Data_Viz_DiagrammeR_and_wordcloud.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# DiagrammeR Viz in Jupyter Python notebook\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-25T07:29:16.918794Z",
     "start_time": "2017-12-25T07:29:15.867904Z"
    }
   },
   "outputs": [],
   "source": [
    "import rpy2\n",
    "\n",
    "# Enables R line magic (%R) and cell magic (%%R)\n",
    "%load_ext rpy2.ipython \n",
    "\n",
    "# # r enables calls to r objects and \n",
    "# # pandas2ri allows conversion both ways\n",
    "# from rpy2.robjects import r, pandas2ri    \n",
    "\n",
    "# Ignore rpy2 RRuntimeWarning complaining\n",
    "# about packages, they work fine!\n",
    "import warnings\n",
    "from rpy2.rinterface import RRuntimeWarning\n",
    "warnings.filterwarnings(\"ignore\", category=RRuntimeWarning)\n",
    "\n",
    "# Enables rendering a string of HTML code \n",
    "# as an image\n",
    "from IPython.display import display, HTML"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The code cell below shows an example of DiagrammeR. It's mostly gleaned from the source below the image with a few added aesthetic changes. \n",
    "\n",
    "\n",
    "To get the code below to work correctly and print inline instead of to a new window on Mac OSX, you will likely need to do the following first:\n",
    "\n",
    "In a terminal \"brew install v8@3.15\"\n",
    "\n",
    "In R \"install.packages(\"v8\");install.packages(\"DiagrammeR\");install.packages(\"DiagrammeRsvg\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-25T07:29:20.602877Z",
     "start_time": "2017-12-25T07:29:16.922375Z"
    }
   },
   "outputs": [],
   "source": [
    "%%R\n",
    "\n",
    "library(DiagrammeR)\n",
    "library(DiagrammeRsvg) # provides export_svg command\n",
    "\n",
    "plot.script <- paste(\"digraph 'ER-D' {\",\n",
    "                         \"graph [layout = neato]\",\n",
    "                         \"node [\",\n",
    "                             \"style = filled,\",\n",
    "                             \"penwidth = 3\",\n",
    "                         \"]\",\n",
    "                         \"node [\",\n",
    "                             \"color = royalblue,\",\n",
    "                             \"fillcolor = cornflowerblue,\",\n",
    "                             \"shape = egg,\",\n",
    "                             \"width = 01.00\",\n",
    "                         \"]\",\n",
    "                         \"aisles\",\n",
    "                         \"departments\",\n",
    "                         \"orders\",\n",
    "                         \"products\",\n",
    "                         \"order_products__\",\n",
    "                         \"node [\",\n",
    "                             \"color = forestgreen,\",\n",
    "                             \"fillcolor = darkseagreen,\",\n",
    "                             \"shape = box,\",\n",
    "                             \"width = 0.40,\",\n",
    "                             \"height = 0.30,\",\n",
    "                             \"fontsize = 9\",\n",
    "                         \"]\",\n",
    "                         \"edge [\",\n",
    "                             \"color = grey,\",\n",
    "                             \"arrowhead = none\",\n",
    "                         \"]\",\n",
    "                         \"aisles -> {\",\n",
    "                             \"aisle_id\",\n",
    "                             \"aisle\",\n",
    "                         \"}\",\n",
    "                         \"departments -> {\",\n",
    "                             \"department_id\",\n",
    "                             \"department\",\n",
    "                         \"}\",       \n",
    "                         \"orders -> {\",\n",
    "                             \"order_id\",\n",
    "                             \"user_id\",\n",
    "                             \"eval_set\",\n",
    "                             \"order_number\",\n",
    "                             \"order_dow\",\n",
    "                             \"order_hour_of_day\",\n",
    "                             \"days_since_prior_order\",\n",
    "                         \"}\",\n",
    "                         \"products -> {\",\n",
    "                             \"product_id\",\n",
    "                             \"product_name\",\n",
    "                             \"aisle_id\",\n",
    "                             \"department_id\",\n",
    "                         \"}\",\n",
    "                         \"order_products__ -> {\",\n",
    "                             \"order_id\",\n",
    "                             \"product_id\",\n",
    "                             \"add_to_cart_order\",\n",
    "                             \"reordered\",\n",
    "                         \"}\",\n",
    "                     \"}\", sep=\" \")\n",
    "svg <- export_svg(grViz(plot.script))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-25T07:29:20.623359Z",
     "start_time": "2017-12-25T07:29:20.605855Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<?xml version=\"1.0\" encoding=\"UTF-8\" standalone=\"no\"?>\n",
       "<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\n",
       " \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\n",
       "<!-- Generated by graphviz version 2.40.1 (20161225.0304)\n",
       " -->\n",
       "<!-- Title: ER&#45;D Pages: 1 -->\n",
       "<svg width=\"669pt\" height=\"364pt\"\n",
       " viewBox=\"0.00 0.00 668.78 363.66\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       "<g id=\"graph0\" class=\"graph\" transform=\"scale(1 1) rotate(0) translate(4 359.6611)\">\n",
       "<title>ER&#45;D</title>\n",
       "<polygon fill=\"#ffffff\" stroke=\"transparent\" points=\"-4,4 -4,-359.6611 664.7786,-359.6611 664.7786,4 -4,4\"/>\n",
       "<!-- aisles -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>aisles</title>\n",
       "<polygon fill=\"#6495ed\" stroke=\"#4169e1\" stroke-width=\"3\" points=\"83.5974,-287.4672 85.9348,-287.5657 88.2479,-287.7132 90.5275,-287.9092 92.7642,-288.1532 94.949,-288.4445 97.0735,-288.7823 99.1292,-289.1657 101.1084,-289.5937 103.0037,-290.0651 104.8084,-290.5785 106.516,-291.1326 108.121,-291.7258 109.6182,-292.3566 111.0033,-293.0232 112.2724,-293.7238 113.4223,-294.4564 114.4506,-295.219 115.3556,-296.0096 116.136,-296.826 116.7913,-297.666 117.3217,-298.5272 117.7279,-299.4073 118.0112,-300.3038 118.1735,-301.2144 118.2171,-302.1365 118.1451,-303.0676 117.9605,-304.0051 117.6672,-304.9465 117.2692,-305.8892 116.7709,-306.8306 116.1768,-307.7681 115.4919,-308.6992 114.7211,-309.6213 113.8694,-310.5319 112.9422,-311.4284 111.9445,-312.3085 110.8814,-313.1697 109.7582,-314.0096 108.5799,-314.826 107.3512,-315.6166 106.0769,-316.3793 104.7614,-317.1119 103.4092,-317.8125 102.0241,-318.479 100.6101,-319.1098 99.1706,-319.7031 97.7089,-320.2572 96.228,-320.7706 94.7307,-321.242 93.2193,-321.6699 91.6962,-322.0534 90.1633,-322.3912 88.6224,-322.6825 87.0751,-322.9265 85.5227,-323.1225 83.9666,-323.27 82.4078,-323.3685 80.8476,-323.4178 79.2868,-323.4178 77.7265,-323.3685 76.1678,-323.27 74.6117,-323.1225 73.0593,-322.9265 71.5119,-322.6825 69.971,-322.3912 68.4381,-322.0534 66.915,-321.6699 65.4037,-321.242 63.9063,-320.7706 62.4255,-320.2572 60.9638,-319.7031 59.5243,-319.1098 58.1102,-318.479 56.7252,-317.8125 55.3729,-317.1119 54.0575,-316.3793 52.7832,-315.6166 51.5545,-314.826 50.3761,-314.0096 49.2529,-313.1697 48.1899,-312.3085 47.1922,-311.4284 46.2649,-310.5319 45.4133,-309.6213 44.6424,-308.6992 43.9575,-307.7681 43.3634,-306.8306 42.8651,-305.8892 42.4671,-304.9465 42.1738,-304.0051 41.9893,-303.0676 41.9172,-302.1365 41.9609,-301.2144 42.1231,-300.3038 42.4065,-299.4073 42.8126,-298.5272 43.343,-297.666 43.9984,-296.826 44.7787,-296.0096 45.6837,-295.219 46.712,-294.4564 47.862,-293.7238 49.131,-293.0232 50.5161,-292.3566 52.0133,-291.7258 53.6183,-291.1326 55.326,-290.5785 57.1306,-290.0651 59.026,-289.5937 61.0052,-289.1657 63.0609,-288.7823 65.1853,-288.4445 67.3702,-288.1532 69.6069,-287.9092 71.8864,-287.7132 74.1996,-287.5657 76.5369,-287.4672 78.8888,-287.4178 81.2455,-287.4178 83.5974,-287.4672\"/>\n",
       "<text text-anchor=\"middle\" x=\"80.0672\" y=\"-301.2178\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#000000\">aisles</text>\n",
       "</g>\n",
       "<!-- aisle_id -->\n",
       "<g id=\"node6\" class=\"node\">\n",
       "<title>aisle_id</title>\n",
       "<polygon fill=\"#8fbc8f\" stroke=\"#228b22\" stroke-width=\"3\" points=\"164.7395,-267.3372 120.7523,-267.3372 120.7523,-245.3372 164.7395,-245.3372 164.7395,-267.3372\"/>\n",
       "<text text-anchor=\"middle\" x=\"142.7459\" y=\"-253.6372\" font-family=\"Times,serif\" font-size=\"9.00\" fill=\"#000000\">aisle_id</text>\n",
       "</g>\n",
       "<!-- aisles&#45;&gt;aisle_id -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>aisles&#45;&gt;aisle_id</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M100.5905,-289.347C109.7304,-282.19 120.2692,-273.9375 128.5231,-267.4743\"/>\n",
       "</g>\n",
       "<!-- aisle -->\n",
       "<g id=\"node7\" class=\"node\">\n",
       "<title>aisle</title>\n",
       "<polygon fill=\"#8fbc8f\" stroke=\"#228b22\" stroke-width=\"3\" points=\"32.7425,-355.6611 -.25,-355.6611 -.25,-333.6611 32.7425,-333.6611 32.7425,-355.6611\"/>\n",
       "<text text-anchor=\"middle\" x=\"16.2463\" y=\"-341.9611\" font-family=\"Times,serif\" font-size=\"9.00\" fill=\"#000000\">aisle</text>\n",
       "</g>\n",
       "<!-- aisles&#45;&gt;aisle -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>aisles&#45;&gt;aisle</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M58.4603,-318.7038C50.0707,-323.8625 40.6167,-329.6758 32.7855,-334.4911\"/>\n",
       "</g>\n",
       "<!-- departments -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>departments</title>\n",
