Occupancy Estimation Single Analysis for Three Interest Areas.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Occupancy Estimation: Single Analysis for Three Interest Areas\n",
    "**(Note: If the results are not satisfiable, we need to check how to further improve the accuracy of estimation of occupancy with Openpose.)**\n",
    "\n",
    "* To run this note, you need to install Openpose following [Steps to Install Openpose in ComputeCanada Cluster](https://docs.google.com/document/d/1vQCQD2iet-K1ZAjTns2aEp-q7R-w8fCQ55jVzriWLV0/edit?usp=sharing)\n",
    "* To set up Jupyter Notebook and submit jobs in Compute Canada Clusters, please refer to [Steps to Use Openpose Analyzing Video](https://docs.google.com/document/d/1tozpc-KpAHVjQOx5XkW0pDYbd8IKKCD45Omk-RvFNcg/edit?usp=sharing)\n",
    "\n",
    "## Purpose\n",
    "This note is to estimate occupancy in each of the [Three Interest Areas]() i.e.\n",
    "1. Whole Interest Area\n",
    "2. Core Interest Area\n",
    "3. Margin Interest Area."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import os\n",
    "os.system('module load nixpkgs/16.09  gcc/5.4.0  cuda/8.0.44  cudnn/7.0 opencv/3.3.0  boost/1.65.1 openblas/0.2.20 hdf5/1.8.18 leveldb/1.18 mkl-dnn/0.14 python/3.5.2')\n",
    "os.system('cd ~')\n",
    "os.system('source openposeEnv_Python3/bin/activate')\n",
    "os.system('export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$HOME/openpose_python_lib/lib:$HOME/openpose_python_lib/python/openpose:$HOME/caffe/build/lib:/cvmfs/soft.computecanada.ca/easybuild/software/2017/avx2/Compiler/gcc5.4/boost/1.65.1/lib')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Maximun Estimation Accuracy\n",
    "#### 1. Openpose Configureation\n",
    "About maximum accuracy configuration, please refer to [Openpose: Maximum Accuracy Configuration](https://github.com/CMU-Perceptual-Computing-Lab/openpose/blob/master/doc/quick_start.md#maximum-accuracy-configuration).\n",
    "\n",
    "Basically, here is the best configuration:\n",
    "* params[\"net_resolution\"] = \"-1x736\" (higher resolution might be better, but will run out of GPU memory)\n",
    "* params[\"model_pose\"] = \"COCO\" (better than \"BODY_25\" but needs 2 GPU)\n",
    "* params[\"scale_gap\"] = 0.25\n",
    "* params[\"scale_number\"] = 4\n",
    "\n",
    "#### 2. Calculate Occupancy According to OpenPose Output\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# From Python\n",
    "# It requires OpenCV installed for Python\n",
    "import sys\n",
    "import csv\n",
    "import cv2\n",
    "import os\n",
    "from sys import platform\n",
    "import argparse\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import pdb\n",
    "from IPython.core.debugger import Tracer\n",
    "\n",
    "# Remember to add your installation path here\n",
    "# Option b\n",
    "# If you run `make install` (default path is `/usr/local/python` for Ubuntu), you can also access the OpenPose/python module from there. This will install OpenPose and the python library at your desired installation path. Ensure that this is in your python path in order to use it.\n",
    "sys.path.insert(0,r'/home/lingheng/openpose_python_lib/python/openpose') \n",
    "\n",
    "# Parameters for OpenPose. Take a look at C++ OpenPose example for meaning of components. Ensure all below are filled\n",
    "try:\n",
    "    from openpose import *\n",
    "except:\n",
    "    raise Exception('Error: OpenPose library could not be found. Did you enable `BUILD_PYTHON` in CMake and have this Python script in the right folder?')\n",
    "params = dict()\n",
    "params[\"logging_level\"] = 3\n",
    "params[\"output_resolution\"] = \"-1x-1\"\n",
    "params[\"net_resolution\"] = \"-1x736\" # if crop video, this should be changged and must be mutplies of 16.\n",
    "params[\"model_pose\"] = \"BODY_25\"#\"COCO\"#\"BODY_25\"\n",
    "params[\"alpha_pose\"] = 0.6\n",
    "params[\"scale_gap\"] = 0.25\n",
    "params[\"scale_number\"] = 4\n",
    "params[\"render_threshold\"] = 0.05\n",
    "# If GPU version is built, and multiple GPUs are available, set the ID here\n",
    "params[\"num_gpu_start\"] = 1\n",
    "params[\"disable_blending\"] = False\n",
    "# Ensure you point to the correct path where models are located\n",
    "params[\"default_model_folder\"] = \"/home/lingheng/openpose/models/\"\n",
    "# Construct OpenPose object allocates GPU memory\n",
    "openpose = OpenPose(params)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Raw frames per second: 30.00063026534171\n",
      "Frame width:1152, Frame height:972.\n"
     ]
    }
   ],
   "source": [
    "# construct the argument parser and parse the arguments\n",
    "args = dict()\n",
    "args['video']='/home/lingheng/project/lingheng/ROM_Video_Process/ROM_raw_videos/test_video.mp4'\n",
    "args['output_directory']='/home/lingheng/project/lingheng/ROM_Video_Process/Openpose_Video_Analysis_Code'\n",
    "\n",
    "if args.get(\"video\", None) is None:\n",
    "    raise Error(\"No input video!!\")\n",
    "# otherwise, we are reading from a video file\n",
    "else:\n",
    "    camera = cv2.VideoCapture(args[\"video\"])\n",
    "# frames per second (fps) in the raw video\n",
    "fps = camera.get(cv2.CAP_PROP_FPS)\n",
    "frame_count = 1\n",
    "print(\"Raw frames per second: {0}\".format(fps))\n",
    "# prepare to save video\n",
    "(grabbed, frame) = camera.read()\n",
    "## downsample frame\n",
    "#downsample_rate = 0.5\n",
    "#frame = cv2.resize(frame,None,fx=downsample_rate, fy=downsample_rate, interpolation = cv2.INTER_LINEAR)\n",
    "# crop frame\n",
    "original_h, original_w, channels= frame.shape\n",
    "top_edge = int(original_h*(1/10))\n",
    "down_edge = int(original_h*1)\n",
    "left_edge = int(original_w*(1/5))\n",
    "right_edge = int(original_w*(4/5))\n",
    "frame_cropped = frame[top_edge:down_edge,left_edge:right_edge,:].copy() # must use copy(), otherwise slice only return address i.e. not hard copy\n",
    "\n",
    "cropped_h, cropped_w, channels = frame_cropped.shape\n",
    "fwidth = cropped_w \n",
    "fheight = cropped_h\n",
    "print(\"Frame width:{}, Frame height:{}.\".format(cropped_w , cropped_h))\n",
    "# Define the polygon of Core Interest Area\n",
    "point_1 = [int(0.17 * cropped_w), int(0.20 * cropped_h)]\n",
    "point_2 = [int(0.17 * cropped_w), int(0.62 * cropped_h)]\n",
    "point_3 = [int(0.44 * cropped_w), int(0.82 * cropped_h)]\n",
    "point_4 = [int(0.61 * cropped_w), int(0.72 * cropped_h)]\n",
    "point_5 = [int(0.61 * cropped_w), int(0.20 * cropped_h)]\n",
    "core_interest_area_polygon = np.array([point_1,point_2,point_3,point_4,point_5])\n",
    "\n",
    "# get output video file name\n",
    "file_path = args[\"video\"].split('/')\n",
    "file_name, _= file_path[-1].split('.')\n",
    "fourcc = cv2.VideoWriter_fourcc(*'XVID')\n",
    "\n",
    "out_camera_frame_whole = cv2.VideoWriter(os.path.join(args['output_directory'],'{}_processed_whole.avi'.format(file_name)),fourcc, fps, (fwidth,fheight))\n",
    "\n",
    "# get output estimated occupancy file name\n",
    "out_occupancy_whole = os.path.join(args['output_directory'],'{}_processed_occupancy_whole.csv'.format(file_name))\n",
    "out_occupancy_core = os.path.join(args['output_directory'],'{}_processed_occupancy_core.csv'.format(file_name))\n",
    "out_occupancy_margin = os.path.join(args['output_directory'],'{}_processed_occupancy_margin.csv'.format(file_name))\n",
    "with open(out_occupancy_whole, 'a') as csv_datafile:\n",
    "    fieldnames = ['Time', 'Occupancy']\n",
    "    writer = csv.DictWriter(csv_datafile, fieldnames = fieldnames)\n",
    "    writer.writeheader()\n",
    "with open(out_occupancy_core, 'a') as csv_datafile:\n",
    "    fieldnames = ['Time', 'Occupancy']\n",
    "    writer = csv.DictWriter(csv_datafile, fieldnames = fieldnames)\n",
    "    writer.writeheader()    \n",
    "with open(out_occupancy_margin, 'a') as csv_datafile:\n",
    "    fieldnames = ['Time', 'Occupancy']\n",
    "    writer = csv.DictWriter(csv_datafile, fieldnames = fieldnames)\n",
    "    writer.writeheader()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1428.0"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "camera.get(cv2.CAP_PROP_FRAME_COUNT)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Process every frame in the video.\n",
    "Overfall procudure:\n",
    "1. Get new frame\n",
    "2. Crop the frame to get Whole Interest Area\n",
    "3. Use Openpose analyze the Whole Interest Area and get\n",
    "   * keypoints_whole_interest_area: information on detected human [(detailes info)](https://github.com/CMU-Perceptual-Computing-Lab/openpose/blob/master/doc/output.md)\n",
