_________________________________________________________________\n",
      "dense_15 (Dense)             (None, 500)               250500    \n",
      "_________________________________________________________________\n",
      "dense_16 (Dense)             (None, 11)                5511      \n",
      "=================================================================\n",
      "Total params: 2,561,022\n",
      "Trainable params: 2,561,022\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "autoencoder.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "#  this model maps an input to its encoded representation\n",
    "encoder = Model(input_img, encoded)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "autoencoder.compile(optimizer='adam', loss='mse')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now let us train our autoencoder model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 8922 samples, validate on 8922 samples\n",
      "Epoch 1/50\n",
      "8922/8922 [==============================] - 5s 616us/step - loss: 1.0001 - val_loss: 1.0000\n",
      "Epoch 2/50\n",
      "8922/8922 [==============================] - 5s 529us/step - loss: 1.0000 - val_loss: 1.0000\n",
      "Epoch 3/50\n",
      "8922/8922 [==============================] - 5s 525us/step - loss: 1.0000 - val_loss: 0.9999\n",
      "Epoch 4/50\n",
      "8922/8922 [==============================] - 5s 525us/step - loss: 0.9997 - val_loss: 0.9988\n",
      "Epoch 5/50\n",
      "8922/8922 [==============================] - 5s 525us/step - loss: 0.9962 - val_loss: 0.9889\n",
      "Epoch 6/50\n",
      "8922/8922 [==============================] - 5s 525us/step - loss: 0.9788 - val_loss: 0.9640\n",
      "Epoch 7/50\n",
      "8922/8922 [==============================] - 5s 526us/step - loss: 0.9555 - val_loss: 0.9461\n",
      "Epoch 8/50\n",
      "8922/8922 [==============================] - 5s 528us/step - loss: 0.9420 - val_loss: 0.9370\n",
      "Epoch 9/50\n",
      "8922/8922 [==============================] - 5s 526us/step - loss: 0.9338 - val_loss: 0.9296\n",
      "Epoch 10/50\n",
      "8922/8922 [==============================] - 5s 525us/step - loss: 0.9266 - val_loss: 0.9227\n",
      "Epoch 11/50\n",
      "8922/8922 [==============================] - 5s 525us/step - loss: 0.9203 - val_loss: 0.9176\n",
      "Epoch 12/50\n",
      "8922/8922 [==============================] - 5s 524us/step - loss: 0.9135 - val_loss: 0.9090\n",
      "Epoch 13/50\n",
      "8922/8922 [==============================] - 5s 530us/step - loss: 0.9064 - val_loss: 0.9032\n",
      "Epoch 14/50\n",
      "8922/8922 [==============================] - 5s 531us/step - loss: 0.9011 - val_loss: 0.8984\n",
      "Epoch 15/50\n",
      "8922/8922 [==============================] - 5s 525us/step - loss: 0.8967 - val_loss: 0.8952\n",
      "Epoch 16/50\n",
      "8922/8922 [==============================] - 5s 525us/step - loss: 0.8949 - val_loss: 0.8920\n",
      "Epoch 17/50\n",
      "8922/8922 [==============================] - 5s 526us/step - loss: 0.8891 - val_loss: 0.8851\n",
      "Epoch 18/50\n",
      "8922/8922 [==============================] - 5s 530us/step - loss: 0.8842 - val_loss: 0.8830\n",
      "Epoch 19/50\n",
      "8922/8922 [==============================] - 5s 530us/step - loss: 0.8817 - val_loss: 0.8790\n",
      "Epoch 20/50\n",
      "8922/8922 [==============================] - 5s 528us/step - loss: 0.8772 - val_loss: 0.8755\n",
      "Epoch 21/50\n",
      "8922/8922 [==============================] - 5s 523us/step - loss: 0.8744 - val_loss: 0.8728\n",
      "Epoch 22/50\n",
      "8922/8922 [==============================] - 5s 530us/step - loss: 0.8716 - val_loss: 0.8699\n",
      "Epoch 23/50\n",
      "8922/8922 [==============================] - 5s 525us/step - loss: 0.8687 - val_loss: 0.8669\n",
      "Epoch 24/50\n",
      "8922/8922 [==============================] - 5s 534us/step - loss: 0.8657 - val_loss: 0.8644\n",
      "Epoch 25/50\n",
      "8922/8922 [==============================] - 5s 530us/step - loss: 0.8634 - val_loss: 0.8622\n",
      "Epoch 26/50\n",
      "8922/8922 [==============================] - 5s 523us/step - loss: 0.8617 - val_loss: 0.8611\n",
      "Epoch 27/50\n",
      "8922/8922 [==============================] - 5s 530us/step - loss: 0.8599 - val_loss: 0.8584\n",
      "Epoch 28/50\n",
      "8922/8922 [==============================] - 5s 527us/step - loss: 0.8573 - val_loss: 0.8557\n",
      "Epoch 29/50\n",
      "8922/8922 [==============================] - 5s 523us/step - loss: 0.8547 - val_loss: 0.8534\n",
      "Epoch 30/50\n",
      "8922/8922 [==============================] - 5s 529us/step - loss: 0.8528 - val_loss: 0.8519\n",
      "Epoch 31/50\n",
      "8922/8922 [==============================] - 5s 526us/step - loss: 0.8519 - val_loss: 0.8501\n",
      "Epoch 32/50\n",
      "8922/8922 [==============================] - 5s 530us/step - loss: 0.8494 - val_loss: 0.8486\n",
      "Epoch 33/50\n",
      "8922/8922 [==============================] - 5s 525us/step - loss: 0.8479 - val_loss: 0.8467\n",
      "Epoch 34/50\n",
      "8922/8922 [==============================] - 5s 524us/step - loss: 0.8460 - val_loss: 0.8450\n",
      "Epoch 35/50\n",
      "8922/8922 [==============================] - 5s 529us/step - loss: 0.8445 - val_loss: 0.8446\n",
      "Epoch 36/50\n",
      "8922/8922 [==============================] - 5s 525us/step - loss: 0.8444 - val_loss: 0.8419\n",
      "Epoch 37/50\n",
      "8922/8922 [==============================] - 5s 528us/step - loss: 0.8413 - val_loss: 0.8402\n",
      "Epoch 38/50\n",
      "8922/8922 [==============================] - 5s 526us/step - loss: 0.8402 - val_loss: 0.8402\n",
      "Epoch 39/50\n",
      "8922/8922 [==============================] - 5s 526us/step - loss: 0.8398 - val_loss: 0.8392\n",
      "Epoch 40/50\n",
      "8922/8922 [==============================] - 5s 531us/step - loss: 0.8385 - val_loss: 0.8371\n",
      "Epoch 41/50\n",
      "8922/8922 [==============================] - 5s 530us/step - loss: 0.8362 - val_loss: 0.8353\n",
      "Epoch 42/50\n",
      "8922/8922 [==============================] - 5s 527us/step - loss: 0.8347 - val_loss: 0.8340\n",
      "Epoch 43/50\n",
      "8922/8922 [==============================] - 5s 531us/step - loss: 0.8339 - val_loss: 0.8337\n",
      "Epoch 44/50\n",
      "8922/8922 [==============================] - 5s 530us/step - loss: 0.8332 - val_loss: 0.8324\n",
      "Epoch 45/50\n",
      "8922/8922 [==============================] - 5s 531us/step - loss: 0.8316 - val_loss: 0.8307\n",
      "Epoch 46/50\n",
      "8922/8922 [==============================] - 5s 531us/step - loss: 0.8292 - val_loss: 0.8261\n",
      "Epoch 47/50\n",
      "8922/8922 [==============================] - 5s 525us/step - loss: 0.8255 - val_loss: 0.8277\n",
      "Epoch 48/50\n",
      "8922/8922 [==============================] - 5s 530us/step - loss: 0.9024 - val_loss: 0.9046\n",
      "Epoch 49/50\n",
      "8922/8922 [==============================] - 5s 531us/step - loss: 0.8990 - val_loss: 0.8989\n",
      "Epoch 50/50\n",
      "8922/8922 [==============================] - 5s 531us/step - loss: 0.8972 - val_loss: 0.8879\n"
     ]
    }
   ],
   "source": [
    "train_history = autoencoder.fit(x_train, x_train, epochs=50, batch_size=1048, validation_data=(x_train, x_train))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_auto_train = autoencoder.predict(x_train) # this is x' = output of auto encoder\n",
