note.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib as mpl\n",
    "import matplotlib.pylab as plt\n",
    "import seaborn as sns\n",
    "%matplotlib inline\n",
    "import time\n",
    "from numpy.random import rand\n",
    "import random\n",
    "import cv2\n",
    "from PIL import Image\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.metrics.cluster import adjusted_rand_score\n",
    "from sklearn.metrics.cluster import normalized_mutual_info_score\n",
    "from sklearn.datasets import fetch_openml\n",
    "from sklearn.datasets import fetch_olivetti_faces"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "from CCE_algorithm import CCE\n",
    "# input (connectivity, max_iteration, target_center, check)\n",
    "# output (n_center, center_id, label, n_cut)\n",
    "from CCE_algorithm import connectivity_matrix\n",
    "# input (X, sigma)\n",
    "# output (similarity matrix)\n",
    "from CCE_algorithm import enhanced_connectivity_matrix\n",
    "# input (X, sigma, epsilon)\n",
    "# output (enhanced matrix)|\n",
    "from CCE_algorithm_new import CCE_experiment"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# mnist data set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Experiment(X,y,sigma,epsilon=0,max_it=1000, check=100, num_cluster = None):\n",
    "    start_time = time.time()\n",
    "    \n",
    "    if num_cluster == None :\n",
    "        n_cluster = len(set(list(y)))\n",
    "    else :\n",
    "        n_cluster = num_cluster\n",
    "    \n",
    "    if epsilon ==0 :\n",
    "        S = connectivity_matrix(X,sigma)\n",
    "    else :\n",
    "        S = enhanced_connectivity_matrix(X,sigma,epsilon)\n",
    "    result = CCE(S,max_it,n_cluster,check)\n",
    "    runtime = (time.time() - start_time)\n",
    "    iteration = len(result[0])\n",
    "    ARI = adjusted_rand_score(result[2], np.array(y))\n",
    "    NMI = normalized_mutual_info_score(result[2], np.array(y))\n",
    "    return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Experiment2(X,y,sigma,epsilon=0,max_it=1000, check=100, num_cluster = None):\n",
    "    start_time = time.time()\n",
    "    \n",
    "    if num_cluster == None :\n",
    "        n_cluster = len(set(list(y)))\n",
    "    else :\n",
    "        n_cluster = num_cluster\n",
    "    \n",
    "    if epsilon ==0 :\n",
    "        S = connectivity_matrix(X,sigma)\n",
    "    else :\n",
    "        S = enhanced_connectivity_matrix(X,sigma,epsilon)\n",
    "    result = CCE_experiment(S,y,max_it,n_cluster,check)\n",
    "    runtime = (time.time() - start_time)\n",
    "    iteration = len(result[0])\n",
    "    ARI = adjusted_rand_score(result[2], np.array(y))\n",
    "    \n",
    "    return np.array([result[2], iteration, result[3], ARI, result[4]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "face = fetch_olivetti_faces()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(400, 4096)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "face['data'].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<PIL.Image.Image image mode=RGB size=64x64 at 0x26E0083A790>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "image = Image.fromarray(face['images'][10]*255)\n",
    "image.convert('RGB')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.87603307"
      ]
     },
     "execution_count": 148,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "face['data'][0].max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 392,
   "metadata": {},
   "outputs": [],
   "source": [
    "sampled_data = face['data'][:40]\n",
    "sampled_label = face['target'][:40]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 393,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "order of connectivity matrix = 1\n",
      "number of cluster center = 40\n",
      "...\n",
      "order of connectivity matrix = 1001\n",
      "number of cluster center = 22\n",
      "...\n",
      "order of connectivity matrix = 2001\n",
      "number of cluster center = 17\n",
      "...\n",
      "order of connectivity matrix = 3001\n",
      "number of cluster center = 15\n",
      "...\n",
      "order of connectivity matrix = 4001\n",
      "number of cluster center = 14\n",
      "...\n",
      "order of connectivity matrix = 5001\n",
      "number of cluster center = 13\n",
      "...\n",
      "order of connectivity matrix = 6001\n",
      "number of cluster center = 11\n",
      "...\n",
      "order of connectivity matrix = 7001\n",
      "number of cluster center = 10\n",
      "...\n",
      "order of connectivity matrix = 8001\n",
      "number of cluster center = 10\n",
      "...\n",
      "order of connectivity matrix = 9001\n",
      "number of cluster center = 8\n",
      "...\n",
      "order of connectivity matrix = 10001\n",
      "number of cluster center = 7\n",
      "...\n",
      "order of connectivity matrix = 11001\n",
      "number of cluster center = 7\n",
      "...\n",
      "order of connectivity matrix = 12001\n",
      "number of cluster center = 7\n",
      "...\n",
      "order of connectivity matrix = 13001\n",
      "number of cluster center = 7\n",
      "...\n",
      "order of connectivity matrix = 14001\n",
      "number of cluster center = 7\n",
      "...\n",
      "order of connectivity matrix = 15001\n",
      "number of cluster center = 6\n",
      "...\n",
      "order of connectivity matrix = 16001\n",
      "number of cluster center = 6\n",
      "...\n",
      "order of connectivity matrix = 17001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 18001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 19001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 20001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 21001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 22001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 23001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 24001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 25001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 26001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 27001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 28001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 29001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 30001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 31001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 32001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 33001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 34001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 35001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 36001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 37001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 38001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 39001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 40001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 41001\n",
      "number of cluster center = 5\n",
      "...\n",
      "order of connectivity matrix = 42001\n",
      "number of cluster center = 5\n",
      "...\n",
      "less then 4 cluster center\n",
      "order 42044\n",
      "[3, 9, 19, 23]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-20-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([array([23,  9, 23,  3,  9,  3, 23, 23, 23,  9, 19, 19, 19, 19, 19, 19, 19,\n",
       "       19, 19, 19, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23,\n",
       "       23,  3, 23, 23, 23, 23]),\n",
       "       42044, 0.36249729963275, 0.5790387231670985, 17.85018229484558, 0],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 393,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Experiment(sampled_data, sampled_label, 2, epsilon=0, max_it= 70000, check=1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "order of connectivity matrix = 1\n",
      "number of cluster center = 40\n",
      "...\n",
      "less then 1000 cluster center\n",
      "order 1\n",
      "[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39]\n",
      "[array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16,\n",
      "       17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33,\n",
      "       34, 35, 36, 37, 38, 39])\n",
      " 1 0.0 0.5463041903939284 0.002991914749145508 0]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 40\n",
      "...\n",
      "less then 1000 cluster center\n",
      "order 1\n",
      "[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39]\n",
      "[array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16,\n",
      "       17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33,\n",
      "       34, 35, 36, 37, 38, 39])\n",
      " 1 0.0 0.5463041903939284 0.001989126205444336 1e-08]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 40\n",
      "...\n",
      "less then 1000 cluster center\n",
      "order 1\n",
      "[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39]\n",
      "[array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16,\n",
      "       17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33,\n",
      "       34, 35, 36, 37, 38, 39])\n",
      " 1 0.0 0.5463041903939284 0.0009975433349609375 5e-08]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 40\n",
      "...\n",
      "less then 1000 cluster center\n",
      "order 1\n",
      "[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39]\n",
      "[array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16,\n",
      "       17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33,\n",
      "       34, 35, 36, 37, 38, 39])\n",
      " 1 0.0 0.5463041903939284 0.0019943714141845703 1e-07]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 40\n",
      "...\n",
      "less then 1000 cluster center\n",
      "order 1\n",
      "[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39]\n",
      "[array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16,\n",
      "       17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33,\n",
      "       34, 35, 36, 37, 38, 39])\n",
      " 1 0.0 0.5463041903939284 0.0019948482513427734 5e-07]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 40\n",
      "...\n",
      "less then 1000 cluster center\n",
      "order 1\n",
      "[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39]\n",
      "[array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16,\n",
      "       17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33,\n",
      "       34, 35, 36, 37, 38, 39])\n",
      " 1 0.0 0.5463041903939284 0.0029931068420410156 1e-06]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 40\n",
      "...\n",
      "less then 1000 cluster center\n",
      "order 1\n",
      "[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39]\n",
      "[array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16,\n",
      "       17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33,\n",
      "       34, 35, 36, 37, 38, 39])\n",
      " 1 0.0 0.5463041903939284 0.0019943714141845703 5e-06]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 40\n",