    "   * output_image_whole_interest_area: marked image\n",
    "4. Calculate occupancy based on detected human in each of the Three Interest Areas\n",
    "5. Write estimation of occupancy in each area into three separate files:\n",
    "   * `test_video_processed_occupancy_whole.csv`: occupancy in Whole Interest Area\n",
    "   * `test_video_processed_occupancy_core.csv`: occupancy in Core Interest Area\n",
    "   * `test_video_processed_occupancy_margin.csv`: occupancy in Margin Interest Area\n",
    "6. Write decorated frame to output video.\n",
    "   * `test_video_processed_whole.avi`: decorated output video.\n",
    "\n",
    "**Note:\n",
    "When we calculate the location of the detected human, we choose the location of keypoint with the hightest confidence as the location of the detected huamn.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Processing frame: 0\n",
      "> <ipython-input-5-47aa40272e29>(33)<module>()\n",
      "-> for people in keypoints_whole_interest_area:\n",
      "(Pdb) keypoints_whole_interest_area\n",
      "array([[[0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [3.0876762e+01, 4.1530106e+02, 6.6540468e-01],\n",
      "        [4.4094513e+01, 4.1663654e+02, 6.3137144e-01],\n",
      "        [4.6716839e+01, 4.3780423e+02, 4.7457364e-01],\n",
      "        [4.1454395e+01, 4.5100604e+02, 1.4897123e-01],\n",
      "        [1.7669554e+01, 4.1268906e+02, 5.9005630e-01],\n",
      "        [4.4829831e+00, 4.3249658e+02, 2.7154326e-01],\n",
      "        [5.7891574e+00, 4.4701697e+02, 7.6491520e-02],\n",
      "        [3.6132339e+01, 4.6025397e+02, 6.4326036e-01],\n",
      "        [3.6158779e+01, 4.9458575e+02, 6.3116223e-01],\n",
      "        [3.6150780e+01, 5.2629749e+02, 5.7809651e-01],\n",
      "        [1.6361681e+01, 4.6021124e+02, 5.6966090e-01],\n",
      "        [1.2380327e+01, 4.9327603e+02, 5.8091313e-01],\n",
      "        [1.2380680e+01, 5.2368555e+02, 4.9761504e-01],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [4.0137348e+01, 3.9419870e+02, 4.4934064e-01],\n",
      "        [2.5613865e+01, 3.9416556e+02, 4.5780569e-01]],\n",
      "\n",
      "       [[2.4490912e+02, 4.1400137e+02, 2.2048141e-01],\n",
      "        [2.2506644e+02, 4.3776529e+02, 8.1044215e-01],\n",
      "        [2.3961478e+02, 4.3908624e+02, 7.4931359e-01],\n",
      "        [2.4885966e+02, 4.6946967e+02, 6.1423236e-01],\n",
      "        [2.5016573e+02, 4.9192310e+02, 4.9737322e-01],\n",
      "        [2.0658670e+02, 4.3646725e+02, 7.1213955e-01],\n",
      "        [1.9604445e+02, 4.6290070e+02, 4.5385358e-01],\n",
      "        [2.0660329e+02, 4.8005377e+02, 2.2266585e-01],\n",
      "        [2.3168697e+02, 4.9460043e+02, 7.4517715e-01],\n",
      "        [2.2639073e+02, 5.3687402e+02, 7.6723194e-01],\n",
      "        [2.2244568e+02, 5.7913007e+02, 6.9073617e-01],\n",
      "        [2.0789926e+02, 4.9325308e+02, 7.1323007e-01],\n",
      "        [2.0656110e+02, 5.3424890e+02, 7.5068426e-01],\n",
      "        [1.9735829e+02, 5.7517548e+02, 7.0799947e-01],\n",
      "        [2.4226846e+02, 4.1004001e+02, 2.1479186e-01],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [2.3697844e+02, 4.1399823e+02, 5.2256668e-01],\n",
      "        [2.2112270e+02, 4.1664764e+02, 1.8364064e-01]],\n",
      "\n",
      "       [[0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [2.7924866e+02, 4.2195218e+02, 6.2203878e-01],\n",
      "        [2.9112527e+02, 4.2192859e+02, 5.5854237e-01],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [2.6734210e+02, 4.2588998e+02, 6.3646483e-01],\n",
      "        [2.6470001e+02, 4.4966809e+02, 5.7461858e-01],\n",
      "        [2.5944296e+02, 4.5100266e+02, 2.7400160e-01],\n",
      "        [2.8845959e+02, 4.6816794e+02, 5.4227352e-01],\n",
      "        [2.8189450e+02, 5.0253448e+02, 5.9857213e-01],\n",
      "        [2.8452609e+02, 5.3554449e+02, 5.8735240e-01],\n",
      "        [2.7001151e+02, 4.6948050e+02, 6.2730968e-01],\n",
      "        [2.7265308e+02, 5.0382294e+02, 5.8462489e-01],\n",
      "        [2.7922104e+02, 5.3684851e+02, 5.6038529e-01],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [2.8059488e+02, 4.0739758e+02, 2.5468197e-01],\n",
      "        [2.6736594e+02, 4.0869815e+02, 3.8304484e-01]],\n",
      "\n",
      "       [[5.4476978e+02, 4.2191525e+02, 7.1075462e-02],\n",
      "        [5.4343347e+02, 4.2722165e+02, 4.2083383e-01],\n",
      "        [5.3292126e+02, 4.2588721e+02, 3.9408067e-01],\n",
      "        [5.2233270e+02, 4.3908490e+02, 4.6072149e-01],\n",
      "        [5.4215643e+02, 4.3909473e+02, 4.7736111e-01],\n",
      "        [5.5270154e+02, 4.2723224e+02, 3.9797547e-01],\n",
      "        [5.6456384e+02, 4.4043845e+02, 3.1136200e-01],\n",
      "        [5.5931708e+02, 4.3911746e+02, 3.0044019e-01],\n",
      "        [5.3555115e+02, 4.6288202e+02, 5.3917766e-01],\n",
      "        [5.3425305e+02, 4.9063388e+02, 5.2914578e-01],\n",
      "        [5.3421722e+02, 5.1441626e+02, 4.8657039e-01],\n",
      "        [5.4873755e+02, 4.6289145e+02, 5.2177852e-01],\n",
      "        [5.5138293e+02, 4.9191458e+02, 5.2632064e-01],\n",
      "        [5.4873999e+02, 5.1572925e+02, 5.0651276e-01],\n",
      "        [5.3951697e+02, 4.1137509e+02, 5.7617910e-02],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [5.3818524e+02, 4.1401306e+02, 1.0771846e-01],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00]]], dtype=float32)\n",
      "(Pdb) plt.imshow(output_image_whole_interest_area)\n",
      "<matplotlib.image.AxesImage object at 0x2b0d03169240>\n",
      "(Pdb) %matplotlib inline\n",
      "*** SyntaxError: invalid syntax\n",
      "(Pdb) c\n",
      "> <ipython-input-5-47aa40272e29>(32)<module>()\n",
      "-> pdb.set_trace()\n",
      "(Pdb) keypoints_whole_interest_area\n",
      "array([[[0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\n",
      "        [3.08766708e+01, 4.15300903e+02, 6.65540218e-01],\n",
      "        [4.40944748e+01, 4.16636536e+02, 6.31612301e-01],\n",
      "        [4.67168159e+01, 4.37804260e+02, 4.75008875e-01],\n",
      "        [4.14548569e+01, 4.51006317e+02, 1.49289429e-01],\n",
      "        [1.76695633e+01, 4.12689117e+02, 5.90099931e-01],\n",
      "        [4.48298359e+00, 4.32496521e+02, 2.71723419e-01],\n",
      "        [5.78900242e+00, 4.47016998e+02, 7.63761178e-02],\n",
      "        [3.61323128e+01, 4.60253845e+02, 6.43427253e-01],\n",
      "        [3.61588173e+01, 4.94585632e+02, 6.31630361e-01],\n",
      "        [3.61508827e+01, 5.26297302e+02, 5.78584671e-01],\n",
      "        [1.63617535e+01, 4.60211121e+02, 5.69718122e-01],\n",
      "        [1.23804150e+01, 4.93275696e+02, 5.81173062e-01],\n",
      "        [1.23808899e+01, 5.23685364e+02, 4.98128057e-01],\n",
      "        [0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\n",
      "        [0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\n",
      "        [4.01372337e+01, 3.94198761e+02, 4.49517608e-01],\n",
      "        [2.56141243e+01, 3.94165314e+02, 4.57757294e-01]],\n",
      "\n",
      "       [[2.46224213e+02, 4.11377411e+02, 2.14444339e-01],\n",
      "        [2.25074631e+02, 4.37764343e+02, 8.18017840e-01],\n",
      "        [2.39622528e+02, 4.39089508e+02, 7.52412677e-01],\n",
      "        [2.48861389e+02, 4.69498169e+02, 6.45880818e-01],\n",
      "        [2.50181747e+02, 4.91961365e+02, 5.64208984e-01],\n",
      "        [2.06589706e+02, 4.36461731e+02, 7.16635227e-01],\n",
      "        [1.96041168e+02, 4.62894104e+02, 4.74480271e-01],\n",
      "        [2.06600098e+02, 4.80070129e+02, 2.38233596e-01],\n",
      "        [2.31678680e+02, 4.95901123e+02, 7.47968614e-01],\n",
      "        [2.26386917e+02, 5.36879089e+02, 7.63882577e-01],\n",
      "        [2.22453766e+02, 5.79132690e+02, 6.89963341e-01],\n",
      "        [2.06623764e+02, 4.93258759e+02, 7.19864547e-01],\n",
      "        [2.06562439e+02, 5.34250305e+02, 7.58575559e-01],\n",
      "        [1.97358871e+02, 5.75174561e+02, 7.06897080e-01],\n",
      "        [2.43563995e+02, 4.08725494e+02, 2.10392237e-01],\n",
      "        [0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\n",
      "        [2.36984268e+02, 4.12691284e+02, 5.24991810e-01],\n",
      "        [2.22437012e+02, 4.16642303e+02, 1.94151223e-01]],\n",
      "\n",
      "       [[0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\n",
      "        [2.77926788e+02, 4.21926300e+02, 6.42961144e-01],\n",
      "        [2.89806061e+02, 4.20619019e+02, 5.85052133e-01],\n",
      "        [0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\n",