    "#pred_auto_train = encoder.predict(x_train) # this is x' = output of auto encoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8922, 11)"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred_auto_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "kmeans_dl = km.fit(pred_auto_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([2, 2, 0, 2, 2, 0, 0, 0, 0, 2], dtype=int32)"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "kmeans_dl.labels_[0:10] #first 10 cluster labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Silhouette Coefficient: 0.986\n"
     ]
    }
   ],
   "source": [
    "print(\"Silhouette Coefficient: %0.3f\"\n",
    "      % metrics.silhouette_score(pred_auto_train, kmeans_dl.labels_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "#import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "XX = pred_auto_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "XX_df = pd.DataFrame({'totfood':XX[:,0],'totalch':XX[:,1], 'totclth':XX[:,2],'tothous':XX[:,3],'totfurn':XX[:,4], 'tothlth':XX[:,5],\n",
    "                       'tottrsp':XX[:,6],'totcmnq':XX[:,7], 'totrcre':XX[:,8],'toteduc':XX[:,9],'totmisc':XX[:,10]})\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "XX_df.to_csv('XX_df.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Adding labels to the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "XX_dfl = pd.DataFrame({'totfood':XX[:,0],'totalch':XX[:,1], 'totclth':XX[:,2],'tothous':XX[:,3],'totfurn':XX[:,4], 'tothlth':XX[:,5],\n",
    "                       'tottrsp':XX[:,6],'totcmnq':XX[:,7], 'totrcre':XX[:,8],'toteduc':XX[:,9],'totmisc':XX[:,10], 'label':kmeans_dl.labels_})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Visualizing the clusters\n",
    "\n",
    "PCA Projection to 2D\n",
    "\n",
    "Now, let's try to visualize the cluster, although we have only one cluster. First, lets use PCA to reduce the dimension of the data to two dimensions\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.decomposition import PCA\n",
    "\n",
    "pca = PCA(n_components=2)\n",
    "\n",
    "principalComponents = pca.fit_transform(XX_dfl)\n",
    "\n",
    "principalDf_dbscan = pd.DataFrame(data = principalComponents\n",
    "             , columns = ['C1', 'C2'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "principalDf_dbscan['label'] = XX_dfl.label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "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>C1</th>\n",
       "      <th>C2</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.698424</td>\n",
       "      <td>-0.292963</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.698424</td>\n",
       "      <td>-0.292963</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         C1        C2  label\n",
       "0  1.698424 -0.292963      2\n",
       "1  1.698424 -0.292963      2"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "principalDf_dbscan.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "'c' argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with 'x' & 'y'.  Please use a 2-D array with a single row if you really want to specify the same RGB or RGBA value for all points.\n",
      "'c' argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with 'x' & 'y'.  Please use a 2-D array with a single row if you really want to specify the same RGB or RGBA value for all points.\n",
      "'c' argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with 'x' & 'y'.  Please use a 2-D array with a single row if you really want to specify the same RGB or RGBA value for all points.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'C2')"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.cm as cm\n",
    "from matplotlib import pyplot as plt \n",
    "n_clusters = 3\n",
    "colors = cm.rainbow(np.linspace(0, 1, n_clusters))\n",
    "cluster_labels = list(range(0, n_clusters))\n",
    "\n",
    "# Create a figure of size 6 inches by 4 inches.\n",
    "plt.figure(figsize=(10,10))\n",
    "pdf = principalDf_dbscan.copy()\n",
    "# plt.scatter(subset.C1, subset.C2)\n",
    "for color, label in zip(colors, cluster_labels):\n",
    "    subset = pdf[pdf.label == label] #subset is a dataframe including all the columns\n",
    "    plt.scatter(subset.C1, subset.C2, c=color, label='cluster'+str(label),alpha=0.5) # plt.scatter(subset.horsepow, subset.mpg)\n",
    "plt.legend()\n",
    "plt.title('Clusters')\n",
    "plt.xlabel('C1')\n",
    "plt.ylabel('C2')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "PARALLEL PLOT"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "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>totfood</th>\n",
       "      <th>totalch</th>\n",
       "      <th>totclth</th>\n",
       "      <th>tothous</th>\n",
       "      <th>totfurn</th>\n",
       "      <th>tothlth</th>\n",
       "      <th>tottrsp</th>\n",
       "      <th>totcmnq</th>\n",
       "      <th>totrcre</th>\n",
       "      <th>toteduc</th>\n",
       "      <th>totmisc</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.187262</td>\n",
       "      <td>0.154462</td>\n",
       "      <td>0.381787</td>\n",
       "      <td>0.160858</td>\n",
       "      <td>0.205355</td>\n",
       "      <td>0.123294</td>\n",
       "      <td>0.125713</td>\n",
       "      <td>0.208678</td>\n",
       "      <td>0.144971</td>\n",
       "      <td>0.284140</td>\n",
       "      <td>0.199687</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.187262</td>\n",
       "      <td>0.154462</td>\n",
       "      <td>0.381787</td>\n",
       "      <td>0.160858</td>\n",
       "      <td>0.205355</td>\n",
       "      <td>0.123294</td>\n",
       "      <td>0.125713</td>\n",
       "      <td>0.208678</td>\n",
       "      <td>0.144971</td>\n",
       "      <td>0.284140</td>\n",
       "      <td>0.199687</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.309691</td>\n",
       "      <td>-0.236916</td>\n",
       "      <td>-0.255200</td>\n",
       "      <td>-0.164006</td>\n",
       "      <td>-0.234674</td>\n",
       "      <td>-0.160486</td>\n",
       "      <td>-0.182042</td>\n",
       "      <td>-0.249690</td>\n",
       "      <td>-0.178124</td>\n",
       "      <td>-0.198115</td>\n",
       "      <td>-0.215703</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    totfood   totalch   totclth   tothous   totfurn   tothlth   tottrsp  \\\n",
       "0  0.187262  0.154462  0.381787  0.160858  0.205355  0.123294  0.125713   \n",
       "1  0.187262  0.154462  0.381787  0.160858  0.205355  0.123294  0.125713   \n",
       "2 -0.309691 -0.236916 -0.255200 -0.164006 -0.234674 -0.160486 -0.182042   \n",
       "\n",
       "    totcmnq   totrcre   toteduc   totmisc  label  \n",
       "0  0.208678  0.144971  0.284140  0.199687      2  \n",
       "1  0.208678  0.144971  0.284140  0.199687      2  \n",
       "2 -0.249690 -0.178124 -0.198115 -0.215703      0  "
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "XX_dfl.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "ddd = XX_dfl.label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "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>TOTFOOD</th>\n",
       "      <th>TOTALCH</th>\n",
       "      <th>TOTCLTH</th>\n",
       "      <th>TOTHOUS</th>\n",
       "      <th>TOTFURN</th>\n",
       "      <th>TOTHLTH</th>\n",
       "      <th>TOTTRSP</th>\n",
       "      <th>TOTCMNQ</th>\n",
