convert_ndjson.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Convert raw ndjson into sketch-rnn format with variable RDP epsilon\n",
    "\n",
    "This notebook will convert `.ndjson` files into the `.npz` files one can use to train [sketch-rnn](https://otoro.net/sketch-rnn). We set the target length to be 200 steps, and vary epsilon parameters to control the granuarity of the RDP algorithm.\n",
    "\n",
    "To download an `.ndjson` file, try:\n",
    "\n",
    "```\n",
    "wget https://storage.googleapis.com/quickdraw_dataset/full/raw/jail.ndjson\n",
    "```\n",
    "\n",
    "To download the `jail` class from [QuickDraw](https://github.com/googlecreativelab/quickdraw-dataset)\n",
    "\n",
    "And we assume it is saved in a subdir called `/ndjson` for this notebook.\n",
    "\n",
    "After the `npz` dataset is generated, we can train sketch-rnn with a command like this:\n",
    "\n",
    "```\n",
    "python sketch_rnn_train.py --data_dir=dir_where_npz_file_is --gpu=0 --log_root=log_jail --hparams=data_set=['jail.200.npz'],num_steps=1000000,conditional=0,dec_rnn_size=1024,recurrent_dropout_prob=0.8\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# notebook to convert raw quickdraw data using incremental RDP to fit everything within, say, 300 datapoints but not sacrifice quality.\n",
    "\n",
    "import numpy as np\n",
    "import util as util\n",
    "import os\n",
    "import json\n",
    "import time\n",
    "import random\n",
    "from rdp import rdp\n",
    "\n",
    "np.set_printoptions(precision=8, edgeitems=6, linewidth=200, suppress=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def raw_to_lines(raw, epsilon=0.25):\n",
    "  result = []\n",
    "  N = len(raw)\n",
    "  for i in range(N):\n",
    "    line = []\n",
    "    rawline = raw[i]\n",
    "    M = len(rawline[0])\n",
    "    if M <= 2:\n",
    "      continue\n",
    "    for j in range(M):\n",
    "      line.append([rawline[0][j], rawline[1][j]])\n",
    "    line = rdp(line, epsilon=epsilon) \n",
    "    result.append(line)\n",
    "  return result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# config\n",
    "NAME = \"jail\"\n",
    "TARGET = 200"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "raw_file = open('ndjson/'+NAME+'.ndjson', 'r') \n",
    "raw_lines = raw_file.readlines() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "120131\n"
     ]
    }
   ],
   "source": [
    "num_drawings = len(raw_lines)\n",
    "print(num_drawings)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "i 0\n",
      "i 1000\n",
      "i 2000\n",
      "i 3000\n",
      "i 4000\n",
      "i 5000\n",
      "i 6000\n",
      "i 7000\n",
      "i 8000\n",
      "i 9000\n",
      "i 10000\n",
      "i 11000\n",
      "i 12000\n",
      "i 13000\n",
      "i 14000\n",
      "i 15000\n",
      "i 16000\n",
      "i 17000\n",
      "i 18000\n",
      "i 19000\n",
      "i 20000\n",
      "i 21000\n",
      "i 22000\n",
      "i 23000\n",
      "i 24000\n",
      "i 25000\n",
      "i 26000\n",
      "i 27000\n",
      "i 28000\n",
      "i 29000\n",
      "i 30000\n",
      "i 31000\n",
      "i 32000\n",
      "i 33000\n",
      "i 34000\n",
      "i 35000\n",
      "i 36000\n",
      "i 37000\n",
      "i 38000\n",
      "i 39000\n",
      "i 40000\n",
      "i 41000\n",
      "i 42000\n",
      "i 43000\n",
      "i 44000\n",
      "i 45000\n",
      "i 46000\n",
      "i 47000\n",
      "i 48000\n",
      "i 49000\n",
      "i 50000\n",
      "i 51000\n",
      "i 52000\n",
      "i 53000\n",
      "i 54000\n",
      "i 55000\n",
      "i 56000\n",
      "i 57000\n",
      "i 58000\n",
      "i 59000\n",
      "i 60000\n",
      "i 61000\n",
      "i 62000\n",
      "i 63000\n",
      "i 64000\n",
      "i 65000\n",
      "i 66000\n",
      "i 67000\n",
      "i 68000\n",
      "i 69000\n",
      "i 70000\n",
      "i 71000\n",
      "i 72000\n",
      "i 73000\n",
      "i 74000\n",
      "i 75000\n",
      "i 76000\n",
      "i 77000\n",
      "i 78000\n",
      "i 79000\n",
      "i 80000\n",
      "i 81000\n",
      "i 82000\n",
      "i 83000\n",
      "i 84000\n",
      "i 85000\n",
      "i 86000\n",
      "i 87000\n",
      "i 88000\n",
      "i 89000\n",
      "i 90000\n",
      "i 91000\n",
      "i 92000\n",
      "i 93000\n",
      "i 94000\n",
      "i 95000\n",
      "i 96000\n",
      "i 97000\n",
      "i 98000\n",
      "i 99000\n",
      "i 100000\n",
      "i 101000\n",
      "i 102000\n",
      "i 103000\n",
      "i 104000\n",
      "i 105000\n",
      "i 106000\n",
      "i 107000\n",
      "i 108000\n",
      "i 109000\n",
      "i 110000\n",
      "i 111000\n",
      "i 112000\n",
      "i 113000\n",
      "i 114000\n",
      "i 115000\n",
      "i 116000\n",
      "i 117000\n",
      "i 118000\n",
      "i 119000\n",
      "i 120000\n"
     ]
    }
   ],
   "source": [
    "all_strokes = []\n",
    "for i in range(num_drawings):\n",
    "  raw_drawing = json.loads(raw_lines[i])['drawing']\n",
    "  lines = raw_to_lines(raw_drawing)\n",
    "  strokes = util.lines_to_strokes(lines)\n",
    "  if i % 1000 == 0:\n",
    "    print(\"i\", i)\n",
    "  if len(strokes) < 20:\n",
    "    continue\n",
    "  strokes[0, 0] = 0\n",
    "  strokes[0, 1] = 0\n",
    "  all_strokes.append(strokes)\n",
    "random.shuffle(all_strokes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "119693"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(all_strokes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.savez_compressed(NAME+\".npz\", rawdata=all_strokes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "20\n",
      "2301\n"
     ]
    }
   ],
   "source": [
    "print(util.get_min_len(all_strokes))\n",
    "print(util.get_max_len(all_strokes))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def further_simplify(temp_strokes, epsilon=0.5):\n",
    "  temp_lines = util.strokes_to_lines(temp_strokes)\n",
    "  new_lines = []\n",
    "  for i in range(len(temp_lines)):\n",
    "    temp_line = temp_lines[i]\n",
    "    new_line = rdp(temp_line, epsilon)\n",
    "    if len(new_line) > 2:\n",
    "      new_lines.append(new_line)\n",
    "  return util.lines_to_strokes(new_lines)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "822\n"
     ]
    }
   ],
   "source": [
    "for i in range(10000):\n",
    "  the_strokes = random.choice(all_strokes)\n",
    "  if len(the_strokes) > 800:\n",
    "    break\n",