      "        [0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\n",
      "        [2.64733429e+02, 4.24565918e+02, 6.66624367e-01],\n",
      "        [2.62090179e+02, 4.48357361e+02, 5.64607918e-01],\n",
      "        [2.59435303e+02, 4.52320923e+02, 2.57017255e-01],\n",
      "        [2.87165100e+02, 4.66861603e+02, 5.47199786e-01],\n",
      "        [2.79288330e+02, 5.02493622e+02, 6.22568607e-01],\n",
      "        [2.80560730e+02, 5.38180481e+02, 6.05678380e-01],\n",
      "        [2.69990753e+02, 4.68167847e+02, 6.30438924e-01],\n",
      "        [2.69986542e+02, 5.02492920e+02, 5.00458360e-01],\n",
      "        [2.77915985e+02, 5.36852539e+02, 3.93920869e-01],\n",
      "        [0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\n",
      "        [0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\n",
      "        [2.80576385e+02, 4.07400604e+02, 3.08827341e-01],\n",
      "        [2.67328217e+02, 4.07427826e+02, 3.95516306e-01]],\n",
      "\n",
      "       [[5.43489990e+02, 4.20596039e+02, 7.31678978e-02],\n",
      "        [5.43427551e+02, 4.27226166e+02, 4.30102348e-01],\n",
      "        [5.32921082e+02, 4.25892426e+02, 4.01474863e-01],\n",
      "        [5.22334717e+02, 4.39089905e+02, 4.62599486e-01],\n",
      "        [5.42157227e+02, 4.39094788e+02, 4.81355190e-01],\n",
      "        [5.52689209e+02, 4.27234375e+02, 4.03666109e-01],\n",
      "        [5.64558533e+02, 4.40444275e+02, 3.09304029e-01],\n",
      "        [5.59309937e+02, 4.39115112e+02, 3.00798088e-01],\n",
      "        [5.35551697e+02, 4.62894409e+02, 5.51015556e-01],\n",
      "        [5.34255798e+02, 4.91905762e+02, 5.40542960e-01],\n",
      "        [5.34228577e+02, 5.15730774e+02, 4.99237657e-01],\n",
      "        [5.48728577e+02, 4.62897644e+02, 5.31040430e-01],\n",
      "        [5.51387573e+02, 4.91931244e+02, 5.39427340e-01],\n",
      "        [5.50083618e+02, 5.18368774e+02, 5.20428479e-01],\n",
      "        [5.39518005e+02, 4.11372040e+02, 6.04081303e-02],\n",
      "        [0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\n",
      "        [5.38181885e+02, 4.14014404e+02, 1.11650094e-01],\n",
      "        [0.00000000e+00, 0.00000000e+00, 0.00000000e+00]]], dtype=float32)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(Pdb) c\n",
      "> <ipython-input-5-47aa40272e29>(33)<module>()\n",
      "-> for people in keypoints_whole_interest_area:\n",
      "(Pdb) keypoints_whole_interest_area\n",
      "array([[[0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [3.0878069e+01, 4.1530646e+02, 6.6519535e-01],\n",
      "        [4.4086868e+01, 4.1664236e+02, 6.2876958e-01],\n",
      "        [4.6714996e+01, 4.3780499e+02, 4.6877766e-01],\n",
      "        [4.1458839e+01, 4.5101196e+02, 1.5582037e-01],\n",
      "        [1.7676708e+01, 4.1269162e+02, 5.8917499e-01],\n",
      "        [4.4834213e+00, 4.3250067e+02, 2.7275273e-01],\n",
      "        [5.7908344e+00, 4.4701544e+02, 7.7307470e-02],\n",
      "        [3.6135269e+01, 4.6025351e+02, 6.4442009e-01],\n",
      "        [3.6165615e+01, 4.9457632e+02, 6.2963223e-01],\n",
      "        [3.6157059e+01, 5.2628680e+02, 5.8030343e-01],\n",
      "        [1.6367445e+01, 4.6021124e+02, 5.6989092e-01],\n",
      "        [1.2388121e+01, 4.9326929e+02, 5.8110547e-01],\n",
      "        [1.2387604e+01, 5.2367212e+02, 5.0096226e-01],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [4.0140316e+01, 3.9420453e+02, 4.4490421e-01],\n",
      "        [2.5618689e+01, 3.9416962e+02, 4.5858100e-01]],\n",
      "\n",
      "       [[2.4621481e+02, 4.1269153e+02, 2.0626946e-01],\n",
      "        [2.2638434e+02, 4.3774835e+02, 8.3430171e-01],\n",
      "        [2.4095132e+02, 4.3778568e+02, 7.6539373e-01],\n",
      "        [2.4888141e+02, 4.6946109e+02, 6.5750563e-01],\n",
      "        [2.5017404e+02, 4.9196564e+02, 5.8428991e-01],\n",
      "        [2.0661122e+02, 4.3645697e+02, 7.2079742e-01],\n",
      "        [1.9604771e+02, 4.6288962e+02, 4.9633181e-01],\n",
      "        [2.0660329e+02, 4.8137070e+02, 2.7428734e-01],\n",
      "        [2.3170158e+02, 4.9460461e+02, 7.6450604e-01],\n",
      "        [2.2640054e+02, 5.3687054e+02, 7.8134012e-01],\n",
      "        [2.2245609e+02, 5.7912769e+02, 7.0006853e-01],\n",
      "        [2.0792068e+02, 4.9325406e+02, 7.3050547e-01],\n",
      "        [2.0657535e+02, 5.3424689e+02, 7.6057500e-01],\n",
      "        [1.9737224e+02, 5.7517389e+02, 7.0763731e-01],\n",
      "        [2.4226761e+02, 4.1003580e+02, 1.9735156e-01],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [2.3699678e+02, 4.1266971e+02, 5.2498597e-01],\n",
      "        [2.2113861e+02, 4.1531723e+02, 2.0676202e-01]],\n",
      "\n",
      "       [[0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [2.7789621e+02, 4.2325052e+02, 6.2233353e-01],\n",
      "        [2.8977823e+02, 4.2063043e+02, 5.4779863e-01],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [2.6472076e+02, 4.2719965e+02, 6.6441053e-01],\n",
      "        [2.6076694e+02, 4.4966721e+02, 4.9035236e-01],\n",
      "        [2.4755150e+02, 4.7873160e+02, 2.1500421e-01],\n",
      "        [2.8059732e+02, 4.6816934e+02, 5.3639191e-01],\n",
      "        [2.7923486e+02, 5.0253253e+02, 6.0136443e-01],\n",
      "        [2.7928937e+02, 5.3685382e+02, 5.6589705e-01],\n",
      "        [2.6736972e+02, 4.6947101e+02, 6.2397987e-01],\n",
      "        [2.6870541e+02, 5.0253433e+02, 5.2314931e-01],\n",
      "        [2.7791748e+02, 5.3556305e+02, 4.2235687e-01],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [2.7925220e+02, 4.0738940e+02, 3.0287012e-01],\n",
      "        [2.6604709e+02, 4.0872113e+02, 3.7616748e-01]],\n",
      "\n",
      "       [[5.4345947e+02, 4.1531491e+02, 6.8117641e-02],\n",
      "        [5.4216284e+02, 4.2326358e+02, 4.1553974e-01],\n",
      "        [5.3288977e+02, 4.2192511e+02, 4.0115628e-01],\n",
      "        [5.2232031e+02, 4.3907404e+02, 4.7124052e-01],\n",
      "        [5.4214587e+02, 4.3909882e+02, 4.8056209e-01],\n",
      "        [5.5272711e+02, 4.2456024e+02, 3.8531643e-01],\n",
      "        [5.6455194e+02, 4.3913196e+02, 3.2047799e-01],\n",
      "        [5.5799615e+02, 4.3913153e+02, 3.0255923e-01],\n",
      "        [5.3554919e+02, 4.6287576e+02, 5.4403085e-01],\n",
      "        [5.3425476e+02, 4.9191861e+02, 5.3733015e-01],\n",
      "        [5.3421619e+02, 5.1702563e+02, 4.8983452e-01],\n",
      "        [5.4875104e+02, 4.6288028e+02, 5.2959555e-01],\n",
      "        [5.5139954e+02, 4.9195221e+02, 5.3193450e-01],\n",
      "        [5.5140350e+02, 5.1967896e+02, 5.1171011e-01],\n",
      "        [5.3952820e+02, 4.0871899e+02, 5.6626417e-02],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [5.3816174e+02, 4.1135553e+02, 1.0254696e-01],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00]]], dtype=float32)\n",
      "(Pdb) c\n",
      "> <ipython-input-5-47aa40272e29>(32)<module>()\n",
      "-> pdb.set_trace()\n",
      "(Pdb) keypoints_whole_interest_area.shape[0]\n",
      "4\n",
      "(Pdb) c\n",
      "> <ipython-input-5-47aa40272e29>(33)<module>()\n",
      "-> for people in keypoints_whole_interest_area:\n",
      "(Pdb) keypoints_whole_interest_area.shape[0]\n",
      "4\n",
      "(Pdb) c\n",
      "Processing frame: 5\n",
      "> <ipython-input-5-47aa40272e29>(32)<module>()\n",
      "-> pdb.set_trace()\n",
      "(Pdb) keypoints_whole_interest_area.shape[0]\n",
      "5\n",
      "(Pdb) keypoints_whole_interest_area\n",
      "array([[[0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [3.0876728e+01, 4.1530463e+02, 6.6813064e-01],\n",
      "        [4.4090340e+01, 4.1664423e+02, 6.3064826e-01],\n",
      "        [4.6729340e+01, 4.3780112e+02, 4.9984318e-01],\n",
      "        [4.2777237e+01, 4.5100610e+02, 1.7523965e-01],\n",
      "        [1.7669796e+01, 4.1268805e+02, 5.9704697e-01],\n",
      "        [4.4808054e+00, 4.3380008e+02, 2.9524088e-01],\n",
      "        [5.7919960e+00, 4.4570959e+02, 8.3067976e-02],\n",
      "        [3.6139183e+01, 4.6025040e+02, 6.4634126e-01],\n",
      "        [3.6172707e+01, 4.9328998e+02, 6.3451666e-01],\n",
      "        [3.6159649e+01, 5.2499231e+02, 5.8941829e-01],\n",
      "        [1.6374342e+01, 4.5894156e+02, 5.7473361e-01],\n",
      "        [1.3686185e+01, 4.9325485e+02, 5.8761555e-01],\n",
      "        [1.3684810e+01, 5.2365826e+02, 5.1744664e-01],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [4.0146427e+01, 3.9420700e+02, 4.6031004e-01],\n",
      "        [2.5623110e+01, 3.9416779e+02, 4.5327657e-01]],\n",
      "\n",
      "       [[2.4358377e+02, 4.1795987e+02, 1.8707269e-01],\n",
      "        [2.2508876e+02, 4.3778177e+02, 8.3150893e-01],\n",