       "      <th>TOTRCRE</th>\n",
       "      <th>TOTEDUC</th>\n",
       "      <th>TOTMISC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>9437.439453</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1978.0</td>\n",
       "      <td>2569.800049</td>\n",
       "      <td>1128.790039</td>\n",
       "      <td>182.5</td>\n",
       "      <td>2248.399902</td>\n",
       "      <td>138.899994</td>\n",
       "      <td>55.25</td>\n",
       "      <td>1481.5</td>\n",
       "      <td>1081.459961</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>6990.479980</td>\n",
       "      <td>0.0</td>\n",
       "      <td>906.0</td>\n",
       "      <td>10808.799805</td>\n",
       "      <td>1693.709961</td>\n",
       "      <td>21.9</td>\n",
       "      <td>192.500000</td>\n",
       "      <td>370.399994</td>\n",
       "      <td>50.00</td>\n",
       "      <td>5079.0</td>\n",
       "      <td>260.200012</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       TOTFOOD  TOTALCH  TOTCLTH       TOTHOUS      TOTFURN  TOTHLTH  \\\n",
       "0  9437.439453      0.0   1978.0   2569.800049  1128.790039    182.5   \n",
       "1  6990.479980      0.0    906.0  10808.799805  1693.709961     21.9   \n",
       "\n",
       "       TOTTRSP     TOTCMNQ  TOTRCRE  TOTEDUC      TOTMISC  \n",
       "0  2248.399902  138.899994    55.25   1481.5  1081.459961  \n",
       "1   192.500000  370.399994    50.00   5079.0   260.200012  "
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('final_capstone_project_data.csv') # original data\n",
    "df.drop(['Unnamed: 0','hid','region','rururb','TOTAL_EXP'], axis = 1, inplace = True)\n",
    "df.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "b1 = pd.merge(ddd, df, how='inner', left_index = True, right_index=True) # we are merging on the household id.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "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>label</th>\n",
       "      <th>TOTFOOD</th>\n",
       "      <th>TOTALCH</th>\n",
       "      <th>TOTCLTH</th>\n",
       "      <th>TOTHOUS</th>\n",
       "      <th>TOTFURN</th>\n",
       "      <th>TOTHLTH</th>\n",
       "      <th>TOTTRSP</th>\n",
       "      <th>TOTCMNQ</th>\n",
       "      <th>TOTRCRE</th>\n",
       "      <th>TOTEDUC</th>\n",
       "      <th>TOTMISC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>9437.439453</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1978.0</td>\n",
       "      <td>2569.800049</td>\n",
       "      <td>1128.790039</td>\n",
       "      <td>182.500000</td>\n",
       "      <td>2248.399902</td>\n",
       "      <td>138.899994</td>\n",
       "      <td>55.250000</td>\n",
       "      <td>1481.5</td>\n",
       "      <td>1081.459961</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>6990.479980</td>\n",
       "      <td>0.0</td>\n",
       "      <td>906.0</td>\n",
       "      <td>10808.799805</td>\n",
       "      <td>1693.709961</td>\n",
       "      <td>21.900000</td>\n",
       "      <td>192.500000</td>\n",
       "      <td>370.399994</td>\n",
       "      <td>50.000000</td>\n",
       "      <td>5079.0</td>\n",
       "      <td>260.200012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>3079.566895</td>\n",
       "      <td>0.0</td>\n",
       "      <td>442.0</td>\n",
       "      <td>1240.099976</td>\n",
       "      <td>379.599976</td>\n",
       "      <td>29.200001</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>146.000000</td>\n",
       "      <td>21.900002</td>\n",
       "      <td>827.0</td>\n",
       "      <td>932.660034</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>6542.259766</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2435.0</td>\n",
       "      <td>1200.800049</td>\n",
       "      <td>580.959961</td>\n",
       "      <td>29.200001</td>\n",
       "      <td>2579.500000</td>\n",
       "      <td>567.700012</td>\n",
       "      <td>12.500000</td>\n",
       "      <td>1198.0</td>\n",
       "      <td>898.919983</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   label      TOTFOOD  TOTALCH  TOTCLTH       TOTHOUS      TOTFURN  \\\n",
       "0      2  9437.439453      0.0   1978.0   2569.800049  1128.790039   \n",
       "1      2  6990.479980      0.0    906.0  10808.799805  1693.709961   \n",
       "2      0  3079.566895      0.0    442.0   1240.099976   379.599976   \n",
       "3      2  6542.259766      0.0   2435.0   1200.800049   580.959961   \n",
       "\n",
       "      TOTHLTH      TOTTRSP     TOTCMNQ    TOTRCRE  TOTEDUC      TOTMISC  \n",
       "0  182.500000  2248.399902  138.899994  55.250000   1481.5  1081.459961  \n",
       "1   21.900000   192.500000  370.399994  50.000000   5079.0   260.200012  \n",
       "2   29.200001     0.000000  146.000000  21.900002    827.0   932.660034  \n",
       "3   29.200001  2579.500000  567.700012  12.500000   1198.0   898.919983  "
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "b1.head(4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "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>TOTFOOD</th>\n",
       "      <th>TOTALCH</th>\n",
       "      <th>TOTCLTH</th>\n",
       "      <th>TOTHOUS</th>\n",
       "      <th>TOTFURN</th>\n",
       "      <th>TOTHLTH</th>\n",
       "      <th>TOTTRSP</th>\n",
       "      <th>TOTCMNQ</th>\n",
       "      <th>TOTRCRE</th>\n",
       "      <th>TOTEDUC</th>\n",
       "      <th>TOTMISC</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>label</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>0</th>\n",
       "      <td>3667.370932</td>\n",
       "      <td>14.542541</td>\n",
       "      <td>655.454817</td>\n",
       "      <td>742.977288</td>\n",
       "      <td>309.075706</td>\n",
       "      <td>31.879222</td>\n",
       "      <td>420.006719</td>\n",
       "      <td>149.034261</td>\n",
       "      <td>83.936939</td>\n",
       "      <td>1171.337822</td>\n",
       "      <td>141.019321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>7086.889629</td>\n",
       "      <td>373.569053</td>\n",
       "      <td>742.789712</td>\n",
       "      <td>658.137652</td>\n",
       "      <td>423.271500</td>\n",
       "      <td>205.307868</td>\n",
       "      <td>647.811045</td>\n",
       "      <td>180.181819</td>\n",
       "      <td>105.770703</td>\n",
       "      <td>1051.703190</td>\n",
       "      <td>202.121383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9597.291821</td>\n",
       "      <td>54.960748</td>\n",
       "      <td>2247.570768</td>\n",
       "      <td>2404.339304</td>\n",
       "      <td>1154.940254</td>\n",
       "      <td>185.995101</td>\n",
       "      <td>3110.407862</td>\n",
       "      <td>783.884044</td>\n",
       "      <td>616.412075</td>\n",
       "      <td>4702.824890</td>\n",
       "      <td>916.639694</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           TOTFOOD     TOTALCH      TOTCLTH      TOTHOUS      TOTFURN  \\\n",
       "label                                                                   \n",
       "0      3667.370932   14.542541   655.454817   742.977288   309.075706   \n",
       "1      7086.889629  373.569053   742.789712   658.137652   423.271500   \n",
       "2      9597.291821   54.960748  2247.570768  2404.339304  1154.940254   \n",
       "\n",
       "          TOTHLTH      TOTTRSP     TOTCMNQ     TOTRCRE      TOTEDUC  \\\n",