       "<polygon fill=\"#6495ed\" stroke=\"#4169e1\" stroke-width=\"3\" points=\"154.0998,-55.5635 157.9,-55.662 161.6609,-55.8095 165.3671,-56.0055 169.0036,-56.2495 172.556,-56.5408 176.01,-56.8786 179.3523,-57.2621 182.5702,-57.69 185.6518,-58.1614 188.5859,-58.6748 191.3623,-59.2289 193.9718,-59.8222 196.4061,-60.4529 198.658,-61.1195 200.7213,-61.8201 202.591,-62.5527 204.2629,-63.3153 205.7342,-64.106 207.003,-64.9224 208.0685,-65.7623 208.9309,-66.6235 209.5913,-67.5036 210.0519,-68.4001 210.3158,-69.3107 210.3867,-70.2328 210.2695,-71.1639 209.9695,-72.1014 209.4926,-73.0428 208.8455,-73.9855 208.0353,-74.9269 207.0695,-75.8644 205.9559,-76.7955 204.7026,-77.7176 203.318,-78.6282 201.8104,-79.5247 200.1882,-80.4048 198.4599,-81.266 196.6337,-82.106 194.7179,-82.9224 192.7202,-83.713 190.6483,-84.4756 188.5096,-85.2082 186.311,-85.9088 184.0591,-86.5754 181.76,-87.2062 179.4196,-87.7994 177.0431,-88.3535 174.6354,-88.8669 172.2009,-89.3383 169.7437,-89.7663 167.2673,-90.1497 164.7751,-90.4875 162.2697,-90.7788 159.754,-91.0228 157.23,-91.2188 154.7,-91.3663 152.1657,-91.4648 149.6289,-91.5142 147.0913,-91.5142 144.5545,-91.4648 142.0202,-91.3663 139.4902,-91.2188 136.9662,-91.0228 134.4504,-90.7788 131.9451,-90.4875 129.4529,-90.1497 126.9765,-89.7663 124.5193,-89.3383 122.0848,-88.8669 119.6771,-88.3535 117.3006,-87.7994 114.9602,-87.2062 112.6611,-86.5754 110.4092,-85.9088 108.2106,-85.2082 106.0719,-84.4756 104,-83.713 102.0023,-82.9224 100.0864,-82.106 98.2603,-81.266 96.532,-80.4048 94.9098,-79.5247 93.4022,-78.6282 92.0176,-77.7176 90.7643,-76.7955 89.6507,-75.8644 88.6849,-74.9269 87.8746,-73.9855 87.2276,-73.0428 86.7507,-72.1014 86.4506,-71.1639 86.3334,-70.2328 86.4044,-69.3107 86.6683,-68.4001 87.1289,-67.5036 87.7893,-66.6235 88.6517,-65.7623 89.7171,-64.9224 90.9859,-64.106 92.4572,-63.3153 94.1292,-62.5527 95.9988,-61.8201 98.0622,-61.1195 100.3141,-60.4529 102.7484,-59.8222 105.3579,-59.2289 108.1343,-58.6748 111.0684,-58.1614 114.1499,-57.69 117.3679,-57.2621 120.7102,-56.8786 124.1642,-56.5408 127.7165,-56.2495 131.3531,-56.0055 135.0593,-55.8095 138.8202,-55.662 142.6204,-55.5635 146.4442,-55.5142 150.276,-55.5142 154.0998,-55.5635\"/>\n",
       "<text text-anchor=\"middle\" x=\"148.3601\" y=\"-69.3142\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#000000\">departments</text>\n",
       "</g>\n",
       "<!-- department_id -->\n",
       "<g id=\"node8\" class=\"node\">\n",
       "<title>department_id</title>\n",
       "<polygon fill=\"#8fbc8f\" stroke=\"#228b22\" stroke-width=\"3\" points=\"223.0343,-152.3532 155.0627,-152.3532 155.0627,-130.3532 223.0343,-130.3532 223.0343,-152.3532\"/>\n",
       "<text text-anchor=\"middle\" x=\"189.0485\" y=\"-138.6532\" font-family=\"Times,serif\" font-size=\"9.00\" fill=\"#000000\">department_id</text>\n",
       "</g>\n",
       "<!-- departments&#45;&gt;department_id -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>departments&#45;&gt;department_id</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M159.0505,-91.3381C166.3967,-103.5862 175.9291,-119.4795 182.2554,-130.0272\"/>\n",
       "</g>\n",
       "<!-- department -->\n",
       "<g id=\"node9\" class=\"node\">\n",
       "<title>department</title>\n",
       "<polygon fill=\"#8fbc8f\" stroke=\"#228b22\" stroke-width=\"3\" points=\"133.8282,-22 76.8546,-22 76.8546,0 133.8282,0 133.8282,-22\"/>\n",
       "<text text-anchor=\"middle\" x=\"105.3414\" y=\"-8.3\" font-family=\"Times,serif\" font-size=\"9.00\" fill=\"#000000\">department</text>\n",
       "</g>\n",
       "<!-- departments&#45;&gt;department -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>departments&#45;&gt;department</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M135.9173,-55.4325C128.5722,-44.7587 119.5005,-31.5758 113.1536,-22.3526\"/>\n",
       "</g>\n",
       "<!-- orders -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>orders</title>\n",
       "<polygon fill=\"#6495ed\" stroke=\"#4169e1\" stroke-width=\"3\" points=\"548.6389,-177.1762 550.9762,-177.2748 553.2894,-177.4222 555.5689,-177.6182 557.8056,-177.8622 559.9905,-178.1535 562.1149,-178.4913 564.1706,-178.8748 566.1498,-179.3027 568.0452,-179.7741 569.8498,-180.2875 571.5575,-180.8416 573.1624,-181.4349 574.6597,-182.0657 576.0447,-182.7322 577.3138,-183.4328 578.4638,-184.1654 579.4921,-184.9281 580.397,-185.7187 581.1774,-186.5351 581.8327,-187.375 582.3631,-188.2362 582.7693,-189.1163 583.0526,-190.0128 583.2149,-190.9234 583.2586,-191.8455 583.1865,-192.7766 583.002,-193.7141 582.7087,-194.6555 582.3107,-195.5982 581.8123,-196.5396 581.2183,-197.4772 580.5334,-198.4082 579.7625,-199.3303 578.9109,-200.2409 577.9836,-201.1375 576.9859,-202.0175 575.9229,-202.8787 574.7997,-203.7187 573.6213,-204.5351 572.3926,-205.3257 571.1183,-206.0883 569.8029,-206.8209 568.4506,-207.5215 567.0656,-208.1881 565.6515,-208.8189 564.212,-209.4121 562.7503,-209.9662 561.2695,-210.4797 559.7721,-210.951 558.2608,-211.379 556.7377,-211.7624 555.2048,-212.1002 553.6639,-212.3915 552.1165,-212.6355 550.5641,-212.8315 549.008,-212.979 547.4493,-213.0775 545.889,-213.1269 544.3282,-213.1269 542.768,-213.0775 541.2092,-212.979 539.6531,-212.8315 538.1007,-212.6355 536.5534,-212.3915 535.0125,-212.1002 533.4796,-211.7624 531.9565,-211.379 530.4451,-210.951 528.9478,-210.4797 527.4669,-209.9662 526.0052,-209.4121 524.5657,-208.8189 523.1517,-208.1881 521.7666,-207.5215 520.4144,-206.8209 519.0989,-206.0883 517.8246,-205.3257 516.5959,-204.5351 515.4175,-203.7187 514.2943,-202.8787 513.2313,-202.0175 512.2336,-201.1375 511.3063,-200.2409 510.4547,-199.3303 509.6839,-198.4082 508.9989,-197.4772 508.4049,-196.5396 507.9066,-195.5982 507.5086,-194.6555 507.2153,-193.7141 507.0307,-192.7766 506.9586,-191.8455 507.0023,-190.9234 507.1646,-190.0128 507.4479,-189.1163 507.8541,-188.2362 508.3845,-187.375 509.0398,-186.5351 509.8202,-185.7187 510.7251,-184.9281 511.7535,-184.1654 512.9034,-183.4328 514.1725,-182.7322 515.5575,-182.0657 517.0548,-181.4349 518.6598,-180.8416 520.3674,-180.2875 522.1721,-179.7741 524.0674,-179.3027 526.0466,-178.8748 528.1023,-178.4913 530.2268,-178.1535 532.4116,-177.8622 534.6483,-177.6182 536.9279,-177.4222 539.241,-177.2748 541.5784,-177.1762 543.9302,-177.1269 546.287,-177.1269 548.6389,-177.1762\"/>\n",
       "<text text-anchor=\"middle\" x=\"545.1086\" y=\"-190.9269\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#000000\">orders</text>\n",
       "</g>\n",
       "<!-- order_id -->\n",
       "<g id=\"node10\" class=\"node\">\n",
       "<title>order_id</title>\n",
       "<polygon fill=\"#8fbc8f\" stroke=\"#228b22\" stroke-width=\"3\" points=\"484.8071,-234.8603 437.8251,-234.8603 437.8251,-212.8603 484.8071,-212.8603 484.8071,-234.8603\"/>\n",
       "<text text-anchor=\"middle\" x=\"461.3161\" y=\"-221.1603\" font-family=\"Times,serif\" font-size=\"9.00\" fill=\"#000000\">order_id</text>\n",
       "</g>\n",
       "<!-- orders&#45;&gt;order_id -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>orders&#45;&gt;order_id</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M516.7403,-204.8547C506.2951,-208.4365 494.5932,-212.4492 484.6431,-215.8612\"/>\n",
       "</g>\n",
       "<!-- user_id -->\n",
       "<g id=\"node11\" class=\"node\">\n",
       "<title>user_id</title>\n",
       "<polygon fill=\"#8fbc8f\" stroke=\"#228b22\" stroke-width=\"3\" points=\"579.5628,-119.6286 536.5723,-119.6286 536.5723,-97.6286 579.5628,-97.6286 579.5628,-119.6286\"/>\n",
       "<text text-anchor=\"middle\" x=\"558.0675\" y=\"-105.9286\" font-family=\"Times,serif\" font-size=\"9.00\" fill=\"#000000\">user_id</text>\n",
       "</g>\n",
       "<!-- orders&#45;&gt;user_id -->\n",
       "<g id=\"edge6\" class=\"edge\">\n",
       "<title>orders&#45;&gt;user_id</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M547.8569,-176.7828C550.4424,-159.525 554.2303,-134.2415 556.3916,-119.8149\"/>\n",
       "</g>\n",
       "<!-- eval_set -->\n",
       "<g id=\"node12\" class=\"node\">\n",
       "<title>eval_set</title>\n",
       "<polygon fill=\"#8fbc8f\" stroke=\"#228b22\" stroke-width=\"3\" points=\"531.0154,-136.4918 485.0349,-136.4918 485.0349,-114.4918 531.0154,-114.4918 531.0154,-136.4918\"/>\n",
       "<text text-anchor=\"middle\" x=\"508.0251\" y=\"-122.7918\" font-family=\"Times,serif\" font-size=\"9.00\" fill=\"#000000\">eval_set</text>\n",
       "</g>\n",
       "<!-- orders&#45;&gt;eval_set -->\n",
       "<g id=\"edge7\" class=\"edge\">\n",
       "<title>orders&#45;&gt;eval_set</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M535.5587,-177.1941C528.7773,-164.4601 519.8786,-147.7501 514.0593,-136.8227\"/>\n",
       "</g>\n",
       "<!-- order_number -->\n",
       "<g id=\"node13\" class=\"node\">\n",
       "<title>order_number</title>\n",
       "<polygon fill=\"#8fbc8f\" stroke=\"#228b22\" stroke-width=\"3\" points=\"660.7684,-207.0363 593.8092,-207.0363 593.8092,-185.0363 660.7684,-185.0363 660.7684,-207.0363\"/>\n",
       "<text text-anchor=\"middle\" x=\"627.2888\" y=\"-193.3363\" font-family=\"Times,serif\" font-size=\"9.00\" fill=\"#000000\">order_number</text>\n",
       "</g>\n",
       "<!-- orders&#45;&gt;order_number -->\n",
       "<g id=\"edge8\" class=\"edge\">\n",
       "<title>orders&#45;&gt;order_number</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M582.3515,-195.539C586.0888,-195.5804 589.8688,-195.6222 593.5666,-195.6631\"/>\n",
       "</g>\n",
       "<!-- order_dow -->\n",
       "<g id=\"node14\" class=\"node\">\n",
       "<title>order_dow</title>\n",
       "<polygon fill=\"#8fbc8f\" stroke=\"#228b22\" stroke-width=\"3\" points=\"637.7089,-263.0592 582.7401,-263.0592 582.7401,-241.0592 637.7089,-241.0592 637.7089,-263.0592\"/>\n",
       "<text text-anchor=\"middle\" x=\"610.2245\" y=\"-249.3592\" font-family=\"Times,serif\" font-size=\"9.00\" fill=\"#000000\">order_dow</text>\n",
       "</g>\n",
       "<!-- orders&#45;&gt;order_dow -->\n",
       "<g id=\"edge9\" class=\"edge\">\n",
       "<title>orders&#45;&gt;order_dow</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M562.5585,-210.3837C573.5587,-220.0015 587.5155,-232.2042 597.4786,-240.9152\"/>\n",
       "</g>\n",
       "<!-- order_hour_of_day -->\n",
       "<g id=\"node15\" class=\"node\">\n",
       "<title>order_hour_of_day</title>\n",
       "<polygon fill=\"#8fbc8f\" stroke=\"#228b22\" stroke-width=\"3\" points=\"605.8712,-285.753 519.9194,-285.753 519.9194,-263.753 605.8712,-263.753 605.8712,-285.753\"/>\n",
       "<text text-anchor=\"middle\" x=\"562.8953\" y=\"-272.053\" font-family=\"Times,serif\" font-size=\"9.00\" fill=\"#000000\">order_hour_of_day</text>\n",
       "</g>\n",
       "<!-- orders&#45;&gt;order_hour_of_day -->\n",
       "<g id=\"edge10\" class=\"edge\">\n",
       "<title>orders&#45;&gt;order_hour_of_day</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M549.1447,-213.1953C552.5886,-228.6126 557.4342,-250.3053 560.3587,-263.3974\"/>\n",
       "</g>\n",
       "<!-- days_since_prior_order -->\n",
       "<g id=\"node16\" class=\"node\">\n",
       "<title>days_since_prior_order</title>\n",
       "<polygon fill=\"#8fbc8f\" stroke=\"#228b22\" stroke-width=\"3\" points=\"657.1788,-150.4443 556.2443,-150.4443 556.2443,-128.4443 657.1788,-128.4443 657.1788,-150.4443\"/>\n",
       "<text text-anchor=\"middle\" x=\"606.7116\" y=\"-136.7443\" font-family=\"Times,serif\" font-size=\"9.00\" fill=\"#000000\">days_since_prior_order</text>\n",
       "</g>\n",
       "<!-- orders&#45;&gt;days_since_prior_order -->\n",
       "<g id=\"edge11\" class=\"edge\">\n",
       "<title>orders&#45;&gt;days_since_prior_order</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M563.5915,-178.4203C573.5149,-169.4506 585.4806,-158.6349 594.287,-150.6748\"/>\n",
       "</g>\n",
       "<!-- products -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>products</title>\n",