    "print(len(the_strokes))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<svg baseProfile=\"full\" height=\"469.6000000000002\" version=\"1.1\" width=\"721.6000000000005\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:ev=\"http://www.w3.org/2001/xml-events\" xmlns:xlink=\"http://www.w3.org/1999/xlink\"><defs/><rect fill=\"white\" height=\"469.6000000000002\" width=\"721.6000000000005\" x=\"0\" y=\"0\"/><path d=\"M261.4000000000002,256.19999999999993 m0.0,0.0 l-1.6,-2.8 1.2,-2.8 l4.0,-11.2 6.4,-12.0 l5.2,-7.6 7.2,-6.8 l9.2,-5.6 10.0,-4.0 l16.4,-2.8 18.0,-0.8 l22.8,2.8 23.2,8.4 l23.2,12.8 21.2,15.2 l19.2,18.4 14.8,18.4 l9.6,17.2 5.2,13.6 l2.4,10.0 1.2,8.4 l0.8,7.2 1.6,6.8 l1.6,4.0 2.8,2.8 l4.4,2.4 7.2,2.0 l11.6,1.2 19.6,-1.6 l20.8,-4.0 20.0,-6.4 l20.0,-7.6 19.6,-8.8 l19.6,-11.6 18.8,-14.0 l19.6,-19.6 15.2,-20.8 l10.0,-21.6 3.2,-22.4 l0.0,-34.8 -3.6,-13.2 l-5.2,-8.8 -5.6,-4.8 l-9.6,-4.0 -13.6,-1.2 l-19.6,2.0 -25.2,5.6 l-35.2,12.0 -38.4,16.8 l-40.0,19.6 -42.8,22.8 l-44.0,24.8 -42.8,24.8 l-42.8,26.8 -97.2,62.8 l-42.8,26.8 -39.6,22.4 l-36.8,17.6 -34.8,12.8 l-29.6,0.8 -22.4,0.0 l0.0,-144.4 12.8,-86.0 l9.6,-27.6 12.8,-26.0 l14.4,-20.4 14.4,-16.0 l18.8,-13.2 21.6,-9.6 l25.6,-8.0 36.0,-4.4 l34.4,-1.6 34.4,0.4 l33.6,3.6 32.8,6.0 l59.2,22.4 50.0,35.6 l18.4,24.8 17.2,30.4 l15.6,36.8 14.8,41.2 l15.2,45.6 17.2,52.8 l16.4,48.0 32.4,9.6 l98.8,0.0 12.4,-3.2 l2.8,-6.4 0.0,-178.4 l-1.2,-31.2 -18.0,-60.8 l-19.2,-33.2 -24.8,-34.4 l-29.6,-30.4 -98.4,-41.6 l-70.4,12.8 -20.8,15.2 l-25.2,24.4 -79.6,86.4 l-29.6,33.6 -30.8,30.8 l-141.6,104.0 -39.6,8.8 l-14.4,-0.4 -15.6,-3.6 l-12.8,-6.4 0.0,-170.0 l55.6,-135.6 54.0,0.0 l33.2,1.2 34.0,6.4 l40.0,14.8 36.0,19.6 l61.6,49.2 24.4,30.4 l22.4,35.2 86.0,168.4 l26.8,58.4 8.8,20.4 l14.0,15.6 20.4,0.0 l9.2,-4.8 12.4,-7.6 l12.8,-11.2 56.4,-80.4 l26.8,-59.6 9.2,-25.6 l6.4,-20.0 3.6,-14.4 l2.4,-10.4 0.4,-2.4 l-32.4,2.4 -23.6,8.4 l-26.8,10.4 -95.2,55.6 l-35.6,25.2 -34.4,26.0 l-96.8,74.8 -54.0,33.6 l-111.6,0.0 -16.0,-1.6 l-12.4,-4.0 -34.8,-46.0 l-4.8,-22.0 -2.4,-26.8 l1.2,-61.6 4.4,-31.2 l7.2,-32.0 8.8,-27.6 l79.6,-70.8 50.4,-10.8 l24.8,-2.8 24.4,-0.8 l62.8,1.2 15.2,3.2 l9.2,2.8 8.0,8.0 l3.2,8.8 6.4,47.6 l3.6,33.6 4.8,39.2 l4.4,41.6 39.2,152.0 l30.8,0.0 59.2,-6.8 l22.0,-11.6 39.2,-40.8 l36.0,-72.0 11.2,-42.0 l6.4,-39.2 0.8,-35.6 l-10.8,-61.2 -11.2,-24.8 l-18.4,-23.2 -90.8,-53.6 l-48.4,-4.4 -90.0,29.6 l-59.6,40.8 -58.8,55.2 l-33.2,36.8 -30.0,34.8 l-52.0,57.2 -42.8,38.8 l-18.8,11.6 -21.2,8.0 l-34.4,4.8 -18.8,-9.6 l0.0,-187.2 2.0,-20.8 l27.6,-42.0 16.4,-9.6 l20.4,-8.0 59.2,-12.4 l32.4,-3.2 76.8,-4.8 l141.6,-3.6 41.2,-0.8 l62.0,-2.0 33.6,0.4 l5.6,1.2 5.6,6.4 l1.2,8.4 -0.4,13.6 l-11.6,80.4 -7.2,78.0 l4.0,83.6 8.0,32.8 l42.0,58.0 16.8,6.8 l20.4,0.8 20.8,-3.6 l22.4,-9.2 30.8,-29.2 l0.0,-182.8 -1.2,-19.2 l-16.0,-33.2 -12.8,-12.0 l-65.6,-24.8 -30.0,-2.4 l-32.4,1.2 -67.2,19.2 l-112.4,69.2 -35.6,30.0 l-72.8,65.2 -72.8,62.4 l-60.8,42.8 -26.0,12.0 l-40.4,8.0 -25.6,-9.6 l0.0,-174.8 0.4,-38.4 l12.4,-25.6 33.6,-27.6 l50.0,-24.0 29.2,-10.4 l95.2,-28.4 29.2,-5.2 l46.0,-2.8 34.0,7.2 l27.2,26.8 20.8,54.8 l8.0,36.0 12.0,82.0 l17.6,83.2 22.0,45.6 l20.4,16.0 25.6,-1.2 l18.0,-7.2 18.4,-9.6 l40.0,-25.6 20.4,-18.0 l45.2,-50.8 46.0,-71.6 l0.0,-26.4 -3.6,-1.6 l-26.0,12.0 -62.4,31.6 l-82.0,42.4 -86.4,48.4 l-78.4,50.0 -58.4,38.8 l-23.2,13.6 -79.6,8.0 l-30.8,-6.8 -23.6,-15.6 l-14.4,-28.0 -2.0,-49.2 l26.4,-92.0 39.6,-58.4 l49.2,-37.6 54.8,-21.2 l58.8,-8.4 62.4,1.2 l46.0,10.8 30.8,18.4 l20.0,30.8 8.8,44.0 l1.2,48.0 -6.8,52.0 l-1.2,51.2 5.6,21.2 l4.8,5.2 10.0,-1.6 l17.6,-12.0 20.8,-22.4 l36.8,-65.6 16.8,-46.0 l11.2,-40.8 3.2,-35.2 l-4.0,-21.6 -7.6,-13.2 l-14.8,-6.4 -21.2,4.8 l-17.2,8.0 -75.6,44.4 l-26.4,18.4 -62.0,42.8 l-110.4,73.2 -70.8,46.4 l-64.8,43.2 -58.4,18.0 l-49.6,0.0 -6.8,-4.8 l-4.8,-22.4 -1.6,-36.8 l5.6,-56.4 12.0,-80.4 l7.6,-44.0 33.2,-89.2 l36.8,-26.0 48.8,-6.8 l56.8,10.8 88.8,52.4 l28.0,25.2 46.8,52.0 l38.0,52.4 38.4,55.2 l40.0,56.0 7.2,8.0 l5.6,5.6 4.8,0.8 l21.2,-15.2 10.4,-9.2 l24.8,-24.0 40.8,-54.0 l18.0,-44.0 6.4,-40.8 l-2.0,-25.6 -1.6,-6.8 l-2.8,-0.4 -62.8,40.0 l-61.2,34.0 -121.2,64.0 l-121.6,62.0 -33.6,15.6 l-83.2,26.4 -18.0,2.0 l-24.8,-2.4 -6.0,-6.0 l-1.6,-15.6 7.6,-30.8 l35.2,-72.4 22.8,-30.0 l79.2,-59.6 65.6,-23.6 l107.2,-18.0 51.2,-2.4 l31.6,0.8 11.2,2.4 l1.6,2.4 0.4,8.8 l-2.4,23.2 1.2,51.2 l2.8,15.2 8.8,26.8 l3.2,3.6 4.4,0.4 l4.4,-2.8 45.6,-34.0 l23.6,-27.6 28.0,-47.2 l8.8,-30.0 1.2,-11.6 l-1.6,-20.8 -19.6,-28.0 l-22.4,-11.6 -48.4,-1.6 l-57.2,14.8 -67.2,28.8 l-75.6,41.2 -128.8,81.2 l-86.8,55.2 -136.8,78.0 l0.0,39.6 2.4,-5.2 l9.2,-8.4 69.6,-47.2 l74.4,-37.2 70.4,-30.0 l71.2,-24.8 95.6,-23.2 l24.4,2.8 12.0,17.2 l14.8,98.0 18.8,84.8 l149.2,0.0 59.6,-18.0 l0.0,-269.2 -4.4,-22.4 l-20.0,3.6 -58.4,28.4 l-162.0,81.2 -126.8,53.2 l-133.6,42.8 -148.4,22.8 l-18.0,-14.0 0.0,-106.8 l2.4,-48.8 12.8,-77.6 l24.4,-78.0 43.2,-16.8 l81.6,10.4 60.4,36.0 l76.0,82.4 64.4,111.2 l39.6,71.2 36.4,52.0 l47.2,32.4 21.6,0.4 l45.6,-33.2 53.6,-64.8 l50.0,-79.2 12.4,-110.4 l0.0,-32.8 -2.8,1.2 l-36.0,28.4 -105.6,78.4 l-96.8,61.2 -151.2,83.2 l-91.2,35.2 -90.4,15.2 l-17.2,-4.4 2.4,-22.0 l33.2,-62.4 56.4,-87.2 l107.2,-131.2 71.6,-46.4 l150.4,17.2 64.0,74.4 l31.2,74.8 21.6,89.2 l13.6,114.4 3.6,24.8 l11.6,-7.2 24.4,-63.2 l0.0,-69.6 -28.4,-175.2 l-39.2,-45.6 -90.4,-26.4 l-68.8,8.4 -72.4,30.4 l-176.4,134.0 -114.8,94.8 l-48.8,26.0 -17.6,4.8 l1.6,-10.0 24.4,-54.0 l62.0,-123.6 99.6,-142.8 l160.8,12.4 55.6,40.4 l74.0,147.6 38.0,156.0 l19.2,62.8 19.2,0.0 l5.6,-14.0 8.4,-53.2 l-28.8,-198.8 -39.2,-84.0 l-48.0,-42.8 -26.0,3.2 l-96.4,81.6 -64.0,71.6 l-103.6,110.8 -98.8,78.4 l-36.4,14.0 -21.2,-8.8 l0.0,-19.6 33.6,-86.0 l46.4,-50.0 112.8,-42.0 l124.8,-10.0 97.2,26.8 l62.8,62.4 21.6,84.0 l6.8,76.4 3.2,0.0 l5.2,-24.0 3.6,-75.6 l-40.4,-149.6 -93.2,-125.6 l-62.8,-26.4 -92.8,95.2 l-80.0,123.6 -82.8,112.0 l-52.0,46.4 -3.6,2.0 l3.6,-8.0 80.8,-109.6 l156.8,-189.2 119.2,-78.8 l127.2,16.0 21.6,35.6 l10.4,128.0 -5.6,87.6 l-16.8,129.2 -2.0,11.2 l-14.0,-36.0 -34.8,-48.8 l-51.2,-32.4 -33.6,14.8 l-47.6,54.0 -48.8,48.4 l-48.0,0.0 -2.0,-1.6 l27.6,-72.4 38.0,-74.0 l76.8,-68.0 15.2,-2.0 l10.8,36.8 0.4,53.6 l-24.4,104.4 -16.8,23.2 l-7.2,-0.4 51.6,-33.6 l70.0,-38.4 43.2,-13.2 l2.4,0.4 -35.2,19.6 l-185.2,62.4 -249.6,3.2 l-64.0,0.0 18.4,-10.4 l123.6,-52.4 200.0,-60.4 l197.6,-19.6 132.0,29.2 l0.0,50.4 -34.0,24.4 l-142.4,38.8 -495.2,0.0 l174.8,0.0 256.0,-2.4 l174.4,-17.2 66.4,-5.2 l-34.8,4.8 -199.6,7.2 l-322.8,-9.2 -114.4,-1.6 l44.4,-5.2 88.4,-1.2 l196.0,6.4 136.4,6.4 l51.2,8.0 -64.8,9.2 l-152.8,0.0 -251.2,-11.6 l-47.6,-13.2 20.0,-9.2 l134.0,-26.8 222.8,-15.2 l100.8,-0.8 10.8,0.4 l-38.4,11.6 -155.6,20.0 l-69.6,5.2 30.8,-3.2 l117.6,-1.6 51.2,2.8 l-6.4,10.8 -48.4,20.4 l-146.0,10.4 59.2,-2.8 l149.6,-25.6 79.6,-14.4 l15.6,-1.2 -11.6,2.0 l-82.4,9.2 -88.0,-2.0 l-16.4,-2.0 10.8,-1.2 l75.6,-0.8 74.0,0.0 l77.6,-0.8 -78.8,18.8 l-156.0,16.8 -180.8,2.8 l-2.0,-2.8 88.8,-43.6 l96.4,-33.2 228.4,-49.6 l108.4,4.4 0.0,8.8 l-58.8,31.2 -196.8,32.0 l-204.8,-5.6 -168.0,-20.0 l-43.2,-19.6 0.4,-2.0 l36.0,-4.0 175.2,-11.2 l108.4,-17.2 235.6,-86.8 l59.6,-38.8 18.4,-14.4 l-15.2,5.6 -152.4,41.2 l-192.8,30.4 -273.2,4.4 l0.0,-12.8 104.8,-13.2 l260.8,1.2 208.0,12.8 l98.0,9.6 -32.8,17.2 l-153.6,48.0 -218.4,43.6 l-266.8,26.0 22.0,-3.6 l149.6,-15.6 201.2,-12.0 l191.2,-18.4 107.6,-8.8 l-66.0,11.2 -236.4,37.6 l-118.0,14.4 -96.4,12.4 l74.4,-4.8 158.4,-21.2 l138.4,-33.6 43.6,-14.8 l-13.2,1.2 -98.8,23.2 l-156.8,19.2 -20.0,0.8 l10.8,-3.6 219.2,-37.2 l77.2,-12.4 49.6,-9.6 l-70.4,2.0 -108.4,2.0 l-274.0,-9.2 -102.0,-17.6 l116.8,-16.4 220.0,-7.2 l121.6,-0.8 10.8,2.0 l-73.2,20.8 -185.2,11.2 l-86.8,-8.8 1.2,-4.8 l182.8,-47.2 141.6,-33.2 l43.2,-10.4 2.0,-1.6 l-92.8,14.0 -248.4,33.2 l-85.2,8.4 66.8,-7.2 l156.8,-42.4 146.8,-76.4 l28.4,-40.0 -20.8,-16.8 l-159.2,69.2 -152.4,103.2 l-75.2,67.2 62.4,0.4 l210.0,-68.8 235.2,-132.8 l21.6,-76.4 -197.6,90.4 l-214.4,122.8 -187.2,174.8 l9.6,31.6 319.6,-41.2 l136.4,-79.6 78.4,-119.2 l-25.2,-134.0 -140.4,-45.6 l-237.6,82.0 -138.4,195.6 l13.6,142.0 298.0,-16.8 l158.4,-90.8 48.8,-121.6 l-44.4,-138.0 -138.8,-52.4 l-258.8,160.8 -79.6,199.6 l48.8,59.2 260.4,-57.6 l77.6,-112.8 6.4,-242.4 l-162.8,32.4 -168.0,156.0 l-58.0,129.2 