      "...\n",
      "less then 1000 cluster center\n",
      "order 1\n",
      "[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39]\n",
      "[array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16,\n",
      "       17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33,\n",
      "       34, 35, 36, 37, 38, 39])\n",
      " 1 0.0 0.5463041903939284 0.001994609832763672 1e-05]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-94-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n",
      "<ipython-input-94-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n",
      "<ipython-input-94-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n",
      "<ipython-input-94-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n",
      "<ipython-input-94-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n",
      "<ipython-input-94-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n",
      "<ipython-input-94-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n",
      "<ipython-input-94-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n"
     ]
    }
   ],
   "source": [
    "epsilon_list = [0, 1e-8, 5e-8, 1e-7, 5e-7, 1e-6, 5e-6, 1e-5]\n",
    "for epsilon in epsilon_list :\n",
    "    print(Experiment(sampled_data, sampled_label, 2, epsilon=epsilon, max_it= 70000, check=1000))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Complex dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "arrhythima = pd.read_csv('./real_data/archive/data_arrhythmia.csv', sep=';')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(452, 280)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "arrhythima.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "nominal_columns = [1] + list(np.array([22, 23, 24, 25, 26, 27]) + 12 -1) \\\n",
    "+ list(np.array([22, 23, 24, 25, 26, 27]) + 12*2 -1) \\\n",
    "+ list(np.array([22, 23, 24, 25, 26, 27]) + 12*3 -1) \\\n",
    "+ list(np.array([22, 23, 24, 25, 26, 27]) + 12*4 -1) \\\n",
    "+ list(np.array([22, 23, 24, 25, 26, 27]) + 12*5 -1) \\\n",
    "+ list(np.array([22, 23, 24, 25, 26, 27]) + 12*6 -1) \\\n",
    "+ list(np.array([22, 23, 24, 25, 26, 27]) + 12*7 -1) \\\n",
    "+ list(np.array([22, 23, 24, 25, 26, 27]) + 12*8 -1) \\\n",
    "+ list(np.array([22, 23, 24, 25, 26, 27]) + 12*9 -1) \\\n",
    "+ list(np.array([22, 23, 24, 25, 26, 27]) + 12*10 -1) \\\n",
    "+ list(np.array([22, 23, 24, 25, 26, 27]) + 12*11 -1) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = MinMaxScaler()\n",
    "\n",
    "lis = arrhythima.columns[norminal_columns].to_list()\n",
    "\n",
    "df1 = arrhythima.iloc[:,[i not in lis for i in arrhythima.columns.to_list()]]\n",
    "df1 = df1.replace({'?' : None})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "def preprocessing_dataset(data, nominal_col_list, ax=0):\n",
    "    scaler = MinMaxScaler()\n",
    "\n",
    "    data = data.replace({'?' : None})\n",
    "    data = data.dropna(axis=ax)\n",
    "    \n",
    "    X = data.iloc[:,:-1]\n",
    "    y = data.iloc[:,-1]    \n",
    "    \n",
    "    numeric_data = X.iloc[:,[i not in nominal_col_list for i in range(X.shape[1])]]\n",
    "    \n",
    "    cat_data = X.iloc[:,nominal_col_list]\n",
    "    cat_data_ = pd.get_dummies(cat_data)\n",
    "    \n",
    "    final_df = pd.concat([cat_data_, numeric_data], axis=1)\n",
    "    final_df = final_df.astype(float)\n",
    "    final_X = scaler.fit_transform(final_df)\n",
    "    \n",
    "    return(final_X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_X, y = preprocessing_dataset(arrhythima, nominal_columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(68, 279)"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_X2, y2 = preprocessing_dataset(arrhythima, nominal_columns, ax=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(452, 274)"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_X2.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "order of connectivity matrix = 1\n",
      "number of cluster center = 452\n",
      "...\n",
      "order of connectivity matrix = 201\n",
      "number of cluster center = 69\n",
      "...\n",
      "order of connectivity matrix = 401\n",
      "number of cluster center = 61\n",
      "...\n",
      "order of connectivity matrix = 601\n",
      "number of cluster center = 58\n",
      "...\n",
      "order of connectivity matrix = 801\n",
      "number of cluster center = 54\n",
      "...\n",
      "order of connectivity matrix = 1001\n",
      "number of cluster center = 49\n",
      "...\n",
      "order of connectivity matrix = 1201\n",
      "number of cluster center = 48\n",
      "...\n",
      "order of connectivity matrix = 1401\n",
      "number of cluster center = 47\n",
      "...\n",
      "order of connectivity matrix = 1601\n",
      "number of cluster center = 45\n",
      "...\n",
      "order of connectivity matrix = 1801\n",
      "number of cluster center = 43\n",
      "...\n",
      "order of connectivity matrix = 2001\n",
      "number of cluster center = 43\n",
      "...\n",
      "order of connectivity matrix = 2201\n",
      "number of cluster center = 43\n",
      "...\n",
      "order of connectivity matrix = 2401\n",
      "number of cluster center = 43\n",
      "...\n",
      "order of connectivity matrix = 2601\n",
      "number of cluster center = 42\n",
      "...\n",
      "order of connectivity matrix = 2801\n",
      "number of cluster center = 41\n",
      "...\n",
      "order of connectivity matrix = 3001\n",
      "number of cluster center = 40\n",
      "...\n",
      "order of connectivity matrix = 3201\n",
      "number of cluster center = 38\n",
      "...\n",
      "order of connectivity matrix = 3401\n",
      "number of cluster center = 38\n",
      "...\n",
      "order of connectivity matrix = 3601\n",
      "number of cluster center = 37\n",
      "...\n",
      "order of connectivity matrix = 3801\n",
      "number of cluster center = 37\n",
      "...\n",
      "order of connectivity matrix = 4001\n",
      "number of cluster center = 36\n",
      "...\n",
      "order of connectivity matrix = 4201\n",
      "number of cluster center = 36\n",
      "...\n",
      "order of connectivity matrix = 4401\n",
      "number of cluster center = 34\n",
      "...\n",
      "order of connectivity matrix = 4601\n",
      "number of cluster center = 34\n",
      "...\n",
      "order of connectivity matrix = 4801\n",
      "number of cluster center = 34\n",
      "...\n",
      "[array([325, 325,   2, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325,  76, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325,  85, 325, 325,  88,  89, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 325, 108, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 141, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 325, 325, 187, 325, 189, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 198, 325, 325, 325, 325, 325, 204, 325, 325, 207,\n",
      "       325, 325, 210, 325, 325, 213, 325, 325, 325, 325, 218, 325, 325,\n",
      "       325, 325, 325, 325, 257, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 257, 325, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 325, 325, 291, 325, 293, 325, 325, 325, 297, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 325, 325, 308, 325, 325, 325,\n",
      "       312, 325, 325, 325, 316, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 327, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 354, 325, 325, 325, 325, 325, 325, 257, 325, 325,\n",
      "       325, 325, 325, 325, 325, 325, 370, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 379, 325, 381, 325, 325, 325, 325, 325, 325, 388, 325,\n",
      "       325, 325, 325, 325, 325, 395, 325, 325, 325, 325, 325, 325, 325,\n",
      "       403, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 325, 424, 325, 325, 325, 325,\n",
      "       325, 430, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325, 325,\n",
      "       325, 325, 325, 325, 325, 325, 325, 449, 325, 325])\n",
      " 5000 0.27931129286406026 0.13054405999435462 11]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-24-9eb1fb91457a>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, result[3], ARI, result[4]])\n"
     ]
    }
   ],
   "source": [
    "print(Experiment2(final_X2, y2, 0.6, epsilon=0, max_it= 5000, check=200, num_cluster=17))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "order of connectivity matrix = 1\n",
      "number of cluster center = 68\n",
      "...\n",
      "order of connectivity matrix = 1001\n",
      "number of cluster center = 31\n",
      "...\n",
      "order of connectivity matrix = 2001\n",
      "number of cluster center = 28\n",
      "...\n",
      "order of connectivity matrix = 3001\n",
      "number of cluster center = 26\n",
      "...\n",
      "order of connectivity matrix = 4001\n",
      "number of cluster center = 26\n",
      "...\n",
      "order of connectivity matrix = 5001\n",
      "number of cluster center = 23\n",
      "...\n",
      "order of connectivity matrix = 6001\n",
      "number of cluster center = 22\n",
      "...\n",
      "order of connectivity matrix = 7001\n",
      "number of cluster center = 22\n",
      "...\n",
      "order of connectivity matrix = 8001\n",
      "number of cluster center = 22\n",
      "...\n",
      "order of connectivity matrix = 9001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 10001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 11001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 12001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 13001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 14001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 15001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 16001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 17001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 18001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 19001\n",
      "number of cluster center = 20\n",
      "...\n",
      "order of connectivity matrix = 20001\n",
      "number of cluster center = 20\n",
      "...\n",
