data_prep.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data download"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "from six.moves.urllib.request import urlretrieve\n",
    "\n",
    "import scipy.io\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from six.moves import cPickle as pickle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Attempting to download: 98.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\98.mat : 7742720\n",
      "\n",
      "Attempting to download: 99.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\99.mat : 15503928\n",
      "\n",
      "Attempting to download: 100.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\100.mat : 7770624\n",
      "\n",
      "Attempting to download: 119.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\119.mat : 2914248\n",
      "\n",
      "Attempting to download: 120.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\120.mat : 2917752\n",
      "\n",
      "Attempting to download: 121.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\121.mat : 2917752\n",
      "\n",
      "Attempting to download: 186.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\186.mat : 2931672\n",
      "\n",
      "Attempting to download: 187.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\187.mat : 2928192\n",
      "\n",
      "Attempting to download: 188.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\188.mat : 2931672\n",
      "\n",
      "Attempting to download: 223.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\223.mat : 2921232\n",
      "\n",
      "Attempting to download: 224.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\224.mat : 2931672\n",
      "\n",
      "Attempting to download: 225.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\225.mat : 2931672\n",
      "\n",
      "Attempting to download: 106.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\106.mat : 2928192\n",
      "\n",
      "Attempting to download: 107.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\107.mat : 2931672\n",
      "\n",
      "Attempting to download: 108.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\108.mat : 2950416\n",
      "\n",
      "Attempting to download: 170.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\170.mat : 2924712\n",
      "\n",
      "Attempting to download: 171.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\171.mat : 2924712\n",
      "\n",
      "Attempting to download: 172.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\172.mat : 2921232\n",
      "\n",
      "Attempting to download: 210.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\210.mat : 2917752\n",
      "\n",
      "Attempting to download: 211.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\211.mat : 2924712\n",
      "\n",
      "Attempting to download: 212.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\212.mat : 2928192\n",
      "\n",
      "Attempting to download: 131.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\131.mat : 2938632\n",
      "\n",
      "Attempting to download: 132.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\132.mat : 2914248\n",
      "\n",
      "Attempting to download: 133.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\133.mat : 2942112\n",
      "\n",
      "Attempting to download: 198.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\198.mat : 2931672\n",
      "\n",
      "Attempting to download: 199.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\199.mat : 2924712\n",
      "\n",
      "Attempting to download: 200.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\200.mat : 2928192\n",
      "\n",
      "Attempting to download: 235.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\235.mat : 2928192\n",
      "\n",
      "Attempting to download: 236.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\236.mat : 2935152\n",
      "\n",
      "Attempting to download: 237.mat\n",
      "\n",
      "Download Complete!\n",
      "./data/download\\237.mat : 2928192\n"
     ]
    }
   ],
   "source": [
    "url_base =  'http://csegroups.case.edu/sites/default/files/bearingdatacenter/files/Datafiles/'\n",
    "download_folder = r'./data/download'\n",
    "\n",
    "file_dictionary = {\n",
    "        \"1-A\" : \"98.mat\",\n",
    "        \"1-B\" : \"99.mat\",\n",
    "        \"1-C\" : \"100.mat\",\n",
    "        \"2-A\" : \"119.mat\",\n",
    "        \"2-B\" : \"120.mat\",\n",
    "        \"2-C\" : \"121.mat\",\n",
    "        \"3-A\" : \"186.mat\",\n",
    "        \"3-B\" : \"187.mat\",\n",
    "        \"3-C\" : \"188.mat\",\n",
    "        \"4-A\" : \"223.mat\",\n",
    "        \"4-B\" : \"224.mat\",\n",
    "        \"4-C\" : \"225.mat\",\n",
    "        \"5-A\" : \"106.mat\",\n",
    "        \"5-B\" : \"107.mat\",\n",
    "        \"5-C\" : \"108.mat\",\n",