      "        [2.3963824e+02, 4.3910620e+02, 7.6130021e-01],\n",
      "        [2.4887427e+02, 4.7080438e+02, 6.2700665e-01],\n",
      "        [2.4886812e+02, 4.9327518e+02, 4.7788772e-01],\n",
      "        [2.0659740e+02, 4.3648267e+02, 7.1263814e-01],\n",
      "        [1.9602548e+02, 4.6419781e+02, 4.8518264e-01],\n",
      "        [2.0658699e+02, 4.8138647e+02, 2.3513818e-01],\n",
      "        [2.3169159e+02, 4.9723880e+02, 7.4215889e-01],\n",
      "        [2.2640463e+02, 5.3817633e+02, 7.5718510e-01],\n",
      "        [2.2246262e+02, 5.7913611e+02, 6.8266845e-01],\n",
      "        [2.0790849e+02, 4.9329227e+02, 7.0073986e-01],\n",
      "        [2.0656606e+02, 5.3552423e+02, 7.4359596e-01],\n",
      "        [1.9737222e+02, 5.7518054e+02, 6.8474770e-01],\n",
      "        [2.4093933e+02, 4.1268213e+02, 1.7206503e-01],\n",
      "        [2.4886122e+02, 4.1796277e+02, 1.3131821e-01],\n",
      "        [2.3698505e+02, 4.1268839e+02, 5.0178635e-01],\n",
      "        [2.2113252e+02, 4.1532132e+02, 2.3072630e-01]],\n",
      "\n",
      "       [[0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [2.7133218e+02, 4.2192209e+02, 6.4008075e-01],\n",
      "        [2.8189813e+02, 4.2061423e+02, 5.5152321e-01],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [2.6077280e+02, 4.2458051e+02, 6.9040519e-01],\n",
      "        [2.6074405e+02, 4.4835825e+02, 4.7616488e-01],\n",
      "        [2.5813568e+02, 4.5627078e+02, 1.9565305e-01],\n",
      "        [2.8057858e+02, 4.6684778e+02, 5.5110180e-01],\n",
      "        [2.7662131e+02, 5.0381421e+02, 6.0143292e-01],\n",
      "        [2.7793576e+02, 5.3817419e+02, 5.3507918e-01],\n",
      "        [2.6865067e+02, 4.6815530e+02, 6.4666682e-01],\n",
      "        [2.6866797e+02, 5.0253207e+02, 5.9767479e-01],\n",
      "        [2.7000327e+02, 5.3424414e+02, 4.8113808e-01],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [2.7263663e+02, 4.0736984e+02, 2.7976924e-01],\n",
      "        [2.6206558e+02, 4.0741425e+02, 3.5961956e-01]],\n",
      "\n",
      "       [[5.4479706e+02, 4.2192133e+02, 8.2135558e-02],\n",
      "        [5.4344556e+02, 4.2722693e+02, 4.6280986e-01],\n",
      "        [5.3292017e+02, 4.2587811e+02, 4.2539316e-01],\n",
      "        [5.2232056e+02, 4.3910455e+02, 4.6521831e-01],\n",
      "        [5.4213849e+02, 4.3912357e+02, 4.7483787e-01],\n",
      "        [5.5272699e+02, 4.2851346e+02, 4.1910270e-01],\n",
      "        [5.6327454e+02, 4.4440155e+02, 3.4206966e-01],\n",
      "        [5.5668958e+02, 4.4042725e+02, 3.1401762e-01],\n",
      "        [5.3557159e+02, 4.6549677e+02, 5.7072175e-01],\n",
      "        [5.3552142e+02, 4.9198053e+02, 5.5931967e-01],\n",
      "        [5.3422089e+02, 5.1835382e+02, 5.1168287e-01],\n",
      "        [5.5004468e+02, 4.6551086e+02, 5.5301058e-01],\n",
      "        [5.5140631e+02, 4.9329163e+02, 5.5214655e-01],\n",
      "        [5.5403467e+02, 5.2228821e+02, 5.3767771e-01],\n",
      "        [5.4213892e+02, 4.1268723e+02, 6.3244171e-02],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [5.3948932e+02, 4.1533035e+02, 1.1217050e-01],\n",
      "        [5.5138104e+02, 4.1795370e+02, 6.0547952e-02]],\n",
      "\n",
      "       [[0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [3.5459064e+02, 3.9550568e+02, 3.2506958e-01],\n",
      "        [3.6380862e+02, 3.9548389e+02, 2.8689560e-01],\n",
      "        [3.7305435e+02, 4.1798611e+02, 9.6882313e-02],\n",
      "        [3.7302228e+02, 4.3647012e+02, 9.5891401e-02],\n",
      "        [3.4665903e+02, 3.9553745e+02, 3.0489981e-01],\n",
      "        [3.4135010e+02, 4.1796164e+02, 1.4691710e-01],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [3.6248355e+02, 4.3909787e+02, 3.5783175e-01],\n",
      "        [3.5851517e+02, 4.7213168e+02, 3.2466760e-01],\n",
      "        [3.5459637e+02, 5.0251462e+02, 2.7996707e-01],\n",
      "        [3.5058633e+02, 4.3911685e+02, 3.9450881e-01],\n",
      "        [3.5319049e+02, 4.7346219e+02, 3.8015863e-01],\n",
      "        [3.5456506e+02, 5.0384897e+02, 3.2062495e-01],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [3.4528082e+02, 3.8098218e+02, 8.3751120e-02]]], dtype=float32)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(Pdb) c\n",
      "> <ipython-input-5-47aa40272e29>(33)<module>()\n",
      "-> for people in keypoints_whole_interest_area:\n",
      "(Pdb) keypoints_whole_interest_area.shape[0]\n",
      "5\n",
      "(Pdb) c\n",
      "> <ipython-input-5-47aa40272e29>(32)<module>()\n",
      "-> pdb.set_trace()\n",
      "(Pdb)  keypoints_whole_interest_area.shape[0]\n",
      "5\n",
      "(Pdb) c\n",
      "> <ipython-input-5-47aa40272e29>(33)<module>()\n",
      "-> for people in keypoints_whole_interest_area:\n",
      "(Pdb)  keypoints_whole_interest_area.shape[0]\n",
      "5\n",
      "(Pdb) c\n",
      "> <ipython-input-5-47aa40272e29>(32)<module>()\n",
      "-> pdb.set_trace()\n",
      "(Pdb)  keypoints_whole_interest_area.shape[0]\n",
      "5\n",
      "(Pdb) c\n",
      "Processing frame: 10\n",
      "> <ipython-input-5-47aa40272e29>(33)<module>()\n",
      "-> for people in keypoints_whole_interest_area:\n",
      "(Pdb)  keypoints_whole_interest_area.shape[0]\n",
      "5\n",
      "(Pdb) c\n",
      "> <ipython-input-5-47aa40272e29>(32)<module>()\n",
      "-> pdb.set_trace()\n",
      "(Pdb)  keypoints_whole_interest_area.shape[0]\n",
      "4\n",
      "(Pdb) c\n",
      "> <ipython-input-5-47aa40272e29>(33)<module>()\n",
      "-> for people in keypoints_whole_interest_area:\n",
      "(Pdb)  keypoints_whole_interest_area.shape[0]\n",
      "4\n",
      "(Pdb) c\n",
      "> <ipython-input-5-47aa40272e29>(32)<module>()\n",
      "-> pdb.set_trace()\n",
      "(Pdb)  keypoints_whole_interest_area.shape[0]\n",
      "4\n",
      "(Pdb) c\n",
      "> <ipython-input-5-47aa40272e29>(33)<module>()\n",
      "-> for people in keypoints_whole_interest_area:\n",
      "(Pdb)  keypoints_whole_interest_area.shape[0]\n",
      "4\n",
      "(Pdb) c\n",
      "Processing frame: 15\n",
      "> <ipython-input-5-47aa40272e29>(32)<module>()\n",
      "-> pdb.set_trace()\n",
      "(Pdb)  keypoints_whole_interest_area.shape[0]\n",
      "4\n",
      "(Pdb) c\n",
      "> <ipython-input-5-47aa40272e29>(33)<module>()\n",
      "-> for people in keypoints_whole_interest_area:\n",
      "(Pdb)  keypoints_whole_interest_area.shape[0]\n",
      "4\n",
      "(Pdb) c\n",
      "> <ipython-input-5-47aa40272e29>(32)<module>()\n",
      "-> pdb.set_trace()\n",
      "(Pdb)  keypoints_whole_interest_area.shape[0]\n",
      "4\n",
      "(Pdb) c\n",
      "> <ipython-input-5-47aa40272e29>(33)<module>()\n",
      "-> for people in keypoints_whole_interest_area:\n",
      "(Pdb)  keypoints_whole_interest_area.shape[0]\n",
      "4\n",
      "(Pdb) c\n",
      "> <ipython-input-5-47aa40272e29>(32)<module>()\n",
      "-> pdb.set_trace()\n",
      "(Pdb)  keypoints_whole_interest_area.shape[0]\n",
      "3\n",
      "(Pdb) keypoints_whole_interest_area\n",
      "array([[[0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [3.0895006e+01, 4.1532303e+02, 6.7090499e-01],\n",
      "        [4.5400295e+01, 4.1665738e+02, 6.3321352e-01],\n",
      "        [4.6736897e+01, 4.3779321e+02, 4.8981860e-01],\n",
      "        [4.1465363e+01, 4.5099893e+02, 1.6954133e-01],\n",
      "        [1.7671200e+01, 4.1399045e+02, 5.9867394e-01],\n",
      "        [4.4751177e+00, 4.3250189e+02, 3.1694832e-01],\n",
      "        [5.7977910e+00, 4.4306241e+02, 9.2666224e-02],\n",
      "        [3.6169003e+01, 4.6025519e+02, 6.5380049e-01],\n",
      "        [3.6187435e+01, 4.9328223e+02, 6.3735837e-01],\n",
      "        [3.6179165e+01, 5.2497150e+02, 6.0326844e-01],\n",
      "        [1.7687927e+01, 4.6022003e+02, 5.7678986e-01],\n",
      "        [1.5004710e+01, 4.9197714e+02, 5.9823370e-01],\n",
      "        [1.5002448e+01, 5.2364502e+02, 5.4825372e-01],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [4.0152000e+01, 3.9547311e+02, 4.3087530e-01],\n",
      "        [2.6889673e+01, 3.9418524e+02, 4.7464016e-01]],\n",
      "\n",
      "       [[2.4225523e+02, 4.1928790e+02, 1.0037558e-01],\n",
      "        [2.2378088e+02, 4.3911823e+02, 7.5059837e-01],\n",
      "        [2.4094576e+02, 4.4173746e+02, 7.1755332e-01],\n",
      "        [2.4887517e+02, 4.7346075e+02, 6.1696136e-01],\n",
      "        [2.5017451e+02, 4.9195203e+02, 4.4119352e-01],\n",
      "        [2.0656770e+02, 4.3908734e+02, 6.8143845e-01],\n",
      "        [1.9865262e+02, 4.6815274e+02, 4.9391323e-01],\n",
      "        [2.0922588e+02, 4.7741165e+02, 2.0036155e-01],\n",