      "order of connectivity matrix = 21001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 22001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 23001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 24001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 25001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 26001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 27001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 28001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 29001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 30001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 31001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 32001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 33001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 34001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 35001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 36001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 37001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 38001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 39001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 40001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 41001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 42001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 43001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 44001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 45001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 46001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 47001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 48001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 49001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 50001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 51001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 52001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 53001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 54001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 55001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 56001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 57001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 58001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 59001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 60001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 61001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 62001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 63001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 64001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 65001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 66001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 67001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 68001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 69001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 70001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 71001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 72001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 73001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 74001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 75001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 76001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 77001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 78001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 79001\n",
      "number of cluster center = 18\n",
      "...\n",
      "[array([ 0, 12, 12, 12, 12,  5, 12,  7, 12, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       17, 12, 12, 20, 12, 22, 12, 12, 25, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       34, 12, 12, 37, 12, 12, 12, 12, 12, 12, 12, 45, 12, 12, 12, 12, 12,\n",
      "       12, 52, 12, 12, 55, 12, 57, 12, 59, 12, 12, 62, 12, 64, 12, 66, 12])\n",
      " 80000 0.2211472213671079 0.11319813627571443 5324]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-24-9eb1fb91457a>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, result[3], ARI, result[4]])\n"
     ]
    }
   ],
   "source": [
    "print(Experiment2(final_X, y, 0.7, epsilon=0, max_it= 80000, check=80000, num_cluster=17))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "order of connectivity matrix = 1\n",
      "number of cluster center = 68\n",
      "...\n",
      "order of connectivity matrix = 2001\n",
      "number of cluster center = 28\n",
      "...\n",
      "order of connectivity matrix = 4001\n",
      "number of cluster center = 26\n",
      "...\n",
      "order of connectivity matrix = 6001\n",
      "number of cluster center = 22\n",
      "...\n",
      "order of connectivity matrix = 8001\n",
      "number of cluster center = 22\n",
      "...\n",
      "order of connectivity matrix = 10001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 12001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 14001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 16001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 18001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 20001\n",
      "number of cluster center = 20\n",
      "...\n",
      "order of connectivity matrix = 22001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 24001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 26001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 28001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 30001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 32001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 34001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 36001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 38001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 40001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 42001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 44001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 46001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 48001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 50001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 52001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 54001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 56001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 58001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 60001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 62001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 64001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 66001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 68001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 70001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 72001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 74001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 76001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 78001\n",
      "number of cluster center = 18\n",
      "...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-24-9eb1fb91457a>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, result[3], ARI, result[4]])\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[array([ 0, 12, 12, 12, 12,  5, 12,  7, 12, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       17, 12, 12, 20, 12, 22, 12, 12, 25, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       34, 12, 12, 37, 12, 12, 12, 12, 12, 12, 12, 45, 12, 12, 12, 12, 12,\n",
      "       12, 52, 12, 12, 55, 12, 57, 12, 59, 12, 12, 62, 12, 64, 12, 66, 12])\n",
      " 80000 0.2211472213671079 0.11319813627571443 5324]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 68\n",
      "...\n",
      "order of connectivity matrix = 2001\n",
      "number of cluster center = 28\n",
      "...\n",
      "order of connectivity matrix = 4001\n",
      "number of cluster center = 26\n",
      "...\n",
      "order of connectivity matrix = 6001\n",
      "number of cluster center = 22\n",
      "...\n",
      "order of connectivity matrix = 8001\n",
      "number of cluster center = 22\n",
      "...\n",
      "order of connectivity matrix = 10001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 12001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 14001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 16001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 18001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 20001\n",
      "number of cluster center = 20\n",
      "...\n",
      "order of connectivity matrix = 22001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 24001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 26001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 28001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 30001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 32001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 34001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 36001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 38001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 40001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 42001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 44001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 46001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 48001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 50001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 52001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 54001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 56001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 58001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 60001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 62001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 64001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 66001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 68001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 70001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 72001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 74001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 76001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 78001\n",
      "number of cluster center = 18\n",
      "...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-24-9eb1fb91457a>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, result[3], ARI, result[4]])\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[array([ 0, 12, 12, 12, 12,  5, 12,  7, 12, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       17, 12, 12, 20, 12, 22, 12, 12, 25, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       34, 12, 12, 37, 12, 12, 12, 12, 12, 12, 12, 45, 12, 12, 12, 12, 12,\n",
      "       12, 52, 12, 12, 55, 12, 57, 12, 59, 12, 12, 62, 12, 64, 12, 66, 12])\n",
      " 80000 0.2211472213671079 0.11319813627571443 5323]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 68\n",
      "...\n",
      "order of connectivity matrix = 2001\n",
      "number of cluster center = 28\n",
      "...\n",
      "order of connectivity matrix = 4001\n",
      "number of cluster center = 26\n",
      "...\n",
      "order of connectivity matrix = 6001\n",
      "number of cluster center = 22\n",
      "...\n",
      "order of connectivity matrix = 8001\n",
      "number of cluster center = 22\n",
      "...\n",
      "order of connectivity matrix = 10001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 12001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 14001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 16001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 18001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 20001\n",
      "number of cluster center = 20\n",
      "...\n",