       "label                                                                 \n",
       "0       31.879222   420.006719  149.034261   83.936939  1171.337822   \n",
       "1      205.307868   647.811045  180.181819  105.770703  1051.703190   \n",
       "2      185.995101  3110.407862  783.884044  616.412075  4702.824890   \n",
       "\n",
       "          TOTMISC  \n",
       "label              \n",
       "0      141.019321  \n",
       "1      202.121383  \n",
       "2      916.639694  "
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "b1.groupby('label').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fd3219bdb70>"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "b1.groupby('label').mean().plot(kind='bar', figsize=(10,10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No handles with labels found to put in legend.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#parallelchord plot\n",
    "import pandas\n",
    "import matplotlib.pyplot as plt\n",
    "from pandas.plotting import parallel_coordinates\n",
    "plt.figure(figsize=(10,10))\n",
    "plt.xticks(rotation='vertical')\n",
    "plt.legend(loc='lower left')\n",
    "parallel_coordinates(XX_dfl, 'label')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This resembles upper, middle  - purple and lower class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "col_list= list(df)\n",
    "b1['TOTAL_EXP'] = df[col_list].sum(axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "b2 = b1[['label','TOTAL_EXP']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fd30bede0b8>"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "b2.groupby('label').mean().plot(kind='bar', figsize=(10,10))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have it here, upper, lower and middle class"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Counting the number of data points in each cluster"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    5657\n",
       "2    1964\n",
       "1    1301\n",
       "Name: label, dtype: int64"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "XX_dfl.label.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "b1.to_csv('b11.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "DEC - deep learning + k means"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [],
   "source": [
    "#from keras_dec import DeepEmbeddingClustering\n",
    "import numpy as np\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "USE DEC AS A FINCTION"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "'''\n",
    "Keras implementation of deep embedder to improve clustering, inspired by:\n",
    "\"Unsupervised Deep Embedding for Clustering Analysis\" (Xie et al, ICML 2016)\n",
    "\n",
    "Definition can accept somewhat custom neural networks. Defaults are from paper.\n",
    "'''\n",
    "import sys\n",
    "import numpy as np\n",
    "import keras.backend as K\n",
    "from keras.initializers import RandomNormal\n",
    "from keras.engine.topology import Layer, InputSpec\n",
    "from keras.models import Model, Sequential\n",
    "from keras.layers import Dense, Dropout, Input\n",
    "from keras.optimizers import SGD\n",
    "from sklearn.preprocessing import normalize\n",
    "from keras.callbacks import LearningRateScheduler\n",
    "from sklearn.utils.linear_assignment_ import linear_assignment\n",
    "from sklearn.cluster import KMeans\n",
    "from sklearn.decomposition import PCA\n",
    "if (sys.version[0] == 2):\n",
    "    import cPickle as pickle\n",
    "else:\n",
    "    import pickle\n",
    "import numpy as np\n",
    "\n",
    "class ClusteringLayer(Layer):\n",
    "    '''\n",
    "    Clustering layer which converts latent space Z of input layer\n",
    "    into a probability vector for each cluster defined by its centre in\n",
    "    Z-space. Use Kullback-Leibler divergence as loss, with a probability\n",
    "    target distribution.\n",
    "    # Arguments\n",
    "        output_dim: int > 0. Should be same as number of clusters.\n",
    "        input_dim: dimensionality of the input (integer).\n",
    "            This argument (or alternatively, the keyword argument `input_shape`)\n",
    "            is required when using this layer as the first layer in a model.\n",
    "        weights: list of Numpy arrays to set as initial weights.\n",
    "            The list should have 2 elements, of shape `(input_dim, output_dim)`\n",
    "            and (output_dim,) for weights and biases respectively.\n",
    "        alpha: parameter in Student's t-distribution. Default is 1.0.\n",
    "    # Input shape\n",
    "        2D tensor with shape: `(nb_samples, input_dim)`.\n",
    "    # Output shape\n",
    "        2D tensor with shape: `(nb_samples, output_dim)`.\n",
    "    '''\n",
    "    def __init__(self, output_dim, input_dim=None, weights=None, alpha=1.0, **kwargs):\n",
    "        self.output_dim = output_dim\n",
    "        self.input_dim = input_dim\n",
    "        self.alpha = alpha\n",
    "        # kmeans cluster centre locations\n",
    "        self.initial_weights = weights\n",
    "        self.input_spec = [InputSpec(ndim=2)]\n",
    "\n",
    "        if self.input_dim:\n",
    "            kwargs['input_shape'] = (self.input_dim,)\n",
    "        super(ClusteringLayer, self).__init__(**kwargs)\n",
    "\n",
    "    def build(self, input_shape):\n",
    "        assert len(input_shape) == 2\n",
    "        input_dim = input_shape[1]\n",
    "        self.input_spec = [InputSpec(dtype=K.floatx(),\n",
    "                                     shape=(None, input_dim))]\n",
    "\n",
    "        self.W = K.variable(self.initial_weights)\n",
    "        self.trainable_weights = [self.W]\n",
    "\n",
    "    def call(self, x, mask=None):\n",
    "        q = 1.0/(1.0 + K.sqrt(K.sum(K.square(K.expand_dims(x, 1) - self.W), axis=2))**2 /self.alpha)\n",
    "        q = q**((self.alpha+1.0)/2.0)\n",
    "        q = K.transpose(K.transpose(q)/K.sum(q, axis=1))\n",
    "        return q\n",
    "\n",
    "    def get_output_shape_for(self, input_shape):\n",
    "        assert input_shape and len(input_shape) == 2\n",
    "        return (input_shape[0], self.output_dim)\n",
    "\n",
    "    def compute_output_shape(self, input_shape):\n",
    "        assert input_shape and len(input_shape) == 2\n",
    "        return (input_shape[0], self.output_dim)\n",
    "\n",
    "    def get_config(self):\n",
    "        config = {'output_dim': self.output_dim,\n",
    "                  'input_dim': self.input_dim}\n",
    "        base_config = super(ClusteringLayer, self).get_config()\n",
    "        return dict(list(base_config.items()) + list(config.items()))\n",
    "\n",
    "\n",
    "class DeepEmbeddingClustering(object):\n",
    "    def __init__(self,\n",
    "                 n_clusters,\n",
    "                 input_dim,\n",
    "                 encoded=None,\n",
    "                 decoded=None,\n",
    "                 alpha=1.0,\n",
    "                 pretrained_weights=None,\n",
    "                 cluster_centres=None,\n",
    "                 batch_size=256,\n",
    "                 **kwargs):\n",
    "\n",
    "        super(DeepEmbeddingClustering, self).__init__()\n",
    "\n",
    "        self.n_clusters = n_clusters\n",
    "        self.input_dim = input_dim\n",
    "        self.encoded = encoded\n",
    "        self.decoded = decoded\n",
    "        self.alpha = alpha\n",