64.0,95.2 l235.6,-27.6 159.2,-147.2 l30.4,-172.8 -56.0,-58.4 l-231.2,92.0 -132.0,171.2 l1.2,100.4 167.6,42.4 l204.4,-50.8 65.2,-109.2 l-7.6,-72.0 -94.0,-60.4 l-111.6,-2.4 -189.6,104.4 l-32.0,88.0 78.4,102.4 l271.2,-13.2 148.4,-93.6 l23.6,-138.8 -58.8,-114.8 l-188.0,-59.2 -388.0,118.8 l-36.8,113.6 48.0,80.0 l159.2,46.8 232.8,-18.8 l79.6,-70.4 -19.6,-163.6 l-122.8,-92.4 -346.8,61.2 l-30.4,86.8 0.0,39.6 l23.2,50.4 156.4,81.6 l96.0,-10.0 89.2,-78.8 l28.0,-74.8 -6.4,-61.6 l-115.2,-32.4 -199.6,20.0 l-71.6,95.2 7.2,120.8 l114.0,107.6 200.8,0.0 l187.2,-58.4 76.0,-73.6 l49.6,-97.2 -12.4,-50.8 l-122.4,-10.8 -151.2,41.2 l-137.2,158.4 18.8,91.2 l254.0,-2.4 106.4,-31.6 l80.8,-138.8 -18.4,-100.0 l-137.6,-55.6 -128.4,16.4 l-156.0,112.0 -16.4,92.0 l96.0,108.0 213.6,-31.2 l78.8,-141.6 -15.2,-169.6 l-70.4,-57.6 -409.2,39.6 l-28.0,82.0 62.4,83.6 l165.6,42.8 138.8,-21.2 l96.8,-87.6 34.0,-82.4 l-47.6,-60.0 -266.0,38.0 l-87.2,120.4 37.2,94.0 l108.4,62.8 104.0,21.6 l128.0,-32.0 62.4,-66.0 l6.8,-60.4 -69.2,-52.0 l-199.2,-15.6 -103.2,52.0 l-5.2,74.0 62.4,67.2 l130.8,21.2 73.2,-29.6 l23.6,-52.0 -24.4,-62.0 l-144.8,-42.4 -138.0,35.2 l-39.2,80.0 33.6,92.0 l132.0,32.0 105.6,-61.2 l30.8,-168.8 -133.2,-111.6 l-225.2,-35.6 -88.8,197.2 l227.2,132.4 196.8,-5.2 l96.8,-100.0 -39.6,-151.6 l-241.6,-90.4 -262.8,230.4 l187.6,175.2 352.4,-24.8 l-81.6,-148.4 -390.8,-96.4 l-84.4,121.6 164.0,161.6 l416.4,-49.2 33.2,-158.8 l-131.6,-204.0 -487.6,98.0 l35.2,198.8 294.8,115.6 l341.6,-155.2 -10.8,-126.4 l-295.2,-138.0 -365.6,188.0 l88.4,215.6 187.6,16.0 l294.4,-243.6 -90.0,-176.0 l-421.6,111.2 -12.8,151.6 l257.6,130.0 117.2,-5.6 l54.4,-68.8 -80.0,-46.4 l-215.6,86.8 -76.4,60.8 l202.4,-16.4 \" fill=\"none\" stroke=\"black\" stroke-width=\"1\"/></svg>"
      ],
      "text/plain": [
       "<IPython.core.display.SVG object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "util.draw_strokes(the_strokes, factor=2.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "682\n"
     ]
    },
    {
     "data": {
      "image/svg+xml": [
       "<svg baseProfile=\"full\" height=\"469.60000000000014\" version=\"1.1\" width=\"721.6000000000006\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:ev=\"http://www.w3.org/2001/xml-events\" xmlns:xlink=\"http://www.w3.org/1999/xlink\"><defs/><rect fill=\"white\" height=\"469.60000000000014\" width=\"721.6000000000006\" x=\"0\" y=\"0\"/><path d=\"M261.4000000000001,256.2000000000001 m0.0,0.0 l-1.6,-2.8 5.2,-14.0 l11.6,-19.6 26.4,-16.4 l34.4,-3.6 22.8,2.8 l23.2,8.4 44.4,28.0 l19.2,18.4 24.4,35.6 l12.8,50.0 14.4,7.2 l31.2,-0.4 40.8,-10.4 l39.6,-16.4 38.4,-25.6 l19.6,-19.6 15.2,-20.8 l10.0,-21.6 3.2,-57.2 l-8.8,-22.0 -15.2,-8.8 l-33.2,0.8 -25.2,5.6 l-73.6,28.8 -82.8,42.4 l-86.8,49.6 -222.4,138.8 l-71.6,30.4 -52.0,0.8 l0.0,-144.4 12.8,-86.0 l22.4,-53.6 28.8,-36.4 l18.8,-13.2 21.6,-9.6 l25.6,-8.0 70.4,-6.0 l34.4,0.4 66.4,9.6 l59.2,22.4 50.0,35.6 l18.4,24.8 17.2,30.4 l15.6,36.8 63.6,187.6 l32.4,9.6 98.8,0.0 l12.4,-3.2 2.8,-6.4 l-1.2,-209.6 -18.0,-60.8 l-19.2,-33.2 -24.8,-34.4 l-29.6,-30.4 -98.4,-41.6 l-70.4,12.8 -20.8,15.2 l-165.2,175.2 -141.6,104.0 l-39.6,8.8 -30.0,-4.0 l-12.8,-6.4 0.0,-170.0 l55.6,-135.6 87.2,1.2 l34.0,6.4 40.0,14.8 l36.0,19.6 61.6,49.2 l24.4,30.4 22.4,35.2 l86.0,168.4 35.6,78.8 l14.0,15.6 20.4,0.0 l9.2,-4.8 25.2,-18.8 l56.4,-80.4 26.8,-59.6 l22.0,-72.8 -32.4,2.4 l-50.4,18.8 -95.2,55.6 l-166.8,126.0 -54.0,33.6 l-111.6,0.0 -28.4,-5.6 l-34.8,-46.0 -7.2,-48.8 l1.2,-61.6 4.4,-31.2 l16.0,-59.6 79.6,-70.8 l75.2,-13.6 87.2,0.4 l24.4,6.0 8.0,8.0 l3.2,8.8 19.2,162.0 l39.2,152.0 30.8,0.0 l59.2,-6.8 22.0,-11.6 l39.2,-40.8 36.0,-72.0 l11.2,-42.0 6.4,-39.2 l0.8,-35.6 -10.8,-61.2 l-11.2,-24.8 -18.4,-23.2 l-90.8,-53.6 -48.4,-4.4 l-90.0,29.6 -59.6,40.8 l-58.8,55.2 -115.2,128.8 l-42.8,38.8 -18.8,11.6 l-21.2,8.0 -34.4,4.8 l-18.8,-9.6 0.0,-187.2 l2.0,-20.8 27.6,-42.0 l36.8,-17.6 59.2,-12.4 l109.2,-8.0 244.8,-6.4 l39.2,1.6 5.6,6.4 l0.8,22.0 -11.6,80.4 l-7.2,78.0 4.0,83.6 l8.0,32.8 42.0,58.0 l16.8,6.8 20.4,0.8 l20.8,-3.6 22.4,-9.2 l30.8,-29.2 -1.2,-202.0 l-16.0,-33.2 -12.8,-12.0 l-65.6,-24.8 -62.4,-1.2 l-67.2,19.2 -112.4,69.2 l-181.2,157.6 -60.8,42.8 l-26.0,12.0 -40.4,8.0 l-25.6,-9.6 0.0,-174.8 l0.4,-38.4 12.4,-25.6 l33.6,-27.6 50.0,-24.0 l124.4,-38.8 29.2,-5.2 l46.0,-2.8 34.0,7.2 l27.2,26.8 20.8,54.8 l20.0,118.0 17.6,83.2 l22.0,45.6 20.4,16.0 l25.6,-1.2 18.0,-7.2 l58.4,-35.2 20.4,-18.0 l45.2,-50.8 46.0,-71.6 l0.0,-26.4 -3.6,-1.6 l-26.0,12.0 -144.4,74.0 l-86.4,48.4 -160.0,102.4 l-79.6,8.0 -30.8,-6.8 l-23.6,-15.6 -14.4,-28.0 l-2.0,-49.2 26.4,-92.0 l39.6,-58.4 49.2,-37.6 l54.8,-21.2 58.8,-8.4 l62.4,1.2 46.0,10.8 l30.8,18.4 20.0,30.8 l8.8,44.0 1.2,48.0 l-6.8,52.0 -1.2,51.2 l5.6,21.2 4.8,5.2 l10.0,-1.6 17.6,-12.0 l20.8,-22.4 36.8,-65.6 l28.0,-86.8 3.2,-35.2 l-4.0,-21.6 -7.6,-13.2 l-14.8,-6.4 -21.2,4.8 l-17.2,8.0 -75.6,44.4 l-334.4,224.0 -58.4,18.0 l-49.6,0.0 -6.8,-4.8 l-4.8,-22.4 -1.6,-36.8 l5.6,-56.4 19.6,-124.4 l33.2,-89.2 36.8,-26.0 l48.8,-6.8 56.8,10.8 l88.8,52.4 28.0,25.2 l46.8,52.0 116.4,163.6 l12.8,13.6 4.8,0.8 l21.2,-15.2 35.2,-33.2 l40.8,-54.0 18.0,-44.0 l6.4,-40.8 -2.0,-25.6 l-1.6,-6.8 -2.8,-0.4 l-62.8,40.0 -61.2,34.0 l-276.4,141.6 -83.2,26.4 l-18.0,2.0 -24.8,-2.4 l-6.0,-6.0 -1.6,-15.6 l7.6,-30.8 35.2,-72.4 l22.8,-30.0 79.2,-59.6 l65.6,-23.6 107.2,-18.0 l51.2,-2.4 42.8,3.2 l0.8,85.6 11.6,42.0 l7.6,4.0 50.0,-36.8 l23.6,-27.6 28.0,-47.2 l8.8,-30.0 -0.4,-32.4 l-19.6,-28.0 -22.4,-11.6 l-48.4,-1.6 -57.2,14.8 l-67.2,28.8 -75.6,41.2 l-215.6,136.4 -136.8,78.0 l0.0,39.6 11.6,-13.6 l69.6,-47.2 74.4,-37.2 l70.4,-30.0 71.2,-24.8 l95.6,-23.2 24.4,2.8 l12.0,17.2 14.8,98.0 l18.8,84.8 149.2,0.0 l59.6,-18.0 0.0,-269.2 l-4.4,-22.4 -20.0,3.6 l-220.4,109.6 -126.8,53.2 l-133.6,42.8 -148.4,22.8 l-18.0,-14.0 2.4,-155.6 l12.8,-77.6 24.4,-78.0 l43.2,-16.8 81.6,10.4 l60.4,36.0 76.0,82.4 l104.0,182.4 36.4,52.0 l47.2,32.4 21.6,0.4 l45.6,-33.2 53.6,-64.8 l50.0,-79.2 12.4,-110.4 l0.0,-32.8 -144.4,108.0 l-96.8,61.2 -151.2,83.2 l-91.2,35.2 -90.4,15.2 l-17.2,-4.4 2.4,-22.0 l33.2,-62.4 56.4,-87.2 l107.2,-131.2 71.6,-46.4 l150.4,17.2 64.0,74.4 l31.2,74.8 21.6,89.2 l17.2,139.2 11.6,-7.2 l24.4,-63.2 0.0,-69.6 l-28.4,-175.2 -39.2,-45.6 l-90.4,-26.4 -68.8,8.4 l-72.4,30.4 -176.4,134.0 l-114.8,94.8 -48.8,26.0 l-17.6,4.8 1.6,-10.0 l24.4,-54.0 62.0,-123.6 l99.6,-142.8 160.8,12.4 l55.6,40.4 74.0,147.6 l38.0,156.0 19.2,62.8 l19.2,0.0 5.6,-14.0 l8.4,-53.2 -28.8,-198.8 l-39.2,-84.0 -48.0,-42.8 l-26.0,3.2 -96.4,81.6 l-167.6,182.4 -98.8,78.4 l-36.4,14.0 -21.2,-8.8 l0.0,-19.6 33.6,-86.0 l46.4,-50.0 112.8,-42.0 l124.8,-10.0 97.2,26.8 l62.8,62.4 21.6,84.0 l6.8,76.4 3.2,0.0 l5.2,-24.0 3.6,-75.6 l-40.4,-149.6 -93.2,-125.6 l-62.8,-26.4 -92.8,95.2 l-80.0,123.6 -82.8,112.0 l-55.6,48.4 84.4,-117.6 l156.8,-189.2 119.2,-78.8 l127.2,16.0 21.6,35.6 l10.4,128.0 -5.6,87.6 l-18.8,140.4 -14.0,-36.0 l-34.8,-48.8 -51.2,-32.4 l-33.6,14.8 -47.6,54.0 l-48.8,48.4 -50.0,-1.6 l27.6,-72.4 38.0,-74.0 l76.8,-68.0 15.2,-2.0 l10.8,36.8 0.4,53.6 l-24.4,104.4 -16.8,23.2 l-7.2,-0.4 51.6,-33.6 l70.0,-38.4 45.6,-12.8 l-35.2,19.6 -185.2,62.4 l-313.6,3.2 18.4,-10.4 l123.6,-52.4 200.0,-60.4 l197.6,-19.6 132.0,29.2 l0.0,50.4 -34.0,24.4 l-142.4,38.8 -495.2,0.0 l430.8,-2.4 240.8,-22.4 l-34.8,4.8 -199.6,7.2 l-437.2,-10.8 44.4,-5.2 l88.4,-1.2 332.4,12.8 l51.2,8.0 -64.8,9.2 l-152.8,0.0 -251.2,-11.6 l-47.6,-13.2 20.0,-9.2 l134.0,-26.8 222.8,-15.2 