      "        [2.3167903e+02, 4.9855316e+02, 6.9846511e-01],\n",
      "        [2.2637807e+02, 5.3819806e+02, 7.0625275e-01],\n",
      "        [2.2244901e+02, 5.7913477e+02, 6.2704080e-01],\n",
      "        [2.0922076e+02, 4.9591663e+02, 6.5951627e-01],\n",
      "        [2.0791612e+02, 5.3685040e+02, 7.0183587e-01],\n",
      "        [2.0132359e+02, 5.7648364e+02, 6.3279927e-01],\n",
      "        [2.4226082e+02, 4.1531857e+02, 9.9917032e-02],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [2.3827722e+02, 4.1797678e+02, 4.5747885e-01],\n",
      "        [2.1714709e+02, 4.1794876e+02, 2.9699805e-01]],\n",
      "\n",
      "       [[5.5273938e+02, 4.2323807e+02, 7.7893861e-02],\n",
      "        [5.5404724e+02, 4.3117877e+02, 3.9610723e-01],\n",
      "        [5.4348260e+02, 4.2852902e+02, 3.8682902e-01],\n",
      "        [5.3421350e+02, 4.4306720e+02, 4.1685602e-01],\n",
      "        [5.5271552e+02, 4.4439859e+02, 4.4126031e-01],\n",
      "        [5.6589600e+02, 4.3382645e+02, 3.7764522e-01],\n",
      "        [5.7254266e+02, 4.4967206e+02, 3.0283302e-01],\n",
      "        [5.6855719e+02, 4.4833691e+02, 2.4059406e-01],\n",
      "        [5.4482111e+02, 4.6685681e+02, 5.2828842e-01],\n",
      "        [5.4214294e+02, 4.9326779e+02, 5.1537818e-01],\n",
      "        [5.3422101e+02, 5.1703229e+02, 4.7345933e-01],\n",
      "        [5.6064557e+02, 4.6816983e+02, 4.9466076e-01],\n",
      "        [5.5800604e+02, 4.9722876e+02, 4.8288602e-01],\n",
      "        [5.5668146e+02, 5.2498767e+02, 4.8485187e-01],\n",
      "        [5.5138287e+02, 4.2059830e+02, 6.0817283e-02],\n",
      "        [0.0000000e+00, 0.0000000e+00, 0.0000000e+00],\n",
      "        [5.5142139e+02, 4.2455493e+02, 8.8327438e-02],\n",
      "        [5.6195032e+02, 4.2323383e+02, 5.6238309e-02]]], dtype=float32)\n",
      "(Pdb) q\n"
     ]
    },
    {
     "ename": "BdbQuit",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mBdbQuit\u001b[0m                                   Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-5-47aa40272e29>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     30\u001b[0m         \u001b[0moccupancy_core\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     31\u001b[0m         \u001b[0moccupancy_margin\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 32\u001b[0;31m         \u001b[0mpdb\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mset_trace\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     33\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mpeople\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mkeypoints_whole_interest_area\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     34\u001b[0m             \u001b[0;31m# Sort all keypoints and pick up the one with the highest confidence\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-5-47aa40272e29>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     30\u001b[0m         \u001b[0moccupancy_core\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     31\u001b[0m         \u001b[0moccupancy_margin\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 32\u001b[0;31m         \u001b[0mpdb\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mset_trace\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     33\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mpeople\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mkeypoints_whole_interest_area\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     34\u001b[0m             \u001b[0;31m# Sort all keypoints and pick up the one with the highest confidence\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/cvmfs/soft.computecanada.ca/nix/store/v29pphgl66qjvjck1mn4pcm5g9agk5kh-python3-3.5.2/lib/python3.5/bdb.py\u001b[0m in \u001b[0;36mtrace_dispatch\u001b[0;34m(self, frame, event, arg)\u001b[0m\n\u001b[1;32m     46\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0;31m# None\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     47\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mevent\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;34m'line'\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 48\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdispatch_line\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mframe\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     49\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mevent\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;34m'call'\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     50\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdispatch_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mframe\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0marg\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/cvmfs/soft.computecanada.ca/nix/store/v29pphgl66qjvjck1mn4pcm5g9agk5kh-python3-3.5.2/lib/python3.5/bdb.py\u001b[0m in \u001b[0;36mdispatch_line\u001b[0;34m(self, frame)\u001b[0m\n\u001b[1;32m     65\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstop_here\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mframe\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbreak_here\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mframe\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     66\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0muser_line\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mframe\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 67\u001b[0;31m             \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mquitting\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;32mraise\u001b[0m \u001b[0mBdbQuit\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     68\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrace_dispatch\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     69\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mBdbQuit\u001b[0m: "
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# loop over the frames of the video\n",
    "total_frame_number = camera.get(cv2.CAP_PROP_FRAME_COUNT)\n",
    "for frame_count in range(int(total_frame_number)):\n",
    "    if frame_count % 5 == 0:\n",
    "        print('Processing frame: {}'.format(frame_count))\n",
    "    (grabbed, frame) = camera.read()\n",
    "    if grabbed == True:\n",
    "        time = camera.get(cv2.CAP_PROP_POS_MSEC) #Current position of the video file in milliseconds.\n",
    "        ## downsample frame\n",
    "        #frame = cv2.resize(frame,None,fx=downsample_rate, fy=downsample_rate, interpolation = cv2.INTER_LINEAR)\n",
    "        # crop frame\n",
    "        frame_cropped = frame[top_edge:down_edge,left_edge:right_edge,:].copy() # must use copy()\n",
    "        #pdb.set_trace()\n",
    "        # 1. Whole Interest Area\n",
    "        # Output keypoints and the image with the human skeleton blended on it\n",
    "        #    (num_people, 25_keypoints, x_y_confidence) = keypoints_whole_interest_area.shape\n",
    "        keypoints_whole_interest_area, output_image_whole_interest_area = openpose.forward(frame_cropped, True)\n",
    "        \n",
    "        # 2. Core Interest Area\n",
    "        core_interest_area_mask = np.zeros(frame_cropped.shape[:2], np.uint8)\n",
    "        cv2.drawContours(core_interest_area_mask, [core_interest_area_polygon], -1, (255, 255, 255), -1, cv2.LINE_AA)\n",
    "        core_interest_area = cv2.bitwise_and(output_image_whole_interest_area, frame_cropped, mask=core_interest_area_mask)\n",
    "            \n",
    "        # 3. Margin Interest Area\n",
    "        margin_interest_area = cv2.bitwise_xor(output_image_whole_interest_area, core_interest_area)\n",
    "        # TODO: infer occupancy from \"keypoints_whole_interest_area\"\n",
    "        \n",
    "        # draw the text and timestamp on the frame\n",
    "        occupancy_whole = keypoints_whole_interest_area.shape[0]\n",
    "        occupancy_core = 0\n",
    "        occupancy_margin = 0\n",
    "        pdb.set_trace()\n",
    "        for people in keypoints_whole_interest_area:\n",
    "            # Sort all keypoints and pick up the one with the highest confidence\n",
    "            # Meaning of keypoints (https://github.com/CMU-Perceptual-Computing-Lab/openpose/blob/master/doc/output.md)\n",
    "            ordered_keypoints = people[people[:,2].argsort(),:] # increasing order\n",
    "            x, y = ordered_keypoints[-1][:2]\n",
    "            #pdb.set_trace()\n",
    "            # Choose the one with higher confidence to calculatate occupancy and location\n",
    "            if cv2.pointPolygonTest(core_interest_area_polygon, (x, y), False) == 1:\n",
    "                occupancy_core += 1\n",
    "            else:\n",
    "                occupancy_margin += 1\n",
    "                \n",
    "        cv2.drawContours(output_image_whole_interest_area, [core_interest_area_polygon], -1, (255, 255, 0), 2, cv2.LINE_AA)\n",
    "        cv2.putText(output_image_whole_interest_area, \"Whole Occupancy: {}, Core Occupancy: {}, Margin Occupancy: {}\".format(occupancy_whole, occupancy_core, occupancy_margin), (10, 80), cv2.FONT_HERSHEY_SIMPLEX, 1, (255, 255, 255), 2)\n",
    "        cv2.putText(core_interest_area, \"Core Occupancy: {}\".format(occupancy_core), (10, 50), cv2.FONT_HERSHEY_SIMPLEX, 1, (255, 255, 255), 2)\n",
    "        cv2.putText(margin_interest_area, \"Margin Occupancy: {}\".format(occupancy_margin), (10, 50), cv2.FONT_HERSHEY_SIMPLEX, 1, (255, 255, 255), 2)\n",
    "        # save estimated occupancy data\n",
    "        fieldnames = ['Time', 'Occupancy']\n",
    "        with open(out_occupancy_whole, 'a') as csv_datafile:\n",
    "            writer = csv.DictWriter(csv_datafile, fieldnames = fieldnames)\n",
    "            writer.writerow({'Time':time, 'Occupancy': occupancy_whole})\n",
    "        with open(out_occupancy_core, 'a') as csv_datafile:\n",
    "            writer = csv.DictWriter(csv_datafile, fieldnames = fieldnames)\n",
    "            writer.writerow({'Time':time, 'Occupancy': occupancy_core})\n",
    "        with open(out_occupancy_margin, 'a') as csv_datafile:\n",
    "            writer = csv.DictWriter(csv_datafile, fieldnames = fieldnames)\n",
    "            writer.writerow({'Time':time, 'Occupancy': occupancy_margin})\n",
    "        # save processed videos\n",
    "        out_camera_frame_whole.write(output_image_whole_interest_area)\n",
    "    else:\n",
    "        # Pass this frame if cannot grab an image.\n",
    "        print('Frame: {}, grabbed={} and frame={}'.format(frame_count, grabbed, frame))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Result\n",
    "After processing, we can get an output video as shown [here (YouTube Video)](https://youtu.be/1iPKDs9P1i4) and output estimation of occupancy in Three Interest Areas."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "occupancy_whole = pd.read_csv('test_video_processed_occupancy_whole.csv', sep=',')\n",
    "occupancy_core = pd.read_csv('test_video_processed_occupancy_core.csv', sep=',')\n",
    "occupancy_margin = pd.read_csv('test_video_processed_occupancy_margin.csv', sep=',')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>Time</th>\n",
       "      <th>Occupancy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>66.665266</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>99.997899</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>133.330532</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>166.663165</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>199.995798</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>233.328431</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>266.661064</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>299.993697</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>333.326331</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>366.658964</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>399.991597</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>433.324230</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>466.656863</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>499.989496</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>533.322129</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>566.654762</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>599.987395</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>633.320028</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>666.652661</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>699.985294</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>733.317927</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>766.650560</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>799.983193</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>833.315826</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>866.648459</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>899.981092</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>933.313725</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>966.646359</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>999.978992</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>1033.311625</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1397</th>\n",
       "      <td>46632.353641</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1398</th>\n",
       "      <td>46665.686275</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1399</th>\n",
       "      <td>46699.018908</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1400</th>\n",
       "      <td>46732.351541</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1401</th>\n",
       "      <td>46765.684174</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1402</th>\n",
       "      <td>46799.016807</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1403</th>\n",
       "      <td>46832.349440</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1404</th>\n",
       "      <td>46865.682073</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1405</th>\n",
       "      <td>46899.014706</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1406</th>\n",
       "      <td>46932.347339</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1407</th>\n",
       "      <td>46965.679972</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1408</th>\n",
       "      <td>46999.012605</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1409</th>\n",
       "      <td>47032.345238</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1410</th>\n",
       "      <td>47065.677871</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1411</th>\n",
       "      <td>47099.010504</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1412</th>\n",
       "      <td>47132.343137</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1413</th>\n",
       "      <td>47165.675770</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1414</th>\n",
       "      <td>47199.008403</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1415</th>\n",
       "      <td>47232.341036</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1416</th>\n",
       "      <td>47265.673669</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1417</th>\n",
       "      <td>47299.006303</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1418</th>\n",
       "      <td>47332.338936</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1419</th>\n",
       "      <td>47365.671569</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1420</th>\n",
       "      <td>47399.004202</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1421</th>\n",
       "      <td>47432.336835</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1422</th>\n",
       "      <td>47465.669468</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1423</th>\n",
       "      <td>47499.002101</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1424</th>\n",
       "      <td>47532.334734</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1425</th>\n",
       "      <td>47565.667367</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1426</th>\n",
       "      <td>47599.000000</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1427 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              Time  Occupancy\n",
       "0        66.665266          4\n",
       "1        99.997899          4\n",
       "2       133.330532          4\n",
       "3       166.663165          4\n",
       "4       199.995798          4\n",
       "5       233.328431          5\n",
       "6       266.661064          5\n",
       "7       299.993697          5\n",
       "8       333.326331          5\n",
       "9       366.658964          5\n",
       "10      399.991597          5\n",
       "11      433.324230          4\n",
       "12      466.656863          4\n",
       "13      499.989496          4\n",
       "14      533.322129          4\n",
       "15      566.654762          4\n",
       "16      599.987395          4\n",
       "17      633.320028          4\n",
       "18      666.652661          4\n",
       "19      699.985294          3\n",
       "20      733.317927          3\n",
       "21      766.650560          3\n",
       "22      799.983193          3\n",
       "23      833.315826          3\n",
       "24      866.648459          3\n",