      "order of connectivity matrix = 22001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 24001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 26001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 28001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 30001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 32001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 34001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 36001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 38001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 40001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 42001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 44001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 46001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 48001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 50001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 52001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 54001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 56001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 58001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 60001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 62001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 64001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 66001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 68001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 70001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 72001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 74001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 76001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 78001\n",
      "number of cluster center = 18\n",
      "...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-24-9eb1fb91457a>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, result[3], ARI, result[4]])\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[array([ 0, 12, 12, 12, 12,  5, 12,  7, 12, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       17, 12, 12, 20, 12, 22, 12, 12, 25, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       34, 12, 12, 37, 12, 12, 12, 12, 12, 12, 12, 45, 12, 12, 12, 12, 12,\n",
      "       12, 52, 12, 12, 55, 12, 57, 12, 59, 12, 12, 62, 12, 64, 12, 66, 12])\n",
      " 80000 0.2211472213671079 0.11319813627571443 5320]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 68\n",
      "...\n",
      "order of connectivity matrix = 2001\n",
      "number of cluster center = 28\n",
      "...\n",
      "order of connectivity matrix = 4001\n",
      "number of cluster center = 26\n",
      "...\n",
      "order of connectivity matrix = 6001\n",
      "number of cluster center = 22\n",
      "...\n",
      "order of connectivity matrix = 8001\n",
      "number of cluster center = 22\n",
      "...\n",
      "order of connectivity matrix = 10001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 12001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 14001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 16001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 18001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 20001\n",
      "number of cluster center = 20\n",
      "...\n",
      "order of connectivity matrix = 22001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 24001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 26001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 28001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 30001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 32001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 34001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 36001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 38001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 40001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 42001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 44001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 46001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 48001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 50001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 52001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 54001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 56001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 58001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 60001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 62001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 64001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 66001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 68001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 70001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 72001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 74001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 76001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 78001\n",
      "number of cluster center = 18\n",
      "...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-24-9eb1fb91457a>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, result[3], ARI, result[4]])\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[array([ 0, 12, 12, 12, 12,  5, 12,  7, 12, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       17, 12, 12, 20, 12, 22, 12, 12, 25, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       34, 12, 12, 37, 12, 12, 12, 12, 12, 12, 12, 45, 12, 12, 12, 12, 12,\n",
      "       12, 52, 12, 12, 55, 12, 57, 12, 59, 12, 12, 62, 12, 64, 12, 66, 12])\n",
      " 80000 0.2211472213671079 0.11319813627571443 5314]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 68\n",
      "...\n",
      "order of connectivity matrix = 2001\n",
      "number of cluster center = 28\n",
      "...\n",
      "order of connectivity matrix = 4001\n",
      "number of cluster center = 26\n",
      "...\n",
      "order of connectivity matrix = 6001\n",
      "number of cluster center = 22\n",
      "...\n",
      "order of connectivity matrix = 8001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 10001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 12001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 14001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 16001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 18001\n",
      "number of cluster center = 20\n",
      "...\n",
      "order of connectivity matrix = 20001\n",
      "number of cluster center = 20\n",
      "...\n",
      "order of connectivity matrix = 22001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 24001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 26001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 28001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 30001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 32001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 34001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 36001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 38001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 40001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 42001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 44001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 46001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 48001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 50001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 52001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 54001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 56001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 58001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 60001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 62001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 64001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 66001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 68001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 70001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 72001\n",
      "number of cluster center = 18\n",
      "...\n",
      "less then 17 cluster center\n",
      "order 72380\n",
      "[0, 5, 7, 12, 17, 22, 25, 34, 37, 45, 52, 55, 57, 59, 62, 64, 66]\n",
      "[array([ 0, 12, 12, 12, 12,  5, 12,  7, 12, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       17, 12, 12, 12, 12, 22, 12, 12, 25, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       34, 12, 12, 37, 12, 12, 12, 12, 12, 12, 12, 45, 12, 12, 12, 12, 12,\n",
      "       12, 52, 12, 12, 55, 12, 57, 12, 59, 12, 12, 62, 12, 64, 12, 66, 12])\n",
      " 72380 0.2211472213671079 0.099088959708088 5254]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 68\n",
      "...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-24-9eb1fb91457a>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, result[3], ARI, result[4]])\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "order of connectivity matrix = 2001\n",
      "number of cluster center = 28\n",
      "...\n",
      "order of connectivity matrix = 4001\n",
      "number of cluster center = 26\n",
      "...\n",
      "order of connectivity matrix = 6001\n",
      "number of cluster center = 22\n",
      "...\n",
      "order of connectivity matrix = 8001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 10001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 12001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 14001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 16001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 18001\n",
      "number of cluster center = 20\n",
      "...\n",
      "order of connectivity matrix = 20001\n",
      "number of cluster center = 20\n",
      "...\n",
      "order of connectivity matrix = 22001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 24001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 26001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 28001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 30001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 32001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 34001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 36001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 38001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 40001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 42001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 44001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 46001\n",
      "number of cluster center = 18\n",
      "...\n",
      "order of connectivity matrix = 48001\n",
      "number of cluster center = 18\n",
      "...\n",
      "less then 17 cluster center\n",
      "order 49685\n",
      "[0, 5, 7, 12, 17, 22, 25, 34, 37, 45, 52, 55, 57, 59, 62, 64, 66]\n",
      "[array([ 0, 12, 12, 12, 12,  5, 12,  7, 12, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       17, 12, 12, 12, 12, 22, 12, 12, 25, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       34, 12, 12, 37, 12, 12, 12, 12, 12, 12, 12, 45, 12, 12, 12, 12, 12,\n",
      "       12, 52, 12, 12, 55, 12, 57, 12, 59, 12, 12, 62, 12, 64, 12, 66, 12])\n",
      " 49685 0.2211472213671079 0.099088959708088 5168]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 68\n",