    "        self.pretrained_weights = pretrained_weights\n",
    "        self.cluster_centres = cluster_centres\n",
    "        self.batch_size = batch_size\n",
    "\n",
    "        self.learning_rate = 0.1\n",
    "        self.iters_lr_update = 20000\n",
    "        self.lr_change_rate = 0.1\n",
    "\n",
    "        # greedy layer-wise training before end-to-end training:\n",
    "\n",
    "        self.encoders_dims = [self.input_dim, 500, 500, 2000, 10]\n",
    "\n",
    "        self.input_layer = Input(shape=(self.input_dim,), name='input')\n",
    "        dropout_fraction = 0.2\n",
    "        init_stddev = 0.01\n",
    "\n",
    "        self.layer_wise_autoencoders = []\n",
    "        self.encoders = []\n",
    "        self.decoders = []\n",
    "        for i  in range(1, len(self.encoders_dims)):\n",
    "            \n",
    "            encoder_activation = 'linear' if i == (len(self.encoders_dims) - 1) else 'relu'\n",
    "            encoder = Dense(self.encoders_dims[i], activation=encoder_activation,\n",
    "                            input_shape=(self.encoders_dims[i-1],),\n",
    "                            kernel_initializer=RandomNormal(mean=0.0, stddev=init_stddev, seed=None),\n",
    "                            bias_initializer='zeros', name='encoder_dense_%d'%i)\n",
    "            self.encoders.append(encoder)\n",
    "\n",
    "            decoder_index = len(self.encoders_dims) - i\n",
    "            decoder_activation = 'linear' if i == 1 else 'relu'\n",
    "            decoder = Dense(self.encoders_dims[i-1], activation=decoder_activation,\n",
    "                            kernel_initializer=RandomNormal(mean=0.0, stddev=init_stddev, seed=None),\n",
    "                            bias_initializer='zeros',\n",
    "                            name='decoder_dense_%d'%decoder_index)\n",
    "            self.decoders.append(decoder)\n",
    "\n",
    "            autoencoder = Sequential([\n",
    "                Dropout(dropout_fraction, input_shape=(self.encoders_dims[i-1],), \n",
    "                        name='encoder_dropout_%d'%i),\n",
    "                encoder,\n",
    "                Dropout(dropout_fraction, name='decoder_dropout_%d'%decoder_index),\n",
    "                decoder\n",
    "            ])\n",
    "            autoencoder.compile(loss='mse', optimizer=SGD(lr=self.learning_rate, decay=0, momentum=0.9))\n",
    "            self.layer_wise_autoencoders.append(autoencoder)\n",
    "\n",
    "        # build the end-to-end autoencoder for finetuning\n",
    "        # Note that at this point dropout is discarded\n",
    "        self.encoder = Sequential(self.encoders)\n",
    "        self.encoder.compile(loss='mse', optimizer=SGD(lr=self.learning_rate, decay=0, momentum=0.9))\n",
    "        self.decoders.reverse()\n",
    "        self.autoencoder = Sequential(self.encoders + self.decoders)\n",
    "        self.autoencoder.compile(loss='mse', optimizer=SGD(lr=self.learning_rate, decay=0, momentum=0.9))\n",
    "\n",
    "        if cluster_centres is not None:\n",
    "            assert cluster_centres.shape[0] == self.n_clusters\n",
    "            assert cluster_centres.shape[1] == self.encoder.layers[-1].output_dim\n",
    "\n",
    "        if self.pretrained_weights is not None:\n",
    "            self.autoencoder.load_weights(self.pretrained_weights)\n",
    "\n",
    "    def p_mat(self, q):\n",
    "        weight = q**2 / q.sum(0)\n",
    "        return (weight.T / weight.sum(1)).T\n",
    "\n",
    "    def initialize(self, X, save_autoencoder=False, layerwise_pretrain_iters=50000, finetune_iters=100000):\n",
    "        if self.pretrained_weights is None:\n",
    "\n",
    "            iters_per_epoch = int(len(X) / self.batch_size)\n",
    "            layerwise_epochs = max(int(layerwise_pretrain_iters / iters_per_epoch), 1)\n",
    "            finetune_epochs = max(int(finetune_iters / iters_per_epoch), 1)\n",
    "\n",
    "            print('layerwise pretrain')\n",
    "            current_input = X\n",
    "            lr_epoch_update = max(1, self.iters_lr_update / float(iters_per_epoch))\n",
    "            \n",
    "            def step_decay(epoch):\n",
    "                initial_rate = self.learning_rate\n",
    "                factor = int(epoch / lr_epoch_update)\n",
    "                lr = initial_rate / (10 ** factor)\n",
    "                return lr\n",
    "            lr_schedule = LearningRateScheduler(step_decay)\n",
    "\n",
    "            for i, autoencoder in enumerate(self.layer_wise_autoencoders):\n",
    "                if i > 0:\n",
    "                    weights = self.encoders[i-1].get_weights()\n",
    "                    dense_layer = Dense(self.encoders_dims[i], input_shape=(current_input.shape[1],),\n",
    "                                        activation='relu', weights=weights,\n",
    "                                        name='encoder_dense_copy_%d'%i)\n",
    "                    encoder_model = Sequential([dense_layer])\n",
    "                    encoder_model.compile(loss='mse', optimizer=SGD(lr=self.learning_rate, decay=0, momentum=0.9))\n",
    "                    current_input = encoder_model.predict(current_input)\n",
    "\n",
    "                autoencoder.fit(current_input, current_input, \n",
    "                                batch_size=self.batch_size, epochs=layerwise_epochs, callbacks=[lr_schedule])\n",
    "                self.autoencoder.layers[i].set_weights(autoencoder.layers[1].get_weights())\n",
    "                self.autoencoder.layers[len(self.autoencoder.layers) - i - 1].set_weights(autoencoder.layers[-1].get_weights())\n",
    "            \n",
    "            print('Finetuning autoencoder')\n",
    "            \n",
    "            #update encoder and decoder weights:\n",
    "            self.autoencoder.fit(X, X, batch_size=self.batch_size, epochs=finetune_epochs, callbacks=[lr_schedule])\n",
    "\n",
    "            if save_autoencoder:\n",
    "                self.autoencoder.save_weights('autoencoder.h5')\n",
    "        else:\n",
    "            print('Loading pretrained weights for autoencoder.')\n",
    "            self.autoencoder.load_weights(self.pretrained_weights)\n",
    "\n",
    "        # update encoder, decoder\n",
    "        # TODO: is this needed? Might be redundant...\n",
    "        for i in range(len(self.encoder.layers)):\n",
    "            self.encoder.layers[i].set_weights(self.autoencoder.layers[i].get_weights())\n",
    "\n",
    "        # initialize cluster centres using k-means\n",
    "        print('Initializing cluster centres with k-means.')\n",
    "        if self.cluster_centres is None:\n",
    "            kmeans = KMeans(n_clusters=self.n_clusters, n_init=20)\n",
    "            self.y_pred = kmeans.fit_predict(self.encoder.predict(X))\n",
    "            self.cluster_centres = kmeans.cluster_centers_\n",
    "\n",
    "        # prepare DEC model\n",
    "        #self.DEC = Model(inputs=self.input_layer,\n",
    "        #                 outputs=ClusteringLayer(self.n_clusters,\n",
    "        #                                        weights=self.cluster_centres,\n",
    "        #                                        name='clustering')(self.encoder))\n",
    "        self.DEC = Sequential([self.encoder,\n",
    "                             ClusteringLayer(self.n_clusters,\n",
    "                                                weights=self.cluster_centres,\n",