l100.8,-0.8 10.8,0.4 l-38.4,11.6 -155.6,20.0 l-69.6,5.2 30.8,-3.2 l117.6,-1.6 51.2,2.8 l-6.4,10.8 -48.4,20.4 l-146.0,10.4 59.2,-2.8 l244.8,-41.2 -94.0,11.2 l-88.0,-2.0 -16.4,-2.0 l10.8,-1.2 227.2,-1.6 l-78.8,18.8 -156.0,16.8 l-180.8,2.8 -2.0,-2.8 l88.8,-43.6 96.4,-33.2 l228.4,-49.6 108.4,4.4 l0.0,8.8 -58.8,31.2 l-196.8,32.0 -204.8,-5.6 l-168.0,-20.0 -42.8,-21.6 l211.2,-15.2 108.4,-17.2 l235.6,-86.8 78.0,-53.2 l-167.6,46.8 -192.8,30.4 l-273.2,4.4 0.0,-12.8 l104.8,-13.2 260.8,1.2 l208.0,12.8 98.0,9.6 l-32.8,17.2 -153.6,48.0 l-218.4,43.6 -266.8,26.0 l171.6,-19.2 201.2,-12.0 l298.8,-27.2 -302.4,48.8 l-214.4,26.8 74.4,-4.8 l158.4,-21.2 138.4,-33.6 l43.6,-14.8 -112.0,24.4 l-156.8,19.2 -20.0,0.8 l10.8,-3.6 346.0,-59.2 l-178.8,4.0 -274.0,-9.2 l-102.0,-17.6 116.8,-16.4 l220.0,-7.2 132.4,1.2 l-73.2,20.8 -185.2,11.2 l-86.8,-8.8 1.2,-4.8 l369.6,-92.4 -341.2,47.2 l-85.2,8.4 66.8,-7.2 l156.8,-42.4 146.8,-76.4 l28.4,-40.0 -20.8,-16.8 l-159.2,69.2 -152.4,103.2 l-75.2,67.2 62.4,0.4 l210.0,-68.8 235.2,-132.8 l21.6,-76.4 -197.6,90.4 l-214.4,122.8 -187.2,174.8 l9.6,31.6 319.6,-41.2 l136.4,-79.6 78.4,-119.2 l-25.2,-134.0 -140.4,-45.6 l-237.6,82.0 -138.4,195.6 l13.6,142.0 298.0,-16.8 l158.4,-90.8 48.8,-121.6 l-44.4,-138.0 -138.8,-52.4 l-258.8,160.8 -79.6,199.6 l48.8,59.2 260.4,-57.6 l77.6,-112.8 6.4,-242.4 l-162.8,32.4 -168.0,156.0 l-58.0,129.2 64.0,95.2 l235.6,-27.6 159.2,-147.2 l30.4,-172.8 -56.0,-58.4 l-231.2,92.0 -132.0,171.2 l1.2,100.4 167.6,42.4 l204.4,-50.8 65.2,-109.2 l-7.6,-72.0 -94.0,-60.4 l-111.6,-2.4 -189.6,104.4 l-32.0,88.0 78.4,102.4 l271.2,-13.2 148.4,-93.6 l23.6,-138.8 -58.8,-114.8 l-188.0,-59.2 -388.0,118.8 l-36.8,113.6 48.0,80.0 l159.2,46.8 232.8,-18.8 l79.6,-70.4 -19.6,-163.6 l-122.8,-92.4 -346.8,61.2 l-30.4,86.8 0.0,39.6 l23.2,50.4 156.4,81.6 l96.0,-10.0 89.2,-78.8 l28.0,-74.8 -6.4,-61.6 l-115.2,-32.4 -199.6,20.0 l-71.6,95.2 7.2,120.8 l114.0,107.6 200.8,0.0 l187.2,-58.4 76.0,-73.6 l49.6,-97.2 -12.4,-50.8 l-122.4,-10.8 -151.2,41.2 l-137.2,158.4 18.8,91.2 l254.0,-2.4 106.4,-31.6 l80.8,-138.8 -18.4,-100.0 l-137.6,-55.6 -128.4,16.4 l-156.0,112.0 -16.4,92.0 l96.0,108.0 213.6,-31.2 l78.8,-141.6 -15.2,-169.6 l-70.4,-57.6 -409.2,39.6 l-28.0,82.0 62.4,83.6 l165.6,42.8 138.8,-21.2 l96.8,-87.6 34.0,-82.4 l-47.6,-60.0 -266.0,38.0 l-87.2,120.4 37.2,94.0 l108.4,62.8 104.0,21.6 l128.0,-32.0 62.4,-66.0 l6.8,-60.4 -69.2,-52.0 l-199.2,-15.6 -103.2,52.0 l-5.2,74.0 62.4,67.2 l130.8,21.2 73.2,-29.6 l23.6,-52.0 -24.4,-62.0 l-144.8,-42.4 -138.0,35.2 l-39.2,80.0 33.6,92.0 l132.0,32.0 105.6,-61.2 l30.8,-168.8 -133.2,-111.6 l-225.2,-35.6 -88.8,197.2 l227.2,132.4 196.8,-5.2 l96.8,-100.0 -39.6,-151.6 l-241.6,-90.4 -262.8,230.4 l187.6,175.2 352.4,-24.8 l-81.6,-148.4 -390.8,-96.4 l-84.4,121.6 164.0,161.6 l416.4,-49.2 33.2,-158.8 l-131.6,-204.0 -487.6,98.0 l35.2,198.8 294.8,115.6 l341.6,-155.2 -10.8,-126.4 l-295.2,-138.0 -365.6,188.0 l88.4,215.6 187.6,16.0 l294.4,-243.6 -90.0,-176.0 l-421.6,111.2 -12.8,151.6 l257.6,130.0 117.2,-5.6 l54.4,-68.8 -80.0,-46.4 l-215.6,86.8 -76.4,60.8 l202.4,-16.4 \" fill=\"none\" stroke=\"black\" stroke-width=\"1\"/></svg>"
      ],
      "text/plain": [
       "<IPython.core.display.SVG object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "simplified_strokes = further_simplify(the_strokes, 5.0)\n",
    "print(len(simplified_strokes))\n",
    "util.draw_strokes(simplified_strokes, factor=2.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.5\n",
      "0.75\n",
      "1.0\n",
      "1.25\n",
      "1.5\n",
      "1.75\n",
      "2.0\n",
      "2.25\n",
      "2.5\n",
      "2.75\n",
      "3.0\n",
      "3.25\n",
      "3.5\n",
      "3.75\n",
      "4.0\n",
      "4.25\n",
      "4.5\n",
      "4.75\n",
      "5.0\n"
     ]
    }
   ],
   "source": [
    "for j in range(1, 20):\n",
    "  eps = j/4.0+0.25\n",
    "  print(eps)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000\n",
      "oops 1163 220\n",
      "2000\n",
      "oops 2024 208\n",
      "oops 2312 215\n",
      "3000\n",
      "4000\n",
      "oops 4407 291\n",
      "5000\n",
      "6000\n",
      "7000\n",
      "oops 7164 272\n",
      "oops 7675 268\n",
      "8000\n",
      "oops 8668 320\n",
      "oops 8789 284\n",
      "9000\n",
      "oops 9429 202\n",
      "oops 9918 226\n",
      "10000\n",
      "oops 10642 209\n",
      "oops 10650 256\n",
      "11000\n",
      "oops 11381 266\n",
      "12000\n",
      "oops 12239 207\n",
      "13000\n",
      "14000\n",
      "oops 14666 253\n",
      "15000\n",
      "oops 15148 383\n",
      "oops 15253 438\n",
      "oops 15381 269\n",
      "oops 15637 212\n",
      "oops 15882 497\n",
      "16000\n",
      "oops 16007 326\n",
      "oops 16871 231\n",
      "17000\n",
      "oops 17692 201\n",
      "18000\n",
      "19000\n",
      "oops 19089 252\n",
      "20000\n",
      "21000\n",
      "oops 21003 521\n",
      "oops 21433 212\n",
      "oops 21614 201\n",
      "oops 21730 386\n",
      "oops 21875 227\n",
      "22000\n",
      "oops 22939 281\n",
      "oops 22945 223\n",
      "23000\n",
      "oops 23209 289\n",
      "oops 23276 243\n",
      "24000\n",
      "oops 24332 208\n",
      "25000\n",
      "oops 25005 228\n",
      "oops 25058 454\n",
      "oops 25861 245\n",
      "26000\n",
      "oops 26543 222\n",
      "oops 26889 461\n",
      "27000\n",
      "28000\n",
      "oops 28279 757\n",
      "29000\n",
      "oops 29475 286\n",
      "oops 29486 333\n",
      "oops 29522 255\n",
      "30000\n",
      "oops 30780 668\n",
      "oops 30877 512\n",
      "oops 30906 237\n",
      "31000\n",
      "oops 31053 614\n",
      "oops 31213 241\n",
      "32000\n",
      "33000\n",
      "34000\n",
      "oops 34758 360\n",
      "35000\n",
      "oops 35027 253\n",
      "oops 35072 233\n",
      "oops 35244 231\n",
      "36000\n",
      "oops 36481 323\n",
      "37000\n",
      "oops 37038 430\n",
      "38000\n",
      "oops 38720 207\n",
      "oops 38741 391\n",
      "39000\n",
      "40000\n",
      "oops 40118 212\n",
      "oops 40648 256\n",
      "41000\n",
      "oops 41055 223\n",
      "42000\n",
      "oops 42514 219\n",
      "oops 42683 246\n",
      "oops 42909 201\n",
      "43000\n",
      "oops 43205 215\n",
      "oops 43506 246\n",
      "44000\n",
      "oops 44180 209\n",
      "oops 44492 222\n",
      "45000\n",
      "oops 45359 250\n",
      "oops 45634 230\n",
      "46000\n",
      "oops 46111 284\n",
      "oops 46733 249\n",
      "47000\n",
      "oops 47995 457\n",
      "48000\n",
      "oops 48054 233\n",
      "oops 48129 277\n",
      "49000\n",
      "oops 49513 247\n",
      "50000\n",
      "oops 50105 722\n",
      "51000\n",
      "52000\n",
      "oops 52164 287\n",
      "oops 52538 351\n",
      "53000\n",
      "oops 53185 222\n",
      "oops 53447 354\n",
      "oops 53881 264\n",
      "54000\n",
      "oops 54107 231\n",
      "oops 54213 228\n",
      "oops 54653 243\n",
      "55000\n",
      "oops 55178 490\n",
      "oops 55949 217\n",
      "56000\n",
      "oops 56110 375\n",
      "oops 56256 202\n",
      "57000\n",
      "oops 57063 201\n",
      "oops 57176 1500\n",
      "oops 57181 543\n",
      "oops 57837 222\n",
      "58000\n",
      "oops 58530 1934\n",
      "oops 58828 211\n",
      "oops 58985 231\n",
      "59000\n",