       "<polygon fill=\"#6495ed\" stroke=\"#4169e1\" stroke-width=\"3\" points=\"224.5081,-201.9348 227.4081,-202.0334 230.2782,-202.1808 233.1066,-202.3768 235.8818,-202.6208 238.5927,-202.9121 241.2287,-203.25 243.7793,-203.6334 246.2351,-204.0614 248.5867,-204.5327 250.8258,-205.0461 252.9446,-205.6002 254.9361,-206.1935 256.7938,-206.8243 258.5123,-207.4909 260.0869,-208.1914 261.5137,-208.924 262.7897,-209.6867 263.9125,-210.4773 264.8808,-211.2937 265.6939,-212.1336 266.352,-212.9948 266.856,-213.8749 267.2075,-214.7715 267.4088,-215.682 267.463,-216.6041 267.3736,-217.5352 267.1446,-218.4727 266.7807,-219.4141 266.2869,-220.3568 265.6686,-221.2982 264.9315,-222.2358 264.0816,-223.1669 263.1252,-224.0889 262.0685,-224.9995 260.918,-225.8961 259.6801,-226.7761 258.3611,-227.6373 256.9675,-228.4773 255.5054,-229.2937 253.9809,-230.0843 252.3998,-230.8469 250.7676,-231.5796 249.0898,-232.2801 247.3712,-232.9467 245.6168,-233.5775 243.8307,-234.1707 242.0171,-234.7248 240.1796,-235.2383 238.3218,-235.7096 236.4466,-236.1376 234.5568,-236.521 232.6548,-236.8588 230.7429,-237.1502 228.823,-237.3941 226.8969,-237.5901 224.9661,-237.7376 223.032,-237.8361 221.0961,-237.8855 219.1596,-237.8855 217.2236,-237.8361 215.2896,-237.7376 213.3588,-237.5901 211.4327,-237.3941 209.5128,-237.1502 207.6009,-236.8588 205.6989,-236.521 203.8091,-236.1376 201.9339,-235.7096 200.076,-235.2383 198.2386,-234.7248 196.425,-234.1707 194.6389,-233.5775 192.8844,-232.9467 191.1659,-232.2801 189.4881,-231.5796 187.8559,-230.8469 186.2748,-230.0843 184.7503,-229.2937 183.2882,-228.4773 181.8946,-227.6373 180.5756,-226.7761 179.3377,-225.8961 178.1871,-224.9995 177.1305,-224.0889 176.174,-223.1669 175.3242,-222.2358 174.5871,-221.2982 173.9688,-220.3568 173.475,-219.4141 173.1111,-218.4727 172.8821,-217.5352 172.7927,-216.6041 172.8468,-215.682 173.0482,-214.7715 173.3997,-213.8749 173.9037,-212.9948 174.5618,-212.1336 175.3749,-211.2937 176.3432,-210.4773 177.466,-209.6867 178.7419,-208.924 180.1687,-208.1914 181.7434,-207.4909 183.4619,-206.8243 185.3196,-206.1935 187.311,-205.6002 189.4298,-205.0461 191.669,-204.5327 194.0206,-204.0614 196.4764,-203.6334 199.027,-203.25 201.6629,-202.9121 204.3739,-202.6208 207.1491,-202.3768 209.9774,-202.1808 212.8475,-202.0334 215.7476,-201.9348 218.6658,-201.8855 221.5899,-201.8855 224.5081,-201.9348\"/>\n",
       "<text text-anchor=\"middle\" x=\"220.1278\" y=\"-215.6855\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#000000\">products</text>\n",
       "</g>\n",
       "<!-- products&#45;&gt;aisle_id -->\n",
       "<g id=\"edge12\" class=\"edge\">\n",
       "<title>products&#45;&gt;aisle_id</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M192.1912,-233.0454C183.1866,-237.2871 173.3378,-241.9265 164.8169,-245.9404\"/>\n",
       "</g>\n",
       "<!-- products&#45;&gt;department_id -->\n",
       "<g id=\"edge13\" class=\"edge\">\n",
       "<title>products&#45;&gt;department_id</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M212.9195,-201.6712C206.8944,-186.4467 198.4968,-165.2276 193.4355,-152.4384\"/>\n",
       "</g>\n",
       "<!-- product_id -->\n",
       "<g id=\"node17\" class=\"node\">\n",
       "<title>product_id</title>\n",
       "<polygon fill=\"#8fbc8f\" stroke=\"#228b22\" stroke-width=\"3\" points=\"333.3731,-242.3035 278.3863,-242.3035 278.3863,-220.3035 333.3731,-220.3035 333.3731,-242.3035\"/>\n",
       "<text text-anchor=\"middle\" x=\"305.8797\" y=\"-228.6035\" font-family=\"Times,serif\" font-size=\"9.00\" fill=\"#000000\">product_id</text>\n",
       "</g>\n",
       "<!-- products&#45;&gt;product_id -->\n",
       "<g id=\"edge14\" class=\"edge\">\n",
       "<title>products&#45;&gt;product_id</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M261.7478,-225.4273C267.3805,-226.1773 273.0422,-226.9311 278.3436,-227.637\"/>\n",
       "</g>\n",
       "<!-- product_name -->\n",
       "<g id=\"node18\" class=\"node\">\n",
       "<title>product_name</title>\n",
       "<polygon fill=\"#8fbc8f\" stroke=\"#228b22\" stroke-width=\"3\" points=\"242.1835,-307.0346 174.2128,-307.0346 174.2128,-285.0346 242.1835,-285.0346 242.1835,-307.0346\"/>\n",
       "<text text-anchor=\"middle\" x=\"208.1982\" y=\"-293.3346\" font-family=\"Times,serif\" font-size=\"9.00\" fill=\"#000000\">product_name</text>\n",
       "</g>\n",
       "<!-- products&#45;&gt;product_name -->\n",
       "<g id=\"edge15\" class=\"edge\">\n",
       "<title>products&#45;&gt;product_name</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M217.3007,-237.9318C215.0131,-252.5335 211.8649,-272.6292 209.934,-284.9543\"/>\n",
       "</g>\n",
       "<!-- order_products__ -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>order_products__</title>\n",
       "<polygon fill=\"#6495ed\" stroke=\"#4169e1\" stroke-width=\"3\" points=\"393.7226,-246.7498 398.8739,-246.8483 403.9719,-246.9958 408.9958,-247.1918 413.9253,-247.4358 418.7406,-247.7271 423.4227,-248.0649 427.9534,-248.4484 432.3154,-248.8763 436.4926,-249.3477 440.4698,-249.8611 444.2334,-250.4152 447.7707,-251.0085 451.0705,-251.6393 454.123,-252.3058 456.92,-253.0064 459.4544,-253.739 461.7207,-254.5017 463.7152,-255.2923 465.4351,-256.1087 466.8794,-256.9486 468.0483,-257.8098 468.9435,-258.6899 469.5679,-259.5864 469.9256,-260.497 470.0218,-261.4191 469.8629,-262.3502 469.4562,-263.2877 468.8098,-264.2291 467.9327,-265.1718 466.8344,-266.1132 465.5252,-267.0508 464.0156,-267.9818 462.3168,-268.9039 460.4398,-269.8145 458.3962,-270.7111 456.1973,-271.5911 453.8545,-272.4523 451.379,-273.2923 448.782,-274.1087 446.074,-274.8993 443.2655,-275.6619 440.3664,-276.3945 437.3861,-277.0951 434.3336,-277.7617 431.2171,-278.3925 428.0446,-278.9857 424.8231,-279.5398 421.5594,-280.0533 418.2594,-280.5246 414.9285,-280.9526 411.5717,-281.336 408.1933,-281.6738 404.7972,-281.9651 401.387,-282.2091 397.9657,-282.4051 394.5361,-282.5526 391.1008,-282.6511 387.6621,-282.7005 384.2222,-282.7005 380.7835,-282.6511 377.3482,-282.5526 373.9186,-282.4051 370.4973,-282.2091 367.087,-281.9651 363.691,-281.6738 360.3126,-281.336 356.9558,-280.9526 353.6249,-280.5246 350.3249,-280.0533 347.0612,-279.5398 343.8397,-278.9857 340.6672,-278.3925 337.5507,-277.7617 334.4982,-277.0951 331.5179,-276.3945 328.6187,-275.6619 325.8103,-274.8993 323.1023,-274.1087 320.5052,-273.2923 318.0298,-272.4523 315.687,-271.5911 313.4881,-270.7111 311.4445,-269.8145 309.5675,-268.9039 307.8687,-267.9818 306.3591,-267.0508 305.0499,-266.1132 303.9516,-265.1718 303.0745,-264.2291 302.4281,-263.2877 302.0214,-262.3502 301.8625,-261.4191 301.9587,-260.497 302.3164,-259.5864 302.9408,-258.6899 303.836,-257.8098 305.0049,-256.9486 306.4492,-256.1087 308.1691,-255.2923 310.1635,-254.5017 312.4299,-253.739 314.9643,-253.0064 317.7612,-252.3058 320.8138,-251.6393 324.1136,-251.0085 327.6509,-250.4152 331.4144,-249.8611 335.3917,-249.3477 339.5689,-248.8763 343.9309,-248.4484 348.4616,-248.0649 353.1437,-247.7271 357.959,-247.4358 362.8884,-247.1918 367.9124,-246.9958 373.0104,-246.8483 378.1617,-246.7498 383.3451,-246.7005 388.5392,-246.7005 393.7226,-246.7498\"/>\n",
       "<text text-anchor=\"middle\" x=\"385.9421\" y=\"-260.5005\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#000000\">order_products__</text>\n",
       "</g>\n",
       "<!-- order_products__&#45;&gt;order_id -->\n",
       "<g id=\"edge16\" class=\"edge\">\n",
       "<title>order_products__&#45;&gt;order_id</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M417.4708,-247.6172C425.3775,-243.333 433.656,-238.8474 440.8629,-234.9425\"/>\n",
       "</g>\n",
       "<!-- order_products__&#45;&gt;product_id -->\n",
       "<g id=\"edge17\" class=\"edge\">\n",
       "<title>order_products__&#45;&gt;product_id</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M346.38,-248.1977C341.6525,-246.2256 336.9162,-244.25 332.435,-242.3807\"/>\n",
       "</g>\n",
       "<!-- add_to_cart_order -->\n",
       "<g id=\"node19\" class=\"node\">\n",
       "<title>add_to_cart_order</title>\n",
       "<polygon fill=\"#8fbc8f\" stroke=\"#228b22\" stroke-width=\"3\" points=\"405.2829,-354.3845 323.3411,-354.3845 323.3411,-332.3845 405.2829,-332.3845 405.2829,-354.3845\"/>\n",
       "<text text-anchor=\"middle\" x=\"364.312\" y=\"-340.6845\" font-family=\"Times,serif\" font-size=\"9.00\" fill=\"#000000\">add_to_cart_order</text>\n",
       "</g>\n",
       "<!-- order_products__&#45;&gt;add_to_cart_order -->\n",
       "<g id=\"edge18\" class=\"edge\">\n",
       "<title>order_products__&#45;&gt;add_to_cart_order</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M380.9254,-282.9499C376.7321,-298.2039 370.8877,-319.464 367.3652,-332.2779\"/>\n",
       "</g>\n",
       "<!-- reordered -->\n",
       "<g id=\"node20\" class=\"node\">\n",
       "<title>reordered</title>\n",
       "<polygon fill=\"#8fbc8f\" stroke=\"#228b22\" stroke-width=\"3\" points=\"443.0705,-354.7206 392.1188,-354.7206 392.1188,-332.7206 443.0705,-332.7206 443.0705,-354.7206\"/>\n",
       "<text text-anchor=\"middle\" x=\"417.5946\" y=\"-341.0206\" font-family=\"Times,serif\" font-size=\"9.00\" fill=\"#000000\">reordered</text>\n",
       "</g>\n",
       "<!-- order_products__&#45;&gt;reordered -->\n",
       "<g id=\"edge19\" class=\"edge\">\n",
       "<title>order_products__&#45;&gt;reordered</title>\n",
       "<path fill=\"none\" stroke=\"#c0c0c0\" d=\"M393.1246,-282.6314C399.2532,-297.9314 407.8764,-319.4591 413.0806,-332.4515\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(HTML(rpy2.robjects.r.svg[0]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I really like this DiagrammeR package. Using the settings above it reminds me of the sklearn cheat sheet.\n",
    "\n",
    "[Source](https://www.kaggle.com/kenny21143/entity-relationship-diagram \"https://www.kaggle.com/kenny21143/entity-relationship-diagram\")\n",
    "[(made with DiagrammeR)](http://rich-iannone.github.io/DiagrammeR/ \"http://rich-iannone.github.io/DiagrammeR/\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# R vs Python for reading data and importing data into the other language\n",
    "\n",
    "Everyone still using pandas.read_csv() needs to read this. A much faster option is fread() from the R package data.table. On my late 2012 MBPr, using fread() instead of read_csv() makes a ~0.5GB data read take a few seconds instead of a few minutes. The reason behind this is that pandas.read_csv() is python based while fread() is C based.\n",
    "\n",
    "Further, reading into R and importing into python is much faster than the reverse process. When you read into python using pandas.read_csv() and then import into R using pandas2ri.py2ri(), you have made two complete copies of the data in memory. When you read into R using \"data <- fread()\" and then import into python using \"py_var = pandas2ri.ri2py(r.data)\" something different happens. A vector pointer is created instead of making a complete copy of the data. The creation of the vector pointer is almost instantaneous, while a complete copy of the data is unbearably slow for anything larger than ~0.5GB.\n",