    "                                                name='clustering')])\n",
    "        self.DEC.compile(loss='kullback_leibler_divergence', optimizer='adadelta')\n",
    "        return\n",
    "\n",
    "    def cluster_acc(self, y_true, y_pred):\n",
    "        assert y_pred.size == y_true.size\n",
    "        D = max(y_pred.max(), y_true.max())+1\n",
    "        w = np.zeros((D, D), dtype=np.int64)\n",
    "        for i in range(y_pred.size):\n",
    "            w[y_pred[i], y_true[i]] += 1\n",
    "        ind = linear_assignment(w.max() - w)\n",
    "        return sum([w[i, j] for i, j in ind])*1.0/y_pred.size, w\n",
    "\n",
    "    def cluster(self, X, y=None,\n",
    "                tol=0.01, update_interval=None,\n",
    "                iter_max=10000,\n",
    "                save_interval=None,\n",
    "                **kwargs):\n",
    "\n",
    "        if update_interval is None:\n",
    "            # 1 epochs\n",
    "            update_interval = X.shape[0]/self.batch_size\n",
    "        print('Update interval', update_interval)\n",
    "\n",
    "        if save_interval is None:\n",
    "            # 50 epochs\n",
    "            save_interval = X.shape[0]/self.batch_size*50\n",
    "        print('Save interval', save_interval)\n",
    "\n",
    "        assert save_interval >= update_interval\n",
    "\n",
    "        train = True\n",
    "        iteration, index = 0, 0\n",
    "        self.accuracy = []\n",
    "\n",
    "        while train:\n",
    "            sys.stdout.write('\\r')\n",
    "            # cutoff iteration\n",
    "            if iter_max < iteration:\n",
    "                print('Reached maximum iteration limit. Stopping training.')\n",
    "                return self.y_pred\n",
    "\n",
    "            # update (or initialize) probability distributions and propagate weight changes\n",
    "            # from DEC model to encoder.\n",
    "            if iteration % update_interval == 0:\n",
    "                self.q = self.DEC.predict(X, verbose=0)\n",
    "                self.p = self.p_mat(self.q)\n",
    "\n",
    "                y_pred = self.q.argmax(1)\n",
    "                delta_label = ((y_pred == self.y_pred).sum().astype(np.float32) / y_pred.shape[0])\n",
    "                if y is not None:\n",
    "                    acc = self.cluster_acc(y, y_pred)[0]\n",
    "                    self.accuracy.append(acc)\n",
    "                    print('Iteration '+str(iteration)+', Accuracy '+str(np.round(acc, 5)))\n",
    "                else:\n",
    "                    print(str(np.round(delta_label*100, 5))+'% change in label assignment')\n",
    "\n",
    "                if delta_label < tol:\n",
    "                    print('Reached tolerance threshold. Stopping training.')\n",
    "                    train = False\n",
    "                    continue\n",
    "                else:\n",
    "                    self.y_pred = y_pred\n",
    "\n",
    "                for i in range(len(self.encoder.layers)):\n",
    "                    self.encoder.layers[i].set_weights(self.DEC.layers[0].layers[i].get_weights())\n",
    "                self.cluster_centres = self.DEC.layers[-1].get_weights()[0]\n",
    "\n",
    "            # train on batch\n",
    "            sys.stdout.write('Iteration %d, ' % iteration)\n",
    "            if (index+1)*self.batch_size > X.shape[0]:\n",
    "                loss = self.DEC.train_on_batch(X[index*self.batch_size::], self.p[index*self.batch_size::])\n",
    "                index = 0\n",
    "                sys.stdout.write('Loss %f' % loss)\n",
    "            else:\n",
    "                loss = self.DEC.train_on_batch(X[index*self.batch_size:(index+1) * self.batch_size],\n",
    "                                               self.p[index*self.batch_size:(index+1) * self.batch_size])\n",
    "                sys.stdout.write('Loss %f' % loss)\n",
    "                index += 1\n",
    "\n",
    "            # save intermediate\n",
    "            if iteration % save_interval == 0:\n",
    "                z = self.encoder.predict(X)\n",
    "                pca = PCA(n_components=2).fit(z)\n",
    "                z_2d = pca.transform(z)\n",
    "                clust_2d = pca.transform(self.cluster_centres)\n",
    "                # save states for visualization\n",
    "                pickle.dump({'z_2d': z_2d, 'clust_2d': clust_2d, 'q': self.q, 'p': self.p},\n",
    "                            open('c'+str(iteration)+'.pkl', 'wb'))\n",
    "                # save DEC model checkpoints\n",
    "                self.DEC.save('DEC_model_'+str(iteration)+'.h5')\n",
    "\n",
    "            iteration += 1\n",
    "            sys.stdout.flush()\n",
    "        return"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [],
   "source": [
    "c = DeepEmbeddingClustering(n_clusters=3, input_dim=11)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "xx = XX_df.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "layerwise pretrain\n",
      "Epoch 1/29\n",
      "8922/8922 [==============================] - 1s 138us/step - loss: 0.0418\n",
      "Epoch 2/29\n",
      "8922/8922 [==============================] - 0s 19us/step - loss: 0.0230\n",
      "Epoch 3/29\n",
      "8922/8922 [==============================] - 0s 33us/step - loss: 0.0084\n",
      "Epoch 4/29\n",
      "8922/8922 [==============================] - 0s 20us/step - loss: 0.0054\n",
      "Epoch 5/29\n",
      "8922/8922 [==============================] - 0s 23us/step - loss: 0.0032\n",
      "Epoch 6/29\n",
      "8922/8922 [==============================] - 0s 18us/step - loss: 0.0022\n",
      "Epoch 7/29\n",
      "8922/8922 [==============================] - 0s 30us/step - loss: 0.0020\n",
      "Epoch 8/29\n",
      "8922/8922 [==============================] - 0s 21us/step - loss: 0.0019\n",
      "Epoch 9/29\n",
      "8922/8922 [==============================] - 0s 25us/step - loss: 0.0018\n",
      "Epoch 10/29\n",
      "8922/8922 [==============================] - 0s 19us/step - loss: 0.0018\n",
      "Epoch 11/29\n",
      "8922/8922 [==============================] - 0s 20us/step - loss: 0.0017\n",
      "Epoch 12/29\n",
      "8922/8922 [==============================] - 0s 18us/step - loss: 0.0017\n",
      "Epoch 13/29\n",
      "8922/8922 [==============================] - 0s 30us/step - loss: 0.0016\n",
      "Epoch 14/29\n",
      "8922/8922 [==============================] - 0s 18us/step - loss: 0.0016\n",
      "Epoch 15/29\n",
      "8922/8922 [==============================] - 0s 22us/step - loss: 0.0016\n",
      "Epoch 16/29\n",
      "8922/8922 [==============================] - 0s 17us/step - loss: 0.0015\n",
      "Epoch 17/29\n",
      "8922/8922 [==============================] - 0s 30us/step - loss: 0.0016\n",
      "Epoch 18/29\n",
      "8922/8922 [==============================] - 0s 18us/step - loss: 0.0016\n",
      "Epoch 19/29\n",
      "8922/8922 [==============================] - 0s 24us/step - loss: 0.0016\n",
      "Epoch 20/29\n",
      "8922/8922 [==============================] - 0s 18us/step - loss: 0.0015\n",
      "Epoch 21/29\n",
      "8922/8922 [==============================] - 0s 34us/step - loss: 0.0014\n",
      "Epoch 22/29\n",
      "8922/8922 [==============================] - 0s 29us/step - loss: 0.0015\n",
      "Epoch 23/29\n",