      "oops 59035 278\n",
      "oops 59167 458\n",
      "oops 59896 406\n",
      "60000\n",
      "oops 60070 298\n",
      "oops 60182 237\n",
      "61000\n",
      "oops 61136 324\n",
      "oops 61504 715\n",
      "62000\n",
      "63000\n",
      "oops 63343 225\n",
      "oops 63446 249\n",
      "oops 63945 231\n",
      "64000\n",
      "oops 64011 265\n",
      "oops 64786 246\n",
      "65000\n",
      "66000\n",
      "oops 66177 484\n",
      "oops 66366 280\n",
      "oops 66445 256\n",
      "67000\n",
      "oops 67048 857\n",
      "oops 67354 205\n",
      "68000\n",
      "oops 68283 271\n",
      "69000\n",
      "oops 69339 369\n",
      "oops 69495 327\n",
      "oops 69836 243\n",
      "70000\n",
      "71000\n",
      "oops 71114 247\n",
      "oops 71341 228\n",
      "72000\n",
      "73000\n",
      "74000\n",
      "75000\n",
      "oops 75315 335\n",
      "76000\n",
      "oops 76208 294\n",
      "77000\n",
      "oops 77001 300\n",
      "oops 77019 301\n",
      "78000\n",
      "oops 78103 260\n",
      "oops 78167 307\n",
      "oops 78442 516\n",
      "oops 78866 433\n",
      "oops 78923 231\n",
      "79000\n",
      "oops 79150 262\n",
      "oops 79659 231\n",
      "80000\n",
      "oops 80331 222\n",
      "oops 80347 252\n",
      "81000\n",
      "oops 81354 233\n",
      "oops 81837 250\n",
      "oops 81917 368\n",
      "82000\n",
      "oops 82084 256\n",
      "oops 82837 216\n",
      "oops 82947 425\n",
      "oops 82971 207\n",
      "83000\n",
      "oops 83993 396\n",
      "84000\n",
      "85000\n",
      "86000\n",
      "oops 86110 216\n",
      "oops 86302 233\n",
      "oops 86688 243\n",
      "87000\n",
      "oops 87845 254\n",
      "88000\n",
      "oops 88252 243\n",
      "89000\n",
      "oops 89120 204\n",
      "90000\n",
      "oops 90091 473\n",
      "oops 90277 203\n",
      "oops 90611 214\n",
      "oops 90654 227\n",
      "91000\n",
      "oops 91533 201\n",
      "oops 91757 388\n",
      "oops 91910 223\n",
      "92000\n",
      "oops 92402 246\n",
      "oops 92758 432\n",
      "93000\n",
      "94000\n",
      "oops 94449 483\n",
      "oops 94631 325\n",
      "oops 94646 250\n",
      "oops 94950 291\n",
      "95000\n",
      "oops 95054 207\n",
      "oops 95244 350\n",
      "oops 95769 347\n",
      "96000\n",
      "oops 96411 340\n",
      "97000\n",
      "98000\n",
      "oops 98249 363\n",
      "oops 98419 228\n",
      "oops 98578 418\n",
      "99000\n",
      "100000\n",
      "oops 100616 203\n",
      "101000\n",
      "oops 101723 218\n",
      "102000\n",
      "103000\n",
      "oops 103030 211\n",
      "104000\n",
      "105000\n",
      "oops 105078 233\n",
      "oops 105855 215\n",
      "106000\n",
      "oops 106942 206\n",
      "107000\n",
      "oops 107385 234\n",
      "oops 107792 208\n",
      "oops 107917 341\n",
      "108000\n",
      "109000\n",
      "oops 109086 220\n",
      "oops 109806 803\n",
      "110000\n",
      "oops 110015 212\n",
      "oops 110387 437\n",
      "oops 110791 278\n",
      "oops 110797 205\n",
      "oops 110982 232\n",
      "111000\n",
      "112000\n",
      "oops 112103 220\n",
      "oops 112518 607\n",
      "oops 112901 531\n",
      "113000\n",
      "114000\n",
      "oops 114068 232\n",
      "oops 114326 264\n",
      "oops 114539 412\n",
      "oops 114564 393\n",
      "115000\n",
      "oops 115425 437\n",
      "116000\n",
      "oops 116326 667\n",
      "oops 116901 209\n",
      "117000\n",
      "oops 117244 274\n",
      "oops 117616 208\n",
      "oops 117897 211\n",
      "118000\n",
      "oops 118167 391\n",
      "119000\n",
      "oops 119173 564\n"
     ]
    }
   ],
   "source": [
    "processed_strokes = []\n",
    "for i in range(len(all_strokes)):\n",
    "  if ((i+1) % 1000 == 0):\n",
    "    print(i+1)\n",
    "  this_strokes = all_strokes[i]\n",
    "  if len(this_strokes) < TARGET:\n",
    "    processed_strokes.append(this_strokes)\n",
    "  else:\n",
    "    for j in range(1, 9):\n",
    "      eps = j/4.0+0.25\n",
    "      simplified_strokes = further_simplify(this_strokes, eps)\n",
    "      if len(simplified_strokes) <= TARGET:\n",
    "        break\n",
    "    if len(simplified_strokes) > TARGET:\n",
    "      print(\"oops\", i, len(simplified_strokes))\n",
    "    else:\n",
    "      processed_strokes.append(simplified_strokes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "119498\n",
      "159\n"
     ]
    },
    {
     "data": {
      "image/svg+xml": [
       "<svg baseProfile=\"full\" height=\"167.99999999999994\" version=\"1.1\" width=\"186.28571428571425\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:ev=\"http://www.w3.org/2001/xml-events\" xmlns:xlink=\"http://www.w3.org/1999/xlink\"><defs/><rect fill=\"white\" height=\"167.99999999999994\" width=\"186.28571428571425\" x=\"0\" y=\"0\"/><path d=\"M30.714285714285694,35.85714285714282 m0.0,0.0 l0.0,1.7142857142857142 -0.5714285714285714,3.4285714285714284 l-0.5714285714285714,9.142857142857142 0.0,7.142857142857143 l-0.2857142857142857,1.7142857142857142 0.0,20.0 l1.1428571428571428,9.428571428571429 0.5714285714285714,10.285714285714286 l0.5714285714285714,4.0 0.0,9.142857142857142 l0.2857142857142857,1.7142857142857142 0.0,5.714285714285714 l0.8571428571428571,4.857142857142857 0.0,1.4285714285714286 l2.2857142857142856,8.571428571428571 0.5714285714285714,3.142857142857143 l0.5714285714285714,1.4285714285714286 0.5714285714285714,2.857142857142857 l0.0,1.4285714285714286 4.285714285714286,-3.7142857142857144 l6.857142857142857,-3.4285714285714284 1.7142857142857142,-0.5714285714285714 l6.571428571428571,-1.1428571428571428 2.0,0.0 l1.7142857142857142,-0.2857142857142857 18.857142857142858,0.0 l1.4285714285714286,0.2857142857142857 4.0,0.0 l13.714285714285714,1.4285714285714286 8.857142857142858,0.0 l2.2857142857142856,0.2857142857142857 13.714285714285714,0.0 l2.0,0.2857142857142857 16.571428571428573,0.0 l5.142857142857143,-0.8571428571428571 2.0,-0.5714285714285714 l3.7142857142857144,-1.7142857142857142 4.285714285714286,-3.142857142857143 l2.857142857142857,-1.7142857142857142 1.1428571428571428,-1.4285714285714286 l0.5714285714285714,-1.4285714285714286 0.2857142857142857,-7.714285714285714 l-0.8571428571428571,-4.285714285714286 -2.5714285714285716,-7.714285714285714 l-3.142857142857143,-6.285714285714286 -1.1428571428571428,-1.7142857142857142 l-2.2857142857142856,-4.571428571428571 -4.0,-6.857142857142857 l-2.5714285714285716,-5.714285714285714 -6.571428571428571,-12.571428571428571 l-1.4285714285714286,-2.2857142857142856 -0.8571428571428571,-2.0 l-2.5714285714285716,-4.0 -1.7142857142857142,-4.0 l-5.714285714285714,-10.857142857142858 -2.0,-3.142857142857143 l-0.5714285714285714,-1.4285714285714286 -3.7142857142857144,-6.285714285714286 l-1.1428571428571428,-1.4285714285714286 -3.4285714285714284,-2.5714285714285716 l-6.0,-3.142857142857143 -4.571428571428571,-1.4285714285714286 l-1.7142857142857142,-0.2857142857142857 -9.142857142857142,0.0 l-9.142857142857142,1.4285714285714286 -1.4285714285714286,0.0 l-12.0,2.0 -11.428571428571429,0.0 l-1.4285714285714286,0.2857142857142857 -4.0,0.0 l-2.0,0.2857142857142857 -10.285714285714286,0.0 l-7.714285714285714,-0.8571428571428571 -7.714285714285714,0.0 l-1.4285714285714286,1.1428571428571428 0.0,1.4285714285714286 m29.142857142857142,-0.8571428571428571 l0.5714285714285714,4.285714285714286 1.4285714285714286,5.428571428571429 l2.5714285714285716,12.571428571428571 0.5714285714285714,1.4285714285714286 l0.5714285714285714,3.4285714285714284 0.5714285714285714,1.4285714285714286 l0.8571428571428571,4.857142857142857 0.2857142857142857,4.285714285714286 l1.4285714285714286,8.0 0.0,1.7142857142857142 l0.5714285714285714,4.857142857142857 0.0,2.5714285714285716 l0.2857142857142857,1.7142857142857142 0.2857142857142857,10.285714285714286 l0.5714285714285714,4.571428571428571 0.0,4.285714285714286 l0.2857142857142857,1.7142857142857142 0.0,16.857142857142858 l1.1428571428571428,7.428571428571429 4.285714285714286,-4.0 l2.0,-1.1428571428571428 2.857142857142857,-1.1428571428571428 l0.0,-4.571428571428571 -0.5714285714285714,-2.857142857142857 l-0.2857142857142857,-6.0 -2.0,-12.285714285714286 l0.0,-1.7142857142857142 -2.0,-9.428571428571429 l0.0,-1.7142857142857142 -2.0,-10.0 l-1.4285714285714286,-8.857142857142858 0.0,-2.0 l-0.5714285714285714,-3.142857142857143 0.0,-4.285714285714286 l0.5714285714285714,-5.714285714285714 0.0,-9.142857142857142 l-0.5714285714285714,-2.857142857142857 0.0,-1.4285714285714286 l-1.4285714285714286,-4.857142857142857 -1.1428571428571428,-2.857142857142857 l-0.8571428571428571,-3.4285714285714284 m11.428571428571429,0.8571428571428571 l0.2857142857142857,2.857142857142857 0.8571428571428571,4.571428571428571 l0.0,26.0 2.0,16.285714285714285 l0.2857142857142857,5.714285714285714 0.8571428571428571,4.571428571428571 l0.5714285714285714,5.142857142857143 0.2857142857142857,10.0 l1.7142857142857142,10.0 2.857142857142857,11.428571428571429 l1.1428571428571428,9.142857142857142 1.4285714285714286,-1.1428571428571428 l2.2857142857142856,-2.857142857142857 1.4285714285714286,-1.1428571428571428 l1.4285714285714286,-0.5714285714285714 1.4285714285714286,0.0 l1.4285714285714286,0.8571428571428571 4.571428571428571,1.7142857142857142 l1.4285714285714286,0.2857142857142857 0.8571428571428571,-2.0 l0.0,-5.428571428571429 -0.5714285714285714,-1.7142857142857142 l-0.5714285714285714,-3.142857142857143 -0.8571428571428571,-6.0 l0.0,-1.7142857142857142 -0.5714285714285714,-2.857142857142857 l-0.5714285714285714,-14.571428571428571 -0.2857142857142857,-3.4285714285714284 l-0.5714285714285714,-3.142857142857143 0.0,-2.857142857142857 l-0.8571428571428571,-5.142857142857143 0.0,-2.2857142857142856 l-2.0,-8.285714285714286 -6.285714285714286,-16.857142857142858 l-0.8571428571428571,-4.857142857142857 -1.1428571428571428,-4.571428571428571 l-0.2857142857142857,-5.714285714285714 -0.8571428571428571,-4.285714285714286 l0.0,-3.142857142857143 \" fill=\"none\" stroke=\"black\" stroke-width=\"1\"/></svg>"
      ],
      "text/plain": [
       "<IPython.core.display.SVG object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(len(processed_strokes))\n",
    "this_strokes = random.choice(processed_strokes)\n",
    "print(len(this_strokes))\n",
    "util.draw_strokes(this_strokes, factor=3.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.savez_compressed(\"processed_\"+NAME+\".npz\", rawdata=processed_strokes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "119498\n"
     ]
    }
   ],
   "source": [
    "N = len(processed_strokes)\n",
    "print(N)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.savez_compressed(NAME+\".\"+str(TARGET)+\".npz\", train=processed_strokes[0:N-5000],valid=processed_strokes[N-5000:N-2000], test=processed_strokes[N-2000:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "filename = NAME+\".\"+str(TARGET)+\".npz\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "20\n",
      "200\n"
     ]
    }
   ],
   "source": [
    "load_data = np.load(filename, allow_pickle=True)\n",
    "train_set = load_data['train']\n",
    "valid_set = load_data['valid']\n",
    "test_set = load_data['test']\n",
    "print(util.get_min_len(train_set))\n",
    "print(util.get_max_len(train_set))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "114498"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train_set)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3000"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(valid_set)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2000"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(test_set)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "86\n"
     ]
    },
    {
     "data": {
      "image/svg+xml": [
       "<svg baseProfile=\"full\" height=\"119.71428571428571\" version=\"1.1\" width=\"109.71428571428571\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:ev=\"http://www.w3.org/2001/xml-events\" xmlns:xlink=\"http://www.w3.org/1999/xlink\"><defs/><rect fill=\"white\" height=\"119.71428571428571\" width=\"109.71428571428571\" x=\"0\" y=\"0\"/><path d=\"M27.285714285714285,27.571428571428555 m0.0,0.0 l-0.2857142857142857,3.142857142857143 0.0,3.142857142857143 l-0.2857142857142857,3.7142857142857144 -0.5714285714285714,4.857142857142857 l-0.8571428571428571,11.142857142857142 -0.2857142857142857,6.571428571428571 l0.0,14.0 0.5714285714285714,2.2857142857142856 l1.1428571428571428,3.7142857142857144 0.5714285714285714,3.4285714285714284 l2.0,1.1428571428571428 4.571428571428571,0.0 l4.857142857142857,-0.2857142857142857 36.57142857142857,0.0 l3.7142857142857144,-0.5714285714285714 1.4285714285714286,-1.1428571428571428 l1.4285714285714286,-2.0 0.8571428571428571,-2.5714285714285716 l0.0,-2.5714285714285716 0.2857142857142857,-3.142857142857143 l0.0,-3.7142857142857144 0.2857142857142857,-3.7142857142857144 l0.5714285714285714,-3.4285714285714284 0.8571428571428571,-10.571428571428571 l0.0,-15.142857142857142 -1.4285714285714286,-4.857142857142857 l-1.1428571428571428,-2.2857142857142856 -1.7142857142857142,-0.8571428571428571 l-4.285714285714286,0.0 -4.857142857142857,0.2857142857142857 l-6.0,0.5714285714285714 -7.714285714285714,0.2857142857142857 l-6.571428571428571,0.0 -9.428571428571429,0.5714285714285714 l-8.571428571428571,0.0 0.0,0.5714285714285714 m-1.4285714285714286,-3.142857142857143 l1.1428571428571428,6.0 0.0,26.571428571428573 l-0.5714285714285714,6.285714285714286 -1.7142857142857142,11.714285714285714 l-0.2857142857142857,4.285714285714286 -0.5714285714285714,3.7142857142857144 l0.0,4.285714285714286 m12.0,-58.857142857142854 l0.8571428571428571,2.0 0.0,6.0 l-2.5714285714285716,17.714285714285715 -0.2857142857142857,6.571428571428571 l-0.5714285714285714,6.285714285714286 -0.2857142857142857,6.571428571428571 l0.0,18.285714285714285 0.5714285714285714,0.2857142857142857 m12.285714285714286,-69.71428571428571 l0.8571428571428571,3.4285714285714284 0.5714285714285714,3.4285714285714284 l0.0,15.714285714285714 -0.8571428571428571,6.285714285714286 l-1.7142857142857142,18.571428571428573 -0.2857142857142857,5.428571428571429 l0.0,8.285714285714286 -0.5714285714285714,5.142857142857143 l0.0,1.4285714285714286 -0.5714285714285714,1.4285714285714286 m15.714285714285714,-67.14285714285714 l-0.2857142857142857,4.285714285714286 0.0,4.857142857142857 l0.5714285714285714,6.857142857142857 0.2857142857142857,5.142857142857143 l0.0,20.285714285714285 -0.2857142857142857,6.857142857142857 l-0.5714285714285714,6.285714285714286 -1.1428571428571428,7.714285714285714 l-0.8571428571428571,3.4285714285714284 m10.0,-57.142857142857146 l0.5714285714285714,2.2857142857142856 1.1428571428571428,3.4285714285714284 l0.5714285714285714,4.285714285714286 0.0,12.285714285714286 l-0.8571428571428571,6.0 -1.4285714285714286,5.428571428571429 l-1.1428571428571428,5.142857142857143 -2.2857142857142856,11.428571428571429 l-0.2857142857142857,2.0 0.0,2.0 \" fill=\"none\" stroke=\"black\" stroke-width=\"1\"/></svg>"