      "...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-24-9eb1fb91457a>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, result[3], ARI, result[4]])\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "order of connectivity matrix = 2001\n",
      "number of cluster center = 28\n",
      "...\n",
      "order of connectivity matrix = 4001\n",
      "number of cluster center = 23\n",
      "...\n",
      "order of connectivity matrix = 6001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 8001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 10001\n",
      "number of cluster center = 20\n",
      "...\n",
      "order of connectivity matrix = 12001\n",
      "number of cluster center = 19\n",
      "...\n",
      "order of connectivity matrix = 14001\n",
      "number of cluster center = 18\n",
      "...\n",
      "less then 17 cluster center\n",
      "order 14061\n",
      "[0, 5, 7, 12, 17, 22, 25, 34, 37, 45, 52, 55, 57, 59, 62, 64, 66]\n",
      "[array([ 0, 12, 12, 12, 12,  5, 12,  7, 12, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       17, 12, 12, 12, 12, 22, 12, 12, 25, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       34, 12, 12, 37, 12, 12, 12, 12, 12, 12, 12, 45, 12, 12, 12, 12, 12,\n",
      "       12, 52, 12, 12, 55, 12, 57, 12, 59, 12, 12, 62, 12, 64, 12, 66, 12])\n",
      " 14061 0.20106113019337488 0.099088959708088 1398]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 68\n",
      "...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-24-9eb1fb91457a>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, result[3], ARI, result[4]])\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "order of connectivity matrix = 2001\n",
      "number of cluster center = 26\n",
      "...\n",
      "order of connectivity matrix = 4001\n",
      "number of cluster center = 21\n",
      "...\n",
      "order of connectivity matrix = 6001\n",
      "number of cluster center = 21\n",
      "...\n",
      "less then 17 cluster center\n",
      "order 7360\n",
      "[0, 5, 7, 12, 17, 22, 25, 34, 37, 45, 52, 55, 57, 59, 62, 64, 66]\n",
      "[array([ 0, 12, 12, 12, 12,  5, 12,  7, 12, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       17, 12, 12, 12, 12, 22, 12, 12, 25, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       34, 12, 12, 37, 12, 12, 12, 12, 12, 12, 12, 45, 12, 12, 12, 12, 12,\n",
      "       12, 52, 12, 12, 55, 12, 57, 12, 59, 12, 12, 62, 12, 64, 12, 66, 12])\n",
      " 7360 0.20106113019337488 0.099088959708088 1448]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-24-9eb1fb91457a>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, result[3], ARI, result[4]])\n"
     ]
    }
   ],
   "source": [
    "epsilon_list = [0, 1e-8, 5e-8, 1e-7, 5e-7, 1e-6, 5e-6, 1e-5]\n",
    "for epsilon in epsilon_list :\n",
    "    print(Experiment2(final_X, y, 0.7, epsilon=epsilon, max_it= 80000, check=2000, num_cluster=17))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "order of connectivity matrix = 1\n",
      "number of cluster center = 68\n",
      "...\n",
      "order of connectivity matrix = 2001\n",
      "number of cluster center = 28\n",
      "...\n",
      "order of connectivity matrix = 4001\n",
      "number of cluster center = 26\n",
      "...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-20-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[array([ 0, 12, 12, 12, 12,  5, 12,  7,  8, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       17, 12, 19, 20, 12, 22, 12, 12, 25, 12, 12, 12, 29, 12, 12, 12, 29,\n",
      "       34, 12, 12, 37, 12, 12, 40, 12, 12, 12, 12, 45, 29, 12, 12, 12, 12,\n",
      "       12, 52, 12, 12, 55, 12, 57, 29, 59, 12, 12, 62, 12, 64, 12, 66, 12])\n",
      " 5325 0.2211472213671079 0.5199063091324911 3.7688920497894287 0]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 68\n",
      "...\n",
      "order of connectivity matrix = 2001\n",
      "number of cluster center = 28\n",
      "...\n",
      "order of connectivity matrix = 4001\n",
      "number of cluster center = 26\n",
      "...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-20-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[array([ 0, 12, 12, 12, 12,  5, 12,  7,  8, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       17, 12, 19, 20, 12, 22, 12, 12, 25, 12, 12, 12, 29, 12, 12, 12, 29,\n",
      "       34, 12, 12, 37, 12, 12, 40, 12, 12, 12, 12, 45, 29, 12, 12, 12, 12,\n",
      "       12, 52, 12, 12, 55, 12, 57, 29, 59, 12, 12, 62, 12, 64, 12, 66, 12])\n",
      " 5324 0.2211472213671079 0.5199063091324911 3.7888176441192627 1e-08]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 68\n",
      "...\n",
      "order of connectivity matrix = 2001\n",
      "number of cluster center = 28\n",
      "...\n",
      "order of connectivity matrix = 4001\n",
      "number of cluster center = 26\n",
      "...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-20-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[array([ 0, 12, 12, 12, 12,  5, 12,  7,  8, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       17, 12, 19, 20, 12, 22, 12, 12, 25, 12, 12, 12, 29, 12, 12, 12, 29,\n",
      "       34, 12, 12, 37, 12, 12, 40, 12, 12, 12, 12, 45, 29, 12, 12, 12, 12,\n",
      "       12, 52, 12, 12, 55, 12, 57, 29, 59, 12, 12, 62, 12, 64, 12, 66, 12])\n",
      " 5321 0.2211472213671079 0.5199063091324911 3.843750476837158 5e-08]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 68\n",
      "...\n",
      "order of connectivity matrix = 2001\n",
      "number of cluster center = 28\n",
      "...\n",
      "order of connectivity matrix = 4001\n",
      "number of cluster center = 26\n",
      "...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-20-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[array([ 0, 12, 12, 12, 12,  5, 12,  7,  8, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       17, 12, 19, 20, 12, 22, 12, 12, 25, 12, 12, 12, 29, 12, 12, 12, 29,\n",
      "       34, 12, 12, 37, 12, 12, 40, 12, 12, 12, 12, 45, 29, 12, 12, 12, 12,\n",
      "       12, 52, 12, 12, 55, 12, 57, 29, 59, 12, 12, 62, 12, 64, 12, 66, 12])\n",
      " 5315 0.2211472213671079 0.5199063091324911 3.787389039993286 1e-07]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 68\n",
      "...\n",
      "order of connectivity matrix = 2001\n",
      "number of cluster center = 28\n",
      "...\n",
      "order of connectivity matrix = 4001\n",
      "number of cluster center = 26\n",
      "...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-20-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[array([ 0, 12, 12, 12, 12,  5, 12,  7,  8, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       17, 12, 19, 20, 12, 22, 12, 12, 25, 12, 12, 12, 29, 12, 12, 12, 29,\n",
      "       34, 12, 12, 37, 12, 12, 40, 12, 12, 12, 12, 45, 29, 12, 12, 12, 12,\n",
      "       12, 52, 12, 12, 55, 12, 57, 29, 59, 12, 12, 62, 12, 64, 12, 66, 12])\n",
      " 5256 0.2211472213671079 0.5199063091324911 3.7998874187469482 5e-07]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 68\n",
      "...\n",
      "order of connectivity matrix = 2001\n",
      "number of cluster center = 28\n",
      "...\n",
      "order of connectivity matrix = 4001\n",
      "number of cluster center = 26\n",
      "...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-20-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[array([ 0, 12, 12, 12, 12,  5, 12,  7,  8, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       17, 12, 19, 20, 12, 22, 12, 12, 25, 12, 12, 12, 29, 12, 12, 12, 29,\n",
      "       34, 12, 12, 37, 12, 12, 40, 12, 12, 12, 12, 45, 29, 12, 12, 12, 12,\n",
      "       12, 52, 12, 12, 55, 12, 57, 29, 59, 12, 12, 62, 12, 64, 12, 66, 12])\n",
      " 5169 0.2211472213671079 0.5199063091324911 3.7087056636810303 1e-06]\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 68\n",
      "...\n",
      "[array([ 0, 12,  2, 12, 12,  5,  6,  7,  8, 12, 12, 12, 12, 12, 12, 12, 12,\n",
      "       17, 12, 19, 20, 12, 22, 12, 12, 25, 12, 12, 12, 29, 12, 12, 12, 29,\n",
      "       34, 12, 36, 37, 12, 12, 40, 12, 12, 12, 12, 45, 29, 12, 12, 49, 50,\n",
      "       12, 52, 12, 12, 55, 56, 57, 29, 59, 12, 12, 62, 12, 64, 12, 66, 12])\n",
      " 1449 0.20106113019337488 0.5217114372931964 1.161499261856079 5e-06]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-20-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n"
     ]
    }
   ],
   "source": [
    "epsilon_list = [(0, 5325), (1e-8, 5324), (5e-8, 5321), (1e-7, 5315), (5e-7, 5256), (1e-6, 5169), (5e-6, 1449)]\n",
    "for epsilon in epsilon_list :\n",
    "    print(Experiment(final_X, y, 0.7, epsilon=epsilon[0], max_it= epsilon[1], check=2000, num_cluster=17))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [],
   "source": [
    "## mnist_fashion dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "mnist_train = pd.read_csv('./real_data/archive/fashion-mnist_train.csv')\n",
    "mnist_test = pd.read_csv('./real_data/archive/fashion-mnist_test.csv') "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [],
   "source": [
    "mnist_data = pd.concat([mnist_train, mnist_test], axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(70000, 785)"
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mnist_data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 350,
   "metadata": {},
   "outputs": [],
   "source": [
    "def sampler(df, t_label, sample_n, r_seed = 1):\n",
    "    random.seed(r_seed)\n",
    "    f_df = pd.DataFrame()\n",
    "    for label in t_label :\n",
    "        t_df = df[df.iloc[:, 0] == label].reset_index(drop=True)\n",
    "        t_df = t_df.iloc[random.sample(range(len(t_df)), sample_n), :].reset_index(drop=True)\n",
    "        f_df = pd.concat([f_df, t_df], axis=0).reset_index(drop=True)\n",
    "    return f_df    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 389,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "order of connectivity matrix = 1\n",
      "number of cluster center = 200\n",
      "...\n",
      "less then 2 cluster center\n",
      "order 37\n",
      "[27, 68]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-20-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([array([27, 27, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
       "       68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68,\n",
       "       68, 68, 68, 68, 68, 27, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
       "       68, 68, 27, 68, 68, 68, 68, 27, 68, 68, 68, 68, 27, 68, 68, 68, 68,\n",
       "       68, 27, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
       "       68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 27, 27,\n",
       "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
       "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
       "       27, 27, 27, 27, 27, 27, 27, 68, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
       "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
