Step1_Preprocessing.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from skimage import measure\n",
    "import os\n",
    "import matplotlib.pyplot as plt\n",
    "import multiprocessing\n",
    "cpu_count = multiprocessing.cpu_count()\n",
    "from multiprocessing import Pool\n",
    "\n",
    "from matplotlib.patches import Rectangle\n",
    "import matplotlib.patches as patches"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Read data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "data_p01.pkl\n",
      "data_p02.pkl\n",
      "data_p03.pkl\n",
      "data_p04.pkl\n",
      "data_p05.pkl\n",
      "data_p06.pkl\n",
      "data_p07.pkl\n",
      "data_p08.pkl\n",
      "data_p09.pkl\n",
      "data_p10.pkl\n",
      "data_p11.pkl\n",
      "data_p12.pkl\n",
      "data_p13.pkl\n",
      "data_p14.pkl\n",
      "data_p15.pkl\n",
      "data_p16.pkl\n",
      "data_p17.pkl\n",
      "data_p18.pkl\n",
      "data_p19.pkl\n",
      "data_p20.pkl\n",
      "data_p21.pkl\n",
      "data_p22.pkl\n",
      "data_p23.pkl\n",
      "data_p24.pkl\n",
      "data_p25.pkl\n",
      "data_p26.pkl\n",
      "data_p27.pkl\n",
      "data_p28.pkl\n",
      "data_p29.pkl\n",
      "data_p30.pkl\n",
      "data_p31.pkl\n",
      "data_p32.pkl\n",
      "data_p33.pkl\n",
      "241120\n"
     ]
    }
   ],
   "source": [
    "directory = \"./data\"\n",
    "dfs =[]\n",
    "for filename in sorted(os.listdir(directory)):\n",
    "    if (filename.startswith(\"data_\") and filename.endswith(\".pkl\")): \n",
    "        print (filename)\n",
    "        dfs.append(pd.read_pickle(os.path.join(directory, filename)))\n",
    "    else:\n",
    "        continue\n",
    "        \n",
    "df = pd.concat(dfs)\n",
    "df = df.sort_index()\n",
    "print (len(df))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Participant</th>\n",
       "      <th>Condition</th>\n",
       "      <th>Matrix</th>\n",
       "      <th>Pitch</th>\n",
       "      <th>Yaw</th>\n",
       "      <th>TimeOffset</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>[[0, 0, -1, 1, 0, 0, -3, 0, -2, -2, -1, -3, -1...</td>\n",
       "      <td>38.856377</td>\n",
       "      <td>-2.493793</td>\n",
       "      <td>0.993750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>[[0, 0, -1, 1, 0, 0, -3, 0, -2, -2, -1, -3, -1...</td>\n",
       "      <td>38.597713</td>\n",
       "      <td>-2.538246</td>\n",
       "      <td>0.999065</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>[[0, 0, -1, 1, 0, 0, -3, 0, -2, -2, -1, -3, -1...</td>\n",
       "      <td>37.787099</td>\n",
       "      <td>-3.215579</td>\n",
       "      <td>-0.005882</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>[[0, 0, -1, 1, 0, 0, -3, 0, -2, -2, -1, -3, -1...</td>\n",
       "      <td>38.310998</td>\n",
       "      <td>-3.407330</td>\n",
       "      <td>-0.001333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>[[0, 0, -1, 1, 0, 0, -3, 0, -2, -2, -1, -3, -1...</td>\n",
       "      <td>38.392563</td>\n",
       "      <td>-3.486370</td>\n",
       "      <td>-0.001205</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Participant  Condition                                             Matrix  \\\n",
       "0            1          2  [[0, 0, -1, 1, 0, 0, -3, 0, -2, -2, -1, -3, -1...   \n",
       "1            1          2  [[0, 0, -1, 1, 0, 0, -3, 0, -2, -2, -1, -3, -1...   \n",
       "2            1          2  [[0, 0, -1, 1, 0, 0, -3, 0, -2, -2, -1, -3, -1...   \n",
       "3            1          2  [[0, 0, -1, 1, 0, 0, -3, 0, -2, -2, -1, -3, -1...   \n",