      ],
      "text/plain": [
       "<IPython.core.display.SVG object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "this_strokes = random.choice(train_set)\n",
    "print(len(this_strokes))\n",
    "util.draw_strokes(this_strokes, factor=3.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "182\n"
     ]
    },
    {
     "data": {
      "image/svg+xml": [
       "<svg baseProfile=\"full\" height=\"196.85714285714286\" version=\"1.1\" width=\"219.7142857142857\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:ev=\"http://www.w3.org/2001/xml-events\" xmlns:xlink=\"http://www.w3.org/1999/xlink\"><defs/><rect fill=\"white\" height=\"196.85714285714286\" width=\"219.7142857142857\" x=\"0\" y=\"0\"/><path d=\"M120.14285714285712,50.42857142857142 m0.0,0.0 l-4.285714285714286,-0.5714285714285714 -3.7142857142857144,0.0 l-2.0,0.2857142857142857 -3.142857142857143,2.0 l-2.0,2.857142857142857 -0.5714285714285714,3.142857142857143 l0.0,5.714285714285714 0.5714285714285714,1.4285714285714286 l1.4285714285714286,1.4285714285714286 1.7142857142857142,1.1428571428571428 l4.0,1.7142857142857142 8.0,0.2857142857142857 l2.5714285714285716,-0.2857142857142857 2.0,-1.1428571428571428 l2.2857142857142856,-2.0 2.5714285714285716,-5.142857142857143 l1.4285714285714286,-4.0 0.2857142857142857,-5.142857142857143 l-2.0,-1.7142857142857142 -2.2857142857142856,-0.8571428571428571 l-5.714285714285714,0.0 -4.857142857142857,-0.8571428571428571 l-4.0,0.0 -3.4285714285714284,0.5714285714285714 l-0.2857142857142857,-10.0 1.7142857142857142,-0.5714285714285714 l11.714285714285714,0.0 0.8571428571428571,1.7142857142857142 l0.2857142857142857,1.7142857142857142 1.7142857142857142,4.857142857142857 m-14.0,17.142857142857142 l2.857142857142857,-3.142857142857143 2.2857142857142856,-1.4285714285714286 l2.2857142857142856,-0.5714285714285714 1.4285714285714286,0.5714285714285714 l1.1428571428571428,3.4285714285714284 m-7.714285714285714,-10.571428571428571 l-0.5714285714285714,2.0 0.2857142857142857,1.4285714285714286 l0.2857142857142857,0.0 m7.142857142857143,-2.857142857142857 l0.5714285714285714,2.5714285714285716 0.0,2.0 m-4.857142857142857,14.857142857142858 l0.0,46.0 -4.857142857142857,19.428571428571427 l-5.714285714285714,16.857142857142858 m12.0,-34.857142857142854 l0.0,1.7142857142857142 3.7142857142857144,8.0 l2.2857142857142856,7.142857142857143 1.7142857142857142,9.142857142857142 l1.7142857142857142,12.571428571428571 1.1428571428571428,1.7142857142857142 m-6.857142857142857,-4.0 l6.571428571428571,-0.2857142857142857 4.571428571428571,-1.7142857142857142 l2.5714285714285716,-1.4285714285714286 m0.8571428571428571,-0.2857142857142857 l2.0,0.8571428571428571 1.7142857142857142,0.0 l5.714285714285714,1.7142857142857142 7.428571428571429,2.857142857142857 l3.142857142857143,0.8571428571428571 3.4285714285714284,0.5714285714285714 l19.142857142857142,1.7142857142857142 m2.0,-11.428571428571429 l-2.857142857142857,2.0 -1.1428571428571428,1.7142857142857142 l-0.5714285714285714,2.2857142857142856 -0.2857142857142857,3.4285714285714284 l0.2857142857142857,5.142857142857143 2.2857142857142856,2.2857142857142856 l2.5714285714285716,0.8571428571428571 4.857142857142857,0.0 l2.5714285714285716,-0.8571428571428571 2.2857142857142856,-2.0 l1.7142857142857142,-2.2857142857142856 1.7142857142857142,-3.7142857142857144 l0.8571428571428571,-3.7142857142857144 0.0,-2.0 l-0.8571428571428571,-1.7142857142857142 -1.7142857142857142,-1.1428571428571428 l-1.4285714285714286,-0.5714285714285714 -6.857142857142857,-0.8571428571428571 l-3.142857142857143,-0.2857142857142857 -1.7142857142857142,0.2857142857142857 m-63.714285714285715,-64.0 l4.571428571428571,2.0 2.857142857142857,0.2857142857142857 l6.0,0.0 6.857142857142857,-2.0 l5.714285714285714,-3.7142857142857144 5.142857142857143,-5.428571428571429 l6.285714285714286,-8.285714285714286 1.4285714285714286,-0.8571428571428571 l5.142857142857143,-0.2857142857142857 1.7142857142857142,0.5714285714285714 l0.8571428571428571,1.1428571428571428 m-46.285714285714285,16.0 l-7.714285714285714,-4.285714285714286 -5.142857142857143,-2.0 l-7.142857142857143,0.2857142857142857 -5.142857142857143,3.4285714285714284 l-4.285714285714286,4.571428571428571 -2.5714285714285716,4.285714285714286 l-0.8571428571428571,3.142857142857143 0.0,1.4285714285714286 m40.857142857142854,-69.71428571428571 l0.2857142857142857,1.7142857142857142 0.0,5.714285714285714 l-4.0,19.428571428571427 -0.8571428571428571,16.285714285714285 l0.0,11.428571428571429 -1.4285714285714286,32.57142857142857 l0.0,56.857142857142854 0.2857142857142857,2.857142857142857 m-27.714285714285715,-143.14285714285714 l-1.7142857142857142,2.0 -1.4285714285714286,5.142857142857143 l0.0,13.428571428571429 1.4285714285714286,43.42857142857143 l0.0,22.0 -2.0,24.857142857142858 l0.0,17.142857142857142 m65.14285714285714,-124.28571428571429 l-0.8571428571428571,2.2857142857142856 0.0,4.0 l0.5714285714285714,4.0 0.2857142857142857,5.714285714285714 l0.2857142857142857,14.285714285714286 -2.857142857142857,27.142857142857142 l-0.8571428571428571,20.0 0.0,19.428571428571427 l2.857142857142857,29.142857142857142 0.0,6.571428571428571 m-100.85714285714286,-135.71428571428572 l-1.1428571428571428,3.142857142857143 -0.5714285714285714,4.285714285714286 l0.0,14.285714285714286 0.8571428571428571,17.428571428571427 l0.8571428571428571,30.571428571428573 0.5714285714285714,54.857142857142854 l0.2857142857142857,8.0 0.5714285714285714,0.5714285714285714 m137.71428571428572,-123.71428571428571 l-1.1428571428571428,3.7142857142857144 -0.2857142857142857,2.2857142857142856 l0.2857142857142857,13.142857142857142 0.8571428571428571,6.0 l1.4285714285714286,23.428571428571427 0.0,27.142857142857142 l-1.4285714285714286,31.142857142857142 0.0,12.285714285714286 m5.142857142857143,-122.85714285714286 l-10.571428571428571,0.0 -25.428571428571427,-1.4285714285714286 l-6.285714285714286,0.0 -37.42857142857143,-3.142857142857143 l-8.285714285714286,-0.2857142857142857 -34.0,0.0 l-15.714285714285714,1.7142857142857142 -7.428571428571429,1.1428571428571428 l-19.714285714285715,4.857142857142857 -4.285714285714286,2.0 l0.8571428571428571,11.714285714285714 0.0,44.57142857142857 l0.5714285714285714,15.714285714285714 2.5714285714285716,22.285714285714285 l0.5714285714285714,8.0 0.5714285714285714,25.428571428571427 l5.714285714285714,-0.2857142857142857 5.428571428571429,-0.8571428571428571 l14.0,-3.7142857142857144 7.714285714285714,-1.1428571428571428 l8.285714285714286,-0.8571428571428571 16.571428571428573,0.0 l18.285714285714285,0.5714285714285714 18.285714285714285,1.7142857142857142 l18.0,2.5714285714285716 26.571428571428573,2.2857142857142856 \" fill=\"none\" stroke=\"black\" stroke-width=\"1\"/></svg>"