       "25      899.981092          3\n",
       "26      933.313725          3\n",
       "27      966.646359          3\n",
       "28      999.978992          3\n",
       "29     1033.311625          3\n",
       "...            ...        ...\n",
       "1397  46632.353641          9\n",
       "1398  46665.686275          9\n",
       "1399  46699.018908          9\n",
       "1400  46732.351541          9\n",
       "1401  46765.684174          9\n",
       "1402  46799.016807          9\n",
       "1403  46832.349440          9\n",
       "1404  46865.682073          9\n",
       "1405  46899.014706          9\n",
       "1406  46932.347339          9\n",
       "1407  46965.679972          9\n",
       "1408  46999.012605          9\n",
       "1409  47032.345238          9\n",
       "1410  47065.677871          9\n",
       "1411  47099.010504          9\n",
       "1412  47132.343137          9\n",
       "1413  47165.675770          9\n",
       "1414  47199.008403          9\n",
       "1415  47232.341036          9\n",
       "1416  47265.673669          9\n",
       "1417  47299.006303          9\n",
       "1418  47332.338936          9\n",
       "1419  47365.671569          9\n",
       "1420  47399.004202          9\n",
       "1421  47432.336835          9\n",
       "1422  47465.669468          9\n",
       "1423  47499.002101          9\n",
       "1424  47532.334734          9\n",
       "1425  47565.667367          9\n",
       "1426  47599.000000          9\n",
       "\n",
       "[1427 rows x 2 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "occupancy_whole"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>Time</th>\n",
       "      <th>Occupancy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>66.665266</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>99.997899</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>133.330532</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>166.663165</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>199.995798</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>233.328431</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>266.661064</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>299.993697</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>333.326331</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>366.658964</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>399.991597</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>433.324230</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>466.656863</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>499.989496</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>533.322129</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>566.654762</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>599.987395</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>633.320028</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>666.652661</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>699.985294</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>733.317927</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>766.650560</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>799.983193</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>833.315826</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>866.648459</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>899.981092</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>933.313725</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>966.646359</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>999.978992</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>1033.311625</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1397</th>\n",
       "      <td>46632.353641</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1398</th>\n",
       "      <td>46665.686275</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1399</th>\n",
       "      <td>46699.018908</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1400</th>\n",
       "      <td>46732.351541</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1401</th>\n",
       "      <td>46765.684174</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1402</th>\n",
       "      <td>46799.016807</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1403</th>\n",
       "      <td>46832.349440</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1404</th>\n",
       "      <td>46865.682073</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1405</th>\n",
       "      <td>46899.014706</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1406</th>\n",
       "      <td>46932.347339</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1407</th>\n",
       "      <td>46965.679972</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1408</th>\n",
       "      <td>46999.012605</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1409</th>\n",
       "      <td>47032.345238</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1410</th>\n",
       "      <td>47065.677871</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1411</th>\n",
       "      <td>47099.010504</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1412</th>\n",
       "      <td>47132.343137</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1413</th>\n",
       "      <td>47165.675770</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1414</th>\n",
       "      <td>47199.008403</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1415</th>\n",
       "      <td>47232.341036</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1416</th>\n",
       "      <td>47265.673669</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1417</th>\n",
       "      <td>47299.006303</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1418</th>\n",
       "      <td>47332.338936</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1419</th>\n",
       "      <td>47365.671569</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1420</th>\n",
       "      <td>47399.004202</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1421</th>\n",
       "      <td>47432.336835</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1422</th>\n",
       "      <td>47465.669468</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1423</th>\n",
       "      <td>47499.002101</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1424</th>\n",
       "      <td>47532.334734</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1425</th>\n",
       "      <td>47565.667367</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1426</th>\n",
       "      <td>47599.000000</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1427 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              Time  Occupancy\n",
       "0        66.665266          3\n",
       "1        99.997899          3\n",
       "2       133.330532          3\n",
       "3       166.663165          3\n",
       "4       199.995798          3\n",
       "5       233.328431          4\n",
       "6       266.661064          4\n",
       "7       299.993697          4\n",
       "8       333.326331          4\n",
       "9       366.658964          4\n",
       "10      399.991597          4\n",
       "11      433.324230          3\n",
       "12      466.656863          3\n",
       "13      499.989496          3\n",
       "14      533.322129          3\n",
       "15      566.654762          3\n",
       "16      599.987395          3\n",
       "17      633.320028          3\n",
       "18      666.652661          3\n",
       "19      699.985294          2\n",
       "20      733.317927          2\n",
       "21      766.650560          2\n",
       "22      799.983193          2\n",
       "23      833.315826          2\n",
       "24      866.648459          2\n",
       "25      899.981092          2\n",
       "26      933.313725          2\n",
       "27      966.646359          2\n",
       "28      999.978992          2\n",
       "29     1033.311625          2\n",
       "...            ...        ...\n",
       "1397  46632.353641          4\n",
       "1398  46665.686275          4\n",
       "1399  46699.018908          4\n",
       "1400  46732.351541          4\n",
       "1401  46765.684174          4\n",
       "1402  46799.016807          4\n",
       "1403  46832.349440          4\n",
       "1404  46865.682073          4\n",
       "1405  46899.014706          4\n",
       "1406  46932.347339          4\n",
       "1407  46965.679972          4\n",
       "1408  46999.012605          4\n",
       "1409  47032.345238          4\n",
       "1410  47065.677871          4\n",