       "4            1          2  [[0, 0, -1, 1, 0, 0, -3, 0, -2, -2, -1, -3, -1...   \n",
       "\n",
       "       Pitch       Yaw  TimeOffset  \n",
       "0  38.856377 -2.493793    0.993750  \n",
       "1  38.597713 -2.538246    0.999065  \n",
       "2  37.787099 -3.215579   -0.005882  \n",
       "3  38.310998 -3.407330   -0.001333  \n",
       "4  38.392563 -3.486370   -0.001205  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Flip data in yaw"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "df[\"IsFlipped\"] = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 2.26 s, sys: 1.39 s, total: 3.65 s\n",
      "Wall time: 3.45 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "def doJob (df):\n",
    "    lst = []\n",
    "    for i, d in df.iterrows():\n",
    "        lst.append(np.fliplr(d.Matrix))\n",
    "    return lst\n",
    "\n",
    "p = Pool(cpu_count)\n",
    "\n",
    "\n",
    "data = np.array_split(df, cpu_count)\n",
    "pool_results = p.map(doJob, data)\n",
    "p.close()\n",
    "del data\n",
    "\n",
    "results = []\n",
    "for result in pool_results:\n",
    "    results.extend(result)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>Pair</th>\n",
       "      <th>Participant</th>\n",
       "      <th>Condition</th>\n",
       "      <th>Matrix</th>\n",
       "      <th>Pitch</th>\n",
       "      <th>Yaw</th>\n",
       "      <th>TimeOffset</th>\n",
       "      <th>IsFlipped</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>[[0, 0, -1, 1, 0, 0, -3, 0, -2, -2, -1, -3, -1...</td>\n",
       "      <td>38.856377</td>\n",
       "      <td>-2.493793</td>\n",
       "      <td>0.993750</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>[[-1, -2, -1, -3, -1, -2, -2, 0, -3, 0, 0, 1, ...</td>\n",
       "      <td>38.856377</td>\n",
       "      <td>2.493793</td>\n",
       "      <td>0.993750</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>[[0, 0, -1, 1, 0, 0, -3, 0, -2, -2, -1, -3, -1...</td>\n",
       "      <td>38.597713</td>\n",
       "      <td>-2.538246</td>\n",
       "      <td>0.999065</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>[[-1, -2, -1, -3, -1, -2, -2, 0, -3, 0, 0, 1, ...</td>\n",
       "      <td>38.597713</td>\n",
       "      <td>2.538246</td>\n",
       "      <td>0.999065</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>[[0, 0, -1, 1, 0, 0, -3, 0, -2, -2, -1, -3, -1...</td>\n",
       "      <td>37.787099</td>\n",
       "      <td>-3.215579</td>\n",
       "      <td>-0.005882</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Pair  Participant  Condition  \\\n",
       "0     0            1          2   \n",
       "1     0            1          2   \n",
       "2     1            1          2   \n",
       "3     1            1          2   \n",
       "4     2            1          2   \n",
       "\n",
       "                                              Matrix      Pitch       Yaw  \\\n",
       "0  [[0, 0, -1, 1, 0, 0, -3, 0, -2, -2, -1, -3, -1...  38.856377 -2.493793   \n",
       "1  [[-1, -2, -1, -3, -1, -2, -2, 0, -3, 0, 0, 1, ...  38.856377  2.493793   \n",
       "2  [[0, 0, -1, 1, 0, 0, -3, 0, -2, -2, -1, -3, -1...  38.597713 -2.538246   \n",
       "3  [[-1, -2, -1, -3, -1, -2, -2, 0, -3, 0, 0, 1, ...  38.597713  2.538246   \n",
       "4  [[0, 0, -1, 1, 0, 0, -3, 0, -2, -2, -1, -3, -1...  37.787099 -3.215579   \n",
       "\n",
       "   TimeOffset  IsFlipped  \n",
       "0    0.993750      False  \n",
       "1    0.993750       True  \n",
       "2    0.999065      False  \n",
       "3    0.999065       True  \n",