    "        \"6-A\" : \"170.mat\",\n",
    "        \"6-B\" : \"171.mat\",\n",
    "        \"6-C\" : \"172.mat\",\n",
    "        \"7-A\" : \"210.mat\",\n",
    "        \"7-B\" : \"211.mat\",\n",
    "        \"7-C\" : \"212.mat\",\n",
    "        \"8-A\" : \"131.mat\",\n",
    "        \"8-B\" : \"132.mat\",\n",
    "        \"8-C\" : \"133.mat\",\n",
    "        \"9-A\" : \"198.mat\",\n",
    "        \"9-B\" : \"199.mat\",\n",
    "        \"9-C\" : \"200.mat\",\n",
    "        \"10-A\" : \"235.mat\",\n",
    "        \"10-B\" : \"236.mat\",\n",
    "        \"10-C\" : \"237.mat\"\n",
    "        }\n",
    "\n",
    "for key,filename in file_dictionary.items():\n",
    "    dest_filename = os.path.join(download_folder, filename)\n",
    "    if not os.path.exists(dest_filename):\n",
    "        print('\\nAttempting to download:', filename) \n",
    "        filename, _ = urlretrieve(url_base + filename, dest_filename)\n",
    "        print('\\nDownload Complete!')\n",
    "        statinfo = os.stat(dest_filename)\n",
    "        print(filename,':', statinfo.st_size)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Formatting arrays"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "Channels = ['DE','FE'] # Drive End , Fan End\n",
    "Vib_mc = {} # multi-channel\n",
    "\n",
    "for c in Channels:\n",
    "    Sen_series = {}\n",
    "    for key,filename in sorted(file_dictionary.items()):\n",
    "        mtfilename = os.path.join(download_folder, filename)\n",
    "        data = scipy.io.loadmat(mtfilename)\n",
    "        for i in data:\n",
    "            if  '_'+c+'_time' in i:\n",
    "                Sen_series[key] = data[i]\n",
    "    Vib_mc[c]= Sen_series"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "checking number of rows and statistics for each of 10 categories(1-10) and 3 HP (A,B,C) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1-A rows: 483903 avg: 0.03 min: -0.22 max: 0.32\n",
      "1-B rows: 485063 avg: 0.03 min: -0.24 max: 0.31\n",
      "1-C rows: 485643 avg: 0.03 min: -0.3 max: 0.38\n",
      "10-A rows: 121991 avg: 0.03 min: -1.66 max: 1.84\n",
      "10-B rows: 122281 avg: 0.03 min: -1.9 max: 1.98\n",
      "10-C rows: 121991 avg: 0.03 min: -1.86 max: 2.09\n",
      "2-A rows: 121410 avg: 0.03 min: -0.31 max: 0.38\n",
      "2-B rows: 121556 avg: 0.03 min: -0.33 max: 0.39\n",
      "2-C rows: 121556 avg: 0.03 min: -0.44 max: 0.48\n",
      "3-A rows: 122136 avg: 0.03 min: -0.31 max: 0.35\n",
      "3-B rows: 121991 avg: 0.03 min: -0.43 max: 0.47\n",
      "3-C rows: 122136 avg: 0.03 min: -0.64 max: 0.65\n",
      "4-A rows: 121701 avg: 0.03 min: -0.59 max: 0.6\n",
      "4-B rows: 122136 avg: 0.03 min: -0.41 max: 0.4\n",
      "4-C rows: 122136 avg: 0.03 min: -0.34 max: 0.41\n",
      "5-A rows: 121991 avg: 0.03 min: -1.15 max: 0.97\n",
      "5-B rows: 122136 avg: 0.03 min: -0.9 max: 0.96\n",
      "5-C rows: 122917 avg: 0.03 min: -1.08 max: 0.9\n",
      "6-A rows: 121846 avg: 0.03 min: -0.67 max: 0.63\n",
      "6-B rows: 121846 avg: 0.03 min: -0.68 max: 0.62\n",
      "6-C rows: 121701 avg: 0.03 min: -0.7 max: 0.65\n",
      "7-A rows: 121556 avg: 0.03 min: -0.86 max: 0.89\n",
      "7-B rows: 121846 avg: 0.03 min: -0.83 max: 1.01\n",
      "7-C rows: 121991 avg: 0.03 min: -0.84 max: 0.87\n",
      "8-A rows: 122426 avg: 0.03 min: -1.04 max: 1.07\n",
      "8-B rows: 121410 avg: 0.03 min: -0.95 max: 1.05\n",
      "8-C rows: 122571 avg: 0.03 min: -0.97 max: 1.23\n",
      "9-A rows: 122136 avg: 0.03 min: -0.38 max: 0.41\n",
      "9-B rows: 121846 avg: 0.03 min: -0.36 max: 0.42\n",
      "9-C rows: 121991 avg: 0.03 min: -0.43 max: 0.49\n",
      "minimum row: 121410\n"
     ]
    }
   ],
   "source": [
    "Channel = Channels[1]\n",
    "\n",
    "minrow = Vib_mc[Channel][list(Vib_mc[Channel].keys())[0]].shape[0]\n",
    "\n",
    "for key,array in sorted(Vib_mc[Channel].items()):\n",
    "    print(key, \"rows:\", array.shape[0]\n",
    "        , \"avg:\",round(np.mean(array),2)\n",
    "        , \"min:\",round(np.min(array),2)\n",
    "        , \"max:\",round(np.max(array),2))    \n",
    "    if array.shape[0] < minrow:\n",
    "        minrow = array.shape[0]\n",
    "\n",
    "print(\"minimum row:\",minrow )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "quick visual check"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sample = Vib_mc['FE']['2-A']\n",
    "plt.plot(sample[:128])\n",
    "plt.show()\n",
    "sample = Vib_mc['DE']['2-A']\n",
    "plt.plot(sample[:128])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We need to create data samples with fixed length (2048 data points) for the 2 accelometers (DE and FE).\n",
    "\n",
    "Let's take test samples without shift(augument) from the 1st half of the data from each file (categories x HP)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test_temp = {}\n",
