Flow Analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Flow Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import netCDF4\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import networkx as nx\n",
    "import random\n",
    "dataset = netCDF4.Dataset('./datasets/world_oscar_vel_5d2019.nc.gz')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "uf = dataset.variables['uf'][0,0].data\n",
    "vf = dataset.variables['vf'][0,0].data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Pre-process data for analysis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "graph = nx.DiGraph()\n",
    "for x in range(0, 360, 5):\n",
    "    for y in range(0, 140, 5):\n",
    "        graph.add_node((x, y))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's define some helper functions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def polar_deg(x, y):\n",
    "    z = x * 1j + y\n",
    "    return np.angle(z)\n",
    "\n",
    "def grid_coord(x, y):\n",
    "    return (x - x % 5, y - y % 5)\n",
    "\n",
    "def in_range(x, y):\n",
    "    return 0 <= x < 360 and 0 <= y < 140"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, calculate the weights."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "deg = polar_deg(uf, vf)\n",
    "deg = np.swapaxes(deg, 1, 0)\n",
    "dx = [5, 5, 0, -5, -5, -5, 0, 5]\n",
    "dy = [0, -5, -5, -5, 0, 5, 5, 5]\n",
    "edges_dict = dict()\n",
    "for x in range(0, 360):\n",
    "    for y in range(0, 140):\n",
    "        if np.isnan(deg[x,y]):\n",
    "            continue\n",
    "        for i in range(8):\n",
    "            if 2 * np.pi / 8 * i <= deg[x,y] < 2 * np.pi * (i + 1):\n",
    "                gridx, gridy = grid_coord(x, y)\n",
    "                nextx = gridx + dx[i]\n",
    "                nexty = gridy + dy[i]\n",
    "                if not in_range(nextx, nexty):\n",
    "                    continue\n",
    "                if ((gridx, gridy), (nextx, nexty)) not in edges_dict:\n",
    "                    edges_dict[((gridx, gridy), (nextx, nexty))] = 1\n",
    "                else:\n",
    "                    edges_dict[((gridx, gridy), (nextx, nexty))] += 1 "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, update the weights and show which part is center in this graph."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "ebunch = [key + ({'capacity': val},) for key, val in edges_dict.items()]\n",