       "4   -0.005882      False  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2 = df.copy(deep=True)\n",
    "df2.Matrix = results\n",
    "df2.Yaw = df2.Yaw *-1\n",
    "df2.IsFlipped = True\n",
    "\n",
    "df = pd.concat([df,df2])\n",
    "df = df.sort_index()\n",
    "df = df.reset_index()\n",
    "df = df.rename(columns={\"index\":\"Pair\"})\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Blob detection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "def loadImg(img):\n",
    "    im = np.copy(img)\n",
    "    \n",
    "    # The Nexus5 capacitive sensor is extreamly noizy in the upper rows \n",
    "    # Since no one touched there we can remove the noise to be able\n",
    "    # to perfercly detect the finger\n",
    "    im[:7] = 0\n",
    "    \n",
    "    # Reduction of noise as already described by Xiao et al.\n",
    "    im[im < 4] = 0\n",
    "    \n",
    "    # Filter sensore overshoots\n",
    "    im[im > 255] = 255\n",
    "    \n",
    "    im = im.astype(np.float64)\n",
    "    return im"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "def findCon (img):\n",
    "    contours = measure.find_contours(img, 35)\n",
    "    points_to_approx = []\n",
    "    highest_val = 0\n",
    "    for n, contour in enumerate(contours):\n",
    "        if (len(contour) > highest_val and len(contour) > 0):\n",
    "            points_to_approx = contour\n",
    "            highest_val = len(contour)\n",
    "\n",
    "    if len(points_to_approx) > 5:\n",
    "        return points_to_approx[:, 1], points_to_approx[:, 0]\n",
    "    else:\n",
    "        return None, None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "def doJob (df):\n",
    "    lst = []\n",
    "    for i, d in df.iterrows():\n",
    "        img = loadImg(d.Matrix)\n",
    "        x,y = findCon(img)\n",
    "        if (not(x is None)):\n",
    "            x1,y1,x2,y2 = int(x.min()-1), int(y.min()-1), int(x.max()+2), int(y.max()+2)\n",
    "            if x1 < 0:\n",
    "                x1 = 0\n",
    "            if y1 < 0:\n",
    "                y1 = 0\n",
    "            if x2 > 15:\n",
    "                x2 = 15\n",
    "            if y2 > 27:\n",
    "                y2 = 27\n",
    "            lst.append([x1, y1, x2 - x1, y2 - y1, img[y1:y2, x1:x2]])\n",
    "        else:\n",
    "            lst.append([0,0,0,0,0])\n",
    "    return lst"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 6.22 s, sys: 1.7 s, total: 7.92 s\n",
      "Wall time: 8.32 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "p = Pool(cpu_count)\n",
    "\n",
    "data = np.array_split(df, cpu_count)\n",
    "pool_results = p.map(doJob, data)\n",
    "p.close()\n",
    "del data\n",
    "results = []\n",
    "for result in pool_results:\n",
    "    results.extend(result)\n",
    "\n",
    "dfNew = pd.DataFrame(results, columns=['TouchX', 'TouchY', 'TouchW', 'TouchH', 'MatrixCropped'])\n",
    "df['TouchX'] = dfNew.TouchX.tolist()\n",
    "df['TouchY'] = dfNew.TouchY.tolist()\n",
    "df['TouchH'] = dfNew.TouchH.tolist()\n",
    "df['TouchW'] = dfNew.TouchW.tolist()\n",
    "df['MatrixCropped'] = dfNew.MatrixCropped.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Data set size: 457268\n"
     ]
    }
   ],
   "source": [
    "# remove no touches\n",
    "df = df[df.TouchH > 2]\n",
    "df = df[df.TouchW > 2]\n",
    "print (\"Data set size: %i\" % len(df))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "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": [
    "fig, ax = plt.subplots()\n",
    "img = loadImg(df.iloc[1234].Matrix)\n",
    "x,y = findCon(img)\n",
    "\n",