       "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
       "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27]),\n",
       "       37, 0.7556871524517946, 0.6891908356320745, 0.048868656158447266,\n",
       "       0], dtype=object)"
      ]
     },
     "execution_count": 389,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mnist_filterd = sampler(mnist_data, [0, 5], 100, 1)\n",
    "X = mnist_filterd.iloc[:,1:]\n",
    "y = mnist_filterd.iloc[:,0] \n",
    "X_mean = X.values.mean()\n",
    "X_std = X.values.std()\n",
    "X_scaled = (X.values - X_mean)/X_std\n",
    "Experiment(X_scaled, y, 10, epsilon=0, max_it= 10000, check=1000, num_cluster=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 390,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 200\n",
      "...\n",
      "order of connectivity matrix = 11\n",
      "number of cluster center = 10\n",
      "...\n",
      "order of connectivity matrix = 21\n",
      "number of cluster center = 6\n",
      "...\n",
      "order of connectivity matrix = 31\n",
      "number of cluster center = 5\n",
      "...\n",
      "less then 2 cluster center\n",
      "order 37\n",
      "[27, 68]\n",
      "[array([27, 27, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 27, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 27, 68, 68, 68, 68, 27, 68, 68, 68, 68, 27, 68, 68, 68, 68,\n",
      "       68, 27, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 68, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27])\n",
      " 37 0.7556871524517946 0.6891908356320745 0.050835609436035156 0]\n",
      "1e-08\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 200\n",
      "...\n",
      "order of connectivity matrix = 11\n",
      "number of cluster center = 10\n",
      "...\n",
      "order of connectivity matrix = 21\n",
      "number of cluster center = 6\n",
      "...\n",
      "order of connectivity matrix = 31\n",
      "number of cluster center = 5\n",
      "...\n",
      "less then 2 cluster center\n",
      "order 37\n",
      "[27, 68]\n",
      "[array([27, 27, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 27, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 27, 68, 68, 68, 68, 27, 68, 68, 68, 68, 27, 68, 68, 68, 68,\n",
      "       68, 27, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 68, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27])\n",
      " 37 0.7556871524517946 0.6891908356320745 0.05086398124694824 1e-08]\n",
      "5e-08\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 200\n",
      "...\n",
      "order of connectivity matrix = 11\n",
      "number of cluster center = 10\n",
      "...\n",
      "order of connectivity matrix = 21\n",
      "number of cluster center = 6\n",
      "...\n",
      "order of connectivity matrix = 31\n",
      "number of cluster center = 5\n",
      "...\n",
      "less then 2 cluster center\n",
      "order 37\n",
      "[27, 68]\n",
      "[array([27, 27, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 27, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 27, 68, 68, 68, 68, 27, 68, 68, 68, 68, 27, 68, 68, 68, 68,\n",
      "       68, 27, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 68, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27])\n",
      " 37 0.7556871524517946 0.6891908356320745 0.04488015174865723 5e-08]\n",
      "1e-07\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 200\n",
      "...\n",
      "order of connectivity matrix = 11\n",
      "number of cluster center = 10\n",
      "...\n",
      "order of connectivity matrix = 21\n",
      "number of cluster center = 6\n",
      "...\n",
      "order of connectivity matrix = 31\n",
      "number of cluster center = 5\n",
      "...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-20-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n",
      "<ipython-input-20-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n",
      "<ipython-input-20-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n",
      "<ipython-input-20-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "less then 2 cluster center\n",
      "order 37\n",
      "[27, 68]\n",
      "[array([27, 27, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 27, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 27, 68, 68, 68, 68, 27, 68, 68, 68, 68, 27, 68, 68, 68, 68,\n",
      "       68, 27, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 68, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27])\n",
      " 37 0.7556871524517946 0.6891908356320745 0.04886937141418457 1e-07]\n",
      "5e-07\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 200\n",
      "...\n",
      "order of connectivity matrix = 11\n",
      "number of cluster center = 10\n",
      "...\n",
      "order of connectivity matrix = 21\n",
      "number of cluster center = 6\n",
      "...\n",
      "order of connectivity matrix = 31\n",
      "number of cluster center = 5\n",
      "...\n",
      "less then 2 cluster center\n",
      "order 37\n",
      "[27, 68]\n",
      "[array([27, 27, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 27, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 27, 68, 68, 68, 68, 27, 68, 68, 68, 68, 27, 68, 68, 68, 68,\n",
      "       68, 27, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 68, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27])\n",
      " 37 0.7556871524517946 0.6891908356320745 0.05186152458190918 5e-07]\n",
      "1e-06\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 200\n",
      "...\n",
      "order of connectivity matrix = 11\n",
      "number of cluster center = 10\n",
      "...\n",
      "order of connectivity matrix = 21\n",
      "number of cluster center = 6\n",
      "...\n",
      "order of connectivity matrix = 31\n",
      "number of cluster center = 5\n",
      "...\n",
      "less then 2 cluster center\n",
      "order 37\n",
      "[27, 68]\n",
      "[array([27, 27, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 27, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 27, 68, 68, 68, 68, 27, 68, 68, 68, 68, 27, 68, 68, 68, 68,\n",
      "       68, 27, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 68, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27])\n",
      " 37 0.7556871524517946 0.6891908356320745 0.04886960983276367 1e-06]\n",
      "5e-06\n",
      "order of connectivity matrix = 1\n",
      "number of cluster center = 200\n",
      "...\n",
      "order of connectivity matrix = 11\n",
      "number of cluster center = 10\n",
      "...\n",
      "order of connectivity matrix = 21\n",
      "number of cluster center = 6\n",
      "...\n",
      "order of connectivity matrix = 31\n",
      "number of cluster center = 5\n",
      "...\n",
      "less then 2 cluster center\n",
      "order 37\n",
      "[27, 68]\n",
      "[array([27, 27, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 27, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 27, 68, 68, 68, 68, 27, 68, 68, 68, 68, 27, 68, 68, 68, 68,\n",
      "       68, 27, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 68, 68, 68, 68,\n",
      "       68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 27, 68, 68, 68, 68, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 68, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,\n",
      "       27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27])\n",
      " 37 0.7556871524517946 0.6891908356320745 0.045877695083618164 5e-06]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-20-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n",
      "<ipython-input-20-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n",
      "<ipython-input-20-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n"
     ]
    }
   ],
   "source": [
    "epsilon_list = [0, 1e-8, 5e-8, 1e-7, 5e-7, 1e-6, 5e-6, 1e-5, 5e-5, ]\n",
    "for epsilon in epsilon_list :\n",
    "    print(epsilon)\n",
    "    print(Experiment(X_scaled, y, 10, epsilon=epsilon, max_it= 10000, check=10, num_cluster=2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 386,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "order of connectivity matrix = 1\n",
      "number of cluster center = 2000\n",
      "...\n",
      "less then 2 cluster center\n",
      "order 9\n",
      "[590, 793]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-20-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([array([590, 590, 590, ..., 590, 590, 590]), 9, 0.5488697674649922,\n",
       "       0.5408884316509882, 1.2467870712280273, 0], dtype=object)"
      ]
     },
     "execution_count": 386,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mnist_filterd = sampler(mnist_data, [0, 5], 1000, 1)\n",
    "X = mnist_filterd.iloc[:,1:]\n",
    "y = mnist_filterd.iloc[:,0] \n",
    "X_mean = X.values.mean()\n",
    "X_std = X.values.std()\n",
    "X_scaled = (X.values - X_mean)/X_std\n",
    "Experiment(X_scaled, y, 13, epsilon=0, max_it= 10000, check=10, num_cluster=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 325,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<PIL.Image.Image image mode=RGB size=28x28 at 0x26E08AF3760>"
      ]
     },
     "execution_count": 325,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "image = Image.fromarray(np.array(mnist_filterd.iloc[107,1:]).reshape((28,28)).astype(np.uint8))\n",
    "image.convert('RGB')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 253,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<PIL.Image.Image image mode=L size=28x28 at 0x26E061C4C10>"
      ]
     },
     "execution_count": 253,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "instance = 10\n",
    "print(mnist_train.iloc[instance,0])\n",
    "Image.fromarray(np.array(mnist_train.iloc[instance,1:]).reshape((28,28)).astype(np.uint8))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.decomposition import PCA\n",
    "pca = PCA(n_components=2)\n",
    "principalComponents = pca.fit_transform(X_scaled)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 255,
   "metadata": {},
   "outputs": [],
   "source": [
    "principalDf = pd.DataFrame(data = principalComponents\n",
    "             , columns = ['principal component 1', 'principal component 2'])\n",
    "\n",
    "principalDf['y'] = y.reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 256,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='principal component 1', ylabel='principal component 2'>"
      ]
     },
     "execution_count": 256,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.scatterplot(data=principalDf, x='principal component 1', y='principal component 2', hue='y')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 1025,
   "metadata": {},
   "outputs": [],
   "source": [
    "sponge_df = pd.read_csv('https://archive.ics.uci.edu/ml/machine-learning-databases/sponge/sponge.data', header=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1112,
   "metadata": {},
   "outputs": [],
   "source": [
    "def preprocessing_dataset(data, numeric_col_list):\n",