    "graph.add_edges_from(ebunch)\n",
    "\n",
    "indeg_centrality = nx.algorithms.in_degree_centrality(graph)\n",
    "outdeg_centrality = nx.algorithms.out_degree_centrality(graph)\n",
    "indeg_centrality_map = np.zeros((360, 140))\n",
    "outdeg_centrality_map = np.zeros((360, 140))\n",
    "for key, val in indeg_centrality.items():\n",
    "    for x in range(key[0], key[0] + 5):\n",
    "        for y in range(key[1], key[1] + 5):\n",
    "            indeg_centrality_map[x][y] = val\n",
    "for key, val in outdeg_centrality.items():\n",
    "    for x in range(key[0], key[0] + 5):\n",
    "        for y in range(key[1], key[1] + 5):\n",
    "            outdeg_centrality_map[x][y] = val"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, let's plot the centrality."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(18, 8))\n",
    "plt.imshow(indeg_centrality_map.swapaxes(1, 0))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABBQAAAGoCAYAAAD2L7Y1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nO3df6zdaX0n9vcHzAw1O7PgELFcwIWOR2HHUeNELraV2qJxtybZbehIKAOqdqcu0jDSpE2cNAts/yCVGimjtuvutmwGWjC0ojDIiwuK6HoZArqu1tfJeOMmeAgbDwlm5gKTbSGZ7nQpkzz9wwfq9fjX89x7v+fe69dLsnzP95zPfT7nnuf7fM9963u+t1prAQAAAOjxonk3AAAAAGw8AgUAAACgm0ABAAAA6CZQAAAAALoJFAAAAIBuAgUAAACg25oFClX1lqr6SlVdqKr3rNU4AAAAwPSqtbb637TqxUn+WZK/luSpJL+T5B2ttSdWfTAAAABgclvW6Pu+KcmF1tpXk6SqPpHkrUmuGijcVre3l+Zla9TKC/3Fy6cbayN50Xf+xbxbAAAA1jG/S21so7/zPZtv//PW2g9fuX2tAoXXJPn6ZbefSrLnWg9+aV6WPXVwjVp5oed+6pqt3NK2njgz7xYAAIB1zO9SG9vo73yPteNfu9r2tQoUbqiqHkjyQJK8NFvn1QYAAAAwYK0uyvh0ktdddvu1s20/0Fr7YGttd2tt90ty+xq1AQAAAKyFtQoUfifJ3VX1hqq6Lcnbk3xmjcYCAAAAJrYmH3lorT1fVT+f5GSSFyf5cGvt/FqMBQAAAExvza6h0Fr7bJLPrtX3BwAAAOZnrT7yAAAAAGxiAgUAAACgm0ABAAAA6CZQAAAAALoJFAAAAIBuAgUAAACgm0ABAAAA6LZl3g0kyW1vfFEWPnpHd93y3meHxtt64sxQHQCsloWl/uNeMn7sm3o8AKbx3L17huqWD9Qqd7I2Duw7P+8Wbsri6Z2Tjrew2IbqRudLPnX8qpudoQAAAAB0EygAAAAA3QQKAAAAQDeBAgAAANBNoAAAAAB0EygAAAAA3QQKAAAAQDeBAgAAANBNoAAAAAB0EygAAAAA3QQKAAAAQDeBAgAAANBNoAAAAAB02zLvBgBgPXju3j2Tjnfh4bG6rTkzVLe899mxAXmB0bmy9cTYaweb0cnlc/NuYU0dvri/u+bCw/cMjbV8oIbqNorF0zvn3cJNObDv/KTjLR8ZO65fOLp3bMBPXX2zMxQAAACAbgIFAAAAoJtAAQAAAOgmUAAAAAC6CRQAAACAbgIFAAAAoJtAAQAAAOgmUAAAAAC6CRQAAACAbgIFAAAAoJtAAQAAAOgmUAAAAAC6CRQAAACAbtVam3cPubO2tT11cN5tAHALu3B071DdwuLYcXT5QA3VjdpxZGmo7uTyuVXu5NoOX9w/2VhJsnh656Tjjc6VrSfOrHInsHFNuSatxJTr2dRr2Ubx5H2PDNUdWtg1VDc6N/c/9K6huqmPDY+142dba7uv3O4MBQAAAKCbQAEAAADoJlAAAAAAugkUAAAAgG4CBQAAAKCbQAEAAADoJlAAAAAAugkUAAAAgG4CBQAAAKCbQAEAAADoJlAAAAAAug0HClX1uqr6QlU9UVXnq+oXZtu3VdXnquoPZ/+/YvXaBQAAANaDlZyh8HySX26t3ZNkb5KHquqeJO9J8vnW2t1JPj+7DQAAAGwi1VpbnW9U9ekk//3s35tba9+oqlcn+WJr7UeuV3tnbWt76uCq9AHA+nJy+dy8W7gphy/un3cLsKoWT+8cqttxZGmVO4EX2ijHhlEjx5TRfXbUgX3nh+qm7nNqC4tjvx9vPXFmlTtZXx5rx8+21nZfuX3Lanzzqnp9kh9PcibJq1pr35jd9c0kr7pGzQNJHkiSl2brarQBAAAATGTFF2Wsqr+U5B8m+cXW2p9dfl+7dPrDVSOe1toHW2u7W2u7X5LbV9oGAAAAMKEVBQpV9ZJcChM+1lr71Gzzt2Yfdcjs/2dW1iIAAACw3qzkrzxUkg8l+XJr7e9edtdnktw/+/r+JJ8ebw8AAABYj1ZyDYWfTPI3k/x+VX3/qip/J8mvJ/lkVb0zydeS/NzKWgQAAADWm+FAobX2vyepa9ztTzYAAADAJrbiizICAAAAtx6BAgAAANBNoAAAAAB0EygAAAAA3QQKAAAAQDeBAgAAANBNoAAAAAB02zLvBgA2i5PL5+bdwqZy+OL+ebcAq+rY9lOTjnfX6Z2TjsetaWHpjnm3sKZGj0Uj+/vhoZGmd2Df+Xm3sKaWjzw7VDe6L4weG+569MGhuh1HlobqrsUZCgAAAEA3gQIAAADQTaAAAAAAdBMoAAAAAN0ECgAAAEA3gQIAAADQTaAAAAAAdBMoAAAAAN0ECgAAAEA3gQIAAADQTaAAAAAAdBMoAAAAAN0ECgAAAEC3aq3Nu4fcWdvanjo47zaAderC0b1DdQuLY+vbqfd/YKju8MX9Q3Wjjm0/Nel4m93o67fZX4ep5/WIzf4aTO3Qwq55t8AKjB4zD+w7v8qdrA37+8Y1ejxZPL1zlTu5vh1HlobqFpbuWOVOrm90X7jr0QeH6v74F/+zs6213Vdud4YCAAAA0E2gAAAAAHQTKAAAAADdBAoAAABAN4ECAAAA0E2gAAAAAHQTKAAAAADdBAoAAABAN4ECAAAA0E2gAAAAAHQTKAAAAADdBAoAAABAN4ECAAAA0G3LvBtIktve+KIsfPSOebdxQ8t7n513C3BL2nFkaaju5PK5Ve7k+hZP7xyqO7Dv/FDd4Yv7h+pGHdt+aqhutM/R8UZNPd6oqV93bj0LS2PvybxPWl0Xju6ddLyp18BDC7vGCpdXtw+mMzrH7pr4/dXoFJt8DVwn+4IzFAAAAIBuAgUAAACgm0ABAAAA6CZQAAAAALoJFAAAAIBuAgUAAACgm0ABAAAA6CZQAAAAALoJFAAAAIBuAgUAAACgm0ABAAAA6LbiQKGqXlxVv1tVvzm7/YaqOlNVF6rq0aq6beVtAgAAAOvJapyh8AtJvnzZ7YeTHG2t7Ujy7STvXIUxAAAAgHWkWmvjxVWvTfLRJL+W5JeS/PtJ/iTJX2mtPV9V+5L8amvt0PW+z521re2pg8N9TOXC0b1DdTuOLE063pP3PTJUd2hh11AdrFcLS3cM1S2e3jlUd2Df+aG6Uce2nxqqO3xx/yp3cn2jfbI+TD1fpmRuXt3oaz7685z6/cfosWFqo8eiUaPvH0c5Fs3f6L733L17VrmTtbH1xJl5t7CpPNaOn22t7b5y+0rPUPhvk/ztJH8xu/1DSb7TWnt+dvupJK+5WmFVPVBVj1fV49/Ld1fYBgAAADCl4UChqv5Gkmdaa2dH6ltrH2yt7W6t7X5Jbh9tAwAAAJiDLSuo/ckkP1tVP5PkpUnuTPL3kry8qrbMzlJ4bZKnV94mAAAAsJ4Mn6HQWntva+21rbXXJ3l7kt9qrf2HSb6Q5G2zh92f5NMr7hIAAABYV1bjrzxc6d1JfqmqLuTSNRU+tAZjAAAAAHO0ko88/EBr7YtJvjj7+qtJ3rQa3xcAAABYn9biDAUAAABgkxMoAAAAAN0ECgAAAEA3gQIAAADQTaAAAAAAdBMoAAAAAN0ECgAAAEC3LfNuYCM5sO/8UN3y4Hg7jiwN1e1ffNdQ3anlDwzVjdr/0FifO979xFDd8t5nh+pGnFw+N9lYSXJoYdek403twtG9Q3XPPdyG6hYyVreYnWPjLY6Nd+jE6Jweqxud14cv7h+qO7b91FDd1KZey0aN/jxH6kZf88XTY/vQqMOTjjZu9LW769EHh+pG3++MWli6Y6hu9Lg++vOcel5P/TqMzpdRUz+/jbAujb4f2HrizNh4g/vehYeHyob7ZH1zhgIAAADQTaAAAAAAdBMoAAAAAN0ECgAAAEA3gQIAAADQTaAAAAAAdBMoAAAAAN0ECgAAAEA3gQIAAADQTaAAAAAAdBMoAAAAAN0ECgAAAEA3gQIAAADQrVpr8+4hd9a2tqcOzruNGzq5fG6obv9D7xqq23rizFDdqNHnd/ji/qG6Cw/fM1S3491PDNUt7322u+bC0b1DYx3Yd36o7tj2U0N1hxZ2DdVNbXSO3fXog0N1O44sDdVxa3ru3j1DdaNr9eh4Uzv1/g9012yU495mt1Hm2OhxfdRmP9bCejX6vtr7ufXhsXb8bGtt95XbnaEAAAAAdBMoAAAAAN0ECgAAAEA3gQIAAADQTaAAAAAAdBMoAAAAAN0ECgAAAEA3gQIAAADQTaAAAAAAdBMoAAAAAN0ECgAAAEA3gQIAAADQTaAAAAAAdKvW2rx7yJ21re2pg/NuY82cXD43VHdoYdcqd7K+PHfvnknH23rizKTjTWl0jo0anZsLS3cM1S2e3jlUt+PI0lAdANO5cHTvUN2BfeeH6pb3PjtUN7XRY+aojfJzYf7ss7emx9rxs6213Vdud4YCAAAA0E2gAAAAAHQTKAAAAADdBAoAAABAN4ECAAAA0E2gAAAAAHQTKAAAAADdBAoAAABAN4ECAAAA0E2gAAAAAHQTKAAAAADdVhQoVNXLq+p4Vf1BVX25qvZV1baq+lxV/eHs/1esVrMAAADA+rDSMxT+XpJ/1Fp7Y5IfS/LlJO9J8vnW2t1JPj+7DQAAAGwi1VobK6z6y0nOJfk32mXfpKq+kuTNrbVvVNWrk3yxtfYj1/ted9a2tqcODvXB/F04unfeLayZHUeW5t3Cmjq5fG7eLWwqhy/uH6pbPL1zlTsBNpMD+85POt7omjR1n6M2ypr75H2PTDreoYVdk443tdH3q6Pzesp5thF6TMb7PLb91FDd1HP6uXv3DNVtPXFmlTtZG4+142dba7uv3L6SMxTekORPkhyrqt+tqv+xql6W5FWttW/MHvPNJK+6WnFVPVBVj1fV49/Ld1fQBgAAADC1lQQKW5L8RJLfaK39eJJ/kSs+3jA7c+Gqp0C01j7YWtvdWtv9kty+gjYAAACAqa0kUHgqyVOtte+fo3E8lwKGb80+6pDZ/8+srEUAAABgvRkOFFpr30zy9ar6/vURDiZ5Islnktw/23Z/kk+vqEMAAABg3dmywvr/JMnHquq2JF9NcjiXQopPVtU7k3wtyc+tcAwAAABgnVlRoNBaO5fkBVd6zKWzFQAAAIBNaiXXUAAAAABuUQIFAAAAoJtAAQAAAOgmUAAAAAC6CRQAAACAbgIFAAAAoJtAAQAAAOi2Zd4NsPpOLp8bqrvr0QdXuZO1cWDf+ekGW7pjqGzx9M6hutHndmz7qaG6qR2+uH+obqM8v1GTzumMz0/YbHYcWZp3Czflwr17hup2vPuJVe7k+jbKse/wUNX48xueZ/eNlXF1o6/D8P53Yrr1ZXm08OhY2UZ5v7ow+D5+1LHtHxiqO/zusffHy3ufHapbbc5QAAAAALoJFAAAAIBuAgUAAACgm0ABAAAA6CZQAAAAALoJFAAAAIBuAgUAAACgm0ABAAAA6CZQAAAAALoJFAAAAIBuAgUAAACgm0ABAAAA6FattXn3kDtrW9tTB+fdxpo5uXxu0vEOX9w/6XjHtp+adLzR5zd1nxvB1D/Lux59cKhu1MLi2Pq29cSZsfGW7hiq2+wWT++cdws35cC+80N1o/vDoYVdQ3XAreG5e/cM1Y0ew2Cz8b7s6pb3PjtU91g7fra1tvvK7c5QAAAAALoJFAAAAIBuAgUAAACgm0ABAAAA6CZQAAAAALoJFAAAAIBuAgUAAACgm0ABAAAA6CZQAAAAALoJFAAAAIBuAgUAAACgm0ABAAAA6CZQAAAAALpVa23ePeTO2tb21MF5t7FmLhzdO+8WbsqT9z0y6XiHFnZNOt6IhaU7Jh3v2PZTk4436vDF/ZOOt3h656TjHdh3fqhuo7x+UxudL6Ov+9Sv39T7g3k2fxvh+HUrOLl8btLxvO6wsTx3755Jx1s+UEN1U79vGV3LHmvHz7bWdl+53RkKAAAAQDeBAgAAANBNoAAAAAB0EygAAAAA3QQKAAAAQDeBAgAAANBNoAAAAAB0EygAAAAA3QQKAAAAQDeBAgAAANBNoAAAAAB0W1GgUFVHqup8VX2pqj5eVS+tqjdU1ZmqulBVj1bVbavVLAAAALA+DAcKVfWaJP9pkt2ttR9N8uIkb0/ycJKjrbUdSb6d5J2r0SgAAACwflRrbazwUqCwlOTHkvxZkv81