    "ax.set_aspect('equal', 'datalim')\n",
    "plt.imshow(img)\n",
    "\n",
    "ax.plot(x,y, color='#ffffff', linewidth=3, alpha=1.0)\n",
    "x1,y1,x2,y2 = int(x.min()-1), int(y.min()-1), int(x.max()+2), int(y.max()+2)\n",
    "ax.add_patch(patches.Rectangle( (x1, y1), x2- x1, y2-y1, fill=False, color='#ffffff'))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### fit all blobs in a 22 x 15 empy image upper left corner"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 7.59 s, sys: 3.36 s, total: 10.9 s\n",
      "Wall time: 10.2 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "def doJob(df2):\n",
    "    lst = []\n",
    "    for i, d in df2.iterrows():\n",
    "        wall = np.zeros((22, 15),dtype=np.float64)\n",
    "        wall[0:d.TouchH, 0:d.TouchW] = d.Matrix[d.TouchY:(d.TouchY+d.TouchH), d.TouchX:(d.TouchX+d.TouchW)]\n",
    "        lst.append(wall)\n",
    "    return lst\n",
    "\n",
    "p = Pool(cpu_count)\n",
    "\n",
    "data =  np.array_split(df, cpu_count)\n",
    "pool_results = p.map(doJob, data)\n",
    "p.close()\n",
    "del data\n",
    "results = []\n",
    "for result in pool_results:\n",
    "    results.extend(result)\n",
    "    \n",
    "df[\"MatrixCroppedSameSize\"] = results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>Pair</th>\n",
       "      <th>Participant</th>\n",
       "      <th>Condition</th>\n",
       "      <th>Matrix</th>\n",
       "      <th>Pitch</th>\n",
       "      <th>Yaw</th>\n",
       "      <th>TimeOffset</th>\n",
       "      <th>IsFlipped</th>\n",
       "      <th>TouchX</th>\n",
       "      <th>TouchY</th>\n",
       "      <th>TouchH</th>\n",
       "      <th>TouchW</th>\n",
       "      <th>MatrixCropped</th>\n",
       "      <th>MatrixCroppedSameSize</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>214</th>\n",
       "      <td>107</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>[[220, 0, 0, -3, -2, -3, -1, -1, -1, 0, 0, -27...</td>\n",
       "      <td>14.614879</td>\n",
       "      <td>-88.972128</td>\n",
       "      <td>-0.000106</td>\n",
       "      <td>True</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "      <td>9</td>\n",
       "      <td>7</td>\n",
       "      <td>[[11.0, 13.0, 13.0, 13.0, 13.0, 7.0, 0.0], [16...</td>\n",
       "      <td>[[11.0, 13.0, 13.0, 13.0, 13.0, 7.0, 1.0, 0.0,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>215</th>\n",
       "      <td>107</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>[[0, 0, -2, -27, 0, 0, -1, -1, -1, -3, -2, -3,...</td>\n",
       "      <td>14.614879</td>\n",
       "      <td>88.972128</td>\n",
       "      <td>-0.000106</td>\n",
       "      <td>False</td>\n",
       "      <td>7</td>\n",
       "      <td>9</td>\n",
       "      <td>9</td>\n",
       "      <td>8</td>\n",
       "      <td>[[0.0, 0.0, 7.0, 13.0, 13.0, 13.0, 13.0, 11.0]...</td>\n",
       "      <td>[[1.0, 1.0, 7.0, 13.0, 13.0, 13.0, 13.0, 11.0,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>216</th>\n",
       "      <td>108</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>[[220, 0, 0, -3, -2, -3, -1, -1, -1, 0, 0, -27...</td>\n",
       "      <td>28.088492</td>\n",
       "      <td>-78.902481</td>\n",
       "      <td>-0.000926</td>\n",
       "      <td>True</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "      <td>9</td>\n",
       "      <td>7</td>\n",
       "      <td>[[8.0, 10.0, 9.0, 10.0, 8.0, 4.0, 0.0], [15.0,...</td>\n",
       "      <td>[[8.0, 10.0, 9.0, 10.0, 8.0, 4.0, 1.0, 0.0, 0....</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>217</th>\n",