      "8922/8922 [==============================] - 0s 20us/step - loss: 0.0015\n",
      "Epoch 24/29\n",
      "8922/8922 [==============================] - 0s 21us/step - loss: 0.0015\n",
      "Epoch 25/29\n",
      "8922/8922 [==============================] - 0s 20us/step - loss: 0.0015\n",
      "Epoch 26/29\n",
      "8922/8922 [==============================] - 0s 18us/step - loss: 0.0014\n",
      "Epoch 27/29\n",
      "8922/8922 [==============================] - 0s 23us/step - loss: 0.0015\n",
      "Epoch 28/29\n",
      "8922/8922 [==============================] - 0s 18us/step - loss: 0.0014\n",
      "Epoch 29/29\n",
      "8922/8922 [==============================] - 0s 31us/step - loss: 0.0014\n",
      "Epoch 1/29\n",
      "8922/8922 [==============================] - 2s 204us/step - loss: 0.0014\n",
      "Epoch 2/29\n",
      "8922/8922 [==============================] - 1s 91us/step - loss: 0.0013\n",
      "Epoch 3/29\n",
      "8922/8922 [==============================] - 1s 93us/step - loss: 0.0011\n",
      "Epoch 4/29\n",
      "8922/8922 [==============================] - 1s 92us/step - loss: 0.0010\n",
      "Epoch 5/29\n",
      "8922/8922 [==============================] - 1s 95us/step - loss: 9.6020e-04\n",
      "Epoch 6/29\n",
      "8922/8922 [==============================] - 1s 93us/step - loss: 8.9909e-04\n",
      "Epoch 7/29\n",
      "8922/8922 [==============================] - 1s 91us/step - loss: 8.5143e-04\n",
      "Epoch 8/29\n",
      "8922/8922 [==============================] - 1s 92us/step - loss: 8.1318e-04\n",
      "Epoch 9/29\n",
      "8922/8922 [==============================] - 1s 95us/step - loss: 7.8259e-04\n",
      "Epoch 10/29\n",
      "8922/8922 [==============================] - 1s 94us/step - loss: 7.5844e-04\n",
      "Epoch 11/29\n",
      "8922/8922 [==============================] - 1s 93us/step - loss: 7.3951e-04\n",
      "Epoch 12/29\n",
      "8922/8922 [==============================] - 1s 99us/step - loss: 7.2457e-04\n",
      "Epoch 13/29\n",
      "8922/8922 [==============================] - 1s 92us/step - loss: 7.1276e-04\n",
      "Epoch 14/29\n",
      "8922/8922 [==============================] - 1s 90us/step - loss: 7.0315e-04\n",
      "Epoch 15/29\n",
      "8922/8922 [==============================] - 1s 91us/step - loss: 6.9512e-04\n",
      "Epoch 16/29\n",
      "8922/8922 [==============================] - 1s 93us/step - loss: 6.8722e-04\n",
      "Epoch 17/29\n",
      "8922/8922 [==============================] - 1s 92us/step - loss: 6.7971e-04\n",
      "Epoch 18/29\n",
      "8922/8922 [==============================] - 1s 94us/step - loss: 6.7316e-04\n",
      "Epoch 19/29\n",
      "8922/8922 [==============================] - 1s 93us/step - loss: 6.6742e-04\n",
      "Epoch 20/29\n",
      "8922/8922 [==============================] - 1s 93us/step - loss: 6.6213e-04\n",
      "Epoch 21/29\n",
      "8922/8922 [==============================] - 1s 89us/step - loss: 6.5725e-04\n",
      "Epoch 22/29\n",
      "8922/8922 [==============================] - 1s 95us/step - loss: 6.5257e-04\n",
      "Epoch 23/29\n",
      "8922/8922 [==============================] - 1s 99us/step - loss: 6.4811e-04\n",
      "Epoch 24/29\n",
      "8922/8922 [==============================] - 1s 95us/step - loss: 6.4336e-04\n",
      "Epoch 25/29\n",
      "8922/8922 [==============================] - 1s 89us/step - loss: 6.3870e-04\n",
      "Epoch 26/29\n",
      "8922/8922 [==============================] - 1s 93us/step - loss: 6.3392e-04\n",
      "Epoch 27/29\n",
      "8922/8922 [==============================] - 1s 91us/step - loss: 6.2917e-04\n",
      "Epoch 28/29\n",
      "8922/8922 [==============================] - 1s 97us/step - loss: 6.2413e-04\n",
      "Epoch 29/29\n",
      "8922/8922 [==============================] - 1s 100us/step - loss: 6.1890e-04\n",
      "Epoch 1/29\n",
      "8922/8922 [==============================] - 4s 409us/step - loss: 1.4695e-04\n",
      "Epoch 2/29\n",
      "8922/8922 [==============================] - 3s 289us/step - loss: 1.3259e-04\n",
      "Epoch 3/29\n",
      "8922/8922 [==============================] - 3s 293us/step - loss: 1.1672e-04\n",
      "Epoch 4/29\n",
      "8922/8922 [==============================] - 3s 289us/step - loss: 1.0307e-04\n",
      "Epoch 5/29\n",
      "8922/8922 [==============================] - 3s 302us/step - loss: 9.2245e-05\n",
      "Epoch 6/29\n",
      "8922/8922 [==============================] - 3s 290us/step - loss: 8.3620e-05\n",
      "Epoch 7/29\n",
      "8922/8922 [==============================] - 3s 296us/step - loss: 7.6580e-05\n",
      "Epoch 8/29\n",
      "8922/8922 [==============================] - 3s 294us/step - loss: 7.1086e-05\n",
      "Epoch 9/29\n",
      "8922/8922 [==============================] - 3s 289us/step - loss: 6.6863e-05\n",
      "Epoch 10/29\n",
      "8922/8922 [==============================] - 3s 290us/step - loss: 6.3503e-05\n",
      "Epoch 11/29\n",
      "8922/8922 [==============================] - 3s 290us/step - loss: 6.0883e-05\n",
      "Epoch 12/29\n",
      "8922/8922 [==============================] - 3s 296us/step - loss: 5.8906e-05\n",
      "Epoch 13/29\n",
      "8922/8922 [==============================] - 3s 294us/step - loss: 5.7399e-05\n",
      "Epoch 14/29\n",
      "8922/8922 [==============================] - 3s 302us/step - loss: 5.6228e-05\n",
      "Epoch 15/29\n",
      "8922/8922 [==============================] - 3s 290us/step - loss: 5.5237e-05\n",
      "Epoch 16/29\n",
      "8922/8922 [==============================] - 3s 288us/step - loss: 5.4433e-05\n",
      "Epoch 17/29\n",
      "8922/8922 [==============================] - 3s 289us/step - loss: 5.3805e-05\n",
      "Epoch 18/29\n",
      "8922/8922 [==============================] - 3s 291us/step - loss: 5.3301e-05\n",
      "Epoch 19/29\n",
      "8922/8922 [==============================] - 3s 293us/step - loss: 5.2876e-05\n",
      "Epoch 20/29\n",
      "8922/8922 [==============================] - 3s 289us/step - loss: 5.2477e-05\n",
      "Epoch 21/29\n",
      "8922/8922 [==============================] - 3s 293us/step - loss: 5.1912e-05\n",
      "Epoch 22/29\n",
      "8922/8922 [==============================] - 3s 292us/step - loss: 5.1255e-05\n",
      "Epoch 23/29\n",
      "8922/8922 [==============================] - 3s 290us/step - loss: 5.0664e-05\n",
      "Epoch 24/29\n",
      "8922/8922 [==============================] - 3s 288us/step - loss: 5.0196e-05\n",
      "Epoch 25/29\n",
      "8922/8922 [==============================] - 3s 296us/step - loss: 4.9781e-05\n",
      "Epoch 26/29\n",
      "8922/8922 [==============================] - 3s 289us/step - loss: 4.9413e-05\n",
      "Epoch 27/29\n",
      "8922/8922 [==============================] - 3s 296us/step - loss: 4.9115e-05\n",
      "Epoch 28/29\n",
      "8922/8922 [==============================] - 3s 289us/step - loss: 4.8887e-05\n",
      "Epoch 29/29\n",
      "8922/8922 [==============================] - 3s 296us/step - loss: 4.8678e-05\n",
      "Epoch 1/29\n",
      "8922/8922 [==============================] - 2s 179us/step - loss: 5.6490e-06\n",
      "Epoch 2/29\n",
      "8922/8922 [==============================] - 1s 65us/step - loss: 5.4500e-06\n",
      "Epoch 3/29\n",
      "8922/8922 [==============================] - 1s 59us/step - loss: 5.1997e-06\n",
      "Epoch 4/29\n",