yX+X5GNJ/kpr7fmq2pfkV1trh673ve6sbW1PHRzqYyM4uXxuqO7wxf1Ddct7nx2qGzX6/Ji/Qwu7huouHN07VPfkfY8M1Y3uC4undw7VjdpxZGmobuo1YtToz3Nhcew4s/XEmbHxlu4Yqpvase2n5t3CTRmZZxvluY0aXTsBYLWMvn8cPYY91o6fba3tvnL78BkKrbWnk/zXSS4m+UaSP01yNsl3WmvPzx72VJLXjI4BAAAArE8r+cjDK5K8NckbkiwkeVmSt3TUP1BVj1fV49/Ld0fbAAAAAOZgJRdl/HeT/FFr7U9aa99L8qkkP5nk5VW1ZfaY1yZ5+mrFrbUPttZ2t9Z2vyS3r6ANAAAAYGorCRQuJtlbVVurqpIcTPJEki8kedvsMfcn+fTKWgQAAADWm5VcQ+FMkuNJ/mmS3599rw8meXeSX6qqC0l+KMmHVqFPAAAAYB3ZcuOHXFtr7X1J3nfF5q8medNKvi8AAACwvq3kIw8AAADALUqgAAAAAHQTKAAAAADdBAoAAABAN4ECAAAA0E2gAAAAAHQTKAAAAADdqrU27x5yZ21re+rgvNtYMyeXzw3VHVrYNVS3sHTHUN3y3meH6jaz0dduaocv7h+qO7b91Cp3sr7c9eiDQ3U7jiytcifXd+Ho3knHe/K+RyYdb3QtG7VR9tuNYnR9mdJmX8tGTb3vASsz9fHLGkGPx9rxs6213Vdud4YCAAAA0E2gAAAAAHQTKAAAAADdBAoAAABAN4ECAAAA0E2gAAAAAHQTKAAAAADdBAoAAABAN4ECAAAA0E2gAAAAAHQTKAAAAADdBAoAAABAN4ECAAAA0K1aa/PuIXfWtranDs67jRs6uXxu0vEOX9w/VHds+6lV7oTNanSObRTLe5+ddwvr0tRrGevDlMeUqY9fjpcb26GFXfNuAdaFjfK7xqjRNfeuRx8cqttxZGmojqt7rB0/21rbfeV2ZygAAAAA3QQKAAAAQDeBAgAAANBNoAAAAAB0EygAAAAA3QQKAAAAQDeBAgAAANBNoAAAAAB0EygAAAAA3QQKAAAAQDeBAgAAANBNoAAAAAB0EygAAAAA3aq1Nu8ecmdta3vq4LzbuKGTy+fm3cKaOnxx/7xbYNCx7afm3cJNuevRB4fqdhxZWuVOGLHZ10BWz9THk42yBo4a/Xlu9p/L1A4t7Jp3C2vquXv3DNVtPXFmlTtZX6Y89k29dl54+J6hulPv/8BQ3ej7wI1is79ffawdP9ta233ldmcoAAAAAN0ECgAAAEA3gQIAAADQTaAAAAAAdBMoAAAAAN0ECgAAAEA3gQIAAADQTaAAAAAAdBMoAAAAAN0ECgAAAEA3gQIAAADQ7YaBQlV9uKqeqaovXbZtW1V9rqr+cPb/K2bbq6r+flVdqKrfq6qfWMvmAQAAgPm4mTMUPpLkLVdse0+Sz7fW7k7y+dntJPnpJHfP/j2Q5DdWp00AAABgPanW2o0fVPX6JL/ZWvvR2e2vJHlza+0bVfXqJF9srf1IVX1g9vXHr3zc9b7/nbWt7amDK3smHU4unxuqO3xx/yp3cn2Lp3cO1R3Yd954qzTe1M9toxj9uYxa3vvspONxdaNr56ip19xj209NOt7Uz2/UyM9l9LmNvgYbZbyNYup9YbO769EHh+oWFm/8Hv1qtp44M1TH1U197Bux2X9HGbXZ34+P2nFkaajusXb8bGtt95XbR6+h8KrLQoJvJnnV7OvXJPn6ZY97arYNAAAA2ERWfFHGdukUh+4ItaoeqKrHq+rx7+W7K20DAAAAmNBooPCt2UcdMvv/mdn2p5O87rLHvXa27QVaax9sre1ure1+SW4fbAMAAACYh9FA4TNJ7p99fX+ST1+2/W/N/trD3iR/eqPrJwAAAAAbz5YbPaCqPp7kzUleWVVPJXlfkl9P8smqemeSryX5udnDP5vkZ5JcSPJcksNr0DMAAAAwZzcMFFpr77jGXS/4swyz6yk8tNKmAAAAgPVtxRdlBAAAAG49AgUAAACgm0ABAAAA6CZQAAAAALoJFAAAAIBuAgUAAACgm0ABAAAA6LZl3g0kyW1vfFEWPnrHvNtYdw7sOz/vFm7K4umdm3a8qZ/b1BYW21DdhcV7huqWD9RQXY6OlT153yNDdXc9+uDYgING+zy0sGuo7uTyuaG6qY3uf6M/z8MX9w/VHdt+aqhu1OTr0sDzm/pnMmrq13yjzLGNYuq1eqN47t49k4639cSZScdbWBr7nWHq/Wh0f9/MNvv76lGj71tG7V9811jhp45fdbMzFAAAAIBuAgUAAACgm0ABAAAA6CZQAAAAALoJFAAAAIBuAgUAAACgm0ABAAAA6CZQAAAAALoJFAAAAIBuAgUAAACgm0ABAAAA6CZQAAAAALoJFAAAAIBu1Vqbdw955V99ZfvrH/3ZebfBJrd4eue8W2CDWFic/7p4M069/wNDdYcv7h+qO7b91FDdqLsefXCo7sn7HlnlTtbG6Oswaso1cHQfWj5Qq9wJIw7sOz9Ut9mPsxvl2DC1rSfODNU9d++eobrRdWIjHBumPi6M2uz7+kYxulZfePieobp/8qlfOdta233ldmcoAAAAAN0ECgAAAEA3gQIAAADQTaAAAAAAdBMoAAAAAN0ECgAAAEA3gQIAAADQTaAAAAAAdBMoAAAAAN0ECgAAAEA3gQIAAADQTaAAAAAAdBMoAAAAAN2qtTbvHnL79te1hV/+xXm3ccs7sO/8vFvYNBZP75x3C2vKXLm6zf66T21hcdrj0/KBGqob3R+ObT81VDfq8MX9k41lX2A9m3pt2Si2njgzVPfcvXtWuZPrG12rn7zvkVXu5PpG1lxrJ1PYcWRpqO6xdvxsa233ldudoQAAAAB0EygAAAAA3QQKAAAAQDeBAgAAANBNoAAAAAB0EygAAAAA3QQKAAAAQDeBAgAAANBNoAAAAAB0EygAAAAA3QQKAAAAQLcbBgpV9eGqeqaqvnTZtv+qqv6gqn6vqk5U1csvu++9VXWhqr5SVYfWqnEAAABgfm7mDIWPJHnLFds+l+RHW2v/ZpJ/luS9SVJV9yR5e5Kds5p/UFUvXrVuAQAAgHVhy40e0FpbrKrXX7HtH192cynJ22ZfvzXJJ1pr303yR1V1Icmbkpy+3hh3vOz/yYF95zvano/F0zvn3cKa2uzPb8RGmJfzMDpXRn+ex7afGqo7fHH/UN1Gsenn575ph1senNej+8Ndm3jN3fRzc4NwXL+65QM17xbWpwN7593Bmtr/0LsmHW8jzLOp12pr0ua0GtdQ+I+T/G+zr1+T5OuX3ffUbNsLVNUDVfV4VT3+L7/zL1ehDQAAAGAqKwoUquo/T/J8ko/11rbWPtha291a2/3Sl790JW0AAAAAE7vhRx6upar+oyR/I8nB1lqbbX46yesue9hrZ9sAAACATWToDIWqekuSv53kZ1trz11212eSvL2qbq+qNyS5O8lvr7xNAAAAYD254RkKVfXxJG9O8sqqeirJ+3LprzrcnuRzVZUkS621B1tr56vqk0meyKWPQjzUWvvztWoeAAAAmI+b+SsP77jK5g9d5/G/luTXVtIUAAAAsL6txl95AAAAAG4xAgUAAACgm0ABAAAA6CZQAAAAALoJFAAAAIBuAgUAAACgm0ABAAAA6FattXn3kFf+1Ve2v/7Rn+2uO7b91NB4hy/uH6rj6hZP75x3C7e8A/vOz7sFVmDqfWijzJeNssZbA+dvo8xp1gf7LJvNZl8D7bOra2Fx7Pf/f/KpXznbWtt95XZnKAAAAADdBAoAAABAN4ECAAAA0E2gAAAAAHQTKAAAAADdBAoAAABAN4ECAAAA0E2gAAAAAHQTKAAAAADdBAoAAABAN4ECAAAA0E2gAAAAAHQTKAAAAADdtsy7gZU4tLBrsPLZVe3jRhaW7ph0vFHLe8d+LjuyNFR34ejeobqFxTZUN2L5QE02FreuA/vOTzre4umdk443+vwOX9w/VHds+6lJx9vsrx+rZ3RubhSj+xCwsYz+zvDcvXuG6hYy3Xv/JNl64syk4210zlAAAAAAugkUAAAAgG4CBQAAAKCbQAEAAADoJlAAAAAAugkUAAAAgG4CBQAAAKCbQAEAAADoJlAAAAAAugkUAAAAgG4CBQAAAKCbQAEAAADoJlAAAAAAulVrbd495M7a1vbUwXm3AcAmsLB0x6TjLe99dtLxpnbh6N7JxtpxZGmysbi2qfehUYund867BWK/hVvFY+342dba7iu3O0MBAAAA6CZQAAAAALoJFAAAAIBuAgUAAACgm0ABAAAA6CZQAAAAALoJFAAAAIBuAgUAAACgm0ABAAAA6CZQAAAAALoJFAAAAIBuAgUAAACgm0ABAAAA6FattXn3kKr6kyRfS/LKJP98zu2wMZgr9DBfuFnmCj3MF26WuUIP84WbNeVc+ddbaz985cZ1ESh8X1U93lrbPe8+WP/MFXqYL9wsc4Ue5gs3y1yhh/nCzVoPc8VHHgAAAIBuAgUAAACg23oLFD447wbYMMwVepgv3CxzhR7mCzfLXKGH+cLNmvtcWVfXUAAAAAA2hvV2hgIAAACwAayLQKGq3lJVX6mqC1X1nnn3w/pTVX9cVb9fVeeq6vHZtm1V9bmq+sPZ/6+Yd59Mr6o+XFXPVNWXLtt21blRl/z92Vrze1X1E/PrnHm4xnz51ap6era+nKuqn7nsvvfO5stXqurQfLpmHqrqdVX1hap6oqrOV9UvzLZbX/hXXGeuWFt4gap