    "    scaler = MinMaxScaler()\n",
    "\n",
    "    data = data.replace({'?' : 0})\n",
    "    data = data.dropna(axis=0)\n",
    "    \n",
    "    X = data.iloc[:,:-1]\n",
    "    y = data.iloc[:,-1]    \n",
    "    \n",
    "    numeric_data = X.iloc[:,numeric_col_list]\n",
    "    \n",
    "    cat_data = X.iloc[:,[i not in numeric_col_list for i in range(X.shape[1])]]\n",
    "    cat_data_ = pd.get_dummies(cat_data)\n",
    "    \n",
    "    final_df = pd.concat([cat_data_, numeric_data], axis=1)\n",
    "    final_df = final_df.astype(float)\n",
    "    final_X = scaler.fit_transform(final_df)\n",
    "    \n",
    "    return(final_X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1123,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='principal component 1', ylabel='principal component 2'>"
      ]
     },
     "execution_count": 1123,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_prep, y = preprocessing_dataset(sponge_df, [7,28,37])\n",
    "\n",
    "\n",
    "pca = PCA(n_components=2)\n",
    "principalComponents = pca.fit_transform(X_prep)\n",
    "principalDf = pd.DataFrame(data = principalComponents\n",
    "             , columns = ['principal component 1', 'principal component 2'])\n",
    "principalDf['y'] = y.reset_index(drop=True)\n",
    "sns.scatterplot(data=principalDf, x='principal component 1', y='principal component 2', hue='y')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1114,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       DURO\n",
       "1       DURO\n",
       "2       DURO\n",
       "3       DURO\n",
       "4       DURO\n",
       "       ...  \n",
       "71    BLANDO\n",
       "72    BLANDO\n",
       "73      DURO\n",
       "74      DURO\n",
       "75      DURO\n",
       "Name: 45, Length: 76, dtype: object"
      ]
     },
     "execution_count": 1114,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1115,
   "metadata": {},
   "outputs": [],
   "source": [
    "#########cylinder band"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1116,
   "metadata": {},
   "outputs": [],
   "source": [
    "cylinder_band_df = pd.read_csv('https://archive.ics.uci.edu/ml/machine-learning-databases/cylinder-bands/bands.data',\n",
    "                               header=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1103,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 541 entries, 0 to 540\n",
      "Data columns (total 40 columns):\n",
      " #   Column  Non-Null Count  Dtype  \n",
      "---  ------  --------------  -----  \n",
      " 0   0       541 non-null    object \n",
      " 1   1       541 non-null    object \n",
      " 2   2       541 non-null    object \n",
      " 3   3       541 non-null    object \n",
      " 4   4       541 non-null    object \n",
      " 5   5       541 non-null    object \n",
      " 6   6       541 non-null    object \n",
      " 7   7       541 non-null    object \n",
      " 8   8       541 non-null    object \n",
      " 9   9       541 non-null    object \n",
      " 10  10      541 non-null    object \n",
      " 11  11      541 non-null    object \n",
      " 12  12      541 non-null    object \n",
      " 13  13      541 non-null    object \n",
      " 14  14      541 non-null    object \n",
      " 15  15      541 non-null    object \n",
      " 16  16      541 non-null    float64\n",
      " 17  17      540 non-null    object \n",
      " 18  18      540 non-null    object \n",
      " 19  19      540 non-null    object \n",
      " 20  20      540 non-null    object \n",
      " 21  21      540 non-null    object \n",
      " 22  22      540 non-null    object \n",
      " 23  23      539 non-null    object \n",
      " 24  24      539 non-null    object \n",
      " 25  25      539 non-null    object \n",
      " 26  26      539 non-null    object \n",
      " 27  27      539 non-null    object \n",
      " 28  28      539 non-null    object \n",
      " 29  29      539 non-null    object \n",
      " 30  30      539 non-null    object \n",
      " 31  31      539 non-null    object \n",
      " 32  32      539 non-null    object \n",
      " 33  33      539 non-null    object \n",
      " 34  34      539 non-null    object \n",
      " 35  35      539 non-null    object \n",
      " 36  36      539 non-null    object \n",
      " 37  37      539 non-null    object \n",
      " 38  38      539 non-null    object \n",
      " 39  39      539 non-null    object \n",
      "dtypes: float64(1), object(39)\n",
      "memory usage: 169.2+ KB\n"
     ]
    }
   ],
   "source": [
    "cylinder_band_df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1104,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "noband    312\n",
       "band      227\n",
       "Name: 39, dtype: int64"
      ]
     },
     "execution_count": 1104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cylinder_band_df[39].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1105,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38]"
      ]
     },
     "execution_count": 1105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list(np.arange(20, 39))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1106,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.        , 0.        , 0.        , ..., 0.75      , 1.        ,\n",
       "        0.        ],\n",
       "       [0.        , 1.        , 0.        , ..., 0.        , 0.33333333,\n",
       "        0.        ],\n",
       "       [0.        , 0.        , 1.        , ..., 0.        , 0.33333333,\n",
       "        0.25      ],\n",
       "       ...,\n",
       "       [0.        , 0.        , 0.        , ..., 0.75      , 0.66666667,\n",
       "        0.        ],\n",
       "       [0.        , 0.        , 0.        , ..., 0.75      , 0.33333333,\n",
       "        0.5       ],\n",
       "       [0.        , 0.        , 0.        , ..., 0.5       , 0.33333333,\n",
       "        0.5       ]])"
      ]
     },
     "execution_count": 1106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_prep"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1108,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>29</th>\n",
       "      <th>30</th>\n",
       "      <th>31</th>\n",
       "      <th>32</th>\n",
       "      <th>33</th>\n",
       "      <th>34</th>\n",
       "      <th>35</th>\n",
       "      <th>36</th>\n",
       "      <th>37</th>\n",
       "      <th>38</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>19910108</td>\n",
       "      <td>X126</td>\n",
       "      <td>TVGUIDE</td>\n",
       "      <td>25503</td>\n",
       "      <td>YES</td>\n",
       "      <td>KEY</td>\n",
       "      <td>YES</td>\n",
       "      <td>BENTON</td>\n",
       "      <td>GALLATIN</td>\n",
       "      <td>UNCOATED</td>\n",
       "      <td>...</td>\n",
       "      <td>50.5</td>\n",
       "      <td>36.4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.5</td>\n",
       "      <td>1</td>\n",
       "      <td>34</td>\n",
       "      <td>40</td>\n",
       "      <td>105</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>19910109</td>\n",
       "      <td>X266</td>\n",
       "      <td>TVGUIDE</td>\n",
       "      <td>25503</td>\n",
       "      <td>YES</td>\n",
       "      <td>KEY</td>\n",
       "      <td>YES</td>\n",
       "      <td>BENTON</td>\n",
       "      <td>GALLATIN</td>\n",
       "      <td>UNCOATED</td>\n",
       "      <td>...</td>\n",
       "      <td>54.9</td>\n",
       "      <td>38.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.5</td>\n",
       "      <td>0.7</td>\n",
       "      <td>34</td>\n",
       "      <td>40</td>\n",
       "      <td>105</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>19910104</td>\n",
       "      <td>B7</td>\n",
       "      <td>MODMAT</td>\n",
       "      <td>47201</td>\n",
       "      <td>YES</td>\n",
       "      <td>KEY</td>\n",
       "      <td>YES</td>\n",
       "      <td>BENTON</td>\n",
       "      <td>GALLATIN</td>\n",
       "      <td>UNCOATED</td>\n",
       "      <td>...</td>\n",
       "      <td>53.8</td>\n",
       "      <td>39.8</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.8</td>\n",
       "      <td>0.9</td>\n",
       "      <td>40</td>\n",
       "      <td>40</td>\n",
       "      <td>103.87</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>19910104</td>\n",
       "      <td>T133</td>\n",
       "      <td>MASSEY</td>\n",
       "      <td>39039</td>\n",
       "      <td>YES</td>\n",
       "      <td>KEY</td>\n",
       "      <td>YES</td>\n",
       "      <td>BENTON</td>\n",
       "      <td>GALLATIN</td>\n",
       "      <td>UNCOATED</td>\n",
       "      <td>...</td>\n",
       "      <td>55.6</td>\n",
       "      <td>38.8</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.5</td>\n",
       "      <td>1.3</td>\n",
       "      <td>40</td>\n",
       "      <td>40</td>\n",
       "      <td>108.06</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>19910111</td>\n",
       "      <td>J34</td>\n",
       "      <td>KMART</td>\n",
       "      <td>37351</td>\n",
       "      <td>NO</td>\n",
       "      <td>KEY</td>\n",
       "      <td>YES</td>\n",
       "      <td>BENTON</td>\n",
       "      <td>GALLATIN</td>\n",
       "      <td>UNCOATED</td>\n",
       "      <td>...</td>\n",
       "      <td>57.5</td>\n",
       "      <td>42.5</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>2.3</td>\n",
       "      <td>0.6</td>\n",
       "      <td>35</td>\n",
       "      <td>40</td>\n",
       "      <td>106.67</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>536</th>\n",
       "      <td>19941005</td>\n",
       "      <td>aa66</td>\n",
       "      <td>kmart</td>\n",
       "      <td>85813</td>\n",
       "      <td>?</td>\n",
       "      <td>key</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>gallatin</td>\n",
       "      <td>super</td>\n",
       "      <td>...</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>?</td>\n",
       "      <td>40</td>\n",
       "      <td>112.5</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>537</th>\n",
       "      <td>19941009</td>\n",
       "      <td>j44</td>\n",
       "      <td>best</td>\n",
       "      <td>38064</td>\n",
       "      <td>?</td>\n",
       "      <td>key</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>gallatin</td>\n",
       "      <td>super</td>\n",
       "      <td>...</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>?</td>\n",
       "      <td>40</td>\n",
       "      <td>110</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>538</th>\n",
       "      <td>19941009</td>\n",
       "      <td>aa58</td>\n",
       "      <td>kmart</td>\n",
       "      <td>85814</td>\n",
       "      <td>?</td>\n",
       "      <td>key</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>gallatin</td>\n",
       "      <td>super</td>\n",
       "      <td>...</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>2.7</td>\n",