    "\n",
    "\n",
    "Like the reading of the data, the importing from one to the other takes a long time from Python to R and seconds from R to python.\n",
    "\n",
    "If you prefer using Python, consider using R as a speed loader for your data.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-25T07:29:24.291890Z",
     "start_time": "2017-12-25T07:29:20.629243Z"
    }
   },
   "outputs": [],
   "source": [
    "%%R\n",
    "\n",
    "library(data.table) # for fread()\n",
    "library(dplyr) # data table manipulation (joins, group_by, etc)\n",
    "\n",
    "orders <- fread('/Users/andrew/Desktop/data/orders.csv')\n",
    "products <- fread('/Users/andrew/Desktop/data/products.csv')\n",
    "aisles <- fread('/Users/andrew/Desktop/data/aisles.csv')\n",
    "departments <- fread('/Users/andrew/Desktop/data/departments.csv')\n",
    "order_products <- fread(paste('/Users/andrew/Desktop/data/',\n",
    "                               'order_products__train.csv', sep=\"\"))\n",
    "\n",
    "\n",
    "# Combine aisle and deparment data with product data\n",
    "tmp <- products %>% \n",
    "    group_by(department_id, aisle_id) %>% \n",
    "    summarise(n_items=n()) %>% \n",
    "    left_join(departments,by=\"department_id\") %>% \n",
    "    left_join(aisles,by=\"aisle_id\")\n",
    "\n",
    "# Combine with order_products__(train) to find total items sold in \n",
    "# each aisle as n_orders\n",
    "goods<-order_products %>% \n",
    "  group_by(product_id) %>% \n",
    "  summarize(count=n()) %>% \n",
    "  left_join(products,by=\"product_id\") %>% \n",
    "  ungroup() %>% \n",
    "  group_by(department_id,aisle_id) %>% \n",
    "  summarize(n_orders = sum(count)) %>% \n",
    "  left_join(tmp, by = c(\"department_id\", \"aisle_id\")) %>% \n",
    "  mutate(onesize = 1)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now let's check out the header on the data frame we just made."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-25T07:29:24.360839Z",
     "start_time": "2017-12-25T07:29:24.295921Z"
    }
   },
   "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>department_id</th>\n",
       "      <th>aisle_id</th>\n",
       "      <th>n_orders</th>\n",
       "      <th>n_items</th>\n",
       "      <th>department</th>\n",
       "      <th>aisle</th>\n",
       "      <th>onesize</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>34</td>\n",
       "      <td>3341</td>\n",
       "      <td>229</td>\n",
       "      <td>frozen</td>\n",
       "      <td>frozen meat seafood</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   department_id  aisle_id  n_orders  n_items department                aisle  \\\n",
       "1              1        34      3341      229     frozen  frozen meat seafood   \n",
       "\n",
       "   onesize  \n",
       "1      1.0  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# \n",
    "%R goods %>% head(1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here you can see how to print just the columns you are interested in viewing.\n",
    "\n",
    "Notice how the %R call prints the data.table object from R into a pandas data-frame output."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-25T07:29:24.434557Z",
     "start_time": "2017-12-25T07:29:24.365904Z"
    }
   },
   "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>aisle</th>\n",
       "      <th>n_orders</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>frozen meat seafood</td>\n",
       "      <td>3341</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ice cream ice</td>\n",
       "      <td>22676</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>frozen meals</td>\n",
       "      <td>18221</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>frozen vegan vegetarian</td>\n",
       "      <td>4727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>frozen breakfast</td>\n",
       "      <td>9729</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     aisle  n_orders\n",
       "1      frozen meat seafood      3341\n",
       "2            ice cream ice     22676\n",
       "3             frozen meals     18221\n",
       "4  frozen vegan vegetarian      4727\n",
       "5         frozen breakfast      9729"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# select n_orders and aisles column from goods in R\n",
    "%R goods_tmp <- goods[ ,c(6,3)] \n",
    "%R goods_tmp %>% head(5) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Wordcloud Viz in Python"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The python version of wordcloud is superior to the R version because you can use images as a mask and you can use the colors in the image as the color scheme!\n",
    "\n",
    "Dependencies:\n",
    "\n",
    "In a terminal run \"conda install -c conda-forge wordcloud=1.2.1; conda install pillow\"\n",
    "\n",
    "[Source](https://amueller.github.io/word_cloud/auto_examples/colored.html \"https://amueller.github.io/word_cloud/auto_examples/colored.html\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-25T07:33:22.446291Z",
     "start_time": "2017-12-25T07:32:42.344182Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<?xml version=\"1.0\" encoding=\"utf-8\" standalone=\"no\"?>\n",
       "<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\n",
       "  \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\n",
       "<!-- Created with matplotlib (http://matplotlib.org/) -->\n",
       "<svg height=\"252pt\" version=\"1.1\" viewBox=\"0 0 245 252\" width=\"245pt\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       " <defs>\n",
       "  <style type=\"text/css\">\n",
       "*{stroke-linecap:butt;stroke-linejoin:round;}\n",
       "  </style>\n",
       " </defs>\n",
       " <g id=\"figure_1\">\n",
       "  <g id=\"patch_1\">\n",
       "   <path d=\"M 0 252.290164 \n",
       "L 245.33932 252.290164 \n",
       "L 245.33932 0 \n",
       "L 0 0 \n",
       "z\n",
       "\" style=\"fill:none;\"/>\n",
       "  </g>\n",
       "  <g id=\"axes_1\">\n",
       "   <g clip-path=\"url(#p5c55a97d7f)\">\n",
       "    <image height=\"218\" id=\"image92aaaa1924\" transform=\"scale(1 -1)translate(0 -218)\" width=\"195\" x=\"39.65\" xlink:href=\"data:image/png;base64,\n",
       "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\" y=\"-10.412039\"/>\n",
       "   </g>\n",
       "  </g>\n",
       " </g>\n",
       " <defs>\n",
       "  <clipPath id=\"p5c55a97d7f\">\n",
       "   <rect height=\"217.44\" width=\"194.98932\" x=\"39.65\" y=\"10.972039\"/>\n",
       "  </clipPath>\n",
       " </defs>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11aa82080>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/svg+xml": [
       "<?xml version=\"1.0\" encoding=\"utf-8\" standalone=\"no\"?>\n",
       "<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\n",
       "  \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\n",
       "<!-- Created with matplotlib (http://matplotlib.org/) -->\n",
       "<svg height=\"252pt\" version=\"1.1\" viewBox=\"0 0 245 252\" width=\"245pt\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       " <defs>\n",
       "  <style type=\"text/css\">\n",
       "*{stroke-linecap:butt;stroke-linejoin:round;}\n",
       "  </style>\n",
       " </defs>\n",
       " <g id=\"figure_1\">\n",
       "  <g id=\"patch_1\">\n",
       "   <path d=\"M 0 252.290164 \n",
       "L 245.33932 252.290164 \n",
       "L 245.33932 0 \n",
       "L 0 0 \n",
       "z\n",
       "\" style=\"fill:none;\"/>\n",
       "  </g>\n",
       "  <g id=\"axes_1\">\n",
       "   <g clip-path=\"url(#pb5e30a6d09)\">\n",
       "    <image height=\"218\" id=\"image33ab498937\" transform=\"scale(1 -1)translate(0 -218)\" width=\"195\" x=\"39.65\" xlink:href=\"data:image/png;base64,\n",
       "iVBORw0KGgoAAAANSUhEUgAAAMMAAADaCAYAAAAbpvb5AAAABHNCSVQICAgIfAhkiAAAIABJREFUeJzsvXeUXNd9Jvjd+96rXF2dcwONnEEkEgwiRYpUljVOkiytZY+lOV7be+wZ75nZ8czsWp499pmzM+Ow67GtsSXnIMuWRCWKSRQjQJAEkXNooHOu7sr13rv37h/frepuAEwSSAJk/87BQXf1q1fvvbrhF77v+wljjMGyLduywa398KPMCSHEdbmYZVu2t9PqkwHlOZiwCugQUFUAEtABRLwJcDzo7CBENAUICSgfMBqisQ+mNAvR2Pf23cGyLdt1soWdwS9yoDsehBuFUQGgHcDxAOkCRgNuDMLxYHQUUAEQVPieZVu2d4AJrbURQtTdpCtdHmM0AAHAQCwa+MYY1N73Sm7SK51z2ZbtRjSpRw/TParkAB3C+EWYsAoTVGAqOejLB2Dmh4FKHiYocwcBYKbO8edqAUYFMOV5mGoeJvRhKjmeEwZ65PCPFI8s27K9VeaamQFovwxTzkI0dEM0dEEPHoDs3Ao9fhzQGiLaADV5FkK6MDqEu/5+mPwYTG4E8BKAMYwhki0QjX3QY8cALwHZtQUmOwh0bQMc97WvZtmW7W00V3ZtA2JpiOIsRLoDRoecFOkOSBUAbgRwY5ACQLQBmL0IABCta20gbQAvDuFEYMpZG2M4EA2dQDkHZ83dgAmxOFZftmW7EU0YY4zOjQMwNhg2EPFmmNIMRCQJE5QArQAIyEw3TGEKiCRhynOAkDagDiBSHTBzg4B0YXLjkE0rAOnA+EUIN2Y/TQLShYhn3sZbXrZlu7a5AGBmByCSLfT3c6Nw+m6FmRsCoikAEjo3AhgDkWiGuvw8ZNt6GL8AqBBGOkBQgUi2AaEPU5kGYKAnTkEkW6CzQ5DNK2FUAOFGgUgSMAqIJJm1Ks8BXhzwi3S5wgpEvPHtfSrL9q40CQCybzdE6zqIeAaydR30yCHuBl4cIt0O2dQPEcsAqgpTnIFo7AX8Ekx+AqachanmARVAtK6FSLZAdm+ny+TGAOs6iXQHRNt6QAfQEyehhw/CGA11+hGYqXNQ55+EGngW6uzjb/czWbZ3qbkAIJwIAMAEFb4aTdHtKUzQtQkrgBsFImmIaNL+nIJMtgJCMtvkeBBCQOcngLBKdyjTA6RaYcrzEF6cqVit6V4lWmBmL0Gk2oF4Bog1QGZ6oXKjb9vDWLZ3t4nF2CTWFADWFQB98RmYah4i1gDZvgFq8AWY+VG4238ScCJANL30ZELAhD5QzcH4Ze4gAAt2QvLvQZmTCa9QeyjNQiRbrvd9LtuyvaZdUT4WdH0mTwHKR3j6exCpdoQnvgWdHQKEA9GyGuryAdYcJk4B1QKEEPXCmnAjEMlWyKa+hWJbbSIYwyq2/V0IAQRlHmJ/XzwRjDHLNYple8ts6WTwC1CH/wkmPwlUC3D6boUeOwp300chu7dBnfoeUJqF038nzPwITHEKwXN/fNVJ9ewlmNIsB/PUWZjcKNToEaA4DTN1Dnr2MvTsJQCAmRuCmTgFnRuHzo1DjRxmsW/yDMzkGU64ZVu2t8CWTAaTn4TxC9DzI9Bjx2BKswyos5cBrSAyXQyMvTiEG4Oeucg6wuJzGA09fgKmMg+UZjkp8uOsVBemYHKj0DMXYHLjfIOQMEEZQkqISJKYJyGZwpUSen7kLXsYy/butiWVMNHQxQEJA5HuAOaGAOmwEm0+ARFvgqkWYOZHoMeOQiRbrf+/5CwQ8UaIdCegfMjOrTwmKDGdmulhuCA9Ht20EiIoEQzoRiDb1wNuFLJzC+DF4TT0sLB3FWZqGfe0bNfXrgigDWBTpSLTAwQVmNwoID2Ixl6Y0gxQLUIkmgEvBjM3DJFoAmINFtlqWEMwYAYqrAK1moFfZNDtRgBjYAqTEA1dRL66UaZyhQSkA56AMYSevczjjOL7heSxRsHMD0O2rHlbHtyyvfNs6c4gBJBohp4ZYB2hWuCKXC3QZercDD19AaaSQ23AmtIsIF3IltUwQQlm6hyLaDpkcJxqh3AjQCTFApsOmV0KKoBfAhwXplqkeyQlB3dYASIpwItBpNpYy5gbhEh3QWS6oYdehGhaCT11HqKpH0I6b8/TW7Z3lF0TMCRiaYhUB5OfQnDwGgO4UYjGXoh4M10dIQGjYfwSEE3BjByGKWfhrH8/IAR3mEiSeCW/BFOZh3BjMEYBwuGkCcoQXgxGgBXsoAxTmIRsI2TDVHKsVBtDeMf8CExQgpAOZPd2ICjDRJJL4OTq8gHAaMi+3Qvu2LI7tWyvYVcxc9Sl/RAN3YAOoUcOEXUqPejRIwAEZNNKiFgaevhlqEv7gaAKM3kapjgNPXGSq39Qgh58ASKWgYg1QHhx6PHjMPMjBPDlJ4BqHqjmYeaGGJ8UpiBiDYSGuFHrLgEi2crzVnJAKQszewmyoQd6+GXo0SMwxRmoc9+HqRagz/8AauA5XnduFCjPQZ9/ktCSZVu217BrTIZ9TG1OnYVo7kd45lGY/BhE+/olx+nx40BpFurSc5A9O6EHnoXxC3VckSnN0s+vnffyAci2dVzZwwrU4AGEpx6CXLkXeuocZOdWqNOPkD8xe4k7EQDhuIw9SjMw1RxdN6MgvBih4jBAcZpcjKAMNfAc8U8qgCllYcpZ6ImTb94TXLZ3jF3N2Qwq0COHIdvWQU+ehrPyNpjZy9Bjx5ccJhp6YHJjkN3boYdehOzdA9m2ganUsMIB6SzKNGlF8F5TPyA9OH174PTfCX35ecimlVAjhyH77wTijXVEbM1McRqieRVk2wZA+cQ5ZXogu7YzMI83EVToROD03QqZ6eE1JpoIJ18G/i3b6zBxpVSMOvcE5Oq7mR3Siu5KWOUf3Wjd9zZGw0yehmjfaI9zAb8IU80RwVrOAvHGOu6J572Hfn9+HMKLA7EMA2oIZrGMJtxbyjphyNQ4E4BN+wpmpKoFGOXzempoWL/Ea42mYMrzQDXH+EZKINZAkGHt/ACgQmKqoqm35GEv241tV02GxZxmo0IAhgPdZo/sUcz+OJEluCP/6T+A070TevocZN9u6NGjcPf83JLgVk+dhZ46D1OZg7v5YxzEYRV6+hwLe14CIt3OoDyowFRz/PywApHu5GqvQujZAV6KdCEiSYimPkLK54YAN8Y0MACRaGGNI94Ekx/jtaa7eN2lWQAGomPzcoC9bFe7SfWJ4BcRHvkq1MnvMtCdOseBVpqBOvLPfL0yD3XsGwvvjaaJdHUikN07oIdeXII9AgARb4KePM3JBCB8+e8ALw7ZsRl64DnoydMw+UmoM49CtKyCmTzDz88O8W9+CcFLf0WJmkQTzOQpqNMPc8J4cYjWNRANndCXD0CPH4eIpaHO/wB68AVOlsI0s2XpdsCNQE+e4b0s27veXlHnRY8dhxAOoRF+EaY4w0FZyUMNvWTdjmnoqbP194hEMyBdOKvuAspZOKveA3XhyaUnjjfaApwHRNPQ4ye430gXRivA8WCqeeiJU4CQ0DMXEZ74Npxt/wLq7OMQqXYI6cFZey9X90jaXoMF+kmXaVvlw934IeiRIxQniCS5W5RmASEIFkx1AF4M8AtvxrNdtpvMXnEyiEw3TFiBHj/GQTl2DHrsGFOl0TRBd0EZcCLQQy8S7VqcgWhaSZLOue/D2fJxYpwWe2LKh2hdw0yQ8gEVQA2/DCgfTv8dQCXHgFdVoUdeZqDcuhZ69ChEpoc7jBB0dxpXAq7HesUVpnNjUCOHITLdcNbcy/Na+IgefBF65gJdLb8I2bPzzXi2y3aT2VUxQ82MMTAzFwE3ApHuItjOL5KtVpiEKc9BtK0DCtMsriVbuLrPjzCgdSIQDV2EXaTal5575iInRcsamOI0B2rrGpjCJH9uWw+TGyM8vKkPEA7rD419gBdH8OTvwrvv3/JcU+eoxNHYW8cvGR0i3P+ncNbdD9G2DiY/QZ5Ey2oG+ZUcA+dEC8zsABBvIuRDh7xA6QI6hHC8N/HRL9uNZq84Gd6IhcceJGV0zb3XDETV0Isw0xfpQq17H4PgH8KMDgGtKWWz6j2vIl6mYbKDkM39/L04Q0hHeY67mnTqGCcTlCAiKYhMD0xxCiY3BtHUDzM/Aqd3p/1M4qKWg+x3tl0XbUhn9XvIXfZLAOyuUpiCKc/BGIPwxb+G7N0FkelBePRrhHnnxhmLBGUYrRYop9aMCqDnRylQFlZ4rskzdGtW3sFUrFasa9T/D22xD5DN/byOUpaD3/EgGlcAkQTdMC8OkWgi7CTWwEJeJAnRtJK7QqYLJvShx47TpQurnFDLZKN3rF0fMaNECwABdfZRuNt+AnrgWeipczDVHNwdn4K782cQHvkniGQLnDXvhb7wFPTUOQCAaOyD7NkBM3kaztr76qcM9/8pRKIZzvoHAOlCj5+g+yInINwoi3SdW5iVCn2YwgRk8yro2cuQrS6RtAD0/Aih59lBiKaVMLlRGK0gUu3QQZnpWx3WdWZl65p6/INUntgmHTL9O36ShcFle0fa9VENDkqA40GdfQzGGAQv/DncW38O7o5PQZ38LrFH0gUiKYjWdVDnfgB3z2fhrHsfMUo6pMLGIlODB+Ds/Bkg2UrtpkQzEEuz9iEdiGiaBbSgTFRsyGAcKmCGqGbVPES8gRkmvwgRa2QmrDTD9Krj8ngdLBQA440EGlZyEF6Uu0JYYZFw2d6xdl12BnXyu3B3/yxxTXNDENEGoDzP4DnRDD34ItwtH0d48ttAeY6YJSEJ07amp89zQrhRVpy14iRzYyQdJVsWVmknwmA61kBQnxuF7NrG/7uTdaQqAMjeXVzx+/Zw4KsAQnoWNyUgEq0LyFxrwnEhVtxqYwWPn+V4lMtZtnesXZedwZTnIDI9cLd8DGZmAN57/w3CY1+HHnoJzsYPQa66E2rwAKAC6OnzkN23UFzAi0E0rYDwEpAtaxAe+gpMbpz858o8wqPfIP1TBYBfJhI1rDI7ZBQhH9Jhmnd+hHDu/AQAQM8MwMxcYJ0kO8jsVCXP8wlBN8gomJmLNr7JMhuWnwC8BIR0OdGqBU6U0OcOlhuDnjrLmGbmInR2iILLxvD18tz1eKTL9jbY69oZTCUHM3uZK6t0gUiCleWwCggJd+/nSQN1oxDJZgCAs/qeuhCZmRuB7N4O0X8nICRk1zaY4hRdnEwPYBQniOOxqNbYB++u/w3Ohg9wBynNsKYRVgnZkBLGr7BwZzR9/qDM4LnmbglJV6i5H3roJcjeXdCzA9ATp+Du/BnooYNAJAHZtg7C8aCGD8FZ/z6ogX3wdn6yfu86O8jsl5BAOQs9cRqyYxOgfKhL+xm3+EVWy3NjvJ9luynt9blJQnIixDIADEWGwypdGulSDKy2qiaaIfwSC3DJFqA0w05AJs6agla2diBhqgUIIWEiCa7MWsEUZyCb+yHbNwBBifCJ5pUQRnN3KE5BZHqBSh7wEjC5cU7AZCuPVwFxTIkmFvWEQ+51QzfBe36RTDopgBrkXLrcafwi37PEDLNkQgKGQmumOAN4MTi9u+naeXEy9yKppUjdZbup7HVNBhFNQXRsqv9urO6qkA5MaRY6exky0ws9NwxRyQGZHroL8Ubocz+AqcxDdm4ll6EwCWf1eyDSHQiHXoKBgZm5RL6EX2J9YO/nmLnp2wN1eT/c5n4I6ULZCri+/DxMUIbUIfTMAESiCbJlNd0rvwgJw2qzNdm8ioC+dDupq2PHIXt2kRMxfqJ+rJ46D9m5eem9J9sswcmFbFtHZXJV5eTKDjEl29gLMzfMTFWmhwSmZbvp7IcKoAmJsLxjLwGUskDLWg64hm7omQsQTX1ktjX1AXNWotKLE41aM+UDEKxkV3KQvbuhwyp99mgKCCuQ6UUDyyi6Pk0rYUaPcPWfGWCadOIkRKYXZuQwBY9nBiBSHOSifQP/b+gmi2/RfcgVt/Hnyjxkzy1YrPRnDFVCRLqj/prs3r7w/tY1EFjDd7Sth9PQdZXKYP08NeQuP/iHeezL9ibbj1yBNsYwwyNdpielt8CDqP1f4zwbBT1yGE7fHr5Xh0SG1/gLtaxOtUhCT8vq+jmFEPTfGzrtZ/B1M3uRv4dVxhAQQEBFb9G0Amb6PK/DjdIt8+IkLs0NW3ECQzhJopnBssU9wfEgIikbJzm8P4u0hXSZNGjoYhBfU/UwFu7ueHXcVf380RRMKUsJHU1tKGFrIct2Y9gb2hmMMTa7wsGFet8FKmhwZFcZYBvFARGE1qcuAJFUfSIAIMK0dl5jJ4ZgwUxYhQ3hRFhhViFE4yLJSifCDM7YMfIcajALFZAaWi3SfcmNs+qsFVCZZyXbYpQAcKKEFSYAGrrsjWrGCU0rgYDHiHjGih+UOYlsLcMUpsiy06S4imiKx82PMIYIalV5pm5NaaY+ORfvIstQj7ff3pibpHwKiE2fo/T8ir0cfHNDBNkZTVWKru31Jid67CgrwzMXqZqxuCBWM6OhLj4Lk5+As+5+wC+QZhpLEwZ+8ruAGyU0fPFqajThFk4EsnMLM0L9d1I6JqxAnfgO4EbgrLgV4bFvQKQ7iVM6/yRk+0aISBx6ZgCy71ZmoaJpSmyefQKAgezbAz38sk0dd0M29UMNvQjhRuFs/BAAChaY8hzvPdbAiegXbHJAsK9F23qmeZVtKxzLMPs1O8CJ1roGryjEvGxvmb2xyWCRqE66nc0Na4UqIZli1IqvSZfUT+VDdu8Awgrh2W7k2ucVEs6K26jMkWpD8Mw/wLvzl6zcPfFJsrG3rgheM1OahWhZC2fVnawd5NmBSAgBNXWW15tsY2BenoPs2r5AOGrqg5EO3TEh6pPMRFKMBRJNgBtDeP5JRO7/9yT3nX4YItlqK95ViEgCJpriDlcrJLpR7kS2mGdq5KZEMyeDEOyfZ4N20lqxPBduAHtDk0EIARNNEruTG6MrADANKh26JDpkVx8vDiEcpkyNhilMcUK8wnl1nXQj2BsiO8hzRJJAWGHaVrrAojy+cGO20eIYoRUQMNPnYeKNELEMdHAeJnsZItUGd+MHYYqz1jUx0LkxiEh+wdVbbOU5INnC8CGSgJkf5bXYXhN66hxk+0YYTTgIpAsIj+6iVjD5CcrxG+4c9f4WNhVNpXFR3zmW7cawN55N0oqrabrDukWoB44i2UKXIpIEQDEyJFs4YBbBHa5lpjJfl4p0t37cdgyNQ3RtY0EurDI9u3jwRFOQnVsoc5lshbPiNrpnxRnIzs2QpSygA4hGZrZgNHtL+CX6+o4Lp3fH0l7WtXuJpgEIuLt/Fmb6AkRDB2TvHujRIwyCrZiBHjvOXaKhk+c0ihmz7GUILwbZfxf04AswfolxhxtnAdKKEizHCjeOveFsUi17VCO+6Klz1EuKpAAYiJbVMFPn+Vqqnep4vbtf80s3oQ8zN0TUqDHQI4dYEAsrDJb9EoQX48/FaYh4EweXdEkfHT1C/zvRZPvTVSESLdDDB5m5STTzfFoB8Sbb1jfkeRr7eG5rOj/JTkOxq9OkVz2LoJYqFkwYON5CZkwFnDR+gS6UkIR1xBqWJ8ENaG8YmySEWMIAE/FGCDdGEFskxYGXbIHIdNMtEJLsMmNgitOEdhhjm6Zb7kMlT1mXllWs7voFvmd+BKYwvRCLaMrdw/Fg/CK5z8rnzhNYeIYKoGcu2d2IpBxTmIYpTMLUGrsXp+qpT50dJLfbpkABQKTamAaeG2ZFG4DxS9Dzo9DZQejcGH19GKaBbaxiSlnuil6C/yywjy5fghMsnqG4gtHX+GeYOTOm/vuyvXX2IwP1RKqN2kmpNshMtyXat0E2dLPZyPww9OQZ6JFDlH4cPggEZejBF3mCoITwyD9BDb4ABBWEL/89mXOt6wAvDnXhSYh4I12L9g1w1t5L9b1TD1Exz2jowRcgOzbRnYo38rMuPM3gOz9BmIZfhEi1QY+fgL58ACLZBj1zgdcydgwwGuGhf+Q9CQE9cgh67BjCg38HAJwolTlmncKqBfoZEpRKWcuBmOKuduk56Ev7oYdf5nsrObpQ1tSFp6AvPgd94Wmos9+HmR+FOvcEmz8OH2TzlgvPvKZruWzX164LanVxG6vFps59H87mj0H27oI68yhg6wJMic5wBcyNMc1YnGZuP9HEVR4Gsnc3cUXpjoXWV5EkgXEN3XA2fYTBdlACjObfvbgt4rHOoU59D872n4Sz7gEAhpwLv8h2XCe+Q79/6iwAwwFeu6fGPpjiFPtZg4GwaN9EjaXGPtY2pAPR3E83sHkVU7RWR0pnB0lgquYZdI8dW3gwXhymPEv1wLkh8r6dCFCjpcbS3L2W7S2168N0ewVz1t8PdfI7EJleOOsf4BdvTD1lqs4+zrRoDSodTTLOmLnw6ip3XgIme4nSlO0bAelBXdoHkWqHXHEbYRlT54BoA7Vcz/0AiKW5e/TsZBziF8hx8OJWmFjAFKeIRO3ZAXXqexBt6yFqk2Gx7L1Y+PnKRcBAUsWjlIXs2QGEVTj9ty9UrwE4a+6hOxlNMZaJNTC1K1zrZraRhyGuD/dq2V6fXRdBgFcyYwxQnGIFOpKiQHCsgUp8YYUuR7INqOT4hlgDUJlnjJBohpkfhjr2ILy7f/Xq8xYmofMTkK1rWQz0S1xh2zfATJ1lMN1B6UszdZYkoXQ71OlH4Gz8IADCLszMBXKjYw0w+XGY4jQ1XcMKAYjprjrGack1WMZdbSeop09Dn0F9Ncc0riDURDi22l5DxQpJpK+2lfoa4Ui6FrtkoSlC8JxuhDumF68H6mJ5slxXuy6TwfhFpiOvamn1o5nOs+/bErAewAHkF4Fo+g1lZUzow2QvQbatf+2DX8PUpechV97GtG66E3rqDIQTgZ4fgdO7ixztllWEnZfnuDsACM8+TrpqJMHjLj1P+EpYhankIRt7YMIK08jT57lzzY8SbTt+ArJnJ9TgC3D671iWsrnOdl2WFnXhKejhQwwU8xMk0UydY/aoNMvsS5WqdaZgg8yJk/Slc+NkpYVcMZmtGWcq1Y0BENAzFyktP33BnmOSAXJulKp+Mxdh8pNkns0OEAB4DRNuZMlEMJU8J9ZrmDEGujC58LsOoeeGCEXJjQFBCcLx2Kgx1U6lj6AMMzfCe6/hoGCDaeXDVIvQ+Qmy5PITJAWFFR5bLVKAYM6KGSifu6W2O0Ulj+WS9fW36xQzCMChgoWIpqHGjjLLFFYo56gV0LkFIpqCnjjFfg6TZyBUABFvgrr4FJwNH4Ro6IIePUoXIN1BaXkdQl18BrJlFVfI1jXcgYISjxXsEqqrecK7I0k4Dd0wWlt9JOty1EwrsDQsoEePkHUXTVqXhbpMdEGWDjZ14tsQuz9bR8aKeIaFwPaNdF1ijZAtLXXpTNm3h8FwtcBWXXYgi0QLnO7tRNH6RVbllU8htN5dQGUOonMzr9koYqs6twDxDGTHRkA6kG3rlmHgb4JdHzepPE8Yhg5Zb7AUTNHQxcEgHRaaHM/2gxMwuRGI5lUcPFpZ+XqPQLrcGHP9kSR9bgui0+PHWWXWysraJwA3QlBcJAk9dowkn/lR9oBLtTMLlO4grsmJQM9egrq8H07/XVT4rhZg/CLc7T8NdfFpUlGb+zkAa/dnNPzv/Sac/tuZNWrs5TW2rgOEhDrzCKBDyLYNkG3rEB79GhDLwFl3P4J9X4RsXcsB7HjQl58HvCScjR+EuvAkzMwAwYk6YPYpKMHZ9pOEvoTVa4qXLamYL9t1s+uTWo1niGJNd5Dkn2iCaFkNEaG8vEi21P1bEWugCnbbBnKmE83cRWp/T7ZAdG4h3COSJCQj3UGl7r5beYx0SNKJN7KYlWqHiKYgV95OBt2Jb8HZ+GHIvt1ElCqfgz4o87VSFrKX+qqybw9k8ypWyjs2MwhelGKt32M0DWfDB5mODX1OQOkwNTv0EiU4rTukx0/UNaCcvt0QyWaIphVEzmZ6WNk2Ck7/ndyFmlaQeJRshc5PsM6RG6ObmR+Hnh+BCcp1GX6TG4POj3PS5yehc6NsRF/JQWcHl4t1P6Rd19SqqTVCtDWBpV/KFf0dICzGX1xxDBb+Zt9ijIEefhnCS8AYDdm1FWbmAiu+NQ2kRCuzSolmVnnz48xaoQYZKVJOxi8xuyMkP6cwBVOYgmxdi+DQV+D07iLh/8p7s2BBEWdVWc8OUMIm3UGXJShBT55m19PQJ1c71QHRtALq5EOsTSTbYIIKU8IrbrUQDVFH2ZrCFNwNH+TnzY9CRBL2wyl6YGpZN7/A9K4OYLyEZQBmGcOEFQj0/cjf5bvRrutk0IMvUDNJOkDbeqI9VZVFLrcmxsWtHzokPKGhi4FnNEWJyLlhpiTDCvs92EyVHn4Z7o5PIjzxLYiWVYRTJ1qAjk1sj5VoYXve7m3w9vwc9PjJulIHdGiVO1phcmNwVt8FgGhbU5i0qd8EnJV7OTlKsxZTZGHqwoG76cPQU+fgbP4YYAxk82ruAIkWOFt/HGZuGM6mj7B4aAWYRaqDrb46NwMqhLvlx6BHj8DZ/FGmUheJD8jWNdB+iQVEgJD4K7JzotHWPFrXMZ6wcHnokO3BsBxW/yh2XesMOj/OANQoFr6yl+ty8UI4ZHnFM4t2D1mHNcNxbaeePGryMFSbiEA4VnmjJg0T+hDxRpj8OGTnFsrVpzthcqOQHZuWiAHo6fNWT3UBdBee/wHcRVKWamAfEEszNomk6IZ1bIa+vJ8VcDdOTndxhoMdsJM5QVi3Cjn4k60WnJeGrLHmXsXUpecpLtC4YoEGahSxWtLlZImm+VnJVtZF8mPWTcsDKuBOZAyvvVaTiDWwou16MNXC67qWZXsTim7GGKizjxNL5MX45YAQbRFvpM6p9IBYCqjkCPlONHNVb10LUy3QxRGSvGY3Rln8RDNb6ObGrMCx4UBNtnHSCcmJV81blW9B9KoxljNwxW0KyddqVFbPKvmpkIMw3sidrDQDEW9iLaDW1qtmob/A3NNqIcNT40sLUW/hyyKfWFIoq6V1zdwwedF2709QAAAgAElEQVQQQDTN+5cOM1zGMA5r6CYxaH6EO08lx/MlWyCEhM5PWmkcwZS0G+Xz9kuQmQURhGV7ZbvucAwhBBCWYSZOwXixeqBpyvMc9GEVwvHYDCU/AZ0dZAA7N8jJkBuDHj3MQDPeTDAdQO60F6cSd7IFKEyRAdd9C/TwQZJpbFN1kW4nNDvZwmB08jR5y27U0jMzzEQZTQyQ9NgSa36E7lG1QMmaTDddsHgjmXjjx3mtuTEgEucuqEOYyjylLVvXwsxeItK1mredRptQbyofb+R11p6VdDjYm1ZYLFWA8OjXIKINzIKtuG0pDASgSjiwIFZmGXYy0XLt72NZEvN125tSz3fWvJcSLE4ETs9OGBXAWf0eToQa8C7Tw0A21cYMjwoRXniK9MhEM4FwLavYIrfv1gXyfLyRgzGahtN/B3nETgSydT1FiueGgaACkemBzk/ynMkW6Llh6PGTFDeTLtTQSwjPPsbVs5ytC42JSJKZscZeTh7LXNP5CZhSFmr0CIPneBO7/0ydo35rqg2mOEuudOtaZq5a10FnLwORBPT0eeiZCzA6hBo5hPDs49DTF2CmzwEwnIhWzEAkmqCGX6Lm6yIQpB48wN00NwqTHYS+tA965Ah3G78EdeEpQlv8IvTFZ7hICMHOSMv2mvamTAYRy0BEElCXnoeeOscGhJDQk6fYGis3DhFJIjz9MGTnVg6C4jREJAWjAvIjomkGvfFGpm7tCilSbRCJFoSnHrJttk5ytZXSruQnIVrXUrb+2Dcgm/qp4O3FuVK3rGEVuDhNcTEvAXX5AF2aeCP/JVv5OfFGyPYNEInmehNE4UbJiQirUGe/D5Hphhp4DsJLQk+e4o4UTUGdehhqYB8nWLoLJigRVmK7jaozj/K8F56icMLQQbo/c8MwxWl4uz+79KEGJYRHvw41+ALUmccQHv1nIJJE8MKfAwDU0Ev1Nl3Bi3/J+MpyI8KjX3szvuZ3nL1pSK8aC0xd3g9nwwdgytmFJuZCsCLrRpiaDKvkHMwPQw08w/dX5m3r3SvOOz8KPXMR3q7PkFgTlFhIGz3KXcWLQaY76mQdkekBhANn9d0E8xlFtyYoQU+dYe6+Jg2j1VVwdD1yGKY4wwwQCO2WjX1AUIa77cfrRUQAMEXbMMUYQjbGjjGg92KcJA2dhKKn21kfSTTRfQrKC81ahGSR0OKy6tcxewnO2nuhx47CFKfh3vIJ3keNBOQX6ruZmRuGKWcRnvg2arKZRoV1uZplu7a9ibBHwyA63gRn5e384mw/Ndm+AXr8GMz8GEkt8SaI5lUwYRXOyjtI0Jk8A3Xi21cXkKRLXzs/DtHYy/SqCuB0bVt6XDUPkxuDOv09DvyRwxBuBOr8k6wj2LSlbF4F2bwK6uIzUBefuvo2anFGKUt3a/hlttTNdDNeKM7AWfteKzYmER57sJ42dnd9mmSmK0yPHIJstr5/0wqo8z+wcUfMCpyNEPsEywQ0BijPw1l5B0S6i9fveEApC6fvNjL2Jk8jPPs4axV7fg4IKkzPCgeyfROf5RUTbNmW2psG4WYWR4M4IAt19gu2+SEHmLr4DKuwtd2iOE26pA4gYhmullaCEpLZJdG8yqrnuTxnxfZ0iCSB0iz8H/x3RD/2X+oCY7J9Q72uUeMmC8ezdE5TFyljZkkzJVm7Bx0uiKPFMjaRpC0k21uAX9cUNoyxv0cRPP478B74T/WmieFLfw33lk9CxNIwKoC6+CzcdffxHEaTk61437WGK7J1re1Y5LCHtary76Gt3XgJauAKCTM/zL9FU3xeYZU7r5CL+tItw75fzd5UPsNiM0YjeOb/Q+Sef7PoNX60Hj5o3SXLK1YhXQnpsQCmAw6IagGic6utbmsEz/4RcULFabi7PsPxXM7SXVh0/h8Wx+Pv/5+QyXaYyhycTR+BXKS5+lqmC1OQqTZeh1asQ2R66hkklGatZAxNXdrPXS/VzglltN1BTyxIV9p2WzUEsIimINrWcUL/iPe6bG8y0+1KM9lBqMEXmClq7GPj87BM/FGml+7M2HGbktwAPXESsmMzuwElW9kgZPggg+rmVaxdtNzD2oSVwDfzI1D5CWKPKjnoidPMXjV0LVHA4AUZqLGjkG3rIbw4TFglEhaA7N1J+MWKvUBuzOoxNUGPHeVA7dkJ4XjEBOXHIXt3swHizABjg1Q7uxGFVaJbe3bAFCap8QpBFRCrOIiwSv628qEnTgOe7UQEQHQsVQU3uVESmuxuZwpTEJluLhAqgBo+yHSuGyXPe+oMY6F0B2TTCpjKPBXO29YD0RT0yGEiYbu2UxztXWxv6Z6pc8QLBc/9MaBDBPu/CMgIgqd+HwD4ugrquqXB81+CKU4jPPxVQDrwn/5/ATcK/8nf5cVnejnYm1ZAZLoRPPtHdJ1UFTAG/jN/CDgewiNfJef4CjM6RPjyV9ipFEDwwl+wd5tRgFZw194HPXKIkOze3Qhf/nuuysbUaZz+039A1w7g53sxdhwqziB48a+AUpbcjenzENKlbH45CwgH6syjFF4TDl2foAx4Ueihl2DmxwAs8Mtr/9SFp5g6FgLwi1AXn1l4voMHYIoz0NlBBs/SplxHDtVxWsHBv19g0ynuuGZ+lJ2V3uX2lk4G2dAJp2cHXYHQh2zfBDM/XB9Yeuos5Kq74Kzca4t3VQRP/Fe4t/68rVnsIF/ZZplMUIaIZ7ibCAdq+GXKuUydszinIcjeXZDtm64qQAMMZN0dn4C6+DQr5wPPEQU6fZ6KgZUc6wljxwE3wuOK06wXzA3yJKEPp3s7hOMxUTAzgPqHqRBy5e1w1r7PYqVaIRp7ycmIpRk4V/O2o9BJiHgj9NR5GK2ZGbuW1eKbmi3CN+m5ETj9d1CTtob9alsH2dAFaXtGcBfbaTNvZNHp3FhdwO3dbG/pZDClLFfGwgTgRdm0JNlCqAMA0dQPPXwQauQIfWDHo7rewHMADJzV91A71Y0w4M6N0q8eOURXo2MjZEMH291KFyLdAT1xCmqxMkXtWoxBePxbkF3bIFfcBpMbhdOzk1ijEldunR2E7NhUV7CQfXt4vVWiRvXcsEWrssebs+K2epaJLpmGHj1CsYLWtVddg1x5u1XuA1Ok2UHAL8JZcRvgXk3prGGj9MyFOoJVT5+HssXEOpHJCpbVmsPXU7/lOcK8p84R2u4XCDZcuffVBRjeJfYWTgYB7+5fhc5PwLvtc4BwWMCqFuHd/WsQQiBy96/S/6/M8y3GQK56D0Rzv4VTRGD8Arw7/lee765fgZkbgbvzZwDlI3L3v4ZRCu6mjwLxRkTu/jX2h7YKF1eau+3HIRwP7sYPA8KBd8cvAhDkFzf2wtv1GWrEbvoQTFCBd+u/pH/dvYPK2WEFcGPMghky64xfYoYs0VzPosl0JyeVMUsWdXfLj0HEGiAzPZD9d0C0rLL4rNw1r9con2DG/BQzbV4C3p7PUiBNBXBW7OXC4niEgtvsluyyRCUdEgxYnOb7o2k46+6j27Zsb1026Ycx/4n/Bu++f/uKGZLXm0FRl/ZDtK6tZ3dey4xfYgrWBtWvJnSgp85eU2DAaIXgyd+Fd9cvM6awyoLBwb+Fu+Xjrwqee6WvRAjBne7y85CdWyl1c8Xfr3pvJQc9O8AAHWBSwIo0v5q9G7NSb/lkqHf6gaH0iRB1TD6kZ/FA/LspzhIw176ex0gXwovZncNK0Lsx+r9GcSV0Isz7V/N8zULITSXH2kBujP3YgvKCwoYbhQmsAIHjMeMlHThr74MaOkjXyLLyzPworwVYaKRoNExQYTYrPw6R7kKw/38S8erFAScCd/tPEWTnF4mDugKAt9jU2HH+PdZAqIpfhJCuTbWe5P0rn+IDWgFGsboPQF8+ANm2jpmtoExBgfIcRDTJ5xJUCEVpXUsFwNCHbFlFsQUdQjatgJ4ZgLPi1jd3INyA9pamVgFw4MwMkKVVzgLRDFe0wiRb0a7YC5O9TJ/WKkeoKiEPsqET6NgMPXTQ9mrI0DWZuWhhCRoimobs2gqdvcwMTkMXZNt6mPETEF1boafOQVhsEqxCn5EeRLIZJuHATJ/jQKwJhakAeuIkg+GurWzMUslTEaNtHesHbeuhBp6Fu+5+9pLLDgHVPCL3/DoAQE+cgv/YbyP6qS+/Zusqo0OgNAsjJVOnU+dIgLK7k54fAoRLrna6C6jM8tnZyWDqhc0IzMghiM4tHPTFKfLKbTxkfNJgYTR1aZVPdyvWQPfvXWhv/WQACFpr7CNG3yJDRaabBSa/wJpAqoMTppKDmR8htMCmB+XK2/mFGQM1cohftHQhV99NsF95jlDqsEwUZyQBuXIvpejzY3A2fZjQkNIs8f617pxxy4OIJDlAAMiurZRtmR+hOkfHZhL7ZwcYB+TGoYdeZBYqP852uIUJysEUpzip3EhdKI3CajPQEydglA/Z0APRsWnBLQl9xgzCgWxexb81rYQePojw7OOQrethKnNWSdCBCcsQXhzq7GNsJLPqLtS4E4gkCZ23yuAy0wu0ric+q9Y03onApDut0mCV2KdSlooknVstN+TdYW95bV5IB7L7FnINEk0UDYilIRLNCI99AyY7SBi1fQ1GIzz4tyy0RRJWb5UK18G+P6baRFAhlqc0Rzma0aNM2QoHeuhFhAe+xB0gKNHtqbk8uTHKX6baqewnXYoaRFNcRcEqLxLNCF/6awT7/oS7SbwRsnsHTGEawRP/D3eHaArq1ENAPEOxg0gKwf4/RXj4K1BnHoX3vv8DAoApTMJ/+P+CHjsGkx1EeOJBunPWwgNfhh58EaYyR+GCjk0Ij30D4cnvANUcwmNfg//If7a9qw38h3+T+rHFaYTP/g/LFY/x734R4cG/gTr9CJmCNkOnzjyC4IUvw5TmWAg9931Wxqt5qIFnqXA+fhz+9/7T69KVeqfY27IzvGJwZhGfS8yYuiz8kpdnLkJPnUX0U19mrIEaFooar5Rr1NAjhxA89Xvw3vOrC5CG+kl0/fdXDRirBZiZi3B3fQbhiW/Bu+1zEEIgeP5P4Wz6CHnRQJ2tJiwDTbZvJAQ800vMkBAIX/xLOGveC2fHp+r3vFi7VV16DpEf+++EmSyCqzibPgJn3f1wqjkqdPAGgFKW2bVYBrpzC4Knfh/OqrtYW7jlE9BzQ5Bt6+HaazQqQLD/TxH79F+xkTxQr1WIZCu8e36dwX5hCpXjD1rpy/grP5t3kL0tk+F6mJ69CAAID/69fYVEIXfzxyggMHKIK2ZuFN5dv/IjfZY69xgH4vr7Uf3HfwV3z89DSIcMukQzzIt/ZS9BcWKk2ulyjB6GKU5Bz1xE5IP/GXAjdazRkvds+ABEo1W0MLruDtYmqLvr0wgPfYU7UCy9QC8F6ILFGHfJ5lWvjUwNykBQRmj5GTQN99ZfYG1ibgjq9MPQk6fgbv/Etdt8vUPt5oAwVnJXYfFFooV+cMsqqu1V5qEuPA0AMJNnEDz3R5B9tyLygS9A9t+18EbbmATgKvlaStfGaIRH/hnOtp8AYo3ESE2dtdfQDBFvql9DePxBGxsYhCe+De99/x7u3s9T/2h+mO9JtkJEG+rvYZ+Jq6Eii0327kHkfb9B6fvGFUsnQw2RCtg+14lXf5ZOhDTR5pWQLasg0u0IX/4H/k358B/6DxDxRnjv/d/h7v3cuyrFekNMBlPJQ40dY8Ft5gJ/riwuPBmoYw9Cjx0nZMIYBtTSISw51UY3ZsuP8ehyFiLRDJnpgZ69BD3y8qLPmrcr32mo0w/XU5KveG1zQ4wTGroghOAq/fLfwRgDd9dnoCdPM76QLmOdxj5QjKCVmKTsIFCaYbBuDDkOk6frrDRIB7JlDUzo875DH2byNNTYMej8BC9CVdly13GtumDDgtCA1lAnvgk9eQbhC3/B5MCrmRuF038nAX7pTpj5UThr77WfE8JUiwQHVnJQZx57VwmS3RBukglK7HJTnmMDwdwoaZ+xBohYA9y9n2cgffI7ELEMZP+dgBtD5P2/CXXquyQFbfgAXwdXUjl7CeGJbwFCsDJrTTR0w5TnEB7/JmT7BjgbPsBrUAEH9KKV0KgApjAF7/ZfrL8uV9xOVweA7H8PHN+6HI5nXaEoq+kf/AKBeEazyeHYMeKkenfDqeQQnv4eIBxEPvAFcjHCCszEKcAoDvxIgs1a0h0MhwpTCMePA5CIfOA3US+aeTG26D3xTYiGTkLZX8WEEPDu+3cIj34N4fFvQqRa6wgAE0kgcvevWblMtSA68C6xG6YCbYyB/9B/hLP+ATir7yZswXFtE5MF/M5VcitXqPgtPh+zNAuqdXr2EvxHfgvRn/kL6g/VUpBhFXryLGTXVpJ7jIZItjKV2b6RjRGDMgW+vBilbvghLNZleihS4HgEEToOEPocTJWchXFX6uw6EUlSZTzdfsW9GITP/iHhJUmbzbL3VL8fqwRILdYKKl/+OGK/+Ij9m3PF8RrB078P2b4R7uaPUW3PxgAL53Ou8dwW+nu/m9ykG2JnAPilu9t/CuGLf8UebS2r4Wz7SagLTxJ2UMrWK7/GNgcxYZV+tLza2xNCsDi12KTDdK4Q0LODVrU7RQGxoEzN1MY+ChIkW/na5GlWwFVA/oWQpF0qn4C+sArZtAJq5BB3nZr/H5TgpDuprDE/TGiHX4AQLtBAYQBYrnZtwAkh4O7+LBBNXTUIr3k/VhsKABuzX2HhgS/BTF8AuncAANSphyFX3FpXKhTxJpjiOJDpqYugmdIstWsdl/3y5oapDFiYpJxOKUv1wxpMJbSNU94B9YgbZmcAuCrVGhIKq1ShL+0DolbBO6zwS5cuK9RCsm3V61y9jArsl93GrIpfsilPx/KZp6kcXhMXq3UZlW5dJEydf4JuVw3laTQVwOeG+F4Vsh2uF+d5baN00jCj5Bhorrz68n7AiUL2385jaip6jsdjpLQ7jWfhKo7934qULYKXXEupm03rDQP9SBLBvi9CNK+yTVWG4G76KNTZx+Dc8gnGTx2bSajq20NRBa2gzjxKGMjESUqH2ucE6cJZcw/Co1+H7Npeb8ZyM9sNszMAdvVzIzCqypWzkueXF2sAfDYEgcX6mEv7IFfeQXehWuCXVNNsreRtClLWpejZhteBTHcgPP0wXZvp83RnUm1kzhWnrOZrFnrmApyeHVbHtZl9HIyhykZhCvr8k3S/kq1w1t7LiZTphZk+TTBc6xryKdo2QKTbYdwo1JF/hrPlxxYG3thxCgKceoif0bEZevKMVfKbpbhYJQfZfQvC49+Eu/XjUJf2wem7lf0tki1MJLzCs1wsWAaAk9VoiM7NEPlxEpniTYBfIH4p3Q6MHyeeqmZ+gWiAmQEWQSNJAIJNVoIyd5l3SDPGG2oyAHb1nj4PHVZRdwOEQ3zN9AVCMcBV22QvQ7SuIRW0dS0LcUGZwZ8XZ5aplIXKXiaMQgjI7ltgbKsp7RepbeQXIQFWXp2I7UNdgs4OUr6ywKyOEAK6Wqhflx47xgGRn2Cv6fw4FQDDKjAXoRhyUGKoq3zyK8aPwxQmoGEY/BrDgZ1opt5Uqg1QIemaXrzesAUq4G5VLUJnL0PPD8NpWfX6nmlQAcIKnNX31Nltct37IBJNFEW2sHWEPpx176sz9yAknE0fBbwY/25ps6Kpz3YRApztP8Ud7B1gN9xkEPFGoHUtpBcHILgTGA3oRqChEyKapNyMlUEBwPSoUUDTSshIAly5cnW2mNO8kq6FbcTubPoIUaZNK+HGMgzQ/RJksplc6vaN5F03raRi3+LcvQrr+kxO+wbuYEEJTt9uysasewAmN8rWudPn6dJZ6X258nZ25kk0U8nPqt/Jvj2LGr+HEJkuOKveQ3fEcanM4bhAogmydxeFBSIpBuI25qijgTUbwS+Oo0xpht1Pa6LP5SyQaCU8ZPYSTCwD4cU4wRZJ7ggh6sQrRBILnyFdCPeGGzo/st1QMcMrmdG17IaA/93fQOTDv72kuV946nsQXqzeIKT+PmOgzjwKde5xDuJkK/kRbhTB819iBRkCItUG785fgog1IDz7GAABd/0DAIDgwJch+/bAqfEBbPbqWhBs//HfgXfnL9cnoTEaeuw49MjLJAYtuSeFcN+fQM9chEi0UGAg3gjZfwfU4a9a6ukoVUPcGLzb/xXxVMZAnfg21IUnATeKyL3/DiLZjPDMo+RyGw0RScHd+7k6qclUCxZWQXkdBr+ROvkIkSTqXUtrMYlf5CLgEAOFyjyVBCfPsCVwNMW4yRgeX56j+verQNNvdLsp9jchZX2lUye/W6+41kxPnqk3P1zy+ugRBM/+IdxbPgHvjl9ccLEAiKYV8PZ+Ht6dv0Tpxpf+BgBgps7VK8wAe06gPL9wLUK+4heuzn6fsOhFx6I4DT300lXHmskzCI99A94dvwhn04cRPPm70NlLEELAWXsf3J2fgrq0H86mj8Ld+SmLqAVgNPxHvgB392fh7f08EE3ys6JpuLt/Ft5dvwLZvgH+w1/g4cZAnXsC6vyTMLbDD/wS9MQZmMIkwpMPUQjajQLVPNTxB4GwivD4N6FHj5AcNHIYauglYr1yo0BQhrq0n26ppdTq0SMUhb6J7Z231y2y8ODfwt3645AtqwEATqqtTrt01r/f1hdOQ2cH4VyDo/xmms5ehuy+BbJ9IyvqPbfU/ybS7VY3yqV4QLxx0RtDiin07FiSQXL676CE5OxFtuFdnOo0mguIG4OMN3MCGQWTG4dsWujyo8dPQLSsZX0hlgGMYkBfq/cEFStUZnvtCYftAFrXQudG4XZufVOf2Ztt75zJoK+BbC3NQJ17nGp1AMzcMNy9n4O76SMwswMInvw9QAVw178fzvafXHjjIsWJ62LX0IytcaYB1FGur9uukUk25Xn43/8vMIVJOH174FmxNiEEnHX3EcVruyYxFolC9u2u8zYAsNuoyz4VsmMj1QsrObYAaFwBJFvgxJuAWBpO705yqJMtgBsh5zqavvrCbiK7aSbDlf3hruQ/q8svwL3tcwstshyPHT3nh+Hd/WuAMaj8w89TYAtA8NTvQa64Fe72n7Jp1wWPUQ29BLemXleaXQgiX8+1mauvTU+zp1xdOtO5lljXFVXg1/UMUP8cdeohQIeIfvx3SRe1HAnjF2Hyk6xVCElfXzgMoqMp/t0vAgackDVMmJeoqwCauWFO3lrfifIsRPMq6IvPQjT28NqNokpg69qlQMKbyG6ayaDOPEr1OeUjeOr3AOnCu/tfEx7hepAdmxA8+n+TE93YB++2fwl3xydR/cavIdj3xTpnWPbuAkC2nB7Yh6A0C5Mbh7fnsxBddpsPfQSP/Q6h1tUCZNuGV702PXaUoLbSLML9XwQiaXh7f4FQbulA9u6C/9hvs9otHUTu/w2bkVkEj3YiCzCP0izCQ/8AXCH2ZCZOIjz1EBAGCJ74bwBQ/xzRsRHm5LcRvPAXQDkL2X0L3Ft+mhVi6Vj+c5Hwi0STTQRomLlhVp4LkxCROMUQghLbZIVE9ZrijNW/bWAwXi3Y3tcN1lUa5ILhF2Fa1960feVuimwSANswZBzVv/k0CT1uhIw56ULPj1LQd36Eq1lTP2QrRbFMeY5BnpBMwUaStTNCj59gxibRAtm5mVXa5/4YMBpy1V1U3m5bz1Tmq6x2pjwPMzeI6j/9Erz3/58Q6XbI9k0QkQR7ZKsqjF+CmbkAkeogl7qUhanMQzb3AwBhD26UtRG/BD15iu5Kz4461MJUizCzA6j8/c8h+sk/AwQg2zdzEBtDCcy5IStavJ4aUFfsWjWcFgAY5UNffJbdj6bOssNoYy/74EUSbAa55l42W7HgRDNzEaJ9Aymwzf2Q3dupChhvhDr/A7i7PnPT4plump1BpDsgU+2s8pZmiOuv5GDcGCAl1S+aVhJTZDT03AigfUAr5vH9IhUjJk5xQKQ7eGxxmv8vriUICadn5+u/tngGIr4NsnMLJ2OjvcZIot4qVwCAHfgAiJFaFOTKxe2tIgnInl0InvivxEFF+DWJaBLo3AL3lp9mvWHRoBNCQLStA9rWLb22xcC9KwepcFjk1ISQiOZ+mNwIZKaH7lNQhtEBZLKV8YYQdIMqObtLUPbfFGcgvChkw83dO+6m2RlqpkYOQx39OgADd+enWRjzC5SrTzRzi1eBLbAJGL9AvsD8KCultYDR+u8IKSEvWtfS9z73BACwEvsGTWcvI3zhL9mFc+299VrFD2N1BOvezxP7lBtjDUF6MMon8M/ei8h0M0g3mnil+VEAhmp9sQxBjXmr6eoXgEQL08NutN41FEYvdGaNZQAnAj16CLJ9s1X3sw0hpa1NQCygW3VIN8zipW7WneGmmwxLunfaoNfMDZL/kGim3InWJMXHMkBxmn3gspch0p384mIZALXG6Pa85Xnii5r76R5ZF4MDS5MIMzfMAFQ6qMO/pQOTn7Ro10jdF4cK7C6xwuozpRaKV9WCXYXH6ItnemFmL3GniDdSFMFLANEkTG6c/enGjvJzbVN3OB53Q+lCriFeqg7eK0zRr5cEMqJaoBaUF+cxiWZ+rsdqvZ4bYnGvNrmaVr4rO4Q6v/Vbv/Vbb/dFvBFbUKQm1t7MDcNMn2NHnr5bETzzh5DtG6HO/QCyaxuCfX9CaIVdNfX5J8lbuIJHEOz7Yysl34HwxT+Hs3IvUJ6DOvsohNEwlTzM/BDUoa/AzI0wQzQ7ANHQBT11FpgfJo3SXlt44EsQiVaIRCPCY18nSK40DXXsQQjbyled/A4EwIzN1FmSbWINhHHoAMJLwkyfg548DWfVXZzMkTiEE4FsX88Ksw5tD70k4Ocp99/Qyd3OGILs3Chk00ruIJEkhPIhW9eyvVamG7K5H8KL1vtDCEsYQmm23nsbfoH9MrKD5GdEkjftDvBKdlNUoF/NTGECom39ApQgKPFL923Wwy+wDW+1AHXhKaYRKzn6yYstrFKtW06day8AACAASURBVDowuXEGvLb1lhrYx+KUMXUtWFOcIbaoWgDKWYjOK/oozI/Q3/cS3F0qOdIqSzNWfHmSFd+hFwEdUgB47X1U9w4qUGceoc9emGQQ7UbrioB65gL05eehTn6HrL0X/gLCiyE8/FVONOlAHf4qyUiwVXPHozz/wb+Fnh+BHjmM8NBXbFYrAuOXEDz7PwC/UG+IGDz/Z/y8iZMID/0jkcHZQYTPfwlXZrreCXbTTwbZuRV69ChFtSrzVJm2mQ91eT+cjR+GawWDnTXvhey+hdCC4lISvtN/F9T5JyixuPJ2qAtPMovU2AfZs8O6ORru1n8Bp3c3ZGMvEaaR5AK3YfH5bvkk449qgbLzk6cgmldC9t/J8zb3Q7Sus6JkttGjVRpHJA5n7f0QkfjVhSwV1rNcauggZP+dVA80xgonsM6hc2OQvUu1WKEVRKwBTs8umLDCTJdhR1DR2Ac47oLwsVZUPNQhoedr710QT+vcgmtW/m5yu+lihmvZ4luoFYr0xachV9+NV/vSltAdy/NkvNWOVz6DcOufM3j0YKbPUtbR8a4aqEvOpzVMcZI9p0cOwVn3wAK61PaTA8RCPCHEAuit1oPBLwAR+xlS0v06+RCgA8i+PTDTF2Aqc8yYrbgN/nf/A7zbfgGyZyeqD/463F2fhrP+/UsySuG+L7K4lmimLmwsA9m5BSKaRvD8nyHywH9kcxgvDjM3DGfjB6lqOHUOzrafQLj/ixBNK+Fs/tg7zk26KSeDCX1bYbVBrIU2CCfCTFKtmSHAdKAbe82qaPD0H8C985frits6N87APJKE0Yoqe419UOefJNkorMBZ895XPF949OukhlpuhbsI7qEGnmM2y6YvhXSBZDOEEyE/Y+QIBZared5CosX2v25mgF4tssglpJ0wScAOdNmxCXLNvdCXnqOggBe/QuTAJ/hu8jTkir0LKFUhoM4+Dnfjh/gMFzHpYDRdT+nVZT2vPO87wW6aOsNi06NHbGBn+OUEZSDZCtm1jbBs5ZN8oqqAE4VzhXT7tcyUsmwfFZRZaBIgjCHTzaA00UT0qV+oV6pNaRbw4tCTZyCb+6EuH2AWJt7MQZ5sgXTj9P+3/Tj05QOWUdYB0dAJfWk/V2Ah4DStBKtoHiXujYIZP0m9JTcGqAwhFLFG6IkTkCv2Qo8cgslPQLatg2hZA9m7E3rsGLsD9d8Bdf5JwGg4fbeSQjt2jLFMUx+54zaOqK2Hzorb+LN06xmzOg22Vh13o7zOOiRGvGMmxU05GWTr2npOHNUcB0voW8hFG3cCyT5piCRfJ1bG8PhqHnrgOajBA2xCPnoEsr0KlGYo97LydgjhQOdGYPwS1OALcDd+GOGxb0C2rCHK043B3fpx+A9/Ac7qe+C999ehRw6zl0P3LST25MZggjJkphvhy/+wtCYRz8AUCTJ0tv8UawrZQRg3BtnViPD5L8Nr6IG68DR9+bAKkx+DGtgHZ/0DfBbGMFVbnmfb3d3/C9SJb8PZ/FEYFVrylL3ziVMWn2SJQW4Uxi9Cdm2D8OKcYOlO6OxlUmdb1/C1WIaT5Y2ADG9guyknw+LKrXGj0IMH2OshkmQPtgqJ8ADoKry+s9bVvcOX/gbO+gfYy80vogbdcDZ+aKEL0PwIwmMPsj9a+wY4QkBd2keXIhWFKecoJV+YZFfP4Zdtc/ZDvNZ4Bs7aeyl6fPDvrroakx+Hs+ruOvNMXT5Q14U1RlNB2y9CjxzmsWvuqTd/N/9/e2/2ZNmVnff99hnvfHOeaq5CoTB0A41Go5tkk2pKnB2SGbQZCtoMyWGHHPaD/wE/O/zmBzsctmwqwlLY4aBDoqSQTLLlVpNsdqMBsDHPKABVWZWVc+adpzPtvf2wzr2VNWBqoABU4n4d6Mw7nXPz1N5n77XWt76vu42/8t9Kb0Z/H/eh38Z0tm6+3t7APfGNm+fqC5FP1Vax3W2ZGEZPOglt/0DiI5OB58nvuQXv0Sar+x335WS4HUKks1KZ9QoSCJvso6UWbzmIh776E+ywgXP627L6RF25A5bmUeVF9Ov/Gnv6Ozh577H36N9Db74ogejsaZzOFqZ5TQZ4OsCZOYXNYuzem7hnf0H6rUtzYo219jj67e/jzJ3jbmlKVT8pso9hecK01e/+ENKR1AWWLsGN51H1VREeWLiIvvJjVH1NWlWtRr//I7xv/QPpY6guo71QuEf7b9/6p689JoVIpcRQBeQ7jdtqL3xPXps5JRPHWpyFi1CavYXte7/jvgygP0vY3PfMtm/IdiPqStVWJ5gbLwrr9PS3QSnM5suQ9CU2aV4HrKQiR22JAW68KCS/098G5chqkMXCQ1p+RGKdZCh9BK6P2XgevIDsZ/+M8Pf+pzu+l21clar3/FmpfQybqKA8Gbx2723ssCmp36CM3X8HO2hIo1BQwe6+KXWKla9BZRF7+J7EGKuPTThT1ogw2XHZ938a3PeTwaYRdtgQmoJyJLDObbGc2dPS5ugVhOSWjITHpByp2GKxaYTZvyx77bgnhaWRpCtt1BGahVKT9zmLF1Gzp0WcC3BuI8Z95PcdNPLgVAp4ygvJXvnn+L/833zo58ze21JJzrv2Pg7M4RXRVZo9LRVpxxXhM5OJUFlhRm4EuXmMaV6TWCwZ5EIMVpITJsuN69fuKlZ2XHD//2XpEEYd2Z6EFSjOSNW5fyDLeuMqE/W6UVvIeUZL55ZO5Lmck2MO3xMdoM6W7ImzRB6TZ450LBTmmVOo+QuAndxZx1KN48b6o7n9cX1BDArfBK8oUvKDBtYvilL3EWW9u0HlJoq3KIePrW5BJm9vR9Kz1WV5b29X3I86W5KKLc3LjSOLJeOmE8QSayAto4ND0Xwq1CEo5oodVlak8kKeWTq+uP83fMUZ1OKDuA/+GmpcFVYedthEv/Pv5PXyvNhK9fdxTn9HBIpdXwTC6idkJRk2ZEtSnJEB5wqtwSYDcQo99S0wRnyYt14RgYJhE/3m/4ttbWCu/Bjb38dceyZXshOYjb+Rn+tPAxKM2uY6yitgjRbhsLsIBtwO5biY9acxu2+g3/k++vV/hbny16IxtfUy+v2/RBXqZG/8m0mq1Fl6CEmPruAsXBTJmLgngfLBu+AVhGOVRuj3figaUYNDkdlxhT9le7uir3TymzeNHY8p7vvJILwbL2+MWRIDwt4uztLDkhlprMu2pn8gzFYvlMKa1cJDam+AjnO3S41pCrvV9vel+6s4g3Jc4QUNDuX51nXpIy7OyKQ6FJNx27gid8/cV9lai9l5Q5p1dt8QLpPRuejwCk5tVZinYwedHNYaTOPqHX/r+Lup2hq2s41prGMaV6XTTDmYw/dyHVjpB7ex+OORxtK4P2jg1E9JE09ZUtAqrAhLtlAX6snMSVEMzJUETeMqZv9duQbHPK6471irHwrHgbgnzeyVRZFXzCLR8ynOTMS7AGl/9Ao4yw+jlCs/HVcKXlEHm0U4CxdRteVJe6Yq1LDdXcnt9/dlAIG8zy+K4WF/X1QkSnPguOI55wVCDXc86aqrrYovHciEHByg6idvDjaToV/6Y/kuXkG2XtZKV1w6winPQWlBCHxhVXhFoyZq/nyuwJcy9p6znc1JT7izdEluFCefkO9SqAEqTwisS5bMDXFXHgWsUDVWHxWjxcL93ez/cXCs1j3leGK0MX5cmsXOnLyFrj15Lfd+ACaM05s/HxUi3aRFVOAsXBAZl7BKdv053KVLqLlzZM//M6FyKIUKq+iN5zGbL+A99Z+LVtHuW5IBMhp97RlUUMJdvIh+9V/g/eJ/hX79X+GeuVW41+y8jr78A9TcGdnKpZEE64Ua2bP/u/CLrMjG2NYGZu9t/LPfJf3x/yhpYXIS49YrkoH6+u+RPfdPZPWsrU7SwyonGbonvwmoScedU5qTv+fT/qPcR7jvt0kfBmuNpEtzj2ibDLFGS7A4fs6aCZ3bGi0Eu1xjyOo0f+/NhJvKyXm2fyhV3qB0S4yg3/5zvCf+AOfML2Cb1/C+849k1Vh7XASGizOi4JEMpZgF0qJ6G9TsGdzHf5/sdRH1sonQsRm1UbU13Md+n+yVf0H2xr/Bffz3JSDPItTsadyHf0cEwt7413hP/Ce4j/++xAvDFowTBLefLx/8k8fHfEt0NxyrleEOGI1e/2neF20lZei4kjKtrjDplht1cC/8Lcz154QiUV2Wgeq4EJRFSMC7Ke9i476IB++9k2u+OpjNF1DFOZFv33wJ01wX0tuYVn7jRdlG9fYwB++Kcnd7M6c6LN3x1e2wIVSQxQcBhdl+RahLiw9imuuo7Vfk3DoVNbv2JmAlHnJ8qY+sfE1k74uzsqpYI5Ryq+843xTHLWa4AwplMgmaCyLUq8IqKiiKSndYka6uYQtn7qyoVZTmJQPlFyCVvmoR+LppKIKOMRvPo4p16Wso1CAegutKTj+LJjGKPbwCYVUa5pcekiLZ4oMS0yw+AFmE9+Bv3CoDrxTO7BkRNH7w12/6Uo/aovxRXpDXLv6aKG10t7GDA9zz35NOtUJVVDhOPDExT1Gzp8X83Q+la+4+1kS9V7jvi24fBmuNtC4q5yaXprIo25TBoTzvF3IeTt4p5/rCHSrNQdS52V+gE0lLzklrp2lvYq79FPfir2G2XxNuT3kB076B++Cvo9/9C+FQeaEcL+lPeiNUeUFYrbVVnDx3b9NIViKd3lyFxr3eY2upnGp9t8KXvvoTaRy6bZDbZDgpuE3x4Tjm26RMBqoXcpOGbFH1k5iD9ybPOXNnsUlfuPuOL7yiC7+at3X2ZFuU9wO7c2cBhaqfwH3878t5vFBijMEhoCa0DOUVhN6x/BDZa/8S98KvSmFv/sJNgQJk0ur3/gLn1JOY9WdEnCufLPbwikjAKIVt3UBVFsU4ZdiUmKVYF3n9+kkh7kUdOXehjirOyHEXLmL9gkxEryD0kUJNHhdnML1dlPKkGam6KvTxUSfPvFlJ/86dvWsi4jjheE8G5QprNZdWRMcTfSQ1f14GhE5Fla80Kw0sY/l7yCuuqyImMO4+O7pdymFH7dyJZ2tCxFPVJWzclTrG6aekOj1eJfKq+TibNTFo9IqSDk1HmMP38ZYfxgwb2H2N8ktSUR8cyCSNutLaWT8hVWU3AKxM/qCEqiwJM9Vk0iTU3xcbMNeXVak0L/SK4ow4CHkhdnCIWxZVEKlzWJTjYaOO9Fsc85j6WG+TPgmslTZLFVYlm5RrEwEiljVsyt03HQmFA1G1IBkKyc/1Jh1nyg3EIiq/u+OF6Cs/wj31lBgE1teEgu0GUiMwmQzI+olJ95q+/hzuxV/PyYNtGbx5HzaFuggRgAzwoHKzNznXSrVJXyRochkdm44kuA4qeXFwDbAyQbME/IIE7Ktfn2wrJx2DWQy3ZZuOI473uvdJoBP029+X34dNYZgivczps3+E2XtnQr3IXvsTGczWkD7zv2LaG2At+tU/QdVWyC7/O6lHzJ+THgvXx6w/e7MVtL9P9vb3MY0r2MEh6dP/C2QR6Y/+B1R1WeRvOjtyF486MGiSvfB/5cH8gsQOxRlZzaorEFakEWfsZlqckYk7bIoXw9bLmJ3X5W7vhdLVt/0qZu9t9Ft/LufwCtIG6peE2OiFecKhgirP37OJMBYw+KT/fdzPfhJMJ8MYuQyM/KqlzxggG6FKc7iXflP4O8sPiR7SwgNy125ek+64zZdkUFkLUS/P2t40WbfpIJeLSYW5OmxC1BNlj+WHRUe1fyBV8NNP4cyflYxTdRnT3c691wRj7aijv98yWMePvVBU94qzqMKMkBb7B6I9W5qV6nyxLqoZR47zea4AFouxhsQkZDYjNRmpScnynwCpSSeva6sxuZJfalJ575Hn5fPyn/2EcjbHO2b4mLDWSj9x67rYSoUV9N6bIp1y9pekNnD1x7hnvyvum7OnSX/yPxP8B/+9MERzI3X8IubKj7Dd7Yn3ws2TGEx/D7ZeFqOUXBFDeSFq6THhAt1Fq9RsvSzV8vqJT/6HBWWctcdvslyXHgQ3wH3ot6SJZzLov7jtj0KBAm88FBWT72VyhyZHufnT+Y0l/6yrbvpbAGhrcD9FkD+NGRg30lwhfe6f4H/rHwpdOuoKUW/pEgybIq47f0Fo3N0diQ0WHshfO0DNPyCB7KgNfumI7Wy+pHd3Ze9tUtTCRQloRx3UwgVZYVrXZSUoLwgB78bzeN/4+9Ip11iHYh1Vms+p4s7EjNz298CYvK4xK4LBg0NR0st7mieaqEoEB4h78vzEw82VFc1oUey7B1mjjzPMJlI6kysnq8Z40H+mUPlEPPrUdDIIrLXod38oIltnvvOB79Mbz0ugvfxI3gykZauRjET0OB1Jsevw/VwYWDIyQrjT0iPhip+E2X1DpGSCshzHL6JmTmGuPydeErms/uTcN16QbJTrC7W6vJAPagXpCPf8r4jXWtTNva+Diauoqq1OWjf1uz9E+SUpAg4OcjXtXUnbnvjGrQXAz/D6GgzaaHzHJzEJnvIwGKy1+I4IFmcmQ6FITYqrXDKb4SgHV7loq3GUg863da5yyYy8rpTCd/x8q5ROBrqxBi+vy2ir8ZRHZjICN7hjOzjdJuVQSuFd+g0AuRsfvCeV49nTMGpJx9jCA2KK8tN/jFdeEAp31BVlbE9LKnVwKJ/NKd9goVAXH4nmumSYVh/Dmgy99SrO/AXwS5iDd6WBZuaUFOtyzSWszXsPQmGdRl2pI8ycmlS6xyrYprMFWYSz9pj0P+y9k9PFl6WK7vqAwhmnlX0xdlHpCFtZFDOVXDfqs4K1drJ3z4zwvMZ3e6UUjnWITYyfS9GM4wGLHSvSYKzBwZGBrTwy5DjjFcNyZ7BsrMEZC1NjUSi00Xie94GxxDSAvhvSETbqkj33R2Ay0mf+N2zUIfvZP5X0adzDHL6PM3MKtfywpB1razgzpzF7b0lRK1f1AyVey7njjaosoxYeEDPBYh1n6UHMjRdQxVn0lb+WukD/QFQsDi6jX/sTTGdT6hVBRbrykgFOfU1SutaKAIFy0K/+idQr3EAyVyaTyTd3Lu/lKEiAPH8Byrk0ZlCG8gJqNg/YP+MtUmYzukkXbTQODoEbTO70KtdcCvMJOF4hPMej4BbwlPwsuAU8x5t8tuAW8B0fV7m4jkvgBAROIN2CKHnsBviOT+AGeMrDVa6sBihC9+4TfjoZ7gYvxBxclpy7NeKEc+YXJd06aovO6cbP8lZN52YWpjSLU13BWf2abFEWH5QKdhZjdl7Dffh3cM59V5bnQQP39LdFjfv6s/J8/eTES8FmMaq2hn7/r2DYQq8/IzGA0Tetf3Umq5i1+XanLzSR7VdE6WNwKEJit+H2zNG9zCB5yqPiV4lNgsHQTXo0oyapSTkYHRDrmESndJIOqUnpp326SQ9jzS3fa5SNyEyGsYZ23GGQSravl/TppT2UUrTiNsNsxCAb5sfpEutYVgaliHUsE9Pqu/6908lwNyRDuQOHVdlaGIO+/P/hXvw76Ks/wXv074oB+7D5gYdQypE7tk5xFi/iXvge2at/gn77z7A6w+y9hV7/Kaa3i3vpt9DvfF8ag/wSZvs1YbIqB/frvwd+ficzGn3jefTG88Kq9YuY9adFNr+6mvcmWFRtDe/R/1Ca+o+wbb8IHO2JGGYjakGVSEf4jk9mNQW3QCduUwtq+I5PK27jKmeSVh0jMckk3ZpZSZumJqXghlR9odVHOqLkFTFWVqFRFt1yHM/xMNZ84DZpGkDfBdZaaaZHTUw87LAhqc+4JxXgvDqrjlK7dUb2zD/G+/Z/gQrL2GFTeqhrJyTI7R9gTSqeDtuviT/C7BmpB3S2JSMUVqSFNO9VRjmT1whKmK1XMVsv4V76TVR5MScQ5u2dUWeSorVRWzJbQQXn5DdlS5Vrt449nJ2lBz+XaznIBjg4pCajFlTZHx2wWFxgb7jPanmFTtzFUYqCV6Adt2V75BUp5h2G1lr2Rwd4jks9qNOK2xhrWCjM0046FNyQ0A3ZHuywVFzEczwG6QBjLb7rYa0ldEMG6QBXuVSCyiSeOIrpZPgMYbMYc+MFnDPf+UhJFZsHih+0R7fWTto1b3neaMy1Z1Gzp4TmfRtM+wa2f4gq1iSDFfdRiw9im1cnftRWJ6iwkguY3VtYayfZHdeROGEc3PbSHoETYKwlsxlBHgfEOiZwhVzpOR6JTtDWYPI+jKJXJNFJnkUSey3f8ekmXSp+BVe5pCbLs0j25pYLRWY0/pFi6FFMJ8OnxO2ez4BklHQGcW9iK+ssXZIV5+CyZKFyS1pVqEvdQiciE9/ZFiUKozEbP0PNnpKMVWUpNyIcSXq1UMt7nV1wfZzaKqa7DWksWabqivgrWIMqzoqr6crX8ly++VwZqOOM0u1349SkbPW3SXVK6IWU/TIKxSgboYDUZKyVV7nR3yRwAhKTELgBq6UV2nGbfjqgGlRxlIOTX//t/g4PzV0iyiJSk8kWTSkKbiiTzqRYLEWvyCiLKB9RXZymVj8l7N5bWKNx1x6bPGea10VUwBpUMsT293JlbSNUjP5+Tv1QOPPnZGtTnBXljM4m5F7Vsj1rYnffFA6SX5LC2uBQ0r2tDSnGeSHUVrGN9dwDbg01expz43lAoebOYTubuCuiHn4vJ0KiY1pxh6JXINUpXl5TmC/McThq4CoHR7kkJiF0A5ZLS5P6QGZSHOUyE9bzxxrP8ThRWcNTntQJHA+FouyXmSnMAIp+0ufFvZeZK8zRSTo8xCVeb7yRrzoGbTVna2eIdQLA3nCfE5U1tvrbfHPppubsdDJ8CKxOMDdeEnW+U09iNn4mTNP582JIGJRQCw+Q/eC/w1n5mmgpgeT1Fy5I+nX5YXR/D73+NM7pp3IXUnDP/hJ6/aeiu3rulyXYdQOMyWTwrzwqfssH76KWHhZKRmXppuIfVuKZoCyCwu0bUklWoCpL6GvP5ibwT0rtIBcJuNfY7W3TSrs4yqPohaQmw1Eu9aDGwXCfclDBWEOsY8p+mXoeOHejNp7j0Uu6nKydyYtzXl4jEASuxGfWWhrDfRbLy/hOQMkr8rX5R6jmwXlmMtbKa1T8Mq5y861WTNWXVaQeVHGUyyNzDxFnMb7r4Sh3mk36UKQxtr2B7WyKhPyoTfbyH2MP30cV65LLVw6U57GN9ycfU/UTUKhLf7MXYDZfksrv1Z+KPlN5Xvb2UUcySF4omSDXxznxhBAClcJsvohauoR++09le5RF4hNx/Vn0+tPoGy+ISeLcWZHLX3hAqs0zJ7F7b6E3X55IzKvS7OdCwIuyESfLa5yrncZDUfPKXKif42Cwy2p5hdXyClWvTN0rMxvUSbOYzc51ykGFQdqnElRpRy0GaY8oG3Ew2GWUDjkc7t9ynmE6oBu1SU3KXn+bxdKiVLbTiIPBHovFeZIsojHcQ5sMbTK2ezdIdUKcRVhr6CddhmmfOL9BTSfDR8CmQ+kZqK1KQOoVcobrABwPc+MF/G/9A/Tlfz/5jFJKagWDQ9kOOa4M0qAkWai4L1pMWTxhyo4/h06EYt3akNUiyv0nlMLsXxZHoWQozyslg722JrFDcUaoHW4gNRKTipbU5wjP8aiFdTKTUfEruQKJxAu1sI6rHDpRk3JQJXBDSZHqBGstZb+CNhndPFvUiVoMkwHVsE52W6o1cAIqYY1Ux1JLyLdZpaBM4IZEWUQ1FF2oRMfUQunaG6YD6oVZSn6JbtRGW003lv6U6TbpI6CCiqjUeSGqtor/C/+l3M1B5N4Byou4D//2rR/Uqez3hy28b/6n2MEhzumnZGuTDFALF3B1IuYmRzH2eG6ui1nK7lv43/rPRCvVDUCneI/9x1IMHAuMuT7uqSdh1JnQt1VtDfeh377puPM5oeiXaEctZgtz9G0Pz9F4jmxVmsNDZovz1AuzJDqmHIhm00xxFs/1JoW1xfIyw2RAJahSDiq0Rg3GzFprrXCUHI9u1GG+tEjZr9AYHTJXmM8HtqLsz9IYHpDqFD/nIVWCKvXCDK3RIQWvxFJllSgdUc5986bZpA+B1SlEHckIxT3Zn4OICRTqN7VHlZO3buZkGgBrJRs0czKnZitp9wzEiHzC0OzvC9EvTyWCMFxtZwvn5Dc/USO/aa5L/WBsqPIFwI5FDMaDN88iHeUoDZIelaA2eV0dee8H/d4aNZgvLdKOO7y4/xJlr8x3Vp7CWMMwG04yUWOMr29jeMB8afGDv29+HqXUdGX4MCjXh1wHVV97VpTqlCO6S6O2eL65nvir3fiZVKt1KtKNubuNM3MKvflSXjzbFJMPrwCzp7HNqyJGptO8Z7ov2yjl4ixexFx/Dvfcdz/29/086gYfhkQnkwqvq1wSkzJIB8yFsxOW6TAb0s9GBF4Ba2GYDSh5JbQ1lP1SXlPQE+6StprUZFSCGolOsNZwvn6Orf42mc1ojJr00h4XZ+5MECilJhPhg+KloxNoOhk+Jpylh4RvNGZ2mkziiPGdW7miLGEy2SI1r0kLp8lQQUX0kpYewg5bOLMz2GYu6FusQ1iRHoRCHYoroMCOWuKjlsvjy5NaGvTTISqsYeOe6K9+SbCXB7njGGGQDigHZdpJh4pfxnEdGlETay0Hw0M8R7ZGoyzC5JNhb7hPalKKXpFqUKWf9OklPYp+kdAJQcFmb4ud4S6bvS1mwjoj7U3YsGM69xifJGkwnQwfE87saawxsifPt0tHL/TRLY3NEvTlH+AU66i4LzIvvV2clUfQ6z8VVbzNF4VFWpiRXuiog/f135scW994EefEN9Cv/Uuxq7WgvED81twA98x3xF30SzQZ5gpzFLxQZJ4wgJJC25jQCKyVV3GU0LEVarKhGm+hVsrLODiTbU5Q8JkvzslWBsUgHeA7PkvFRc7WzjDKRrjKxWLpJB1mwhk57l2KoR+F6WT4BFAfkpm5fW+v6idEhmX7NVlVnFzIbNiSWMMNvAY6fgAAHNhJREFUpDDWuCo+a++/L3ykfDLYxlWsF0pRTeWdaIBTnsfsX8bG/S/VRABuVnMVwJHrcWQ8ekqG3N24QdZaPHWEKqHA5dbrGrhSiV4uLaGUop10SLXQPRzlTBqDxjwkqU1IRgmQWkQWk5pkEisU3IJQRaYB9GcDm6veKcedtJGq+Qs3TcfHP4dNUesbP7Za4ox0JKnRcUPK2EVn6eGbgym/jdrD96V19C7cpS8bpAJsADvpYvMcj1jHFLwCsU6kuGYtbzXf5lztLPWw/oHHa4waWGTiFb0isU4YpgNmwpkJAbCf9idbtlpQk3Ru3EVbw9naGdpxm8QkjLIRoDhRXqXkl6Yrw2cGo2XrU1uVHoZRW6rF6Qhn9iymcUWkV6orUktoXBV6htHY7o74Q1iDLc1jdl6DLBHqReua1CNyAxJ37bGJpRVf8okA0E26XG69x0ppmb3hPjNhHaUUh6MGX5t/hFcPXxdCnYW94R5nqneSD4+i7Jd5evsZTlRO8PDcJTzl4h3pkrNA1a9SqUvadhwgzxXmJo8Xiwu30LjH75kW3T4rKAWOJ/RrnYj6BUqsr/p74qOgMzFYzNW/rcmkA600J6vDuPF9vFibTFKxY0r5fbiIO8qln/YJ3ZCZcIZ+fhdPTUrgBswE9bwn2We+OP+R/vXDbEg76bA72AXgYHTIRm8DbXXe6SZxgpD3bjZefdDjowH3dJv0GeFjX0aj75SROYJbFCLSkVShQThNXnjfqdqNszzDbMQwG1Ly5O9JTcZMWOdgdEjFL6OUI9IwOcN03NU2Gaj5/yyw3l1nmI14cukJukmP7cEOl2YuEumIglv40GtkrZ1s0W7HdJv0GeFjD1L3oy/55FjBEVN3/85/vPsB4/rCVn9ronhhMHmjz6M0ogY7g10WiwsM0gGtuMXXF77Odn8bYw3VoEon7uR0jhquctno3SDSMU8uPUGiEyp+GWMNoywidENSnXIwOqSTdAmcAC/vo8hMxkp5mdcbb/Kd5aem6hhTfDE4Xx8XBNXk/13lcrH+AJnVGKsJHJ+iV6TghtSCWt53UMBTLo5ymAln0NbgOz7LJTF4KXoFWnELRzlEWQRhfZJq7cZdVsrLHIwOKXkl9kf7WCxz4SyZyfBvi7mmk2GKzwVj+vXt8F0fz3psDbbZGeyireZEZY0rnatU/ApKwSAdUvJKzBXmuNK5SjWocq17nW8sPj5RwwDRRYp1TOAELBYXWSuvolCcrp6Sn+kpyn7pDvGwMaaT4SuKMW1i3D7pO36+J7dHmEV3rwfcC6yUllkprUzy/99e/hapSScCYGPoXCTsdPUUIMW6kl8iyxt/xtXro/WJ8eCv5sTAD8J0MnxFcTA6JDWpiP7m+27X8QjdgEE6ZJgOKfklVssr9/y7KKUmxTjPkQB7rIl0OxaK82z1txilI0Amx/7wgFK+var4Hz7gPwzTyfAVRT2oMcpGlLwSqUknKcbxKlENKjh8+ayvqn6Fb688xeHoEGCSHi16RYpe8QO3QB8H08nwFUXBK0yU5cIjkpI/D6fnXmL8fbTVEh+4Ab7rU/HPTF4vegW00Qyz0aTV8+fBdDJ8hWCtZb17jVjHnKmexsmFfSt+GWst24MdukmXtfIamUlpRE08x+Nc7Szr3WuAZG9Wy6v37Pu14w5KqYlIsMVQcAtcbr2Hk4sL99Ies+EsZ2tnJulaRznoT2npO61Af4Vgsby0/zILxQWe2/0Zg2xAK2oBcud9cf8l0SQyCS8fvErBDXlm5zkyk7HeuUbZL/F28/JHnOXT4XrvOhu9Da51r7M/2qeX9BlmI1bLK5T8EhY4Wz3DfGEebUXR23d8nFyF79OsaNOV4SuG0C0wG87QT6X3up/20UZcb6JMtEivdK7y5NI3JS+ft1mW/RL1oP6J3XA+KR6df/QWw5Fxb4RSikUrjVa3D/h+Kt9plEWf6tzTleErhkE24OWDV3hy6ZsEjk+Uxbx08Aq+43O+fpZYJzw69zCe49JPB5ytnSF0QzKreav5zqcKUD8K4rHg3cIbch33DsuuowjdcJLxKt6FYvGJzj/lJn11YKzhmZ3n+M7Kt/GOBJmZ1ZPH45bL8fNuPijHYlw/232B76794hfy/T8MEm+0qYf1n7s2Mp0MXyFYaxllEaWf05nHWssoV7r+ssFa+7GIeh+G6WT4iiHKokn2JTXppBE/1SntpEPg+PTTAbOhSDf6jlhNRVnEjf4mJysn2ehtsFhcpBk1OVM9zdZgm+XSEhu9TZZLS3ST7l0b9L/smAbQXzG823oPpRxWystc616n4IZ8feFrvNl8G1c5E9XrdtzmYNRgvjBHN+lS8orMFmZ5q/k2Z2tn2Ojd4ErnKhdnHuDl/Vf43Qt/j1cPXuNM7TTbgx1OV099oEPOlxXTyfAVw9naGRzHwVMel2YfpBt3AagFVZaKIquSWY22mpXSCkopzjqnORgd0o47PDR7iZJXwkGx2d8i1jGhVyDRCU8sPU4tqDFfmONgdMjJys9h1/sFYrpN+grjaLX51mEgVL2je+8oE7edo3whUavzJz/vd0wnwzHF2AzF6AzH9cnSAY7j54PeoBwPlTtuKuWQJgOxnk1HWKvx/DLKcfMOtAzXDdHZKPewc+U5L5TXtXgeeF5BNE/TIa4biqK4NTiuj9FJ/lyKd8QT4cuE6TbpmEJnMWAZ9HYoVZZJ4j6uF2B0ymhwQKE0h+sV8f0Srhcy6O0IUc8N0OmINBmQJgMKpXmSqEN9/gGG/T38oIrrBQy625QqK7heAZ1FpOmAav0U1lr63S1qM2dJkz5J3MsnpqVYXiIaHlKfu/BRX/8LwXQyHFMYnZIkPfygkt/NHTyvRKI7FEoLeL4MYicUzdNiaR6UQmcxQVBF6xgvKOM4Xq7oB0FYw/NLaJ0QFmZwvZAsHeK4HmFhBpA+iCCsAwrPL2GtQWcxynFxHJdCce4LuiIfjek26ZjCWovOIjy/eIv0IkiMMDEfz9Xu8k/dIvZ781hGSH06wT/i2Tw+5tFjw01DcmsNBtjpb3GyeupLw4T9IExXhmMKpRReXlzTVnOtu0E1qJDohJOVE7zTepfF4gL7w/1J3aHgFZgrzBG6Ae+1r6BNxonKGjuDXR6cvcj2cI8zNVH43h7sEOuYil+mHXdZKi3m/RFF9kcHnKycZJgOmCvMsTXcZTW3ovoyY8pN+grAWIPnuHSTLodRg1jH9JIeB8MDil6Rfjpgb7g/sZHqJwNqQXXCGA3dAoN0SOgGk6xTP+1TcENxyzEJqUnZHuxMOucG6YBG1JxIQR7kzThfZny5p+oUnwkU0la5Vl1llEVYLA/NPkjgBqQmY7kkBbh6UGOYjWjGTdbKq1T8CqEbkugEz3FJj/Qjh25IwStOVCx8x+eh2QcJ3ZA1yE3O5f2nK6eITfIF/fUfH9OV4RjDWkszb9CZL86JuV9Yyw3Ch3STXr7NaVMPaxS9IiWvyGJxATd33Sx6BeqhTJKFwjytuD1xzXTy2KPiVyZmIc2omZsaOvTTAa2ojaOcLyWf6XZMJ8Mxx9vNyyil6CTdCd9/s79FP+3LKmElaH639Z6YBQ730dZM+h3GeO3wdZRSvNO6zEv7LwMwTEcMsiG9pI+jxHsh0jHvd65M1PAut9695z0QnxWmk+EYI7OazGRkRoOFQTog0QnD3Dvt/fYVLJbrvQ0emJHcfy93v7y9byHKYhItMi79tE9qUt5uXQYs/bQvrjpYKn6ZwAmwWNY71zlTO31PeyA+S0wnwzHGweiATtKlGTUo+SVG2YitwTbnamdwlMuZ2unJ1inJtz7DdMhh1KCam/6NYbFc7V7j4swDPLn0TbQ1PDL3EKEboq3hRm8TgL3hwaQpf7YwMznu/YBpneEYw1rL83svslJe5lTl5OT5o1yko79nNuNwdMhKaWViTDjGS/uv8MTi4x96voPRIfWwNlG4O3q++wHTyXDMMZZqv12Z7m4YO3I6ymG9c42KX6bgFcnyvgdHKVKTTTRRG1FzYkIYuAGLxQWiLOYwOiTRKQUvRBudm5LEuMolNWku9GUx1rJQnP/cVPs+CtPU6jGHq1w+asuuTUaSDvG9Il7OPg3cgEE2JMv1S/3cEMRaSz9qMxPWaUZNAjfAWEvBiitPM27Sjbu4jkscR4BknPppH537uEU6JjMZZa/EQnH+Hl+Bj4/pynBMYK1FmxRjNL5XIM0iHMfFdXxMLsCllCLNYpGJd3yUctAmY7/5PoWgQnewz9m1J+84nuv6WGtwHR9tUnYO3yHwitQqK/heAdfxyGwmxizW5OcfkVkjK4tOhBvlBiTZCM8NMEbLT6txlEuajfC9AplOsNbiez9/++bPiy/H+jTFZ4Jr2y/S7N4gSYfsHL7N9sFbGKs5bK2z27iMsYbD9jX6wwa94T7GarYP3qBSmqc/alApLUyOpZQiSYdc33mRTCdsH7yFNinDqE0Ud9Emw3cDtg/eFLarhf3W+/QG+8TpgFZvm+291zBWs7n/GlHSo9XbotvfYxS12T54E4CD1hUyHdPs3sAYTX94yF7zPcynFAT7eTCdDMcIFsNMdY046RMGZTIdg7U0e5sUgiqOcimGVarlRTKdYkxGnAyw1lIp3WkhVQzrKOUQ+iWipIe1GmMyUApjNa4bEOUU7SQboXVKf9RgFHWYra7Jexwf1/GpFOdpdTex1tDqbRMlfSyWNIvJdEK1tIjjeKRZjO+GR8iDnx+mk+EYIc0iWt1NSsVZPLeA7xVRyqFeXiYzCWAZxT3RLPVLtLpblItzBF4BrWViHEWcDijkKdbQr9DobFAqzFAuzGGNJtMxxUKdg9Y6gV/C9wp4XoFKaZ5G5wZZFk+ajACW5y5iMVSKc5SLsxy21pmprNIfNmj3ttE6QZsEx/G+kMrENGY4JrDWstu4zPLcxfyuOvZZUDS7G8zlbNMo7hLmPQ72KGXbiuH70cyOtZZWb4u52kmsNUfo3vYW6re15sjxABTWGnYbl1ldeJhOf5d6ZeWW9x49hrEarVP8XARs/J7PO2aYToZjAmvtLYPo6D/rOF261XyXtdkHuL2/+YOUt8fHdHJv67u95/b3Hz33rZPkTjW8DzrGR53nXmGaWj0mSHVMq79L6JcIvCK9UYNyoU6axaRa9uGHvS0c5bFQO4mf20pZa9jrXMNay8rMeZr9bearJ9BGc9DdwHV8Fqon2O9cJ8wbhWbKyxx0b7BUPz05/07rKkop5iqr9EZNUh2zUD3Jfmcd1/GZK6/QGuxhrWGhdpL9zgaeGzBbXmavs04xqFIOZ2gNdikGFVKdsFA9wUH3Bou1U5/LNZzGDMcEruOhlKIftdhsvsN8dY0bjXfYaV9ltrxCe7iPwmGuskKzvzP5nLWWUTLAd0OSLKLZF3/l3qhBtThPL+9DaPS3qBbn6QwPADjM6RdjHHQ3mKusctDdoNnfpl5apDXYI0oGeI5PZlLKYY1CUMZRLp7rUy3OSlpWZ9SKC+y2rzJfWWOr+R7N/o6wbo9813uN6WQ4JkiyiCgdMltZoRhU6UdtSkEN1/Hx3AA3F/TtR23C2+jUSTZilPZxHY9Ux7QHB/heyDDuYIFMJ6RZQpQMMNbSGe4T3Cbya6ymH7Up+BKPDOIOBb9E4BUIc1PBOBvl9Q2F63i4yst/F8uqYlChH7cn7+8MD+74rvcS023SMUHBLzFXWcF3Q07MXmSYdDk5f4lUxwCszJxnsXYabVLK4czNDypFKagyV1nFdTzOLz9OphNKQQ0s1IoLOI5LvbwoA9hx8dyQMwuP3nL+wCsSeCHlsE61OEuSxZTDOqFfmqxamUkIvFCMydMh1eIc1lqSnFq+UDvFMO5wZuERtMlIdcLJ+Uuf2zWcBtBfcYwHY/gBYsTGGpIsIvAK4rWcDOTOfVuAO4y7lHKljTE6owb1I3QLY3QezLv0oxblwgwKNdmSfdGEvuk26ZggyRt1rLXEmThhWmvpjJoM4h6pTuiMGqQ6YZhI484oFw4bT4RR0qcXtUl1Qj/u0Bk1yXTKQX+bQdyRY+to8vmjMFhGSZ84G9GNWkTpkI3me3RGDbTRJFlEN24DEGVDSmEdaw3dqIWTB/PdUZN+fp5u1KIfddAmozNqTP6me4npZDgmaA73J97OBz0JOtujQwZxh1Hax1pDlI7Y6Vxno/kuAJvtK5PPW2tZb7xDkkX0orb4HQwPQEGcjWgM9vKtS8q1xjt3nH+j+S4oxWbrCkkWsd25RmZSAPa6G7x/8DpJFpGZlO32OkkWMYi77Pc22Wy9T6JjulGLrfY6g6RLe3jIKB1MVq4bzffv+TWcToZjAGstqU4m8cEo7ZNkEalO0Eaz07lOqhOibMhCZZVUp3e90wrJLsRiCbyQ0C/i4BC4BTzHx2JpDnZZuS3VqU2GsRbfCViunSYzGUAuP1Mi0TFROiLORvSiNqN0QJyNcByPajiT08wNvhviKIfMZGibsdvdQFtNL+6wXDt5x/f9rDENoI8BLJYoHdIaHrBSO8VsaZHOqMl8eZnD/i4nZy8QeAUKnmicKiX7+bX6uckxlFKs1s8wTHr4bsgoHVIKqqKClyvjeY5P/S57+1QnWGsYpn0KXhEFrNROUw3rtEcN1upnmSstMUoHzBQX0CYjSocsFWaJ0yHz5WVCr4jFYqymGtZJshG1cBbP8SgHVT6Sh/4ZYBpAHxNE6ZDmcJ+VvJXzw7Df3WTpA+602mTE2YiCXyZKh5SCyl3fl+iEQTak4pfxHZ84i3CUyzAbkuR2tMYaAtfHWIunXCpBhX46INaxeLbhTLZ2c4VZQCZprTD3hQTT08lwTHD0n/GTDqSfhwLxT9/6P3mgfoHrvQ3+8NIfAJCYhIPRIeuda4RuQMkrkdkMbQ1Vv8LFmQd4dvdvKLpF6bFQ0uhTD2o8Ov/IF27IPo0ZjgnGTpgfZyD99eZPJr8fjA754Y2/5Nmd5wBojBq086zPevc6USZxSGYyfrz1NN+/9gMao4bEKSajE3cx1vDj7af5yxs/ouKXKftlLs48wNnaGU5VTrI72GN/dMAwG3GycoJrves4yuGRuYfZGexyvbdBohNeO3ydH239mL+48Vf34Ap9NKaT4Zgj1glxHliP8ZPtn05+/7/f+X+YL8wTegWstTSiFp3czWejuzH5rKvcXE7GUg/rnKufZZD2+a0zv8HWYJuD4QFr5VWshSudq5T9MpWgQiNqUfAKXO9uMMpGzBdEhfvB2Qd5r/0+Ja/IWnmV1KT86fqf88TiN6aTYYp7g2bUpBE1b3lumA0ZpANSkxLrmH7a58/Wv4+xhljHRDpGG+k0OxgdMEilHhE4AUWviOd4XG69yxNL3+AHG/8ebTWRjrne22C9ew2Aw6jBMB1yrn6Gql8hNSmOchhlIwpuiOe4ZFYzyiIut95jb7iPBQLni3MAmk6GY47UpDi3/TP/0uov8PT2MzRGDf7woT9gkA753olfQVvNm823ePngFdpxh7XKKnvDfZ7d+RsA3my+xUppGYDvnfgVLrfeYzac5WTlxESS8uG5SzxQv8DVzjov7L+Eg0MrblMPatSCKpdb7/F+5ypvN9/h0bmHKXpFKn6ZU5WT/Pqpv83TO8+gvqBhOQ2gjznebLxFNahyunqzNtBNuhO1bNfxGGUjfMdnqbhIM27xVzf+mm8tf5MzVaFoj/shRlnE9mCbC/XzWGvJbDbxeQCwFkI3YJgN889ByStOGojGvhDbgx1OlNfuqt9ksfzRG/8H//XX/9HndIVuYjoZjjn2hwfEOuZUVVKp1lr+cvNHfGf5Kf7t+p9xceYC+8NDGlGDf/jwH0rlOelQcAsUJ51nlkbUZL4wR2YyfNfHWMN65xplv0xqEkw+Oc7VzvJe+318x58EzM2oyUw4M0mffhQG6ZDyF+D7Ni26HXOMspFoJx3BweiQftpnsbjAMB1xrn6WQu7Io5TKDdFvxV9v/YTHF75OLajSTXosFBd4p3WZizMPMFeYY3ewy/n6ORpRkwv185MV40Z/k63+Nm833+Fr84+SmIRTlVNc617jVPUUW/0tLJa18iqb/W0eqJ//QiYCTCfDsUfZLxPcZkv7O2d+k0pQ4e+UfhWFbE8emn1w0rZ5N1hrePngVVbLKxOF7ZJX4rndn6FQnK+fI7OaNxtv8asn/9bkc6vlVV7ce4lfXP0F/mb3eZZLS+wO9rg0e5Fndp5llI0YpAOWzy6z0dtgtbzCjFu/p9fkgzANoI85HOVQdG/SsxOT4Dk+iU7oxB32hvt0kx4A691rbPRusDvYoxW12B3sToSDXeXiOz6Pzj1ClEVU/AqJSakHNU5WT0zikHbc4Xp3Y3I+YzQnKifY7G9NZCaXS0vsDHZZKa0Aim7Sp5f0CPJC3ReFacxwzPFO8zLLpSVm8/36zmCXNKdL9NM+mRFZyIfmHuLVw9dEzS5X4av4FVbLKxS9IqNsNEmv9tM+taDGMBviOR6e8hikA6pBlVjHGGso5VudsbSk5/i4yiEzmpJXpJt0qQU1RtmIF/df5rtrv0g36VELql+Y9up0MhxzbPa3UChOVNYAqSSLx5s3eQwKiyEzGk+5jHSENhkz4QyDPDMUOL50uSmP1GR4jnvHoN0f7mOsxXWEl1RwQ1pxmzPV03SSLomOKbgFUIrMZLjKoexXGKQDPMel7Jc/12tzO6YxwzHH7ba0t6tx99MB/aTHq4dv0Ek6/PaZ3+RP1/+cc7Wz/PLaL/Fm4y1+tvcCv3v+73IwOuTS7EW+f+0H/EcXfhfHvXUyrHevU/ZL9NPBJBsVuiGJSWnFLQ6GhywUFwhcnziLsVgSnWCweMabToYp7i366YCqf3fmKcjKEOl4UqleKM6zVFriqeUncR2X76w8xWuHr3Omepr17jX++PI/52+f/N5dJe6fWHwcz5GVw3e8Wxx7zlaFp3R7ZmtscfVlcPeZBtDHHKcrJ6mHH5ydUUpxvbdBJ+kAlsxkNKMmt/cPjLNMTyx+g+u9G1iYtJmOg2zf8UW6Mv95O45OhDGp0FEOzhegnnc3TFeGY47KB/QjjFEPapyvn+Oh2UtkVhr2L9TP47syNEa3dcQ9PHeJZtTievc6sY7R1pDZjMfmv8bbzXeo5aIAqU7xHY8sl8OvBBVaUQtXuZytnbk3f+ynxHQyfMVwe77Ed3zO184xyAaUPNErGvOPrLW823qfxxa+jrWWk5UT+E7A6eoprvc2mC3M0k26zHg1MpsxW5gldKV4lzopgRNgEM8GV7nUgtod3+fLhOlk+IohsxlRFnOjd4N6WMd3PFpxm63+DgvFOWpBjdAL2Rvs4SjJ8BS8kMxmDNMRVzpXJ9qp2mg85bHR2yR0CxTcApv9TUp+CWMNmclITcpyaZl2rsx3vn7uS7EluhumMcNXDAolNGqvwPZgh8SkhG7I2dppBumQzGi6SY9+OmSUjegkHZpRi8xk1IIqi8UFEh3jOx7NWHyfy36JXtIj0hFbgx0CJ5B+Br9CySuJAIGOCdzgSxEofxCmK8NXDJ7jsVRaZIlFTlVP5gU24RGtd6/hOx5X2+t8Y+ExMpPS6/UI3ZDN/hbna+dwlMOVzhXWymtEOmJnsIOjhKYdOiFYy4nK2h01iBN27Uu7IowxnQxfYYxNCx3lTGKCmXCG87WzokrhlzldPUU1qNKKWpMBPsxGOEpRcAs8NHeJKIvITEYn6XKqeuqud/8v+0SAaQV6CuBa9zqJTlBKUQ/qJDohNjHWWpZKS6QmJdEJkY7QRksKNR/ca+XVSc1hlEUT2vf9iOnKMMVkMPuOTzfpUvSKhE4ASpHqBM/xMI5HYAO0EsqG7/ikJiXSMZX88/fzRIDpyjDFbTC5V/PNzjNw7uLoA/fH1ueT4P8HrAfZgpgeSnMAAAAASUVORK5CYII=\" y=\"-10.412039\"/>\n",
       "   </g>\n",
       "  </g>\n",
       " </g>\n",
       " <defs>\n",
       "  <clipPath id=\"pb5e30a6d09\">\n",
       "   <rect height=\"217.44\" width=\"194.98932\" x=\"39.65\" y=\"10.972039\"/>\n",
       "  </clipPath>\n",
       " </defs>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113e08860>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/svg+xml": [
       "<?xml version=\"1.0\" encoding=\"utf-8\" standalone=\"no\"?>\n",
       "<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\n",
       "  \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\n",
       "<!-- Created with matplotlib (http://matplotlib.org/) -->\n",
       "<svg height=\"252pt\" version=\"1.1\" viewBox=\"0 0 245 252\" width=\"245pt\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       " <defs>\n",
       "  <style type=\"text/css\">\n",
       "*{stroke-linecap:butt;stroke-linejoin:round;}\n",
       "  </style>\n",
       " </defs>\n",
       " <g id=\"figure_1\">\n",
       "  <g id=\"patch_1\">\n",
       "   <path d=\"M 0 252.290164 \n",
       "L 245.33932 252.290164 \n",
       "L 245.33932 0 \n",
       "L 0 0 \n",
       "z\n",
       "\" style=\"fill:none;\"/>\n",
       "  </g>\n",
       "  <g id=\"axes_1\">\n",
       "   <g clip-path=\"url(#p30a142c487)\">\n",
       "    <image height=\"218\" id=\"image4deee4433a\" transform=\"scale(1 -1)translate(0 -218)\" width=\"195\" x=\"39.65\" xlink:href=\"data:image/png;base64,\n",
       "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\" y=\"-10.412039\"/>\n",
       "   </g>\n",
       "  </g>\n",
       " </g>\n",
       " <defs>\n",
       "  <clipPath id=\"p30a142c487\">\n",
       "   <rect height=\"217.44\" width=\"194.98932\" x=\"39.65\" y=\"10.972039\"/>\n",
       "  </clipPath>\n",
       " </defs>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113e7a5c0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from PIL import Image\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from wordcloud import WordCloud, STOPWORDS, ImageColorGenerator\n",
    "\n",
    "%matplotlib inline\n",
    "%config InlineBackend.figure_format = 'svg'\n",
    "\n",
    "path = '/Users/andrew/Desktop/data/'\n",
    "im_path = path + 'instacart_color.png'\n",
    "im_out_path = 'instacart_wc.png'\n",
    "\n",
    "# select n_orders and aisles column from goods in R\n",
    "%R goods_tmp <- goods[ ,c(6,3)] \n",
    "\n",
    "# make goods_tmp into pandas data frame in Python\n",
    "goods = rpy2.robjects.pandas2ri.ri2py(rpy2.robjects.r.goods_tmp) \n",
    "\n",
    "# convert data-frame to list of lists for consumption\n",
    "#  by wordcloud.generate_from_frequencies(frequencies=d)\n",
    "d = goods.values.tolist()\n",
    "\n",
    "# set stop words\n",
    "stopwords = set(STOPWORDS)\n",
    "stopwords.add(\"said\")\n",
    "\n",
    "# read the mask image\n",
    "instacart_image = np.array(Image.open(im_path))\n",
    "\n",
    "# create coloring from image\n",
    "image_colors = ImageColorGenerator(instacart_image)\n",
    "\n",
    "# create wordcloud object\n",
    "wc = WordCloud(background_color=\"white\", max_words=134, \n",
    "               mask=instacart_image, stopwords=stopwords)\n",
    "\n",
    "# get word counts directly from list of lists\n",
    "wc.generate_from_frequencies(frequencies=d)\n",
    "\n",
    "\n",
    "\n",
    "# plot figure\n",
    "plt.figure()\n",
    "plt.imshow(wc, interpolation=\"bilinear\")\n",
    "plt.axis(\"off\")\n",
    "\n",
    "# plot figure\n",
    "plt.figure()\n",
    "plt.imshow(wc.recolor(color_func=image_colors), \n",
    "           interpolation=\"bilinear\")\n",
    "plt.axis(\"off\")\n",
    "\n",
    "# store to file\n",
    "wc.to_file(im_out_path)\n",
    "\n",
    "# plot figure\n",
    "plt.figure()\n",
    "plt.imshow(instacart_image, cmap=plt.cm.gray, \n",
    "           interpolation='bilinear')\n",
    "plt.axis(\"off\")\n",
    "\n",
    "# show plot\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [default]",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}