    "sample_len = 2048\n",
    "sample_test_cnt = 25\n",
    "\n",
    "for key,sample in sorted(Vib_mc[Channels[0]].items()):\n",
    "    X_test_temp[key] = np.concatenate((Vib_mc[Channels[0]][key][:sample_len*sample_test_cnt].reshape(sample_test_cnt,sample_len,1)\n",
    "                            ,Vib_mc[Channels[1]][key][:sample_len*sample_test_cnt].reshape(sample_test_cnt,sample_len,1)\n",
    "                            ),axis=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, training data with overlap (augmentation). Shifting 103 data points."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_temp = {}\n",
    "sample_train_cnt = 660\n",
    "sample_train_shift = 103\n",
    "for key,sample in sorted(Vib_mc[Channels[0]].items()):\n",
    "    train_array = np.concatenate((Vib_mc[Channels[0]][key][sample_len*sample_test_cnt:sample_len*sample_test_cnt+sample_len].reshape(1,sample_len,1)\n",
    "                                 ,Vib_mc[Channels[1]][key][sample_len*sample_test_cnt:sample_len*sample_test_cnt+sample_len].reshape(1,sample_len,1)\n",
    "                                 ),axis=2)\n",
    "    for i in range(sample_train_cnt-1):\n",
    "        train_array = np.concatenate((train_array\n",
    "                                 ,np.concatenate((\n",
    "                                         Vib_mc[Channels[0]][key][sample_len*sample_test_cnt+sample_train_shift*(i+1):sample_len*sample_test_cnt+sample_len+sample_train_shift*(i+1)].reshape(1,sample_len,1)\n",
    "                                        ,Vib_mc[Channels[1]][key][sample_len*sample_test_cnt+sample_train_shift*(i+1):sample_len*sample_test_cnt+sample_len+sample_train_shift*(i+1)].reshape(1,sample_len,1)\n",
    "                                ),axis=2)),axis=0)\n",
    "    X_train_temp[key] = train_array"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create training and test dataset with randomization, both with X and Y(lable...at this point not one-hot encoded yet).\n",
    "\n",
    "For each, let user get A,B,C (HP) data by key. A, B, C contains 10 categories data for each HP. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def randomize(dataset, labels):\n",
    "  permutation = np.random.permutation(labels.shape[0])\n",
    "  shuffled_dataset = dataset[permutation,:]\n",
    "  shuffled_labels = labels[permutation]\n",
    "  return shuffled_dataset, shuffled_labels\n",
    "\n",
    "def create_dataset(X_temp):\n",
    "    X = {\"A\":np.empty((0,sample_len,2),float),\"B\":np.empty((0,sample_len,2),float),\"C\":np.empty((0,sample_len,2),float)}\n",
    "    Y = {\"A\":np.empty((0,1),int),\"B\":np.empty((0,1),int),\"C\":np.empty((0,1),int)}\n",
    "    for key,sample in sorted(X_temp.items()):\n",
    "        X[key.split(sep=\"-\")[1]]=np.concatenate((X[key.split(sep=\"-\")[1]],sample),axis=0)\n",
    "        y = np.empty(len(sample),int)\n",
    "        y.fill(key.split(sep=\"-\")[0])\n",
    "        y = y.reshape((len(sample),1))\n",
    "        Y[key.split(sep=\"-\")[1]]=np.concatenate((Y[key.split(sep=\"-\")[1]],y),axis=0)\n",
    "    #shuffling\n",
    "    for key,samplearray in sorted(X.items()):\n",
    "        X[key],Y[key] = randomize(samplearray,Y[key])\n",
    "    return X,Y\n",
    "\n",
    "X_test,Y_test = create_dataset(X_test_temp) \n",
    "X_train,Y_train = create_dataset(X_train_temp)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "dataset D contains all. Vertically stack A, B and C and shuffle."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test[\"D\"],Y_test[\"D\"] = randomize(np.vstack((X_test[\"A\"],X_test[\"B\"],X_test[\"C\"]))\n",
    "    ,np.vstack((Y_test[\"A\"],Y_test[\"B\"],Y_test[\"C\"])))\n",
    "\n",
    "X_train[\"D\"],Y_train[\"D\"] = randomize(np.vstack((X_train[\"A\"],X_train[\"B\"],X_train[\"C\"]))\n",
    "    ,np.vstack((Y_train[\"A\"],Y_train[\"B\"],Y_train[\"C\"])))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Saving the dataset as pickle format"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Compressed pickle size: 1346930161\n"
     ]
    }
   ],
   "source": [
    "\n",
    "pickle_file = os.path.join('./data/', 'ABCD_Datasets.pickle')\n",
    "\n",
    "try:\n",
    "  f = open(pickle_file, 'wb')\n",
    "  save = {\n",
    "    'train_datasets': X_train,\n",
    "    'train_labels': Y_train,\n",
    "    'test_datasets': X_test,\n",
    "    'test_labels': Y_test,\n",
    "    }\n",
    "  pickle.dump(save, f, pickle.HIGHEST_PROTOCOL)\n",
    "  f.close()\n",
    "except Exception as e:\n",
    "  print('Unable to save data to', pickle_file, ':', e)\n",
    "  raise\n",
    "\n",
    "statinfo = os.stat(pickle_file)\n",
    "print('Compressed pickle size:', statinfo.st_size)\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}