       "1411  47099.010504          4\n",
       "1412  47132.343137          4\n",
       "1413  47165.675770          4\n",
       "1414  47199.008403          4\n",
       "1415  47232.341036          4\n",
       "1416  47265.673669          4\n",
       "1417  47299.006303          4\n",
       "1418  47332.338936          4\n",
       "1419  47365.671569          4\n",
       "1420  47399.004202          4\n",
       "1421  47432.336835          4\n",
       "1422  47465.669468          4\n",
       "1423  47499.002101          4\n",
       "1424  47532.334734          4\n",
       "1425  47565.667367          3\n",
       "1426  47599.000000          3\n",
       "\n",
       "[1427 rows x 2 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "occupancy_core"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>Time</th>\n",
       "      <th>Occupancy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>66.665266</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>99.997899</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>133.330532</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>166.663165</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>199.995798</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>233.328431</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>266.661064</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>299.993697</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>333.326331</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>366.658964</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>399.991597</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>433.324230</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>466.656863</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>499.989496</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>533.322129</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>566.654762</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>599.987395</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>633.320028</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>666.652661</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>699.985294</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>733.317927</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>766.650560</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>799.983193</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>833.315826</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>866.648459</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>899.981092</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>933.313725</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>966.646359</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>999.978992</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>1033.311625</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1397</th>\n",
       "      <td>46632.353641</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1398</th>\n",
       "      <td>46665.686275</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1399</th>\n",
       "      <td>46699.018908</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1400</th>\n",
       "      <td>46732.351541</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1401</th>\n",
       "      <td>46765.684174</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1402</th>\n",
       "      <td>46799.016807</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1403</th>\n",
       "      <td>46832.349440</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1404</th>\n",
       "      <td>46865.682073</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1405</th>\n",
       "      <td>46899.014706</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1406</th>\n",
       "      <td>46932.347339</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1407</th>\n",
       "      <td>46965.679972</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1408</th>\n",
       "      <td>46999.012605</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1409</th>\n",
       "      <td>47032.345238</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1410</th>\n",
       "      <td>47065.677871</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1411</th>\n",
       "      <td>47099.010504</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1412</th>\n",
       "      <td>47132.343137</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1413</th>\n",
       "      <td>47165.675770</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1414</th>\n",
       "      <td>47199.008403</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1415</th>\n",
       "      <td>47232.341036</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1416</th>\n",
       "      <td>47265.673669</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1417</th>\n",
       "      <td>47299.006303</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1418</th>\n",
       "      <td>47332.338936</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1419</th>\n",
       "      <td>47365.671569</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1420</th>\n",
       "      <td>47399.004202</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1421</th>\n",
       "      <td>47432.336835</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1422</th>\n",
       "      <td>47465.669468</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1423</th>\n",
       "      <td>47499.002101</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1424</th>\n",
       "      <td>47532.334734</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1425</th>\n",
       "      <td>47565.667367</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1426</th>\n",
       "      <td>47599.000000</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1427 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              Time  Occupancy\n",
       "0        66.665266          1\n",
       "1        99.997899          1\n",
       "2       133.330532          1\n",
       "3       166.663165          1\n",
       "4       199.995798          1\n",
       "5       233.328431          1\n",
       "6       266.661064          1\n",
       "7       299.993697          1\n",
       "8       333.326331          1\n",
       "9       366.658964          1\n",
       "10      399.991597          1\n",
       "11      433.324230          1\n",
       "12      466.656863          1\n",
       "13      499.989496          1\n",
       "14      533.322129          1\n",
       "15      566.654762          1\n",
       "16      599.987395          1\n",
       "17      633.320028          1\n",
       "18      666.652661          1\n",
       "19      699.985294          1\n",
       "20      733.317927          1\n",
       "21      766.650560          1\n",
       "22      799.983193          1\n",
       "23      833.315826          1\n",
       "24      866.648459          1\n",
       "25      899.981092          1\n",
       "26      933.313725          1\n",
       "27      966.646359          1\n",
       "28      999.978992          1\n",
       "29     1033.311625          1\n",
       "...            ...        ...\n",
       "1397  46632.353641          5\n",
       "1398  46665.686275          5\n",
       "1399  46699.018908          5\n",
       "1400  46732.351541          5\n",
       "1401  46765.684174          5\n",
       "1402  46799.016807          5\n",
       "1403  46832.349440          5\n",
       "1404  46865.682073          5\n",
       "1405  46899.014706          5\n",
       "1406  46932.347339          5\n",
       "1407  46965.679972          5\n",
       "1408  46999.012605          5\n",
       "1409  47032.345238          5\n",
       "1410  47065.677871          5\n",
       "1411  47099.010504          5\n",
       "1412  47132.343137          5\n",
       "1413  47165.675770          5\n",
       "1414  47199.008403          5\n",
       "1415  47232.341036          5\n",
       "1416  47265.673669          5\n",
       "1417  47299.006303          5\n",
       "1418  47332.338936          5\n",
       "1419  47365.671569          5\n",
       "1420  47399.004202          5\n",
       "1421  47432.336835          5\n",
       "1422  47465.669468          5\n",
       "1423  47499.002101          5\n",
       "1424  47532.334734          5\n",
       "1425  47565.667367          6\n",
       "1426  47599.000000          6\n",
       "\n",
       "[1427 rows x 2 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "occupancy_margin"
   ]
  },
  {
   "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}