6aVX9dlX9H7P58l/Mtr+hqs7M5sWjVXXbbPvts9sXZve/fp79M53rzJWPVNUfXba27Jptn8txaO6BQlW9OMn7k/x0knuSvKOq7plvV2Pbkn8AAASVSURBVKxT/05rbddlfxrlPUk+31q7O8nnZ7e59XwkyVuu2HatufHTSe6e/XsgyW9M1CPrx0fywvmSJEdn68uu1tpnk2R2LHp7kp2zmn8wO2Zxa3g+yS+31u5JsjfJQ7M5YX3hSteaK4m1hRf6bpKfaq39WJJdSd5SVXuTPJxL82VHkm8neefs8e9M8u3Z9qOzx3FruNZcSZJfuWxtOTfbNpfj0NwDhSRvSnKhtfbV1tr/m+QTSd46557YGN6a5KOzrz+a5D+YYy/MSWttMcn/dcXma82Ntyb5n9olS0leXlWvnqZT1oNrzJdreWuST7TWvtta+6MkF3LpmMUtoLX2jdbaP519/WySLyd5TawvXOE6c+VarC23sNka8X/Pbr5k9q8l+akkx2fbr1xbvr/mHE9ysKpqonaZo+vMlWuZy3FoPQQKr0ny9ctuP5XrL8LcmlqSf1xVZ6vqgdm2V7XWvjH7+ptJXjWf1liHrjU3rDdcy8/PTg/88GUfnzJfSJLMTjH+8SRnYn3hOq6YK4m1hauoqhdX1bkkzyT5XJInk3yntfb87CGXz4kfzJfZ/X+a5Iem7Zh5uXKutNa+v7b82mxtOVpVt8+2zWVtWQ+BAtyMf7u19hO5dCrPQ1V14PI726U/V+JPlvAC5gY34TeS3JVLpxN+I8l/M992WE+q6i8l+YdJfrG19meX32d94XJXmSvWFq6qtfbnrbVdSV6bS2envHHOLbFOXTlXqupHk7w3l+bMv5VkW5J3z7HFdREoPJ3kdZfdfu1sG/xAa+3p2f/PJDmRS4vvt75/Gs/s/2fm1yHrzLXmhvWGF2itfWt2wP6LJP9D/v9Tj82XW1xVvSSXfkH8WGvtU7PN1hde4GpzxdrCjbTWvpPkC0n25dLp6Vtmd10+J34wX2b3/+Uk/+fErTJnl82Vt8w+ZtVaa99NcixzXlvWQ6DwO0nunl3Z9LZcukjNZ+bcE+tIVb2squ74/tdJ/r0kX8qleXL/7GH3J/n0fDpkHbrW3PhMkr81uwru3iR/etmpy9yirvh84b25tL4kl+bL22dX2H5DLl3k6Len7o/5mH1G+UNJvtxa+7uX3WV94V9xrblibeFqquqHq+rls6//tSR/LZeuu/GFJG+bPezKteX7a87bkvzW7OwoNrlrzJU/uCzUrly61sbla8vkx6EtN37I2mqtPV9VP5/kZJIXJ/lwa+38nNtifXlVkhOz689sSfK/tNb+UVX9TpJPVtU7k3wtyc/NsUfmpKo+nuTNSV5ZVU8leV+SX8/V58Znk/xMLl0A67kkhydvmLm6xnx58+xPLrUkf5zkXUnSWjtfVZ9M8kQuXcX9odban8+jb+biJ5P8zSS/P/v8apL8nVhfeKFrzZV3WFu4ilcn+ejsL3u8KMknW2u/WVVPJPlEVf2XSX43l0KqzP7/n6vqQi5dVPjt82iaubjWXPmtqvrhJJXkXJIHZ4+fy3GoBFwAAABAr/XwkQcAAABggxEoAAAAAN0ECgAAAEA3gQIAAADQTaAAAAAAdBMoAAAAAN0ECgAAAEA3gQIAAADQ7f8DfAsZljd9yWYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1296x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(18, 8))\n",
    "plt.imshow(outdeg_centrality_map.swapaxes(1, 0))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The 'sources' will be the points where has maximum $(outdegree) - (indegree)$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(20, 55),\n",
       " (20, 70),\n",
       " (35, 50),\n",
       " (40, 80),\n",
       " (115, 80),\n",
       " (130, 45),\n",
       " (235, 75),\n",
       " (255, 65),\n",
       " (340, 15),\n",
       " (350, 35),\n",
       " (350, 80)]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "outdeg = nx.algorithms.out_degree_centrality(graph)\n",