       "      <td>2.8</td>\n",
       "      <td>?</td>\n",
       "      <td>40</td>\n",
       "      <td>108</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>539</th>\n",
       "      <td>19941010</td>\n",
       "      <td>aa70</td>\n",
       "      <td>kmart</td>\n",
       "      <td>85814</td>\n",
       "      <td>?</td>\n",
       "      <td>key</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>gallatin</td>\n",
       "      <td>super</td>\n",
       "      <td>...</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>1.5</td>\n",
       "      <td>2.3</td>\n",
       "      <td>?</td>\n",
       "      <td>40</td>\n",
       "      <td>108</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>540</th>\n",
       "      <td>19941010</td>\n",
       "      <td>j70</td>\n",
       "      <td>best</td>\n",
       "      <td>38064</td>\n",
       "      <td>?</td>\n",
       "      <td>key</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>gallatin</td>\n",
       "      <td>super</td>\n",
       "      <td>...</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>?</td>\n",
       "      <td>2.5</td>\n",
       "      <td>1</td>\n",
       "      <td>?</td>\n",
       "      <td>40</td>\n",
       "      <td>108.1</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>541 rows × 39 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           0     1        2      3    4    5    6       7         8   \\\n",
       "0    19910108  X126  TVGUIDE  25503  YES  KEY  YES  BENTON  GALLATIN   \n",
       "1    19910109  X266  TVGUIDE  25503  YES  KEY  YES  BENTON  GALLATIN   \n",
       "2    19910104    B7   MODMAT  47201  YES  KEY  YES  BENTON  GALLATIN   \n",
       "3    19910104  T133   MASSEY  39039  YES  KEY  YES  BENTON  GALLATIN   \n",
       "4    19910111   J34    KMART  37351   NO  KEY  YES  BENTON  GALLATIN   \n",
       "..        ...   ...      ...    ...  ...  ...  ...     ...       ...   \n",
       "536  19941005  aa66    kmart  85813    ?  key    ?       ?  gallatin   \n",
       "537  19941009   j44     best  38064    ?  key    ?       ?  gallatin   \n",
       "538  19941009  aa58    kmart  85814    ?  key    ?       ?  gallatin   \n",
       "539  19941010  aa70    kmart  85814    ?  key    ?       ?  gallatin   \n",
       "540  19941010   j70     best  38064    ?  key    ?       ?  gallatin   \n",
       "\n",
       "           9   ...    29    30 31 32   33   34  35  36      37   38  \n",
       "0    UNCOATED  ...  50.5  36.4  0  0  2.5    1  34  40     105  100  \n",
       "1    UNCOATED  ...  54.9  38.5  0  0  2.5  0.7  34  40     105  100  \n",
       "2    UNCOATED  ...  53.8  39.8  0  0  2.8  0.9  40  40  103.87  100  \n",
       "3    UNCOATED  ...  55.6  38.8  0  0  2.5  1.3  40  40  108.06  100  \n",
       "4    UNCOATED  ...  57.5  42.5  5  0  2.3  0.6  35  40  106.67  100  \n",
       "..        ...  ...   ...   ... .. ..  ...  ...  ..  ..     ...  ...  \n",
       "536     super  ...     ?     ?  ?  ?    1    1   ?  40   112.5  100  \n",
       "537     super  ...     ?     ?  ?  ?    0    0   ?  40     110  100  \n",
       "538     super  ...     ?     ?  ?  ?  2.7  2.8   ?  40     108  100  \n",
       "539     super  ...     ?     ?  ?  ?  1.5  2.3   ?  40     108  100  \n",
       "540     super  ...     ?     ?  ?  ?  2.5    1   ?  40   108.1  100  \n",
       "\n",
       "[541 rows x 39 columns]"
      ]
     },
     "execution_count": 1108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1117,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='principal component 1', ylabel='principal component 2'>"
      ]
     },
     "execution_count": 1117,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "X = cylinder_band_df.iloc[:,:-1]\n",
    "y = cylinder_band_df.iloc[:,-1]\n",
    "#y = np.array([1 if i == 'band' else 0 for i in list(y)])\n",
    "\n",
    "X_prep, y = preprocessing_dataset(cylinder_band_df, list(np.arange(20, 39)))\n",
    "\n",
    "pca = PCA(n_components=2)\n",
    "principalComponents = pca.fit_transform(X_prep)\n",
    "principalDf = pd.DataFrame(data = principalComponents\n",
    "             , columns = ['principal component 1', 'principal component 2'])\n",
    "principalDf['y'] = y\n",
    "sns.scatterplot(data=principalDf, x='principal component 1', y='principal component 2', hue='y')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1120,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_ = [1 if i == 'band' else 0 for i in list(y)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1122,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "order of connectivity matrix = 1\n",
      "number of cluster center = 539\n",
      "...\n",
      "one cluster center\n",
      "order 5\n",
      "[96]\n",
      "[array([96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,\n",
      "       96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96])\n",
      " 5 0.0 2.6097543071848243e-15 0.07777261734008789 0]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-574-55ba005d6ae0>:18: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
      "  return np.array([result[2], iteration, ARI, NMI, runtime, epsilon])\n"
     ]
    }
   ],
   "source": [
    "print(Experiment(X_prep, y_, 2, epsilon=0, max_it= 1000, check=100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 754,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "ename": "ParserError",
     "evalue": "Error tokenizing data. C error: Expected 2 fields in line 18, saw 3\n",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mParserError\u001b[0m                               Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-754-a831561f12a3>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0msponge_info\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mpd\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mread_csv\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'https://archive.ics.uci.edu/ml/machine-learning-databases/sponge/sponge.info'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\pandas\\io\\parsers.py\u001b[0m in \u001b[0;36mread_csv\u001b[1;34m(filepath_or_buffer, sep, delimiter, header, names, index_col, usecols, squeeze, prefix, mangle_dupe_cols, dtype, engine, converters, true_values, false_values, skipinitialspace, skiprows, skipfooter, nrows, na_values, keep_default_na, na_filter, verbose, skip_blank_lines, parse_dates, infer_datetime_format, keep_date_col, date_parser, dayfirst, cache_dates, iterator, chunksize, compression, thousands, decimal, lineterminator, quotechar, quoting, doublequote, escapechar, comment, encoding, dialect, error_bad_lines, warn_bad_lines, delim_whitespace, low_memory, memory_map, float_precision)\u001b[0m\n\u001b[0;32m    684\u001b[0m     )\n\u001b[0;32m    685\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 686\u001b[1;33m     \u001b[1;32mreturn\u001b[0m \u001b[0m_read\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfilepath_or_buffer\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkwds\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    687\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    688\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\pandas\\io\\parsers.py\u001b[0m in \u001b[0;36m_read\u001b[1;34m(filepath_or_buffer, kwds)\u001b[0m\n\u001b[0;32m    456\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    457\u001b[0m     \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 458\u001b[1;33m         \u001b[0mdata\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mparser\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mread\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnrows\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    459\u001b[0m     \u001b[1;32mfinally\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    460\u001b[0m         \u001b[0mparser\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mclose\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\pandas\\io\\parsers.py\u001b[0m in \u001b[0;36mread\u001b[1;34m(self, nrows)\u001b[0m\n\u001b[0;32m   1194\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mread\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnrows\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mNone\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1195\u001b[0m         \u001b[0mnrows\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0m_validate_integer\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m\"nrows\"\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnrows\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1196\u001b[1;33m         \u001b[0mret\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_engine\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mread\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnrows\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1197\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1198\u001b[0m         \u001b[1;31m# May alter columns / col_dict\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\pandas\\io\\parsers.py\u001b[0m in \u001b[0;36mread\u001b[1;34m(self, nrows)\u001b[0m\n\u001b[0;32m   2153\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mread\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnrows\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mNone\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   2154\u001b[0m         \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 2155\u001b[1;33m             \u001b[0mdata\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_reader\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mread\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnrows\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   2156\u001b[0m         \u001b[1;32mexcept\u001b[0m \u001b[0mStopIteration\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   2157\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_first_chunk\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32mpandas\\_libs\\parsers.pyx\u001b[0m in \u001b[0;36mpandas._libs.parsers.TextReader.read\u001b[1;34m()\u001b[0m\n",
      "\u001b[1;32mpandas\\_libs\\parsers.pyx\u001b[0m in \u001b[0;36mpandas._libs.parsers.TextReader._read_low_memory\u001b[1;34m()\u001b[0m\n",
      "\u001b[1;32mpandas\\_libs\\parsers.pyx\u001b[0m in \u001b[0;36mpandas._libs.parsers.TextReader._read_rows\u001b[1;34m()\u001b[0m\n",
      "\u001b[1;32mpandas\\_libs\\parsers.pyx\u001b[0m in \u001b[0;36mpandas._libs.parsers.TextReader._tokenize_rows\u001b[1;34m()\u001b[0m\n",
      "\u001b[1;32mpandas\\_libs\\parsers.pyx\u001b[0m in \u001b[0;36mpandas._libs.parsers.raise_parser_error\u001b[1;34m()\u001b[0m\n",
      "\u001b[1;31mParserError\u001b[0m: Error tokenizing data. C error: Expected 2 fields in line 18, saw 3\n"
     ]
    }
   ],
   "source": [
    "sponge_info = pd.read_csv('https://archive.ics.uci.edu/ml/machine-learning-databases/sponge/sponge.info')"
   ]
  },
  {
   "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}