       "      <td>108</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>[[0, 0, -2, -27, 0, 0, -1, -1, -1, -3, -2, -3,...</td>\n",
       "      <td>28.088492</td>\n",
       "      <td>78.902481</td>\n",
       "      <td>-0.000926</td>\n",
       "      <td>False</td>\n",
       "      <td>7</td>\n",
       "      <td>9</td>\n",
       "      <td>9</td>\n",
       "      <td>8</td>\n",
       "      <td>[[0.0, 0.0, 4.0, 8.0, 10.0, 9.0, 10.0, 8.0], [...</td>\n",
       "      <td>[[2.0, 1.0, 4.0, 8.0, 10.0, 9.0, 10.0, 8.0, 0....</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>218</th>\n",
       "      <td>109</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>[[0, 0, -2, -27, 0, 0, -1, -1, -1, -3, -2, -3,...</td>\n",
       "      <td>28.257708</td>\n",
       "      <td>79.029943</td>\n",
       "      <td>-0.001010</td>\n",
       "      <td>False</td>\n",
       "      <td>7</td>\n",
       "      <td>10</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>[[0.0, 0.0, 6.0, 8.0, 12.0, 12.0, 11.0, 11.0],...</td>\n",
       "      <td>[[2.0, 3.0, 6.0, 8.0, 12.0, 12.0, 11.0, 11.0, ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Pair  Participant  Condition  \\\n",
       "214   107            1          2   \n",
       "215   107            1          2   \n",
       "216   108            1          2   \n",
       "217   108            1          2   \n",
       "218   109            1          2   \n",
       "\n",
       "                                                Matrix      Pitch        Yaw  \\\n",
       "214  [[220, 0, 0, -3, -2, -3, -1, -1, -1, 0, 0, -27...  14.614879 -88.972128   \n",
       "215  [[0, 0, -2, -27, 0, 0, -1, -1, -1, -3, -2, -3,...  14.614879  88.972128   \n",
       "216  [[220, 0, 0, -3, -2, -3, -1, -1, -1, 0, 0, -27...  28.088492 -78.902481   \n",
       "217  [[0, 0, -2, -27, 0, 0, -1, -1, -1, -3, -2, -3,...  28.088492  78.902481   \n",
       "218  [[0, 0, -2, -27, 0, 0, -1, -1, -1, -3, -2, -3,...  28.257708  79.029943   \n",
       "\n",
       "     TimeOffset  IsFlipped  TouchX  TouchY  TouchH  TouchW  \\\n",
       "214   -0.000106       True       0       9       9       7   \n",
       "215   -0.000106      False       7       9       9       8   \n",
       "216   -0.000926       True       0       9       9       7   \n",
       "217   -0.000926      False       7       9       9       8   \n",
       "218   -0.001010      False       7      10       8       8   \n",
       "\n",
       "                                         MatrixCropped  \\\n",
       "214  [[11.0, 13.0, 13.0, 13.0, 13.0, 7.0, 0.0], [16...   \n",
       "215  [[0.0, 0.0, 7.0, 13.0, 13.0, 13.0, 13.0, 11.0]...   \n",
       "216  [[8.0, 10.0, 9.0, 10.0, 8.0, 4.0, 0.0], [15.0,...   \n",
       "217  [[0.0, 0.0, 4.0, 8.0, 10.0, 9.0, 10.0, 8.0], [...   \n",
       "218  [[0.0, 0.0, 6.0, 8.0, 12.0, 12.0, 11.0, 11.0],...   \n",
       "\n",
       "                                 MatrixCroppedSameSize  \n",
       "214  [[11.0, 13.0, 13.0, 13.0, 13.0, 7.0, 1.0, 0.0,...  \n",
       "215  [[1.0, 1.0, 7.0, 13.0, 13.0, 13.0, 13.0, 11.0,...  \n",
       "216  [[8.0, 10.0, 9.0, 10.0, 8.0, 4.0, 1.0, 0.0, 0....  \n",
       "217  [[2.0, 1.0, 4.0, 8.0, 10.0, 9.0, 10.0, 8.0, 0....  \n",
       "218  [[2.0, 3.0, 6.0, 8.0, 12.0, 12.0, 11.0, 11.0, ...  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "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": [
    "plt.imshow(df.iloc[1234].MatrixCroppedSameSize)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "df.to_pickle(\"./data/all_data.pkl\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}