    "indeg = nx.algorithms.in_degree_centrality(graph)\n",
    "delta = dict()\n",
    "for k in outdeg.keys():\n",
    "    delta[k] = outdeg[k] - indeg[k]\n",
    "source_nodes = [k for k, v in delta.items() if v == max(delta.values())]\n",
    "source_nodes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The 'sinks' will be points where has minimum $(outdegree) - (indegree)$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(25, 85),\n",
       " (35, 85),\n",
       " (45, 75),\n",
       " (80, 80),\n",
       " (85, 55),\n",
       " (100, 55),\n",
       " (105, 70),\n",
       " (135, 85),\n",
       " (160, 70),\n",
       " (200, 80),\n",
       " (215, 85),\n",
       " (230, 90),\n",
       " (235, 45)]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sink_nodes = [k for k, v in delta.items() if v == min(delta.values())]\n",
    "sink_nodes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, let's use min-cut max-flow theorm to determine points to deploy robots."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{(25, 70),\n",
       " (55, 80),\n",
       " (75, 65),\n",
       " (75, 70),\n",
       " (75, 80),\n",
       " (80, 60),\n",
       " (85, 60),\n",
       " (90, 60),\n",
       " (100, 70),\n",
       " (100, 75),\n",
       " (105, 75),\n",
       " (110, 75),\n",
       " (155, 70),\n",
       " (155, 75),\n",
       " (225, 60),\n",
       " (230, 45)}"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cut_points = set()\n",
    "for source_node in source_nodes:\n",
    "    for sink_node in sink_nodes:\n",
    "        cut_points |= set(nx.minimum_node_cut(graph, source_node, sink_node))\n",
    "cut_points"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, plot the cut points."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "nan_mask = dataset.variables['uf'][0,0].mask\n",
    "x_axis = np.argwhere(nan_mask)[:,1]\n",
    "y_axis = 139 - np.argwhere(nan_mask)[:,0]\n",
    "area_points = []\n",
    "for x0, y0 in cut_points:\n",
    "    for x in range(x0, x0 + 5):\n",
    "        for y in range(y0, y0 + 5):\n",
    "            area_points.append((x, y))\n",
    "plt.figure(figsize=(18, 8))\n",
    "plt.scatter(x_axis, y_axis, 0.1)\n",
    "plt.scatter(np.array(area_points)[:,0], np.array(area_points)[:,1], 5)\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}