      "8922/8922 [==============================] - 1s 73us/step - loss: 4.9436e-06\n",
      "Epoch 5/29\n",
      "8922/8922 [==============================] - 1s 62us/step - loss: 4.6971e-06\n",
      "Epoch 6/29\n",
      "8922/8922 [==============================] - 1s 64us/step - loss: 4.4634e-06\n",
      "Epoch 7/29\n",
      "8922/8922 [==============================] - 1s 71us/step - loss: 4.2439e-06\n",
      "Epoch 8/29\n",
      "8922/8922 [==============================] - 1s 62us/step - loss: 4.0378e-06\n",
      "Epoch 9/29\n",
      "8922/8922 [==============================] - 1s 68us/step - loss: 3.8445e-06\n",
      "Epoch 10/29\n",
      "8922/8922 [==============================] - 1s 61us/step - loss: 3.6637e-06\n",
      "Epoch 11/29\n",
      "8922/8922 [==============================] - 1s 62us/step - loss: 3.4945e-06\n",
      "Epoch 12/29\n",
      "8922/8922 [==============================] - 1s 67us/step - loss: 3.3357e-06\n",
      "Epoch 13/29\n",
      "8922/8922 [==============================] - 1s 69us/step - loss: 3.1877e-06\n",
      "Epoch 14/29\n",
      "8922/8922 [==============================] - 1s 60us/step - loss: 3.0495e-06\n",
      "Epoch 15/29\n",
      "8922/8922 [==============================] - 1s 63us/step - loss: 2.9209e-06\n",
      "Epoch 16/29\n",
      "8922/8922 [==============================] - 1s 72us/step - loss: 2.8005e-06\n",
      "Epoch 17/29\n",
      "8922/8922 [==============================] - 1s 61us/step - loss: 2.6885e-06\n",
      "Epoch 18/29\n",
      "8922/8922 [==============================] - 1s 69us/step - loss: 2.5837e-06\n",
      "Epoch 19/29\n",
      "8922/8922 [==============================] - 1s 83us/step - loss: 2.4859e-06\n",
      "Epoch 20/29\n",
      "8922/8922 [==============================] - 1s 62us/step - loss: 2.3946e-06\n",
      "Epoch 21/29\n",
      "8922/8922 [==============================] - 1s 61us/step - loss: 2.3096e-06\n",
      "Epoch 22/29\n",
      "8922/8922 [==============================] - 1s 62us/step - loss: 2.2300e-06\n",
      "Epoch 23/29\n",
      "8922/8922 [==============================] - 1s 60us/step - loss: 2.1556e-06\n",
      "Epoch 24/29\n",
      "8922/8922 [==============================] - 1s 74us/step - loss: 2.0864e-06\n",
      "Epoch 25/29\n",
      "8922/8922 [==============================] - 1s 61us/step - loss: 2.0219e-06\n",
      "Epoch 26/29\n",
      "8922/8922 [==============================] - 1s 63us/step - loss: 1.9616e-06\n",
      "Epoch 27/29\n",
      "8922/8922 [==============================] - 1s 60us/step - loss: 1.9056e-06\n",
      "Epoch 28/29\n",
      "8922/8922 [==============================] - 1s 69us/step - loss: 1.8531e-06\n",
      "Epoch 29/29\n",
      "8922/8922 [==============================] - 1s 66us/step - loss: 1.8043e-06\n",
      "Finetuning autoencoder\n",
      "Epoch 1/29\n",
      "8922/8922 [==============================] - 5s 559us/step - loss: 0.0426\n",
      "Epoch 2/29\n",
      "8922/8922 [==============================] - 4s 430us/step - loss: 0.0426\n",
      "Epoch 3/29\n",
      "8922/8922 [==============================] - 4s 427us/step - loss: 0.0426\n",
      "Epoch 4/29\n",
      "8922/8922 [==============================] - 4s 426us/step - loss: 0.0426\n",
      "Epoch 5/29\n",
      "8922/8922 [==============================] - 4s 428us/step - loss: 0.0426\n",
      "Epoch 6/29\n",
      "8922/8922 [==============================] - 4s 437us/step - loss: 0.0426\n",
      "Epoch 7/29\n",
      "8922/8922 [==============================] - 4s 426us/step - loss: 0.0426\n",
      "Epoch 8/29\n",
      "8922/8922 [==============================] - 4s 434us/step - loss: 0.0426\n",
      "Epoch 9/29\n",
      "8922/8922 [==============================] - 4s 432us/step - loss: 0.0426\n",
      "Epoch 10/29\n",
      "8922/8922 [==============================] - 4s 423us/step - loss: 0.0426\n",
      "Epoch 11/29\n",
      "8922/8922 [==============================] - 4s 431us/step - loss: 0.0426\n",
      "Epoch 12/29\n",
      "8922/8922 [==============================] - 4s 430us/step - loss: 0.0426\n",
      "Epoch 13/29\n",
      "8922/8922 [==============================] - 4s 431us/step - loss: 0.0426\n",
      "Epoch 14/29\n",
      "8922/8922 [==============================] - 4s 430us/step - loss: 0.0426\n",
      "Epoch 15/29\n",
      "8922/8922 [==============================] - 4s 430us/step - loss: 0.0426\n",
      "Epoch 16/29\n",
      "8922/8922 [==============================] - 4s 430us/step - loss: 0.0425\n",
      "Epoch 17/29\n",
      "8922/8922 [==============================] - 4s 430us/step - loss: 0.0425\n",
      "Epoch 18/29\n",
      "8922/8922 [==============================] - 4s 424us/step - loss: 0.0427\n",
      "Epoch 19/29\n",
      "8922/8922 [==============================] - 4s 430us/step - loss: 0.0425\n",
      "Epoch 20/29\n",
      "8922/8922 [==============================] - 4s 456us/step - loss: 0.0426\n",
      "Epoch 21/29\n",
      "8922/8922 [==============================] - 4s 437us/step - loss: 0.0425\n",
      "Epoch 22/29\n",
      "8922/8922 [==============================] - 4s 432us/step - loss: 0.0424\n",
      "Epoch 23/29\n",
      "8922/8922 [==============================] - 4s 430us/step - loss: 0.0423\n",
      "Epoch 24/29\n",
      "8922/8922 [==============================] - 4s 437us/step - loss: 0.0421\n",
      "Epoch 25/29\n",
      "8922/8922 [==============================] - 4s 430us/step - loss: 0.0418\n",
      "Epoch 26/29\n",
      "8922/8922 [==============================] - 4s 429us/step - loss: 0.0404\n",
      "Epoch 27/29\n",
      "8922/8922 [==============================] - 4s 431us/step - loss: 0.0300\n",
      "Epoch 28/29\n",
      "8922/8922 [==============================] - 4s 430us/step - loss: 0.0099\n",
      "Epoch 29/29\n",
      "8922/8922 [==============================] - 4s 437us/step - loss: 0.0090\n",
      "Initializing cluster centres with k-means.\n"
     ]
    }
   ],
   "source": [
    "c.initialize(xx, finetune_iters=1000, layerwise_pretrain_iters=1000)\n",
    "#cl = c.cluster(xx, y=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Update interval 34.8515625\n",
      "Save interval 1742.578125\n",
      "100.0% change in label assignment\n",
      "Reached maximum iteration limit. Stopping training.\n"
     ]
    }
   ],
   "source": [
    "cl = c.cluster(xx, y=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([2, 2, 1, ..., 1, 1, 1])"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cl"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8922, 11)"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "xx.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8922,)"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cl.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cl.max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cl.min()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Silhouette Coefficient: 0.417\n"
     ]
    }
   ],
   "source": [
    "#from sklearn.metrics import silhouette_samples, silhouette_score\n",
    "#from sklearn import metrics\n",
    "print(\"Silhouette Coefficient: %0.3f\"\n",
    "      % metrics.silhouette_score(xx, cl))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "DEC DOES NOT IMPROVE THE CLUSTERING"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We stick autoencoder + kmeans"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}