      ],
      "text/plain": [
       "<IPython.core.display.SVG object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "this_strokes = random.choice(train_set)\n",
    "print(len(this_strokes))\n",
    "util.draw_strokes(this_strokes, factor=3.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "167\n"
     ]
    },
    {
     "data": {
      "image/svg+xml": [
       "<svg baseProfile=\"full\" height=\"166.11427525111606\" version=\"1.1\" width=\"154.45715332031247\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:ev=\"http://www.w3.org/2001/xml-events\" xmlns:xlink=\"http://www.w3.org/1999/xlink\"><defs/><rect fill=\"white\" height=\"166.11427525111606\" width=\"154.45715332031247\" x=\"0\" y=\"0\"/><path d=\"M25,26.828569684709805 m0.0,0.0 l0.4571533203125,2.0571419852120534 0.9142892020089286,2.74285888671875 l1.3714250837053572,3.1999991280691966 2.0571463448660716,7.314287458147321 l2.9714181082589284,15.085710797991071 1.6000104631696428,8.914289202008929 l0.9142717633928571,11.885707310267858 0.0,26.514290945870535 l-0.6857038225446429,2.28570556640625 -0.4571533203125,3.4285714285714284 l-0.6857038225446429,3.8857247488839284 -2.971435546875,13.714285714285714 l-1.1428571428571428,10.971418108258929 1.3714250837053572,-0.22856794084821427 l1.8285784040178572,-1.5999930245535714 9.599993024553571,-6.628574916294643 l6.628574916294643,-3.4285714285714284 6.628574916294643,-2.7428501674107144 l5.714285714285714,-2.5142996651785716 7.314278738839286,-2.5142822265625 l7.542864118303571,-1.8285609654017858 7.314278738839286,-0.4571533203125 l13.257149832589286,0.0 5.4857177734375,1.3714250837053572 l5.257132393973214,1.6000104631696428 6.400006975446429,1.5999930245535714 l5.257132393973214,1.6000104631696428 4.342860630580357,1.5999930245535714 m4.571428571428571,0.45713588169642855 l-1.3714250837053572,0.0 m-5.028564453125,0.22858537946428573 l1.3714250837053572,-1.3714425223214286 0.9142892020089286,-1.5999930245535714 l0.6857038225446429,-1.8285784040178572 0.4571533203125,-2.05712890625 l0.45713588169642855,-4.342860630580357 0.6857212611607143,-2.2857142857142856 l0.45713588169642855,-2.971435546875 0.9142892020089286,-3.657139369419643 l0.22856794084821427,-3.4285714285714284 0.0,-5.714285714285714 l0.22856794084821427,-1.8285696847098214 0.0,-27.200003487723215 l0.9142892020089286,-7.314278738839286 0.0,-2.2857142857142856 l0.22856794084821427,-2.0571463448660716 0.0,-5.25714111328125 l0.4571533203125,-2.74285888671875 0.0,-13.25714111328125 l-0.22858537946428573,-2.514286586216518 -0.9142717633928571,-2.0571419852120534 l-2.2857142857142856,-2.74285888671875 -0.4571533203125,-1.371429443359375 l-1.3714250837053572,-0.22857230050223215 -4.799996512276786,0.0 l-4.799996512276786,0.4571446010044643 -7.771432059151786,1.8285696847098214 l-2.0571463448660716,0.22857230050223215 -1.3714250837053572,0.0 l-1.5999930245535714,0.22857230050223215 -6.628574916294643,0.0 l-2.2857142857142856,0.22857230050223215 -2.0571463448660716,0.45714024135044645 l-4.114292689732143,0.4571446010044643 -2.2857142857142856,0.0 l-2.7428501674107144,0.6857125418526786 -3.2000034877232144,0.4571446010044643 l-4.799996512276786,1.371429443359375 -3.4285714285714284,0.45714024135044645 l-2.971435546875,0.6857125418526786 -2.7428501674107144,0.22857666015625 l-2.2857142857142856,0.4571446010044643 -10.5142822265625,0.0 l-4.571428571428571,-2.2857186453683034 -1.3714250837053572,-0.9142848423549107 l-5.4857177734375,-2.74285888671875 -1.8285784040178572,-0.45714024135044645 l-1.3714250837053572,-0.6857169015066964 -1.8285784040178572,-0.22856794084821427 l-2.05712890625,-0.6857169015066964 -1.6000104631696428,-0.22857230050223215 l-1.8285609654017858,0.0 -1.6000104631696428,-0.22856794084821427 m37.02858189174107,57.142857142857146 l-3.4285714285714284,3.657139369419643 -0.6857212611607143,1.3714250837053572 l-0.45713588169642855,2.0571463448660716 -0.22858537946428573,1.8285696847098214 l0.0,4.342860630580357 0.6857212611607143,1.8285696847098214 l1.3714250837053572,1.3714250837053572 1.3714250837053572,0.9142892020089286 l3.2000034877232144,0.6857125418526786 4.342860630580357,0.0 l1.5999930245535714,-0.9142804827008929 1.3714425223214286,-1.1428571428571428 l0.9142717633928571,-1.8285784040178572 0.4571533203125,-1.3714250837053572 l0.0,-1.6000017438616072 -0.6857212611607143,-1.6000017438616072 l-3.4285714285714284,-2.0571376255580356 -3.4285714285714284,-2.74285888671875 l-0.9142892020089286,-0.9142892020089286 m-1.5999930245535714,0.22857666015625 l0.0,2.9714268275669644 -0.22856794084821427,1.8285696847098214 m4.342860630580357,1.8285696847098214 l0.22856794084821427,-1.3714250837053572 -0.22856794084821427,-1.1428571428571428 m-3.6571568080357144,3.4285714285714284 l0.0,3.657139369419643 m-18.05712890625,-68.11427961077008 l0.0,29.714281354631698 2.05712890625,10.057146344866071 l0.9142892020089286,3.657139369419643 0.6857212611607143,5.25714111328125 l1.1428571428571428,7.085719517299107 0.6857038225446429,2.9714268275669644 l0.22858537946428573,3.4285714285714284 1.1428571428571428,5.028573172433036 l0.6857038225446429,4.342851911272321 0.6857212611607143,2.2857142857142856 l0.22856794084821427,2.0571463448660716 1.1428571428571428,3.657139369419643 l0.22856794084821427,1.6000104631696428 0.9142892020089286,2.05712890625 l1.1428571428571428,4.800013950892857 0.9142892020089286,7.5428466796875 l0.22856794084821427,8.914289202008929 0.45713588169642855,2.5142822265625 l0.0,4.799996512276786 m4.800013950892857,-106.28570556640625 l0.0,3.657139369419643 0.45713588169642855,3.2000034877232144 l3.657139369419643,20.114283970424108 0.6857212611607143,5.942862374441964 l0.6857038225446429,4.342851911272321 0.6857212611607143,3.4285714285714284 l0.6857212611607143,2.74285888671875 1.8285609654017858,5.25714111328125 l0.4571533203125,4.342860630580357 0.45713588169642855,1.5999930245535714 l0.45713588169642855,2.971435546875 0.22858537946428573,2.0571376255580356 l0.0,1.6000017438616072 0.9142717633928571,4.342860630580357 l0.0,1.3714250837053572 0.9142892020089286,2.74285888671875 l0.0,1.3714338030133928 0.6857212611607143,4.114275251116071 l0.0,3.2000034877232144 m7.5428466796875,-82.97142682756696 l0.0,5.942853655133929 0.22858537946428573,2.971435546875 l0.6857038225446429,3.1999947684151784 2.2857142857142856,7.314287458147321 l2.0571463448660716,11.200003487723214 1.8285784040178572,7.085710797991071 l1.5999930245535714,6.857142857142857 1.5999930245535714,9.600001743861608 l0.6857212611607143,6.628566196986607 0.6857212611607143,3.2000034877232144 l0.22856794084821427,1.8285696847098214 0.6857038225446429,1.8285696847098214 l0.4571533203125,1.6000017438616072 0.0,0.6857125418526786 \" fill=\"none\" stroke=\"black\" stroke-width=\"1\"/></svg>"
      ],
      "text/plain": [
       "<IPython.core.display.SVG object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "this_strokes = random.choice(train_set)\n",
    "print(len(this_strokes))\n",
    "util.draw_strokes(this_strokes, factor=3.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "this_strokes = random.choice(train_set)\n",
    "print(len(this_strokes))\n",
    "util.draw_strokes(this_strokes, factor=3.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}